PROGRESS IN BRAIN RESEARCH

VOLUME 161

NEUROTRAUMA: NEW INSIGHTS INTO PATHOLOGY AND TREATMENT

Other volumes in PROGRESS IN BRAIN RESEARCH

Volume 124: Cerebellar Modules: Molecules, Morphology and Function, by N.M. Gerrits, T.J.H. Ruigrok and C.E. De Zeeuw (Eds.)
– 2000, ISBN 0-444-50108-8.
Volume 125: Transmission Revisited, by L.F. Agnati, K. Fuxe, C. Nicholson and E. Syková (Eds.) – 2000, ISBN 0-444-50314-5.
Volume 126: Cognition, Emotion and Autonomic Responses: the Integrative Role of the Prefrontal Cortex and Limbic Structures, by
H.B.M. Uylings, C.G. Van Eden, J.P.C. De Bruin, M.G.P. Feenstra and C.M.A. Pennartz (Eds.) – 2000, ISBN 0-444-
50332-3.
Volume 127: Neural Transplantation II. Novel Cell Therapies for CNS Disorders, by S.B. Dunnett and A. Björklund (Eds.) – 2000,
ISBN 0-444-50109-6.
Volume 128: Neural Plasticity and Regeneration, by F.J. Seil (Ed.) – 2000, ISBN 0-444-50209-2.
Volume 129: Nervous System Plasticity and Chronic Pain, by J. Sandkühler, B. Bromm and G.F. Gebhart (Eds.) – 2000, ISBN 0-444-
50509-1.
Volume 130: Advances in Neural Population Coding, by M.A.L. Nicolelis (Ed.) – 2001, ISBN 0-444-50110-X.
Volume 131: Concepts and Challenges in Retinal Biology, by H. Kolb, H. Ripps and S. Wu (Eds.) – 2001, ISBN 0-444-50677-2.
Volume 132: Glial Cell Function, by B. Castellano López and M. Nieto-Sampedro (Eds.) – 2001, ISBN 0-444-50508-3.
Volume 133: The Maternal Brain. Neurobiological and Neuroendocrine Adaptation and Disorders in Pregnancy and Post Partum, by
J.A. Russell, A.J. Douglas, R.J. Windle and C.D. Ingram (Eds.) – 2001, ISBN 0-444-50548-2.
Volume 134: Vision: From Neurons to Cognition, by C. Casanova and M. Ptito (Eds.) – 2001, ISBN 0-444-50586-5.
Volume 135: Do Seizures Damage the Brain, by A. Pitkänen and T. Sutula (Eds.) – 2002, ISBN 0-444-50814-7.
Volume 136: Changing Views of Cajal’s Neuron, by E.C. Azmitia, J. DeFelipe, E.G. Jones, P. Rakic and C.E. Ribak (Eds.) – 2002,
ISBN 0-444-50815-5.
Volume 137: Spinal Cord Trauma: Regeneration, Neural Repair and Functional Recovery, by L. McKerracher, G. Doucet and S.
Rossignol (Eds.) – 2002, ISBN 0-444-50817-1.
Volume 138: Plasticity in the Adult Brain: From Genes to Neurotherapy, by M.A. Hofman, G.J. Boer, A.J.G.D. Holtmaat, E.J.W.
Van Someren, J. Verhaagen and D.F. Swaab (Eds.) – 2002, ISBN 0-444-50981-X.
Volume 139: Vasopressin and Oxytocin: From Genes to Clinical Applications, by D. Poulain, S. Oliet and D. Theodosis (Eds.) – 2002,
ISBN 0-444-50982-8.
Volume 140: The Brain’s Eye, by J. Hyönä, D.P. Munoz, W. Heide and R. Radach (Eds.) – 2002, ISBN 0-444-51097-4.
Volume 141: Gonadotropin-Releasing Hormone: Molecules and Receptors, by I.S. Parhar (Ed.) – 2002, ISBN 0-444-50979-8.
Volume 142: Neural Control of Space Coding, and Action Production, by C. Prablanc, D. Pélisson and Y. Rossetti (Eds.) – 2003,
ISBN 0-444-509771.
Volume 143: Brain Mechanisms for the Integration of Posture and Movement, by S. Mori, D.G. Stuart and M. Wiesendanger (Eds.) –
2004, ISBN 0-444-513892.
Volume 144: The Roots of Visual Awareness, by C.A. Heywood, A.D. Milner and C. Blakemore (Eds.) – 2004, ISBN 0-444-50978-X.
Volume 145: Acetylcholine in the Cerebral Cortex, by L. Descarries, K. Krnjević and M. Steriade (Eds.) – 2004, ISBN 0-444-51125-3.
Volume 146: NGF and Related Molecules in Health and Disease, by L. Aloe and L. Calzà (Eds.) – 2004, ISBN 0-444-51472-4.
Volume 147: Development, Dynamics and Pathology of Neuronal Networks: From Molecules to Functional Circuits, by J. Van Pelt,
M. Kamermans, C.N. Levelt, A. Van Ooyen, G.J.A. Ramakers and P.R. Roelfsema (Eds.) – 2005, ISBN 0-444-51663-8.
Volume 148: Creating Coordination in the Cerebellum, by C.I. De Zeeuw and F. Cicirata (Eds.) – 2005, ISBN 0-444-51754-5.
Volume 149: Cortical Function: A View from the Thalamus, by V.A. Casagrande, R.W. Guillery and S.M. Sherman (Eds.) – 2005,
ISBN 0-444-51679-4.
Volume 150: The Boundaries of Consciousness: Neurobiology and Neuropathology, by Steven Laureys (Ed.) – 2005, ISBN 0-444-51851-7.
Volume 151: Neuroanatomy of the Oculomotor System, by J.A. Büttner-Ennever (Ed.) – 2006, ISBN 0-444-51696-4.
Volume 152: Autonomic Dysfunction after Spinal Cord Injury, by L.C. Weaver and C. Polosa (Eds.) – 2006, ISBN 0-444-51925-4.
Volume 153: Hypothalamic Integration of Energy Metabolism, by A. Kalsbeek, E. Fliers, M.A. Hofman, D.F. Swaab, E.J.W. Van
Someren and R. M. Buijs (Eds.) – 2006, ISBN 978-0-444-52261-0.
Volume 154: Visual Perception, Part 1, Fundamentals of Vision: Low and Mid-Level Processes in Perception, by S. Martinez-Conde,
S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-52966-4.
Volume 155: Visual Perception, Part 2, Fundamentals of Awareness, Multi-Sensory Integration and High-Order Perception, by
S. Martinez-Conde, S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-51927-6.
Volume 156: Understanding Emotions, by S. Anders, G. Ende, M. Junghofer, J. Kissler and D. Wildgruber (Eds.) – 2006, ISBN 978-0-
444-52182-8.
Volume 157: Reprogramming of the Brain, by A.R. Møller (Ed.) – 2006, ISBN 978-0-444-51602-2.
Volume 158: Functional Genomics and Proteomics in the Clinical Neurosciences, by S.E. Hemby and S. Bahn (Eds.) – 2006, ISBN 978-
0-444-51853-8.
Volume 159: Event-Related Dynamics of Brain Oscillations, by C. Neuper and W. Klimesch (Eds.) – 2006, ISBN 978-0-444-52183-5
Volume 160: GABA and the Basal Ganglia: From Molecules to Systems, by J.M. Tepper, E.D. Abercrombie and J.P. Bolam
(Eds.) – 2007, ISBN 978-0-444-52184-2
 PROGRESS IN BRAIN RESEARCH

VOLUME 161

NEUROTRAUMA:
NEW INSIGHTS INTO PATHOLOGY AND
TREATMENT

EDITED BY

JOHN T. WEBER
Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands
and
School of Pharmacy, Memorial University of Newfoundland, St. John’s, NL, Canada

ANDREW I.R. MAAS

Department of Neurosurgery, Erasmus Medical Centre, Rotterdam, The Netherlands

AMSTERDAM – BOSTON – HEIDELBERG – LONDON – NEW YORK – OXFORD

PARIS – SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO
2007
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands
Linacre House, Jordan Hill, Oxford OX2 8DP, UK

First edition 2007

Copyright r 2007 Elsevier B.V. All rights reserved

No part of this publication may be reproduced, stored in a retrieval system

or transmitted in any form or by any means electronic, mechanical, photocopying,
recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights
Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333;
e-mail: permissions@elsevier.com. Alternatively you can submit your request online by
visiting the Elsevier web site at http://www.elsevier.com/locate/permissions, and selecting
Obtaining permission to use Elsevier material

Notice
No responsibility is assumed by the publisher for any injury and/or damage to persons
or property as a matter of products liability, negligence or otherwise, or from any use
or operation of any methods, products, instructions or ideas contained in the material
herein. Because of rapid advances in the medical sciences, in particular, independent
verification of diagnoses and drug dosages should be made

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing in Publication Data

Neurotrauma: new insights into pathology and treatment. –

(Progress in brain research; v. 161)
1. Nervous system – Wounds and injuries
I. Weber, John II. Maas, Andrew I.R.
617.40 8044

ISBN-13: 9780444530172

ISBN: 978-0-444-53017-2 (this volume)

ISSN: 0079-6123 (Series)

For information on all Elsevier publications

visit our website at books.elsevier.com

Printed and bound in The Netherlands

07 08 09 10 11 10 9 8 7 6 5 4 3 2 1
 List of Contributors

J. Ai, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada
G.J. Amelink, Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical
Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
A.J. Baker, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada
D.C. Baptiste, Division of Cellular and Molecular Biology, Toronto Western Research Institute and
Krembil Neuroscience Centre, Toronto Western Hospital, 12th Floor Room 407, McLaughlin Pavilion,
399 Bathurst Street, Toronto, ON M5 T 2S8, Canada
M.D. Baumann, Department of Chemical Engineering and Applied Chemistry, University of Toronto,
Terrence Donnelly Centre for Cellular and Biomolecular Research, 164 College Street, Toronto, ON
M5S 3G9, Canada
S.J. Bernstein, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia,
PA 19104-6392, USA
N. Biasca, Clinic of Orthopaedic, Sports Medicine and Traumatology, Department of Surgery, Spital
Oberengadin, CH-7503 Samedan/St. Moritz, Switzerland
O. Bloch, Department of Neurological Surgery, University of California, San Francisco, CA, USA
and Brain and Spinal Injury Center, University of California, 1001 Potrero Avenue, Room 101, San
Francisco, CA 94110, USA
H.M. Bramlett, The Miami Project to Cure Paralysis, Department of Neurological Surgery, University of
Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA
M. Buchfelder, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg,
Schwabachanlage 6, D-91054 Erlangen, Germany
A. Buki, Department of Neurosurgery, Medical Faculty, Pécs University, Rét St. 2, Pécs, H-7624, Hungary
P. Bukovics, Department of Neurosurgery, University Medical School, Pécs University, Rét St. 2, Pécs,
H-7624, Hungary
M.R. Bullock, Department of Neurosurgery, Virginia Commonwealth University Medical Center,
Richmond, VA, USA
A.S. Cohen, Department of Pediatrics, University of Pennsylvania School of Medicine and Division of
Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA 19104-4318, USA
D.K. Cullen, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of
Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive,
Atlanta, GA 30332-0535, USA
E. Czeiter, Department of Neurosurgery, University Medical School, Pécs University, Rét St. 2, Pécs,
H-7624, Hungary
W.D. Dietrich, The Miami Project to Cure Paralysis, Department of Neurological Surgery, University of
Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA
T. Doczi, Department of Neurosurgery, University Medical School, Pécs University, Rét St. 2, Pécs,
H-7624, Hungary
J.J. Donkin, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide,
Adelaide, SA 5005, Australia

v
vi

A.-C. Duhaime, Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medi-
cal Center, Lebanon, NH 03756, USA
S. Durham, Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medical
Center, Lebanon, NH 03756, USA
O. Farkas, Department of Anatomy and Neurobiology, Medical College of Virginia Campus, Virginia
Commonwealth University, 1101 E. Marshall St., 12-048 P.O. Box 980709, Richmond, VA 23298, USA
M.G. Fehlings, Division of Cellular and Molecular Biology, Toronto Western Research Institute and
Krembil Neuroscience Centre, Toronto Western Research Institute, 4th floor West Wing Room 449, 399
Bathurst Street, Toronto, ON M5 T 2S8, Canada
C.L. Floyd, Department of Physical Medicine and Rehabilitation, Center for Glial Biology in Medicine,
547 Spain Rehabilitation Center, University of Alabama at Birmingham, Birmingham, AL 35249, USA
B.F. Fuller, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the
McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA
D.M. Geddes-Klein, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk,
Philadelphia, PA 19104-6392, USA
P.B. Goforth, Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Common-
wealth University, Richmond, VA, USA
M.S. Grady, Department of Neurosurgery, University of Pennsylvania School of Medicine, Philadelphia,
PA, USA
B.C. Hains, Department of Neurology and Center for Neuroscience and Regeneration Research, LCI-707,
Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA and Rehabilitation
Research Center, VA Connecticut Healthcare System, West Haven, CT 06516, USA
I.K. Haitsma, Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE
Rotterdam, The Netherlands
I. Hassan, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide,
SA 5005, Australia
R.L. Hayes, Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight
Brain Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA
F. Hesse, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Schwa-
bachanlage 6, D-91054 Erlangen, Germany
M. Horowitz, Laboratory of Environmental Physiology, Faculty of Dental Medicine, Hebrew University
of Jerusalem, Jerusalem 91120, Israel
P. Jendelova, Institute of Experimental Medicine ASCR, Videnska 1083, 14220 Prague 4, Czech Republic
C. Kang, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Terrence
Donnelly Centre for Cellular and Biomolecular Research, 164 College Street, Toronto, ON M5S 3G9,
Canada
Y. Katayama, Department of Neurological Surgery, Nihon University School of Medicine, 30-1 Oyaguchi
Kamimachi, Itabashi-ku, Tokyo 173-8610, Japan
T. Kawamata, Department of Neurological Surgery, Nihon University School of Medicine, 30-1 Oyaguchi
Kamimachi, Itabashi-ku, Tokyo 173-8610, Japan
A. Kleindienst, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg,
Schwabachanlage 6, D-91054 Erlangen, Germany
F.H. Kobeissy, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the
McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA
J.D. Kocsis, Yale University School of Medicine, Neuroscience Research Center (127A), VA Connecticut
Health Care System, West Haven, CT 06516, USA
E.J.O. Kompanje, Department of Intensive Care, Erasmus MC University Medical Center Rotterdam,
’s-Gravendijkwal 230, P.O. Box 2040, 3015 CE Rotterdam, The Netherlands
 vii

E. Kovesdi, Department of Neurosurgery, University Medical School, Pécs University, Rét St. 2, Pécs,
H-7624, Hungary
K.L. Lankford, Rehabilitation Research Center, Veterans Affairs Connecticut Healthcare System, West
Haven, CT 06516, USA
M.C. LaPlaca, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of
Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive,
Atlanta, GA 30332-0535, USA
B. Li, Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing,
P.R. of China
M.C. Liu, Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight Brain
Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA
N. Lukáčová, Institute of Neurobiology, Slovak Academy of Sciences, Šoltésovej 4, 040 01 Košice, Slovak
Republic
B.G. Lyeth, Department of Neurological Surgery, University of California, 1515 Newton Court, One
Shields Avenue, Davis, CA 95816, USA
A.I.R. Maas, Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE
Rotterdam, The Netherlands
G.T. Manley, Department of Neurosurgery, University of California, 1001 Potrero Avenue, Room 101,
San Francisco, CA 94110, USA
J. Maršala, Institute of Neurobiology, Slovak Academy of Sciences, Šoltésovej 4, 040 01 Košice, Slovak
Republic
W.L. Maxwell, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK
D.F. Meaney, Departments of Bioengineering and Neurosurgery, University of Pennsylvania, 3320 Smith
Walk, Philadelphia, PA 19104-6392, USA
M. Mesfin, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA
19104-6392, USA
W.J. Miller, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia,
PA 19104-6392, USA
J.P. Muizelaar, Department of Neurosurgery, University of California at Davis Medical Center, 4860 Y
Street, Suite 3740, Sacramento, CA 95817, USA
M.W. Oli, Banyan Biomarkers Inc., 12085 Research Drive, Alachua, FL 32615, USA
J. Orendáčová, Institute of Neurobiology, Slovak Academy of Sciences, Šoltésovej 4, 040 01 Košice, Slovak
Republic
A.K. Ottens, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the
McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA
J. Pal, Neurobiology Research Group of the Hungarian Academy of Sciences, University Medical School,
Pécs University, Rét St. 2, Pécs, H-7624, Hungary
E. Park, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada
B.J. Pfister, Department of Biomedical Engineering, New Jersey Institute of Technology, Newark, NJ,
USA
N. Plesnila, Laboratory of Experimental Neurosurgery, Department of Neurosurgery and Institute
for Surgical Research, University of Munich Medical Center, Grosshadern, Marchioninistr 15, 81377
Munich, Germany
P. Poon, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Donnelly
Center for Cellular and Biomolecular Research, Toronto, ON M5S 3E1, Canada
J.T. Povlishock, Department of Anatomy and Neurobiology, Medical College of Virginia Campus,
Virginia Commonwealth University, 1101 E. Marshall St., 12-048 Sanger Hall, P.O. Box 980709,
Richmond, VA 23298, USA
viii

G.R. Prado, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department
of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive,
Atlanta, GA 30332-0535, USA
C. Radtke, Department of Plastic, Hand and Reconstructive Surgery, Hannover Medical School, Hann-
over, Germany
D. Reglodi, Department of Anatomy, University Medical School, Pécs University, Szigeti u. 12, Pécs,
H-7624, Hungary
P. Reilly, Department of Neurosurgery, Royal Adelaide Hospital, Adelaide, SA 5000, Australia
M. Sasaki, Department of Neurology and Center for Neuroscience and Regeneration Research, Yale
University School of Medicine, New Haven, CT 06510, USA
L.S. Satin, Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Common-
wealth University, Richmond, VA, USA
E. Schwarzbach, Department of Pharmacology, University of Pennsylvania School of Medicine,
Philadelphia, PA, USA
N.A. Shein, Department of Pharmacology, School of Pharmacy, Hebrew University of Jerusalem,
Jerusalem 91120, Israel
E. Shohami, Department of Pharmacology, School of Pharmacy, Hebrew University of Jerusalem,
Jerusalem, 91120, Israel
M.S. Shoichet, Department of Chemical Engineering and Applied Chemistry, University of Toronto,
Terrence Donnelly Centre for Cellular and Biomolecular Research, 160 College Street, Room 514,
Toronto, ON M5S 3E1, Canada
C.M. Simon, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of
Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive,
Atlanta, GA 30332-0535, USA
P. Singh, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA
19104-6392, USA
F.J.A. Slieker, Department of Neurosurgery, Erasmus MC University Medical Center Rotterdam,
’s-Gravendijkwal 230, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands
J.M. Spaethling, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadel-
phia, PA 19104-6392, USA
D.G. Stein, Department of Emergency Medicine, Emory University School of Medicine, Brain Research
Laboratory, Suite 5100, 1365B Clifton Road N.E., Atlanta, GA 30322, USA
N. Stocchetti, Ospedale Policlinico IRCCS, Milan University, Milan, Italy
E. Sykova, Institute of Experimental Medicine ASCR, EU centre of Excellence, Videnska 1083, 14220
Prague 4, Czech Republic
D. Szellar, Department of Neurosurgery, University Medical School, Pécs University, Rét St. 2, Pécs,
H-7624, Hungary
A. Tamas, Department of Anatomy, University Medical School, Pécs University, Szigeti u. 12, Pécs,
H-7624, Hungary
C.H. Tator, Krembil Neuroscience Centre, Toronto Western Research Institute and Department of
Surgery, University of Toronto, 399 Bathurst Street, Toronto, ON M5 T 2S8, Canada
E. Thornton, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide,
Adelaide, SA 5005, Australia
R.J. Turner, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide,
Adelaide, SA 5005, Australia
C. Van Den Heuvel, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide,
Adelaide, SA 5005, Australia
 ix

I. Vanický, Institute of Neurobiology, Slovak Academy of Sciences, Šoltésovej 4, 040 01 Košice, Slovak
Republic
B.H. Verweij, Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical
Center Utrech, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
R. Vink, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide,
SA 5005, Australia
C.R. Von Reyn, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadel-
phia, PA 19104-6392, USA
K.K.W. Wang, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the
McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA
S.G. Waxman, Department of Neurology and Center for Neuroscience and Regeneration Research, LCI-
707, Yale School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA and Rehabilitation
Research Center, VA Connecticut Healthcare System, West Haven, CT 06516, USA
J.T. Weber, Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands and
School of Pharmacy, Memorial University of Newfoundland, Health Sciences Centre, 300 Prince Philip
Drive, St. John’s, NL A1B 3V6, Canada
X. Yao, Department of Neurological Surgery, University of California, San Francisco, CA, USA and Brain
and Spinal Injury Center, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA
Z. Zador, Department of Neurological Surgery, University of California, San Francisco, CA, USA and
Brain and Spinal Injury Center, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA
This page intentionally left blank
 Preface

This current volume of Progress in Brain Research, entitled ‘‘Neurotrauma: New insights into pathology and
treatment’’ is a compilation of chapters on major topics which were discussed at the 8th International
Neurotrauma Symposium held in Rotterdam, The Netherlands in May of 2006. Neurotrauma, which
encompasses both traumatic brain injury (TBI) and spinal cord injury (SCI), is the leading cause of death
and disability in young adults, and the incidence in older patients is increasing. As such, neurotrauma is a
field in medicine with one of the highest unmet needs. Concentrated, focused and multidisciplinary efforts
are required to combat this important disease. Therefore, one of the major goals of the symposium was to
stimulate cross-talk between basic researchers and clinicians, as well as between individuals in the fields of
TBI and SCI. We feel that TBI and SCI researchers have much to learn from one another in their approach
to treating these disorders. In addition, researchers have an obligation to design and conduct clinically
relevant research, just as clinicians have an obligation to learn about the latest approaches to treating
neurotrauma. Exciting findings from basic research open opportunities for improving treatment results,
and this volume represents a unique and comprehensive overview of the latest findings and insights on
translational research in neurotrauma.
Experts who delivered presentations at the symposium have submitted the majority of chapters in this
volume. In general, these individuals were selected because the topics that they, and their research groups,
are currently pursuing are pertinent for understanding the pathology and treatment of neurotrauma. We
are grateful also to other experts, who were not in attendance in Rotterdam, for writing chapters in specific
areas of neurotrauma, which we felt were needed to make this a more comprehensive volume. The result,
we think, is a report on some of the most relevant and up-to-date topics in the neurotrauma field. Also, we
are proud that the author base is a true representation of the international neurotrauma society (INTS),
with contributors from Western and Eastern Europe, The United States, Canada, Asia and Australia. The
global problem of neurotrauma can best be served by a worldwide representation of neurotrauma
specialists combining and exchanging expertise.
We are extremely grateful to all of the participants, assistants and sponsors who made the 8th Inter-
national Neurotrauma Symposium a resounding success, and are especially gracious to all of the authors
who have taken the time to contribute to this timely volume. In addition, we would like to apologize to
some individuals or groups who may feel excluded, which was primarily due to space limitations. This
exclusion is by no means intentional, and we have no doubt that new and additional topics, which are not
covered in the current volume, will soon be represented elsewhere on the international stage.

John T. Weber
Andrew I.R. Maas

xi
This page intentionally left blank
 Contents

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Section I. Introduction

1. The impact of neurotrauma on society: an international perspective

P. Reilly (Adelaide, South Australia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Section II. Biomechanics of Injury

2. CNS injury biomechanics and experimental models

M.C. LaPlaca, C.M. Simon, G.R. Prado and D.K. Cullen (Atlanta, GA, USA). . . . 13

3. Linking impact to cellular and molecular sequelae of CNS injury: modelling in vivo complexity
with in vitro simplicity
J.M. Spaethling, D.M. Geddes-Klein, W.J. Miller, C.R. von Reyn, P. Singh, M. Mesfin,
S.J. Bernstein and D.F. Meaney (Philadelphia, PA, USA) . . . . . . . . . . . . . . . . . . . 27

Section III. Pathological Mechanisms of Injury

4. Cellular and subcellular change evoked by diffuse traumatic brain injury: a complex web of
change extending far beyond focal damage
O. Farkas and J.T. Povlishock (Richmond, VA, USA) . . . . . . . . . . . . . . . . . . . . . 43

5. Astroglia: Important mediators of traumatic brain injury

C.L. Floyd and B.G. Lyeth (Davis, CA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

6. Rescuing neurons and glia: is inhibition of apoptosis useful?

E. Kovesdi, E. Czeiter, A. Tamas, D. Reglodi, D. Szellar, J. Pal, P. Bukovics,
T. Doczi and A. Buki (Pécs, Hungary) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

7. Substance P in traumatic brain injury

J.J. Donkin, R.J. Turner, I. Hassan and R. Vink (Adelaide, South Australia) . . . . . 97

xiii
xiv

8. Current concepts of cerebral oxygen transport and energy metabolism after severe traumatic
brain injury
B.H. Verweij, G.J. Amelink and J.P. Muizelaar (Utrecht, The Netherlands and
Sacramento, CA, USA). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

9. Progressive damage after brain and spinal cord injury: pathomechanisms and treatment strategies
H.M. Bramlett and W.D. Dietrich (Miami, FL, USA) . . . . . . . . . . . . . . . . . . . . . . 125

10. Injury-induced alterations in CNS electrophysiology

A.S. Cohen, B. J. Pfister, E. Schwarzbach, M.S. Grady, P.B. Goforth and
L.S. Satin (Philadelphia, PA, Newark, NJ and Richmond, VA, USA) . . . . . . . . . . 143

11. Traumatic injury of the spinal cord and nitric oxide

J. Mars̆ala, J. Orendáčová, N. Lukáčová and I. Vanický (Košice, Slovak Republic) 171

12. Aquaporins: role in cerebral edema and brain water balance

Z. Zador, O. Bloch, X. Yao and G.T. Manley (San Francisco, CA, USA) . . . . . . . 185

13. Sodium channel expression and the molecular pathophysiology of pain after SCI
B.C. Hains and S.G. Waxman (West Haven, CT, USA) . . . . . . . . . . . . . . . . . . . . 195

Section IV. Novel Aspects of Clinical Research in CNS Injury

14. Monitoring cerebral oxygenation in traumatic brain injury

I.K. Haitsma and A.I.R. Maas (Rotterdam, The Netherlands) . . . . . . . . . . . . . . . 207

15. Update on the treatment of spinal cord injury

D.C. Baptiste and M.G. Fehlings (Toronto, ON, Canada) . . . . . . . . . . . . . . . . . . 217

16. Cerebral contusion: a role for lesion progression

T. Kawamata and Y. Katayama (Tokyo, Japan) . . . . . . . . . . . . . . . . . . . . . . . . . 235

17. Ethical implications of time frames in a randomized controlled trial in acute severe traumatic
brain injury
E.J.O. Kompanje, A.I.R. Maas, F.J.A. Slieker and N. Stocchetti
(Rotterdam, The Netherlands and Milan, Italy). . . . . . . . . . . . . . . . . . . . . . . . . . 243

Section V. Emerging Topics in CNS Trauma

18. Experimental models of repetitive brain injuries

J.T. Weber (Rotterdam, The Netherlands) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

19. Minor traumatic brain injury in sports: a review in order to prevent neurological sequelae
N. Biasca and W.L. Maxwell (Samedan/St. Moritz, Switzerland
and Glasgow, UK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
 xv

20. Traumatic brain injury in infants: the phenomenon of subdural hemorrhage with
hemispheric hypodensity (‘‘Big Black Brain’’)
A.-C. Duhaime and S. Durham (Lebanon, NH, USA) . . . . . . . . . . . . . . . . . . . . . 293

21. Traumatic brain injury and Alzheimer’s disease: a review

C. Van Den Heuvel, E. Thornton and R. Vink (Adelaide, Australia). . . . . . . . . . . 303

22. The neurotrophic protein S100B: value as a marker of brain damage and possible therapeutic
implications
A. Kleindienst, F. Hesse, M.R. Bullock and M. Buchfelder (Erlangen-Nuremberg,
Germany and Richmond, VA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

23. Cerebellar injury: clinical relevance and potential in traumatic brain injury research
E. Park, J. Ai and A.J. Baker (Toronto, ON, Canada) . . . . . . . . . . . . . . . . . . . . . 327

24. Sex differences in brain damage and recovery of function: experimental and clinical findings
D.G. Stein (Atlanta, GA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

25. Heat acclimation: a unique model of physiologically mediated global preconditioning

against traumatic brain injury
N.A. Shein, M. Horowitz and E. Shohami (Jerusalem, Israel). . . . . . . . . . . . . . . . 353

Section VI. The Future of Neurotrauma: Developing Novel Treatment Strategies

26. In vivo tracking of stem cells in brain and spinal cord injury
E. Sykova and P. Jendelova (Prague, Czech Republic) . . . . . . . . . . . . . . . . . . . . . 367

27. Intrathecal drug delivery strategy is safe and efficacious for localized delivery to the spinal cord
M.S. Shoichet, C.H. Tator, P. Poon, C. Kang and M.D. Baumann
(Toronto, ON, Canada). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385

28. Decompression craniectomy after traumatic brain injury: recent experimental results
N. Plesnila (Munich, Germany) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

29. Novel neuroproteomic approaches to studying traumatic brain injury

A.K. Ottens, F.H. Kobeissy, B.F. Fuller, M.C. Liu, M.W. Oli,
R.L. Hayes and K.K.W. Wang (Gainesville and Alachua, FL, USA) . . . . . . . . . . 401

30. Remyelination of the injured spinal cord

M. Sasaki, B. Li, K.L. Lankford, C. Radtke and J.D. Kocsis (New Haven, CT, USA,
Chongqing, People’s Republic of China and Hannover, Germany) . . . . . . . . . . . . 419

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

This page intentionally left blank
 SECTION I

Introduction
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 1

The impact of neurotrauma on society: an

international perspective

Peter Reilly

Department of Neurosurgery, Royal Adelaide Hospital, Adelaide, South Australia 5000

Abstract: Neurotrauma, in many countries and particularly in the younger age group kills more people
than AIDS or cancer but unlike these diseases the causes are known and it is preventable. The costs to
communities in terms of suffering and economics are enormous. The common causes are road traffic
accidents, falls and violence. Neurotrauma affects particularly the developing world where it consumes
already over stretched health resources. In the developed world steps to reduce the incidence of neuro-
trauma and to treat the victims have had some effect nevertheless it stills remains an endemic problem
which does not receive the public awareness or the political support it deserves. For the victims there is
general agreement on the principles of clinical management but often difficulties in applying early and
effective care in countries with the greatest need because of shortage of facilities and expertise. To reduce
the overall burden of neurotrauma demands actions which extend from the political to basic patient care.
There have been remarkable advances in the understanding of acute brain and spinal cord injury and
encouraging possibilities for effective neuroprotection, repair and regeneration but in the broader context
prevention of neurotrauma is the urgent imperative. In this endeavour the neuroscientist has knowledge
which informs and encourages policy makers to take the steps necessary to reduce injury. These steps
require political will and community support for hard decisions which impact on the way people conduct
their daily lives. The WHO predicts that unless there are changes in present policies and if there are no
additional road safety countermeasures put in place, there will be a major increase in road traffic fatalities
over the next 20 years and beyond (World Health Organisation. (2004). www.who.int).

Keywords: epidemiology; prevention; brain and spinal injury statistics

Introduction firearms that are easily available; for poor indus-

It is a sad paradox that one of the greatest health
problems of this age is largely avoidable — neuro-
trauma of the brain and spinal cord is caused by
injury inflicted by misadventure, violence or care-
lessness. The principle causes of neurotrauma
reflect choices by individuals and by society for
fast, freely available and personal transport; for
Corresponding author. Tel.: +61-8-8222-5232;
Fax: +61-8-8222-5668; E-mail: p.reilly@adelaide.edu.au
trial safety standards and for sports such as boxing
which aim to inflict neurotrauma or other sports
which regularly do so.
Some accidents are unavoidable and unforesee-
able. Much of neurotrauma is all too predictable
and avoidable.
In this introduction to the Proceedings of the
8th International Neurotrauma Symposium I wish
to consider some global aspects of neurotrauma
and the role of the International Neurotrauma
Society (INTS) in the overall effort to reduce the
impact of neurotrauma on societies.

DOI: 10.1016/S0079-6123(06)61001-7 3
4
Neurotrauma as a global problem
The global incidence of traumatic brain injury
(TBI) is generally reported as 200 in 100,000 with
a mortality of 20 per 100,000. Incidences reported
from different countries range from 91 to 430 per
100,000 with mortalities of 9–89 per 100,000
(Fearnside and Simpson, 2005).
Despite awareness of the causes and of the eco-
nomic and human costs of neurotrauma the inci-
dence remains depressingly high and in developing
countries continues to rise. One of the difficulties in
describing accurately the global impact of neuro-
trauma is the variable way in which statistics are
recorded. Even in countries with well-developed
data collection systems, information on the full
impact of head and spinal cord injury is difficult to
obtain. Mortality figures are often based on A&E
admissions or hospital separations and do not in-
clude pre hospital deaths which account for 50%
of road trauma deaths.
The incidence of spinal cord injury ranged from
14.5 to 57.8 cases per million in a review of reported
studies; the low figure of 14.5 from Australia did
not include deaths at the scene of trauma (Ackery
et al., 2004).
Death and serious injury can be measured with
reasonable accuracy but the greater numbers of
mild and moderate head injuries are less likely to
be encompassed, yet they represent a major com-
ponent of the total burden of neurotrauma. Chang-
ing practices in hospital admissions also affect
comparisons between studies. The greater use of
CT scanning has lead to fewer hospital admissions,
increasing the proportion of severe head injury ad-
missions and reducing the mild and moderate head
injury admissions (Thurman and Guerrero, 1999).
A true measure of incidence and prevalence of ne-
urotrauma would require that all head and spinal
cord injuries be included.
In fact there is no fully comprehensive study of
the incidence of TBI in a defined population
(Fearnside and Simpson, 2005).
Similar problems affect spinal injury statistics
(Ackery et al., 2004).
Despite these difficulties and limitations accurate
data are essential in order to develop prevention
and management strategies.
These wide ranges no doubt reflect a combina-
tion of actual differences, varying definitions of
injury, different populations of inclusion and sam-
pling errors (Bruns and Hauser, 2003).
A number of studies indicate important regional
variations. A higher mortality for patients injured
in rural areas compared with urban areas has been
recorded in Australia and Taiwan (Chiu et al.,
1995, 1997; Hillier et al., 1997).
Injury and mortality rates are bimodal affecting
particularly the under 25 and over 65 years.
In the UK the TBI incidence of 10:100,000 rep-
resents 1% of all deaths but 20% of deaths in age
range 5–45 years. As in other developed countries,
TBI is the common cause of death in this age range.
The causes of neurotrauma
The main causes of neurotrauma, transport acci-
dents, falls and gunshot wounds, reflect societal
behaviour.
From the beginning motorised transport has in-
flicted injury to a degree which would undoubtedly
have shocked its inventors.
By the 1930s the motor car ‘‘had emerged as the
most persistent killer in the western world’’ (Gilbert,
1997). It is estimated that by 1997 25 million people
had died on the roads (Odero et al., 1997). In 2002
the global death rate from road traffic accidents was
1.2 million and between 20 and 50 million people
are estimated to be injured or disabled each year
(World Health Organisation, 2000, 2004). Of road
trauma deaths it is estimated that 1/3–1/2 are due to
brain injury. Vehicular accidents are also the major
cause of spinal cord injury worldwide (Ackery et al.,
2004).
The cost of road traffic injuries to society is
estimated as 2% of a country’s gross domestic
product (World Health Organisation, 2004).
Global figures encompass important regional
variations and trends (Soderlund and Zwi, 1995)
(Table 1). Ninety percent of road traffic accident
deaths occur in low and middle income countries
and at nearly twice the population adjusted rate of
high-income countries. Road traffic fatalities per
100,000 population in the year 2000 ranged from
28.3 in the African region to 5.9 in Great Britain
 5

Table 1. Mechanisms of TBI

RTA Falls Violence Sport Occupation

Australia (Tate et al., 1998) 40 21 8 25

India (Gururaj, 2002) 45–60 20–30 10–20
China (Wang et al., 1986) 32 22 1 15 24
USA (CDC, 1999) 49 26 17

Note: Numbers are percentages of trauma victims in each of the respective categories. RTA, road traffic accidents.

Fig. 1. Road traffic fatality trends in three high-income countries (Australia, United Kingdom, United States of America). Sources:
Transport Safety Bureau, Australia; Department of Transport, United Kingdom; Fatality Analysis Reporting System, United States of
America (World Health Organisation, 2004).

(World Health Organisation, 2004). Furthermore
TBI from road traffic accidents has been falling in
high-income countries for some years. In Australia
there has been an average yearly decline in TBI
of 5%, most due to an 8% decline in vehicle
occupancy injuries (O’Connor and Cripps, 1999).
Between 1968 and 1983 road traffic accident
mortality declined by more than 20% in Europe
(Fig. 1) In contrast it increased by more than
150% in Asian countries and more than 200% in
African countries (Soderlund, 1995) (Fig. 2).
Traffic type has important influences on injury
patterns. In Beijing one third of traffic deaths oc-
cur among bicyclists. In India pedestrians account
for 15–35% of injuries, 2-wheeler occupants
20–40% and bicyclists 5–15% (Gururaj, 2002).
Taiwan had a high rate of TBI from road traffic
accidents reported in 1991 to be 89 per 100,000
(Chiu et al., 1995). This was considered to be
related to the rapidly increasing use of motor
cycles, the most dangerous form of road transport
in current use anywhere in the world, and an
increasing component of traffic in developing
countries (World Health Organisation, 2004).
Other causes of neurotrauma also show marked
regional variation and changing trends. In the USA
firearm injuries exceeded road traffic accidents for
the first time in 1990 and this trend contrasted with
a fall in road traffic accident deaths (Sosin et al.,
1995). Other trends in injury causation have been
documented in several countries. A review of head
injury mortality from 1987 to 2000 in Sweden
showed a constant rate of injury over this period,
however a fall in transport-related injury was bal-
anced by an increase in falls below the age of 25 and
over the age of 65 (Kleiven et al., 2003). In the UK
6
Fig. 2. Global and regional road fatality trends, 1987–1995. Data are displayed according to regional classification of TRL Ltd.,
United Kingdom. The World Health Organisation, 2004 document from which this comes states that reproduced with permission of
the authors namely, Jacobs G, Aeron-Thomas A, Astrop A. Estimating Global Road Fatalities. Crowthorne, Transport Research
Laboratory, 2000 (TRL Report 445).
there were regional differences between the inci-
dence of road traffic accident versus falls and of
alcohol-related accidents (Kay and Teasdale, 2001).
Surveys from different countries also indicate
variation in the causation of spinal cord injury
between countries and over time. In most countries
vehicular accidents are the leading cause of spinal
cord injury in the younger age group but, as with
head injury, this figure is falling while the incidence
of injury due to falls in the older age group is
increasing (Cripps, 2006). Falls and violence are
more prevalent in developing countries and in the
lower economic strata of developed countries
(Ackery et al., 2004).
The community costs
Most studies of TBI focus on severe injury. Ninety
percent of patients presenting to A&E with head
injury in the UK had minor, 5% moderate and 5%
severe head injury. For every severe head injury
there were 17 mild or moderate injuries (Kay and
Teasdale, 2001). Most patients who suffer from mi-
nor TBI make uneventful recoveries but many
suffer continuing, disabling symptoms. WHO
figures indicate a possible treatment rate of moder-
ate head injury of 100–300 per 100,000 but the
actual rate may be greater than 600 per 100,000
(Cassidy et al., 2004). Thornhill et al. (2000)
reported that following mild head injury 51% had
continuing symptoms and after moderate head in-
jury 54%. Following mild head injury (GCS 13–15)
79% show persistent headache, 59% memory prob-
lems and 34% remained unemployed. Guerrero
et al. (2000) found that at 1 year 15% still had
symptoms. This represents an enormous and often
under treated group of potential disabilities.
A pathophysiological basis has been identified
for continuing symptoms (Blumbergs et al., 1994;
Bigler, 2001). The ensuing neuropsychological
effects and the value of psychological therapy have
been documented (Ponsford et al., 2000; Carroll
et al., 2004; Cassidy et al., 2004; Ponsford, 2005).
Providing care
Since the 1970s routine intracranial pressure
recording, the CT scan and Intensive Care Units
have focused treatment on severe head injury. The
general principles of management of brain and

spinal injury may be agreed upon but there are
major difficulties in delivering care to patients in
developing countries through shortage of neuro-
surgeons, a lack of trained personal for primary
care, particularly in rural areas, basic facilities and
transport systems for trauma victims. These pres-
sures are often compounded by the need to address
many other health problems with limited resources.
In such situations general surgeons may need to
be trained in acute neurosurgery.
In Australia distance imposes special needs and
the general surgeons in Darwin, a city of 70,000 and
2000 km from a neurosurgical centre, have man-
aged acute neurosurgery for many years with sup-
port by neurosurgeons and teleradiology (Treacy
et al., 2005). To meet these needs protocols for on
site management of neurotrauma by non neurosur-
geons have been developed (Neurosurgical Society
of Australasia Guidelines, 2000).
What can be done?
Neurotrauma is a major public health challenge.
Acknowledging that most TBI and spinal cord
injury is avoidable clearly emphasises the need for
strategies to identify causes and prevent injury.
Motor vehicle accident rates have fallen in many
countries in response to public awareness cam-
paigns and legislation (McDermott et al., 1996).
Each new legislated measure, for random alcohol
testing, speed cameras, seat belts or helmets, tends
to be followed by a modest fall in accident rates
(Campbell, 1987; MacLennan et al., 2004).
In response to the high number of motorcycle-
related head injuries in Taiwan a concerted effort
was made to introduce motorcycle helmets from
1997. In the first year motorcycle-related head in-
juries fell by 33%. There was a better than 90%
compliance rate with helmet wearing which was
mandated by law (Chiu et al., 2000a, b).
Motorcycle helmets have now been introduced via
legislation in many countries including Australia.
In the US the repeal of helmet legislation in
several states, provided a human experiment un-
likely to be approved by an ethics committee.
In Arkansas repeal of the state helmet law in 1997
led to a fall in helmet usage from 97 to 52% over
7
9 months. There was a marked increase in crash
scene fatalities and in admissions of non helmeted
crash survivors. It was concluded that helmet us-
age decreased fatalities by 20–40% (Bledsoe et al.,
2002). Other studies have supported the benefit of
motor cycle helmets in reducing head and facial
injury (Wagle et al., 1993; Liu et al., 2003).
Bicycle helmets have been mandatory in Australia
since 1990 and several studies have indicated their
benefit (Attewell et al., 2001; Thompson et al.,
1999).
There is great potential benefit for helmets in
developing countries where bicycle and increas-
ingly motor bikes are a major cause of TBI.
Societal attitudes are clearly key to any injury
prevention campaign (Weinstein, 1987). Cars are
undoubtedly safer now than they were decades
ago. Standard safety ratings are accepted by most
manufacturers, but they can be safer. It is often
remarked that safety does not sell compared with
appearance and performance.
There appears to be an acceptance of the risk of
neurotrauma even though there must be few
families who do not have a victim of neuro-
trauma of some degree. Those at highest risk have
the least perception of risk (Gilk et al., 1999).
Community support for research
Funding levels are a barometer of the value a so-
ciety puts on a particular field of research.
The Brain Injury Association of America esti-
mates that TBI in the US costs the community
$48.3 billion per year, $31.7 billion in hospitalisa-
tion costs and $16.6 billion from fatal TBI. Spinal
cord injury has been estimated to cost the US $9.7
billion each year (Berkowitz, 1998). In 2006 the US
National Institutes of Health, the major govern-
ment funding body, allocated $86 million for TBI
research and $87 million for spinal cord injury re-
search. Whether these are appropriate amounts
can be argued.
In Australia the principle commonwealth funded
body, the National Health and Medical Research
Council provided $600,000 towards neurotrauma
research in 2005–2006 from a total budget of $60
million, that is to say 1% of the research budget.
8
Researchers in neurotrauma in some states have
been more successful in gaining substantial grants
from the state government third party insurers.
One of the more innovative funding initiatives oc-
cured in Western Australia for 5 years fines raised
by speed cameras amounting to $500,000 per year
were allocated to trauma research.
In the US and Australia the funds directed to-
wards neurotrauma research appear small com-
pared with the societal costs and in comparison
with the funds directed to other major diseases.
Spinal research has profited greatly from patient
advocacy. A visit by Christopher Reeves to Sydney
in 1999 lead to the NSW Premier’s Initiative re-
search fund for spinal research. There are now
moves to establish national platforms to coordinate
research and develop funds for spinal injury re-
search — important given the need to support high-
quality research in a relatively small population.
The challenge
Neurotrauma affects all communities, all age
groups and walks of life. Despite understanding
of the causes and being able to treat victims better
it is an increasing global problem with important
regional variations. There is a lack of adequate
comprehensive data of the incidence of neurotrau-
ma in any community, most of all in those com-
munities which have the major burden of
neurotrauma. Neurotrauma is a particular burden
on the developing countries which have the least
capacity to manage it.
Reducing the incidence of neurotrauma can
only occur by greater political action and the en-
gagement of societies in order to see neurotrauma
as avoidable.
The role of the INTS and neuroscience
The INTS comprises a worldwide body of experts
in neurotrauma from clinical, non clinical and basic
science backgrounds. Many are actively involved in
acute neurotrauma care or working with victims of
the delayed effects of traumatic brain or spinal cord
injury or in injury prevention. The INTS aims to
encourage research in all aspects of neurotrauma by
providing an international forum for basic, preclini-
cal and clinical research. It aims to encourage
young neuroscientists to pursue their interests and
in this way to foster research around the globe.
Scientists have a role in neurotrauma prevention
and treatment in every level. Scientific evidence
underpins effective political and social action. The
neuroscientist can play a vital role in informing
and stimulating policy makers towards action that
will reduce the incidence of neurotrauma.
The INTS plays a highly significant role in en-
couraging the worldwide search for better treat-
ments for victims of neurotrauma.
It is important that the INTS liaises with other
neurotrauma organisations such as the WFNS,
National Neurotrauma Societies and regional
consortia so that the international neuroscience
community can use its wide experience and exper-
tise to the best effect.
References
Ackery, A., Tator, C. and Krassioukov, A. (2004) A global
perspective on spinal cord injury epidemiology. J. Neuro-
trauma, 21: 1355–1370.
Attewell, R.G., Glase, K. and McFadden, M. (2001) Bicycle
helmet efficacy: a meta-analysis. Accid. Anal. Prev., 33:
345–352.
Berkowitz, M., O’Leary, P., Kruse, D. and Harvey, C. (1998)
Spinal cord injury: an analysis of medical and social costs.
Demos Medical Publishing Inc., New York.
Bigler, E. (2001) The lesion(s) in traumatic brain injury:
implications for clinical neuropsychology. Arch. Clin.
Neuropsychol., 16: 95–131.
Bledsoe, G.H., Schexnayder, S.M., Carey, M.J., Dobbins,
W.N., Gibson, W.D., Hindman, J.W., Collins, T., Wallace,
B.H., Cone, J.B. and Ferrer, T.J. (2002) The negative impact
of the repeal of the Arkansas motorcycle helmet law. J.
Trauma, 53: 1078–1087.
Blumbergs, P.C., Scott, G. and Manavis, J. (1994) Staining of
amyloid precursor protein to study axonal injury in mild
head injury. Lancet, 344: 1055–1056.
Bruns Jr., J. and Hauser, W.A. (2003) The epidemiology of
traumatic brain injury: a review. Epilepsia, 44(Suppl 10):
2–10.
Campbell, B.J. (1987) Safety belt injury reduction related to
crash severity and front seated position. J. Trauma, 27:
733–739.
Carroll, L.J., Cassidy, J.D. and Holm, L. (2004) Methodological
issues and research recommendations for mild traumatic brain
injury: the WHO Collaborating Centre Task Force on Mild
Traumatic brain injury. J. Rehabil. Med., 43(Suppl): 113–135.

Cassidy, J.D., Carroll, L.J. and Peloso, P.M. (2004) Incidence,
risk factors and prevention of mild traumatic brain injury:
results of the WHO Collaborating Centre Task Force on
Mild Traumatic brain injury. J. Rehabil. Med., 43(Suppl):
26–28.
Center for Disease Control and Prevention (CDC). (1999)
Traumatic brain injury in the United States. A Report
to Congress Current Data on Traumatic Brain Injury and
Mortality.
Chiu, W.T., Ho, Y.S. and Lee, Y.S. (2000a) Sharp decline of
injury mortality rate in a developing country. Am. J. Public
Health, 90(5): 793–796.
Chiu, W.T., Hung, C.C., Le, L.S., Lin, L.S., Shih, C.J. and
LaPorte, R.E. (1997) Head injury in urban and rural
populations in a developing country. J. Clin. Neurosci., 4:
469–472.
Chiu, W.T., Hung, C.C. and Shih, C.J. (1995) Epidemiology
of head injury in rural Taiwan: a four year survey. J. Clin.
Neurosci., 2: 210–215.
Chiu, W.T., Kuo, C.Y., Hung, C.C. and Chen, M. (2000b) The
effect of the Taiwan motorcycle helmet use law on head
injuries. Am. J. Public Health, 5(90): 793–796.
Cripps, R.A. (2006) Spinal Cord Injury, Australia 2003–04.
Australian Institute of Health and Welfare, Canberra.
Fearnside, M.R. and Simpson, D.A. (2005) Epidemiology.
In: Reilly P.L. and Bullock R.B. (Eds.), Head Injury:
Pathophysiology and Management (2nd ed). Hodder
Arnold, London, pp. 3–25.
Gilbert, M. (1997) A History of the Twentieth Century, Vol. 1:
1900–1933. Harper Collins, London, 41p.
Gilk, D.C., Kronenfeld, J.J., Jackson, K. and Zhang, W. (1999)
Comparison of traffic accident and chronic disease risk
perceptions. Am. J. Health. Behav., 23: 198–209.
Guerrero, J.L., Thurman, D.J. and Sniezek, J.E. (2000)
Emergency department visits associated with traumatic brain
injury: United States, 1995–1996. Brain Inj., 14: 181–186.
Gururaj, G. (2002) Epidemiology of traumatic brain injuries:
Indian scenario. Neurol. Res., 24: 24–28.
Hillier, S., Hillier, J. and Metzer, J. (1997) Epidemiology of
traumatic brain injury in South Australia. Brain Inj., 11:
649–659.
Kay, A. and Teasdale, G. (2001) Head injury in the United
Kingdom. World J. Surg., 25: 1210–1220.
Kleiven, S., Peloso, P.M. and von Holst, H. (2003) The epide-
miology of head injuries in Sweden from 1987 to 2000. Inj.
Control Saf. Promot., 10: 173–180.
Liu, B., Norton, R., Blows, S. and Lo, S.K. (2003) Helmets for
preventing injury in motorcycle riders. The Cochrane Data-
base of Systemic Reviews, Issue 4.
MacLennan, P.A., McGwin, G., Metzger, J., Moran, S.G. and
Rue III, L.W. (2004) Risk of injury for occupants of motor
vehicle collisions from unbelted occupants. Inj. Prevent., 10:
363–367.
McDermott, F.T., Cordner, S.M. and Tremayne, A.B. (1996)
Evaluation of the medical management and preventability of
9
death in 137 road traffic fatalities in Victoria, Australia: an
overview. J. Trauma, 40: 520–535.
Neurosurgical Society of Australasia Guidelines. (2000) The
Management of Acute Neurotrauma in Rural and Remote
Locations. A Set of Guidelines for the Care of Head and
Spinal Injuries. The Neurosurgical Society of Australasia,
Royal Australasian College of Surgeons, Melbourne.
O’Connor, P.J. and Cripps, R.A. (1999) Traumatic Brain Injury
(TBI) Surveillance Issues. AIHW National Injury Surveil-
lance Unit, Flinders University Research Centre for Injury
Studies, Adelaide.
Odero, W., Garner, P. and Zwi, A. (1997) Road traffic injuries
in developing countries: a comprehensive review of epidemio-
logical studies. Trop. Med. Int. Health, 2: 445–460.
Ponsford, J. (2005) Rehabilitation interventions after mild head
injury. Curr. Opin. Neurol., 18: 692–697.
Ponsford, J., Wilmott, C., Rothwell, A., Cameron, P., Kelly,
A.M., Nelms, R., Curran, C. and Ng, K.T. (2000) Factors
influencing outcomes following mild traumatic brain injury
in adults. J. Int. Neuropsychol. Soc., 6: 568–597.
Soderlund, N. and Zwi, A.B. (1995) Traffic related mortality in
industrialized and less developed countries. World Health
Organisation Bull., 73: 175–182.
Sosin, D.M., Sniezek, J.E. and Waxweiler, R.J. (1995) Trends in
death associated with traumatic brain injury, 1979 through
1992. JAMA, 273: 1778–1780.
Tate, R.L., McDonald, S. and Lulham, J.M. (1998) Incidence
of hospital-treated traumatic brain injury in an Australian
community. Aust. N.Z. J. Public Health, 22: 419–423.
Thompson, D.C., Rivara, F.P. and Thompson, R. (1999) Hel-
mets for preventing head and facial injuries in bicyclists. The
Cochrane Database of Systemic Reviews, Issue 4.
Thornhill, S., Teasdale, G.M., Murray, G.D., McEwen, J.,
Roy, C.W. and Penny, K.I. (2000) Disability in young people
and adults one year after head injury: prospective cohort
study. BMJ, 320: 1631–1635.
Thurman, D. and Guerrero, J. (1999) Trends in hospitalization
associated with traumatic brain injury. JAMA, 282(10):
954–957.
Treacy, P.J., Reilly, P. and Brophy, B.P. (2005) Emergency
neurosurgery by general surgeons at a remote major hospital.
ANZ J. Surg., 75: 852–857.
Wagle, V.G., Perkins, C. and Vallera, A. (1993) Is helmet use
beneficial to motorcyclists? J. Trauma, 34: 120–122.
Wang, C.C., Schoenberg, B.S., Li, S., Yang, Y., Cheng, X. and
Bolis, L. (1986) Brain injury due to head trauma. Epidemio-
logy in urban areas of the People’s Republic of China. Arch.
Neurol., 43: 570–572.
Weinstein, N.D. (1987) Unrealistic optimism about susceptibi-
lity to health problems: conclusions from a community-wide
sample. J. Behav. Med., 10: 481–500.
World Health Organisation. (2000) Injury: leading cause of the
global burden of disease. www.who.int
World Health Organisation. (2004) World report on road
traffic injury prevention. www.who.int
 SECTION II

Biomechanics of Injury
This page intentionally left blank
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 2

CNS injury biomechanics and experimental models

M.C. LaPlaca, C.M. Simon, G.R. Prado and D.K. Cullen

Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia
Institute of Technology and Emory University, 313 Ferst Dr., Atlanta, GA 30332-0535, USA

Abstract: Traumatic brain injury (TBI) and traumatic spinal cord injury (SCI) are acquired when an
external physical insult causes damage to the central nervous system (CNS). Functional disabilities re-
sulting from CNS trauma are dependent upon the mode, severity, and anatomical location of the me-
chanical impact as well as the mechanical properties of the tissue. Although the biomechanical insult is the
initiating factor in the pathophysiology of CNS trauma, the anatomical loading distribution and the
resulting cellular responses are currently not well understood. For example, the primary response phase
includes events such as increased membrane permeability to ions and other molecules, which may initiate
complex signaling cascades that account for the prolonged damage and dysfunction. Correlation of insult
parameters with cellular changes and subsequent deficits may lead to refined tolerance criteria and facilitate
the development of improved protective gear. In addition, advancements in the understanding of injury
biomechanics are essential for the development and interpretation of experimental studies at both the in
vitro and in vivo levels and may lead to the development of new treatment approaches by determining
injury mechanisms across the temporal spectrum of the injury response. Here we discuss basic concepts
relevant to the biomechanics of CNS trauma, injury models used to experimentally simulate TBI and SCI,
and novel multilevel approaches for improving the current understanding of primary damage mechanisms.

Keywords: traumatic brain injury; traumatic spiral cord injury; neurotrauma; biomechanics; membrane
permeability; finite element analysis; injury tolerance criteria

Introduction mechanical input that has exceeded structural

Traumatic brain injury (TBI) and spinal cord in-
jury (SCI) result in a range of deficits depending
on the insult severity and the anatomical region(s)
affected. In traumatic central nervous injury
(CNS) injury, a mechanical impact (caused by
motor vehicle accidents, gunshot wounds, blows to
the head or spine, etc.) induces a mechanical re-
sponse at the cell and tissue level that ultimately
causes a pathophysiological injury response (as
shown in Fig. 1). In the acute phase of injury, pri-
mary damage occurs as a direct result of a
Corresponding author.
E-mail: michelle.laplaca@bme.gatech.edu
limits of cells and tissue. Primary damage is
characterized by nonspecific cell loss as well as
sublethal injury, which activates a cascade of seco-
ndary responses leading to prolonged cell death,
network dysfunction, and system level changes
(Fig. 2). Although the mechanical impact is the
initiating event in traumatic CNS injury, the rela-
tionship between biomechanical inputs and the
downstream pathological effects are not well un-
derstood. Investigation of relevant loading param-
eters and the resulting cell and tissue responses in a
variety of model systems is imperative for deci-
phering injury-induced pathophysiological mecha-
nisms and developing experimental models that
hold fidelity to the human clinical situation.

DOI: 10.1016/S0079-6123(06)61002-9 13
14

Mechanical Mechanical Injury

Input Response Response

Fig. 1. Steps in CNS trauma. Traumatic brain and spinal cord injuries result from mechanical loading to the tissue. Pathophysiological
events are initiated by the mechanical tissue response to impact.
insult

secondary response

primary lifespan
response

Fig. 2. Temporal aspects of injury. Mechanical loading causes an acute primary phase followed by a prolonged secondary phase. The
primary response is characterized by nonspecific cell loss, which initiates a cascade of complex secondary events such as inflammation,
excitotoxicity, and neurodegeneration.

A notable application of the study of bio-
mechanics in CNS trauma is the determination
of accurate tissue tolerances. Tissue tolerances are
defined as the point at which structural and/or
physiological failure occurs. An improved under-
standing of injury biomechanics and the resulting
brain and spinal cord responses will ultimately fa-
cilitate the development of improved protective
gear (e.g., helmets and seat belts). Determination
of tolerance criteria requires information about the
forces and deformations that lead to failure, but
the mechanical parameters (i.e., magnitude and
rate of force and deformation) are only partially
understood. Tissue response and tolerance criteria
for humans are largely based on cadaveric studies,
but may not accurately represent the properties of
living tissue. Basic cell and animal studies, in
which a defined mechanical insult can be applied
to live cells in culture or in an intact animal, have
an advantage for the determination of tissue tol-
erance and may lead to the refinement of human
tolerance criteria. These tolerance criteria must be
model-independent and represent inherent system
properties.
We will discuss basic biomechanical concepts as
they relate to traumatic brain and spinal cord in-
juries and present experimental models that have
been developed and characterized in an attempt to
mimic the forces and deformations occurring in
human CNS trauma. The mechanisms by which
the mechanical response to a traumatic insult leads
to dysfunction are complex, yet can be simplified
using controlled cellular injury models that ac-
count for deformation magnitude and rate. Bio-
mechanically relevant in vitro TBI models, used in
combination with animal studies and computer
simulations, may lead to improved cellular and
tissue injury tolerance criteria as well as a more
complete understanding of the relationship bet-
ween the biomechanical input and pathophysio-
logical changes. This multilevel approach will be
discussed with respect to selection of experimental
models, development of mechanistically driven
treatment strategies, and future research priorities.
Basic biomechanics
Biomechanics is the study of forces and physical
responses in stationary (static) and moving (dy-
namic) biological systems. A system (in the case of
traumatic CNS injury — the brain or spinal cord)
reacts in a specific way when a force, or load, is
placed on it. These external loads may result in
initial damage or lead to delayed damage. The
point at which loading causes tissue damage is the
threshold (or the tolerance) of the system and is
dependent on the type and duration of the load.
The basic terms, or descriptors, that biomechani-
cians use to describe applied loads are force and
stress and the resulting responses are deformations
and strains.

Force is defined as the action of one body (a
physical entity in the system, such as a windshield)
on another (as a result of an impact), which will
cause acceleration of the second body (e.g., the
head) unless acted upon by an equal and opposite
action counteracting the effect of the first body.
The unit is a Newton (N); 1 N is the force that will
give 1 kg an acceleration of 1 m/s2 (English unit is
pound-force, lbf). When forces are generated in
tissue, deformation may ensue depending on the
material properties and the nature of the force itself.
Deformation is defined as the change in shape of a
body undergoing a force. A rigid body, for exam-
ple, would experience extremely small deforma-
tions, while biological tissue (usually referred to as
deformable or nonrigid) can often undergo sub-
stantially large deformations.
Stress is another term frequently used in bio-
mechanical analysis and refers to the distribution
of force relative to the area on which it acts.
Normal stresses (designated by the Greek letter
sigma (s)) act perpendicular to the surface, while
shear stresses (designated by the Greek letter tau
(t)) act tangential to the surface. The unit is the
Pascal (Pa); 1 Pa ¼ 1 N/m2. A given force acting
on a small surface produces greater stress than the
same force acting over a larger surface. In other
words, the amount of mechanical stress created by
a force is dependent on the size of the area over
which the force is applied. The resulting strain
that occurs relates the deformed state of the body
to the undeformed state and is unitless. Exten-
sional strain is the change in length divided by the
original length (designated by the Greek letter
epsilon) (e ¼ Dl/lo) and can be further classified as
being in tension (positive strain) or compression
(negative strain). Extensional strain results from
stresses generated from linear (or translational)
loads. Shear strain, often resulting from rotational
loads, is also the change in length divided by the
original length (designated by the Greek letter
gamma) (g ¼ Dl/lo). Brain tissue is thought to be
more sensitive to shear strain than extensional
strain (Holbourn, 1943). Therefore loading that
involves rotation of the head has been thought to
result in more severe injuries, although this as-
sumption has recently been questioned (King
et al., 2003). The relationships between stress
15
and strain are referred to as constitutive relation-
ships and the resulting equations are used to de-
fine behavior of the tissue (or the mechanical
response).
The basic mechanics terms defined above are
valuable in describing the conditions that lead to
injuries, although several factors surround bio-
mechanical analysis of damage prediction. Me-
chanical conditions can be referred to as the insult
parameters and the result as the injury (Fig. 1).
Two broad categories of insults can be defined as
static and dynamic loading, with dynamic loading
being the most common. The mechanical response
to insult is the tissue deformation or strain and
will initiate the ensuing pathological events. The
insult parameters and the mechanical response
will dictate the types of injury (focal and/or
diffuse). We will consider the categories of insults,
the mechanical response to traumatic insult,
the types of injuries produced, as well as two
overlapping response phases (primary and second-
ary) in light of the biomechanical fidelity of
experimental models used to simulate these con-
ditions.
Traumatic mechanical insults
Loads are described as direct (e.g., physical con-
tact between the head and another object) or in-
direct (e.g., as the result of motion of the head). In
indirect loading, acceleration of the second body
(e.g., the head) can act analogously to applied
forces. Loads can be translational (linear), rota-
tional, or angular (a combination of translational
and rotational). The type of force and the direc-
tion, or plane, of loading, will also affect the re-
sulting mechanical response in the tissue. The
extent and severity of deformation increases with
increasing force, and this relationship is nonlinear.
In other words, the increase in tissue damage may
be greater than the proportional increase in force.
Static loading is a very slowly applied direct load.
Usually there are no deficits until there is sub-
stantial tissue deformation. These loading condi-
tions are relatively rare and often occur in human
entrapment situations (e.g., earthquakes). Dy-
namic loading, on the other hand, can occur quite
16
rapidly (under 1 s, often o50 ms) and is the most
common cause of TBI and SCI. Dynamic loading
can further be broken down into impact loading
(direct loading where an impact occurs with an
object hitting the head or the head hitting an
object) or impulsive loading (indirect loading
where no contact occurs). Impact loading can be
either focal or diffuse, depending on the magni-
tude of the force and area of impact. Although
pure impact would involve contact with no head
movement, impact loading is usually a combina-
tion of contact forces — from the impact itself —
and inertial forces — from the motion of the head
and the brain within the skull. It is important to
consider the size, mass, and hardness of the im-
pacting object as well as the surface area and ve-
locity at which contact occurs. For example,
impact with smaller objects (i.e., o2 in. in diam-
eter) results in high local stress concentrations and
therefore is associated with a greater risk for more
local and severe damage and is more likely to re-
sult in tissue penetration. Impulsive loading is due
to inertial forces alone and leads to diffuse
brain injuries. Models of impulsive loading in-
clude angular acceleration of the head, yet many
of the models utilized for impact loading are de-
signed to deliver a rapid bulk insult that has in-
ertial components. Ultimately, the response is
dictated by the mechanical response of the tissue
or cells.
Loads, in particular rotational inputs to the
brain, however, do not linearly scale between hu-
mans and animal, as the mass of the brain is much
smaller. In fact, to produce an equivalent rota-
tional load in a rodent brain as in a human brain
the angular acceleration would need to be approx-
imately two orders of magnitude higher. This an-
atomical complexity introduces difficulty in
directly linking pathological consequences to the
biomechanical input. In addition to these con-
straints in animal modeling, the regional stresses
and strains have yet to be well characterized. Fu-
ture investigations to determine the relationships
between biomechanical parameters and cellular
responses will require a detailed spatial characteri-
zation of local cellular stresses and strains in ani-
mal models of CNS trauma.
Mechanical response to traumatic insult
A traumatic insult to brain or spinal cord will lead
to a mechanical response of the tissue that is de-
pendent on the mode, severity, and anatomical lo-
cation of the impact as well as the mechanical
properties of the tissue. The mechanical properties
of a tissue vary from individual to individual, as
well as with age and previous injuries or disease
(Prange and Margulies, 2002). In addition, cellular
orientation and tissue composition varies among
anatomical regions of the brain and spinal cord,
creating nonuniform (or heterogeneous) mechani-
cal properties that directly affect structural and
functional tolerances as well as the load distribution
throughout the tissue upon mechanical loading.
Because of the properties of soft tissues, like
brain and spinal cord, both the rate and the du-
ration of the insult will also influence the response.
Loads that are applied quickly may incur more
damage due to the material properties of CNS tis-
sue. When loads are applied at a high rate, the
tissue cannot absorb (or reduce) the force fast
enough and can fail both structurally and func-
tionally. In contrast, slowly applied loads give the
tissue ‘‘time’’ to reduce the force and generally re-
sult in less damage. For short durations of force,
much of the effects of the force are reduced. As the
duration of force increases, less reduction occurs
and therefore less force is needed to produce tissue
deformation. These behaviors are defined by a
mechanical property termed viscoelasticity.
Types of traumatic CNS injury
Focal injuries result from direct loading and can
often occur without widespread, or diffuse, dam-
age. Focal injuries are typically induced when an
object penetrates the skull or vertebral column as a
result of a motor vehicle accident, gunshot wound,
or a blow. As a result, macroscopically visible
damage is typically visible at the site of impact,
and the clinical symptoms are often very specific to
the area that is directly injured. Focal injuries to
the brain include epidural hematomas and skull
fracture (with or without brain damage). When

there is osteal or dural compromise, this is often
termed open head injury in the clinical setting.
Contact loading can also result in coup (at the site
of impact) and contra-coup (away from the site of
impact) contusions to the brain, involving both
cellular and vascular components. Focal injuries
account for one-half of all severe head injuries, but
two-third of all deaths in this group (Thurman and
Guerrero, 1999; Adekoya et al., 2002).
SCI is most commonly caused by fracture and
dislocation of the spinal column, resulting in a fo-
cal injury. The mechanical impact causes displace-
ment of bone fragments, intervertebral discs, or
ligaments, resulting in transient compression or
contusion of spinal cord tissue. Spinal cord is com-
pressed at the site of impact that causes the sur-
rounding tissue to lengthen in the longitudinal
direction. Tissue near the center of the spinal cord
is most vulnerable, suggesting that the mechanical
loads are highest in this anatomical region. Large
myelinated axons in the surrounding white matter
are also highly susceptible to mechanical damage,
due to stress concentrations at the nodes of
Ranvier (Maxwell, 1996). As in TBI, the rate,
magnitude, and duration of the biomechanical in-
sult can dictate the injury response and may affect
functional outcome. Slow stretching of the spinal
cord results in very little tissue damage. In fact,
increasing the length of the spinal cord up to twice
the original length results in very little damage if
the elongation is applied slowly (Shi and White-
bone, 2006). However, biomechanical inputs ap-
plied rapidly or for an extended duration (longer
than 20–30 min) may surpass tissue thresholds and
result in irreversible damage.
Diffuse injuries are most often caused by inertial
loading, which describes the motion of objects.
The acceleration (velocity change divided by
change in time) is an important parameter in de-
termining tissue response. Higher accelerations
correspond to higher forces (force equals mass
times acceleration, Newton’s second law). This
must be taken into account when establishing
thresholds for tissue damage. Because of the com-
plex head-neck dynamics, the brain can undergo
high acceleration when subjected to an external
load and therefore TBI often manifests as a diffuse
17
injury. When the acceleration is translational, in-
juries tend to be localized to a smaller area. Ro-
tational acceleration, on the other hand, can lead
to large strains deep within the brain, resulting in
diffuse axonal injury (DAI) (Gennarelli et al.,
1982). Most injuries seen clinically are a combina-
tion of translational and rotational accelerations
(referred to as angular acceleration). Diffuse inju-
ries are thought to occur as a result of not only the
acceleration portion of loading, but also from the
deceleration portion of the insult, creating very
fast moving, uneven load distributions (Margulies
et al., 1990). Diffuse strains can lead to differential
movement of the skull relative to the brain, caus-
ing parasagittal bridging vein injury, as well as in-
tracerebral hemorrhage. Diffuse injury to the brain
tends to lead to widespread dysfunction, making
these injuries the most prevalent cause of persistent
neurological disability. Clinically, diffuse injury is
often seen in closed head injury and arises most
often from motor vehicle accidents.
Experimental modeling of traumatic CNS injury
Experimental models of CNS injury have been in-
valuable in the investigation of pathological mech-
anisms and treatment strategies. However, due to
the variable nature of clinical traumatic CNS in-
jury (e.g., inconsistencies in the anatomical loca-
tion of impact and the magnitude and duration of
loading), experimental models must simplify the
human condition in order to create a reproducible
injury that can be utilized for controlled experi-
mental testing. Although relevance to the clinic
may be sacrificed, these simplifications allow the
assessment of various outcome measures at the
cellular, tissue, and organism level in response to
defined bulk loading parameters.
In vivo animal models preserve much of the
complexity associated with human traumatic CNS
injury while allowing the investigator to experi-
mentally manipulate certain parameters (e.g.,
treatment variables, time of sacrifice) that are not
possible in humans. In the study of injury biome-
chanics, in vivo models provide a more complete
representation of the human brain and spinal cord
18
because they more closely mimic the material
properties and anatomical architecture. Therefore,
the load distribution and structural failure in an-
imal models are expected to be similar to human
injury when clinically relevant biomechanical load-
ing parameters are applied in a scale-appropriate
manner.
In vivo models commonly used in TBI and SCI
research have been used to experimentally repre-
sent aspects of the biomechanics of CNS trauma.
Direct loading has been mimicked using contu-
sion, weight drop, fluid percussion, or compression
injuries. Contusion or weight drop involves brief,
rapid loading of CNS tissue using a piston or a
weight dropped from various heights (Dixon et al.,
1991; Anderson and Stokes, 1992; Marmarou
et al., 1994; Young, 2002; Scheff et al., 2003).
These models are designed to deliver a rapid bulk
insult that has both impact and inertial compo-
nents. Compression injury is also used to experi-
mentally replicate mechanical loads applied to
spinal cords over long durations (e.g., due to ab-
normal, prolonged twisting of the spine during an
automobile accident) (Rivlin and Tator, 1978;
Dolan and Tator, 1979). Inertial loading experi-
enced during TBI is modeled with fluid percussion
injury (Dixon et al., 1987; McIntosh et al., 1989;
Thibault et al., 1992) (which has components of
impact loads as well) and angular acceleration of
the head (Gennarelli et al., 1982; Smith et al.,
1997), which results in characteristic pathophysio-
logical changes such as DAI.
In vitro TBI models offer several advantages
over whole animal models, including control over
cellular components and real-time measurement of
acute responses. Neural cultures and tissue explants
have been subjected to compression, tension, or
shear to experimentally mimic aspects of CNS
trauma (see Morrison et al., 1998b). Models in-
clude deformable membranes that are stretched
biaxially (Ellis et al., 1995; Cargill and Thibault,
1996; Geddes and Cargill, 2001; Morrison et al.,
1998a) or uniaxially (Pfister et al., 2003; Lusardi
et al., 2004) to transfer strain to attached cells,
some with the capability of deforming neurites
aligned longitudinally to the strain field (Galbraith
and Thibault, 1993; Smith et al., 1999). These in
vitro models allow for isolation of specific
biomechanical parameters (e.g., deformation mode,
rate, and magnitude), allowing for systematic as-
sessment of cellular responses to defined inputs.
The recent development of a three-dimensional (3-
D) model in which neural cells are cultured in a
hydrogel offers an intermediate degree of complex-
ity, as bulk deformation of the culture results
in heterogeneous strain fields at the cellular
level depending on the orientation of the cell
within the matrix (LaPlaca et al., 2005; Cullen
and LaPlaca, 2006).
Tolerance criteria for CNS injury
To date, several cellular tolerance criteria have
been established to describe the contribution of
both acceleration and pulse duration for a specific
head injury (e.g., skull fracture, concussion), in-
cluding the Wayne State Tolerance Criteria
(Lissner et al., 1960), the Gadd Severity Index
(Gadd, 1966), and the Head Injury Criterion (Ver-
sace, 1971). The basic overlying principle is that
short pulses of high acceleration can produce in-
jury, while lower accelerations require longer
pulses to produce injury. These criteria have con-
tributed to the development of a fundamental
foundation; however, the tolerance stipulations
have been based on cadaver or primate data in
which the measure of injury did not consider
damage at the cellular level. The efforts at the
National Highway Traffic & Safety Administra-
tion (NHTSA) have produced models of the head
in the SIMon project. The predictive capability of
SIMon and other computational models hinge on
adoption of rational and experimentally verified
thresholds for damage. Because different regions
of CNS tissue have different cellular orientations
and tissue composition, resulting in nonuniform
(or heterogeneous) mechanical properties, struc-
tural and functional tolerances of the brain and
spinal cord differ depending on the region af-
fected. More complex and realistic computer mod-
els have been developed to provide more accurate
information relevant to the biomechanics of injury
(e.g., Zhang et al., 2004). Iterative verification of
these models is imperative to their successful ap-
plication. Measurement of brain tissue strain

during a dynamic mechanical event is exceedingly
difficult in the intact animal or postmortem human
subject (Hardy et al., 2003). Consequently, it has
proven challenging to determine quantitative tol-
erances to be used for the damage measures em-
bedded in computer models. Current efforts have
utilized existing experimental data and scaling re-
lationships to empirically derive thresholds to pre-
dict physiological outcome in animal experiments
(Takhounts et al., 2003). Therefore, experimental
models that enable the correlation of strain and
acute injury could potentially determine detailed
cellular tolerances.
Response phases of traumatic CNS injury
Acute cellular response
The initial damage that is a direct result of loading
to the brain is defined as the primary phase of in-
jury. Biomechanicians study this phase in order to
determine tissue tolerances to mechanical loading
because the effects of the mechanical insult can be
more easily isolated from biochemical events oc-
curring in the secondary or more chronic phase.
Our understanding of tolerances at the cellular
level is vital to developing better safety equipment
and understanding mechanotransduction in the
pathological range. At the time of the insult there
may be a varying amount of primary damage that
results from the physical force itself. This includes
compromised skin, bony fractures, tissue tearing,
cellular rupture, and reorientation of the tissue
components. If a deformation threshold is sur-
passed, these structural failures result and can se-
verely compromise brain function.
Due to the heterogeneity of CNS tissue, it is
likely that loads and deformations experienced by
cells in various anatomical regions are not con-
sistent and cannot be accurately estimated by sim-
plistic models assuming homogeneity. Certain
anatomical regions may be subjected to more se-
vere loading during impact because of differences
in the material properties in that particular loca-
tion (due to variations in cellular orientation, mye-
lination, etc.). Anatomical regions experiencing
larger strains would therefore be expected to be
19
more susceptible to primary damage caused by the
mechanical insult itself. Although identification of
these regions would allow more accurate correla-
tions between the mechanical input and patho-
physiological responses, very little is currently
known about local cellular strains in animal mod-
els of CNS trauma, mainly due to limitations in
detection techniques.
One approach for addressing these technical
limitations is the development of more sensitive
methods for the detection of mechanically induced
damage. Although detection of structural failures
can be relatively obvious in some instances (such
as the presence of large focal lesions), more subtle
damage may also be present and can provide a
unique opportunity for assessment of local cellular
strains after trauma. Visualization of the anatom-
ical localization of this mechanical damage can
provide a more sensitive measure of the load dis-
tribution throughout the tissue. We and others
have investigated nonspecific plasma membrane
damage as an indicator of mechanical damage in
various models of TBI and SCI (Pettus et al., 1994;
LaPlaca et al., 1997; Shi and Borgens, 2000;
Geddes et al., 2003; Farkas et al., 2006). This type
of cellular damage occurs as a direct result of me-
chanical loading, creating rips or tears in the
plasma membrane at regions of high local strain.
We have utilized Lucifer yellow as an indicator
of acute biophysical membrane failure after TBI
and SCI. Lucifer yellow is normally membrane-
impermeable; therefore, cellular presence of this
molecule can be used to detect plasma membrane
compromise. In these experiments, Lucifer yellow
was injected intrathecally 3 h prior to brain or spi-
nal cord contusion, and animals were sacrificed
10 min after injury (a schematic of the injury de-
vices are illustrated in Fig. 3). Histological evi-
dence demonstrated heterogeneous uptake of the
permeability marker in various anatomical loca-
tions (as shown in Fig. 4), indicating that the dis-
tribution of mechanical loading in CNS tissue is
complex and not well understood. Although we
have focused on acute membrane damage as an
indicator of the load distribution throughout the
brain and spinal cord, others have explored mem-
brane compromise as an initiator of downstream
pathological events. Cell membrane damage can
20

A. Infinite Horizons spinal cord contusion device B. Cortical contusion impact device

Displacement
sensor and motor LVDT
controller

Impactor tip
Impactor tip Load cell
Air Stereotactic
Spinal clamps
tank frame
Control Box
500 µm

Fig. 3. In vivo contusion injury devices. Injury devices are used to experimentally deliver prescribed injury parameters to the exposed
brain or spinal cord. For example, the Infinite Horizons spinal cord contusion device (A) allows the user to select an impact force for
injury, while the controlled cortical impact device (B) utilizes a pneumatic system to injure the brain at a defined tissue displacement.

A. TBI B. SCI

100 µm

500 µm 500 µm

Fig. 4. Acute cellular permeability following TBI and SCI. In the acute phase of traumatic injury, the plasma membrane becomes
damaged due to local cellular strains that exceed structural thresholds. Lucifer yellow uptake in the injured brain (A) and spinal cord
(B) demonstrates a heterogeneous distribution of membrane failure, suggesting that loading is not evenly distributed throughout the
CNS parenchyma.

lead to abnormal ion movement across the mem-
brane, resulting in pathophysiological changes
such as conduction block, neurofilament compact-
ion, and impaired axonal transport (Pettus et al.,
1994; Shi and Pryor, 2002). Thus, mechanical
loading may directly result in pathophysiological
changes.
Experimental evidence has demonstrated that
the extent of membrane compromise is dependent
on the magnitude and rate of strain (LaPlaca et al.,
1997; Geddes et al., 2003; Shi and Whitebone,
2006). In addition, others have suggested that the
mode of injury may play a critical role in dictating
the extent of mechanically induced cell membrane
damage (Geddes-Klein et al., 2006). After TBI,
membrane disruption has been shown to occur
after focal injury in a contusion model (Fig. 4) as
well as diffuse loading after impact acceleration
injury (Farkas et al., 2006), with patterns of
marker uptake specific to the mode of impact. Be-
cause there is a correlation between injury severity
and membrane compromise, permeability markers
can therefore be used as an indicator of the extent
of local cellular loading parameters. For example,
experiments conducted in our laboratory have
demonstrated more extensive permeability marker
uptake in specific hippocampal regions after con-
tusion injury, suggesting that local cellular loading
is more severe in certain anatomical locations.
These data may explain the preferential cell death
seen in these regions in the subacute and chronic
phases, as mechanical damage during the initial
 21

3-D Neural
Co-Cultures
piston

confocal microscope

3-D Cell Shearing Device 3-D Cell Compression Device

Trapezoidal Input

undeformed undeformed
Strain

0.50
0.25

10 20 30 40
Time (ms)
deformed deformed

Fig. 5. In vitro injury devices for shear and compression injuries. Neural cells cultured in a 3-D configuration were subjected to either
shear or compression injury with a prescribed strain magnitude and rate. This experimental model provides control over bulk material
deformation, while local strains may vary due to cellular orientation within the matrix.

impact in these regions may make the cells more
susceptible to death and/or dysfunction during the
secondary phase of injury.
Although in vivo models can provide a more
anatomically accurate representation of the struc-
tural and functional damage associated with hu-
man CNS injury, in vitro models allow for more
thorough investigation of tissue tolerances because
biomechanical insult parameters can be more pre-
cisely controlled and manipulated. In a recent
study, the effects of both shear and compression
modes of impact were investigated (Fig. 5). This is
an example of how strain estimations derived from
finite element analysis (FEA) can be applied to
simplified culture environments to isolate compo-
nents of the heterogeneous mechanical response
(Fig. 6). Briefly, mixed cultures consisting of neu-
rons and astrocytes were plated in a 3-D matrix
and subjected to either shear or compressive load-
ing (0.50 strain at strain rates of 1, 10, or 30 s
1).
Both types of loading resulted in significant in-
creases in membrane permeability in a strain rate
dependent manner, with no differences in the den-
sity or percentage of permeabilized cells based on
mode of deformation. However, the degree of
permeability marker uptake per permeabilized cell,
potentially a gauge of local cellular strain/stress
concentrations, was greater following shear defor-
mation (Fig. 7). Interestingly, the density of dead
cells was also significantly greater following shear
deformation (5–7 fold increase) compared to com-
pression (2-fold increase), suggesting that there is a
correlation between the degree of membrane per-
meability and the extent of cell death. This study
agrees with previous work demonstrating that
shear deformation is the primary mode of tissue
failure (Holbourn, 1943; Sahay et al., 1992).
Although this study evaluated cellular responses
based on different modes of bulk deformation, lo-
cal cellular strains are heterogeneous, and may be
a function of cell orientation with respect to the
bulk strain field (amongst other factors) (LaPlaca
et al., 2005; Cullen and LaPlaca, 2006). We have
demonstrated that neuronal response to loading
depends on cell orientation, and hence local cellu-
lar strain, where maximal neurite loss occurred at
shear-dominated strain regimes (LaPlaca et al.,
2005). Ongoing in vitro studies are aimed at de-
fining the biomechanical parameters (deformation
mode, rate, and magnitude) that lead to structural
22

A
B

Finite element modeling of strain propagation following a focal insult

(controlled cortical impact) in a rat.
in vivo simulations in vitro
A.

Shear-dominated

shear

B.

Compression-dominated

compression

Unloaded region

static

Fig. 6. Finite element model (FEM) simulations and corresponding isolation of loading components in vitro. Traumatic loading to the
brain results in the generation of complex, heterogeneous strain patterns at tissue and cell levels. Heterogeneity in the cellular response
to traumatic loading may be due to several factors, including mode of deformation, cell population, and cell orientation. Neural cell
tolerances to traumatic loading may therefore be elucidated based on these parameters.

Static Control 1 s-1 rate
Compression
Shear
Fig. 7. Acute cellular permeability increases in vitro depend on mode of bulk loading. Representative confocal reconstructions of
calcein+ cells following static control conditions or mechanical loading (0.50 strain at 1, 10, or 30 s
1 strain rate). Calcein, a normally
cell-impermeant molecule, was added to the extracellular space prior to loading but becomes intracellularly sequestered following
loading. Reconstructions from 50 mm thick z-stacks are shown here.
failure at the cellular level. Models of neural
trauma that represent the related biomechanics
and pathophysiology are important for the eluci-
dation of cellular tolerances and the development
of mechanistically driven intervention strategies.
Secondary response
Primary damage initiates a cascade of secondary
responses, leading to cell death, network dysfunc-
tion, and system-level changes (Fig. 8). While there
is no absolute time when primary damage evolves
into delayed effects, the secondary phase of injury
can be defined as any consequence of the primary
insult. This may be in the acute (minutes to hours)
period or in a more delayed fashion (days to
months) and is dependent on the severity of the
initial insult, as well as the health and age of the
individual. There is a role for biomechanics in de-
termining injury mechanisms in both the primary
and secondary phases of the injury response by
utilizing laboratory models that best mimic the
forces/stresses and deformations/strains that occur
23
10 s-1 rate 30 s-1 rate
during a traumatic insult. The response (whether
cellular or whole organism) can better represent
the clinical setting and therefore potential treat-
ments can be evaluated in a more relevant setting.
Future directions
Determination of tolerance criteria for traumatic
CNS injury will likely require a multilevel approach
that incorporates both existing data and new
knowledge from animal and cellular studies with
more refined computer modeling. Computer mode-
ling in the form of FEA can provide estimates of
the mechanical response of tissue to a large range
of traumatic insult parameters, allowing parametric
analysis. These models need to contain anatomical
detail (for both human and animals) and corre-
sponding mechanical property data to maintain the
highest possible fidelity. In addition, they should be
able to simulate large, high rate deformations for
both impact and inertial insult conditions. These
estimated strain and stress patterns should be ver-
ified with in situ measurements when possible. This
24
MECHANICAL INSULT
MEMBRANE STRAIN
DEPOLARIZATION
abnormal cell
signaling energy
deficits
abnormal gene
expression
PERSISTENT DYSFUNCTION OR DEATH
Fig. 8. Simplified schematic of injury cascades initiated by mechanical loading. Mechanical injury may directly initiate downstream
pathophysiological events, but the cause-and-effect relationship has not been thoroughly explored. Plasma membrane damage is
hypothesized to trigger cell death or dysfunction through the inability to regulate ion flux.
represents an experimental challenge and is worthy
of consideration with new advances in nano- and
micro-fabrication techniques, which permit elec-
tromechanical sensors to be instrumented. Animal
models provide an opportunity to study the acute
phase of injury and therefore can be correlated with
estimated strain patterns in order to improve our
current understanding of mechanotransduction. In
addition, parallel long-term studies of delayed cell
death and functional outcome can provide correl-
ative data to acute responses. Furthermore, the cell
response can be studied under very controlled con-
ditions, and in vitro models of traumatic injury can
be used to isolate elements of the mechanical re-
sponse and refine our understanding of cellular
tolerances. Altogether, these data (with known
temporal responses) can be applied to human mod-
els of traumatic injury (with unknown temporal
responses) and tolerance criteria for humans ex-
tracted and predicted for specific scenarios.
Conclusion
Given the tremendous consequences that TBI and
SCI have on society, it is important to better un-
derstand the biomechanical circumstances as they
relate to the physiological and clinical implica-
tions. Biomechanics can play a role in improving
preventative measures such as safety design in
automobiles and sports equipment, as well as
MEMBRANE
PERMEABILITY
NON-SPECIFIC ION FLUX
(e.g., Ca2+, Na+, K+)
loss of
structural proteolytic
integrity degradation
highway and road safety by determining loading
thresholds to the soft tissue of the brain and spinal
cord. In addition to preventative strategies, bio-
mechanics plays an important role in experimental
modeling which, in turn, is vital to the develop-
ment and application of mechanistically inspired
pharmaceutical agents. By applying consistent and
clinically relevant mechanical parameters (e.g.,
shear strain applied at high rates) to isolated
neural cells or animal tissue, the response to me-
chanical disturbances can be assessed. The strain
response is dependent on the tissue heterogeneity,
namely the region-specific material properties and
tissue orientation, therefore making elucidation of
the cellular-level response to mechanical-trauma
complex. The correlation of the injury response
with strain enables detailed cellular tolerances that
can be used to predict human injury criteria using
FEA. In addition to cellular-level investigations,
biomechanical models can be utilized at the animal
level to achieve preclinical testing settings. Taken
together, multilevel investigations can be used to
eventually decrease the incidence of traumatic
CNS injury and improve clinical outcomes.
Acknowledgements
We acknowledge Liying Zhang and King Yang
from Wayne State University for the FEA com-
puter simulations. Partial funding for the results

presented was provided by NSF (BES-0093830)
and by Cooperative Agreement No. DTNH22-
01-H-07551 from the U.S. Department of Trans-
portation — National Highway Traffic Safety
Administration to the University of Alabama at
Birmingham, Southern Consortium for Injury
Biomechanics.
References
Adekoya, N., Thurman, D.J., White, D.D. and Webb, K.W.
(2002) Surveillance for traumatic brain injury deaths —
United States, 1989–1998. MMWR Surveill. Summ., 51: 1–14.
Anderson, T.E. and Stokes, B.T. (1992) Experimental models
for spinal cord injury research: physical and physiological
considerations. J. Neurotrauma, 9(Suppl 1): S135–S142.
Cargill, R.S. and Thibault, L.E. (1996) Acute alterations in
[Ca2+]i in NG108-15 cells subjected to high strain rate de-
formation and chemical hypoxia: an in vitro model for neural
trauma. J. Neurotrauma, 13: 395–407.
Cullen, D.K. and LaPlaca, M.C. (2006) Neuronal response to
high rate shear deformation depends on heterogeneity of the
local strain field. J. Neurotrauma, 23: 1304–1319.
Dixon, C.E., Clifton, G.L., Lighthall, J.W., Yaghmai, A.A. and
Hayes, R.L. (1991) A controlled cortical impact model of
traumatic brain injury in the rat. J. Neurosci. Methods, 39:
253–262.
Dixon, C.E., Lyeth, B.G., Povlishock, J.T., Findling, R.L.,
Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L.
(1987) A fluid percussion model of experimental brain injury
in the rat. J. Neurosurg., 67: 110–119.
Dolan, E.J. and Tator, C.H. (1979) A new method for testing
the force of clips for aneurysms or experimental spinal cord
compression. J. Neurosurg., 51: 229–233.
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-in-
duced injury of cells in culture: characterization of the model
using astrocytes. J. Neurotrauma, 12: 325–339.
Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechano-
poration induced by diffuse traumatic brain injury: an irre-
versible or reversible response to injury? J. Neurosci., 26:
3130–3140.
Gadd, C.W. (1966) Use of a weighted impulse criteria for es-
timating injury hazard. In: 10th Stapp Car Crash Conference
1966, pp. 164–174.
Galbraith, J. and Thibault, L.E. (1993) Mechanically induced
depolarizations in the squid giant axon. J. Biomech. Eng.,
115: 13–22.
Geddes, D.M. and Cargill II., R.S. (2001) An in vitro model of
neural trauma: device characterization and calcium response
to mechanical stretch. J. Biomech. Eng., 123: 247–255.
Geddes, D.M., Cargill II., R.S. and LaPlaca, M.C. (2003) Me-
chanical stretch to neurons results in a strain rate and
25
magnitude-dependent increase in plasma membrane perme-
ability. J. Neurotrauma, 20: 1039–1049.
Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F. (2006)
Mechanisms and consequences of neuronal stretch injury in
vitro differ with the model of trauma. J. Neurotrauma, 23:
193–204.
Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I.,
Thompson, C.J. and Marcincin, R.P. (1982) Diffuse axonal
injury and traumatic coma in the primate. Ann. Neurol., 12:
564–574.
Hardy, W.N., Foster, C., Mason, M., Yang, K.H., King, A.I.
and Tashman, S. (2003) Investigation of head injury mech-
anisms using neutral density technology and high-speed
biplanar X-ray. Stapp Car Crash J., 45: 337–368.
Holbourn, A.H. (1943) Mechanics of head injuries. Lancet, 2:
438–441.
King, A.I., Yang, K.H., Zhang, L., Hardy, W. and Viano, D.
(2003) Is head injury caused by linear or angular acceleration?
In: 2003 IRCOBI Conference. Lisbon, Portugal, pp. 1–12.
LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II.,
R.S. (2005) High rate shear strain of three-dimensional neu-
ral cell cultures: a new in vitro traumatic brain injury model.
J. Biomech., 38: 1093–1105.
LaPlaca, M.C., Lee, V.M. and Thibault, L.E. (1997) An in vitro
model of traumatic neuronal injury: loading rate-dependent
changes in acute cytosolic calcium and lactate dehydrogenase
release. J. Neurotrauma, 14: 355–368.
Lissner, H.R., Lebow, M. and Evans, F.G. (1960) Experimental
studies on the relation between acceleration and intracranial
pressure changes in man. Surg. Gynecol. Obstet., 111:
329–338.
Lusardi, T.A., Rangan, J., Sun, D., Smith, D.H. and Meaney,
D.F. (2004) A device to study the initiation and propagation
of calcium transients in cultured neurons after mechanical
stretch. Ann. Biomed. Eng., 32: 1546–1558.
Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990)
Physical model simulations of brain injury in the primate.
J. Biomech., 23: 823–836.
Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J.,
Kita, H. and Demetriadou, K. (1994) A new model of diffuse
brain injury in rats. Part I: pathophysiology and biomechan-
ics. J. Neurosurg., 80: 291–300.
Maxwell, W.L. (1996) Histopathological changes at central
nodes of Ranvier after stretch-injury. Microsc. Res. Tech.,
34: 522–535.
McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak,
S., Soares, H. and Faden, A.L. (1989) Traumatic brain injury
in the rat: characterization of a lateral fluid percussion model.
Neuroscience, 28: 233–244.
Morrison III., B., Meaney, D.F. and McIntosh, T.K. (1998a)
Mechanical characterization of an in vitro device designed to
quantitatively injure living brain tissue. Ann. Biomed. Eng.,
26: 381–390.
Morrison III., B., Saatman, K.E., Meaney, D.F. and McIntosh,
T.K. (1998b) In vitro central nervous system models of me-
chanically induced trauma: a review. J. Neurotrauma, 15:
911–928.
26
Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock,
J.T. (1994) Traumatically induced altered membrane perme-
ability: its relationship to traumatically induced reactive ax-
onal change. J. Neurotrauma, 11: 507–522.
Pfister, B.J., Weihs, T.P., Betenbaugh, M. and Bao, G. (2003)
An in vitro uniaxial stretch model for axonal injury. Ann.
Biomed. Eng., 31: 589–598.
Prange, M.T. and Margulies, S.S. (2002) Regional, directional,
and age-depedent properties of the brain undergoing large
deformation. J. Biomed. Eng., 124: 244–252.
Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute
spinal cord compression in a new acute cord injury model in
the rat. Surg. Neurol., 10: 38–43.
Sahay, K.B., Mehrotra, R., Sachdeva, U. and Banerji, A.K.
(1992) Elastomechanical characterization of brain tissues. J.
Biomech., 25: 319–326.
Scheff, S.W., Rabchevsky, A.G., Fugaccia, I., Main, J.A. and
Lumpp Jr., J.E. (2003) Experimental modeling of spinal cord
injury: characterization of a force-defined injury device. J.
Neurotrauma, 20: 179–193.
Shi, R. and Borgens, R.B. (2000) Anatomical repair of nerve
membranes in crushed mammalian spinal cord with polyeth-
ylene glycol. J. Neurocytol., 29: 633–643.
Shi, R. and Pryor, J.D. (2002) Pathological changes of isolated
spinal cord axons in response to mechanical stretch. Neuro-
science, 110: 765–777.
Shi, R. and Whitebone, J. (2006) Conduction deficits and
membrane disruption of spinal cord axons as a function of
magnitude and rate of strain. J. Neurophysiol., 95:
3384–3390.
Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennar-
elli, T.A. and Meaney, D.F. (1997) Characterization of dif-
fuse axonal pathology and selective hippocampal damage
following inertial brain trauma in the pig. J. Neuropathol.
Exp. Neurol., 56: 822–834.
Smith, D.H., Wolf, J.A., Lusardi, T.A., Lee, V.M. and Meaney,
D.F. (1999) High tolerance and delayed elastic response of
cultured axons to dynamic stretch injury. J. Neurosci., 19:
4263–4269.
Takhounts, E.G., Eppinger, R.H., Campbell, J.Q., Tannous,
R.E. and Power, E.D. (2003) On the development of the SI-
Mon finite element head model. Stapp Car Crash J., 47:
107–133.
Thibault, L.E., Meaney, D.F., Anderson, B.J. and Marmarou,
A. (1992) Biomechanical aspects of a fluid percussion model
of brain injury. J. Neurotrauma, 9: 311–322.
Thurman, D. and Guerrero, J. (1999) Trends in hospitalization
associated with traumatic brain injury. JAMA, 282: 954–957.
Versace, J. (1971) A review of the severity index. In: Proceed-
ings of the 15th Stapp Car Crash Conference, Society of
Automotive Engineers, New York, pp. 771–796.
Young, W. (2002) Spinal cord contusion models. Prog. Brain
Res., 137: 231–255.
Zhang, L., Yang, K.H. and King, A.I. (2004) A proposed injury
threshold for mild traumatic brain injury. J. Biomech. Eng.,
126: 226–236.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 3

Linking impact to cellular and molecular sequelae of

CNS injury: Modeling in vivo complexity with in
vitro simplicity

Jennifer M. Spaethling, Donna M. Geddes-Klein, William J. Miller, Catherine R. von Reyn,

Pallab Singh, Mahlet Mesfin, Steven J. Bernstein and David F. Meaney

Departments of Bioengineering and Neurosurgery, University of Pennsylvania, 3320 Smith Walk, Philadelphia,
PA, 19104-6392, USA

Abstract: Traumatic brain injury (TBI) represents one of most common disorders to the central nervous
system (CNS). Despite significant efforts, though, an effective clinical treatment for TBI is not yet available.
The complexity of human TBI is modeled with a broad group of experimental models, with each model
matching some aspect of the human condition. In the past 15 years, these in vivo models were complemented
with a group of in vitro models, with these in vitro models allowing investigators to more precisely identify
the mechanism(s) of TBI, the different intracellular events that occur in acute period following injury, and
the possible treatment of this injury in vitro. In this paper, we review the available in vitro models to study
TBI, discuss their biomechanical basis for human TBI, and review the findings from these in vitro models.
Finally, we synthesize the current knowledge and point out possible future directions for this group of
models, especially in the effort toward developing new therapies for the traumatically brain injured patient.

Keywords: traumatic brain injury; biomechanics; in vitro models

Introduction Sattin, 2005). In the elderly, TBI is only eclipsed by

The enormous consequences of traumatic brain in-
jury (TBI) continue in society, despite the rapid ad-
vance of technology to reduce the severity of injuries
and new approaches in trauma patient care. Even
with these changes, traumatic brain injuries remain
the leading cause of death in people less than 45
years old (Thurman et al., 1999). The incidence rate
is equally startling — the number of people hospi-
talized each year for traumatic brain injuries exceed
those diagnosed with multiple sclerosis, breast can-
cer, and spinal cord injury combined (Langlois and
Corresponding author. Tel.: +1 215 573 3155; Fax: +1 215
573 2071; E-mail: dmeaney@seas.upenn.edu
cardiovascular disease and cancer as a major cause
of death.
Clinically, brain injuries are categorized broadly
as focal or diffuse (Gennarelli et al., 1982a). Focal
injuries are readily observable lesions that appear
on standard CT or MRI scans, and include injuries
to the vasculature (epidural, subdural hematoma),
the microvasculature (cerebral contusions), and vis-
ible tears in the brain parenchyma (intracerebral
hemorrhage). Diffuse injuries are diagnosed when
no visible lesions are present using conventional
imaging, yet the patient has clear neurological
impairment. A major substrate of diffuse brain
injuries is diffuse axonal injury (DAI), but this cat-
egory of injury also includes diffuse brain swelling

DOI: 10.1016/S0079-6123(06)61003-0 27
28
and brain edema (Graham et al., 1995; Povlishock
and Katz, 2005).
Generations of investigators directed their efforts
toward understanding the mechanisms of TBI.
From these efforts, researchers and clinicians reco-
gnize that TBI for a given patient is not captured
well with only a ‘diffuse’ or ‘focal’ description.
Rather, many different injuries are grouped with
the broad clinical subtypes and leads to a tremen-
dous diversity of injuries in the patient population.
Injury mechanisms share a similar diversity — the
mechanisms of injury for one specific traumatic in-
jury may not apply universally to other injury
types, and vice versa. As a field, we often use a
reductionist approach to understand individual
components of clinical TBI.
Models to study the in vivo complexity of TBI
exist in different species, along different length
scales that span from the single cell to the whole
organism. The purpose of this review is to provide
a current synthesis of the findings from models in-
tended to study TBI in vitro, with an emphasis on
discussing how these models relate to experimental
efforts using animal models of TBI. In turn, we will
summarize the findings from these models and
point out new areas of opportunity where these
models may prove invaluable in developing new
treatments for TBI.
Biomechanical mechanisms that cause TBI in vivo
A discussion of the linkages between in vitro and in
vivo experimental TBI studies would be incomplete
without a brief review of the underlying biome-
chanical mechanisms that cause TBI. One major
aim of in vitro models is to replicate the underlying
physical forces that the tissue experiences during
traumatic injury. However, the forces experienced
by the tissue during even a single head impact can
vary greatly with impact direction, force, and the
impacting surface. A person designing an in vitro
model must ask — which of these forces to the
tissue are important for causing injury? In addi-
tion, one asks — how does one control the inherent
variability of these forces?
Pressures developed within the brain during in-
jury were considered one of the primary
mechanisms for causing immediate impairment
after brain injury, with the first modern studies
dating back over six decades (Denny-Brown and
Russel, 1941; Denny-Brown, 1945). These studies
were soon followed by a number of in vivo experi-
mental models that simulated these pressures to
cause a concussive insult in animals (Stalhammar
and Olsson, 1975; Sullivan et al., 1976). The intra-
cranial pressure patterns that are generated in the
human brain during impact is known (Nahum et
al., 1977) and is predicted accurately with compu-
tational models (e.g., Zhang et al., 2004). There-
fore, one can draw a clear relationship between the
pressures generated in the in vivo models and
within the brain during TBI. The effect of transient
pressure changes on tissue function, though, is less
clear. Of the possible mechanisms, there is only
consensus that pressure cause damages is when the
pressure drops below the cavitation threshold for
tissue, thereby causing immediate (primary) tissue
damage (Nusholtz et al., 1995). Much less agree-
ment exists on whether high positive pressures will
cause impairment, or an additional physical injury
mechanism is needed.
Tissue deformation during impact is the
second major physical mechanism explaining
the primary patterns of injury in the brain.
Although the brain shape under pressure load-
ing largely does not change, the compliant
properties of the tissue make the brain susceptible
to shearing deformations during impact. Until
recently, we knew little about the properties of
the brain during situations causing injury. Recent
information suggests that the brain softens as it
deforms (Prange and Margulies, 2002; Coats
and Margulies, 2006), an intriguing material
property that may lead to unexpected patterns
of tissue damage. Although there is wide recog-
nition that brain deformation/stress is a major
mechanism leading to cellular damage, exactly
how the stresses are transferred from the tissue
to the cellular structures within the material are
not well known. Estimates of the tissue strain
needed to cause injury are now available (Shreiber
et al., 1997; Bain and Meaney, 2000; Zhu et al.,
2006), but there remains considerable discussion
about how these strains should be modeled in
either in vivo or in vitro models, and if these

models can truly capture the complete simulation
of injury.
Reproducing injury mechanisms with in vivo TBI
models
Much like other diseases and disorders, an inves-
tigator faces several challenges when modeling hu-
man TBI in an animal model. The clinical mix of
focal and diffuse injuries conflicts with the need to
develop a consistent, reliable, and repeatable
model in the laboratory. Inevitably, the investiga-
tor chooses which components of clinical TBI to
study in the laboratory. Here, we briefly categorize
the different experimental models and describe
their uses. We use this description as background
material for the in vitro models discussed in the
next section.
For nonpenetrating injuries, an early area of
study was to model the effects of pressure loading
on the brain, as the concussive effects of pressure
were reported in the early literature (Denny-
Brown, 1945). The effect of pressure led to the so
called ‘percussion concussion’ models (Gennarelli,
1994) that used a pulse of either air or fluid on the
exposed cortex to cause neurological impairment.
The most common model in current use is the fluid
percussion model, available in many species and
used in different configurations (Stalhammar and
Olsson, 1975; Sullivan et al., 1976; Dixon et al.,
1988; McIntosh et al., 1989a). Early work with
fluid percussion showed how the pressure applied
to the cortex was dissipated throughout the brain
and the spinal canal (Stalhammar and Olsson,
1975). Later work showed that this pressure also
caused a movement of intracranial tissue, leading
to pressure and strain throughout the brain
(Thibault et al., 1992). Possibly due to the com-
plex mechanics of this model, different variations
of the percussion model can lead to forms of injury
that resemble components of human TBI.
A more recent technique to study brain injury in
vivo is to directly deform the brain with a solid
indentor, often referred to as the cortical impact
model (Lighthall, 1988). The advantages of the
model include a highly quantified impact condi-
tion, an ability to easily scale the impact condition
29
across species (Smith et al., 1995), and modifying
the model to injure different areas of the cortex.
Recent modifications of the model used a direct
skull impact, leading to a closed head impact
model that could be considered closer to the clin-
ical condition. With the precise control of the
model, the cortical impact model is frequently the
choice when studying TBI in transgenic animals.
From an injury mechanism standpoint, the cortical
model is qualitatively similar to the fluid percus-
sion technique — apply a mechanical input locally
to the cortex and study the progressive pattern of
injury throughout the brain.
A final method for TBI models is using acceler-
ation to injure the brain, reproducing a common
mechanical loading that causes TBI in humans.
Studies show the necessary acceleration to cause
injury increases quickly with decreasing brain mass
(Ommaya et al., 1967); therefore, the acceleration
method is most readily used in large animal species
that include the nonhuman primate and the min-
iature pig (Gennarelli et al., 1982b; Meaney et al.,
1995). Although models using acceleration input in
small animals (e.g., rat, ferret) appear in the liter-
ature (Marmarou et al., 1994; Xiao-Sheng et al.,
2000; Gutierrez et al., 2001), it is difficult to sep-
arate the effects of the acceleration from the pos-
sible shape changes in the skull caused by the
relatively large force needed to create these accel-
erations. A combination of the acceleration input
and the higher proportion of white matter in the
gyrencephalic brains of the nonhuman primate and
pig makes these models ideal for studying injury to
the white matter.
Broad categorization of in vivo models — what do
they reproduce?
The different in vivo models described above are
generally designed to simulate the mechanisms of
injury that occur in humans. Results from in vivo
models are sometimes difficult to interpret, though,
as measurements of intracellular signaling, single
cell function, and the role of different cell types are
difficult. Moreover, the effect of mechanical and
hypoxic injury are not easily separable with in vivo
models, but are easily divided with in vitro
30

approaches. The use of in vitro models to simulate cuts can lead to gene expression changes over time
TBI began several decades ago and has signifi- (Raghupathi et al., 1998). This remains probably
cantly evolved in complexity (Morrison et al., the most precise model to study distal and proximal
1998b). In this section, we describe the in vitro effects of transection. The model best corresponds
models that are currently in use to study TBI, and to the physical disruption that can occur to some
provide an in vivo correlate for these in vitro neuronal processes following injury (Maxwell et al.,
methods. 1997), especially at the severe levels. The most im-
portant utility of the model, though, may be the
ability to distinguish the role of local and remote
Scratch model/laceration signaling on cell survival/death following injury
(Singleton et al., 2002).
The most direct method for mechanical injury is to
tear or lacerate cultures with a stylus or punch. The
Weight drop/compression
direct mechanical disruption of cultures is one of
the earliest methods to study the progression of in-
An additional and straightforward technique to
jury, beginning with tissue chunks (Epstein, 1971)
use on cultures is to mechanically compress the
but soon moving to mixed cultures of neurons and
cultures with a weight, akin to the weight drop
glia (Tecoma et al., 1989; Regan and Choi, 1994;
method developed initially by Allen (1911) to
Regan and Panter, 1995). The scratch method is
study spinal cord injury in vivo. Indeed, one of the
well suited for high throughput drug screening, and
first in vitro models for central nervous system
has been used by Faden and colleagues to examine
(CNS) injury used this technique of spinal cord
both inhibitors and antisense oligonucleotide treat-
cultures (Balentine et al., 1988). The technique is
ment (Faden et al., 1997; Mukhin et al., 1997,
well suited to organotypic cultures that have a de-
1998). This technology also scales easily to slice
fined thickness and more realistic 3D architecture,
culture tissue (Sieg et al., 1999), and has suggested
and can be used to study the effects of both me-
factors that influence the vulnerability or surviv-
chanical injury and a superimposed hypoxic injury
ability of neurons in different regions of the brain.
(Adamchik et al., 2000). The order of the injuries
The direct mechanical injury remains in use, now
can be changed, so that the mechanical injury can
with an emphasis that includes the release of mole-
be considered the secondary injury, or vice versa.
cules that could be considered biomarkers of the
The technique to compress the tissue construct can
injury (Yang et al., 2006) and the potential factors
also change; a dropped weight can be replaced by a
that can regulate the migratory activity of glial cells
rolling stainless steel bar, or a composite foam in-
to the injury site (Barral-Moran et al., 2003). The
dentor over a region of the culture (Adamchik
technique best correlates to the primary tissue tear-
et al., 2000). Recently, this type of model showed
ing that can occur in different brain regions fol-
the potential spreading depression that can occur
lowing severe head injury, a penetrating ballistic
from mechanical injury (Church and Andrew,
injury, or the local tissue damage that occurs with
2005). One primary disadvantage of this tech-
depressed skull fractures.
nique, though, is drawing the direct in vivo cor-
An interesting variation of the tearing model is
relate for this method. Crush injuries to the brain
using a laser to focally disrupt or transect the proc-
parenchyma are rare, and are complicated by
esses of neurons in culture (Gross et al., 1983). The
overlying skull fracture.
relative position of the laser cut can be controlled
and can be close to or distant from the neuronal
soma, yielding distinct differences in cell fate Cell/substrate stretch model
(Lucas, 1987). The distance from the lesion to the
soma also changes the subsequent ultrastructure Based on the number of completed studies, the
response, as well as the electrophysiological prop- most commonly used in vitro technique to study
erties of the cell (Lucas et al., 1985). Moreover, the the consequences of TBI in vitro is the cell stretch

(or substrate deformation) model. These models
replicate the magnitude and rate of tissue defor-
mation that occurs in vivo during injury (Meaney
et al., 1995), but often simplify the multiple oscil-
lations of tissue deformation into a single, transient
stretch insult. One early feature of the model was
using a design where the cultured cells, plated to an
elastic substrate, were simultaneously deformed in
two directions. Although first used in astrocytes,
this technique was rapidly tested in many different
cell types of the CNS including neuron and neuron-
like cell lines, endothelial cells, and, more recently,
microglia (Cargill and Thibault, 1996; Ellis et al.,
1995; McKinney et al., 1996; Rzigalinski et al.,
1997). A variation of the initial stretch models is
now available, where the cultures are stretched
only in one direction (Lusardi et al., 2004). The
stretch can be confined to just cultured axons to
model diffuse axonal injury (Smith et al., 1999).
Methods to stretch dissociated cultures transfer
easily to studying the effects of stretch on tissue
slice cultures (Morrison et al., 1998a), where the
transfer between the substrate stretch and the re-
sulting tissue stretch are defined (Cater et al., 2006).
More recent work shows the cell stretch technique
can scale to study the effect of more complex, 3D
strain patterns on the morphology and viability of
cells in tissue constructs (LaPlaca et al., 2005).
With the increasing number of controlled mechan-
ical manipulations available on cultures and tissue
constructs, it is now possible to identify if CNS
cells are uniquely vulnerable to a certain type of
mechanical deformation (Geddes-Klein et al.,
2006a), if the effects of injury are cumulative
(Slemmer et al., 2002), or if the mechanism(s) of
injury will change if the cells experience stretch,
compression, or fluid shear deformation (LaPlaca
et al., 1997).
Mechanoactivation — the first therapeutic target
The increasing diversity of models used to study
TBI in vitro leads one to a common question —
what processes do these forces activate rapidly fol-
lowing injury, and how do these ‘mechanoactivated’
signals lead to the pathological changes observed
hours to days following the initial injury?
31
Identifying these early changes in CNS cells pro-
vides initial therapeutic targets. Knowing how the
mechanoactivated targets lead to subsequent intra-
cellular events and/or consequences will naturally
generate new therapeutic targets, is perhaps as
important. In this section, we review the current
knowledge on the early changes that occur follow-
ing mechanical injury and their relative utility as a
target for reversing the effects of the injury.
NMDAR are the most commonly studied
mechanoactivated target
Alterations in ionic homeostasis have been ob-
served across nearly all in vitro and in vivo models
of TBI and therefore provide a natural starting
point for identifying mechanoactivated receptors.
Due to the role of glutamate receptors in both
physiologic (learning and memory) and pathologic
(stroke and epilepsy) conditions, N-methyl
D-aspartate receptors (NMDAR) are among the
most widely studied receptors responsible for the
cytosolic calcium overload (Gagliardi, 2000;
Arundine and Tymianski, 2004). Multiple models
of TBI have shown that the bulk of the loss of ionic
homeostasis can be attributed to activation of
NMDARs. The connectivity of NMDARs to the
actin cytoskeleton may provide a potential mech-
anism as to how these receptors are activated by
stretch (Geddes-Klein et al., 2006b). One unique
feature of mechanical injury to neurons is the
mechanically induced reduction in the normal mag-
nesium block of the NMDAR (Zhang et al., 1996).
Altering the magnesium block of NMDARs
changes the relative influence of NMDAR even
under normal neurotransmission signaling and
provides a therapeutic target that has been found
to be somewhat effective in vivo (McIntosh et al.,
1989a, b). It is worth noting that mechanically in-
duced changes in the NMDAR properties are not
observed across all models of TBI, including mixed
cultures (Faden et al., 2001), or when the cultures
are subject to a sublethal stretch (Arundine et al.,
2003, 2004). Additionally, the mechanically induced
activation or modulation of NMDARs leads to
the propagation of intracellular calcium into adja-
cent, uninjured cells (Lusardi et al., 2004). While
32

NMDAR antagonists have not been effective in the 2005), these mechanically initiated changes may
clinic, recent data suggests that targeting the loca- have long lasting effects on synaptic physiology
tion of NMDAR (synaptic vs. extrasynaptic) may following injury.
differentiate between the protective and excitotoxic
processes of NMDARs (DeRidder et al., 2006).
Voltage-gated sodium channels can contribute to
regional changes in neuronal axons
Are other glutamate receptors involved?
Voltage-gated sodium channels have been shown
to indirectly contribute to shifts in intracellular
The roles of other glutamate receptors are also be-
calcium following mechanical injury. The relative
ginning to generate serious attention in treating
role of the voltage-gated sodium channel only
traumatic brain injuries. a-Amino-3-hydroxy-5-
appears when the axonal segment of the neuron is
methyl-4-isoxazolepropionic acid receptors (AM-
deformed, potentially because the NMDAR and
PAR) undergo a unique transformation, losing their
AMPAR mediated changes dominate for dendritic
rapid desensitization property following mechanical
processes and the neuronal soma. Moreover, it is
injury (Goforth et al., 1999). The loss in desensiti-
worth noting, that little in vitro work exists for
zation is persistent for at least 24 h following injury,
myelinated axons due to the difficulties of creating
and can be avoided if the NMDARs are inhibited
myelinated cultures in vitro, even though it has
prior to mechanical injury (Goforth et al., 2004).
been shown that myelinated axons respond differ-
The impact of this loss in desensitization of the
ently to stretch than myelinated axon (Reeves
AMPARs on neuronal function is not yet well de-
et al., 2005). Studies show the sodium channel ac-
scribed, though. As AMPARs are centrally involved
tivation following mechanical injury leads to a
in fundamental processes such as synaptic plasticity
dramatic and sudden rise in axoplasmic calcium,
and metaplasticity, even a transient alteration in the
but not from calcium entering directly through the
normal desensitizing properties of the AMPAR can
sodium channel (Wolf et al., 2001). Rather, the
lead to immediate and long term changes in neu-
source of increased axoplasmic calcium is through
ronal network behavior.
voltage-gated calcium channels and through re-
Although there is evidence that inhibition of me-
versal of the sodium calcium exchanger. These
tabotropic glutamate receptors (mGluRs) will pro-
changes in the sodium channel occur simultane-
vide some protection against the effects of
ously with a larger, but reversible change in the
mechanical injury, it is not known if these changes
morphology of axons subjected to mechanical in-
are directly caused by mechanical modulation of
jury (Smith et al., 1999). Over time, these sodium
the mGluRs, or if these protective effects are from
channels are the targets of proteolysis, which can
simply inhibiting the activation of these receptors
lead to a sustained change in neuronal activity
from enhanced glutamate levels following injury
(Iwata et al., 2004).
(Faden et al., 2001; Movsesyan et al., 2001;
Movsesyan and Faden, 2006). One interesting
recent theory is the role of potential physical cou- Changes in mechanical permeability
pling between the group I mGluR receptors, IP3,
and phospholipase C (PLC) (Floyd et al., 2004). One final consequence of mechanical force is an
The role of different mGluR subtypes also reveals a immediate, but transient, change in the plasma
complex interplay between mGluRs after injury, membrane permeability termed ‘mechanoporation’.
where the inhibition of one subtype (type III) can In a series of studies, both the relative size and du-
eliminate the protection offered by an antagonist ration of these transient pores in the membrane were
directed against a second subtype (type II) estimated following neuronal stretch (Geddes and
(Movsesyan and Faden, 2006). Given the role of Cargill, 2001; Geddes et al., 2003). These effects are
mGluRs in either enhancing or modulating synap- also measured following traumatic injury in vivo,
tic efficacy (Gubellini et al., 2004; Simonyi et al., although these changes appear to occur over a

longer time period than in vitro preparations (Pettus
et al., 1994; Stone et al., 2004; Farkas et al., 2006). It
is not clear, though, how these changes in perme-
ability can change for different neuronal regions
(axon, dendrite, soma), or if this process is limited to
one or more cell types. The effect of these transient
pores can be minimized by resealing the membrane
with surfactants (Serbest et al., 2006). Even though
these changes may be brief, recent evidence suggests
this mechanism may be capable of specifically stim-
ulating different components of the mitogen-
activated protein kinase (MAPK) cascade, and can
therefore play a role in the ensuing cell death that
occurs after injury (Serbest et al., 2006). However,
the role of permeability increases in viability seems
to be inconsistent across models, as some studies
reveal no change in membrane resistance or perme-
ability to a fluorescent dye (Tavalin et al., 1995;
Zhang et al., 1996; Smith et al., 1999).
Mechanoactivation cascade — the next target
Even though the early events of mechanical injury
in vitro are becoming clear, these efforts only begin
to establish the complex cascade of events that will
influence neuronal or glial survival after TBI. The
initial mechanoactivated receptors and ensuing loss
of ionic homeostasis activates a cascade of cellular
processes that ultimately result in loss of function
or cell death. Understanding these cascades will be
essential to developing novel therapeutic targets
that are admissible in a clinically relevant thera-
peutic window. We term these next events as the
‘mechanoactivation cascade’. These cascades can
occur rapidly following the initial mechanoactiva-
tion step, or can progress more slowly over time. In
this section, we group these cascades according to
the cell types studied to date in the CNS.
Neuronal cascades
The changes in neurons following mechanical in-
jury include alterations in the electrophysiological
properties, organelle function, receptor profile, and
intracellular calcium buffering. Some of these
changes lead to neuronal vulnerability, while oth-
ers will transiently impair function. Early after
33
injury, there is a delayed, but persistent depolari-
zation that is linked to the alteration in the elect-
rogenic sodium/potassium exchanger (Tavalin
et al., 1995, 1997). The depolarization is triggered
with a mild glutamate stimulation, and can be
attenuated by restoring intracellular ATP levels.
Findings from several groups show that mechan-
ical injury leads to an enhanced response to gluta-
mate up to 24 h following injury (Weber et al.,
1999; Arundine et al., 2004; Geddes-Klein et al.,
2006a), in turn leading to an enhanced vulnerabil-
ity to glutamate excitotoxicity (Arundine et al.,
2003). Although most reports focus on the role of
the NMDAR as the underlying factor for this
enhanced glutamate response, this is not the only
ligand-gated receptor that changes following in-
jury. The loss in desensitization of the AMPAR
appears in parallel with an enhancement of AMPA
currents after injury, as does the enhancement of
g-aminobutyric acid (GABA) currents. Both the
AMPA and GABA current changes are linked to
the NMDAR, as inhibiting the NMDA prior to
injury eliminates the enhancement of both currents
(Goforth et al., 2004; Kao et al., 2004). These
changes may also contribute to the changes in
neural network activity that appears following me-
chanical injury (Prado et al., 2005), although these
changes remain to be explored in detail. These
changes are likely among the initiating factors
that contribute to failure to induce LTP in in
vivo models and ultimately result in learning and
memory deficits.
Neuronal mitochondria are also affected, with a
reduction in the mitochondria membrane potential
that can persist and which is dependent on the
presence of surrounding glia (Ahmed et al., 2000,
2002). A series of reports show that the normal
regulation of intracellular calcium stores are altered
soon after mechanical injury, and this change in
capacitive calcium influx can alter the homeostatic
mechanisms for calcium induced calcium release
(Weber et al., 2001; Chen et al., 2004; Weber, 2004).
Cysteine protease activation is dependent on the
severity of the injury, and can be affected by
NMDAR activation inhibition/activation (Pike
et al., 2000; DeRidder et al., 2006). Inhibiting the
activation of one cysteine protease, caspase-3, will
have a short therapeutic window but also reveals
34
some crosstalk among the classic apoptotic and
necrotic pathways (Knoblach et al., 2004). Gene
expression changes correlate with the mechanical
input applied to the culture (Morrison et al., 2000);
some of these changes reflect changes observed in
single neurons following TBI in vivo (O’Dell et al.,
2000). The role of MAPKs in vitro are similar to
the role observed in vivo — extracellular signal
regulated kinase (ERK) inhibition may be a
potential target (Mori et al., 2002), but the timing
of the therapy needs to be better defined (Dash
et al., 2002).
Astrocytic cascades
The role of astrocyte-based changes can influence
not only glial reactivity in the injured brain, but
also neuronal activity through the coupling of as-
trocytes and neurons at the synapse. Intracellular
calcium signaling in astrocytes after injury is
coupled to the presence of extracellular calcium,
and the absence of extracellular calcium will in-
fluence astrocytic death (Rzigalinski et al., 1997).
The complex intracellular regulation of calcium
also changes, as stimulation of mGluRs is less
coupled to IP3 mediated calcium release and is
regulated, in part, by PLC (Rzigalinski et al., 1998;
Floyd et al., 2001). Changes in intracellular
sodium occur in parallel with these calcium
changes and are linked with glutamate uptake in
astrocytes, but may also be loosely coupled to
intracellular calcium changes through changes in
the sodium calcium exchanger (Floyd et al., 2005).
These changes in ion homeostasis lead to altera-
tions in MAPK signaling (Neary et al., 2003),
some of which can be linked to the eventual glial
reactivity phase that forms an important part of
the neuropathology found in human TBI.
Perhaps more than studies using cultured neu-
rons, there are now several reports focusing on the
‘crossover’ effect of mechanically injured astrocytes
on other CNS cell types. For example, injured as-
trocytes generate isoprostanes that can influence
vasoconstriction and reduce blood flow after
trauma (Hoffman et al., 2000). Generation of re-
active oxygen species from mechanically injured
astrocytes can affect not only metabolic changes
within astrocytes, but also mediate damage in
adjacent neurons (Lamb et al., 1997). Changes in
mitochondria membrane potential and ATP levels
occur only transiently in astrocytes after injury, but
the presence of the astrocytes in cultures change the
time course of neuronal mitochondria changes
(Ahmed et al., 2000). The presence of astrocytes
also appears to influence the release of matrix met-
alloproteinases following injury in vitro, a factor
that may significantly affect the regeneration phase
after injury (Wang et al., 2002). Gene expression
changes can be altered in the presence or absence of
glia (Katano et al., 1999), indicating that the cross-
over effect extends to the genomic level.
What therapies have emerged?
A primary advantage of in vitro models is the
ability to quickly test the efficacy of new or existing
compounds in reducing the effects of mechanical
injury. Most work to date, though, has focused on
the mechanisms that contribute to either early
changes in intracellular signaling or cell function.
Surprisingly few studies have moved this work
toward studying therapies that may be used for in
vivo TBI treatment. Early work showed the effec-
tiveness of NMDA antagonists (Regan and Panter,
1995), free radical scavengers (Lamb et al., 1997;
Shah et al., 1997), and mGluR activation or inhi-
bition (Allen et al., 1999; Movsesyan et al., 2001).
More recent work includes the development of new
peptides to reduce cell death after injury, even if the
compound is delivered hours following injury
(Faden et al., 2003, 2004, 2005). These peptides
offer a strategic advantage over receptor antago-
nists by offering more selective inhibition of intra-
cellular signaling triggered by a target receptor.
Customized peptide designs can interfere with crit-
ical regulatory points in the mechanoactivation
cascade, such as the point where NMDAR activa-
tion can lead to the formation of reactive oxygen
species. Using a peptide to inhibit the linkage
between the NMDAR and PSD-95, a prominent
postsynaptic density protein, leads to an interrup-
tion in the ROS cascade and will lead to neuronal
protection (Arundine et al., 2004). Delayed efficacy
is also possible by using subunit specific

antagonists for the NMDAR (DeRidder et al.,
2006), perhaps due to the recent data showing the
dual role for NMDAR in cell survival and
apoptosis (Hardingham. and Bading, 2003).
In vitro models are also useful in determining
more specific biomarkers for both TBI diagnosis
and treatment. A group of released factors have
been found in the media following mechanical in-
jury, with some contributing to the ensuing neu-
ronal death. Injury using a stylus transection
method is inhibited with NMDAR antagonists
and aminosteroids (Regan and Panter, 1995), and
the surrounding media can contain free radicals
that, if applied to uninjured cultures, would cause
apoptotic cell death within hours (Shah et al.,
1997). A physical scratching injury also activates
inflammatory mediated injury in neuronal cultures,
as well as astrocytic migration (Fitch et al., 1999),
MAPK activation, and the release of matrix
metalloproteinases (MMPs) (Wang et al., 2002).
Mechanical stretch injury will cause the synaptic
release of zinc (Cho et al., 2003), an increase in
s100b in the media (Willoughby et al., 2004). It is
interesting to note that at least some of these
molecules are being investigated as potential
biomarkers for in vivo injury (Pineda et al., 2004).
Pointing toward the future
Collectively, this work shows the extensive efforts
completed in understanding how mechanical forces
are transmitted to cells of the CNS, what processes
are activated by these mechanical forces, and how
this information reveals different intervention
points for treating the effects of mechanical injury
in the CNS. Motivated by an understanding of the
physical forces that occur within the tissue during
injury, there are now models to study the effect of
both simple and more complex mechanical loading
to cultures. These models are well controlled, can
be easily extended to understand more realistic
loading conditions, and are even approaching the
very complex question of how these forces are
transferred to cells in 3D tissue. One needs to con-
sider if additional in vitro models are needed, or if
the complex mechanical environment is suitably
modeled with existing technology. Certainly, the
35
current work does not comprehensively address the
initial mechanoactivation process for all cell types
in the brain; notable exceptions are the oligoden-
drocytes, brain endothelial cells, and microglia.
Even within a given cell type such as neurons, there
are differences that appear among neurons from
different brain regions. Moreover, there is little
consensus information on how the culture platform
— nearly pure or mixed cultures, tissue constructs,
or organotypic slice cultures — contribute to the
measured response; information to date show the
potential synergistic interactions that appear bet-
ween different cell types. These areas, as well as
many others, can provide more insight into how
these models will move toward representing the in
vivo environment following injury.
In closing, it is worth noting that the natural
division between in vivo and in vitro models of
injury is slowly disappearing. The culturing tech-
niques are becoming more sophisticated, allowing
one to construct functional tissues with highly pre-
cise and more ‘in vivo like’ architectures. The
methods to injure these cultures is now developed
enough to mechanically load these constructs with
any desired loading profile, mimicking the loading
profiles occurring in vivo following injury. In com-
parison, rapid advances with in vivo imaging tech-
nology now allows one to use fluorescent indicator
dyes in vivo to monitor shifts ion homeostasis, and
technology is already emerging to monitor the
electrical activity in living, awake animals over
time. As a result, technologies and measurement
normally restricted to in vitro systems are now
available for testing in the in vivo animal. The in-
creasing convergence of these different approaches
will likely make comparative approaches in the fu-
ture more common, and will lead to a more rapid
translation of in vitro findings to the in vivo set-
ting. Ultimately, this will translate into the more
rapid testing of therapeutics in vivo that originated
from in vitro systems, and lead to more therapeutic
options for the traumatically brain injured patient.
Acknowledgments
Funds for this work were provided by grants from
the National Institutes of Health (RO1 HD-41699
36
and NS-35712) and the Commonwealth of Penn-
sylvania.
References
Adamchik, Y., Frantseva, M.V., Weisspapir, M., Carlen, P.L.
and Perez Velazquez, J.L. (2000) Methods to induce primary
and secondary traumatic damage in organotypic hippocam-
pal slice cultures. Brain Res. Brain Res. Protoc., 5(2):
153–158.
Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding,
H.A. and Ellis, E.F. (2000) Stretch-induced injury alters mi-
tochondrial membrane potential and cellular ATP in cultured
astrocytes and neurons. J. Neurochem., 74(5): 1951–1960.
Ahmed, S.M., Weber, J.T., Liang, S., Willoughby, K.A., Sit-
terding, H.A., Rzigalinski, B.A. and Ellis, E.F. (2002)
NMDA receptor activation contributes to a portion of the
decreased mitochondrial membrane potential and elevated
intracellular free calcium in strain-injured neurons. J. Neuro-
trauma, 19(12): 1619–1629.
Allen, A. (1911) Surgery of experimental lesion of spinal cord
equivalent to crush injury of fracture dislocation of spinal
column — a preliminary report. JAMA, 57(11): 878–890.
Allen, J.W., Ivanova, S.A., Fan, L., Espey, M.G., Basile, A.S.
and Faden, A.I. (1999) Group II metabotropic glutamate
receptor activation attenuates traumatic neuronal injury and
improves neurological recovery after traumatic brain injury.
J. Pharmacol. Exp. Ther., 290(1): 112–120.
Arundine, M., Aarts, M., Lau, A. and Tymianski, M. (2004)
Vulnerability of central neurons to secondary insults after in
vitro mechanical stretch. J. Neurosci., 24(37): 8106–8123.
Arundine, M., Chopra, G.K., Wrong, A., Lei, S., Aarts, M.M.,
MacDonald, J.F. and Tymianski, M. (2003) Enhanced vul-
nerability to NMDA toxicity in sublethal traumatic neuronal
injury in vitro. J. Neurotrauma, 20(12): 1377–1395.
Arundine, M. and Tymianski, M. (2004) Molecular mecha-
nisms of glutamate-dependent neurodegeneration in ischemia
and traumatic brain injury. Cell Mol. Life Sci., 61(6):
657–668.
Bain, A.C. and Meaney, D.F. (2000) Tissue-level thresholds for
axonal injury in an experimental model of CNS white matter
injury. J. Biomech. Eng., 122: 615–622.
Balentine, J.D., Greene, W.B. and Bornstein, M. (1988) In vitro
spinal cord trauma. Lab. Invest., 58(1): 93–99.
Barral-Moran, M.J., Calaora, V., Vutskits, L., Wang, C.,
Zhang, H., Durbec, P., Rougon, G. and Kiss, J.Z. (2003)
Oligodendrocyte progenitor migration in response to injury
of glial monolayers requires the polysialic neural cell-
adhesion molecule. J. Neurosci. Res., 72(6): 679–690.
Cargill II, R.S. and Thibault, L.E. (1996) Acute alterations in
[Ca2+]i in NG108-15 cells subjected to high strain rate de-
formation and chemical hypoxia: an in vitro model for neural
trauma. J. Neurotrauma, 13(7): 395–407.
Cater, H.L., Sundstrom, L.E. and Morrison III, B. (2006)
Temporal development of hippocampal cell death is
dependent on tissue strain but not strain rate. J. Biomech.,
39(15): 2810–2818.
Chen, T., Willoughby, K.A. and Ellis, E.F. (2004) Group I
metabotropic receptor antagonism blocks depletion of cal-
cium stores and reduces potentiated capacitative calcium en-
try in strain-injured neurons and astrocytes. J. Neurotrauma,
21(3): 271–281.
Cho, I.H., Im, J.Y., Kim, D., Kim, K.S., Lee, J.K. and Han,
P.L. (2003) Protective effects of extracellular glutathione
against Zn2+-induced cell death in vitro and in vivo. J.
Neurosci. Res., 74(5): 736–743.
Church, A.J. and Andrew, R.D. (2005) Spreading depression
expands traumatic injury in neocortical brain slices. J.
Neurotrauma, 22(2): 277–290.
Coats, B. and Margulies, S.S. (2006) Material properties of
porcine parietal cortex. J. Biomech., 39(13): 2521–2525.
Dash, P.K., Mach, S.A. and Moore, A.N. (2002) The role of
extracellular signal-regulated kinase in cognitive and motor
deficits following experimental traumatic brain injury. Neuro-
science, 114(3): 755–767.
Denny-Brown, D. (1945) Cerebral concussion. Physiol. Rev.,
25: 296–325.
Denny-Brown, D. and Russel, W.R. (1941) Experimental
cerebral concussion. Brain, 64(2–3): 92–164.
DeRidder, M.N., Simon, M.J., Siman, R., Auberson, Y.P.,
Raghupathi, R. and Meaney, D.F. (2006) Traumatic mechan-
ical injury to the hippocampus in vitro causes regional ca-
spase-3 and calpain activation that is influenced by NMDA
receptor subunit composition. Neurobiol. Dis., 22(1): 165–176.
Dixon, C.E., Lighthall, J.W. and Anderson, T.E. (1988) Physio-
logic, histopathologic, and cineradiographic characterization
of a new fluid-percussion model of experimental brain injury
in the rat. J. Neurotrauma, 5(2): 91–104.
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-
induced injury of cells in culture: characterization of the
model using astrocytes. J. Neurotrauma, 12(3): 325–339.
Epstein, M.H. (1971) Relative susceptibility of elements of the
cerebral cortex to mechanical trauma in the rat. J. Neu-
rosurg., 35(5): 517–522.
Faden, A.I., Fox, G.B., Di, X., Knoblach, S.M., Cernak, I.,
Mullins, P., Nikolaeva, M. and Kozikowski, A.P. (2003)
Neuroprotective and nootropic actions of a novel cyclized
dipeptide after controlled cortical impact injury in mice. J.
Cereb. Blood Flow Metab., 23(3): 355–363.
Faden, A.I., Ivanova, S.A., Yakovlev, A.G. and Mukhin, A.G.
(1997) Neuroprotective effects of group III mGluR in trau-
matic neuronal injury. J. Neurotrauma, 14(12): 885–895.
Faden, A.I., Knoblach, S.M., Movsesyan, V.A. and Cernak, I.
(2004) Novel small peptides with neuroprotective and noo-
tropic properties. J. Alzheimers Dis., 6(Suppl 6): S93–S97.
Faden, A.I., Movsesyan, V.A., Knoblach, S.M., Ahmed, F. and
Cernak, I. (2005) Neuroprotective effects of novel small pep-
tides in vitro and after brain injury. Neuropharmacology,
49(3): 410–424.
Faden, A.I., O’Leary, D.M., Fan, L., Bao, W., Mullins, P.G.
and Movsesyan, V.A. (2001) Selective blockade of the

mGluR1 receptor reduces traumatic neuronal injury in vitro
and improves outcome after brain trauma. Exp. Neurol.,
167(2): 435–444.
Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanop-
oration induced by diffuse traumatic brain injury: an irre-
versible or reversible response to injury? J. Neurosci., 26(12):
3130–3140.
Fitch, M.T., Doller, C., Combs, C.K., Landreth, G.E. and
Silver, J. (1999) Cellular and molecular mechanisms of glial
scarring and progressive cavitation: in vivo and in vitro anal-
ysis of inflammation-induced secondary injury after CNS
trauma. J. Neurosci., 19(19): 8182–8198.
Floyd, C.L., Gorin, F.A. and Lyeth, B.G. (2005) Mechanical
strain injury increases intracellular sodium and reverses Na+/
Ca2+ exchange in cortical astrocytes. Glia, 51(1): 35–46.
Floyd, C.L., Rzigalinski, B.A., Sitterding, H.A., Willoughby,
K.A. and Ellis, E.F. (2004) Antagonism of group I metabo-
tropic glutamate receptors and PLC attenuates increases in
inositol trisphosphate and reduces reactive gliosis in strain-
injured astrocytes. J. Neurotrauma, 21(2): 205–216.
Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A.,
Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury
of cultured astrocytes alters inositol (1,4,5)-trisphosphate-
mediated signaling. Glia, 33(1): 12–23.
Gagliardi, R.J. (2000) Neuroprotection, excitotoxicity and
NMDA antagonists. Arq. Neuropsiquiatr., 58(2B): 583–588.
Geddes, D.M. and Cargill II, R.S. (2001) An in vitro model of
neural trauma: device characterization and calcium response
to mechanical stretch. J. Biomech. Eng., 123(3): 247–255.
Geddes, D.M., LaPlaca, M.C. and Cargill II, R.S. (2003) Sus-
ceptibility of hippocampal neurons to mechanically induced
injury. Exp. Neurol., 184(1): 420–427.
Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F.
(2006a) Mechanisms and consequences of neuronal stretch
injury in vitro differ with the model of trauma. J. Neuro-
trauma, 23(2): 193–204.
Geddes-Klein, D.M., Serbest, G., Mesfin, M.N., Cohen, A.S.
and Meaney, D.F. (2006b) Pharmacologically induced cal-
cium oscillations protect neurons from increases in cytosolic
calcium after trauma. J. Neurochem., 97(2): 462–474.
Gennarelli, T.A. (1994) Animate models of human head injury.
J. Neurotrauma, 11(4): 357–368.
Gennarelli, T.A., Spielman, G.M., Langfitt, T.W., Gildenberg,
P.L., Harrington, T., Jane, J.A., Marshall, L.F., Miller, J.D.
and Pitts, L.H. (1982a) Influence of the type of intracranial
lesion on outcome from severe head injury. J. Neurosurg., 56:
26–32.
Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I.,
Thompson, C.J. and Marcincin, R.P. (1982b) Diffuse axonal
injury and traumatic coma in the primate. Ann. Neurol., 12:
564–574.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement
of AMPA-mediated current after traumatic injury in cortical
neurons. J. Neurosci., 19(17): 7367–7374.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical in-
jury modulates AMPA receptor kinetics via an NMDA
receptor-dependent pathway. J. Neurotrauma, 21(6): 719–732.
37
Graham, D.I., Adams, J.H., Nicoll, J.A., Maxwell, W.L. and
Gennarelli, T.A. (1995) The nature, distribution and causes
of traumatic brain injury. Brain Pathol., 5(4): 397–406.
Gross, G.W., Lucas, J.H. and Higgins, M.L. (1983) Laser mi-
crobeam surgery: ultrastructural changes associated with
neurite transection in culture. J. Neurosci., 3(10): 1979–1993.
Gubellini, P., Pisani, A., Centonze, D., Bernardi, G. and Ca-
labresi, P. (2004) Metabotropic glutamate receptors and
striatal synaptic plasticity: implications for neurological dis-
eases. Prog. Neurobiol., 74(5): 271–300.
Gutierrez, E., Huang, Y., Haglid, K., Bao, F., Hansson, H.A.,
Hamberger, A. and Viano, D. (2001) A new model for diffuse
brain injury by rotational acceleration: I model, gross appear-
ance, and astrocytosis. J. Neurotrauma, 18(3): 247–257.
Hardingham, G.E. and Bading, H. (2003) The Yin and Yang of
NMDA receptor signalling. Trends Neurosci., 26(2): 81–89.
Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis,
E.F. (2000) Astrocytes generate isoprostanes in response to
trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420.
Iwata, A., Stys, P.K., Wolf, J.A., Chen, X.H., Taylor, A.G.,
Meaney, D.F. and Smith, D.H. (2004) Traumatic axonal in-
jury induces proteolytic cleavage of the voltage-gated sodium
channels modulated by tetrodotoxin and protease inhibitors.
J. Neurosci., 24(19): 4605–4613.
Kao, C.Q., Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004)
Potentiation of GABA(A) currents after mechanical injury of
cortical neurons. J. Neurotrauma, 21(3): 259–270.
Katano, H., Fujita, K., Kato, T., Asai, K., Kawamura, Y.,
Masago, A. and Yamada, K. (1999) Traumatic injury in vitro
induces IEG mRNAs in cultured glial cells, suppressed by co-
culture with neurons. Neuroreport, 10(12): 2439–2448.
Knoblach, S.M., Alroy, D.A., Nikolaeva, M., Cernak, I.,
Stoica, B.A. and Faden, A.I. (2004) Caspase inhibitor
z-DEVD-fmk attenuates calpain and necrotic cell death in
vitro and after traumatic brain injury. J. Cereb. Blood Flow
Metab., 24(10): 1119–1132.
Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A.
and Ellis, E.F. (1997) Alterations in phosphatidylcholine me-
tabolism of stretch-injured cultured rat astrocytes. J. Neuro-
chem., 68(5): 1904–1910.
Langlois, J.A. and Sattin, R.W. (2005) Traumatic brain injury
in the United States: research and programs of the Centers
for Disease Control and Prevention (CDC). J. Head Trauma
Rehabil., 20(3): 187–188.
LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II,
R.S. (2005) High rate shear strain of three-dimensional neu-
ral cell cultures: a new in vitro traumatic brain injury model.
J. Biomech., 38(5): 1093–1105.
LaPlaca, M.C., Lee, V.M. and Thibault, L.E. (1997) An in vitro
model of traumatic neuronal injury: loading rate-dependent
changes in acute cytosolic calcium and lactate dehydrogenase
release. J. Neurotrauma, 14(6): 355–368.
Lighthall, J.W. (1988) Controlled cortical impact: a new experi-
mental brain injury model. J. Neurotrauma, 5: 1–15.
Lucas, J.H. (1987) Proximal segment retraction increases the
probability of nerve cell survival after dendrite transection.
Brain Res., 425(2): 384–387.
38
Lucas, J.H., Gross, G.W., Emery, D.G. and Gardner, C.R.
(1985) Neuronal survival or death after dendrite transection
close to the perikaryon: correlation with electrophysiologic,
morphologic, and ultrastructural changes. Cent. Nerv. Syst.
Trauma, 2(4): 231–255.
Lusardi, T.A., Rangan, J., Sun, D., Smith, D.H. and Meaney,
D.F. (2004) A device to study the initiation and propagation
of calcium transients in cultured neurons after mechanical
stretch. Ann. Biomed. Eng., 32(11): 1546–1558.
Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J.,
Kita, H. and Demetriadou, K. (1994) A new model of diffuse
brain injury in rats. Part I: Pathophysiology and biomechan-
ics. J. Neurosurg., 80(2): 291–300.
Maxwell, W.L., Povlishock, J.T. and Graham, D.L. (1997) A
mechanistic analysis of nondisruptive axonal injury: a review.
J. Neurotrauma, 14(7): 419–440.
McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak,
S., Soares, H. and Faden, A.L. (1989a) Traumatic brain in-
jury in the rat: characterization of a lateral fluid-percussion
model. Neuroscience, 28(1): 233–244.
McIntosh, T.K., Vink, R., Yamakami, I. and Faden, A.I.
(1989b) Magnesium protects against neurological deficit after
brain injury. Brain Res., 482(2): 252–260.
McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F.
(1996) Stretch-induced injury of cultured neuronal, glial, and
endothelial cells. Effect of polyethylene glycol-conjugated
superoxide dismutase. Stroke, 27(5): 934–940.
Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller,
R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical
analysis of experimental diffuse axonal injury. J. Neurotrau-
ma, 12(4): 689–694.
Mori, T., Wang, X., Jung, J.C., Sumii, T., Singhal, A.B., Fini,
M.E., Dixon, C.E., Alessandrini, A. and Lo, E.H. (2002)
Mitogen-activated protein kinase inhibition in traumatic
brain injury: in vitro and in vivo effects. J. Cereb. Blood Flow
Metab., 22(4): 444–452.
Morrison III, B., Eberwine, J.H., Meaney, D.F. and McIntosh,
T.K. (2000) Traumatic injury induces differential expression
of cell death genes in organotypic brain slice cultures deter-
mined by complementary DNA array hybridization. Neuro-
science, 96(1): 131–139.
Morrison III, B., Meaney, D.F. and McIntosh, T.K. (1998a)
Mechanical characterization of an in vitro device designed to
quantitatively injure living brain tissue. Ann. Biomed. Eng.,
26(3): 381–390.
Morrison III, B., Saatman, K.E., Meaney, D.F. and McIntosh,
T.K. (1998b) In vitro central nervous system models of
mechanically induced trauma: a review. J. Neurotrauma,
15(11): 911–928.
Movsesyan, V.A. and Faden, A.I. (2006) Neuroprotective
effects of selective group II mGluR activation in brain
trauma and traumatic neuronal injury. J. Neurotrauma,
23(2): 117–127.
Movsesyan, V.A., O’Leary, D.M., Fan, L., Bao, W., Mullins,
P.G., Knoblach, S.M. and Faden, A.I. (2001) mGluR5
antagonists 2-methyl-6-(phenylethynyl)-pyridine and (E)-2-
methyl-6-(2-phenylethenyl)-pyridine reduce traumatic neuronal
injury in vitro and in vivo by antagonizing N-methyl-D-aspart-
ate receptors. J. Pharmacol. Exp. Ther., 296(1): 41–47.
Mukhin, A.G., Ivanova, S.A., Allen, J.W. and Faden, A.I.
(1998) Mechanical injury to neuronal/glial cultures in micro-
plates: role of NMDA receptors and pH in secondary neu-
ronal cell death. J. Neurosci. Res., 51(6): 748–758.
Mukhin, A.G., Ivanova, S.A., Knoblach, S.M. and Faden, A.I.
(1997) New in vitro model of traumatic neuronal injury:
evaluation of secondary injury and glutamate receptor-me-
diated neurotoxicity. J. Neurotrauma, 14(9): 651–663.
Nahum, A.M., Smith, R. and Ward, C.C. (1977) Intracranial
pressure dynamics during head impact. In: 21st Annual Stapp
Car Crash Conference, Society of Automotive Engineers. San
Diego, CA.
Neary, J.T., Kang, Y., Willoughby, K.A. and Ellis, E.F. (2003)
Activation of extracellular signal-regulated kinase by stretch-
induced injury in astrocytes involves extracellular ATP and
P2 purinergic receptors. J. Neurosci., 23(6): 2348–2356.
Nusholtz, G.S., Wylie, E.B. and Glascoe, L.G. (1995) Internal
cavitation in simple head impact model. J. Neurotrauma,
12(4): 707–714.
O’Dell, D.M., Raghupathi, R., Crino, P.B., Eberwine, J.H.
and McIntosh, T.K. (2000) Traumatic brain injury alters
the molecular fingerprint of TUNEL-positive cortical neurons
in vivo: a single-cell analysis. J. Neurosci., 20(13): 4821–4828.
Ommaya, A.K., Yarnell, P., Hirsch, A.E. and Harris, E.H.
(1967) Scaling of experimental data on cerebral concussion in
sub-human primates to concussion threshold for man. In:
13th Stapp Car Crash Conference, Anaheim, CA, Society of
Automotive Engineers.
Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock,
J.T. (1994) Traumatically induced altered membrane perme-
ability: its relationship to traumatically induced reactive axo-
nal change. J. Neurotrauma, 11(5): 507–522.
Pike, B.R., Zhao, X., Newcomb, J.K., Glenn, C.C., Anderson,
D.K. and Hayes, R.L. (2000) Stretch injury causes calpain
and caspase-3 activation and necrotic and apoptotic cell
death in septo-hippocampal cell cultures. J. Neurotrauma,
17(4): 283–298.
Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers
of proteolytic damage following traumatic brain injury. Brain
Pathol., 14(2): 202–209.
Povlishock, J.T. and Katz, D.I. (2005) Update of neuropathol-
ogy and neurological recovery after traumatic brain injury.
J. Head Trauma Rehabil., 20(1): 76–94.
Prado, G.R., Ross, J.D., DeWeerth, S.P. and LaPlaca, M.C.
(2005) Mechanical trauma induces immediate changes in
neuronal network activity. J. Neural Eng., 2(4): 148–158.
Prange, M.T. and Margulies, S.S. (2002) Regional, directional,
and age-dependent properties of the brain undergoing large
deformation. J. Biomech. Eng., 124(2): 244–252.
Raghupathi, R., Grants, I., Rosenberg, L.J., McIntosh, T.K.
and Lucas, J.H. (1998) Increased jun immunoreactivity in an
in vitro model of mammalian spinal neuron physical injury.
J. Neurotrauma, 15(7): 555–561.
Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005)
Myelinated and unmyelinated axons of the corpus callosum

differ in vulnerability and functional recovery following
traumatic brain injury. Exp. Neurol., 196(1): 126–137.
Regan, R.F. and Choi, D.W. (1994) The effect of NMDA,
AMPA/kainate, and calcium channel antagonists on trau-
matic cortical neuronal injury in culture. Brain Res., 633(1–2):
236–242.
Regan, R.F. and Panter, S.S. (1995) Traumatic neuronal injury
in cortical cell culture is attenuated by 21-aminosteroids.
Brain Res., 682(1–2): 144–150.
Rzigalinski, B.A., Liang, S., McKinney, J.S., Willoughby, K.A.
and Ellis, E.F. (1997) Effect of Ca2+ on in vitro astrocyte
injury. J. Neurochem., 68(1): 289–296.
Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis,
E.F. (1998) Intracellular free calcium dynamics in stretch-
injured astrocytes. J. Neurochem., 70(6): 2377–2385.
Serbest, G., Horwitz, J., Jost, M. and Barbee, K. (2006) Mech-
anisms of cell death and neuroprotection by poloxamer 188
after mechanical trauma. FASEB J., 20(2): 308–310.
Shah, P.T., Yoon, K.W., Xu, X.M. and Broder, L.D. (1997)
Apoptosis mediates cell death following traumatic injury in
rat hippocampal neurons. Neuroscience, 79(4): 999–1004.
Shreiber, D., Bain, A.C. and Meaney, D.F. (1997) In vivo
thresholds for mechanical injury to the blood brain barrier. J.
Passenger Cars, 106: 3792–3806.
Sieg, F., Wahle, P. and Pape, H.C. (1999) Cellular reactivity to
mechanical axonal injury in an organotypic in vitro model of
neurotrauma. J. Neurotrauma, 16(12): 1197–1213.
Simonyi, A., Schachtman, T.R. and Christoffersen, G.R. (2005)
The role of metabotropic glutamate receptor 5 in learning
and memory processes. Drug News Perspect., 18(6): 353–361.
Singleton, R.H., Zhu, J., Stone, J.R. and Povlishock, J.T. (2002)
Traumatically induced axotomy adjacent to the soma does not
result in acute neuronal death. J. Neurosci., 22(3): 791–802.
Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T.
(2002) Repeated mild injury causes cumulative damage to
hippocampal cells. Brain, 125(Pt 12): 2699–2709.
Smith, D.H., Soares, H.D., Pierce, J.S., Pearlman, K., Saatman,
K.E., Meaney, D.F., Dixon, C.E. and McIntosh, T.K. (1995)
A model of parasagittal controlled cortical impact in the
mouse: cognitive and histopathologic effects. J. Neurotrauma,
12(2): 169–178.
Smith, D.H., Wolf, J.A., Lusardi, T.A., Lee, V.M. and Meaney,
D.F. (1999) High tolerance and delayed elastic response of
cultured axons to dynamic stretch injury. J. Neurosci., 19(11):
4263–4269.
Stalhammar, D. and Olsson, Y. (1975) Experimental brain
damage from fluid pressures due to impact acceleration. Acta
Neurol. Scand., 52: 38–55.
Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu,
L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired ax-
onal transport and altered axolemmal permeability occur in
distinct populations of damaged axons following traumatic
brain injury. Exp. Neurol., 190(1): 59–69.
Sullivan, H.G., Martinez, J., Becker, D.P., Miller, J.D.,
Griffith, R. and Wist, A.O. (1976) Fluid-percussion model
of mechanical brain injury in the cat. J. Neurosurg., 45(5):
521–534.
39
Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical
perturbation of cultured cortical neurons reveals a stretch-
induced delayed depolarization. J. Neurophysiol., 74(6):
2767–2773.
Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1997) Inhibition of
the electrogenic Na pump underlies delayed depolarization
of cortical neurons after mechanical injury or glutamate.
J. Neurophysiol., 77(2): 632–638.
Tecoma, E.S., Monyer, H., Goldberg, M.P. and Choi, D.W.
(1989) Traumatic neuronal injury in vitro is attenuated by
NMDA antagonists. Neuron, 2(6): 1541–1545.
Thibault, L.E., Meaney, D.F., Anderson, B.J. and Marmarou,
A. (1992) Biomechanical aspects of a fluid percussion model
of brain injuiry. J. Neurotrauma, 9(4): 311–322.
Thurman, D.J., Alverson, C., Dunn, K.A., Guerrero, J. and
Sniezek, J.E. (1999) Traumatic brain injury in the United
States: a public health perspective. J. Head Trauma Rehabil.,
14(6): 602–615.
Wang, X., Mori, T., Jung, J.C., Fini, M.E. and Lo, E.H. (2002)
Secretion of matrix metalloproteinase-2 and -9 after mechan-
ical trauma injury in rat cortical cultures and involvement of
MAP kinase. J. Neurotrauma, 19(5): 615–625.
Weber, J.T. (2004) Calcium homeostasis following traumatic
neuronal injury. Curr. Neurovasc. Res., 1(2): 151–171.
Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2001) Traumatic
injury of cortical neurons causes changes in intracellular
calcium stores and capacitative calcium influx. J. Biol.
Chem., 276(3): 1800–1807.
Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F.
and Ellis, E.F. (1999) Alterations in calcium-mediated signal
transduction after traumatic injury of cortical neurons. Cell
Calcium, 26(6): 289–299.
Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir,
J.K. and Ellis, E.F. (2004) S100B protein is released by in
vitro trauma and reduces delayed neuronal injury. J. Neuro-
chem., 91(6): 1284–1291.
Wolf, J.A., Stys, P.K., Lusardi, T., Meaney, D. and Smith, D.H.
(2001) Traumatic axonal injury induces calcium influx mod-
ulated by tetrodotoxin-sensitive sodium channels. J. Neuro-
sci., 21(6): 1923–1930.
Xiao-Sheng, H., Sheng-Yu, Y., Xiang, Z., Zhou, F. and Jian-
ning, Z. (2000) Diffuse axonal injury due to lateral head ro-
tation in a rat model. J. Neurosurg., 93(4): 626–633.
Yang, X.F., Cao, F., Pan, D.S., Liu, W.G., Hu, W.W., Zheng,
X.J., Zhao, X.Q. and Lu, S.T. (2006) Establishment of a
mechanical injury model of rat hippocampal neurons in vitro.
Chin. J. Traumatol., 9(1): 29–33.
Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996)
Reduction of voltage-dependent Mg2+ blockade of NMDA
current in mechanically injured neurons. Science, 274(5294):
1921–1923.
Zhang, L., Yang, K.H. and King, A.I. (2004) A proposed injury
threshold for mild traumatic brain injury. J. Biomech. Eng.,
126(2): 226–236.
Zhu, Q., Prange, M. and Margulies, S. (2006) Predicting un-
consciousness from a pediatric brain injury threshold. Dev.
Neurosci., 28(4–5): 388–395.
 SECTION III

Pathological Mechanisms of Injury

This page intentionally left blank
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 4

Cellular and subcellular change evoked by diffuse

traumatic brain injury: a complex web of change
extending far beyond focal damage

Orsolya Farkas and John T. Povlishock

Department of Anatomy and Neurobiology, Medical College of Virginia Campus, Virginia Commonwealth University,
P.O. Box 980709, Richmond, VA 23298, USA

Abstract: Until recently, our understanding of the cellular and subcellular changes evoked by diffuse
traumatic brain injury has been framed in the context of primary focal injury. In this regard, the ensuing
cell death cascades were linked to contusional-mediated changes associated with frank hemorrhage and
ischemia, and these were assumed to contribute to the observed apoptotic and necrotic neuronal death.
Little consideration was given to the potential that other non-contusional cell death cascades could
have been triggered by the diffuse mechanical forces of injury. While the importance of these classical,
contusion-related apoptotic and necrotic cell death cascades cannot be discounted with diffuse injury,
more recent information suggests that the mechanical force of injury itself can diffusely porate the
neuronal plasmalemma and its axolemmal membranes, evoking other forms of cellular response that can
contribute to cell injury or death. In this regard, the duration of the membrane alteration appears to be a
dependent factor, with enduring membrane change, potentially leading to irreversible damage, whereas
more transient membrane perturbation can be followed by cell membrane resealing associated with
recovery and/or adaptive change. With more enduring mechanical membrane perturbation, it appears
that some of the traditional death cascades involving the activation of cysteine proteases are at work.
Equally important, non-traditional pathways involving the lysosomal dependent release of hydrolytic
enzymes may also be players in the ensuing neuronal death. These mechanically related factors that
directly impact upon the neuronal somata may also be influenced by concomitant and/or secondary
axotomy-mediated responses. This axonal injury, although once thought to involve a singular intraaxo-
nal response to injury, is now known to be more complex, reflecting differential responses to injuries of
varying severity. Moreover, it now appears that fiber size and type may also influence the axon’s reaction
to injury. In sum, this review explicates the complexity of the cellular and subcellular responses evoked by
diffuse traumatic brain injury in both the neuronal somata and its axonal appendages. This review
further illustrates that our once simplistic views framed by evidence based upon contusional and/or
ischemic change do not fully explain the complex repertoire of change evoked by diffuse traumatic brain
injury.

Keywords: diffuse traumatic brain injury; neuronal injury; axonal injury; necrosis; apoptosis

Corresponding author. E-mail: jtpovlis@vcu.edu

DOI: 10.1016/S0079-6123(06)61004-2 43
44
Introduction
Over the last 40 years, our understanding of the
complex pathobiology of traumatic brain injury
(TBI) has improved significantly. From both the
clinical and basic science perspective, now most
consider the pathobiology of TBI in the context of
focal as well as diffuse change (Povlishock and
Katz, 2005). It is also recognized in the clinical
setting that focal injuries are typically embedded
within concomitant diffuse pathologies, with the
diffuse pathologies being the major determinants
of the adverse outcomes associated with TBI
(Povlishock and Katz, 2005). Despite this recogni-
tion of the importance of diffuse change, a dispro-
portionate number of basic science studies have
focused on focal injuries, such as contusional change
without a parallel consideration of any diffuse
changes that accompany the injury. In this review,
we attempt to address what is known as well as
what is controversial in our understanding of diffuse
change within the traumatically injured brain paren-
chyma, examining both the cellular and subcellular
responses evoked by diffuse traumatic brain injury
(DTBI) to the neurons and their axonal extensions.
In this review, we attempt also to consider new
experimental directions for this important area of
scientific inquiry.
Neuronal damage associated with DTBI
Neuronal damage and death associated with several
types of brain injury including TBI in humans and
animals have been widely studied and, as such, have
been the focus of several major reviews (Adams
et al., 1980; Cervos-Navarro and Lafuente, 1991;
Kotapka et al., 1992, 1993; Ross et al., 1993;
Kermer et al., 1999; Raghupathi, 2004; Yakovlev
and Faden, 2004). The majority of these studies and
reviews, however, have focused on focal injuries,
addressing neuronal death in localized contusional
or pericontusional regions. Because all these studies
provided excellent, detailed descriptions of the
necrotic and apoptotic neuronal change associ-
ated with focal TBI, the present review follows a
different track, focusing primarily on the under
appreciated phenomenon of diffuse traumatic
neuronal change.
Morphological characteristics of neuronal injury
after DTBI
As noted above, unlike the wide body of literature
on neuronal death after focal TBI, there are rela-
tively few descriptions of neuronal injury and
death occurring remote from contusional regions
(Dietrich et al., 1994a, b; Colicos et al., 1996;
Hicks et al., 1996), or after DTBI (Smith et al.,
1997; Runnerstam et al., 2001; Cernak et al.,
2002). These studies, similar to those focusing on
focal TBI, describe two distinct forms of cell
death, namely necrosis and apoptosis, within the
neocortex, hippocampus and diencephalon. With
diffuse injury, damaged/dying neurons have been
described not only in the neocortex in or near
contusions, but also in regions remote from the
impact/contusional site, such as the C1, C2, C3
layers and the dentate hilus of hippocampus and
thalamus. Additional neuronal damage/death has
been described scattered in the caudate/putamen
and the inferior and superior colliculi, consistent
with the diffuse nature of the injury. Like previ-
ous descriptions of necrosis, these damaged neu-
rons were dark and shrunken with Nissl staining
(Dietrich et al., 1994b; Hicks et al., 1996) or
eosinophilic with acid fuchsin staining (Cortez
et al., 1989; Dietrich et al., 1994a; Hicks et al.,
1996), revealing distorted profiles and vacuolizat-
ion. At the ultrastructural level, such neurons
demonstrated increased cytoplasmic and nuclear
electron density, swollen mitochondria, vacuolated
cytoplasm, plasma membrane disruption, pyknotic
nuclei and perisomatic glial swelling (Dietrich et al.,
1994b). Importantly, neuronal necrosis has also
been observed after DTBI in the absence of any
focal lesion, such as contusion or hematoma for-
mation. Specifically, neuronal necrosis bilaterally
scattered in the neocortex (Singleton et al., 2002;
Singleton and Povlishock, 2004; Farkas et al., 2006)
as well as in the cerebellum, and the C1 and C3
pyramidal layers and dentate hilus of the hippo-
campus has been described (Kotapka et al., 1991;
Smith et al., 1997; Singleton et al., 2002; Singleton

and Povlishock, 2004). While this limited literature
supports the occurrence necrotic cell death with
DTBI, to date, the mechanisms underlying this
necrosis are not well understood. The presence of
necrotic neurons remote from or in the absence of
contusion suggests that the pathogenesis of this ne-
crosis does not follow from any overt destructive or
ischemic processes comparable to those occurring
within contusional foci. Unfortunately, our current
beliefs regarding the pathogenesis of DTBI-induced
necrotic change have been biased by the ischemic
literature, wherein disproportionate emphasis has
been placed on membrane pump failure and sub-
sequent ionic dysregulation and calpain-mediated
proteolysis. It is likely, however, that other poten-
tially important factors and mechanisms associated
with direct mechanical perturbation and its sequel-
ae may be at work in the pathogenesis of this
diffuse necrotic neuronal death and, as such, merit
consideration (vide infra).
Like necrosis, scattered apoptosis has also
been described following DTBI. Although under-
appreciated and originally thought to take place
only in ‘‘physiologic’’ programmed neuronal death
processes and/or slowly progressive neurodegene-
rative diseases, its role in the pathobiology of
DTBI has become more apparent. Rink et al.
(1995) first identified this process in TBI, positing
that it played a role in delayed neuronal damage
in contrast to necrosis that was linked to both
acute and delayed neuronal death. Neurons with
apoptotic morphology have been described in con-
tusional foci, but more commonly have been iden-
tified either in the absence of contusions or in
foci remote from contusional change, such as the
hippocampus and/or diencephalon (Colicos and
Dash, 1996; Colicos et al., 1996; Pravdenkova et al.,
1996; Conti et al., 1998; Fox et al., 1998; Newcomb
et al., 1999; Lin et al., 2001; Runnerstam et al.,
2001; Cernak et al., 2002, 2004; Raghupathi et al.,
2002). At ultrastructural level, these apoptotic
neurons demonstrated cytoplasmic condensation,
nuclear pyknosis, chromatin condensation, cell
rounding, membrane blebbing, cytoskeletal colla-
pse and, ultimately, disintegration of the cell into
small fragments forming apoptotic bodies. Similar
to our understanding of necrosis, the pathogenesis
of apoptosis occurring after DTBI is not well
45
appreciated. It is believed that the balance between
the anti- and pro-apoptotic signals determines if
the injured neuron dies or survives. At present,
apoptotic cell death is considered to play a role in
delayed neuronal death occurring several hours to
weeks after DTBI, when the dominance of pro-
apoptotic factors in the presence of a persistent
energy supply results in the activation of cysteine
proteases, such as caspases that are regulators and
effectors of apoptotic cell death.
Mechanisms eliciting neuronal injury and death
after DTBI
Similar to their morphological differences, apop-
tosis and necrosis have also been distinguished by
the differences in their pathological mechanisms.
As noted, apoptosis is an active process requir-
ing energy while necrosis is a passive event that
results from energy failure and consequent loss of
ionic homeostasis with the ultimate recruitment of
an inflammatory response. Although both proc-
esses were once considered distinct, recent evi-
dence demonstrates that necrosis and apoptosis
can evolve in parallel in the same injured tissue.
Further, it has been recently suggested based upon
the cellular microenvironment and energy supply
that a switch between necrosis and apoptosis can
occur within an individual cell (Kermer et al., 1999;
Kitanaka and Kuchino, 1999; Nicotera et al., 1999;
Raghupathi, 2004). This hybrid form of cell death,
sometimes referred as aponecrosis, reflects the fact
that these cells show the morphological character-
istics of both types of cell death (Formigli et al.,
2000). To date, multiple reviews have considered
those cellular, subcellular and pathophysiological
cascades that may contribute to the above described
necrotic and apoptotic neuronal cell death cascades
(Kermer et al., 1999; Raghupathi et al., 2000; Zipfel
et al., 2000; Keane et al., 2001; Raghupathi, 2004;
Yakovlev and Faden, 2004). Accordingly, it is un-
necessaryto revisit these details. Rather, we provide
here only a summary of these well-characterized
processes, while providing, in subsequent passages,
more detail on other not well appreciated mecha-
nisms related to the pathogenesis of DTBI.
46
Neuroexcitation and calcium dysregulation
Calcium dysregulation has been linked to both
necrotic and apoptotic processes, although differ-
ent intracellular cascades are involved. To date,
it is known that the traumatic episode can be
associated with generalized neuroexcitotoxicity,
involving excitatory amino acid (EAA) release
that can trigger several subcellular alterations via
Na+ and Ca2+ influx from the extracellular space.
Using microdialysis, the EAA glutamate has been
found to be elevated after TBI both in animals and
humans. While this glutamate elevation initially
was described after focal TBI (Faden et al., 1989;
Katayama et al., 1990; Palmer et al., 1993), recent
studies also showed the involvement of EAA
release after DTBI (Runnerstam et al., 2001; Goda
et al., 2002; Fei et al., 2005). The subsequent mas-
sive Ca2+ influx via glutamate/NMDA channels
triggers further Ca2+ release from intracellular
stores and thereby dramatically elevates free intra-
cellular Ca2+. This increased cytosolic Ca2+ can
lead to activation of proteases, such as calpains,
phosphatases, protein kinases or nitric oxide synt-
hase, all which have been linked in various scenar-
ios to either necrotic or apoptotic cell death
(Kermer et al., 1999).
Free radicals
Concurrent with and influenced by the glutamate
NMDA receptor activation and Ca2+ dysregula-
tion described above, the generation of oxygen
free radicals can also occur after DTBI. Either
via the generation of nitric oxide, the accelerated
metabolism of arachidonate acid via multiple
pathways or the generation of superoxide anions
via an evoked inflammatory response, destruc-
tive radical species can be generated with lethal
results for the brain (Kontos and Povlishock,
1986; Povlishock and Kontos, 1992). With more
localized radical generation and Ca2+ dysregula-
tion, this creates the permissive environment for
mitochondrial damage, cytochrome C release and
apoptotic cell death. With more generalized radi-
cal generation, the oxygen radicals participate
in highly destructive processes including lipid
peroxidation, protein nitrosylation and DNA
degradation (Braughler and Hall, 1992).
Calpains and caspases
As noted above, increased intracellular Ca2+ can
activate proteases, such as calpain. Calpain, a
member of cysteine proteases, and its activation
are well known to be associated with several
different types of brain injury, including DTBI
(Kupina et al., 2001, 2002; Buki et al., 2003;
Farkas et al., 2006). It has several different iso-
types, including m-calpain and m-calpain that are
ubiquitous in the central nervous system. Although
calpains have several different substrates within the
cell, such as cytoskeletal elements, neurofilaments,
protein kinase C, calmodulin binding proteins and
transcription factors (for review, see Wang (2000);
Huang and Wang (2001)), one of the most widely
investigated of the calpain substrates is spectrin.
Spectrin is the primary component of the neuronal
cytoskeleton and its cleavage by calpain results
in specific and stable breakdown products, such as
150 and 145 kDa spectrin breakdown products
(SBDPs). The presence of calpain-specific SBDPs
as a result of calpain-mediated spectrin proteolysis
(CMSP) has been routinely observed both after
experimental (Kampfl et al., 1996; Saatman et al.,
1996; Pike et al., 1998, 2001; Farkas et al., 2006)
and human TBI (McCracken et al., 1999; Pineda
et al., 2004; Farkas et al., 2005), and now is being
used as one of the markers of Ca2+-induced cell
injury and death. Excessive Ca2+ influx and the
concomitant calpain activation lead to the proteoly-
tic degradation of intracellular proteins and mem-
branes. The ensuing cytoskeletal and membrane
damage can destroy cell integrity, cause increased
membrane permeability, compromise the transport
of essential cell products, induce aberrant signaling
cascades and finally lead to cell death (Kermer
et al., 1999). Although calpain activation can lead
to either necrotic or apoptotic change, most concur
that in TBI it is primarily responsible for necrotic
cell death (Wang, 2000; Raghupathi, 2004).
Like calpains, caspases are also members of the
cysteine protease family, with these proteases
thought to play a major role in apoptotic cell

death (Eldadah and Faden, 2000; Wang, 2000;
Raghupathi, 2004). Caspase-3, the common apop-
tosis effector, is activated via two major pathways,
either the mitochondrial intrinsic or the death
receptor, extrinsic pathways. Briefly, the apoptotic
stimuli can result in the permeabilization of the
mitochondrial membrane via different pathways,
subsequently triggering the release of cytocrome-c
into the cytoplasm. Cytocrome-c binds to apopto-
sis activating protein-1 (Apaf-1) and the complex
activates caspase-9, which then activates caspase-3
via cleaving its proenzyme form. During the extrin-
sic pathways, apoptotic stimuli activate the death
receptors, such as TNFa1-receptor or Fas receptor.
The receptors recruit adapter proteins and form a
complex, which then induces the autolytic activa-
tion of caspase-8, which activates caspase-3 (for
review, see Wang (2000); Yakovlev and Faden
(2004)). Caspase-3, similar to calpain, cleaves spec-
trin in a specific manner resulting in a 120 kDa
SBDP that can be used for the specific detection
of caspase-3 activity and apoptosis after TBI. The
activation of caspase-3 has been observed in wide-
spread regions both after focal and DTBI (Pike
et al., 1998, 2001; Beer et al., 2000; Clark et al.,
2000; Keane et al., 2001; Yakovlev et al., 2001;
Cernak et al., 2002), as well as in human head injury
(Clark et al., 1999; Pineda et al., 2004; Farkas et al.,
2005).
The Bcl-2 oncogene family
Integral to the apoptotic death cascade is the Bcl-2
proto-oncogene family, which is one of the key
regulators of apoptosis. This family consists of
members with anti-apoptotic (Bcl-2 and BclXL)
and pro-apoptotic (Bax, Bad, Bid, BclXS) activity.
Typically, following apoptotic stimuli, Bax trans-
locates from the cytoplasm to the mitochondrial
membrane resulting in cytocrome-c release, which
by binding with Apaf-1 activates caspases. Bax
is also suggested to directly activate caspase-3.
In contrast, anti-apoptotic Bcl-2 and BclXL can
prevent the mitochondrial permeabilization by
inactivating Bax via heterodimerization. Bcl-2
and BclXL can also inhibit caspase-3 activation
(Merry and Korsmeyer, 1997; Yang et al., 1997).
47
Pro-apoptotic Bax overexpression has been
demonstrated after DTBI (Cernak et al., 2002).
Concomitant decrease in the expression of the
anti-apoptotic Bcl-2 and BclXL has also been
reported (Felderhoff-Mueser et al., 2002). How-
ever, some studies have demonstrated increased
Bcl-2 expression after DTBI (Cernak et al., 2002).
Overall, it has been suggested that the shift between
pro- and anti-apoptotic oncogene expression regu-
lates the fate of the injured neuron, resulting in cell
death if the expression of pro-apoptotic proteins
exceeds the expression of anti-apoptotic proteins,
while promoting cell survival if the anti-apoptotic
Bcl protein expression is upregulated over the pro-
apoptotic members of the Bcl-family (Merry and
Korsmeyer, 1997).
Role of hydrolytic enzymes
In the previous passages, attention has been focused
on those pathways that have received emphasis in
terms of the apoptotic/necrotic pathways occurring
with TBI. While it is clear that calcium dysregula-
tion and cysteine protease activation are integral
to traumatically induced neuronal death, it is of
note that other potentially important pathways and
processes may be operant, although, they have been
give little attention in the context of TBI. Foremost
among these are the calcium-mediated activation of
lysosomal hydrolytic pathways that have received
considerable attention in non-TBI-related neuronal
death. Further, it is well known that lysosomes and
their enzymes are involved in different cell death
pathways in various non-CNS systems as reviewed
by Kroemer and Jaattela (2005) and Yamashima
(2004). Multiple damaging insults are known to
result in the disruption of the lysosomal membrane
leading to the release of several lysosomal enzymes,
such as cathepsins. It has been posited that low
levels of stress result in limited lysosomal mem-
brane rupture and cathepsin release that in turn
elicit apoptosis. In contrast, high levels of stress
can cause generalized lysosomal membrane rupture
resulting in necrosis (Brunk et al., 1997). In the
later scenario, targeted Ca2+-mediated calpain
activation and the likely concomitant production
of free radicals are posited to destroy the lysosomal
48
membrane leading to the leakage of damaging
hydrolytic lysosomal enzymes into the cytoplasm to
cause the digestion of the cell’s structural proteins
(Yamashima, 2000; Yamashima et al., 2003).
Despite the consistent involvement of lysosomes
and cathepsin in cell death in multiple organs,
their contribution to neuronal death following
DTBI has not been rigorously evaluated and
obviously requires attention.
Mechanically related factors
While the previous passages briefly reviewed some
of the neurochemical cascades triggered by the
forces of injury, until recently there has not been
any parallel consideration that the mechanical
force of injury itself could directly perturb the
neuronal cell membrane or its appendages lead-
ing to mechanically induced neuronal death cas-
cades. In this regard, we comment here on recently
published data evaluating the potential damaging
neuronal somatic consequences of traumatically
induced axotomy and/or the direct disruption of
the neuronal plasmalemma.
Axotomy-related changes
Neuronal somatic changes occurring adjacent
to sites of traumatic axonal injury (TAI) have
been recognized previously (Van den Heuvel et al.,
1998, 1999; Cernak et al., 2004); yet, the rela-
tion of these events to any subsequent injury cas-
cades has not been appreciated. Recently, using
immunohistochemical techniques recognizing
amyloid precursor protein (APP), a well accepted
marker of TAI, neuronal somata directly linked to
TAI were identified following DTBI in the cerebral
cortex, hippocampus and thalamus (Singleton
et al., 2002). These studies showed that contrary
to contemporary thought, TAI, even in immediate
proximity to the neuronal soma, did not result in
acute neuronal death. Rather, neurons sustaining
TAI revealed signs of reactive change, including
the loss and degranulation of the rough endoplas-
mic reticulum, disaggregation of polysomes and
dispersal of the Golgi apparatus without any cyto-
skeletal or mitochondrial alterations (Singleton
et al., 2002). These TAI-linked neurons also re-
vealed a transient suppression of protein synthesis.
Taken together, these transient reactive changes
suggest a potential neuronal attempt at reorgan-
ization and repair, rather than the initiation of any
prenecrotic/apoptotic change.
Mechanoporation and the fate of neurons with
membrane disruption
Using extracellular tracer infusion techniques,
multiple in vitro and in vivo studies have demon-
strated that immediately following DTBI, other
non-axotomized neurons can take up either high
molecular weight tracers, such as horseradish per-
oxidase or different molecular weight dextrans
or other molecules normally excluded from the
neuronal cytoplasm by the intact cell membrane.
This immediate tracer uptake suggests that the me-
chanical force of the injury itself evoked neuronal
cell membrane disruption (mechanoporation), that
then most likely allowed the influx of damaging
ions through the compromised cell membrane
(Geddes et al., 2003; Prado and LaPlaca, 2004;
Prado et al., 2004, 2005; Singleton and Povlishock,
2004).
Following these initial descriptions of neu-
ronal membrane perturbation, subsequent studies
revealed a rather heterogeneous neuronal response
to this disruption (Singleton and Povlishock, 2004;
Farkas et al., 2006). Some neurons sustaining cell
membrane disruption revealed Fluoro Jade posit-
ivity and ultrastructural evidence of overt neuron-
al necrosis. In contrast, other neurons sustaining
membrane disruption did not reveal comparable
signs of overt cell damage. Further, at the same
time and in the same brain foci, different popula-
tions of injured neurons did not reveal evidence
of membrane disruption, yet, they demonstrated
either perisomatic axonal injury resulting in
increased neuronal somatic APP-positivity, or the
non-axotomy-related induction of heat shock
protein expression (Singleton and Povlishock,
2004), and/or CMSP (Farkas et al., 2006, vide
infra). In part, because of these divergent neuronal
responses, some of which were suggestive of cell
death whereas others suggested recovery, the

potential for post-traumatic neuronal cell membrane
resealing and recovery became a consideration. Via
the administration of different extracellular tracers
at varied times post-injury, in vitro studies have
revealed that the majority of neurons sustaining
mechanoporation could reseal their disrupted mem-
branes in the first minutes post-TBI (Geddes et al.,
2003; Prado and LaPlaca, 2004; Prado et al., 2005).
In contrast, in vivo studies did not confirm these
observations. Rather, in vivo studies revealed
that more than 50% of the tracer-containing corti-
cal neurons contained both pre- and post-injury
administered tracers at several hours post-DTBI,
suggesting enduring membrane permeability (Fig. 1)
(Farkas et al., 2006). Resealing was observed after
DTBI, yet, to a lesser extent than suggested in vitro
(Fig. 2) (Farkas et al., 2006). To further complicate
this issue, the recent finding of neurons demon-
strating enduring membrane permeability without
evidence of overt cell damage points to the occur-
rence of potentially delayed membrane resealing,
rather than the rapid resealing/repair observed in
vitro (Farkas et al., 2006). Further complexity was
also associated with the finding that other neurons
only revealed post-injury tracer uptake without
evidence of the pre-injury tracer, suggesting the
potential for delayed membrane disruption (Fig. 3)
(Farkas et al., 2006). Although the underlying
mechanism(s) of this delayed membrane disruption
is (are) unknown, the possibility exists that the sus-
tained, elevated intracranial pressure occurring after
injury and persisting for several hours post-TBI
(Farkas et al., 2006) could contribute to this pheno-
menon. Statistical analysis of the neuronal popu-
lations demonstrating different tracer uptake over
time post-DTBI has revealed significant tracer
distribution/redistribution consistent with the com-
plex neuronal membrane changes described above.
Specifically, between 4 and 8 h post-injury, the
proportion of neurons with delayed membrane per-
turbation was significantly different from the pro-
portion of resealed neurons as well as the proportion
of those neurons with enduring membrane pertur-
bation. The change in the proportion of neurons
with resealed and enduring membrane perturbation
over time was not significant. This redistribution of
membrane perturbation types between 4 and 8 h
post-injury may be the result of resealed neurons
49
reopening their membranes to now manifest endur-
ing damage and/or additional neurons suffering
delayed membrane perturbation (Fig. 4) (Farkas
et al., 2006). Collectively, these in vivo studies clearly
confirm the existence of DTBI-injured neuronal
membrane perturbation, while further illustrating
the complex pathobiology associated with DTBI.
This complexity was further highlighted by re-
cent in vitro studies that explored the potential re-
lationship of the above described altered membrane
permeability to CMSP positing that CMSP co-ex-
isted with increased membrane permeability (Pike
et al., 2000; Liu and Schnellmann, 2003; Liu et al.,
2004). In contrast, with in vivo DTBI, the majority
of neurons sustaining either membrane disruption
and/or demonstrating necrotic change did not re-
veal CMSP, suggesting that membrane disruption
itself did not lead to calpain activation. Conversely,
with DTBI, neurons demonstrating CMSP did not
reveal membrane disruption, and, as the majority of
these CMSP-positive neurons revealed only limited
ultrastructural damage, this suggested that CMSP
was not associated with neuronal death after DTBI
(Fig. 5) (Farkas et al., 2006). Taken together with
other studies (Saatman et al., 1996; Brana et al.,
1999), these findings emphasize the need for caution
in interpreting the occurrence of CMSP and its
overall implications for the neuronal injury and
death associated with DTBI.
Axonal damage associated with DTBI
In addition to diffuse neuronal perturbation and
death described above, diffuse axonal injury is also
a distinguishing feature of DTBI, occurring across
the spectrum of brain injury ranging from mild
through severe. Historically, the histological identi-
fication of diffuse axonal injury was based upon
the use of silver salts to detect, within the first days
of injury, grossly swollen axonal bulbs that lacked
continuity with their downstream axonal partners
(Strich, 1956). This suggested their mechanical
transsection resulting in axonal retraction and
axoplasmic pooling at the site of disconnection
(Strich, 1956; Adams, 1982). More contempo-
rary studies, however, have shown that in large
part, this premise of transsection and retraction
50

Fig. 1. This figure illustrates the phenomenon of mechanically induced neuronal membrane disruption and its consequences for the
neuron. Via double labeled confocal images, panels A–C show neurons flooding with both dextrans with evidence of concomitant
cellular injury, reflected in their irregular, distorted profiles and vacuolization (arrows). In those cells showing the most severe damage,
note that the dextrans are also typically found within the nucleus (arrowhead). Note that other double-labeled neurons (D) dem-
onstrate little or no pathological damage and that despite homogenous tracer uptake no nuclear accumulation or vacuolization occurs
(arrows). Scale bars: 100 mm. Panel E illustrates three tracer flooded neurons, confirmed by routine fluorescent microscopy and
followed via EM. The most severely damaged neuron (asterisk) demonstrates increased electron density, organelle vacuolization
(arrows) and perisomatic glial ensheathment (arrowheads). The two other cells (double and triple asterisks) demonstrate little or no
pathological change. Note that the surrounding neuropil demonstrates little overt pathologic change consistent with the confocal
observations. Scale bar: 5 mm.

is not correct. Rather, it has been shown in mul- impairment of axonal transport, with progressive
tiple animal studies, as well as limited human in- local axonal swelling followed by detachment over
vestigations, that the forces of injury diffusely alter a post-traumatic course ranging from several hours
focal axonal segments. This results in a local up to a day (Povlishock and Jenkins, 1995). Given
 51

Fig. 2. This figure illustrates the finding that although mechanoporation can occur, some neurons reseal their perturbated membranes
with increased post-traumatic survival. Here confocal images of pre-injury infused dextrans (A), post-injury-infused dextrans (B) and
their overlay (C) demonstrate some cortical neurons flooding with the pre-injury infused dextran alone without concomitant flooding
with the post-injury administrated tracer (arrows), all of which is suggestive of cell membrane closure/recovery. Note that one double-
flooded neuron demonstrating severe damage (arrowhead) and one double-labeled neuron demonstrating less severe pathology (double
arrowhead) are also shown. Scale bar: 50 mm. Panel D illustrates a routine fluorescent image that reveals a single labeled cell. In panel
E, the same neuron is visualized through the use of antibodies to the fluorophore and then carried to the EM level (F–G). Note that
neurons flooding with the pre-injury dextran alone do not show overt pathological damage. Immunoreactive products (anti-Alexa
Fluor IR) are labeled with arrows in panel G, which is an enlargement of the area, blocked out in panel F. Scale bar 2 mm. (Adapted
with permission from Farkas et al. (2006); Copyright 2006 by the Society for Neuroscience).

the fact that these reactive axonal changes were
found in scattered axons related to other intact ax-
ons and their vascular elements, this precluded the
potential for direct mechanical renting, suggesting
that more subtle intraaxonal changes were at work
in the pathogenesis of this progressive axonal
change leading to disconnection.
Cellular and subcellular factors related to the
initiating pathogenesis of DTBI-associated axonal
injury
Appreciating that the above described cascades of
axonal perturbation leading to impaired transport
and disconnection evolved over a relatively long
52

Fig. 3. In contrast to the phenomenon of resealing shown in Fig. 2, this figure provides evidence of delayed neuronal plasmalemmal
opening. Here confocal images of pre-injury infused dextran (A), post-injury-infused dextran (B) and their overlay (C) demonstrate
scattered neurons flooding with the post-injury infused dextran alone (arrows), among with double-flooded neurons (arrowheads).
Scale bar: 100 mm. (Adapted with permission from Farkas et al. (2006); Copyright 2006 by the Society for Neuroscience).

post-traumatic course, emphasis has been placed
upon identifying the initiating intraaxonal cellular
and subcellular factors both to understand better
the pathobiology of this axonal injury and to
develop therapies targeting these cellular and sub-
cellular changes. While immediate physical trans-
section of the axon cylinder has been ruled out,
the potential for focal disturbances in the axo-
lemma leading to local ionic dysregulation was
evaluated in the experimental setting using extra-
cellular tracers normally excluded by the intact/un-
altered axolemma. Through this approach, our
laboratory showed that discreet axonal foci scat-
tered throughout the injured brain revealed evi-
dence of altered focal axolemmal permeability to
various normally excluded extracellular tracers
(Pettus and Povlishock, 1996; Povlishock and Pet-
tus, 1996). Moving on the premise that this alter-
ation in axolemmal permeability would be
accompanied by local calcium dysregulation, we
probed the same axonal segments with antibodies
targeting calcium-activated, CMSP. In these stud-
ies, CMSP was observed initially in the subaxo-
lemmal domain followed by its activation in the
axon’s core, with a particular predilection for the
mitochondria that appeared swollen with disrupted
cristae (Buki et al., 1999). Because such mi-
tochondrial damage appeared consistent with
local calcium overloading and caspase activation,
our laboratory probed these same segments with
antibodies targeting cytochrome C and caspase-
mediated spectrin proteolysis (Buki et al., 2000).
Via this approach, we demonstrated that the dam-
aged mitochondria released cytochrome C that, in
turn, activated caspase-mediated spectrin degrada-
tion (Buki et al., 2000; Buki and Povlishock, 2006).
Collectively, both cysteine proteases, calpain and
caspase were recognized to participate in the deg-
radation of the axonal cytoskeleton associated
with concomitant neurofilament side-arm cleavage,
neurofilament compaction and microtubular loss
(Buki and Povlishock, 2006). Because of these local
cytoskeletal abnormalities, these compacted axonal
segments could also be routinely identified by use
of antibodies targeting altered neurofilament sub-
units (RMO14) that, in our hands, provided a sur-
rogate marker for identifying such sites of injury
(Povlishock et al., 1997). Based on these findings in
our lab, as well as others, it was assumed that these
intraaxonal changes universally led to an upstream
impairment of axonal transport leading to swelling
and disconnection (Maxwell et al., 1997). However,
continued investigation failed to establish a routine
correlation between these two axonal events, lead-
ing us to question whether neurofilament compact-
ion and axonal swelling, consistently occurred
within the same axonal segment. More recent stud-
ies, using multiple strategies to qualitatively and

Fig. 4. This figure provides a semiquantitative, graphic repre-
sentation of those different forms and prognosis of membrane
perturbation following DTBI. Here the distribution of neurons
with resealed (gray column), enduring (white column) or delayed
membrane perturbation (black column) are shown after DTBI-
induced membrane perturbation. As membranes remain closed,
open or reseal in response to injury, tracer availability deter-
mines the type of membrane perturbation assigned categorically
to each neuron. The pre-injury administration of Alexa Fluor
488 and post-injury administration of Texas Red conjugated
dextrans permit the evaluation of neuronal membrane pertur-
bation distribution at 4 and 8 h after injury. Across time points,
the proportion of neurons with delayed membrane perturbation
is significantly different from the proportion of resealed neurons
and the proportion of enduring membrane perturbation (w2,
po0.016). The change in the proportion of neurons with re-
sealed and enduring membrane perturbation over time was not
significant (w2, p ¼ 0.19). The redistribution of the types of
membrane perturbation between 4 and 8 h post-injury may result
from resealed neurons reopening (gray arrow) to become neu-
rons with enduring damage and/or additional neurons (black
arrow) suffering delayed membrane perturbation.
quantitatively assess axonal numbers as well as the
spatial relationship of axons showing neurofila-
ment compaction and impaired transport have now
convincingly shown that the above described sites
of altered axonal permeability, cysteine protease
activation and cytoskeletal collapse do not rou-
tinely correlate with sites of impaired axonal trans-
port (Stone et al., 2004; Marmarou et al., 2005).
This suggested that the neurofilament-compacted
axonal segments and swollen axonal segments dem-
onstrating impaired axonal transport most likely
represent two different populations of injured ax-
ons responding to the traumatic episode in different
fashions. For those axonal segments showing
53
focally altered axolemmal permeability it appears,
as noted previously, that local calcium dysregula-
tion with the activation of the cysteine proteases,
causes local degradation of the axonal cytoskele-
ton, leading to axonal failure and disconnection
(Stone et al., 2004; Marmarou et al., 2005). Why
these sites of axonal injury do not reveal impaired
axonal transport and axonal swelling is unclear,
yet, it is conceivable that the suprathreshold
calcium uptake occurring at these sites most likely
converts anterograde to retrograde transport,
thereby precluding the development of reactive
axonal swelling (Sahenk and Lasek, 1988; Martz
et al., 1989). The validity of the above is supported
by the use of various therapeutic strategies target-
ing calpain inhibition that, as such, significantly
reduce the numbers of axonal profiles showing the
above described cysteine protease activation and
cytoskeletal collapse (Buki et al., 2003; Buki and
Povlishock, 2006). In contrast, it appears that those
axons showing impaired axonal transport and local
swelling do not sustain any alteration in local
axolemmal permeability or any activation of the
cysteine proteases (Povlishock and Stone, 2001).
Rather, it is posited that other mechanisms are at
work and that these are linked to more subtle
forms of calcium dysregulation. These potentially
involve the activation of micromolar calpains to
trigger the activation of calcineurin that, in turn,
alters the microtubular network to disrupt local
axonal transport kinetics and thereby elicit the
swellings described above (Povlishock and Stone,
2001). Although limited direct evidence exists to
support this pathway, the use of calcineurin
antagonists, such as FK506, directly attenuate the
numbers of axons showing impaired axonal
transport and swelling while having no effect upon
those axons showing neurofilament compaction
and disconnection (Marmarou and Povlishock,
2006). This supports the premise that calcineurin
is integral to the pathogenesis of impaired axonal
transport and swelling. Collectively, these studies
illustrate the complexity of the pathogenesis
of diffuse axonal injury, suggesting at least two
differing types of initiating mechanisms, with the
caveat that both populations of injured axons will
not likely be amenable to one form of therapeutic
intervention.
54
 55

Other potential cellular and subcellular factors
related to diffuse axonal injury
In the above passages, we have focused upon the
contemporary appreciation of the cellular and
subcellular changes involved in diffuse axonal in-
jury. Importantly, all these changes were believed
to be ongoing only in myelinated nerve fiber popu-
lations with no involvement of unmyelinated
axons that, in general, have received virtually no
consideration either in the context of TBI (Jafari
et al., 1998) or, for that matter, any other CNS
disorder. Recently, however, this perception was
changed by work conducted by Reeves et al. (2005)
who, through the use of electrophysiological meth-
ods, provided compelling evidence for unmyelin-
ated nerve fiber damage and dysfunction within
the corpus callosum of traumatically brain-injured
animals. Using an analysis of compound action
potentials by examining two specific wave forms
that can be independently related to either myelin-
ated or unmyelinated axons populations, Reeves
and colleagues showed significant and sustained
depression of the compound action potentials
associated with the unmyelinated axon population.
These electrophysiological studies were accompa-
nied by routine morphological analysis using elec-
tron microscopy. Although these studies were
preliminary, they suggested that the morphological
progression of unmyelinated fiber change observed
was quite dissimilar from that described above for
myelinated axons. This potential difference is also
partially supported by excellent in vitro studies that
have examined non-myelinated neurites subjected
to mechanical strains that ultimately led to local
axonal beading and disconnection (Wolf et al.,
2001; Iwata et al., 2004). In this pathology, how-
ever, no evidence of overt axolemmal change or
axolemmal disruption was discerned. Rather,
the depolarization associated with this injury
evoked sodium influx, the activation of voltage
gated calcium channels and the concomitant
activation of sodium/calcium exchangers, all of
which contributed to local intraaxonal calcium
overloading (Wolf et al., 2001). These calcium-me-
diated changes, comparable with some of the
changes described in myelinated axons, were linked
with the activation of proteases. These, in turn,
contributed to subsequent proteolysis of its NaCh
subunit to promote a persistent elevation in
intracellular calcium, fueling additional pathologi-
cal changes through many of the pathways ad-
dressed above (Iwata et al., 2004). While these
changes obviously remain to be confirmed in vivo,
they are intriguing and speak to yet another, differ-
ent form of cellular and subcellular change that,
in this case, involves a potential channelopathy as
major player in the ensuing unmyelinated axonal
perturbation.
Concluding comments
In this brief review, we have attempted to critically
evaluate current thought on the cellular and sub-
cellular change evoked by DTBI at both the neu-
ronal somatic and axonal fronts. Recognizing that
our appreciation of cellular and subcellular change
in the context of TBI has been framed primarily by
our understanding of the pathobiology of contu-
sional injury, this review cautions against moving
on the assumption that these contusional cellular
and subcellular cascades must transfer to those
neuronal cell bodies and their appendages that
sustain diffuse injury. Basic differences between
focal and diffuse injury give rise to the possibility

Fig. 5. This figure illustrates the relation between CMSP and tracer uptake. Panel A shows triple labeled confocal image demon-
strating dextran flooded neurons as well as CMSP immunopositive neurons. Note that at both 4 and 8 h post-injury CMSP immuno-
reactive neurons were observed throughout the neocortex (arrows). Scale bar: 100 mm. In panels B–E confocal images of pre-injury
dextran flooding (B), post-injury dextran flooding (C), CMSP immunopositivity (D) and their overlay (E) demonstrate neurons
showing enduring membrane disruption reflected in their content of both tracers. Note that some neurons colocalize with CMSP-
immunopositivity (arrow), whereas other neurons demonstrate tracer flooding without CMSP (arrowheads). Also note that at these
same time points, several CMSP immunoreactive neurons also could be identified without concomitant tracer flooding (double
arrowhead). Lastly note that in the same region, a neuron demonstrating only the initial tracer flooding (big arrow), as well as a neuron
revealing only secondary tracer flooding (triple arrowhead) can also be seen. Scale bar: 50 mm. (Adapted with permission from Farkas
et al. (2006); Copyright 2006 by the Society for Neuroscience).
56
for the involvement of other cell perturbation/
death cascades that may not be operant with focal
injury. Specifically, the pathogenesis of diffuse
injury highlights the potential for mechanically
induced membrane perturbation to serve as a more
significant player in the ensuing cellular and sub-
cellular change. The potential for transient versus
more enduring mechanically induced membrane
perturbation may also influence subsequent neuron-
al somatic and axonal perturbation and/or death.
Additionally, the activation of other cell death
cascades including, but not limited to, activation
of lysosomal-mediated change through the release
of hydrolytic enzymes must now be considered.
Collectively, this review illustrates the complexity
of the pathobiology of DTBI and the danger
in assuming that a singular mechanistic view of
cellular and subcellular change can translate to the
diffusely injured brain.
Abbreviations
CMSP calpain-mediated spectrin proteolysis
DTBI diffuse traumatic brain injury
SBDP spectrin breakdown product
TBI traumatic brain injury
Acknowledgments
This work was supported in part by NIH grants
NS045824, NS047463 and HD055813.
References
Adams, J.H. (1982) Diffuse axonal injury in non-missile head
injury. Injury, 13(5): 444–445.
Adams, J.H., Graham, D.I., Scott, G., Parker, L.S. and Doyle,
D. (1980) Brain damage in fatal non-missile head injury.
J. Clin. Pathol., 33(12): 1132–1145.
Beer, R., Franz, G., Srinivasan, A., Hayes, R.L., Pike, B.R.,
Newcomb, J.K., Zhao, X., Schmutzhard, E., Poewe, W. and
Kampfl, A. (2000) Temporal profile and cell subtype distri-
bution of activated caspase-3 following experimental trau-
matic brain injury. J. Neurochem., 75(3): 1264–1273.
Brana, C., Benham, C.D. and Sundstrom, L.E. (1999) Calpain
activation and inhibition in organotypic rat hippocampal
slice cultures deprived of oxygen and glucose. Eur. J. Neurosci.,
11(7): 2375–2384.
Braughler, J.M. and Hall, E.D. (1992) Involvement of lipid per-
oxidation in CNS injury. J. Neurotrauma, 9(Suppl 1): S1–S7.
Brunk, U.T., Dalen, H., Roberg, K. and Hellquist, H.B. (1997)
Photo-oxidative disruption of lysosomal membranes causes
apoptosis of cultured human fibroblasts. Free Radic. Biol.
Med., 23(4): 616–626.
Buki, A., Farkas, O., Doczi, T. and Povlishock, J.T. (2003) Pre-
injury administration of the calpain inhibitor MDL-28170 at-
tenuates traumatically induced axonal injury. J. Neurotrauma,
20(3): 261–268.
Buki, A., Okonkwo, D.O., Wang, K.K. and Povlishock, J.T.
(2000) Cytochrome c release and caspase activation in trau-
matic axonal injury. J. Neurosci., 20(8): 2825–2834.
Buki, A. and Povlishock, J.T. (2006) All roads lead to discon-
nection? Traumatic axonal injury revisited. Acta Neurochir.
(Wien.), 148(2): 181–193.
Buki, A., Siman, R., Trojanowski, J.Q. and Povlishock, J.T.
(1999) The role of calpain-mediated spectrin proteolysis in
traumatically induced axonal injury. J. Neuropathol. Exp.
Neurol., 58(4): 365–375.
Cernak, I., Chapman, S., Hamlin, G. and Vink, R. (2002)
Temporal characterisation of pro- and anti-apoptotic mech-
anisms following diffuse traumatic brain injury in rats.
J. Clin. Neurosci., 9(5): 565–572.
Cernak, I., Vink, R., Zapple, D.N., Cruz, M.I., Ahmed, F.,
Chang, T., Fricke, S.T. and Faden, A.I. (2004) The patho-
biology of moderate diffuse traumatic brain injury as iden-
tified using a new experimental model of injury in rats.
Neurobiol. Dis., 17(1): 29–43.
Cervos-Navarro, J. and Lafuente, J.V. (1991) Traumatic brain
injuries: structural changes. J. Neurol. Sci., 103(Suppl):
S3–S14.
Clark, R.S., Kochanek, P.M., Chen, M., Watkins, S.C.,
Marion, D.W., Chen, J., Hamilton, R.L., Loeffert, J.E. and
Graham, S.H. (1999) Increases in Bcl-2 and cleavage of
caspase-1 and caspase-3 in human brain after head injury.
FASEB J., 13(8): 813–821.
Clark, R.S., Kochanek, P.M., Watkins, S.C., Chen, M., Dixon,
C.E., Seidberg, N.A., Melick, J., Loeffert, J.E., Nathaniel,
P.D., Jin, K.L. and Graham, S.H. (2000) Caspase-3 medi-
ated neuronal death after traumatic brain injury in rats.
J. Neurochem., 74(2): 740–753.
Colicos, M.A. and Dash, P.K. (1996) Apoptotic morphology of
dentate gyrus granule cells following experimental cortical
impact injury in rats: possible role in spatial memory deficits.
Brain Res., 739(1–2): 120–131.
Colicos, M.A., Dixon, C.E. and Dash, P.K. (1996) Delayed,
selective neuronal death following experimental cortical im-
pact injury in rats: possible role in memory deficits. Brain
Res., 739(1–2): 111–119.
Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh,
T.K. (1998) Experimental brain injury induces regionally
distinct apoptosis during the acute and delayed post-traumatic
period. J. Neurosci., 18(15): 5663–5672.

Cortez, S.C., McIntosh, T.K. and Noble, L.J. (1989) Experi-
mental fluid percussion brain injury: vascular disruption and
neuronal and glial alterations. Brain Res., 482(2): 271–282.
Dietrich, W.D., Alonso, O., Busto, R., Globus, M.Y. and
Ginsberg, M.D. (1994a) Post-traumatic brain hypothermia
reduces histopathological damage following concussive brain
injury in the rat. Acta Neuropathol. (Berl.), 87(3): 250–258.
Dietrich, W.D., Alonso, O. and Halley, M. (1994b) Early
microvascular and neuronal consequences of traumatic
brain injury: a light and electron microscopic study in rats.
J. Neurotrauma, 11(3): 289–301.
Eldadah, B.A. and Faden, A.I. (2000) Caspase pathways,
neuronal apoptosis, and CNS injury. J. Neurotrauma, 17(10):
811–829.
Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989)
The role of excitatory amino acids and NMDA receptors in
traumatic brain injury. Science, 244(4906): 798–800.
Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanop-
oration induced by diffuse traumatic brain injury: an irre-
versible or reversible response to injury? J. Neurosci., 26(12):
3130–3140.
Farkas, O., Polgar, B., Szekeres-Bartho, J., Doczi, T., Povlishock,
J.T. and Buki, A. (2005) Spectrin breakdown products in the
cerebrospinal fluid in severe head injury: preliminary observa-
tions. Acta Neurochir. (Wien.), 147(8): 855–861.
Fei, Z., Zhang, X., Jiang, X.F., Huang, W.D. and Bai, H.M.
(2005) Altered expression patterns of metabotropic glutamate
receptors in diffuse brain injury. Neurosci. Lett., 380(3):
280–283.
Felderhoff-Mueser, U., Sifringer, M., Pesditschek, S., Kuckuck,
H., Moysich, A., Bittigau, P. and Ikonomidou, C. (2002)
Pathways leading to apoptotic neurodegeneration following
trauma to the developing rat brain. Neurobiol. Dis., 11(2):
231–245.
Formigli, L., Papucci, L., Tani, A., Schiavone, N., Tempestini,
A., Orlandini, G.E., Capaccioli, S. and Orlandini, S.Z. (2000)
Aponecrosis: morphological and biochemical exploration of
a syncretic process of cell death sharing apoptosis and
necrosis. J. Cell Physiol., 182(1): 41–49.
Fox, G.B., Fan, L., LeVasseur, R.A. and Faden, A.I. (1998)
Sustained sensory/motor and cognitive deficits with neuronal
apoptosis following controlled cortical impact brain injury in
the mouse. J. Neurotrauma, 15(8): 599–614.
Geddes, D.M., Cargill, R.S. and LaPlaca, M.C. (2003)
Mechanical stretch to neurons results in a strain rate and
magnitude-dependent increase in plasma membrane perme-
ability. J. Neurotrauma, 20(10): 1039–1049.
Goda, M., Isono, M., Fujiki, M. and Kobayashi, H. (2002) Both
MK801 and NBQX reduce the neuronal damage after impact-
acceleration brain injury. J. Neurotrauma, 19(11): 1445–1456.
Hicks, R., Soares, H., Smith, D. and McIntosh, T. (1996)
Temporal and spatial characterization of neuronal injury
following lateral fluid-percussion brain injury in the rat. Acta
Neuropathol. (Berl.), 91(3): 236–246.
Huang, Y. and Wang, K.K. (2001) The calpain family and
human disease. Trends Mol. Med., 7(8): 355–362.
57
Iwata, A., Stys, P.K., Wolf, J.A., Chen, X.H., Taylor, A.G.,
Meaney, D.F. and Smith, D.H. (2004) Traumatic axonal in-
jury induces proteolytic cleavage of the voltage-gated sodium
channels modulated by tetrodotoxin and protease inhibitors.
J. Neurosci., 24(19): 4605–4613.
Jafari, S.S., Nielson, M., Graham, D.I. and Maxwell, W.L.
(1998) Axonal cytoskeletal changes after nondisruptive
axonal injury. II. Intermediate sized axons. J. Neurotrauma,
15(11): 955–966.
Kampfl, A., Posmantur, R., Nixon, R., Grynspan, F., Zhao, X.,
Liu, S.J., Newcomb, J.K., Clifton, G.L. and Hayes, R.L.
(1996) mu-Calpain activation and calpain-mediated cytoskele-
tal proteolysis following traumatic brain injury. J. Neurochem.,
67(4): 1575–1583.
Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A.
(1990) Massive increases in extracellular potassium and the
indiscriminate release of glutamate following concussive
brain injury. J. Neurosurg., 73(6): 889–900.
Keane, R.W., Kraydieh, S., Lotocki, G., Alonso, O.F., Aldana,
P. and Dietrich, W.D. (2001) Apoptotic and antiapoptotic
mechanisms after traumatic brain injury. J. Cereb. Blood
Flow Metab., 21(10): 1189–1198.
Kermer, P., Klocker, N. and Bahr, M. (1999) Neuronal death
after brain injury. Models, mechanisms, and therapeutic
strategies in vivo. Cell Tissue Res., 298(3): 383–395.
Kitanaka, C. and Kuchino, Y. (1999) Caspase-independent
programmed cell death with necrotic morphology. Cell
Death Differ., 6(6): 508–515.
Kontos, H.A. and Povlishock, J.T. (1986) Oxygen radicals in
brain injury. Cent. Nerv. Syst. Trauma, 3(4): 257–263.
Kotapka, M.J., Gennarelli, T.A., Graham, D.I., Adams, J.H.,
Thibault, L.E., Ross, D.T. and Ford, I. (1991) Selective
vulnerability of hippocampal neurons in acceleration-induced
experimental head injury. J. Neurotrauma, 8(4): 247–258.
Kotapka, M.J., Graham, D.I., Adams, J.H., Doyle, D. and
Gennarelli, T.A. (1993) Hippocampal damage in fatal pae-
diatric head injury. Neuropathol. Appl. Neurobiol., 19(2):
128–133.
Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli,
T.A. (1992) Hippocampal pathology in fatal non-missile human
head injury. Acta Neuropathol. (Berl.), 83(5): 530–534.
Kroemer, G. and Jaattela, M. (2005) Lysosomes and autophagy
in cell death control. Nat. Rev. Cancer, 5(11): 886–897.
Kupina, N.C., Detloff, M.R., Dutta, S. and Hall, E.D. (2002)
Neuroimmunophilin ligand V-10,367 is neuroprotective after
24-hour delayed administration in a mouse model of diffuse
traumatic brain injury. J. Cereb. Blood Flow Metab., 22(10):
1212–1221.
Kupina, N.C., Nath, R., Bernath, E.E., Inoue, J., Mitsuyoshi,
A., Yuen, P.W., Wang, K.K. and Hall, E.D. (2001) The novel
calpain inhibitor SJA6017 improves functional outcome after
delayed administration in a mouse model of diffuse brain
injury. J. Neurotrauma, 18(11): 1229–1240.
Lin, X., Zhi, D. and Zhang, S. (2001) Inhibiting effect of mod-
erate hypothermia on cell apoptosis after diffuse brain injury
in rats. Chin. J. Traumatol., 4(1): 14–19.
58
Liu, X. and Schnellmann, R.G. (2003) Calpain mediates
progressive plasma membrane permeability and proteolysis
of cytoskeleton-associated paxillin, talin, and vinculin during
renal cell death. J. Pharmacol. Exp. Ther., 304(1): 63–70.
Liu, X., Van Vleet, T. and Schnellmann, R.G. (2004) The role
of calpain in oncotic cell death. Annu. Rev. Pharmacol.
Toxicol., 44: 349–370.
Marmarou, C.R. and Povlishock, J.T. (2006) Administration
of the immunophilin ligand FK506 differentially attenuates
neurofilament compaction and impaired axonal transport in
injured axons following diffuse traumatic brain injury. Exp.
Neurol., 197(2): 353–362.
Marmarou, C.R., Walker, S.A., Davis, C.L. and Povlishock,
J.T. (2005) Quantitative analysis of the relationship between
intra-axonal neurofilament compaction and impaired axonal
transport following diffuse traumatic brain injury. J. Neuro-
trauma, 22(10): 1066–1080.
Martz, D., Garner, J. and Lasek, R.J. (1989) Protein changes
during anterograde-to-retrograde conversion of axonally
transported vesicles. Brain Res., 476(1): 199–203.
Maxwell, W.L., Povlishock, J.T. and Graham, D.L. (1997) A
mechanistic analysis of nondisruptive axonal injury: a review.
J. Neurotrauma, 14(7): 419–440.
McCracken, E., Hunter, A.J., Patel, S., Graham, D.I. and
Dewar, D. (1999) Calpain activation and cytoskeletal protein
breakdown in the corpus callosum of head-injured patients.
J. Neurotrauma, 16(9): 749–761.
Merry, D.E. and Korsmeyer, S.J. (1997) Bcl-2 gene family in
the nervous system. Annu. Rev. Neurosci., 20: 245–267.
Newcomb, J.K., Zhao, X., Pike, B.R. and Hayes, R.L. (1999)
Temporal profile of apoptotic-like changes in neurons and
astrocytes following controlled cortical impact injury in the
rat. Exp. Neurol., 158(1): 76–88.
Nicotera, P., Leist, M. and Manzo, L. (1999) Neuronal cell
death: a demise with different shapes. Trends Pharmacol.
Sci., 20(2): 46–51.
Palmer, A.M., Marion, D.W., Botscheller, M.L., Swedlow,
P.E., Styren, S.D. and DeKosky, S.T. (1993) Traumatic brain
injury-induced excitotoxicity assessed in a controlled cortical
impact model. J. Neurochem., 61(6): 2015–2024.
Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a
distinct set of intra-axonal ultrastructural changes associated
with traumatically induced alteration in axolemmal perme-
ability. Brain Res., 722(1–2): 1–11.
Pike, B.R., Flint, J., Dutta, S., Johnson, E., Wang, K.K.
and Hayes, R.L. (2001) Accumulation of non-erythroid
alpha II-spectrin and calpain-cleaved alpha II-spectrin break-
down products in cerebrospinal fluid after traumatic brain
injury in rats. J. Neurochem., 78(6): 1297–1306.
Pike, B.R., Zhao, X., Newcomb, J.K., Glenn, C.C., Anderson,
D.K. and Hayes, R.L. (2000) Stretch injury causes calpain
and caspase-3 activation and necrotic and apoptotic cell
death in septo-hippocampal cell cultures. J. Neurotrauma,
17(4): 283–298.
Pike, B.R., Zhao, X., Newcomb, J.K., Posmantur, R.M.,
Wang, K.K. and Hayes, R.L. (1998) Regional calpain and
caspase-3 proteolysis of alpha-spectrin after traumatic brain
injury. Neuroreport, 9(11): 2437–2442.
Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers
of proteolytic damage following traumatic brain injury. Brain
Pathol., 14(2): 202–209.
Povlishock, J.T. and Jenkins, L.W. (1995) Are the pathobio-
logical changes evoked by traumatic brain injury immediate
and irreversible? Brain Pathol., 5(4): 415–426.
Povlishock, J.T. and Katz, D.I. (2005) Update of neuropa-
thology and neurological recovery after traumatic brain
injury. J. Head Trauma Rehabil., 20(1): 76–94.
Povlishock, J.T. and Kontos, H.A. (1992) The role of oxygen
radicals in the pathobiology of traumatic brain injury. Hum.
Cell, 5(4): 345–353.
Povlishock, J.T., Marmarou, A., McIntosh, T., Trojanowski,
J.Q. and Moroi, J. (1997) Impact acceleration injury in the
rat: evidence for focal axolemmal change and related neuro-
filament sidearm alteration. J. Neuropathol. Exp. Neurol.,
56(4): 347–359.
Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced
axonal damage: evidence for enduring changes in axolemmal
permeability with associated cytoskeletal change. Acta
Neurochir. Suppl. (Wien.), 66: 81–86.
Povlishock, J.T. and Stone, J.R. (2001) Traumatic axonal
injury. In: Miller L.P., Hayes R.L. and Newcomb J.K. (Eds.),
Head Trauma: Basic, Preclinical and Clinical Directions.
Wiley, pp. 281–302.
Prado, G.R., Ross, J.D., DeWeerth, S.P. and LaPlaca,
M.C. (2005) Mechanical trauma induces immediate chan-
ges in neuronal network activity. J. Neural Eng., 2(4):
148–158.
Prado, G., Caltagirone, C.L. and LaPlaca, M.C. (2004)
Neuronal plasma membrane disruption in the traumatically
injured brain. J. Neurotrauma, 21: 1331.
Prado, G. and LaPlaca, M.C. (2004) Neuronal plasma
membrane is transiently disrupted by mechanical trauma:
an insight into the mechanisms involved. J. Neurotrauma, 21:
1331.
Pravdenkova, S.V., Basnakian, A.G., James, S.J. and Andersen,
B.J. (1996) DNA fragmentation and nuclear endonuclease
activity in rat brain after severe closed head injury. Brain Res.,
729(2): 151–155.
Raghupathi, R. (2004) Cell death mechanisms following trau-
matic brain injury. Brain Pathol., 14(2): 215–222.
Raghupathi, R., Conti, A.C., Graham, D.I., Krajewski, S.,
Reed, J.C., Grady, M.S., Trojanowski, J.Q. and McIntosh,
T.K. (2002) Mild traumatic brain injury induces apoptotic
cell death in the cortex that is preceded by decreases in cellular
Bcl-2 immunoreactivity. Neuroscience, 110(4): 605–616.
Raghupathi, R., Graham, D.I. and McIntosh, T.K. (2000)
Apoptosis after traumatic brain injury. J. Neurotrauma,
17(10): 927–938.
Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005)
Myelinated and unmyelinated axons of the corpus callosum
differ in vulnerability and functional recovery following
traumatic brain injury. Exp. Neurol., 196(1): 126–137.

Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M.,
Neugebauer, E. and McIntosh, T.K. (1995) Evidence of
apoptotic cell death after experimental traumatic brain injury
in the rat. Am. J. Pathol., 147(6): 1575–1583.
Ross, D.T., Graham, D.I. and Adams, J.H. (1993) Selective
loss of neurons from the thalamic reticular nucleus following
severe human head injury. J. Neurotrauma, 10(2): 151–165.
Runnerstam, M., Bao, F., Huang, Y., Shi, J., Gutierrez, E.,
Hamberger, A., Hansson, H.A., Viano, D. and Haglid, K.
(2001) A new model for diffuse brain injury by rotational
acceleration: II. Effects on extracellular glutamate, intracra-
nial pressure, and neuronal apoptosis. J. Neurotrauma, 18(3):
259–273.
Saatman, K.E., Bozyczko-Coyne, D., Marcy, V., Siman, R. and
McIntosh, T.K. (1996) Prolonged calpain-mediated spectrin
breakdown occurs regionally following experimental brain
injury in the rat. J. Neuropathol. Exp. Neurol., 55(7):
850–860.
Sahenk, Z. and Lasek, R.J. (1988) Inhibition of proteolysis
blocks anterograde-retrograde conversion of axonally trans-
ported vesicles. Brain Res., 460(1): 199–203.
Singleton, R.H. and Povlishock, J.T. (2004) Identification and
characterization of heterogeneous neuronal injury and death
in regions of diffuse brain injury: evidence for multiple in-
dependent injury phenotypes. J. Neurosci., 24(14):
3543–3553.
Singleton, R.H., Zhu, J., Stone, J.R. and Povlishock, J.T. (2002)
Traumatically induced axotomy adjacent to the soma does not
result in acute neuronal death. J. Neurosci., 22(3): 791–802.
Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli,
T.A. and Meaney, D.F. (1997) Characterization of diffuse ax-
onal pathology and selective hippocampal damage following
inertial brain trauma in the pig. J. Neuropathol. Exp. Neurol.,
56(7): 822–834.
Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu,
L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired
axonal transport and altered axolemmal permeability occur
in distinct populations of damaged axons following traumatic
brain injury. Exp. Neurol., 190(1): 59–69.
Strich, S.J. (1956) Diffuse degeneration of the cerebral white
matter in severe dementia following head injury. J. Neurol.
Neurosurg. Psychiatry, 19: 163–185.
59
Van den Heuvel, C., Blumbergs, P.C., Finnie, J.W., Manavis,
J., Jones, N.R., Reilly, P.L. and Pereira, R.A. (1999) Upreg-
ulation of amyloid precursor protein messenger RNA in re-
sponse to traumatic brain injury: an ovine head impact
model. Exp. Neurol., 159(2): 441–450.
Van den Heuvel, C., Lewis, S., Wong, M., Manavis, J., Finnie,
J., Blumbergs, P.C., Jones, N. and Reilly, P. (1998) Diffuse
neuronal perikaryon amyloid precursor protein immunore-
activity in a focal head impact model. Acta Neurochir. Suppl.
(Wien.), 71: 209–211.
Wang, K.K. (2000) Calpain and caspase: can you tell the
difference? Trends Neurosci., 23(1): 20–26.
Wolf, J.A., Stys, P.K., Lusardi, T., Meaney, D. and Smith,
D.H. (2001) Traumatic axonal injury induces calcium
influx modulated by tetrodotoxin-sensitive sodium channels.
J. Neurosci., 21(6): 1923–1930.
Yakovlev, A.G. and Faden, A.I. (2004) Mechanisms of neural
cell death: implications for development of neuroprotective
treatment strategies. NeuroRx, 1(1): 5–16.
Yakovlev, A.G., Ota, K., Wang, G., Movsesyan, V., Bao, W.L.,
Yoshihara, K. and Faden, A.I. (2001) Differential expression
of apoptotic protease-activating factor-1 and caspase-3 genes
and susceptibility to apoptosis during brain development and
after traumatic brain injury. J. Neurosci., 21(19): 7439–7446.
Yamashima, T. (2000) Implication of cysteine proteases
calpain, cathepsin and caspase in ischemic neuronal death
of primates. Prog. Neurobiol., 62(3): 273–295.
Yamashima, T. (2004) Ca2+-dependent proteases in ischemic
neuronal death: a conserved ‘calpain-cathepsin cascade’ from
nematodes to primates. Cell Calcium, 36(3–4): 285–293.
Yamashima, T., Tonchev, A.B., Tsukada, T., Saido, T.C.,
Imajoh-Ohmi, S., Momoi, T. and Kominami, E. (2003)
Sustained calpain activation associated with lysosomal rup-
ture executes necrosis of the postischemic CA1 neurons in
primates. Hippocampus, 13(7): 791–800.
Yang, J., Liu, X., Bhalla, K., Kim, C.N., Ibrado, A.M., Cai, J.,
Peng, T.I., Jones, D.P. and Wang, X. (1997) Prevention
of apoptosis by Bcl-2: release of cytochrome c from
mitochondria blocked. Science, 275(5303): 1129–1132.
Zipfel, G.J., Babcock, D.J., Lee, J.M. and Choi, D.W. (2000)
Neuronal apoptosis after CNS injury: the roles of glutamate
and calcium. J. Neurotrauma, 17(10): 857–869.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 5

Astroglia: Important mediators of traumatic brain

injury

Candace L. Floyd1, and Bruce G. Lyeth2

1
Department of Physical Medicine and Rehabilitation, Center for Glial Biology in Medicine, 547 Spain Rehabilitation
Center, University of Alabama at Birmingham, Birmingham, AL 35249, USA
2
Department of Neurological Surgery, University of California, 1515 Newton Court, One Shields Avenue, Davis,
CA 95816, USA

Abstract: Traumatic brain injury (TBI) research to date has focused almost exclusively on the patho-
physiology of injured neurons with very little attention paid to non-neuronal cells. However in the past
decade, exciting discoveries have challenged this century-old view of passive glial cells and have led to a
reinterpretation of the role of glial cells in central nervous system (CNS) biology and pathology. In this
chapter we review several lines of evidence, indicating that glial cells, particularly astrocytes, are active
partners to neurons in the brain, and summarize recent findings that detail the significance of astrocyte
pathology in traumatic brain injury.

Keywords: astrocyte; glutamate; sodium; calcium; acidosis; mechanical strain injury

Introduction the significance of astrocyte pathology in traumatic

brain injury.
Traumatic brain injury (TBI) research to date has
focused almost exclusively on the pathophysiology
Astrocytes in normal brain function: recasting an
of injured neurons with very little attention paid to
old ‘‘star’’
non-neuronal cells. This near-exclusive focus on
neuroprotection likely reflects the predominant
One line of evidence that astrocytes are more than
paradigm of the neuroscience community as a
supportive connective tissue comes from recent de-
whole, which characterized glial cells as a special-
tailed analyses of the cellular morphology. A
ized type of connective tissue that merely provided
unique morphological feature of astrocytes is that
support for the neurons. However in the past dec-
nearly the entire cell surface is covered with proc-
ade, exciting discoveries have challenged this
esses that extend and become lamellae or filopodia
century-old view of passive glial cells and have
as indicated by three-dimensional reconstruction of
led to a reinterpretation of the role of glial cells in
electron micrographs. Most lamellae and filopodia
central nervous system (CNS) biology and patho-
originate from processes, which, like the cell
logy. In this chapter we review several lines of
body, contain organelles and cytoskeletal elements;
evidence, indicating that glial cells, particularly
however, the cell surface extensions do not contain
astrocytes, are active partners to neurons in the
organelles and intermediate filaments themselves
brain, and summarize recent findings which detail
and are therefore not visualized with GFAP
immunohistochemistry (Chao et al., 2002). Thus,
Corresponding author. E-mail: clfloyd@uab.edu GFAP immunohistochemistry provides only a

DOI: 10.1016/S0079-6123(06)61005-4 61
62
limited view of the total three-dimensional area and
morphological complexity of a given astrocyte’s
domain. In proprotoplasmic astrocytes of the cor-
tex, cell surface extensions account for 50–60% of
the cytoplasmic volume of the cell (the remainder is
the cell body and processes) and cover
80% of
the cell surface. Structurally, astrocytes are highly
interdigitated such that neighboring processes often
correspond to different astrocytic cell bodies. Also
astrocyte processes frequently surround neuronal
synapses; for example in the rat neocortex, 56% of
all synapses are contacted by astrocytes (Chao
et al., 2002). Large variations in the structural
characteristics of glia-synaptic contacts exist such
that common morphologies run the gambit from
partial to entire encasement of a single axo-
dendritic synapse or a triadic/serial synapse to
complete coverage of complex synapses including
centro-dendritic, centro-axonic, and multicenteric
glomeruli synapses (Chao et al., 2002). The exten-
sive astrocyte–synapse connections have led to the
hypothesis that many synapses are actually of a
tripartite nature, comprising pre- and post-synaptic
neuronal elements and a third glial component
(Haydon, 2001; Chao et al., 2002).
A second line of evidence for the active role of
astrocytes in brain function is that astrocytes ex-
press diverse neurotransmitter receptors that regu-
late the concentration of intercellular calcium
([Ca2+]i). Astrocytes express receptors for gluta-
mate (Glaum et al., 1990; Holzwarth et al., 1994),
acetylcholine (Kondou et al., 1994; Shao and
McCarthy, 1995), ATP (Neary and Zhu, 1994),
and histamine (Peakman and Hill, 1995). Addi-
tionally, astrocytes express voltage-gated Ca2+
channels similar to those in neurons (L- and
N-type) and Ca2+ entry through these channels
alters [Ca2+]i (MacVicar et al., 1991; D’Ascenzo
et al., 2004). Activation of voltage-gated Ca2+
channels requires depolarization that has been
shown in cultured astrocytes, organotypic cultures,
and acutely isolated astrocytes following
local changes in extracellular K+ concentration
(Verkhratsky et al., 1998). Importantly, astrocytes
respond to neuronal transmitter release by gener-
ation of calcium waves (Dani et al., 1992; Porter
and McCarthy, 1996). Thus astrocytes have intact
cyto-machinery to sense and respond to the
extracellular environment in ways that were tradi-
tionally the domain of neurons, the ‘‘excitable
cells’’ of the CNS.
Not only do astrocytes have the cyto-machinery
of excitable cells, but also they use it. Astrocyte
excitability comprises activation by external or in-
ternal signals followed by subsequent transfer of a
specific message to nearby cells and has been
termed ‘‘gliotransmission’’ (Bezzi and Volterra,
2001). The foundation of astrocyte excitability is
calcium signaling, both intracellular and intercel-
lular; and since astrocytes do not generate action
potentials, imaging of calcium transients and os-
cillations has been more effective than traditional
electrophysiology techniques in studying astrocyte
excitation (Volterra and Meldolesi, 2005). Two
well-documented forms of astrocyte excitation are
(a) neuron-dependent excitation and (b) neuron-
independent excitation. Neuron-dependent excita-
tion, mainly through elevations in transmitter
outside of the synaptic cleft, was originally de-
scribed for glutamate transmission in hippocam-
pus and cerebellum but now includes multiple
brain circuits as well as multiple receptors systems
including GABA, acetylcholine, dopamine, ATP,
BDNF, etc. (Haydon, 2001). Neuron-independent,
or spontaneous, excitation of astrocytes occurs
mainly during development, but has also been
demonstrated in adult preparations (Aguado et al.,
2002) and may occur more frequently under path-
ological conditions such as CNS injury as dis-
cussed later in this chapter. Neuron-independent
calcium oscillations principally involve release of
calcium from intracellular stores, but calcium en-
try via voltage-gate channels may also play a role
(Parri et al., 2001; Nett et al., 2002; Parri and
Crunelli, 2003). Spontaneous excitation of as-
trocytes results in subsequent activation of neigh-
boring cells, including astrocytes and neurons,
which indicates that both cell types are sources of
excitation and communication networks may be
far more complex than an astrocytic response to
neuronal excitation (Volterra et al., 2005).
The cellular mechanisms of intercellular calcium
signaling in astrocytes remain an area of active in-
quiry. Comprehensive reviews detailing findings
leading to the current understanding of the mech-
anisms of astrocyte intercellular calcium waves

have recently been published (Haydon, 2001;
Volterra et al., 2005), but the most prevalent work-
ing hypotheses for these mechanisms are briefly de-
scribed here. Intercellular calcium signaling in as-
trocyte comprises a calcium wave that is essentially
a sequence of increases in intracellular calcium that
spreads, like a wave, throughout a group of as-
trocytes. Astrocyte calcium waves were initially
characterized in response to glutamate application
or mechanical stimulation to single astrocytes in
cell culture (Cornell-Bell et al., 1990; Cornell-Bell
and Finkbeiner, 1991), and these findings have
been corroborated in organotypic hippocampal
slice cultures (Harris-White et al., 1998), Müller
glia of acutely isolated retina (Newman and Zahs,
1998), and more recently in vivo (Nimmerjahn
et al., 2004). In astrocytes, a ligand (such as gluta-
mate) binds to a G-protein receptor coupled to
phosphoinositide-specific phospholipase C (PLC)
which cleaves phosphatidylinositol 4,5 bisphos-
phate (PIP2) yielding the second messenger
inositol-1,4,5 trisphosphate (IP3). IP3 then binds
to the IP3 receptor (IP3R) located on intracellular
calcium stores, causing increased intracellular cal-
cium. Elevated intracellular calcium in a single cell
can translate into an intercellular calcium wave by
at least two known mechanisms, working inde-
pendently or in concert. First, an intercellular cal-
cium wave can spread between astrocytes via gap-
junction mediated metabolic coupling. In this
model, IP3 diffuses between gap junctions and
stimulates the release of intracellular calcium from
adjacent astrocytes resulting in a wave of calcium.
Additionally, an extracellular component may also
be involved since studies show that calcium waves
can travel between cells separated by a cell-free
zone (Hassinger et al., 1996; Newman and Zahs,
1997) and that waves follow the direction of extra-
cellular perfusion (Hassinger et al., 1996). Since
waves of extracellular ATP can accompany inter-
cellular calcium waves, the most likely candidate
for the extracellular message for calcium wave
propagation is ATP (Cotrina et al., 1998b, 2000;
Guthrie et al., 1999). In this second model, as-
trocytes release ATP that then activates the
purinergic receptor P2Y1, thereby increasing IP3
and producing a calcium signal in nearby cells.
These two pathways may work together to produce
63
both shorter-range (gap-junction mediated) and
longer-range (ATP-mediated) calcium waves (for
review, see Haydon, 2001).
One functional consequence of astrocyte cal-
cium signaling is glutamate release (Parpura et al.,
1994; Bezzi et al., 1998). Using live-cell fluorescent
imaging and cell culture, glutamate release from
astrocyte was visualized by applying the cofactors
NAD+ and the enzyme glutamate dehydrogenase
in the bath so that any released glutamate is con-
verted to a-ketogluatarate and the NAD+ is re-
duced to the fluorescent NADH. Combination of
this imaging technique with live-cell calcium im-
aging demonstrated that a wave of extracellular
glutamate accompanies an intercellular calcium
wave (Innocenti et al., 2000). Subsequent charac-
terization of glutamate release from astrocytes has
shown that this release is not only related to in-
tracellular calcium increases, but is actually cal-
cium dependent and is very likely regulated by
exocytosis (Parpura and Haydon, 2000; Anlauf
and Derouiche, 2005). In contrast to neurons
where extracellular calcium is a main component
of calcium-dependent exocytosis, both the IP3 and
ryanodine-sensitive internal calcium stores seem to
be key components in calcium-dependent gluta-
mate release from astrocytes (Araque et al., 1999).
In further support of the hypothesis that astrocytes
exocytose glutamate, distinct vesicular compart-
ments containing glutamate have been localized in
astrocytes (Kreft et al., 2004); and astrocytes ex-
press key elements of vesicular exocytosis,
SNARE-family proteins (Parpura et al., 1995;
Crippa et al., 2006). However, astrocyte exocytosis
of glutamate is slower (
2 orders of magnitude)
than neuronal counterparts (Kreft et al., 2004).
The primary implication of astrocyte release of
transmitters, such as glutamate, is modulation of
synaptic transmission in nearby neurons and this
modulation has been shown in several experimen-
tal preparations. In cell culture, astrocytes were
shown to either evoke or depress synaptic activa-
tion and this activity was regulated by NMDA or
metabotropic glutamate receptors (Araque et al.,
1999). In hippocampal slice cultures, glutamate
release from astrocytes augmented interneuron
pyramidal synaptic connections (Kang et al.,
1998; Liu et al., 2004a) and activated kainate
64
receptors on interneurons (Liu et al., 2004b).
Functional consequences of astrocyte release of
D-serine have also been demonstrated in that as-
trocyte-derived D-serine, through activation at the
allosteric glycine-binding site, facilitated NMDA
receptor activity (Mothet et al., 2005; Panatier
et al., 2006). Also, as described above, astrocytes
can release ATP. ATP can be converted to adeno-
sine, which then can hyperpolarize adjacent neu-
rons as has been shown in the retina (Newman,
2003). Taken together, these studies demonstrate
that under physiological conditions astrocytes re-
lease transmitters that alter neuronal signaling.
While the effects of CNS injury on intra- and
intercellular calcium signaling as well as on as-
trocyte neuronal signaling have yet to be fully elu-
cidated, a strong body of evidence suggests that
astrocytes are not merely passive support cells but
are active communication partners to neurons in
the uninjured brain. Perhaps a better understand-
ing of the effect of trauma on these elements of
astrocyte communication will provide the infra-
structure to develop therapeutic interventions that
target multiple cell types in the injured brain.
Revisiting and revising the ‘‘Classical View’’ of
astrocytes in CNS trauma
As discussed in the introduction, TBI research to
date has focused mainly on neuroprotection and
given little consideration to the role of glial cells in
injury. This has significant implications for the
understanding of brain pathology after TBI
since the population of glial cells in brain is actu-
ally much larger than neurons. Furthermore, the
importance of the complex communication and
interaction between astrocytes and neurons de-
scribed above are becoming more apparent in
normal brain function and increasingly appear to
play a critical role in such perturbed brain func-
tions as seizures and ischemia injury (Vernadakis,
1996; Vesce et al., 1999). Increasingly, research
studies are indicating that astroglia cells are ad-
versely affected by trauma and ischemia and that
damage to astrocytes also affects the fate of neu-
rons in the traumatically injured brain. Early im-
pairment of astrocyte function after TBI may
compromise critical neuronal–glia interactions and
thus, may play a significant role in outcome after
injury. Perturbations in astrocyte function can also
have indirect effects on normal brain function. For
example, alterations in glial glutamate transporter
number or function could indirectly alter gluta-
mate-mediated excitability. The recent findings
that D-serine, which is exclusively released from
astrocytes in the synapse, preferentially acts as
the NMDA receptor co-agonist (Boehning and
Snyder, 2003) would suggest that any injury-
induced alteration in astrocyte function could ad-
versely impact normal glutamatergic signaling.
The majority of TBI laboratory studies that have
investigated astrocyte response to injury have fo-
cused on the phenomena of reactive astrocytes char-
acterized by increased GFAP immunoreactivity,
hypertrophy, and hyperplasia. The CNS responds
within days to injury by producing reactive as-
trogliosis and glial scarring (McGraw et al., 2001).
Although not completely understood, glial scarring
is thought to be an attempt by the CNS to restore
homeostasis by isolating the damaged region (Fitch
et al., 1999). However, the glial scar may interfere
with any subsequent neural repair or axonal regen-
eration (Ridet et al., 1997). Two days after midline
fluid percussion TBI in the rat, increased GFAP
staining and thickened glia consistent with reactive
astrocytes were observed in the CA3 hippocampus
(D’Ambrosio et al., 1999) even though midline fluid
percussion TBI is generally not associated with CA3
pyramidal cell loss (Lyeth et al., 1990). Reactive as-
trogliosis and the formation of a glial scar have been
viewed as both a promoter of and an impediment to
CNS regeneration (Reier et al., 1983; Muller et al.,
1995; Ridet et al., 1997). This chapter focuses on
events that involve TBI-induced astrocyte damage
or death that likely occur prior to the reactive as-
trocytic responses. Therapeutic interventions tar-
geted at these early mechanisms of astrocyte injury
may rescue both astrocytes and neurons and as a
consequence may reduce glial scarring.
Role of glutamate in astrocyte pathology in CNS
trauma
During normal neuronal activity extracellular
glutamate in the synaptic cleft must be rapidly

cleared to optimize the signal-to-noise ratio as well
as to prevent neuronal damage from excitotoxic-
ity. Glutamate is removed from the synapse by
glutamate transporters located in the plasma mem-
brane of neurons and adjacent astrocytes (Kanner
and Schuldiner, 1987). The major glutamate re-
moval mechanism is through the astrocyte-specific
sodium-dependent glutamate transporters, GLT-1
and GLAST (Danbolt, 1994, 2001; Takahashi
et al., 1997). GLT-1 is located primarily in the rat
forebrain while GLAST is found primarily in the
cerebellum (Tanaka et al., 1997; Danbolt, 2001).
Glutamate taken up by astrocytes is converted to
glutamine by glutamine synthetase, an energy-de-
pendent process requiring one molecule of ATP
for each molecule of glutamate converted to gluta-
mine (Daikhin and Yudkoff, 2000). Glutamine
then diffuses out of the astrocyte and is taken up
by the presynaptic neuron for recycling back into
glutamate via the enzyme glutaminase.
Glutamate excitotoxicity is a significant deter-
minant of TBI pathophysiology. Numerous labo-
ratory studies (Faden et al., 1989; Katayama et al.,
1990; Nilsson et al., 1990; Zhong et al., 2006) and
several clinical studies (Zauner et al., 1996; Bullock
et al., 1998; Koura et al., 1998) have documented
an excessive glutamate release following TBI. The
elevated extracellular concentration of glutamate
results in an excitotoxicity from excessive activa-
tion of ion-channel linked (Hayes et al., 1988;
McIntosh et al., 1989, 1998; Hicks et al., 1994)
and G-protein linked glutamate receptors (Hayes
et al., 1988; Hicks et al., 1994; Mukhin et al., 1996,
1997; Faden et al., 1997; Lyeth et al., 2001;
Zwienenberg et al., 2001). Experimental cerebral
ischemia is also associated with a large, rapid
increase in extracellular glutamate (Choi and
Rothman, 1990).
Energy demands and acidosis in CNS trauma
The uptake of glutamate rapidly increases energy
demands on astrocytes through several mecha-
nisms. Mechanical damage to astrocytes elicits el-
evated [Na+]i in large part from cotransport with
glutamate (Floyd et al., 2005). The sodium co-
transported into astrocytes along with glutamate
65
triggers activation of the Na+-K+ ATPase, which
is fueled by the ATP produced by glycolysis (Silver
and Erecinska, 1997). Further energy demand is
created by the conversion of glutamate to gluta-
mine by glutamine synthetase. Energy demands of
the brain are met through several pathways includ-
ing the oxidative phosphorylation, glycolytic, and
glycogen pathways. Recent studies indicate that
astrocytic glycolysis plays a major role in supplying
rapid energy demands of astrocytes (Schurr et al.,
1999; Paemeleire and Leybaert, 2000). Glycolysis
also provides an important source for neuronal
energy substrates in the form of lactate for subse-
quent conversion to pyruvate for use in the TCA
and oxidative phosphorylation pathways (Schurr
et al., 1999; Paemeleire and Leybaert, 2000). The
majority of excess lactate measured following TBI
is most likely of astrocytic origin and contributes to
lactic acidosis (Kawamata et al., 1992, 1995; Lyeth
et al., 1996). The lactate released as a by-product of
glycolysis in astrocytes is taken up by neurons and
used as an aerobic energy substrate. In vitro slice
experiments show that adjacent neurons use lactate
generated in astrocytes. The elevated levels of lac-
tate in slices exposed to high glutamate and ample
glucose are significantly increased when lactate
transport into neurons is inhibited (Schurr et al.,
1999). Furthermore, neuronal functional viability is
also lost when lactate transport into neurons is in-
hibited (Schurr et al., 1999). Elevated tissue levels
of lactate associated with hypoxia or ischemia have
often been considered an indicator of anaerobic
metabolism. However, many studies now suggest
that brain tissue also can produce lactate aerobi-
cally under certain physiological conditions when
stimulated (Raichle, 1991; Wilsch et al., 1994;
Hyder et al., 1996; Hu and Wilson, 1997). Taken
together, these results indicate that mechanical in-
jury to astrocytes rapidly elevates [Na+]i, followed
by a rapid energy demand, lactate production, and
intracellular acidosis.
The role of lactacidosis and a fall in intracellular
pH in brain injury and ischemia is well docu-
mented (Siesjo, 1988; Kraig and Chesler, 1990;
Staub et al., 1990; Nedergaard et al., 1991a, b).
Electron microscopy studies show that acidosis
promotes cytotoxic edema that is localized pre-
dominantly in astrocytic endfeet (Garcia et al.,
66
1977; Jenkins et al., 1982). Acidosis-induced glial
swelling is attributed to several membrane trans-
porters including Na+/H+ antiporter (Grinstein
and Rothstein, 1986), Cl/HCO3 exchanger
(Kimelberg et al., 1990), and the Na+/HCO3 co-
transporter (Newman, 1991). Activation of these
pH-stabilizing mechanisms during TBI-induced
acidosis would elicit increases in glial intracellular
Na+ and Cl along with osmotically obligated
water leading to astrocyte swelling (Kempski et al.,
1988; Jakubovicz and Klip, 1989). Recent studies
have demonstrated that amiloride analogues that
inhibit the NHE-1 significantly reduce hippocam-
pal extracellular glutamate levels following tran-
sient cerebral ischemia in rat (Phillis et al., 1998)
and reduce CA1 hippocampal neuronal cell death
assessed at 6 days after transient cerebral ischemia
in the gerbil (Phillis et al., 1999). Phillis et al.
(1999) attributed their results to the actions of
amiloride in reducing astrocyte swelling and
thereby increasing extracellular space. In a study
by Plesnila et al. (1999), astrocyte cultures exposed
to lactic acid exhibited cell volume increases of
21% when pHi dropped to 6.8–6.2, levels observed
in TBI. Swelling was completely blocked with the
NHE-1 inhibitor, amiloride. Ringel et al. (2000)
found that inhibition of anion transports reduced
lactacidosis-induced swelling in astrocyte cultures.
Activation of the pH-regulated membrane trans-
porters during periods of intracellular acidosis
contribute to further elevations in [Na+]i and may
result in reversal of the sodium/calcium exchanger
as discussed below.
Acute astrocyte loss following CNS insult
Astrocyte swelling is a prominent feature of both
cerebral ischemia (Jenkins et al., 1982) and TBI
(Castejon, 1998) and is considered to be an ‘‘exag-
gerated extension of normal astrocyte function’’
(Kimelberg, 1992). Astrocyte swelling is likely a
prominent component of raised ICP after both is-
chemia and TBI and resulting in part from alter-
ations in ion-transport mechanisms that leads to
osmotically obligated water entry (Kimelberg
et al., 1990). Cortical biopsies of human TBI com-
plicated with subdural hematoma have shown
severe astrocytic edema, dissociation of the peri-
vascular astrocyte endfeet from the capillary base-
ment membrane, and disruption of inter-astrocyte
gap junctions (Castejon, 1998). Electron micro-
scopy of brain tissue from human cerebral contu-
sion patients has revealed evidence of astrocytic
swelling as early as 3 h after injury, persisting for as
long as 3 days (Bullock et al., 1991). It is important
to consider that human tissue is essentially never
available for analysis during the first hours after
insult. Thus, there is a distinct possibility that hu-
man astrocyte swelling may occur very rapidly after
injury as in animal models of ischemia and TBI. In
addition to cell swelling and raised intracranial
pressure, the alterations in astrocyte ionic homeo-
stasis can likely initiate cell damage and death.
There has been a dogma that astrocytes are
highly resistant to hypoxic/ischemic insults com-
pared to neurons (Lukaszevicz et al., 2002). This
dogma has been challenged by early suggestions
and by later experiments demonstrating that as-
trocytes are indeed vulnerable to ischemic condi-
tions. Over 20 years ago, Plum (Plum, 1983)
suggested that the degree of infarction from focal
ischemia might depend upon the viability of as-
trocytes. This proved to be an insightful prediction
and subsequent studies indicate that this sugges-
tion applies to TBI as well as cerebral ischemia. A
number of in vitro studies have demonstrated that
astrocytes are indeed vulnerable to ischemia. As-
trocytes exposed to hypoxia alone can require up
to 24 h exposure to produce widespread death
(Yu et al., 1989; Sochocka et al., 1994) even in the
absence of glucose (Kelleher et al., 1993). How-
ever, combining hypoxia with acidosis, a more re-
alistic model, accelerates astrocyte death to just a
few hours (Swanson et al., 1997). Recent in vitro
studies have examined this issue utilizing a novel
experimental condition in which cultured as-
trocytes were exposed to hypoxic, acidic, ion-
shifted Ringers (HAIR) solution that mimics the
environmental characteristics of ischemia and TBI
(Bondarenko and Chesler, 2001a). Their microen-
vironment included glucose and thus mimicked
incomplete ischemia as well as severe TBI. They
found that as little as 15–20 min exposure to their
HAIR solution resulted in death of 60% of the
astrocytes (Bondarenko and Chesler, 2001a). The

same group subsequently reported that the NHE-1
inhibitor amiloride protected astrocytes exposed to
the HAIR protocol. Furthermore, blocking the
Na+/Ca2+ exchanger with benzamil or KB-
R7943, which preferentially inhibits reverse Na+/
Ca2+ exchange, protected astrocytes from the
HAIR protocol (Bondarenko and Chesler,
2001b; Bondarenko et al., 2005). These studies
provide evidence that astrocytic intracellular aci-
dosis activates NHE-1 leading to Na+ entry and
the reversal of the Na+/Ca2+ exchanger, trigger-
ing astrocyte cell loss after ischemia.
Recent in vivo studies provide further compelling
evidence of early astrocyte loss following cere-
bral ischemia. Liu et al. (1999) reported a 50% de-
crease in GFAP mRNA and protein in the ischemic
core within 5 h after permanent middle cerebral
artery occlusion, and almost complete absence of
GFAP mRNA by 12 h. The loss of GFAP mRNA
preceded the loss of the mRNA for GLUT3, a
neuron-specific glucose transporter. Immunohisto-
chemical experiments using other astrocyte markers
including S100beta, GSTYb1, and vimentin pro-
duced similar results. Interestingly, mRNA began
to increase in the peri-infarct area by 3 h and
peaked at 48 h after ischemia, suggestive of reactive
astrocyte proliferation in the ischemic penumbra.
Using a rat subdural hematoma-induced ischemia
model, Jiang et al. (2000) reported complete loss of
GFAP immunostaining at 4 h after the insult in
the ischemic core. Protoplasmic astrocytes appear
to be more vulnerable to ischemia than fibrous as-
trocytes. For example, following permanent middle
cerebral artery occlusion in mice, Lukaszevicz et al.
(2002) found that protoplasmic astrocytes (type 1,
accounting for most of the astrocytes in gray
matter) lost GFAP immunolabeling within minutes
in the ischemic core. In contrast, fibrous astrocytes
(type 2, found predominantly in white matter)
displayed a transient hypertrophy with no conspic-
uous cell death (Lukaszevicz et al., 2002). These
studies provide compelling evidence that ischemia
is associated with an early loss of astrocytes.
Glutamate excitotoxicity can lead to both ne-
crotic cell death and apoptotic programed cell
death following application of glutamate agonists
(Du et al., 1997; Nicotera et al., 1997) ischemia
(Namura et al., 1998; Graham and Chen, 2001)
67
and TBI (Rink et al., 1995; Newcomb et al., 1999).
Most of these studies have focused on apoptotic
cell death of neurons. Astrocytes are also vulner-
able to apoptotic cell death that may contribute to
the pathogenesis of many acute and chronic ne-
urodegenerative disorders (Takuma et al., 2004).
Hypoxia/ischemia induces mitochondrial depolari-
zation in astrocytes (Smith et al., 2003), that trig-
gers caspase-dependent cell death pathways and by
poly(ADP-ribose) polymerase-1 cell death path-
way (Giffard and Swanson, 2005). Astrocyte
apoptosis can be detected by caspase activation
by a variety of in vitro insults including staur-
osporin (Keane et al., 1997), cycloheximide (Tsuc-
hida et al., 2002), and ischemic conditions (Yu
et al., 2001; Gabryel et al., 2002). Activation of
apoptotic cascades in astrocytes has also been re-
ported after in vivo TBI. For example, accumula-
tion of active caspase-3 was observed for 5 days
after controlled cortical impact TBI in rats in
both neurons and astrocytes but, interestingly
with a higher proportion of activated caspase-3
in astrocytes than in neurons (Johnson et al.,
2005).
TBI and ischemia share a number of patho-
physiological similarities including excessive gluta-
mate release and selective neuronal cell loss. The
role of astrocytes in uptake of glutamate may have
particular importance to the traumatically injured
brain where large fluxes of extracellular glutamate
(Faden et al., 1989; Katayama et al., 1990) likely
contribute to acute excitotoxic processes (Olney
and Ho, 1970; Hayes et al., 1988). Specifically,
early impairment of astrocyte function after TBI
may compromise maintenance of homeostasis in
the extracellular microenvironment and disrupt
critical neuronal–glia interactions. Zhao et al.
(2003) demonstrated selective loss of GFAP and
glutamine synthetase immunoreactivity in rat
brains as early as 1 h after lateral fluid percussion
TBI. The loss of immunoreactivity was limited to
regions directly adjacent to areas of selective neu-
ronal vulnerability, the hippocampus CA3. The
loss of immunoreactivity for astrocytes markers
preceded fluoro-jade detection of neuronal degen-
eration by several hours suggesting that loss of
supporting astrocytes may contribute to subse-
quent neuronal cell loss.
68
Unraveling the mechanisms of acute astrocyte
damage with in vitro mechanical injury
It is clear from the studies described above that in
vivo experimental TBI or ischemia produces rapid
and profound damage to astrocytes. Moreover,
several in vitro models of ischemia also show acute
astrocyte pathology. Additionally, several of the
astrocytic intracellular events occurring acutely
after TBI have been characterized in a series of in
vitro studies using a mechanical injury model de-
veloped by Ellis et al. (1995). In this model, cells
are grown on a membrane that is subjected to a
rapid mechanical deformation that approximates
the tensile strain or stretch that is a key component
in the acceleration/deceleration type of closed hu-
man TBI (Meaney et al., 1995) and in vivo brain
injury (Margulies et al., 1990). This model has
been validated by demonstrating that in vitro
strain injury produces many of the post-traumatic
responses observed with in vivo injury models in-
cluding intracellular lesions to the mitochondria,
Golgi, and cytoskeletal elements in astrocytes and
neurons (Dietrich et al., 1994; Ellis et al., 1995;
McKinney et al., 1996), increases in astrocyte and
neuronal permeability (Ellis et al., 1995; Weber
et al., 1999; Willoughby et al., 2004), activates
phospholipases (Lamb et al., 1997; Floyd et al.,
2001), and induces free radical and isoprostane
formation (McKinney et al., 1996; Lamb et al.,
1997; Hoffman et al., 2000) as well as release of the
injury markers S-100b and neuron-specific enolase
(NSE; Slemmer et al., 2002, 2004). Strain-injury of
cultured neurons also alters the Mg2+ block of the
NMDA receptor (Zhang et al., 1996) and produces
a novel stretch-induced delayed neuronal depo-
larization, which may be related to transient neu-
ronal dysfunction observed in vivo (Hamm et al.,
1994; Tavalin et al., 1995). Strain-injury in neurons
also increases activation of the AMPA receptor
by decreasing AMPA receptor desensitization
(Goforth et al., 1999). Thus, this well character-
ized mechanical injury model recapitulates many
aspects of in vivo TBI and is a useful tool to ex-
amine injury-induced changes in acute astrocyte
pathology.
A critical element in astrocytic calcium signa-
ling is regulation of intracellular free Ca2+.
Mechanical strain injury to astrocytes produces a
rapid rise in [Ca2+]i that returns to near-basal lev-
els within 15 min. However, application of either
glutamate or the metabotropic glutamate group I
(mGluRI) agonist trans-1-aminocyclopentyl-1-3-
dicarboxylic acid (tACPD) after injury results in a
significantly blunted rise in [Ca2+]i for up to 24 h
post-injury, suggesting that the calcium signaling
machinery is disturbed by injury (Rzigalinski et al.,
1998). Alterations in intracellular calcium dynam-
ics after mechanical strain injury may be related to
alterations in inositol trisphosphate (IP3) signaling
as strain injury to astrocytes causes a significant
increase in IP3 at 5, 15, and 30 min and at 24 and
48 h post-injury (Floyd et al., 2001). Interestingly,
the injury-induced increases in IP3 at 15 and
30 min post-injury correspond to time points when
[Ca2+]i had returned to near normal levels; how-
ever, high IP3 should produce increased [Ca2+]i if
Ca2+ signaling were functioning normally. Instead
the opposite was observed — elevated IP3 at time
points when intracellular calcium has returned to
basal levels which indicates that IP3 may be un-
coupled from its target, the intracellular Ca2+
store. Additionally, antagonism of mGluRIs and
inhibition of PLC attenuated injury-induced un-
coupling of IP3-mediated Ca2+ signaling, reduced
astrocyte damage (Floyd et al., 2001, 2004), and
reduced injury-induced depletion of intracellular
calcium stores (Chen et al., 2004). Thus, mechan-
ical strain injury to astrocytes causes acute eleva-
tions in intracellular calcium and disruption of
IP3-mediated intracellular calcium signaling.
Mechanical strain injury to astrocytes also sig-
nificantly increases intracellular sodium (Floyd
et al., 2005). Mild and moderate strain injury pro-
duced a rapid rise in intracellular sodium that
returned to near-basal levels by 20–25 min post-
injury while severe injury produced increased in-
tracellular sodium for at least 50 min (the duration
of the experiment). These elevations in intracellu-
lar sodium were similar to those induced by exo-
genous glutamate application and were reduced by
pre-injury application of the sodium-dependent
glutamate uptake inhibitor TBOA, indicating that
sodium which accompanies glutamate uptake
contributes to injury-induced increases in intra-
cellular sodium. Comparison of the time course of

injury-induced elevations in intracellular sodium
and calcium indicated that there is significant
overlap in the first 30 min post-injury, which raises
the possibility that sodium/calcium exchange may
be involved in the acute astrocyte pathology after
injury. The magnitude and direction of calcium
flux of the sodium/calcium exchanger is depend-
ent, in part, on the Na+ electrochemical gradient,
and under normal conditions the sodium/calcium
exchanger operates in the ‘‘forward’’ or calcium-
efflux mode to extrude calcium. Importantly,
manipulations in uninjured astrocytes that raise
intracellular sodium have been shown to increase
intracellular calcium by ‘‘reversal’’ of the sodium/
calcium exchanger or calcium-influx operation
(Goldman et al., 1994). Thus, it was predicted
that the elevated intracellular sodium observed af-
ter mechanical strain injury may be sufficient to
reverse the sodium/calcium exchanger and cause
an influx of calcium. It was found that pre-injury
blockade of the calcium-influx (reversed) mode
with the compound KB-R7943 significantly re-
duced intracellular calcium in the moderate and
high levels of injury (but not at the low levels of
injury), the magnitudes of injury with the largest
and more sustained elevations in intracellular so-
dium (Floyd et al., 2005). Additionally, this sever-
ity of injury-dependent reversal of the sodium/
calcium exchanger was predicted by mathematical
modeling of the reversal potential of the sodium/
calcium exchanger using experimental values of
intercellular sodium and calcium concentrations
after mechanical strain injury. Taken together,
these data suggest that increased intracellular so-
dium can cause a reversal of the sodium/calcium
exchanger to cause influx of calcium into the cells,
and that this only occurs with moderate or severe
injury.
Mechanical strain injury to astrocytes causes
release of ATP (Ahmed et al., 2000; Neary et al.,
2005). ATP release following mechanical injury
not only has consequences for cellular energetics
(Bambrick et al., 2004), but also may alter inter-
cellular calcium signaling via P2 purinergic re-
ceptors. Cortical astrocytes express both the
ionotropic P2X receptors and the G-protein cou-
pled P2Y receptors (Lenz et al., 2000; Weisman
et al., 2005); however, P2Y expression is much
69
more prevalent in astrocytes than P2X (Illes and
Ribeiro, 2004). Nucleotides released from injured
cells can activate either the P2X or P2Y and cause
an increase in intracellular calcium (Illes and
Ribeiro, 2004) as well as activation of mitogen-
activated protein kinase (MAPK) pathways
(Gendron et al., 2003). Importantly, activation of
P2X receptors on astrocytes is linked to neurode-
generative events and astrogliosis (for review, see
Neary et al., 1996); yet, activation of P2Y receptors
on astrocytes is protective. For example, activation
of P2Y2 increases Bcl-2 family genes (Chorna et al.,
2004), and activates extracellular signal-regulated
protein kinase (ERK) which has been linked to the
protective activation of CREB (Neary et al., 1996;
Chorna et al., 2004). In an eloquent series of stud-
ies, Neary and colleagues have detailed the effects
of mechanical strain injury to astrocytes on P2 re-
ceptor activation and found that injury induces
a rapid release of ATP which, via P2 receptors,
activates ERK (Neary et al., 2003) and Akt
(Neary et al., 2005); increases expression and re-
lease of a protein that induces synapse formation,
thrombospondin-1 (Tran and Neary, 2006); and
also phosphorylates GSK3b at the Ser-9 location,
thereby inhibiting activity (Neary and Kang, 2006).
Thus, release of ATP from injured astrocytes can
act not only as a stimulus for gliosis, but also as a
promoter of cell survival and synaptic plasticity.
Further work is needed in this exciting area to bet-
ter elucidate the complex relationship between P2
activation, gliosis, and cell survival.
Mechanical strain injury to astrocytes disrupts
many of the critical features of intra- and inter-
cellular calcium signaling in astrocytes. The known
changes in astrocyte signaling after mechanical in-
jury are summarized schematically in Fig. 1. Strain
injury to astrocytes causes a rapid increase in in-
tracellular calcium and sodium that is dependent
on the magnitude of injury (Floyd et al., 2005).
There are several mechanisms by which calcium is
elevated following mechanical injury. Excessive
glutamate following injury agonizes mGluRs to
cause IP3-mediated calcium release (Floyd et al.,
2004). Theoretically, elevated glutamate following
injury could also activate the AMPA ionotropic
glutamate channel as has been shown in neurons
(Goforth et al., 2004). Although the effects of
70

ATP
P2YR Hemichannel?
Ca2+ mGluR ATP
Propidium
Iodide ?
AMPA MITOCHONDRIA
Na+ PLC

X
(reversed)

NC
IP3
Ca2+
1
T-
GL

Adjacent
Na+
Astrocyte TSP-2
Ca2+ Store Akt
ERK
Ca2+
Gap Junction

BCL-2
NUCLEUS

Fig. 1. Pathological events known to occur in astrocytes after mechanical strain injury. Mechanical strain injury to astrocytes results in
a rapid rise in intracellular calcium and sodium. Sources of injury-induced increases in intracellular calcium include activation of the
mGluRs, release of calcium from IP3-mediated intracellular calcium stores, reversal of the sodium/calcium exchanger, and possibly
AMPA receptor activation. Increases in intracellular sodium after mechanical injury are largely due to sodium-dependent uptake of
excessive glutamate. Injured astrocytes also release ATP. ATP could activate P2YRs and further increase intracellular calcium but also
could initiate cell survival pathways such as Bcl-2. Astrocytes are coupled by gap junctions, and small molecules such as IP3 or Ca2+
could travel between cells after injury. Propidium iodide, a marker for dead or damaged cells, could enter the astrocyte through
disruptions in the cell membrane and intercalate into the DNA. (See text for further details.) Abbreviations: protein kinase B/Akt
(Akt): (RS)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionate (AMPA), extracellular signal-regulated protein kinase (ERK),
glutamate transporter (GLT-1), inositol trisphosphate (IP3), metabotropic glutamate receptor (mGluR), sodium/calcium exchanger
(NCX), purinergic 2Y receptor (P2YR), phospholipase C (PLC).

mechanical injury on the subunit composition in
astrocytes has not been examined, yet, if the
GluR2 subunit is present, injury induced activa-
tion would likely contribute to increase intracellu-
lar Na+. However, if the GluR2 subunit were
absent, the channel would also be permeable to
Ca2+ (Frandsen and Schousboe, 2003). Activation
of either configuration of the AMPA receptor
would likely contribute to the pathophysiology of
astrocyte injury, as both increased intracellular
calcium and sodium are intimately involved. Ex-
cessive glutamate from mechanical injury also af-
fects the astrocyte by activation of sodium-
dependent glutamate uptake that significantly
elevates intracellular sodium. At moderate and se-
vere levels of injury, the elevation in intracellular
sodium is sufficient to cause reversal of the so-
dium/calcium exchanger and produce calcium

Fig. 2. Effect of mechanical injury on functional coupling in astrocytes. Confluent primary astrocyte cultures we mechanically injured
and then functional coupling was analyzed using fluorescence recovery after photobleach (FRAP). Panel (A) shows representative
micrographs of uninjured (left column) or moderately injured (right column) astrocytes before, immediately after, and 15 s after
photobleach. Regions of interest (ROIs) were drawn around four cells per field. The ROIs outlined in red and green were bleached but
the ROIs in yellow and blue were not. A representative trace of a mildly injured astrocyte at 24 h post-injury is shown in panel (B).
Panel (C) shows quantification of the maximal recovery at various post-injury times. Mild injury increased and moderate injury
decreased functional coupling in astrocytes (n ¼ 3 separate experiments).
71
72
influx, thereby further contributing to elevated in-
tracellular calcium (Floyd et al., 2005). Mechanical
strain injury causes release of ATP from damaged
mitochondria (Ahmed et al., 2000; Neary et al.,
2005), some of which is released into the extracel-
lular space. Although the mechanism of ATP re-
lease after injury is still unknown, candidates for
release include hemichannels, ATP transporters,
or vesicular exocytosis (Schwiebert, 2001). Extra-
cellular ATP following injury can activate pur-
inergic receptors including the P2YRs. Activation
of the P2YRs can increase intracellular calcium,
and also can have beneficial effects such as acti-
vation of cytoprotective pathways and increased
expression of Bcl-2 (Neary et al., 2003, 2005).
Propidium iodide, a marker of cell damage and
death, could enter the cell via perturbations in the
cell membrane. As astrocytes are extensively cou-
pled via gap junctions, small molecules such as
Ca2+ and IP3 could travel between cells to poten-
tially propagate an injury signal, as will be dis-
cussed below.
Work in uninjured astrocytes shows that gap
junction communication is an important element
in intercellular calcium signaling in astrocytes
which facilitates intercellular transfer of ions (i.e.,
Ca2+ or Na+) or second messengers (i.e., IP3).
Gap junctions are connections of intercellular
channels formed by hemichannels from adjoining
cell membranes, and the adult brain astrocytes are
so abundantly interconnected that the network is
regarded as a syncytium. Gap junctions are
formed by the joining of 2 hemichannels, each of
which comprises six protein subunits termed con-
nexins. The connexin43 protein (Cx43) is highly
expressed in astrocyte gap junctions while other
connexin proteins comprise neuronal gap junc-
tions (Yamamoto et al., 1990a, b; Giaume and
McCarthy, 1996). Gap junctions are not simply
passive channels between cells, but functionally
respond to alterations in the cellular milieu that
alters gap junction assembly and permeability,
mainly via phosphorylation (Musil et al., 1990;
Nagy and Li, 2000; Rouach et al., 2002; VanSlyke
and Musil, 2005). Most studies of the effect of in-
jury on gap junction coupling have used models of
ischemic brain injury with divergent findings. In-
creases in intracellular calcium, decreases in pH,
and production of oxygen-free radicals can inhibit
gap junction coupling (Bolanos and Medina, 1996;
Martinez and Saez, 2000); yet, other evidence sug-
gests that gap junctions remain open during is-
chemia (Cotrina et al., 1998a). The role of gap
junction communication in cell death following an
ischemic insult is also unclear. Some groups report
that inhibition of astrocyte gap junctions during
ischemic injury decreases the area of infarct and is
neuroprotective (Lin et al., 1998; Frantseva et al.,
2002), suggesting that astrocyte gap junctions am-
plify damage and spread cell death signals to oth-
erwise viable cells. An alternative hypothesis
suggests that gap junctions enable astrocytes to
use the syncytium to dissipate otherwise toxic con-
centrations of glutamate or ions (Blanc et al., 1998;
Siushansian et al., 2001). Part of the controversy
over the role of gap junction communication in
ischemic injury may be related to heterogeneity in
gap junction coupling as well as regional or cel-
lular differences in intracellular regulation of gap
junctions (Zvalova et al., 2004). For example, gap
junction hemichannels that open to the extracel-
lular space have been characterized (Contreras
et al., 2002) and may be involved in calcium signa-
ling or ATP release (Stout et al., 2002; Bennett
et al., 2003; Dermietzel et al., 2003).
The effect of traumatic brain injury on astrocyte
gap junction communication has only recently
been investigated. Ohsumi et al. (2006) have shown
that CX43 immunohistochemistry was increased in
the hippocampus 24 and 72 h after lateral fluid
percussion injury. We evaluated the effect of me-
chanical strain injury on functional connectivity of
gap junctions in astrocytes. In this experiment,
primary astrocyte cultures were grown on Flexs
plates and subjected to either a mild or moderate
rapid mechanical strain injury as previously de-
scribed (Floyd et al., 2005). Gap junction coupling
was evaluated at 30 min, 4 h, or 24 h after me-
chanical strain injury using fluorescence recovery
after photobleach (FRAP). Astrocytes were
bath loaded with 0.05 mg/ml 5-(and 6-) car-
boxyfluorescein diacetate-AM (CFDA) which is
a cell membrane and gap junction permeable
dye that becomes membrane impermeable after
de-esterification. Thus, when cells are visualized,
the CFDA is trapped within the cells but can

transfer between coupled cells via gap junctions.
After the fluorescent moieties of the dye are
bleached by intense laser excitation in a cell, any
subsequent increase in fluorescent intensity, or re-
covery, is attributed to transfer of dye between gap
junctions and represents the extent of gap junction
coupling. In this experiment, cells were visualized
on an upright Zeiss Axioscop 2 Laser scanning
confocal microscope (LSM510) using a Zeiss
Achroplan 20  dipping objective. The dye was
excited at 400 nm with emitted fluorescence above
520 nm captured by a photo-multiplier tube. Two
cells near the middle of the field were selected for
bleach and regions of interest (ROIs) were drawn.
Two additional ROIs were drawn for other cells in
the field that were not subjected to photobleach.
Three field scans were obtained prior to bleach.
Selected cells were bleached with 100 iterations at
100% laser power. To assess recovery after bleach,
field scans were acquired every 2 s for the next
2.5 min (75 scans total). Digital images obtained
during experiments were saved on a PC for off-line
analysis. Mean intensity values for each ROI were
divided by mean intensity values of reference (non-
bleached) cells to correct for photobleach of the
entire field. Figure 2(A) shows representative
micrographs of uninjured astrocytes (left column)
and astrocyte subjected to a moderate mechanical
strain injury (6.5 mm deformation) 30 min prior to
FRAP analysis. The ROIs outlined in red and
green were bleached but the ROIs outlined in yel-
low and blue were not. A representative trace of
the percentage of fluorescence recovery over time
for an astrocyte mildly injured 24 h prior to FRAP
is shown in Fig. 2(B). Recovery of fluorescence is
rapid and begins to plateau by
400 s after bleach.
Quantification of the maximal recovery at various
post-injury times is shown in Fig. 2(C). Uninjured
astrocytes were moderately coupled and recovery
after photobleach was between 30 and 40%. Neu-
rons are reportedly not extensively coupled and
were used as a procedural control with less than
5% fluorescence recovery detected in adjacent
neurons in co-cultures. In mildly injured as-
trocytes, a significant increase in recovery after
photobleach was seen at 30 min, and 4 and 24 h
after mechanical strain injury. However, with
moderate strain injury, recovery after photobleach
73
was decreased at these same time-points. Recovery
of photobleach was assessed after severe strain in-
jury, but many severely injured cells leaked dye
into the extracellular compartment, rendering the
severe injury data inconclusive (data not shown).
In summary, we found that mild injury increases
functional coupling and moderate injury decreases
coupling. Additional experiments will be con-
ducted to better understand the mechanisms of
these changes in gap junction coupling after me-
chanical strain injury and their role in the patho-
physiology of traumatic brain injury.
Conclusion
With some notable exceptions such as the study of
gliosis and astrocytic swelling, most TBI research
has generally focused on the pathophysiology of
neurons. However, many recent developments in
the field of glial biology demonstrate the active
nature of neuronal-astrocyte signaling indicating
the vital importance of astrocytes in normal brain
function. Furthermore, recent studies in Neuro-
trauma demonstrate the vulnerability of astrocytes
to CNS insults. Therefore, defining the time course
of astrocyte damage after TBI and uncovering the
mechanisms mediating that damage are critical to
a more complete understanding of TBI patho-
physiology. Collectively, these findings are leading
to a novel departure from traditional approaches
to TBI pathophysiology that have generally con-
centrated on injury to neurons.
Acknowledgments
Support provided by UCD Health Systems Re-
search Award (CLF), NIH NS29995 (BGL), NIH
NS45136 (BGL).
References
Aguado, F., Espinosa-Parrilla, J.F., Carmona, M.A. and
Soriano, E. (2002) Neuronal activity regulates correlated
network properties of spontaneous calcium transients in as-
trocytes in situ. J. Neurosci., 22: 9430–9444.
74
Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding,
H.A. and Ellis, E.F. (2000) Stretch-induced injury alters
mitochondrial membrane potential and cellular ATP in cul-
tured astrocytes and neurons. J. Neurochem., 74: 1951–1960.
Anlauf, E. and Derouiche, A. (2005) Astrocytic exocytosis ves-
icles and glutamate: a high-resolution immunofluorescence
study. Glia, 49: 96–106.
Araque, A., Sanzgiri, R.P., Parpura, V. and Haydon, P.G.
(1999) Astrocyte-induced modulation of synaptic transmis-
sion. Can. J. Physiol. Pharmacol., 77: 699–706.
Bambrick, L., Kristian, T. and Fiskum, G. (2004) Astrocyte
mitochondrial mechanisms of ischemic brain injury and
neuroprotection. Neurochem. Res., 29: 601–608.
Bennett, M.V., Contreras, J.E., Bukauskas, F.F. and Saez, J.C.
(2003) New roles for astrocytes: gap junction hemichannels
have something to communicate. Trends Neurosci., 26:
610–617.
Bezzi, P., Carmignoto, G., Pasti, L., Vesce, S., Rossi, D.,
Rizzini, B.L., Pozzan, T. and Volterra, A. (1998) Prostaglan-
dins stimulate calcium-dependent glutamate release in astro-
cytes. Nature, 391: 281–285.
Bezzi, P. and Volterra, A. (2001) A neuron-glia signalling net-
work in the active brain. Curr. Opin. Neurobiol., 11:
387–394.
Blanc, E.M., Bruce-Keller, A.J. and Mattson, M.P. (1998)
Astrocytic gap junctional communication decreases neuronal
vulnerability to oxidative stress-induced disruption of Ca2+
homeostasis and cell death. J. Neurochem., 70: 958–970.
Boehning, D. and Snyder, S.H. (2003) Novel neural modu-
lators. Annu. Rev. Neurosci., 26: 105–131.
Bolanos, J.P. and Medina, J.M. (1996) Induction of nitric oxide
synthase inhibits gap junction permeability in cultured rat
astrocytes. J. Neurochem., 66: 2091–2099.
Bondarenko, A. and Chesler, M. (2001a) Calcium dependence
of rapid astrocyte death induced by transient hypoxia, aci-
dosis, and extracellular ion shifts. Glia, 34: 143–149.
Bondarenko, A. and Chesler, M. (2001b) Rapid astrocyte death
induced by transient hypoxia, acidosis, and extracellular ion
shifts. Glia, 34: 134–142.
Bondarenko, A., Svichar, N. and Chesler, M. (2005) Role of
Na+-H+ and Na+-Ca2+ exchange in hypoxia-related acute
astrocyte death. Glia, 49: 143–152.
Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and
Adams, J.H. (1991) Glial swelling following human cerebral
contusion: an ultrastructural study. J. Neurol. Neurosurg.
Psychiatry, 54: 427–434.
Bullock, R., Zauner, A., Woodward, J.J., Myseros, J., Choi,
S.C., Ward, J.D., Marmarou, A. and Young, H.F. (1998)
Factors affecting excitatory amino acid release following se-
vere human head injury. J. Neurosurg., 89: 507–518.
Castejon, O.J. (1998) Morphological astrocytic changes in
complicated human brain trauma: a light and electron micro-
scopic study. Brain Inj., 12: 409–427.
Chao, T.I., Rickmann, M. and Wolff, J.R. (2002) The synapse-
astrocyte boundry: an anatomical basis for an inegrative role
of glia in synaptic transmission. In: Volterra A., Magistretti P.
and Haydon P.G. (Eds.), The Tripartite Synapse: Glia in
Synaptic Transmission. Oxford University Press, Oxford,
pp. 3–23.
Chen, T., Willoughby, K.A. and Ellis, E.F. (2004) Group I
metabotropic receptor antagonism blocks depletion of cal-
cium stores and reduces potentiated capacitative calcium en-
try in strain-injured neurons and astrocytes. J. Neurotrauma,
21: 271–281.
Choi, D.W. and Rothman, S.M. (1990) The role of glutamate
neurotoxicity in hypoxic-ischemic neuronal death. Annu.
Rev. Neurosci., 13: 171–182.
Chorna, N.E., Santiago-Perez, L.I., Erb, L., Seye, C.I., Neary,
J.T., Sun, G.Y., Weisman, G.A. and Gonzalez, F.A. (2004)
P2Y receptors activate neuroprotective mechanisms in as-
trocytic cells. J. Neurochem., 91: 119–132.
Contreras, J.E., Sanchez, H.A., Eugenin, E.A., Speidel, D.,
Theis, M., Willecke, K., Bukauskas, F.F., Bennett, M.V. and
Saez, J.C. (2002) Metabolic inhibition induces opening of
unapposed connexin 43 gap junction hemichannels and re-
duces gap junctional communication in cortical astrocytes in
culture. Proc. Natl. Acad. Sci. U.S.A., 99: 495–500.
Cornell-Bell, A.H. and Finkbeiner, S.M. (1991) Ca2+ waves in
astrocytes. Cell Calcium, 12: 185–204.
Cornell-Bell, A.H., Finkbeiner, S.M., Cooper, M.S. and Smith,
S.J. (1990) Glutamate induces calcium waves in cultured as-
trocytes: long-range glial signaling. Science, 247: 470–473.
Cotrina, M.L., Kang, J., Lin, J.H., Bueno, E., Hansen, T.W.,
He, L., Liu, Y. and Nedergaard, M. (1998a) Astrocytic gap
junctions remain open during ischemic conditions. J. Neuro-
sci., 18: 2520–2537.
Cotrina, M.L., Lin, J.H., Alves-Rodrigues, A., Liu, S., Li, J.,
Azmi-Ghadimi, H., Kang, J., Naus, C.C. and Nedergaard,
M. (1998b) Connexins regulate calcium signaling by control-
ling ATP release. Proc. Natl. Acad. Sci. U.S.A., 95:
15735–15740.
Cotrina, M.L., Lin, J.H., Lopez-Garcia, J.C., Naus, C.C. and
Nedergaard, M. (2000) ATP-mediated glia signaling. J.
Neurosci., 20: 2835–2844.
Crippa, D., Schenk, U., Francolini, M., Rosa, P., Verderio, C.,
Zonta, M., Pozzan, T., Matteoli, M. and Carmignoto, G.
(2006) Synaptobrevin2-expressing vesicles in rat astrocytes:
insights into molecular characterization, dynamics and
exocytosis. J. Physiol., 570: 567–582.
Daikhin, Y. and Yudkoff, M. (2000) Compartmentation of
brain glutamate metabolism in neurons and glia. J. Nutr.,
130: 1026S–1031S.
D’Ambrosio, R., Maris, D.O., Grady, M.S., Winn, H.R. and
Janigro, D. (1999) Impaired K(+) homeostasis and altered
electrophysiological properties of post-traumatic hippo-
campal glia. J. Neurosci., 19: 8152–8162.
Danbolt, N.C. (1994) The high affinity uptake system for ex-
citatory amino acids in the brain. Prog. Neurobiol., 44:
377–396.
Danbolt, N.C. (2001) Glutamate uptake. Prog. Neurobiol., 65:
1–105.
Dani, J.W., Chernjavsky, A. and Smith, S.J. (1992) Neuronal
activity triggers calcium waves in hippocampal astrocyte net-
works. Neuron, 8: 429–440.

D’Ascenzo, M., Vairano, M., Andreassi, C., Navarra, P., Azzena,
G.B. and Grassi, C. (2004) Electrophysiological and molecular
evidence of L-(Cav1), N-(Cav2.2), and R-(Cav2.3) type Ca2+
channels in rat cortical astrocytes. Glia, 45: 354–363.
Dermietzel, R., Meier, C., Bukauskas, F. and Spray, D.C.
(2003) Following tracks of hemichannels. Cell Commun.
Adhes., 10: 335–340.
Dietrich, W.D., Alonso, O., Busto, R., Globus, M.Y. and
Ginsberg, M.D. (1994) Post-traumatic brain hypothermia
reduces histopathological damage following concussive brain
injury in the rat. Acta Neuropathol., 87: 250–258.
Du, Y., Bales, K.R., Dodel, R.C., Hamilton-Byrd, E., Horn,
J.W., Czilli, D.L., Simmons, L.K., Ni, B. and Paul, S.M.
(1997) Activation of a caspase 3-related cysteine protease is
required for glutamate-mediated apoptosis of cultured cere-
bellar granule neurons. Proc. Natl. Acad. Sci. U.S.A., 94:
11657–11662.
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-
induced injury of cells in culture: characterization of the
model using astrocytes. J. Neurotrauma, 12: 325–339.
Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989)
The role of excitatory amino acids and NMDA receptors in
traumatic brain injury. Science, 244: 798–800.
Faden, A.I., Ivanova, S.A., Yakovlev, A.G. and Mukhin, A.G.
(1997) Neuroprotective effects of group III mGluR in trau-
matic neuronal injury. J. Neurotrauma, 14: 885–895.
Fitch, M.T., Doller, C., Combs, C.K., Landreth, G.E. and
Silver, J. (1999) Cellular and molecular mechanisms of glial
scarring and progressive cavitation: in vivo and in vitro anal-
ysis of inflammation-induced secondary injury after CNS
trauma. J. Neurosci., 19: 8182–8198.
Floyd, C.L., Gorin, F.A. and Lyeth, B.G. (2005) Mechanical
strain injury increases intracellular sodium and reverses Na(+)/
Ca(2+) exchange in cortical astrocytes. Glia, 51: 35–46.
Floyd, C.L., Rzigalinski, B.A., Sitterding, H.A., Willoughby,
K.A. and Ellis, E.F. (2004) Antagonism of group I meta-
botropic glutamate receptors and PLC attenuates increases in
inositol trisphosphate and reduces reactive gliosis in strain-
injured astrocytes. J. Neurotrauma, 21: 205–216.
Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A.,
Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury of
cultured astrocytes alters inositol (1,4,5)- trisphosphate-me-
diated signaling. Glia, 33: 12–23.
Frandsen, A. and Schousboe, A. (2003) AMPA receptor-
mediated neurotoxicity: role of Ca2+ and desensitization.
Neurochem. Res., 28: 1495–1499.
Frantseva, M.V., Kokarovtseva, L. and Perez Velazquez, J.L.
(2002) Ischemia-induced brain damage depends on specific
gap-junctional coupling. J. Cereb. Blood Flow Metab., 22:
453–462.
Gabryel, B., Adamczyk, J., Huzarska, M., Pudelko, A. and
Trzeciak, H.I. (2002) Aniracetam attenuates apoptosis of as-
trocytes subjected to simulated ischemia in vitro. Neurotoxi-
cology, 23: 385–395.
Garcia, J.H., Kalimo, H., Kamijyo, Y. and Trump, B.F. (1977)
Cellular events during partial cerebral ischemia. I. Electron
75
microscopy of feline cerebral cortex after middle-cerebral-
artery occlusion. Virchows Arch. B. Cell Pathol., 25: 191–206.
Giaume, C. and McCarthy, K.D. (1996) Control of gap-junc-
tional communication in astrocytic networks. Trends Neuro-
sci., 19: 319–325.
Giffard, R.G. and Swanson, R.A. (2005) Ischemia-induced
programmed cell death in astrocytes. Glia, 50: 299–306.
Glaum, S.R., Holzwarth, J.A. and Miller, R.J. (1990) Gluta-
mate receptors activate Ca2+ mobilization and Ca2+ influx
into astrocytes. Proc. Natl. Acad. Sci. U.S.A., 87: 3454–3458.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement
of AMPA-mediated current after traumatic injury in cortical
neurons. J. Neurosci., 19: 7367–7374.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical
injury modulates AMPA receptor kinetics via an NMDA
receptor-dependent pathway. J. Neurotrauma, 21: 719–732.
Goldman, W.F., Yarowsky, P.J., Juhaszova, M., Krueger, B.K.
and Blaustein, M.P. (1994) Sodium/calcium exchange in rat
cortical astrocytes. J. Neurosci., 14: 5834–5843.
Graham, S.H. and Chen, J. (2001) Programmed cell death in
cerebral ischemia. J. Cereb. Blood Flow Metab., 21: 99–109.
Grinstein, S. and Rothstein, A. (1986) Mechanisms of regula-
tion of the Na+/H+ exchanger. J. Membr. Biol., 90: 1–12.
Guthrie, P.B., Knappenberger, J., Segal, M., Bennett, M.V.,
Charles, A.C. and Kater, S.B. (1999) ATP released from as-
trocytes mediates glial calcium waves. J. Neurosci., 19:
520–528.
Hamm, R.J., Pike, B.R., O’Dell, D.M. and Lyeth, B.G. (1994)
Traumatic brain injury enhances the amnesic effect of an
NMDA antagonist in rats. J. Neurosurg., 81(2): 267–271.
Harris-White, M.E., Zanotti, S.A., Frautschy, S.A. and
Charles, A.C. (1998) Spiral intercellular calcium waves in
hippocampal slice cultures. J. Neurophysiol., 79: 1045–1052.
Hassinger, T.D., Guthrie, P.B., Atkinson, P.B., Bennett, M.V.
and Kater, S.B. (1996) An extracellular signaling component
in propagation of astrocytic calcium waves. Proc. Natl. Acad.
Sci. U.S.A., 93: 13268–13273.
Haydon, P.G. (2001) Glia: listening and talking to the synapse.
Nat. Rev. Neurosci., 2: 185–193.
Hayes, R.L., Jenkins, L.W., Lyeth, B.G., Balster, R.L.,
Robinson, S.E., Clifton, G.L., Stubbins, J.F. and Young,
H.F. (1988) Pretreatment with phencyclidine, an N-methyl-D-
aspartate antagonist, attenuates long-term behavioral deficits
in the rat produced by traumatic brain injury. J. Neurotrau-
ma, 5: 259–574.
Hicks, R.R., Smith, D.H., Gennarelli, T.A. and McIntosh, T.K.
(1994) Kynurenate is neuroprotective following experimental
brain injury in the rat. Brain Res., 655: 91–96.
Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis,
E.F. (2000) Astrocytes generate isoprostanes in response to
trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420.
Holzwarth, J.A., Gibbons, S.J., Brorson, J.R., Philipson, L.H.
and Miller, R.J. (1994) Glutamate receptor agonists stimulate
diverse calcium responses in different types of cultured rat
cortical glial cells. J. Neurosci., 14: 1879–1891.
Hu, Y. and Wilson, G.S. (1997) A temporary local energy pool
coupled to neuronal activity: fluctuations of extracellular
76
lactate levels in rat brain monitored with rapid-response
enzyme-based sensor. J. Neurochem., 69: 1484–1490.
Hyder, F., Chase, J.R., Behar, K.L., Mason, G.F., Siddeek, M.,
Rothman, D.L. and Shulman, R.G. (1996) Increased
tricarboxylic acid cycle flux in rat brain during forepaw stim-
ulation detected with 1H[13C]NMR. Proc. Natl. Acad. Sci.
U.S.A., 93: 7612–7617.
Illes, P. and Ribeiro, J.A. (2004) Molecular physiology of P2
receptors in the central nervous system. Eur. J. Pharmacol.,
483: 5–17.
Innocenti, B., Parpura, V. and Haydon, P.G. (2000) Imaging
extracellular waves of glutamate during calcium signaling in
cultured astrocytes. J. Neurosci., 20: 1800–1808.
Jakubovicz, D.E. and Klip, A. (1989) Lactic acid-induced
swelling in C6 glial cells via Na+/H+ exchange. Brain Res.,
485: 215–224.
Jenkins, L.W., Becker, D.P. and Coburn, T.H. (1982) A quan-
titative analysis of glial swelling and ischemic neuronal injury
following complete cerebral ischemia. In: Go K.G. and Baeth-
mann A. (Eds.), Recent Progress in the Study and Therapy of
Brain Edema. Plenum Press, New York, pp. 523–537.
Jiang, X.B., Ohno, K., Qian, L., Tominaga, B., Kuroiwa, T.,
Nariai, T. and Hirakawa, K. (2000) Changes in local cerebral
blood flow, glucose utilization, and mitochondrial function
following traumatic brain injury in rats. Neurol. Med. Chir.,
40: 16–28.
Johnson, E.A., Svetlov, S.I., Wang, K.K., Hayes, R.L. and
Pineda, J.A. (2005) Cell-specific DNA fragmentation may be
attenuated by a survivin-dependent mechanism after trau-
matic brain injury in rats. Exp. Brain Res., 167: 17–26.
Kang, J., Jiang, L., Goldman, S.A. and Nedergaard, M. (1998)
Astrocyte-mediated potentiation of inhibitory synaptic trans-
mission. Nat. Neurosci., 1: 683–692.
Kanner, B.I. and Schuldiner, S. (1987) Mechanism of transport
and storage of neurotransmitters. CRC Crit. Rev. Biochem.,
22: 1–38.
Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A.
(1990) Massive increases in extracellular potassium and the
indiscriminate release of glutamate following concussive
brain injury. J. Neurosurg., 73: 889–900.
Kawamata, T., Katayama, Y., Hovda, D.A., Yoshino, A. and
Becker, D.P. (1992) Administration of excitatory amino acid
antagonists via microdialysis attenuates the increase in glu-
cose utilization seen following concussive brain injury. J.
Cereb. Blood Flow Metab., 12: 12–24.
Kawamata, T., Katayama, Y., Hovda, D.A., Yoshino, A. and
Becker, D.P. (1995) Lactate accumulation following concus-
sive brain injury: the role of ion fluxes induced by excitatory
amino acids. Brain Res., 674: 196–204.
Keane, R.W., Srinivasan, A., Foster, L.M., Testa, M.P., Ord,
T., Nonner, D., Wang, H.G., Reed, J.C., Bredesen, D.E. and
Kayalar, C. (1997) Activation of CPP32 during apoptosis of
neurons and astrocytes. J. Neurosci. Res., 48: 168–180.
Kelleher, J.A., Chan, P.H., Chan, T.Y. and Gregory, G.A.
(1993) Modification of hypoxia-induced injury in cultured rat
astrocytes by high levels of glucose. Stroke, 24: 855–863.
Kempski, O., Staub, F., Jansen, M., Schodel, F. and Baeth-
mann, A. (1988) Glial swelling during extracellular acidosis
in vitro. Stroke, 19: 385–392.
Kimelberg, H.K. (1992) Astrocytic edema in CNS trauma. J.
Neurotrama, 9(Suppl. 1): S71–S81.
Kimelberg, H.K., Goderie, S.K., Higman, S., Pang, S. and
Waniewski, R.A. (1990) Swelling-induced release of gluta-
mate, aspartate, and taurine from astrocyte cultures. J.
Neurosci., 10: 1583–1591.
Kondou, H., Inagaki, N. and Fukui, H. (1994) Formation of
inositol phosphates mediated by M3 muscarinic receptors in
type-1 and type-2 astrocytes from neonatal rat cerebral cor-
tex. Neurosci. Lett., 180: 131–134.
Koura, S.S., Doppenberg, E.M., Marmarou, A., Choi, S.,
Young, H.F. and Bullock, R. (1998) Relationship between
excitatory amino acid release and outcome after severe hu-
man head injury. Acta Neurochir. Suppl., 71: 244–246.
Kraig, R.P. and Chesler, M. (1990) Astrocytic acidosis in
hyperglycemic and complete ischemia. J. Cereb. Blood Flow
Metab., 10: 104–114.
Kreft, M., Stenovec, M., Rupnik, M., Grilc, S., Krzan, M.,
Potokar, M., Pangrsic, T., Haydon, P.G. and Zorec, R.
(2004) Properties of Ca(2+)-dependent exocytosis in cul-
tured astrocytes. Glia, 46: 437–445.
Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A.
and Ellis, E.F. (1997) Alterations in phosphatidylcholine
metabolism of stretch-injured cultured rat astrocytes. J.
Neurochem., 68(5): 1904–1910.
Lenz, G., Gottfried, C., Luo, Z., Avruch, J., Rodnight, R., Nie,
W.J., Kang, Y. and Neary, J.T. (2000) P(2Y) purinoceptor
subtypes recruit different mek activators in astrocytes. Br. J.
Pharmacol., 129: 927–936.
Lin, J.H., Weigel, H., Cotrina, M.L., Liu, S., Bueno, E.,
Hansen, A.J., Hansen, T.W., Goldman, S. and Nedergaard,
M. (1998) Gap-junction-mediated propagation and amplifi-
cation of cell injury. Nat. Neurosci., 1: 494–500.
Liu, D., Smith, C.L., Barone, F.C., Ellison, J.A., Lysko, P.G.,
Li, K. and Simpson, I.A. (1999) Astrocytic demise precedes
delayed neuronal death in focal ischemic rat brain. Brain Res.
Mol. Brain Res., 68: 29–41.
Liu, Q.S., Xu, Q., Arcuino, G., Kang, J. and Nedergaard, M.
(2004a) Astrocyte-mediated activation of neuronal kainate
receptors. Proc. Natl. Acad. Sci. U.S.A., 101: 3172–3177.
Liu, Q.S., Xu, Q., Kang, J. and Nedergaard, M. (2004b) As-
trocyte activation of presynaptic metabotropic glutamate re-
ceptors modulates hippocampal inhibitory synaptic
transmission. Neuron Glia Biol., 1: 307–316.
Lukaszevicz, A.C., Sampaio, N., Guegan, C., Benchoua, A.,
Couriaud, C., Chevalier, E., Sola, B., Lacombe, P. and On-
teniente, B. (2002) High sensitivity of protoplasmic cortical
astroglia to focal ischemia. J. Cereb. Blood Flow Metab., 22:
289–298.
Lyeth, B.G., Gong, Q.Z., Dhillon, H.S. and Prasad, M.R.
(1996) Effects of muscarinic receptor antagonism on the
phosphatidylinositol bisphosphate signal transduction path-
way after experimental brain injury. Brain Res., 742: 63–70.

Lyeth, B.G., Gong, Q.Z., Shields, S., Muizelaar, J.P. and
Berman, R.F. (2001) Group I metabotropic glutamate
antagonist reduces acute neuronal degeneration and behavi-
oral deficits after traumatic brain injury in rats. Exp. Neurol.,
169: 191–199.
Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips,
L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990)
Prolonged memory impairment in the absence of hippocam-
pal cell death following traumatic brain injury in the rat.
Brain Res., 526: 249–258.
MacVicar, B.A., Hochman, D., Delay, M.J. and Weiss, S.
(1991) Modulation of intracellular Ca++ in cultured as-
trocytes by influx through voltage-activated Ca++ channels.
Glia, 4: 448–455.
Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990)
Physical model simulations of brain injury in the primate. J.
Biomech., 23: 823–836.
Martinez, A.D. and Saez, J.C. (2000) Regulation of astrocyte
gap junctions by hypoxia-reoxygenation. Brain Res. Brain
Res. Rev., 32: 250–258.
McGraw, J., Hiebert, G.W. and Steeves, J.D. (2001) Modulat-
ing astrogliosis after neurotrauma. J. Neurosci. Res., 63:
109–115.
McIntosh, T.K., Juhler, M. and Wieloch, T. (1998) Novel
pharmacologic strategies in the treatment of experimental
traumatic brain injury: 1998. J. Neurotrauma, 15: 731–769.
McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak,
S., Soares, H. and Faden, A.I. (1989) Traumatic brain injury
in the rat: characterization of a lateral fluid-percussion
model. Neuroscience, 28: 233–244.
McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F.
(1996) Stretch-induced injury of cultured neuronal, glial, and
endothelial cells: effect of polyethylene glycol-conjugated
superoxide dismutase. Stroke, 27: 934–940.
Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller,
R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical
analysis of experimental diffuse axonal injury. J. Neurotrau-
ma, 12: 689–694.
Mothet, J.P., Pollegioni, L., Ouanounou, G., Martineau, M.,
Fossier, P. and Baux, G. (2005) Glutamate receptor activa-
tion triggers a calcium-dependent and SNARE protein-de-
pendent release of the gliotransmitter D-serine. Proc. Natl.
Acad. Sci. U.S.A., 102: 5606–5611.
Mukhin, A., Fan, L. and Faden, A.I. (1996) Activation of me-
tabotropic glutamate receptor subtype mGluR1 contributes
to post-traumatic neuronal injury. J. Neurosci., 16:
6012–6020.
Mukhin, A.G., Ivanova, S.A. and Faden, A.I. (1997) mGluR
modulation of post-traumatic neuronal death: role of
NMDA receptors. Neuroreport, 8: 2561–2566.
Muller, H.W., Junghans, U. and Kappler, J. (1995) Astroglial
neurotrophic and neurite-promoting factors. Pharmacol.
Ther., 65: 1–18.
Musil, L.S., Cunningham, B.A., Edelman, G.M. and Good-
enough, D.A. (1990) Differential phosphorylation of the gap
junction protein connexin43 in junctional communication-
77
competent and -deficient cell lines. J. Cell Biol., 111:
2077–2088.
Nagy, J.I. and Li, W.E. (2000) A brain slice model for in vitro
analyses of astrocytic gap junction and connexin43 regula-
tion: actions of ischemia, glutamate and elevated potassium.
Eur. J. Neurosci., 12: 4567–4572.
Namura, S., Zhu, J., Fink, K., Endres, M., Srinivasan, A.,
Tomaselli, K.J., Yuan, J. and Moskowitz, M.A. (1998) Ac-
tivation and cleavage of caspase-3 in apoptosis induced by
experimental cerebral ischemia. J. Neurosci., 18: 3659–3668.
Neary, J.T. and Kang, Y. (2006) P2 purinergic receptors signal
to glycogen synthase kinase-3beta in astrocytes. J. Neurosci.
Res., 84: 515–524.
Neary, J.T., Kang, Y., Tran, M. and Feld, J. (2005) Traumatic
injury activates protein kinase B/Akt in cultured astrocytes:
role of extracellular ATP and P2 purinergic receptors. J. Ne-
urotrauma, 22: 491–500.
Neary, J.T., Kang, Y., Willoughby, K.A. and Ellis, E.F. (2003)
Activation of extracellular signal-regulated kinase by stretch-
induced injury in astrocytes involves extracellular ATP and
P2 purinergic receptors. J. Neurosci., 23: 2348–2356.
Neary, J.T., Rathbone, M.P., Cattabeni, F., Abbracchio, M.P.
and Burnstock, G. (1996) Trophic actions of extracellular
nucleotides and nucleosides on glial and neuronal cells.
Trends Neurosci., 19: 13–18.
Neary, J.T. and Zhu, Q. (1994) Signaling by ATP receptors in
astrocytes. Neuroreport, 5: 1617–1620.
Nedergaard, M., Goldman, S.A., Desai, S. and Pulsinelli, W.A.
(1991a) Acid-induced death in neurons and glia. J. Neurosci.,
11: 2489–2497.
Nedergaard, M., Kraig, R.P., Tanabe, J. and Pulsinelli,
W.A. (1991b) Dynamics of interstitial and intracellular
pH in evolving brain infarct. Am. J. Physiol., 260:
R581–R588.
Nett, W.J., Oloff, S.H. and McCarthy, K.D. (2002) Hippo-
campal astrocytes in situ exhibit calcium oscillations that
occur independent of neuronal activity. J. Neurophysiol., 87:
528–537.
Newcomb, J.K., Zhao, X., Pike, B.R. and Hayes, R.L. (1999)
Temporal profile of apoptotic-like changes in neurons and
astrocytes following controlled cortical impact injury in the
rat. Exp. Neurol., 158: 76–88.
Newman, E.A. (1991) Sodium-bicarbonate cotransport in reti-
nal Muller (glial) cells of the salamander. J. Neurosci., 11:
3972–3983.
Newman, E.A. (2003) Glial cell inhibition of neurons by release
of ATP. J. Neurosci., 23: 1659–1666.
Newman, E.A. and Zahs, K.R. (1997) Calcium waves in retinal
glial cells. Science, 275: 844–847.
Newman, E.A. and Zahs, K.R. (1998) Modulation of neuronal
activity by glial cells in the retina. J. Neurosci., 18:
4022–4028.
Nicotera, P., Ankarcrona, M., Bonfoco, E., Orrenius, S. and
Lipton, S.A. (1997) Neuronal necrosis and apoptosis: two
distinct events induced by exposure to glutamate or oxidative
stress. Adv. Neurol., 72: 95–101.
78
Nilsson, P., Hillered, L., Ponten, U. and Ungerstedt, U. (1990)
Changes in cortical extracellular levels of energy-related
metabolites and amino acids following concussive brain in-
jury in rats. J. Cereb. Blood Flow Metab., 10: 631–637.
Nimmerjahn, A., Kirchhoff, F., Kerr, J.N. and Helmchen, F.
(2004) Sulforhodamine 101 as a specific marker of astroglia
in the neocortex in vivo. Nat. Methods, 1: 31–37.
Ohsumi, A., Nawashiro, H., Otani, N., Ooigawa, H., Toyooka,
T., Yano, A., Nomura, N. and Shima, K. (2006) Alteration
of gap junction proteins (connexins) following lateral fluid
percussion injury in rats. Acta Neurochir. Suppl., 96:
148–150.
Olney, J.W. and Ho, O.L. (1970) Brain damage in infant mice
following oral intake of glutamate, aspartate or cysteine.
Nature, 227: 609–611.
Paemeleire, K. and Leybaert, L. (2000) Ionic changes accom-
panying astrocytic intercellular calcium waves triggered by
mechanical cell damaging stimulation. Brain Res., 857:
235–245.
Panatier, A., Theodosis, D.T., Mothet, J.P., Touquet, B., Pol-
legioni, L., Poulain, D.A. and Oliet, S.H. (2006) Glia-derived
D-serine controls NMDA receptor activity and synaptic
memory. Cell, 125: 775–784.
Parpura, V., Basarsky, T.A., Liu, F., Jeftinija, K., Jeftinija, S.
and Haydon, P.G. (1994) Glutamate-mediated astrocyte-
neuron signalling. Nature, 369: 744–747.
Parpura, V. and Haydon, P.G. (2000) Physiological astrocytic
calcium levels stimulate glutamate release to modulate adja-
cent neurons. Proc. Natl. Acad. Sci. U.S.A., 97: 8629–8634.
Parpura, V., Liu, F., Brethorst, S., Jeftinija, K., Jeftinija, S. and
Haydon, P.G. (1995) Alpha-latrotoxin stimulates glutamate
release from cortical astrocytes in cell culture. FEBS Lett.,
360: 266–270.
Parri, H.R. and Crunelli, V. (2003) The role of Ca2+ in the
generation of spontaneous astrocytic Ca2+ oscillations. Ne-
uroscience, 120: 979–992.
Parri, H.R., Gould, T.M. and Crunelli, V. (2001) Spontaneous
astrocytic Ca2+ oscillations in situ drive NMDAR-mediated
neuronal excitation. Nat. Neurosci., 4: 803–812.
Peakman, M.C. and Hill, S.J. (1995) Adenosine A1 receptor-
mediated changes in basal and histamine-stimulated levels of
intracellular calcium in primary rat astrocytes. Br. J. Pharma-
col., 115: 801–810.
Phillis, J.W., Estevez, A.Y., Guyot, L.L. and O’Regan, M.H.
(1999) 5-(N-Ethyl-N-isopropyl)-amiloride, an Na(+)-H(+)
exchange inhibitor, protects gerbil hippocampal neurons
from ischemic injury. Brain Res., 839: 199–202.
Phillis, J.W., O’Regan, M.H. and Song, D. (1998) 5-(N-Ethyl-
N-isopropyl)-amiloride inhibits amino acid release from the
ischemic rat cerebral cortex: role of Na+-H+ exchange.
Brain Res., 812: 297–300.
Plesnila, N., Haberstok, J., Peters, J., Kolbl, I., Baethmann, A.
and Staub, F. (1999) Effect of lactacidosis on cell volume and
intracellular pH of astrocytes. J. Neurotrauma, 16: 831–841.
Plum, F. (1983) What causes infarction in ischemic brain?: the
Robert Wartenberg lecture. Neurology, 33: 222–233.
Porter, J.T. and McCarthy, K.D. (1996) Hippocampal as-
trocytes in situ respond to glutamate released from synaptic
terminals. J. Neurosci., 16: 5073–5081.
Raichle, M.E. (1991) The metabolic requirements of functional
activity in the human brain: a positron emission tomography
study. Adv. Exp. Med. Biol., 291: 1–4.
Reier, P.J., Stensaas, L.J. and Guth, L. (1983) The astrocytic
scar as an impediment to regeneration in the central nervous
system. In: Kao C.C., Bunge R.P. and Reier P.J. (Eds.),
Spinal Cord Reconstruction. Raven Press, New York,
pp. 163–198.
Ridet, J.L., Malhotra, S.K., Privat, A. and Gage, F.H. (1997)
Reactive astrocytes: cellular and molecular cues to biological
function. Trends Neurosci., 20: 570–577.
Ringel, F., Chang, R.C., Staub, F., Baethmann, A. and Ples-
nila, N. (2000) Contribution of anion transporters to the ac-
idosis-induced swelling and intracellular acidification of glial
cells. J. Neurochem., 75: 125–132.
Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Ne-
ugebauer, E. and McIntosh, T.K. (1995) Evidence of apo-
ptotic cell death after experimental traumatic brain injury in
the rat. Am. J. Pathol., 147: 1575–1583.
Rouach, N., Avignone, E., Meme, W., Koulakoff, A., Venance,
L., Blomstrand, F. and Giaume, C. (2002) Gap junctions and
connexin expression in the normal and pathological central
nervous system. Biol. Cell, 94: 457–475.
Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis,
E.F. (1998) Intracellular free calcium dynamics in stretch-
injured astrocytes. J. Neurochem., 70: 2377–2385.
Schurr, A., Miller, J.J., Payne, R.S. and Rigor, B.M. (1999) An
increase in lactate output by brain tissue serves to meet the
energy needs of glutamate-activated neurons. J. Neurosci.,
19: 34–39.
Schwiebert, E.M. (2001) ATP release mechanisms, ATP recep-
tors and purinergic signalling along the nephron. Clin. Exp.
Pharmacol. Physiol., 28: 340–350.
Shao, Y. and McCarthy, K.D. (1995) Receptor-mediated cal-
cium signals in astroglia: multiple receptors, common stores
and all-or-nothing responses. Cell Calcium, 17: 187–196.
Siesjo, B.K. (1988) Acidosis and ischemic brain damage. Ne-
urochem. Pathol., 9: 31–88.
Silver, I.A. and Erecinska, M. (1997) Energetic demands of the
Na+/K+ ATPase in mammalian astrocytes. Glia, 21: 35–45.
Siushansian, R., Bechberger, J.F., Cechetto, D.F., Hachinski,
V.C. and Naus, C.C. (2001) Connexin43 null mutation in-
creases infarct size after stroke. J. Comp. Neurol., 440:
387–394.
Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T.
(2002) Repeated mild injury causes cumulative damage to
hippocampal cells. Brain, 125: 2699–2709.
Slemmer, J.E., Weber, J.T. and De Zeeuw, C.I. (2004) Cell
death, glial protein alterations and elevated S-100 beta release
in cerebellar cell cultures following mechanically induced
trauma. Neurobiol. Dis., 15: 563–572.
Smith, I.F., Plant, L.D., Boyle, J.P., Skinner, R.A., Pearson,
H.A. and Peers, C. (2003) Chronic hypoxia potentiates

capacitative Ca2+ entry in type-I cortical astrocytes. J.
Neurochem., 85: 1109–1116.
Sochocka, E., Juurlink, B.H., Code, W.E., Hertz, V., Peng, L.
and Hertz, L. (1994) Cell death in primary cultures of mouse
neurons and astrocytes during exposure to and ‘recovery’
from hypoxia, substrate deprivation and simulated ischemia.
Brain Res., 638: 21–28.
Staub, F., Baethmann, A., Peters, J., Weigt, H. and Kempski,
O. (1990) Effects of lactacidosis on glial cell volume and
viability. J. Cereb. Blood Flow Metab., 10: 866–876.
Stout, C.E., Costantin, J.L., Naus, C.C. and Charles, A.C.
(2002) Intercellular calcium signaling in astrocytes via ATP
release through connexin hemichannels. J. Biol. Chem., 277:
10482–10488.
Swanson, R.A., Farrell, K. and Stein, B.A. (1997) Astrocyte
energetics, function, and death under conditions of incom-
plete ischemia: a mechanism of glial death in the penumbra.
Glia, 21: 142–153.
Takahashi, M., Billups, B., Rossi, D., Sarantis, M., Hamann, M.
and Attwell, D. (1997) The role of glutamate transporters in
glutamate homeostasis in the brain. J. Exp. Biol., 200: 401–409.
Takuma, K., Baba, A. and Matsuda, T. (2004) Astrocyte
apoptosis: implications for neuroprotection. Prog. Ne-
urobiol., 72: 111–127.
Tanaka, J., Ichikawa, R., Watanabe, M., Tanaka, K. and In-
oue, Y. (1997) Extra-junctional localization of gluta-
mate transporter EAAT4 at excitatory Purkinje cell synapses.
Neuroreport, 8: 2461–2464.
Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical
perturbation of cultured cortical neurons reveals a stretch-
induced delayed depolarization. J. Neurophysiol., 74(6):
2767–2773.
Tran, M.D. and Neary, J.T. (2006) Purinergic signaling induces
thrombospondin-1 expression in astrocytes. Proc. Natl.
Acad. Sci. U.S.A., 103: 9321–9326.
Tsuchida, T., Kato, T., Yamada, A. and Kawamoto, K. (2002)
Cycloheximide induces apoptosis of astrocytes. Pathol. Int.,
52: 181–185.
VanSlyke, J.K. and Musil, L.S. (2005) Cytosolic stress reduces
degradation of connexin43 internalized from the cell surface
and enhances gap junction formation and function. Mol.
Biol. Cell, 16: 5247–5257.
Verkhratsky, A., Orkand, R.K. and Kettenmann, H. (1998)
Glial calcium: homeostasis and signaling function. Physiol.
Rev., 78: 99–141.
Vernadakis, A. (1996) Glia-neuron intercommunications and
synaptic plasticity. Prog. Neurobiol., 49: 185–214.
Vesce, S., Bezzi, P. and Volterra, A. (1999) The active role of
astrocytes in synaptic transmission. Cell Mol. Life Sci., 56:
991–1000.
Volterra, A. and Meldolesi, J. (2005) Astrocytes, from brain
glue to communication elements: the revolution continues.
Nat. Rev. Neurosci., 6: 626–640.
Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F.
and Ellis, E.F. (1999) Alterations in calcium-mediated signal
79
transduction after traumatic injury of cortical neurons. Cell
Calcium, 26: 289–299.
Weisman, G.A., Wang, M., Kong, Q., Chorna, N.E., Neary,
J.T., Sun, G.Y., Gonzalez, F.A., Seye, C.I. and Erb, L. (2005)
Molecular determinants of P2Y2 nucleotide receptor func-
tion: implications for proliferative and inflammatory path-
ways in astrocytes. Mol. Neurobiol., 31: 169–183.
Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir,
J.K. and Ellis, E.F. (2004) S100B protein is released by in
vitro trauma and reduces delayed neuronal injury. J. Ne-
urochem., 91: 1284–1291.
Wilsch, V.W., Pidoplichko, V.I., Opitz, T., Shinozaki, H. and
Reymann, K.G. (1994) Metabotropic glutamate recep-
tor agonist DCG-IV as NMDA receptor agonist in imma-
ture rat hippocampal neurons. Eur. J. Pharmacol., 262:
287–291.
Yamamoto, T., Hertzberg, E.L. and Nagy, J.I. (1990a)
Epitopes of gap junctional proteins localized to neuronal
subsurface cisterns. Brain Res., 527: 135–139.
Yamamoto, T., Ochalski, A., Hertzberg, E.L. and Nagy, J.I.
(1990b) On the organization of astrocytic gap junctions in rat
brain as suggested by LM and EM immunohistochemistry of
connexin43 expression. J. Comp. Neurol., 302: 853–883.
Yu, A.C., Gregory, G.A. and Chan, P.H. (1989) Hypoxia-in-
duced dysfunctions and injury of astrocytes in primary cell
cultures. J. Cereb. Blood Flow Metab., 9: 20–28.
Yu, A.C., Wong, H.K., Yung, H.W. and Lau, L.T. (2001)
Ischemia-induced apoptosis in primary cultures of astrocytes.
Glia, 35: 121–130.
Zauner, A., Bullock, R., Kuta, A.J., Woodward, J. and Young,
H.F. (1996) Glutamate release and cerebral blood flow
after severe human head injury. Acta Neurochir. Suppl., 67:
40–44.
Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996)
Reduction of voltage-dependent Mg2+ blockade of NMDA
current in mechanically injured neurons. Science, 274(5294):
1921–1923.
Zhao, X., Ahram, A., Berman, R.F., Muizelaar, J.P. and Lyeth,
B.G. (2003) Early loss of astrocytes after experimental trau-
matic brain injury. Glia, 44: 140–152.
Zhong, C., Zhao, X., Van, K.C., Bzdega, T., Smyth, A., Zhou,
J., Kozikowski, A.P., Jiang, J., O’Connor, W.T., Berman,
R.F., Neale, J.H. and Lyeth, B.G. (2006) NAAG peptidase
inhibitor increases dialysate NAAG and reduces glutamate,
aspartate and GABA levels in the dorsal hippocampus fol-
lowing fluid percussion injury in the rat. J. Neurochem., 97:
1015–1025.
Zvalova, D., Cordier, J., Mesnil, M., Junier, M.P. and Chneiw-
eiss, H. (2004) p38/SAPK2 controls gap junction closure in
astrocytes. Glia, 46: 323–333.
Zwienenberg, M., Gong, Q.Z., Berman, R.F., Muizelaar, J.P.
and Lyeth, B.G. (2001) The effect of groups II and III
metabotropic glutamate receptor activation on neuronal in-
jury in a rodent model of traumatic brain injury. Neurosur-
gery, 48: 1119–1127.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 6

Rescuing neurons and glia: is inhibition of apoptosis

useful?

E. Kovesdi1, E. Czeiter1, A. Tamas2, D. Reglodi2, D. Szellar1, J. Pal3, P. Bukovics1,3,

T. Doczi1 and A. Buki1,

1
Department of Neurosurgery, University Medical School, Pécs University, Pécs, Hungary, Rét u. 2. H-7624, Hungary
2
Department of Anatomy, University Medical School, Pécs University, Pécs, Hungary, Szigeti u. 12. H-7624, Hungary
3
Neurobiology Research Group of the Hungarian Academy of Sciences, University Medical School, Pécs University, Pécs,
Hungary, Rét u. 2. H-7624, Hungary

Abstract: Traumatic brain injury (TBI) represents a leading cause of death in western countries. Despite all
research efforts we still lack any pharmacological agent that could effectively be utilized in the clinical
treatment of TBI. Detailed unraveling of the pathobiological processes initiated by/operant in TBI is a
prerequisite to the development of rational therapeutic interventions. In this review we provide a summary
of those therapeutic interventions purported to inhibit the cell death (CD) cascades ignited in TBI. On
noxious stimuli three major forms of CD, apoptosis, autophagia and necrosis may occur. Apoptosis can be
induced either via the mitochondrial (intrinsic) or the receptor mediated (extrinsic) pathway; endoplasmic
reticular stress is the third trigger of caspase-mediated apoptotic processes. Although, theoretically pan-
caspase inhibition could be an efficient tool to limit apoptosis and thereby the extent of TBI, potential
cross-talk between various avenues of CD suggests that more upstream events, particularly the preservation
of the cellular energy homeostasis (cyclosporine-A, poly ADP ribose polymerase (PARP) inhibition, hy-
pothermia treatment) may represent more efficient therapeutic targets hopefully also translated to the
clinical care of the severely head injured.

Keywords: cell death; caspases; calpain; traumatic brain injury; traumatically induced axonal injury;
necrosis; apoptosis

Introduction per year for TBI is $60 billion. The TBI-related

Traumatic brain injury (TBI) puts an extreme
burden on societies worldwide. More than five
million Americans have a TBI-related long-term or
lifelong need for help for daily activities. Long-
term consequences include functional changes
such as cognitive deficit, epilepsy, hypopytuitar-
ism and increased risk for Alzheimer’s disease and
Parkinson’s disease. The cumulative societal cost
Corresponding author. Tel.: +36-30-411-3349; Fax: +36-72-
535931; E-mail: andras.buki@aok.pte.hu
death rate of the population under 35 years of age
is 3.5 times that of cancer and heart disease to-
gether (Lewin, 1991; Thurman et al., 1999). De-
spite of the social and economical significance of
TBI we still lack any efficient specific pharmaco-
logical approach to interfere with the damaging
consequence of injury (Narayan et al., 2002). The
therapeutic approaches utilized to date are basi-
cally not different from those of the late seventies
to early eighties; a noteworthy reduction in mor-
tality was only achieved via the implementation of
scientific evidence based therapeutic guidelines.

DOI: 10.1016/S0079-6123(06)61006-6 81
82
While such reorganization of the system and eve-
ryday practice of TBI care led to marked reduction
in mortality and morbidity, the clinical care for the
head injured has reached its limits (Vukic et al.,
1999; Ghajar, 2000). Thereby, now basic scientists
and/or pharmaceutical research should provide
further ammunition for clinicians to fight the silent
epidemic represented by TBI. The importance of
translational studies in TBI is further highlighted
by data from the field of health care management,
indicating that effective treatment of TBI is con-
sidered one of the most cost-efficient medical inter-
ventions (Saltman and Findling, 1997).
Although detailed analysis of the pathobiolog-
ical processes activated in TBI are beyond the
scope of this work, the authors will provide a short
theoretical background to each section addressing
various approaches to rescue neurons and glia in
TBI, while also referring to other chapters of this
work for further relevant information.
Cell death routes
On noxious influences or agents three major forms
of cell death (CD) may occur: autophagic, necrotic
and apoptotic.
Autophagy is characterized by lysosomal se-
questration and degradation of cellular compo-
nents. This mode of cell death has its potential role
in physiological processes as well as in neurode-
generative disorders (for review, see Assuncao and
Linden, 2004).
While primary brain injury leads to focal and
diffuse alterations in brain parenchyma, which are
predominantly necrotic by nature, secondary brain
damage practically triggered and initiated at the
moment of TBI is also associated with apopto-
tic damage. In fact, as time elapses post-injury,
apoptotic processes become more dominant and in
late phases post-injury autophagy may also partici-
pate particularly in axonal demise (Lockshin and
Zakeri, 2004; Ringger et al., 2004; Gallyas et al.,
2006).
Necrosis is characterized with swelling and rup-
ture of the cell (or its axon); membrane damage
and activation of inflammatory reactions is a con-
sistent feature of necrotic cell death. Such necrotic
processes are also associated with energy deple-
tion/negative energetic state of the cell (Cai et al.,
1998; Arden and Betenbaugh, 2004). Recent ob-
servations proved that both in focal and diffuse
forms of experimental TBI in injured soma as well
as in axons, various proteolytic processes are trig-
gered and are responsible for necrotic demise
(Bartus, 1997; Yamashima, 2004; Farkas et al.,
2006). These studies demonstrated that mechanic
or enzymatic alterations of membrane integrity in
conjunction with altered function of ionic pumps
and channels should lead to intracellular calcium
accumulation which in turn results in the activa-
tion of neutral proteases, primarily the cysteine-
aspartate protease calpain (Bartus, 1997; Stys and
Jiang, 2002). Besides direct proteolysis associated
with calpain activation, indirect proteolytic conse-
quences such as calpain-mediated lysosomal rup-
ture and cathepsin activation (Yamashima, 2004)
also contribute to cellular demise.
Although the association and feedback connec-
tion between permeability changes, activation of
proteolytic processes, cellular demise and func-
tional alterations following TBI is not entirely de-
veloped, preliminary data suggest that inhibition
of necrotic enzymes may facilitate functional re-
covery (Saatman et al., 1996).
Regardless of the initiating factors and the route
leading to altered intracellular ionic homeostasis,
increased level of intracellular Ca++ represents a
permissive environment for the activation of an-
other form of CD that is apoptosis (Cai et al., 1998).
Apoptosis, frequently referred to as programmed
cell death plays a pivotal role in embryogenesis
and is frequently initiated by a non-lethal stim-
ulus leading to cell death via the activation of
proteolytic cascades. Characteristic features of
apoptosis include DNA laddering (detectable by
electrophoresis), DNA fragmentation (TUNEL/
terminal deoxynucleotidyl-transferase-mediated de-
oxyuridine triphosphate-biotin nick-end labeling),
sub-2N DNA (FACS), rounding/blebbing of the
cell, fragmentation of nuclei with marginalization
of chromatin, exteriorization of phosphatidylserine
(annexin V binding) (for review, see Arden and
Betenbaugh, 2004; Lockshin and Zakeri, 2004).
As apoptosis requires a relatively intact energetic
state of the cell, most probably it will not be

localized to the core of tissue laceration (focal dam-
age) rather it will involve remote/diffuse (‘‘penum-
bra’’) area of the central nervous system (CNS). In
the last decade multiple lines of evidences pointed to
the activation of apoptotic processes in various
forms of TBI (Rink et al., 1995; Conti et al., 1998;
Buki et al., 2000; Ginis et al., 2000).
Studies in more severe, focal forms of TBI have
proven cytochrome c (cyto c) release and caspase-3
activation in the cerebral cortex 6–24 h post-injury
(Sullivan et al., 2002). Caspase-12 activation was
also demonstrated in cortex, hippocampus and
cortical astrocytes in this model of TBI implicating
endoplasmic reticulum-associated apoptotic proc-
esses (vide infra) in the pathogenesis of cell loss
(Larner et al., 2004) These authors also demon-
strated capsase-7 activation in the same model
peaking at day five post-injury (Larner et al., 2005).
Fig. 1. The three major routes of apoptosis. Note that different caspases participate at initiation and the execution of the cell death
processes. Also note the crucial role of calcium and the mitochondrion (for further details, see text).
83
Knoblach et al. (2002), in the model of lateral fluid
percussion TBI in accordance with other, previous
reports (Yakovlev et al., 1997; Beer et al., 2000;
Clark et al., 2000), demonstrated caspase-3 and -9
activation in the cerebral cortex and the hippo-
campus 1–12 h post-injury. Although these workers
were not able to demonstrate considerable activa-
tion of caspase-8, in other models including cortical
impact TBI caspase-8 (representing the induction
of the extrinsic route of apoptosis) was also found
activated in both neurons and glia (Beer et al.,
2001).
Apoptotic pathways
Apoptosis could be initiated via three major path-
ways (Fig. 1), and in most of these pathways
84
intracellular Ca++, mitochondria and the caspase
enzyme family play a crucial role. The extrinsic
receptor-mediated route utilizes specific cell sur-
face receptors belonging to the tumor necrosis
factor (TNF) super family where binding of the
‘‘death signal’’ (such as Fas by Fas ligand) triggers
the activation of the caspase enzyme cascade
(Chinnaiyan et al., 1995). Caspases are cysteine-
aspartate proteases which are present in the cell in
the inactive proenzyme (zymogen) form. Specifi-
cally, ligand–receptor interaction in the extrinsic
route will activate pro-caspase-8, which, in turn
activates pro-caspase-3.
Similarly to this pathway, the final executor of
the second, intrinsic pathway is also caspase-3,
however, this route is mediated by mitochondrial
alterations (Cai et al., 1998; Kruman and Mattson,
1999; Adams and Cory, 2002; Kroemer, 2003).
Mitochondria play a key role in apoptosis. Seve-
ral studies demonstrated both in vitro and in vivo
that TBI leads to a perturbation of the energy
homeostasis (Giza and Hovda, 2001). Partially due
to excessive firing of neurons and/or ionic imbal-
ance due to altered pump/channel function as well
as mechanoporation in at least a subpopulation of
cell bodies and axons, a clear metabolic dysfunc-
tion occurs, leading to a shift toward anaerobic
glycolysis, lactate-production and the accumula-
tion of reactive oxygen species (ROS) (Pettus et al.,
1994; Stys and Jiang, 2002; Singleton and Povli-
shock, 2004; Farkas et al., 2006). N-methyl-D-
aspartate (NMDA) receptor activation in neurons,
failure of the Na-K-ATPase in axons and me-
chanoporation in some soma and axons in con-
junction with altered energy homeostasis leads to
the accumulation of intracellular (intraaxonal)
Ca++ (Siesjo et al., 1999; Stys and Jiang, 2002).
Mitochondria accumulate the excess amount
of Ca++ leading to swelling and dissipation of
their membrane potential, initiating the perm-
ealization of their outer membrane for solutes
under 1500 Da, widely explained by the opening
of the mitochondrial permeability transition
pore (MPTP) (Zoratti and Szabo, 1995; Susin
et al., 1998; Kruman and Mattson, 1999). This
purported pore is consisting of the adenine nuc-
leotide translocase (ANT), voltage dependent an-
ion channel (VDAC), cyclophilin D, peripheral
benzodiazepine receptor, hexokinase, creatine kin-
ase and Bcl-2 proteins (Susin et al., 1999) (some of
these Bcl-2 proteins harbor anti-apoptotic poten-
tial, while others possess pro-apoptotic activity
(Reed, 1998; Scorrano and Korsmeyer, 2003)).
Opening of the MPTP leads to the release of pro-
apoptotic substances — cyto c, second mito-
chondrial activator of caspases (SMAC)/direct
inhibitor of apoptosis protein (IAP) binding pro-
tein with low pI (DIABLO) and apoptosis induc-
ing factor (AIF). When released, cyto c together
with the apoptosis activating factor 1 (Apaf-1),
dATP and cytosolic pro-caspase-9 forms the ‘‘apo-
ptosome’’ resulting in the generation of active ca-
spase-9 which, in turn triggers the transition of
pro-caspase-3 to active caspase-3 (Cai et al., 1998;
Adams and Cory, 2002).
The third, most recently described route of
apoptosis is initiated by endoplasmatic reticulum
(ER)-stress leading to caspase-12 activation. This
latter enzyme can act either as an executor caspase
or via direct or indirect activation of caspase-3
(Oyadomari et al., 2002; Zong et al., 2003).
Crossroads of cell death
To develop rationally targeted therapeutic strate-
gies beyond understanding these basic mechanisms
of cell death, it is also mandatory to unravel those
potential routes via the pathways initiated by noxi-
ous stimuli may communicate or interact.
Such connections or ‘‘detours’’ exist within the
apoptotic machinery itself as well as between var-
ious major pathways of cell death that is necrosis,
apoptosis and autophagy (Denecker et al., 2001;
Lockshin and Zakeri, 2004). In addition to their
role in the maintenance of the ionic and energy
homeostasis and triggering the intrinsic pathway
of apoptosis, mitochondria also serve as a cross-
road or common pathway for all routes of apo-
ptotic cell death (Cai and Jones, 1998; Saikumar
et al., 1998) (Fig. 1).
To this end, besides pro-apoptotic members of
the Bcl-2 family the mitochondrial membrane also
contains other pro-apoptotic proteins such as Bax
and Bak. These are responsible for the cross-talk
between mitochondria (that is the intrinsic

pathway of apoptosis) and the other two avenues
of apoptotic cell death; the extrinsic route is con-
verging to these pro-apoptotic proteins via the ca-
spase-8 Bid–Bax/Bak interaction while the
endoplasmatic reticulum stress also interacts with
these proteins as well as via sending direct Ca++
signals to the mitochondria (Jurgensmeier et al.,
1998; Schendel et al., 1998; Ferri and Kroemer,
2001).
Not only the major intracellular ‘‘death-ave-
nues’’ but also the redundancy of the effector ca-
spases provide considerable versatility for cell
death; studies with knock-out mice as well as ca-
spase-inhibitor studies have proven that an injured
cell has the potential to select between various ef-
fector caspases and such decision is most probably
influenced by various, physiologically active in-
hibitors of apoptosis (Assuncao and Linden,
2004).
Alternative, protease-mediated pathways may
also contribute to cell death. Ubiquitination of
some caspases and their subsequent destruction by
the proteosome may participate in the regulation
of apoptosis (Lockshin and Zakeri, 2002; Ditzel
et al., 2003; Varshavsky, 2003). Recent observa-
tions have proven that disconnection of the cell
from its neighboring structures specifically from
the extracellular matrix should contribute to the
initiation of cell death mechanisms. This process,
primarily mediated by matrix metalloproteases is
frequently described with the Greek word ‘‘an-
oikis’’ (homelessness) (Lockshin and Zakeri, 2002;
Abe et al., 2004).
The above-detailed complexity of the cell death
machinery provides multiple choices for a cell to
react to a noxious stimulus. Apoptotic cell death
requires energy and phagocytes that scavenge the
dying cells. In cell cultures, in vitro in the absence
of phagocytes the final mechanisms of cell death
are always necrotic, despite of the initiating agent
and the route the cell entered to die (Arden and
Betenbaugh, 2004; Assuncao and Linden, 2004).
Limited sources of ATP can easily halt the apo-
ptotic processes and the cell may fail to display
apoptotic morphology, however, this does not
mean an arrest of cell death rather the initiation of
autophagic or necrotic processes (Cai et al., 1998;
Reed, 1998; Saikumar et al., 1998).
85
When we consider, how to inhibit apoptosis, we
also have to answer, whether inhibition of
apoptosis is useful. Theoretically, in a philosophi-
cal approach, we may conclude that a lethal or
even non-lethal stimulus may ignite any of the
above-detailed pathways. Due to the cross-talk
between these avenues and the internal diversity of
the individual pathways it is quite possible that
pharmacological intervention to halt a death proc-
ess may just shift the cell death mechanism to an-
other pathway of cellular demise. Inhibition of
apoptotic cell death can lead to autophagy instead
of resulting in cell survival (Lang-Rollin et al.,
2003). Similarly, although the application of ca-
spase inhibitors is a popular way to interfere with
apoptosis, more upstream targets should be advo-
cated for pharmacological intervention. Specifi-
cally, as caspase-9 and -3 represent the post-
mitochondrial phase of cell death their activation
requires the release of cyto c and other pro-apo-
ptotic substances. Such excessive mitochondrial
damage also points to a considerable disruption of
the electron transport chain leading to mass gene-
ration of free radicals as well as to altered energy
homeostasis both facilitating the activation of ne-
crotic processes in the cell. In traumatic axonal
injury (TAI) the activation of caspases leads to
irreversible digestion of the membrane skeleton
and other cytoskeletal structures indicative of im-
minent, irreversible axonal demise (Wang et al.,
1998; Buki et al., 2000) (Fig. 2). The relevance of
such co-activation of the ‘‘necrotic’’ and ‘‘apopto-
tic’’ machinery was also demonstrated in an ele-
gant study by Liu et al. (2006) in hippocampal
lysates from rats which have undergone controlled
cortical impact injury, where degradome-studies
identified various common targets of calpain-2 and
caspase-3 including beta-II-spectrin, synaptotag-
min-1 and striatin. Similar ‘‘collaboration’’ bet-
ween calpain and caspase-3 was also demonstrated
by Warren et al. (2005) in metamphetamine
induced experimental neurotoxicity.
Inhibition of apoptosis
Several studies with caspase-inhibitors corroborate
— at least from some aspects — the above-described
86
Fig. 2. Images of damaged axonal profiles from the medial le-
mniscus of rat exposed to impact acceleration brain injury
(60 min post-injury). Cytochrome c release from damaged mito-
chondria (immunofluorescent label on A) and caspase-3 acti-
vation (immunofluorescent label on B) are co-localised in the
same damaged axons (digital overlay on C) pointing to the
contribution of calcium-induced mitochondrial damage and
caspase activation to the pathogenesis of traumatically induced
diffuse brain injury (axonal injury).
critical attitude to the usefulness of inhibiting the
effector-phase of apoptosis alone. Abrahamson
et al. (2006) demonstrated that a pan-caspase in-
hibitor (BAF) significantly reduced the formation
of caspase-cleaved amyloid precursor protein frag-
ments also leading to a reduced volume of hippo-
campal cell loss, however, in their experiment
performed in mice applying the cortical impact
model of injury significant necrotic areas where
still noted in the cortex and the hippocam-
pus. In another work, the pan-caspase inhibitor
FK-011 was not able to achieve reduction in
cortical lesion size measured 7 days post-injury
(Sullivan et al., 2002). Although Knoblach et al.
(2004) were able to prove significant improvement
of motor and neurological function using the
pan-caspase inhibitor z-VAD-fmk, they also stress
that this beneficial effect was purportedly also
attributable to the inhibition of calpain that is
co-inhibition of necrosis in the dose they applied.
In their review Lockshin and Zakeri (2004) also
stress that co-inhibition of lysosomal proteases
may lead to overestimation of the efficacy of
caspase-inhibition.
Keeping these warnings in mind one should
conclude that inhibition of apoptosis alone might
be ‘‘too little, too late’’. Thus, the most promising
way to interfere with the progression of the patho-
biological processes evoked by/operant in TBI
must either be the application of therapeutic
agents with multiple targets like the above men-
tioned pan-caspase inhibitors or targeting pre-
mitochondrial/mitochondrial events in the course
of cellular demise.
As far as multitargetic therapeutic approaches
are considered, sex steroids should be included to
our list; several studies suggested that steroids
could have beneficial effects both on neurons and
glia. In ischemic and TBI sex steroids have been
proven to inhibit necrotic as well as apoptotic cell
death (Roof and Hall, 2000). Although the exact
mechanism of their action still requires further in-
vestigation, sex steroids are thought to inhibit lipid
peroxydation and the formation of ROS while also
exerting anti-apoptotic properties facilitating the
expression of anti-apoptotic proteins including
bcl-2 and Bax (Soustiel et al., 2005). Similarly, est-
rogen treatment has a positive effect on oligoden-
droglia function (Curry and Heim, 1966) and
astrocytic scar formation (Garcia-Estrada et al.,
1999). Testosterone can influence the expression of
the astrocytic water channel protein aquaporin-4
indicating that sex steroids may also exert their
action on the formation of brain edema — a major
player in secondary brain injury (Gu et al., 2003).
In the last two decades several studies suggested
that poly(ADP-ribose)polymerase-1 (PARP-1, EC
2.4.2.30) inhibition might be beneficial in the in-
hibition of CNS injury of various origin (Burkle,
2001; Besson et al., 2003, 2005). This enzyme was
originally associated with DNA-repair function:
stress induced strand DNA breaks activate PARP

that will transfer ADP-ribose units to nuclear pro-
teins from NAD. This energy consuming process
was recognized to lead to NAD depletion and loss
of ATP leading to a collapse in cellular energy
homeostasis and cell death (Burkle, 2001). While
long-term inhibition of PARP could theoretically
lead to mutagenesis and carcinogenesis, in the
acute phase such intervention would be beneficial
in terms of restoring/preserving the cellular energy
pool. As it has recently been stressed by Komjati et
al. (2005) PARP inhibition is not only interfering
with ROS-generated necrosis. Beyond representing
a potential pathway in initiating and/or modulat-
ing necrosis via AIF, PARP has a substantial role
in the induction of apoptosis. Via the NF-kappaB-
pathway PARP is also capable of influencing the
expression of inflammatory cytokines and media-
tors (Komjati et al., 2005).
In the last few years several PARP inhibitors
have been proven to inhibit TBI. In the lateral
fluid percussion head injury model LaPlaca et al.
(2001) have demonstrated that the PARP inhibitor
GPI 1650 significantly reduced the lesion volume
while not influencing the number of TUNEL posi-
tive cells.
Other experiments proved that PARP inhibitors
were able to improve the neurological score of in-
jured animals without reducing the extent of the
cellular lesion (Besson et al., 2005). These data in
conjunction with those results indicating that
PARP inhibitors reduce the extent of TAI as-
sessed by beta amyloid precursor protein immuno-
reactivity implicate the inhibition of TAI in the
neuroprotective effects of PARP inhibitors.
Fluid percussion TBI is not the only model
where PARP inhibitors proved their neuroprotec-
tive effect: 3-aminobenzanide significantly reduced
the lesion size in cold injury (Hortobagyi et al.,
2003) and proved beneficial functional effects
(PJ34) as well as improved effectiveness of neural
stem cell transportation (Lacza et al., 2003).
In the last decades, several studies implicated the
cytosolic neutral cysteine protease calpain (EC
3.4.22.17) in the pathogenesis of both diffuse and
focal TBI (Bartus, 1997; Kampfl et al., 1997; Buki
et al., 1999b). This Ca++-induced enzyme is ca-
pable of reversibly cleaving spectrin and other
constituents of the cytoskeleton (Wang et al.,
87
1998). While classic thought appreciated calpain as
an agent contributing to necrotic processes several
lines of evidences demonstrated that calpain can
participate in the induction of the mitochondrial
(intrinsic) pathway of apoptosis. This communi-
cation between the classic, Ca++-induced necrotic
processes and the apoptotic enzyme system is par-
tially executed that way that calcium accumulation
leads to axolemmal permeability alteration
through proteolytic modification of the subaxo-
lemmal network (cortical cytoskeleton) which in
turn at least in some neurons and in axons con-
tributes to further membrane-permeability
changes (Buki et al., 1999b). As it has been de-
scribed in the above passages, excessive influx of
Ca++ induces mitochondrial damage via MPTP-
opening. In addition to this pathway, calpain itself
is capable — as it has been demonstrated in liver
— to open the MPTP via its direct proteolytic
modification (Aguilar et al., 1996). Once MPTP is
open and the intrinsic route, thus caspase-3 is
activated, it can cleave calpastatin, the major
inhibitor of calpain, contributing to further un-
controlled activation of calpain. Taking into con-
sideration such interactions and the temporal/
causative relationship between calpain and caspase
activation a logical way to influence the release
of apoptotic enzymes should be the inhibition of
calcium-mediated proteolytic alterations that is
the inhibition of calpain.
Recent observations indicated that systemic ad-
ministration of calpain inhibitors reduced the
overall degree of the cerebral ischemia (Bartus,
1997). Further, Saatman et al. (1996) demon-
strated that the use of calpain inhibitors resulted in
improved behavioral outcome in a contusional
model of TBI. Intriguingly enough, in an extension
of this work Saatman et al. failed to detect any
significant change in the size of contusion, con-
cluding that the beneficial therapeutic effects of
AK-295 could be explained by inhibiting TAI in
remote areas (Saatman et al., 2000). The correct-
ness of this assumption was further supported by
the finding that in a rodent model of TAI the cell
permeable peptidyl-aldehyde calpain inhibitor
MDL-28170 (carbobenzylzoxy-Val-Phe-H) proved
to prevent axonal injury in the brainstem fiber
tracts (Buki et al., 2003). Most recent observations
88
indicate that this agent is also potent in terms of
reducing the length of axonal segments displaying
traumatically induced axolemmal permeability
changes frequently associated with the activation
of the Ca++-induced protease calpain coupled in
some axons to the initial, strain-induced axolem-
mal mechanoporation (Büki et al., unpublished)
(vide supra). Although physical and chemical
properties of MDL-28170 do not favor its appli-
cation in the clinical care, the family of such cell-
permeable calpain-inhibitors that readily cross
the blood brain barrier and exert their beneficial
effects in a relatively wide therapeutic window
(Bartus, 1997; Markgraf et al., 1998) should be
considered candidates for further pharmacological
modifications and studies.
Calpain inhibitors are of particular interest as
recent reports demonstrated that besides decreas-
ing the formation of cleavage products of this
cysteine protease primarily participating in necro-
tic processes, they are capable of inhibiting the
formation of caspase-cleaved byproducts, thereby
apoptosis too (Kawamura et al., 2005).
In the line of therapeutic approaches aiming to
inhibit the mitochondrial/premitochondrial phase
of apoptosis one of the most promising interven-
tions is the use of the inhibitors of mPTP opening.
Among these interventions the immunophilin lig-
and cyclosporine A (CsA) has been used in a set of
experiments targeting TAI in the rodent impact
acceleration model of TBI. First, immuno-electron
microscopic studies by Okonkwo and Povlishock
(1999) demonstrated that CsA significantly atten-
uated mitochondrial swelling and rupture, trans-
lating into reduced numbers of damaged,
disconnected axonal segments displaying immonu-
reactivity for the transport protein beta amyloid
precursor protein. Subsequent studies also proved
that both pre- and post-injury administration
of CsA significantly reduced the extent of TAI
detected by immunohistochemical markers of
calpain-mediated spectrin proteolysis, axonal
disconnection and cytoskeletal alterations — ne-
urofilament compaction (NFC). Dose response
curves for the beneficial effects of CsA in the
experimental model of TBI have also been estab-
lished (Okonkwo et al., 2003). More recent stud-
ies also demonstrated that CsA limits/improves
some aspects of functional outcome following TBI
(Riess et al., 2001). Preliminary data regarding the
use of CsA in head injury in man indicate that this
compound is safe and potentially beneficial (Alves
et al., 2003).
Another immunophilin ligand FK506, a sub-
stance that has no known effect on MPT and ex-
erts its action via inhibition of calcineurin activity
(Liu et al., 1991; Kuroda et al., 1999) has also
displayed beneficial effects on TAI in a pre-injury
administration paradigm (Singleton et al., 2001)
and reduced the complications associated with
rapid re-warming following hypothermic interven-
tion for TAI (Suehiro et al., 2001).
Inhibition of the Ca++-induced phosphatase
calcineurin is supposed to be beneficial in axonal
injury by preventing dephosphorylization of the
neurofilament side arms that should decrease the
repelling forces between the neurofilaments
thereby permitting NFC leading to impairment
of axoplasmatic transport (Povlishock et al., 1997;
Okonkwo et al., 1998). Such dephosphorylated
neurofilaments would be more susceptible for pro-
teolytic degradation by calpain and caspase (Pant,
1988). It is of note that calcineurine might also
participate in the signaling pathways regulating
proliferative responses of astrocytes and its inhi-
bition by CsA as well as FK-506 diminishes as-
trocytic reactions (Pyrzynska et al., 2001).
Calcineurin also contributes to the control of
synaptic activity by inactivating dynamin I and
synapsin I, proteins participating in neurotrans-
mitter release (Cousin and Robinson, 2001) and by
dephosporylation-mediated induction of nitric ox-
ide synthase leading to increased nitric oxide level
and neurotransmitter release (Dawson et al.,
1993).
Other important aspects of TBI might also be
influenced by immunophyllin ligands. Specifically,
CsA inhibited ischemia-induced edema in vitro,
suggesting that such a beneficial effect should also
be worth investigating in secondary brain injury
where hypoperfusion-hypoxya-induced edema and
tissue damage is an issue, too (MacGregor et al.,
2003).
CsA also proved neuroprotective in the hands of
Gabbita et al. (2005) in the controlled cortical im-
pact model where it inhibited the TBI-induced

increase of cleaved tau (cytoskeletal protein) level
in the ipsilateral hippocampus and its accumula-
tion in the CSF of injured rodents. Van Den et al.
(2004) have demonstrated another beneficial effect
of CsA in TBI proving that 30 min post-injury
administration of CsA reduced APP mRNA at 2
and 6 h post-injury and likewise reduced APP ac-
cumulation in injured cell bodies. Nevertheless,
they did not prove axonal preservation in terms of
APP-immunoreactivity in their experiments. While
these results might reflect the complexity in the
pathogenesis of TBI, it is of note that several
studies have proved that APP immunoreactivity
alone roughly underestimates the extent of TBI
and, in some axon-populations other mechanisms
than impaired axoplasmatic transport and axonal
swelling should be held accountable for axonal
demise (Singleton and Povlishock, 2004; Farkas
et al., 2006).
While, without doubt, a continuous develop-
ment of various anti-apoptotic or, even more
promising, anti-apoptotic, anti-necrotic drugs
should be a major goal of pharmaceutical re-
search, neuroprotective substances physiologically
produced in the brain itself should also be taken
into account. To this end, a bulk of recently col-
lected evidences point to the neuroprotective ef-
fects of pituitary adenylate cyclase activating
polypeptide (PACAP).
This member of the vasoactive intestinal peptide
(VIP)/secretin/glucagon peptide family (Miyata et
al., 1989) was discovered as a hypothalamic pep-
tide on its potential of increasing adenylate cyclase
activity in the pituitary gland and proved to exert
various effects in the central and peripheral nerv-
ous systems including neurotrophic and neuropro-
tective ones (Arimura, 1998; Vaudry et al., 2000).
In the extradural static weight compression model
of spinal cord injury in rat, post-injury PACAP
treatment significantly reduced the number of
apoptotic cells assessed by TUNEL in the spinal
cord (Katahira et al., 2003).
Further, PACAP has been shown to eliminate
the increase of TNF after spinal cord transection
(Kim et al., 2000) and in experimental TBI post-
injury induction of PACAP-mRNA and its upreg-
ulation paralleled the decrease in the number of
apoptotic cells (Skoglosa et al., 1999). These
89
studies indicate that PACAP is a promising ther-
apeutic agent in traumatic brain and spinal cord
injuries, which is further supported by our obser-
vations.
Unfortunately, the exact mechanism underlying
the in vivo neuroprotective effect of PACAP has
not been clarified so far. In vitro studies assign
both anti-apoptotic and anti-inflammatory actions
to PACAP. In cerebellar granule cells, PACAP
significantly inhibited the activation of caspase-3
(Vaudry et al., 2004). PACAP was proven to be a
potent inactivator of induced microglial release of
pro-inflammatory cytokines and nitric oxide
(Kong et al., 1999; Kim et al., 2000; Delgado et
al., 2003). It is also believed to influence mi-
tochondrial integrity as PACAP inhibited the mi-
tochondrial Ca++ uptake induced inactivation of
aconitase, a key mitochondrial enzyme influencing
the viability of neurons (Tabuchi et al., 2003).
In the last few years our laboratories extensively
investigated the purported neuroprotective role of
PACAP in a diffuse model of TBI. Intriguingly
enough, we were not able to reproduce neuropro-
tective effects in terms of preserving axonal integ-
rity assessed by APP-immunoreactivity (vide
supra) with the administration paradigm (125 mg
pre-injury, iv) that had worked in a middle cere-
bral artery occlusion model of stroke (Reglodi et
al., 2002). Nevertheless, we were able to establish
the dose-response curve for intracerebroventricu-
lar (icv) administration of PACAP, proving that
100 mg PACAP significantly reduced the density of
damaged, immunoreactive axons in the cortico-
spinal tract (Farkas et al., 2004). Our most recent
observations proved that a considerable therapeu-
tic window exists for post-injury PACAP treat-
ment demonstrating significant axonal protection
in an icv administration paradigm 2 h following
impact acceleration TBI (Tamas et al., 2006).
A new avenue of pharmacotherapy targeting
apoptosis in TBI is represented by the inhibition
of cell cycle proteins. Di Giovanni et al. (2005)
demonstrated that the cell cycle inhibitor flavopir-
idol was not only capable of inhibiting etoposide-
induced, capase-mediated cell death in cultured
primary cortical neurons but also inhibited the
proliferation of astrocytes and microglia and de-
creased the extension of TBI-induced tissue lesion
90
and improved functional recovery in the lateral
fluid percussion rodent model.
As proteolytic processes are key players in TBI-
evoked apoptosis, the list of potential therapeutic
measures would not be complete without men-
tioning the neuroprotective capacity of therapeutic
hypothermia. This intervention is purported to act
via general inhibition or slowing of the proteolytic
processes which could provide an extended ther-
apeutic window for further, specific therapeutic
approaches while also enabling the neurons to
keep their mitochondria up and running and their
local energy homeostasis intact (Buki et al.,
1999a). Although the use of controlled hypother-
mia in the treatment of human TBI remains con-
troversial, on the basis of multiple lines of evidence
this therapeutic modality might be worth further
investigations and probably further clinical studies
planned on the basis of past experience could re-
solve the controversy concerning the role of hypo-
thermia in the treatment of the severely head
injured (Clifton, 2004). Specifically, on the basis of
our current knowledge on its action, hypothermia
should be introduced relatively earlier in the care
of TBI paying special attention to gradual re-
warming in conjunction with introduction of other
treatment strategies such as FK-506 that has al-
ready proved its potential to inhibit diffuse brain
injury associated with post-hypothermic rewarm-
ing (Suehiro et al., 2001).
Concluding remarks
There are several controversies in the development
of novel pharmacotherpeutic agents aimed at the
inhibition of pathobiological processes evoked by
TBI. While everybody in the field considers the
treatment of TBI an extremely cost efficient medi-
cal procedure, we still lack any novel, efficient
therapeutic agent. Despite of the promising results
achieved with various pre-clinical studies all can-
didates failed at the clinical phase and so far none
of them proved useful for the everyday practice.
While detailed analysis of such failure is well be-
yond the scope of this review, the authors feel ap-
propriate to note that more detailed understanding
and appreciation of cell death mechanisms and
experimental models applied should be a prereq-
uisite for successful development of neuroprotec-
tive agents. As far as study design is considered we
should appreciate that the rather complex and
versatile nature of TBI occurring in human is hard
to copy in any experimental model existing cur-
rently in our hands. In the overwhelming majority
of cases human TBI is a mixture of focal and
diffuse injuries complicated with secondary brain
injury primarily caused by hypoxia and hypoper-
fusion. The most frequently used models like con-
trolled cortical impact or moderate/severe forms of
fluid percussion brain injury produce mass tissue
destruction where strain and shearing-force distri-
bution leads to excessive influx and release of
Ca++, and excitatory amino acids, activation and
generation of ROS. Although such models may
mimic the situation in the most severe forms of
human TBI, they ignite such a bulk of various
pathways that could hardly be controlled with a
drug exerting its effect on a single target.
While diffuse brain injury models seem to pro-
vide a better environment to study more subtle
forms of TBI and theoretically provide a more
convenient field to investigate the effect of pharma-
ceutical agents, recent observations shed light on
the extremely complex and heterogenous nature of
both diffuse axonal and neuronal injury. Specifi-
cally, in different fiber populations considerably
different markers of axonal damage may appear
representing heterogeneity in the pathobiology as
well as in therapeutic efficacy (Stone et al., 2004).
Similarly, acceleration–deceleration injury evokes
various forms of neuronal-cellular damage indi-
cating a likewise complex field to target therapeu-
tically (Singleton and Povlishock, 2004; Farkas
et al., 2006).
In light of the above passages and the diversity
and interactivity of cell death pathways, a fatalistic
approach would be to conclude that when a cell is
committed to die due to an extrinsic or intrinsic
death signal it will do so regardless of whatever
therapeutic measures have been taken; inhibition
of apoptosis will ignite autophagic or necrotic
cell death, ultimately leading to cellular demise.
Nevertheless, the authors still feel appropriate to
address this very issue from a different point of
view with the believe that combined therapeutic
 91

approaches or ‘‘polyhparmacia’’ might success- References

fully target intermingled death pathways with re-
sulting improved morphological as well as Abe, Y., Yamamoto, T., Sugiyama, Y., Watanabe, T., Saito,
functional outcome. To this end one can envision N., Kayama, H. and Kumagai, T. (2004) ‘‘Anoikis’’ of
oligodendrocytes induced by Wallerian degeneration: ultra-
a scenario where ultra-early (pre-hospital) appli-
structural observations. J. Neurotrauma, 21: 119–124.
cation of therapeutic measures aimed to slow Abrahamson, E.E., Ikonomovic, M.D., Ciallella, J.R., Hope,
down proteolytic processes (such as hypothermia C.E., Paljug, W.R., Isanski, B.A., Flood, D.G., Clark, R.S.
treatment) and preserving energy homeostasis and DeKosky, S.T. (2006) Caspase inhibition therapy abol-
(PARP-inhibition, CsA-treatment) is followed or ishes brain trauma-induced increases in Abeta peptide: im-
plications for clinical outcome. Exp. Neurol., 197: 437–450.
joined by caspase and calpain inhibitors, or
Adams, J.M. and Cory, S. (2002) Apoptosomes: engines for
PACAP treatment. caspase activation. Curr. Opin. Cell Biol., 14: 715–720.
As far as the title of the present article is con- Aguilar, H.I., Botla, R., Arora, A.S., Bronk, S.F. and Gores,
sidered, to date it is hard to predict, whether in- G.J. (1996) Induction of the mitochondrial permeability
hibition of apoptosis is useful; one can only transition by protease activity in rats: a mechanism of he-
patocyte necrosis. Gastroenterology, 110: 558–566.
conclude that it is most probably beneficial in
Alves, O., Tolias, C.M., Lewis, C., Hayes, R.L., Choi, S.C.,
conjunction with other measures taken to inhibit Gilman, C., Enriquez, P., Povlishock, J.T. and Bullock, M.R.
various different pathways of traumatically in- (2003) Brain neurochemical alterations in patients with severe
duced neural demise. brain injury treated with cyclosporin-a. A placebo-controlled
study. Preliminary report. J. Neurotrauma, 20: 1125.
Arden, N. and Betenbaugh, M.J. (2004) Life and death in
Abbreviations mammalian cell culture: strategies for apoptosis inhibition.
Trends Biotechnol., 22: 174–180.
Arimura, A. (1998) Perspectives on pituitary adenylate cyclase
AIF apoptosis inducing factor activating polypeptide (PACAP) in the neuroendocrine, en-
CD cell death docrine, and nervous systems. Jpn. J. Physiol., 48: 301–331.
CsA cyclosporine A Assuncao, G.C. and Linden, R. (2004) Programmed cell deaths.
ER endoplasmatic reticulum Apoptosis and alternative deathstyles. Eur. J. Biochem., 271:
1638–1650.
IAP inhibitor of apoptosis protein
Bartus, R. (1997) The calpain hypothesis of neurodegeneration:
Icv intracerebroventricular (icv) evidence for a common cytotoxic pathway. Neuroscientist, 3:
MPTP mitochondrial permeability tran- 314–327.
sition pore Beer, R., Franz, G., Krajewski, S., Pike, B.R., Hayes, R.L.,
NMDA N-methyl-D-aspartate Reed, J.C., Wang, K.K., Klimmer, C., Schmutzhard, E., Po-
ewe, W. and Kampfl, A. (2001) Temporal and spatial profile
PARP-1 poly(ADP-ribose)polymerase 1
of caspase 8 expression and proteolysis after experimental
TAI traumatic axonal injury traumatic brain injury. J. Neurochem., 78: 862–873.
TBI traumatic brain injury Beer, R., Franz, G., Srinivasan, A., Hayes, R.L., Pike, B.R.,
TNF tumor necrosis factor Newcomb, J.K., Zhao, X., Schmutzhard, E., Poewe, W. and
Kampfl, A. (2000) Temporal profile and cell subtype distri-
bution of activated caspase-3 following experimental trau-
matic brain injury. J. Neurochem., 75: 1264–1273.
Besson, V.C., Croci, N., Boulu, R.G., Plotkine, M. and Marc-
Aknowledgments hand-Verrecchia, C. (2003) Deleterious poly(ADP-ribose)po-
lymerase-1 pathway activation in traumatic brain injury in
rat. Brain Res., 989: 58–66.
Supported by the Hungarian Science Found Besson, V.C., Zsengeller, Z., Plotkine, M., Szabo, C. and
(OTKA T048724) and the Bekesi Scholarship of Marchand-Verrecchia, C. (2005) Beneficial effects of PJ34
the Hungarian Academy of Sciences. The authors and INO-1001, two novel water-soluble poly(ADP-ribose)
thank for the invaluable advices and help of Pro- polymerase inhibitors, on the consequences of traumatic
brain injury in rat. Brain Res., 1041: 149–156.
fessor John T. Povlishock, Department of Anat-
Buki, A., Farkas, O., Doczi, T. and Povlishock, J.T. (2003)
omy and Neurobiology, Medical College of Preinjury administration of the calpain inhibitor MDL-28170
Virginia, Virginia Commonwealth University, attenuates traumatically induced axonal injury. J. Neuro-
Richmond, VA 23298-0709. trauma, 20: 261–268.
92
Buki, A., Koizumi, H. and Povlishock, J.T. (1999a) Moderate
posttraumatic hypothermia decreases early calpain-mediated
proteolysis and concomitant cytoskeletal compromise in
traumatic axonal injury. Exp. Neurol., 159: 319–328.
Buki, A., Okonkwo, D.O., Wang, K.K. and Povlishock, J.T.
(2000) Cytochrome c release and caspase activation in trau-
matic axonal injury. J. Neurosci., 20: 2825–2834.
Buki, A., Siman, R., Trojanowski, J.Q. and Povlishock, J.T.
(1999b) The role of calpain-mediated spectrin proteolysis in
traumatically induced axonal injury. J. Neuropathol. Exp.
Neurol., 58: 365–375.
Burkle, A. (2001) Physiology and pathophysiology of poly
(ADP-ribosyl)ation. Bioessays, 23: 795–806.
Cai, J. and Jones, D.P. (1998) Superoxide in apoptosis. Mito-
chondrial generation triggered by cytochrome c loss. J. Biol.
Chem., 273: 11401–11404.
Cai, J., Yang, J. and Jones, D.P. (1998) Mitochondrial control
of apoptosis: the role of cytochrome c. Biochim. Biophys.
Acta, 1366: 139–149.
Chinnaiyan, A.M., O’Rourke, K., Tewari, M. and Dixit, V.M.
(1995) FADD, a novel death domain-containing protein, in-
teracts with the death domain of Fas and initiates apoptosis.
Cell, 81: 505–512.
Clark, R.S., Kochanek, P.M., Watkins, S.C., Chen, M., Dixon,
C.E., Seidberg, N.A., Melick, J., Loeffert, J.E., Nathaniel,
P.D., Jin, K.L. and Graham, S.H. (2000) Caspase-3 mediated
neuronal death after traumatic brain injury in rats. J. Neuro-
chem., 74: 740–753.
Clifton, G.L. (2004) Is keeping cool still hot? An update on
hypothermia in brain injury. Curr. Opin. Crit. Care, 10:
116–119.
Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh,
T.K. (1998) Experimental brain injury induces regionally
distinct apoptosis during the acute and delayed post-trau-
matic period. J. Neurosci., 18: 5663–5672.
Cousin, M.A. and Robinson, P.J. (2001) The dephosphins:
dephosphorylation by calcineurin triggers synaptic vesicle
endocytosis. Trends Neurosci., 24: 659–665.
Curry III, J.J. and Heim, L.M. (1966) Brain myelination after
neonatal administration of oestradiol. Nature, 209: 915–916.
Dawson, T.M., Steiner, J.P., Dawson, V.L., Dinerman, J.L.,
Uhl, G.R. and Snyder, S.H. (1993) Immunosuppressant
FK506 enhances phosphorylation of nitric oxide synthase
and protects against glutamate neurotoxicity. Proc. Natl.
Acad. Sci. U.S.A., 90: 9808–9812.
Delgado, M., Leceta, J. and Ganea, D. (2003) Vasoactive in-
testinal peptide and pituitary adenylate cyclase-activating
polypeptide inhibit the production of inflammatory media-
tors by activated microglia. J. Leukoc. Biol., 73: 155–164.
Denecker, G., Vercammen, D., Declercq, W. and Vandena-
beele, P. (2001) Apoptotic and necrotic cell death induced by
death domain receptors. Cell Mol. Life Sci., 58: 356–370.
Di Giovanni, S., Movsesyan, V., Ahmed, F., Cernak, I., Schin-
elli, S., Stoica, B. and Faden, A.I. (2005) Cell cycle inhibition
provides neuroprotection and reduces glial proliferation and
scar formation after traumatic brain injury. Proc. Natl. Acad.
Sci. U.S.A., 102: 8333–8338.
Ditzel, M., Wilson, R., Tenev, T., Zachariou, A., Paul, A.,
Deas, E. and Meier, P. (2003) Degradation of DIAP1 by the
N-end rule pathway is essential for regulating apoptosis. Nat.
Cell Biol., 5: 467–473.
Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechano-
poration induced by diffuse traumatic brain injury: an irre-
versible or reversible response to injury? J. Neurosci., 26:
3130–3140.
Farkas, O., Tamas, A., Zsombok, A., Reglodi, D., Pal, J., Buki,
A., Lengvari, I., Povlishock, J.T. and Doczi, T. (2004) Effects
of pituitary adenylate cyclase activating polypeptide in a rat
model of traumatic brain injury. Regul. Pept., 123: 69–75.
Ferri, K.F. and Kroemer, G. (2001) Organelle-specific initiation
of cell death pathways. Nat. Cell Biol., 3: E255–E263.
Gabbita, S.P., Scheff, S.W., Menard, R.M., Roberts, K., Fug-
accia, I. and Zemlan, F.P. (2005) Cleaved-tau: a biomarker of
neuronal damage after traumatic brain injury. J. Neurotrau-
ma, 22: 83–94.
Gallyas, F., Pal, J., Farkas, O. and Doczi, T. (2006) The fate of
axons subjected to traumatic ultrastructural (neurofilament)
compaction: an electron-microscopic study. Acta Neuro-
pathol. (Berl.), 111: 229–237.
Garcia-Estrada, J., Luquin, S., Fernandez, A.M. and Garcia-
Segura, L.M. (1999) Dehydroepiandrosterone, pregnenolone
and sex steroids down-regulate reactive astroglia in the male
rat brain after a penetrating brain injury. Int. J. Dev. Neuro-
sci., 17: 145–151.
Ghajar, J. (2000) Traumatic brain injury. Lancet, 356: 923–929.
Ginis, I., Hallenbeck, J.M., Liu, J., Spatz, M., Jaiswal, R. and
Shohami, E. (2000) Tumor necrosis factor and reactive ox-
ygen species cooperative cytotoxicity is mediated via inhibi-
tion of NF-kappaB. Mol. Med., 6: 1028–1041.
Giza, C.C. and Hovda, D.A. (2001) The neurometabolic cas-
cade of concussion. J. Athl. Train, 36: 228–235.
Gu, F., Hata, R., Toku, K., Yang, L., Ma, Y.J., Maeda, N.,
Sakanaka, M. and Tanaka, J. (2003) Testosterone up-
regulates aquaporin-4 expression in cultured astrocytes.
J. Neurosci. Res., 72: 709–715.
Hortobagyi, T., Gorlach, C., Benyo, Z., Lacza, Z., Hortobagyi,
S., Wahl, M. and Harkany, T. (2003) Inhibition of neuronal
nitric oxide synthase-mediated activation of poly(ADP-rib-
ose) polymerase in traumatic brain injury: neuroprotection
by 3-aminobenzamide. Neuroscience, 121: 983–990.
Jurgensmeier, J.M., Xie, Z., Deveraux, Q., Ellerby, L., Brede-
sen, D. and Reed, J.C. (1998) Bax directly induces release of
cytochrome c from isolated mitochondria. Proc. Natl. Acad.
Sci. U.S.A., 95: 4997–5002.
Kampfl, A., Posmantur, R.M., Zhao, X., Schmutzhard, E.,
Clifton, G.L. and Hayes, R.L. (1997) Mechanisms of calpain
proteolysis following traumatic brain injury: implications for
pathology and therapy: implications for pathology and ther-
apy: a review and update. J. Neurotrauma, 14: 121–134.
Katahira, M., Yone, K. and Arishima, Y. (2003) The neuro-
protective effects of PACAP on spinal cor injury (SCI) in
rats. Regul. Pept., 15: 49.
Kawamura, M., Nakajima, W., Ishida, A., Ohmura, A., Miura,
S. and Takada, G. (2005) Calpain inhibitor MDL 28170

protects hypoxic-ischemic brain injury in neonatal rats by
inhibition of both apoptosis and necrosis. Brain Res., 1037:
59–69.
Kim, W.K., Kan, Y., Ganea, D., Hart, R.P., Gozes, I. and
Jonakait, G.M. (2000) Vasoactive intestinal peptide and pi-
tuitary adenylyl cyclase-activating polypeptide inhibit tumor
necrosis factor-alpha production in injured spinal cord and
in activated microglia via a cAMP-dependent pathway.
J. Neurosci., 20: 3622–3630.
Knoblach, S.M., Alroy, D.A., Nikolaeva, M., Cernak, I., Sto-
ica, B.A. and Faden, A.I. (2004) Caspase inhibitor z-DEVD-
fmk attenuates calpain and necrotic cell death in vitro and
after traumatic brain injury. J. Cereb. Blood Flow Metab.,
24: 1119–1132.
Knoblach, S.M., Nikolaeva, M., Huang, X., Fan, L., Krajew-
ski, S., Reed, J.C. and Faden, A.I. (2002) Multiple caspases
are activated after traumatic brain injury: evidence for in-
volvement in functional outcome. J. Neurotrauma, 19:
1155–1170.
Komjati, K., Besson, V.C. and Szabo, C. (2005) Poly (ADP-
ribose) polymerase inhibitors as potential therapeutic agents
in stroke and neurotrauma. Curr. Drug Targets CNS Neurol.
Disord., 4: 179–194.
Kong, L.Y., Maderdrut, J.L., Jeohn, G.H. and Hong, J.S.
(1999) Reduction of lipopolysaccharide-induced neurotoxic-
ity in mixed cortical neuron/glia cultures by femtomolar
concentrations of pituitary adenylate cyclase-activating poly-
peptide. Neuroscience, 91: 493–500.
Kroemer, G. (2003) Mitochondrial control of apoptosis: an in-
troduction. Biochem. Biophys. Res. Commun., 304: 433–435.
Kruman, I.I. and Mattson, M.P. (1999) Pivotal role of mito-
chondrial Ca++ uptake in neural cell apoptosis and necrosis.
J. Neurochem., 72: 529–540.
Kuroda, S., Janelidze, S. and Siesjo, B.K. (1999) The immuno-
suppressants cyclosporin A and FK506 equally ameliorate
brain damage due to 30-min middle cerebral artery occlusion
in hyperglycemic rats. Brain Res., 835: 148–153.
Lacza, Z., Horvath, E.M., Komjati, K., Hortobagyi, T., Szabo,
C. and Busija, D.W. (2003) PARP inhibition improves the
effectiveness of neural stem cell transplantation in experi-
mental brain trauma. Int. J. Mol. Med., 12: 153–159.
Lang-Rollin, I.C., Rideout, H.J., Noticewala, M. and Stefanis,
L. (2003) Mechanisms of caspase-independent neuronal
death: energy depletion and free radical generation. J. Neuro-
sci., 23: 11015–11025.
LaPlaca, M.C., Zhang, J., Raghupathi, R., Li, J.H., Smith, F.,
Bareyre, F.M., Snyder, S.H., Graham, D.I. and McIntosh,
T.K. (2001) Pharmacologic inhibition of poly(ADP-ribose)
polymerase is neuroprotective following traumatic brain in-
jury in rats. J. Neurotrauma, 18: 369–376.
Larner, S.F., Hayes, R.L., McKinsey, D.M., Pike, B.R. and
Wang, K.K. (2004) Increased expression and processing of
caspase-12 after traumatic brain injury in rats. J. Neuro-
chem., 88: 78–90.
Larner, S.F., McKinsey, D.M., Hayes, R.L. and Wang, K.K.
(2005) Caspase 7: increased expression and activation after
traumatic brain injury in rats. J. Neurochem., 94: 97–108.
93
Lewin, I.C.F. (1991) Update of Max, W., Mackenzie, E.J.,
Rice, D.P. Head injuries: costs and consequences. J. Head
Trauma Rehabil., 67: 76–91.
Liu, J., Farmer Jr., J.D., Lane, W.S., Friedman, J., Weissman,
I. and Schreiber, S.L. (1991) Calcineurin is a common target
of cyclophilin-cyclosporin A and FKBP-FK506 complexes.
Cell, 66: 807–815.
Liu, M.C., Akle, V., Zheng, W., Dave, J.R., Tortella, F.C.,
Hayes, R.L. and Wang, K.K. (2006) Comparing calpain- and
caspase-3-mediated degradation patterns in traumatic brain in-
jury by differential proteome analysis. Biochem. J., 394:
715–725.
Lockshin, R.A. and Zakeri, Z. (2002) Caspase-independent cell
deaths. Curr. Opin. Cell Biol., 14: 727–733.
Lockshin, R.A. and Zakeri, Z. (2004) Apoptosis, autophagy,
and more. Int. J. Biochem. Cell Biol., 36: 2405–2419.
MacGregor, D.G., Avshalumov, M.V. and Rice, M.E. (2003)
Brain edema induced by in vitro ischemia: causal factors and
neuroprotection. J. Neurochem., 85: 1402–1411.
Markgraf, C.G., Velayo, N.L., Johnson, M.P., McCarty, D.R.,
Medhi, S., Koehl, J.R., Chmielewski, P.A. and Linnik, M.D.
(1998) Six-hour window of opportunity for calpain in-
hibition in focal cerebral ischemia in rats. Stroke, 29:
152–158.
Miyata, A., Arimura, A., Dahl, R.R., Minamino, N., Uehara,
A., Jiang, L., Culler, M.D. and Coy, D.H. (1989) Isolation of
a novel 38 residue-hypothalamic polypeptide which stimu-
lates adenylate cyclase in pituitary cells. Biochem. Biophys.
Res. Commun., 164: 567–574.
Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A.,
Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C., Clif-
ton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D.,
Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D.,
Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller,
N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W.,
Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E.,
Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P.,
Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G.,
Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich,
M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L.
(2002) Clinical trials in head injury. J. Neurotrauma, 19:
503–557.
Okonkwo, D.O., Melon, D.E., Pellicane, A.J., Mutlu, L.K.,
Rubin, D.G., Stone, J.R. and Helm, G.A. (2003) Dose-
response of cyclosporin A in attenuating traumatic axonal
injury in rat. Neuroreport, 14: 463–466.
Okonkwo, D.O., Pettus, E.H., Moroi, J. and Povlishock, J.T.
(1998) Alteration of the neurofilament sidearm and its rela-
tion to neurofilament compaction occurring with traumatic
axonal injury. Brain Res., 784: 1–6.
Okonkwo, D.O. and Povlishock, J.T. (1999) An intrathecal
bolus of cyclosporin A before injury preserves mitochondrial
integrity and attenuates axonal disruption in traumatic brain
injury. J. Cereb. Blood Flow Metab., 19: 443–451.
Oyadomari, S., Araki, E. and Mori, M. (2002) Endoplasmic
reticulum stress-mediated apoptosis in pancreatic beta-cells.
Apoptosis, 7: 335–345.
94
Pant, H.C. (1988) Dephosphorylation of neurofilament proteins
enhances their susceptibility to degradation by calpain. Bio-
chem. J., 256: 665–668.
Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock,
J.T. (1994) Traumatically induced altered membrane perme-
ability: its relationship to traumatically induced reactive axo-
nal change. J. Neurotrauma, 11: 507–522.
Povlishock, J.T., Marmarou, A., McIntosh, T., Trojanowski,
J.Q. and Moroi, J. (1997) Impact acceleration injury in the
rat: evidence for focal axolemmal change and related neuro-
filament sidearm alteration. J. Neuropathol. Exp. Neurol.,
56: 347–359.
Pyrzynska, B., Lis, A., Mosieniak, G. and Kaminska, B. (2001)
Cyclosporin A-sensitive signaling pathway involving calcine-
urin regulates survival of reactive astrocytes. Neurochem.
Int., 38: 409–415.
Reed, J.C. (1998) Bcl-2 family proteins. Oncogene, 17:
3225–3236.
Reglodi, D., Tamas, A., Somogyvari-Vigh, A., Szanto, Z.,
Kertes, E., Lenard, L., Arimura, A. and Lengvari, I. (2002)
Effects of pretreatment with PACAP on the infarct size and
functional outcome in rat permanent focal cerebral ischemia.
Peptides, 23: 2227–2234.
Riess, P., Bareyre, F.M., Saatman, K.E., Cheney, J.A., Lifshitz,
J., Raghupathi, R., Grady, M.S., Neugebauer, E. and McIn-
tosh, T.K. (2001) Effects of chronic, post-injury Cyclosporin
A administration on motor and sensorimotor function fol-
lowing severe, experimental traumatic brain injury. Restor.
Neurol. Neurosci., 18: 1–8.
Ringger, N.C., O’Steen, B.E., Brabham, J.G., Silver, X.,
Pineda, J., Wang, K.K., Hayes, R.L. and Papa, L. (2004)
A novel marker for traumatic brain injury: CSF alphaII-
spectrin breakdown product levels. J. Neurotrauma, 21:
1443–1456.
Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neuge-
bauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic
cell death after experimental traumatic brain injury in the rat.
Am. J. Pathol., 147: 1575–1583.
Roof, R.L. and Hall, E.D. (2000) Gender differences in acute
CNS trauma and stroke: neuroprotective effects of estrogen
and progesterone. J. Neurotrauma, 17: 367–388.
Saatman, K.E., Murai, H., Bartus, R.T., Smith, D.H., Hay-
ward, N.J., Perri, B.R. and McIntosh, T.K. (1996) Calpain
inhibitor AK295 attenuates motor and cognitive deficits fol-
lowing experimental brain injury in the rat. Proc. Natl. Acad.
Sci. U.S.A., 93: 3428–3433.
Saatman, K.E., Zhang, C., Bartus, R.T. and McIntosh, T.K.
(2000) Behavioral efficacy of posttraumatic calpain inhibi-
tion is not accompanied by reduced spectrin proteolysis,
cortical lesion, or apoptosis. J. Cereb. Blood Flow Metab.,
20: 66–73.
Saikumar, P., Dong, Z., Weinberg, J.M. and Venkatachalam,
M.A. (1998) Mechanisms of cell death in hypoxia/reoxygen-
ation injury. Oncogene, 17: 3341–3349.
Saltman, R.B. and Findling, R.L. (1997) European Health Care
Reform. Analysis of current strategies. In: WHO Regional
Publications European Series 72 Copenhagen: WHO.
Schendel, S.L., Montal, M. and Reed, J.C. (1998) Bcl-2 family
proteins as ion-channels. Cell Death Differ., 5: 372–380.
Scorrano, L. and Korsmeyer, S.J. (2003) Mechanisms of cyto-
chrome c release by proapoptotic BCL-2 family members.
Biochem. Biophys. Res. Commun., 304: 437–444.
Siesjo, B.K., Hu, B. and Kristian, T. (1999) Is the cell death
pathway triggered by the mitochondrion or the endoplasmic
reticulum? J. Cereb. Blood Flow Metab., 19: 19–26.
Singleton, R.H. and Povlishock, J.T. (2004) Identification and
characterization of heterogeneous neuronal injury and death
in regions of diffuse brain injury: evidence for multiple in-
dependent injury phenotypes. J. Neurosci., 24: 3543–3553.
Singleton, R.H., Stone, J.R., Okonkwo, D.O., Pellicane, A.J.
and Povlishock, J.T. (2001) The immunophilin ligand FK506
attenuates axonal injury in an impact acceleration model of
traumatic brain injury. J. Neurotrauma, 18: 607–614.
Skoglosa, Y., Lewen, A., Takei, N., Hillered, L. and Lindholm,
D. (1999) Regulation of pituitary adenylate cyclase activating
polypeptide and its receptor type 1 after traumatic brain in-
jury: comparison with brain-derived neurotrophic factor and
the induction of neuronal cell death. Neuroscience, 90:
235–247.
Soustiel, J.F., Palzur, E., Nevo, O., Thaler, I. and Vlodavsky,
E. (2005) Neuroprotective anti-apoptosis effect of estrogens
in traumatic brain injury. J. Neurotrauma, 22: 345–352.
Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu,
L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired ax-
onal transport and altered axolemmal permeability occur in
distinct populations of damaged axons following traumatic
brain injury. Exp. Neurol., 190: 59–69.
Stys, P.K. and Jiang, Q. (2002) Calpain-dependent neurofila-
ment breakdown in anoxic and ischemic rat central axons.
Neurosci. Lett., 328: 150–154.
Suehiro, E., Singleton, R.H., Stone, J.R. and Povlishock, J.T.
(2001) The immunophilin ligand FK506 attenuates the ax-
onal damage associated with rapid rewarming following
posttraumatic hypothermia. Exp. Neurol., 172: 199–210.
Sullivan, P.G., Keller, J.N., Bussen, W.L. and Scheff, S.W.
(2002) Cytochrome c release and caspase activation after
traumatic brain injury. Brain Res., 949: 88–96.
Susin, S.A., Lorenzo, H.K., Zamzami, N., Marzo, I., Snow,
B.E., Brothers, G.M., Mangion, J., Jacotot, E., Costantini,
P., Loeffler, M., Larochette, N., Goodlett, D.R., Aebersold,
R., Siderovski, D.P., Penninger, J.M. and Kroemer, G.
(1999) Molecular characterization of mitochondrial apopto-
sis-inducing factor. Nature, 397: 441–446.
Susin, S.A., Zamzami, N. and Kroemer, G. (1998) Mito-
chondria as regulators of apoptosis: doubt no more. Biochim.
Biophys. Acta, 1366: 151–165.
Tabuchi, A., Funaji, K., Nakatsubo, J., Fukuchi, M., Tsuchiya,
T. and Tsuda, M. (2003) Inactivation of aconitase during the
apoptosis of mouse cerebellar granule neurons induced by a
deprivation of membrane depolarization. J. Neurosci. Res.,
71: 504–515.
Tamas, A., Zsombok, A., Farkas, O., Reglodi, D., Pal, J., Buki,
A., Lengvari, I., Povlishock, J.T. and Doczi, T. (2006) Post-
injury administration of pituitary adenylate cyclase activating

polypeptide (PACAP) attenuates traumatically induced ax-
onal injury in rats. J. Neurotrauma, 23: 686–695.
Thurman, D.J., Alverson, C., Dunn, K.A., Guerrero, J. and
Sniezek, J.E. (1999) Traumatic brain injury in the United
States: A public health perspective. J. Head Trauma Re-
habil., 14: 602–615.
Van Den, H.C., Donkin, J.J., Finnie, J.W., Blumbergs, P.C.,
Kuchel, T., Koszyca, B., Manavis, J., Jones, N.R., Reilly, P.L.
and Vink, R. (2004) Downregulation of amyloid precursor
protein (APP) expression following post-traumatic cyclo-
sporin-A administration. J. Neurotrauma, 21: 1562–1572.
Varshavsky, A. (2003) The N-end rule and regulation of
apoptosis. Nat. Cell Biol., 5: 373–376.
Vaudry, D., Cottet-Rousselle, C., Basille, M., Falluel-Morel,
A., Fournier, A., Vaudry, H. and Gonzalez, B.J. (2004) Pi-
tuitary adenylate cyclase-activating polypeptide inhibits ca-
spase-3 activity but does not protect cerebellar granule
neurons against beta-amyloid (25–35)-induced apoptosis.
Regul. Pept., 123: 43–49.
Vaudry, D., Gonzalez, B.J., Basille, M., Yon, L., Fournier, A.
and Vaudry, H. (2000) Pituitary adenylate cyclase-activating
polypeptide and its receptors: from structure to functions.
Pharmacol. Rev., 52: 269–324.
Vukic, M., Negovetic, L., Kovac, D., Ghajar, J., Glavic, Z. and
Gopcevic, A. (1999) The effect of implementation of guide-
lines for the management of severe head injury on patient
95
treatment and outcome. Acta Neurochir. (Wien.), 141:
1203–1208.
Wang, K.K., Posmantur, R., Nath, R., McGinnis, K., Whitton,
M., Talanian, R.V., Glantz, S.B. and Morrow, J.S. (1998)
Simultaneous degradation of alphaII- and betaII-spectrin by
caspase 3 (CPP32) in apoptotic cells. J. Biol. Chem., 273:
22490–22497.
Warren, M.W., Kobeissy, F.H., Liu, M.C., Hayes, R.L., Gold,
M.S. and Wang, K.K. (2005) Concurrent calpain and ca-
spase-3 mediated proteolysis of alpha II-spectrin and tau in
rat brain after methamphetamine exposure: a similar profile
to traumatic brain injury. Life Sci., 78: 301–309.
Yakovlev, A.G., Knoblach, S.M., Fan, L., Fox, G.B., Good-
night, R. and Faden, A.I. (1997) Activation of CPP32-like
caspases contributes to neuronal apoptosis and neurological
dysfunction after traumatic brain injury. J. Neurosci., 17:
7415–7424.
Yamashima, T. (2004) Ca2+-dependent proteases in ischemic
neuronal death: a conserved ‘calpain-cathepsin cascade’ from
nematodes to primates. Cell Calcium, 36: 285–293.
Zong, W.X., Li, C., Hatzivassiliou, G., Lindsten, T., Yu, Q.C.,
Yuan, J. and Thompson, C.B. (2003) Bax and Bak can lo-
calize to the endoplasmic reticulum to initiate apoptosis. J.
Cell Biol., 162: 59–69.
Zoratti, M. and Szabo, I. (1995) The mitochondrial permeabi-
lity transition. Biochim. Biophys. Acta, 1241: 139–176.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 7

Substance P in traumatic brain injury

James J. Donkin1, Renee J. Turner1, Islam Hassan1 and Robert Vink1,2,

1
Discipline of Pathology, University of Adelaide, Adelaide, South Australia
2
Centre for Neurological Diseases, The Hanson Institute, Adelaide, South Australia

Abstract: Recent evidence has suggested that neuropeptides, and in particular substance P (SP), may play a
critical role in the development of morphological injury and functional deficits following acute insults to the
brain. Few studies, however, have examined the role of SP, and more generally, neurogenic inflammation,
in the pathophysiology of traumatic brain injury and stroke. Those studies that have been reported suggest
that SP is released following injury to the CNS and facilitates the increased permeability of the blood brain
barrier, the development of vasogenic edema and the subsequent cell death and functional deficits that are
associated with these events. Inhibition of the SP activity, either through inhibition of the neuropeptide
release or the use of SP receptor antagonists, have consistently resulted in profound decreases in edema
formation and marked improvements in functional outcome. The current review summarizes the role of SP
in acute brain injury, focussing on its properties as a neurotransmitter and the potential for SP to adversely
affect outcome.

Keywords: neurotrauma; traumatic brain injury; edema; neuropeptides

Introduction such as helmets, airbags and seatbelts, little can be

Traumatic brain injury is the leading cause of
death and disability in people under the age of 40
years (Fleminger and Ponsford, 2005) with inci-
dence rates estimated at 150–250 cases per 100,000
populations per year (Leon-Carrion et al., 2005).
The cost for rehabilitation and care of such indi-
viduals to the community runs into billion of dol-
lars annually. Despite the enormity of this public
health problem, no effective treatment currently
exists. It is now accepted that brain injury results
in the development of neurologic deficits through
two main mechanisms. Firstly, the primary event
includes the mechanical processes such as shear-
ing, laceration and stretching of nerve fibres that
occurs at the time of the injury (Graham et al.,
1992, 1996). Besides the use of preventive measures
Corresponding author. Tel.: +61-8–8222-3092; Fax: +61-8-
8222-3093; E-mail: Robert.Vink@adelaide.edu.au
done to prevent primary injury, and such injury
may be regarded as irreversible. In contrast, seco-
ndary injury is made up of the delayed biochemical
and physiological factors that are initiated by the
primary event, and these secondary injury factors
are thought to account for much of the morbidity
following brain injury (McIntosh et al., 1996). This
secondary injury cascade evolves over minutes to
days and even months after the initial event, and
as such, there are opportunities for interventional
pharmacology to prevent further injury and im-
prove outcome. As a result, research has focused
on the identification of secondary injury factors
and the development of novel therapies that
attenuate, or even prevent, their action.
A number of secondary injury factors have been
identified to date including blood brain barrier
(BBB) opening, edema formation, release of neuro-
transmitters such as excitatory amino acids, ion
changes, oxidative stress and bioenergetic failure,

DOI: 10.1016/S0079-6123(06)61007-8 97
98
amongst others. At the cellular level, the initial
effect of mechanical impact is to increase the se-
lective permeability of the cell membrane and this
occurs to varying degrees depending on the severity
of injury. This effect, known as mechanoporation
(Gennarelli and Graham, 1998), allows for the in-
creased movement of ions into and out of cells
along their natural concentration gradients. Thus,
calcium (Ca2+), sodium (Na+) and chloride (Cl
)
ions enter cells whilst potassium (K+) and magne-
sium (Mg2+) ions are lost from the cells. From this
point, the pathological changes might be con-
sidered to differentiate into two subroutines ac-
cording to whether these alterations in ion
concentration cause effects due to their chemical
properties (the enzymatic subroutine) or due to
their physical properties (the osmotic subroutine).
The enzymatic subroutine revolves around the
influx of calcium ions, which activates several cel-
lular enzyme cascades. These enzyme cascades me-
diate cellular dysfunction, including activation of
calpains, axonal injury, accumulation of free rad-
ical species, increased production of nitric oxide
and induction of proinflammatory gene expres-
sion, which can potentially culminate in cell death
(Obrenovitch and Urenjak, 1997; Xiong et al.,
1997; Vespa et al., 1998). Among these different
mechanisms of delayed cell damage in TBI, in-
flammation is the predominant mechanism in the
case of contusions (Graham et al., 2002). The in-
flammatory reaction consists of various compo-
nents that evolve at their own specific rate and
according to their own specific pattern as the age
of the lesion increases (Oehmichen and Raff, 1980;
Oehmichen et al., 1986; Cervos-Navarro and
Lafuente, 1991). For example, in terms of inflam-
matory cell infiltration, several microscopic studies
of human injury have demonstrated a distinct time
course (Holmin et al., 1998; Hausmann et al.,
1999; Engel et al., 2000). In lesions aged up to 24 h,
the cellular component of inflammation was
represented by margination of neutrophils (also
referred to polymorphonuclear leukocytes or
PMNLs) in the vessels, whereas at 3–5 days of
survival, the inflammatory cell reaction consisted
of tissue infiltration of not only neutrophils, but
also monocyte/macrophages and CD4- and CD8-
positive T-lymphocytes, as well as an activation of
resident microglia. Changes in inflammatory cells
are paralleled by proliferation of astrocytes
(Hausmann and Betz, 2000), proliferation of cap-
illaries, swelling of their endothelium and by the
formation of perivascular edema (Bullock et al.,
1991; Vaz et al., 1997). The changes often culmi-
nate in a gliotic scar studded with hemosiderin-
laden macrophages.
The osmotic subroutine occurs because the net
influx of ions is much greater than the net efflux of
ions. Consequently, water is osmotically obligated
to follow the passage of ions into cells. This leads
to cellular swelling, referred to as cytotoxic edema.
Glia also swell due to the fact that they function in
the uptake of the K+ accumulating in the extra-
cellular fluid (Reilly, 2001). This glial swelling may
further compromise cerebral perfusion by com-
pressing the small blood vessels running amidst the
glial cells. Alternatively, water may be obligated to
follow an osmotic gradient generated by the pas-
sage of proteins and ions from the vasculature to
the brain interstitium. This edema is known as
vasogenic edema and is associated with an in-
creased permeability of the BBB, best observed in
the first 5 h after the TBI (O’Connor et al., 2003).
The microvasculature in the injury zone is affected
such that capillaries exhibit increased permeability
and arterioles lose their capacity to regulate blood
flow (Dietrich et al., 1994). Although the exact
mechanisms of BBB disruption are unknown, it is
hypothesised that inflammatory mediators play a
role, possibly through receptor-mediated actions.
Among these inflammatory mediators, neuropep-
tides such as substance P (SP), released from peri-
vascular axons, are prime candidates.
It is clear that the development of edema is
common to both the enzymatic and osmotic sub-
routines of injury following TBI, and its adverse
consequences on outcome through effects on in-
tracranial pressure (ICP) have been well described
(Marmarou et al., 2000). Current protocols for the
management of raised ICP include pharmacologi-
cal regimens such as administration of hyperos-
motic agents and barbiturates, or induction
of hyperventilation and hypothermia, as well as
surgical procedures such as drainage of cerebro-
spinal fluid (CSF) and decompressive craniotomy
(Graham et al., 2002). Unfortunately, in terms of

improving patient survival rates and functional
outcome, these interventions have essentially been
inadequate, largely because they do not address
the fundamental issue of what specific mechanisms
are associated with edema development after TBI.
Recent studies have suggested that neuropeptides,
and in particular SP, may play a critical role in
edema formation, not only in terms of vasogenic
edema associated with increased BBB permeabil-
ity, but also in the later cytotoxic phase of edema
development (Nimmo et al., 2004). Its involvement
in the pathophysiology of TBI therefore seems to
straddle both the enzymatic and osmotic subrou-
tines of injury.
Substance P
SP was first identified in the early part of 1930
(Von Euler and Gaddum, 1931), initially as a crude
extract isolated from equine brain and gut. The
letter P derives from the ‘powder’ they extracted
that contained the active substance. It was found to
have potent hypotensive and smooth muscle
contractile properties (Von Euler and Gaddum,
1931), and was identified in high concentrations in
the dorsal root of the spinal cord, leading to the
proposal that it was a neuronal sensory transmitter
associated with pain transmission (Lembeck,
1953). Today, it is accepted that SP is released
from both central and peripheral endings of pri-
mary afferent neurons and functions as a neuro-
transmitter (Otsuka and Yoshioka, 1993).
Susan Leeman and colleagues (Chang et al.,
1971) identified SP as an undecapeptide in the early
1970s, and were the first to synthesise the com-
pound (Tregear et al., 1971) and set up radio-
immunoassays (Powell et al., 1973). Such advances
allowed the effects of SP to be tested in physiolog-
ical models (Takahashi et al., 1974; Henry, 1976),
using antibodies to monitor SP by radioimmuno-
assay and immunohistochemistry (Hokfelt et al.,
1975; Nilsson et al., 1975; Takahashi and Otsuka,
1975; Cuello and Kanazawa, 1978; Ljungdahl et al.,
1978; Costa et al., 1980; Schultzberg et al., 1980)
and demonstrating controlled neuronal release
(Otsuka and Konishi, 1976; Olgart et al., 1977).
It’s role as a neurotransmitter was subsequently
99
widely accepted (Otsuka and Takahashi, 1977;
Nicoll et al., 1980; Pernow, 1983; Otsuka and
Yoshioka, 1993). In terms of structure, the fact that
certain neuropeptides share specific amino acid
sequences allows this vast collection of molecules
to be sorted into families, such as the tachykinin
family which includes SP, calcitonin gene-related
peptide (CGRP) and neurokinin A (NKA). The
tachykinin peptide family certainly represents one
of the largest peptide families described in animals
(Severini et al., 2002).
Synthesis
The formation of the peptide bonds of neuropep-
tides necessitates that they be synthesised on ribo-
somes, structures present exclusively in the cell
body. It is common for the mRNA encoding
neuropeptides to initially be translated into a
larger protein precursor. Two genes exist, the
preprotachykinin A (PPTA) gene and PPTB gene.
The PPTA gene can express four different forms of
mRNA through alternative splicing, two of which
(the b and g forms) encode synthesis of both SP
and NKA. Expression of SP and its mRNA is
widely abundant in both CNS and PNS (Harrison
and Geppetti, 2001). aPPTA expression is more
abundant in the brain, whilst bPPT-A and gPPTA
mRNAs predominate in peripheral tissues (Kotani
et al., 1986). The b and g forms of PPTA mRNA
also encode synthesis of neuropeptide K (NPK)
and neuropeptide g (NPg), which are elongated
forms of NKA. However their function has not
been fully elucidated. The PPTB gene gives rise to
neurokinin B (NKB) (Hokfelt et al., 2001).
Localisation
SP immunoreactivity has been demonstrated in
the rhinencephalon, telencephalon, basal ganglia,
hippocampus, amygdala, septal areas, diencepha-
lon, hypothalamus, mesencephalon, metencepha-
lon, pons, myelencephalon and spinal cord
(Shults et al., 1984). Nerve fibres containing SP-
like immunoreactivity are common in most
autonomic ganglia (Helke et al., 1982; Helke and
Phillips, 1988; Bergner et al., 2000), and SP
100
immunoreactivity has also been described in trige-
minal and dorsal root ganglia (Lee et al., 1985;
Gibbins et al., 1987) and intrinsic neurons of the
gut (Sternini et al., 1995). In the autonomic ganglia
SP is thought to play a modulatory role, the best
characterised response being observed in guinea pig
inferior mesenteric ganglion. In these neurons SP
mimics a slow depolarisation, which can be evoked
by repetitive afferent nerve stimulation (Dun and
Minota, 1981). Peripheral inflammation also leads
to an increase in SP immunoreactivity within the
superficial spinal cord (Marlier et al., 1991) and
increased SP release (Schaible et al., 1990). Acti-
vation or damage to neurons leads to changes
in neuropeptide biosynthesis that results from
induction of neuropeptide gene expression
(Hokfelt et al., 1994). Specifically, the expression
of PPT mRNA (Noguchi et al., 1988) and
SP (NK1) receptor mRNA (McCarson, 1999) are
upregulated in the periphery during noxious stim-
ulation or neurogenic inflammation (Harrison and
Geppetti, 2001).
SP is released from its precursor by the actions
of proteases called convertases. Cleavage points
for the convertases on the PPT gene are doublets
of cationic residues (Harrison and Geppetti, 2001).
After release, the only mechanisms to terminate
the action of neuropeptides are diffusion away
from the receptor site or degradation by extracel-
lular peptidases. The slow nature of these proc-
esses accounts for the prolonged effects of
neuropeptides (Kandel and Squire, 2000).
Metabolism
Enzymes involved in metabolising SP include neu-
tral endopeptidase (NEP) (Matsas et al., 1984),
SP-degrading enzyme (Probert and Hanley, 1987),
angiotensin-converting enzyme (ACE) (Skidgel
and Erdos, 1987), dipeptidyl aminopeptide IV
(Heymann and Mentlein, 1978), post-proline end-
opeptidase (Blumberg et al., 1980), cathepsin-D
(Azaryan and Galoyan, 1988) and cathepsin-E
(Kageyama, 1993). While all of these enzymes
are known to cleave SP in vitro, their individual
cellular localisation suggests that NEP and/or
ACE are most likely to be involved in the
cleavage of SP in vivo (Nadel, 1991). NEP has
been demonstrated to metabolise SP in the brain
(Hooper and Turner, 1987), spinal cord (Sakurada
et al., 1990) and peripheral tissues (Di Maria et al.,
1998), while ACE has been reported to degrade SP
in plasma (Wang et al., 1991), CSF and substantia
nigra. ACE has also been shown to contribute to
the degradation of fragments released from NEP.
Both NEP and ACE catalyse the hydrolysis bonds
of SP, leaving the peptide lacking the carboxyl
terminal regions required to bind to the tachykinin
receptors (Skidgel and Erdos, 1987).
Receptors
The biological actions of SP are mediated by
tachykinin (neurokinin: NK) receptors (Harrison
and Geppetti, 2001), rhodopsin-like membrane
structures consisting of seven hydrophobic trans-
membrane domains, connected by extracellular
and intracellular loops and coupled to G-proteins
(Nakanishi, 1991; Gerard et al., 1993; Maggi and
Schwartz, 1997). There are three types of mam-
malian tachykinin receptors that have been cloned:
NK1, NK2 and NK3 exhibiting preferences for SP,
NKA and NKB, respectively (Regoli et al., 1994).
Endogenous tachykinins are not highly selective
for any given receptor, and may act on all three
receptors with varying affinities under certain
conditions such as receptor availability or high
peptide concentrations. For this reason SP acti-
vates not only NK1 receptors, but also NK2
and NK3 receptors in a number of tissues (Regoli
et al., 1994).
Functions
SP has been implicated in memory and reinforce-
ment processes. The amino terminus (NH2) of SP
has been found to be involved in the memory pro-
moting effects of SP while the carboxy terminus
(COOH) is involved in reinforcing properties. An
NK1 antagonist (WIN 51708) was found to block
these actions, indicating that the behavioural effects
of SP are mediated by NK1 receptor (Hasenohrl
et al., 2000). SP is also expressed widely in areas of
the brain involved in fear producing pathways,

including the amygdala, septum, hippocampus,
hypothalamus and periaqueductal gray (Rupniak
and Kramer, 1999), and accordingly is released in
response to aversive stimuli. As expected, injection
of SP into regions such as the periaqueductal gray
modulates defensive reactions, while administration
of an NK1 antagonist inhibits long-lasting audible
vocalisation, which is a sign of anxiety and fear
(Severini et al., 2002). Intracerebroventricular
injection of SP results in many diverse effects in
rodents including increased blood pressure and
heart rate, increased hindlimb rearing behaviour,
scratching, skin biting and grooming.
In bronchial smooth muscle, activation of sen-
sory neurons leads to vasodilation and increased
vascular permeability, effects that are abolished by
pretreatment with capsaicin, an agent that causes
neuropeptide depletion, or by treatment with a SP
antagonist. Histamine antagonists also blocked
nearly all of the stimulatory effects of SP. There-
fore, SP containing neurons play a major role in
the local regulation of airway resistance and in-
terstitial fluid transfer in various pathological con-
ditions (Lundberg et al., 1983).
Trigeminovascular system
Cerebral blood vessels are innervated by a combi-
nation of sympathetic, parasympathetic and trige-
minal somatic nerve fibres all of which are
important in cerebrovascular regulation (Atalay
et al., 2002). The trigeminal component of this in-
nervation is commonly referred to as the trige-
minovascular system. This system has been shown
to transmit pain sensation from the dura mater
and cranial vessels (Huber, 1899; Penfield, 1932,
1934, 1940; Feindel et al., 1960). The perivascular
endings of trigeminovascular fibres contain several
neurotransmitters including SP (Edvinsson et al.,
1989), CGRP (Uddman et al., 1985; McCulloch
et al., 1986), NKA (Edvinsson et al., 1988), nitric
oxide and amylin (Edvinsson et al., 2001).
Neurogenic inflammation
Bayliss (1901) initially described vasodilatation of
lower limb vessels following stimulation of the
101
dorsal root ganglia. The concept of neurogenic
inflammation has since evolved to encompass
vasodilatation, plasma extravasation and neuron-
al hypersensitivity caused by the release of neuro-
peptides, including SP and CGRP, from sensory
neurons (Black, 2002). The effects of sensory ne-
uropeptides are particularly prominent at the level
of the vasculature where they cause vasodilation
of arterioles, plasma protein extravasation in
post-capillary venules and leukocyte adhesion to
endothelial cells of venules (Geppetti et al., 1995).
Additional tissue-specific responses produced by
neurogenic inflammation include smooth muscle
relaxation/contraction in the urinary bladder, ure-
ter and iris, inotropic and chronotropic effect on
the heart and bronchoconstriction in the airways
amongst others (Geppetti et al., 1995).
Peptide-containing primary sensory neurons are
characterised by their unique sensitivity to caps-
aicin, the pungent ingredient found in capsicum
(Szallasi and Blumberg, 1999). The recent cloning
of the channel operated by capsaicin, the transient
receptor potential vanilloid receptor-1 (TRPVR-1)
(Caterina et al., 1997), has clarified the molecular
basis of the selective action of capsaicin on sensory
neurons. This seven transmembrane domain pro-
tein is a non-selective cation channel, whose en-
dogenous stimulants include heat (>431C) and
protons. Subsets of primary sensory neurons are
stimulated selectively by capsaicin, causing the
release of sensory neuropeptides and promoting
neurogenic inflammation. At higher concentra-
tions capsaicin kills neurons, blocking the genesis
of subsequent neurogenic inflammatory responses
(Szallasi and Blumberg, 1999). These neurons are
defined as ‘‘capsacin-sensitive’’ due to the specific
excitatory/desensitisation effect of capsaicin
(Szolcsanyi and Mozsik, 1984).
Edema
Of all the secondary injury factors involved in
the development of neuronal dysfunction, edema
formation is thought to be central to outcome fol-
lowing injury (Lobato et al., 1988; Sarabia et al.,
1988). This is particularly the case in younger vic-
tims of TBI where formation of edema within the
102
brain has been found to be responsible for 50% of
all death and disability (Feickert et al., 1999). The
mechanisms associated with cerebral edema for-
mation remain largely unknown, however, inves-
tigation of peripheral tissues (Woie et al., 1993)
has demonstrated an association between neuro-
peptides, development of increased vascular per-
meability and subsequent edema formation.
Termed neurogenic inflammation, the process in-
volves the stimulation of the slow velocity C-fibres
(nociceptors), initiating the release of neuropep-
tides (Woie et al., 1993) including SP, CGRP and
NKA, which produce vasodilation, edema and
tissue swelling. While SP has been recognised as
being primarily associated with increased vascular
permeability, SP is stored and co-released from
sensory nerve endings with CGRP, which is a
most potent endogenous vasodilator, and displays
potent edema producing activity in the presence of
SP (Severini et al., 2002).
When left unchecked cerebral edema results in
an increase in ICP that may lead to a decrease in
tissue perfusion, localised hypoxia and ischemia,
and in severe cases brain herniation and death. It is
therefore of paramount importance that edema
genesis is inhibited following TBI. Whilst a
number of studies have investigated the role of
classical inflammation in edema formation follow-
ing TBI (Stahel et al., 1998; Lenzlinger et al., 2001;
Stein and Hoffman, 2003; Besson et al., 2005), to
date only one study has examined the role of ne-
urogenic inflammation post-trauma. Nimmo et al.
(2004) demonstrated that capsaicin administered
prior to TBI significantly attenuated BBB opening,
edema formation and the development of both
motor and cognitive deficits. The authors con-
cluded that a capsaicin-induced depletion of ne-
uropeptides clearly prevented the development of
neurogenic inflammation, and inhibited develop-
ment of vasogenic edema (Fig. 1). Although, this
study demonstrated a role for neurogenic inflam-
mation in TBI, the identification of which neuro-
peptide is primarily involved in the formation of
increased BBB permeability and edema formation
was not established.
Previous studies of peripheral edema have
shown that SP is the neuropeptide most closely
associated with capsaicin sensitivity (Saria, 1984;
Fig. 1. Alterations in brain water content at 5 h following
diffuse traumatic brain injury in rats. Vehicle treated controls
showed significantly more (po0.001) edema than either sham
or capsaicin pre-treated animals.
Yonehara et al., 1987; De and Ghosh, 1990; Laird
et al., 2000). It is also well known that SP is the
neuropeptide responsible for increased vascular
permeability, whereas CGRP is primarily associ-
ated with vasodilation. Finally, Kramer et al.
(2003) have demonstrated in studies of cardiac is-
chaemia that SP release is increased with magne-
sium depletion; declines in magnesium
concentration have been widely described follow-
ing TBI (Vink et al., 1987, 1988). Therefore it is
feasible to propose that drugs acting on SP recep-
tors, and in particular NK1 receptor (Hokfelt
et al., 2001), may be beneficial in disease treat-
ment. Studies of neurogenic inflammation follow-
ing acute brain injury have provided evidence
supporting such a possibility.
Acute CNS injury
Virtually all blood vessels of the body are sur-
rounded by sensory nerve fibres that contain both
CGRP and SP. Cerebral arteries, in particular,
appear to receive a dense supply of these neurones,
and it is therefore consistent that these neurones
have a role as mediators of the inflammatory
process following injury. Studies of migraine
(Ferrari, 1998) have indeed demonstrated that
neuropeptides are a therapeutic target to reduce
vascular permeability. However, only a limited
number of studies have demonstrated that

neurogenic inflammation may play a role in acute
injury. Significant amounts of SP release has been
detected in the nervous system following both pe-
ripheral nerve injury (Malcangio et al., 2000),
traumatic spinal cord injury (Sharma et al., 1990)
and more recently, in vitro studies of endothelium
stimulation (Annunziata et al., 2002). In our own
TBI studies, perivascular SP immunoreactivity has
been shown to increase after TBI (Fig. 2), irre-
spective of the injury model (focal versus diffuse)
or severity of injury. In ischaemia, SP immunore-
activity has been shown to increase in GABAergic
interneurons around regions of infarction, and
transiently expressed in cerebrovenular endotheli-
um (Stumm et al., 2001). Our own studies have
shown increased SP immunoreactivity following
ischaemic brain injury in the rat, which was asso-
ciated with a significant increase in edema forma-
tion (Turner et al., 2006). With respect to the glial
response after acute injury, receptor-binding sites
for SP have been shown to increase on glia after
neuronal injury (Mantyh et al., 1989). Because SP
is known to regulate inflammatory and immune
responses in peripheral tissue, it therefore may
regulate the glial response to injury. Subsequent
studies have confirmed that SP receptors are
expressed on astrocytes after injury and may there-
fore be linked to their transformation to reactive
astrocytes (Lin, 1995). This increase was not ob-
served in undamaged areas.
Several studies using NK1 receptor antagonists
also support a role for SP in neurogenic inflam-
mation following ischaemic injury, although few
have been applied to studies of the CNS. For
Fig. 2. SP immunoreactivity at 5 h following diffuse traumatic brain injury in the rat in (A) sham, uninjured rat cortex and (B) injured
rat cortex. Note the qualitative increase in perivascular SP immunoreactivity (bar ¼ 100 mm).
103
example, in a rodent cerebral stroke model, the
NK1 receptor antagonist (SR140333) reduced inf-
arct volume after focal ischaemia, implying that
that SP might play a role in exacerbating ischaemic
damage (Yu et al., 1997). However, despite these
positive findings there has been no further work
published in this area. Interestingly, serum levels
of SP have been measured in humans with both
transient ischaemic attack and complete stroke,
and these have been found to be significantly el-
evated compared with controls (Bruno et al.,
2003). In other organs, post-ischaemic blockade
of tachykinin receptors have been shown to inhibit
vascular permeability, neutrophil recruitment,
intestinal haemorrhage and neutropaenia follow-
ing ischaemia and reperfusion of the superior
mesenteric artery in the rat (Souza et al., 2002).
Similarly, SP antagonists reduced post-ischaemic
myocardial injury in rats with dietary Mg
deficiency (Kramer et al., 1997), and the authors
suggest that SP may play an early critical role
in inflammatory/pro-oxidant responses following
ischaemia. This finding is of particular interest to
TBI as traumatic injury produces sustained decline
in intracellular magnesium, implying that SP may
significantly contribute to the detrimental effects
of magnesium deficiency (Heath and Vink, 1996).
NK1 receptor antagonists
A number of groups have hypothesised that tachy-
kinin receptor antagonists may have several ther-
apeutic applications (Watling, 1992; Lowe
104
et al., 1994; Rupniak et al., 2000). The notion of
antagonising SP was first raised by Leban et al.
(1979) when examining the effects of SP agonists
in the guinea pig ileum. Subsequently, Folkers
et al. (1981) discussed the chemical design of SP
antagonists, before Engberg et al. (1981) develop-
ed the first synthetic peptide antagonist (D-Pro,
D-Trp)-SP for use in the CNS, specifically the
block of locus coeruleus (LC) neurones. However,
the affinity and metabolic instability of peptides
limited their usefulness for in vivo studies. It was
the development of the first non-peptide SP anta-
gonist by Snider et al. (1991), who showed that
CP-96345 was a potent, competitive and highly
selective antagonist of the NK1 receptor, that ini-
tiated a new wave of interest in these compounds.
This antagonist was further refined to create the
highly potent n-acetyl-L-tryptophan benzyl esters
(MacLeod et al., 1993, 1995), and it was the 3,5-bis
(trifluoromethyl) benzyl ester (L-732138) that led
to the inhibition of SP-induced inositol phosphate
accumulation in Chinese hamster ovarian cells
expressing the human NK1 receptor. L-732138
was then used as a starting point to identify high-
affinity SP receptor antagonists with improved in
vivo activity. Rupniak and Kramer (1999) first
described the efficacy of SP antagonists in the
treatment of experimental depression and emesis,
with the NK1 receptor antagonists being able to
decrease anxiety (Santarelli et al., 2001) and de-
pression with fewer side effects than other drugs
of choice for the treatment of depression (Severini
et al., 2002). Ranga and Krishnan (2002) subse-
quently published the first use of the SP antagonist
MK-0869 in the treatment of clinical depression
and anxiety. NK1 receptor antagonists have also
been tested in dental pain, osteoarthritis, ne-
uropathic pain and migraine however no analge-
sic effects have been reported in studies to date
(Hokfelt et al., 2001).
An alternative to inhibiting SP binding is to
decrease neuropeptide release and depleting neuro-
peptide stores using the vanilloid agonist capsaicin
(Cadieux et al., 1986; Kashiba et al., 1997). Nimmo
et al. (2004) were able to demonstrate that neuro-
peptide-depletion prior to induction of TBI
attenuated the development of BBB dysfunction
and vasogenic edema formation associated with
neurogenic inflammation, as well as improve motor
and cognitive function after TBI. In guinea pig
skin, administration of the NK1 receptor antago-
nist RP 67580 was able to inhibit SP-induced ed-
ema formation and white blood cell accumulation
(Campos and Calixto, 2000). In vitro studies of
endothelial injury have supported a potential role
for SP antagonists in the treatment of BBB dys-
function by demonstrating that such antagonists
neutralised increased BBB permeability, upregula-
tion of MHC-II molecules, reduced expression of
ICAM-1 and prevented associated cell morpholog-
ical changes (Annunziata et al., 2002).
Conclusion
While a role for neurogenic inflammation in vas-
cular permeability and edema formation has been
described in peripheral tissues for a number of
years, few studies have examined the potential for
neurogenic inflammation to influence BBB perme-
ability and edema formation after traumatic brain
injury. Those studies that have investigated a role
for neuropeptides in acute brain injury have dem-
onstrated that inhibition of release attenuates BBB
permeability and edema formation after injury,
and results in an associated improvement in func-
tional outcome. Immunohistochemistry studies
have demonstrated that increased SP levels are
observed perivascularly, confirming a potential
role in vasogenic edema formation. Given the
apparent lack of side effects from this class of
compound, and the potential to improve post-
traumatic anxiety and depression, inhibition of the
SP pathway using NK1 receptor antagonists is ex-
pected to provide a novel approach to the man-
agement of edema formation and improvement of
functional outcome after TBI.
References
Annunziata, P., Cioni, C., Santonini, R. and Paccagnini, E.
(2002) Substance P antagonist blocks leakage and reduces
activation of cytokine-stimulated rat brain endothelium.
J. Neuroimmunol., 131: 41–49.
Atalay, B., Bolay, H., Dalkara, T., Soylemezoglu, F., Oge, K.
and Ozcan, O.E. (2002) Transcorneal stimulation of trigemi-
nal nerve afferents to increase cerebral blood flow in rats with

cerebral vasospasm: a noninvasive method to activate the
trigeminovascular reflex. J. Neurosurg., 97: 1179–1183.
Azaryan, A.V. and Galoyan, A.A. (1988) Substrate specificity
of cerebral cathepsin D and high-Mr aspartic endopeptidase.
J. Neurosci. Res., 19: 268–271.
Bayliss, W.M. (1901) On the origin from the spinal cord of the
vaso-dilator of the hindlimb, and on the nature of these fib-
ers. J. Physiol. (Lond.), 32: 1025–1043.
Bergner, A.J., Murphy, S.M. and Anderson, C.R. (2000) After
axotomy, substance P and vasoactive intestinal peptide ex-
pression occurs in pilomotor neurons in the rat superior cer-
vical ganglion. Neuroscience, 96: 611–618.
Besson, V.C., Chen, X.R., Plotkine, M. and Marchand-
Verrecchia, C. (2005) Fenofibrate, a peroxisome prolifera-
tor-activated receptor alpha agonist, exerts neuroprotective
effects in traumatic brain injury. Neurosci. Lett., 388: 7–12.
Black, P.H. (2002) Stress and the inflammatory response: a re-
view of neurogenic inflammation. Brain Behav. Immun., 16:
622–653.
Blumberg, S., Teichberg, V.I., Charli, J.L., Hersh, L.B.
and Mckelvy, J.F. (1980) Cleavage of substance P to an
N-terminal tetrapeptide and a C-terminal heptapeptide by a
post-proline cleaving enzyme from bovine brain. Brain Res.,
192: 477–486.
Bruno, G., Tega, F., Bruno, A., Graf, U., Corelli, F., Molfetta,
R. and Barucco, M. (2003) The role of substance P in cere-
bral ischemia. Int. J. Immunopathol. Pharmacol., 16: 67–72.
Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and
Adams, J.H. (1991) Glial swelling following human cerebral
contusion: an ultrastructural study. J. Neurol. Neurosurg.
Psychiatry, 54: 427–434.
Cadieux, A., Springall, D.R., Mulderry, P.K., Rodrigo, J.,
Ghatei, M.A., Terenghi, G., Bloom, S.R. and Polak, J.M.
(1986) Occurrence, distribution and ontogeny of CGRP
immunoreactivity in the rat lower respiratory tract: effect of
capsaicin treatment and surgical denervations. Neuroscience,
19: 605–627.
Campos, M.M. and Calixto, J.B. (2000) Neurokinin mediation
of edema and inflammation. Neuropeptides, 34: 314–322.
Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen,
T.A., Levine, J.D. and Julius, D. (1997) The capsaicin re-
ceptor: a heat-activated ion channel in the pain pathway.
Nature, 389: 816–824.
Cervos-Navarro, J. and Lafuente, J.V. (1991) Traumatic brain
injuries: structural changes. J. Neurol. Sci., 103(Suppl): S3–S14.
Chang, M.M., Leeman, S.E. and Niall, H.D. (1971) Amino-
acid sequence of substance P. Nat. New Biol., 232: 86–87.
Costa, M., Cuello, A.C., Furness, J.B. and Franco, R. (1980)
Distribution of enteric neurons showing immunoreactivity
for substance P in the guinea-pig ileum. Neuroscience, 5:
323–331.
Cuello, A.C. and Kanazawa, I. (1978) The distribution of sub-
stance P immunoreactive fibers in the rat central nervous
system. J. Comp. Neurol., 178: 129–156.
De, A.K. and Ghosh, J.J. (1990) Inflammatory responses in-
duced by substance P in rat paw. Indian J. Exp. Biol., 28:
946–948.
105
Di Maria, G.U., Bellofiore, S. and Geppetti, P. (1998) Regu-
lation of airway neurogenic inflammation by neutral endo-
peptidase. Eur. Respir. J., 12: 1454–1462.
Dietrich, W.D., Alonso, O. and Halley, M. (1994) Early
microvascular and neuronal consequences of traumatic
brain injury: a light and electron microscopic study in rats.
J. Neurotrauma, 11: 289–301.
Dun, N.J. and Minota, S. (1981) Effects of substance P on
neurones of the inferior mesenteric ganglia of the guinea-pig.
J. Physiol., 321: 259–271.
Edvinsson, L., Brodin, E., Jansen, I. and Uddman, R. (1988)
Neurokinin A in cerebral vessels: characterization, localiza-
tion and effects in vitro. Regul. Pept., 20: 181–197.
Edvinsson, L., Ekman, R. and Thulin, T. (1989) Reduced levels
of calcitonin gene-related peptide (CGRP) but not substance
P during and after treatment of severe hypertension in man.
J. Hum. Hypertens, 3: 267–270.
Edvinsson, L., Elsas, T., Suzuki, N., Shimizu, T. and Lee, T.J.
(2001) Origin and co-localization of nitric oxide synthase,
CGRP, PACAP, and VIP in the cerebral circulation of the
rat. Microsc. Res. Tech., 53: 221–228.
Engberg, G., Svensson, T.H., Rosell, S. and Folkders, K. (1981)
A synthetic peptide as an antagonist of substance P. Nature,
293: 222–223.
Engel, S., Schluesener, H., Mittelbronn, M., Seid, K., Adjodah,
D., Wehner, H.D. and Meyermann, R. (2000) Dynamics of
microglial activation after human traumatic brain injury are
revealed by delayed expression of macrophage-related proteins
MRP8 and MRP14. Acta Neuropathol. (Berl.), 100: 313–322.
Feickert, H.J., Drommer, S. and Heyer, R. (1999) Severe head
injury in children: impact of risk factors on outcome.
J. Trauma, 47: 33–38.
Feindel, W., Penfield, W. and Mc, N.F. (1960) The tentorial
nerves and localization of intracranial pain in man. Neuro-
logy, 10: 555–563.
Ferrari, M.D. (1998) Migraine. Lancet, 351: 1043–1051.
Fleminger, S. and Ponsford, J. (2005) Long term outcome after
traumatic brain injury. BMJ, 331: 1419–1420.
Folkers, K., Horig, J., Rosell, S. and Bjorkroth, U. (1981)
Chemical design of antagonists of substance P. Acta Physiol.
Scand., 111: 505–506.
Gennarelli, T.A. and Graham, D.I. (1998) Neuropathology of
the head injuries. Semin. Clin. Neuropsychiatry, 3: 160–175.
Geppetti, P., Bertrand, C., Ricciardolo, F.L. and Nadel, J.A.
(1995) New aspects on the role of kinins in neurogenic in-
flammation. Can. J. Physiol. Pharmacol., 73: 843–847.
Gerard, N.P., Bao, L., Xiao-Ping, H. and Gerard, C. (1993)
Molecular aspects of the tachykinin receptors. Regul. Pept.,
43: 21–35.
Gibbins, I.L., Furness, J.B. and Costa, M. (1987) Pathway-
specific patterns of the co-existence of substance P, calcitonin
gene-related peptide, cholecystokinin and dynorphin in neu-
rons of the dorsal root ganglia of the guinea-pig. Cell Tissue
Res., 248: 417–437.
Graham, D.I., Clark, J.C., Adams, J.H. and Gennarelli,
T.A. (1992) Diffuse axonal injury caused by assault. J. Clin.
Pathol., 45: 840–841.
106
Graham, D.I., Gennarelli, T.A. and McIntosh, T.K. (2002)
Greenfield’s neuropathology. Arnold, New York.
Graham, D.I., Gentleman, S.M., Nicoll, J.A., Royston, M.C.,
Mckenzie, J.E., Roberts, G.W. and Griffin, W.S. (1996)
Altered beta-APP metabolism after head injury and its
relationship to the aetiology of Alzheimer’s disease. Acta
Neurochir. Suppl., 66: 96–102.
Harrison, S. and Geppetti, P. (2001) Substance P. Int. J. Bio-
chem. Cell Biol., 33: 555–576.
Hasenohrl, R.U., Souza-Silva, M.A., Nikolaus, S., Tomaz, C.,
Brandao, M.L., Schwarting, R.K. and Huston, J.P. (2000)
Substance P and its role in neural mechanisms governing
learning, anxiety and functional recovery. Neuropeptides, 34:
272–280.
Hausmann, R. and Betz, P. (2000) The time course of the vas-
cular response to human brain injury: an immunohisto-
chemical study. Int. J. Legal Med., 113: 288–292.
Hausmann, R., Kaiser, A., Lang, C., Bohnert, M. and Betz, P.
(1999) A quantitative immunohistochemical study on the
time-dependent course of acute inflammatory cellular re-
sponse to human brain injury. Int. J. Legal Med., 112:
227–232.
Heath, D.L. and Vink, R. (1996) Traumatic brain axonal injury
produces sustained decline in intracellular free magnesium
concentration. Brain Res., 738: 150–153.
Helke, C.J., Neil, J.J., Massari, V.J. and Loewy, A.D. (1982)
Substance P neurons project from the ventral medulla to the
intermediolateral cell column and ventral horn in the rat.
Brain Res., 243: 147–152.
Helke, C.J. and Phillips, E.T. (1988) Substance P antagonist-
induced spinal cord vasoconstriction: effects of thyrotropin-
releasing hormone and substance P agonists. Peptides, 9:
1307–1315.
Henry, J.L. (1976) Effects of substance P on functionally iden-
tified units in cat spinal cord. Brain Res., 114: 439–451.
Heymann, E. and Mentlein, R. (1978) Liver dipeptidyl amino-
peptidase IV hydrolyzes substance P. FEBS Lett., 91: 360–364.
Hokfelt, T., Kellerth, J.O., Nilsson, G. and Pernow, B. (1975)
Experimental immunohistochemical studies on the localiza-
tion and distribution of substance P in cat primary sensory
neurons. Brain Res., 100: 235–252.
Hokfelt, T., Pernow, B. and Wahren, J. (2001) Substance P: a
pioneer amongst neuropeptides. J. Intern. Med., 249: 27–40.
Hokfelt, T., Zhang, X. and Wiesenfeld-Hallin, Z. (1994)
Messenger plasticity in primary sensory neurons following
axotomy and its functional implications. Trends Neurosci.,
17: 22–30.
Holmin, S., Soderlund, J., Biberfeld, P. and Mathiesen, T.
(1998) Intracerebral inflammation after human brain contu-
sion. Neurosurgery, 42: 291–298; discussion 298–299.
Hooper, N.M. and Turner, A.J. (1987) Isolation of two differ-
entially glycosylated forms of peptidyl-dipeptidase A (angio-
tensin converting enzyme) from pig brain: a re-evaluation of
their role in neuropeptide metabolism. Biochem. J., 241:
625–633.
Huber, G.C. (1899) Observations on the innervation of intra-
cranial vessels. J. Comp. Neurol., 9: 1–34.
Kageyama, T. (1993) Rabbit procathepsin E and cathepsin E.
Nucleotide sequence of cDNA, hydrolytic specificity for bio-
logically active peptides and gene expression during devel-
opment. Eur. J. Biochem., 216: 717–728.
Kandel, E.R. and Squire, L.R. (2000) Neuroscience: breaking
down scientific barriers to the study of brain and mind.
Science, 290: 1113–1120.
Kashiba, H., Ueda, Y. and Senba, E. (1997) Systemic capsaicin
in the adult rat differentially affects gene expression for ne-
uropeptides and neurotrophin receptors in primary sensory
neurons. Neuroscience, 76: 299–312.
Kotani, Y., Hirota, Y., Sugiyama, K., Joh, S., Shibutani, T.,
Matsuura, H. and Inoki, R. (1986) Effects of noxious stimuli
and anesthetic agents on substance P content in rat central
nervous system. Jpn. J. Pharmacol., 40: 143–147.
Kramer, J.H., Mak, I.T., Phillips, T.M. and Weglicki, W.B.
(2003) Dietary magnesium intake influences circulating pro-
inflammatory neuropeptide levels and loss of myocardial tol-
erance to postischemic stress. Exp. Biol. Med. (Maywood),
228: 665–673.
Kramer, S.D., Jakits-Deiser, C. and Wunderli-Allenspach, H.
(1997) Free fatty acids cause pH-dependent changes in
drug–lipid membrane interactions around physiological pH.
Pharm. Res., 14: 827–832.
Laird, J.M., Olivar, T., Roza, C., De Felipe, C., Hunt, S.P. and
Cervero, F. (2000) Deficits in visceral pain and hyperalgesia
of mice with a disruption of the tachykinin NK1 receptor
gene. Neuroscience, 98: 345–352.
Leban, J., Rackur, G., Yamaguchi, I., Folkers, K., Bjorkroth,
U., Rosell, S., Yanaihara, N. and Yanaihara, C. (1979) Syn-
thesis of substance P analogs and agonistic and antagonistic
activities. Acta Chem. Scand. B, 33: 664–668.
Lee, Y., Kawai, Y., Shiosaka, S., Takami, K., Kiyama, H.,
Hillyard, C.J., Girgis, S., Macintyre, I., Emson, P.C. and
Tohyama, M. (1985) Coexistence of calcitonin gene-related
peptide and substance P-like peptide in single cells of the
trigeminal ganglion of the rat: immunohistochemical analy-
sis. Brain Res., 330: 194–196.
Lembeck, F. (1953) Central transmission of afferent impulses.
III. Incidence and significance of the substance P in the dorsal
roots of the spinal cord. Naunyn Schmiedebergs Arch. Exp.
Pathol. Pharmakol., 219: 197–213.
Lenzlinger, P.M., Morganti-Kossmann, M.C., Laurer, H.L.
and McIntosh, T.K. (2001) The duality of the inflammatory
response to traumatic brain injury. Mol. Neurobiol., 24:
169–181.
Leon-Carrion, J., Dominguez-Morales Mdel, R., Barroso,
Y.M.J.M. and Murillo-Cabezas, F. (2005) Epidemiology
of traumatic brain injury and subarachnoid hemorrhage.
Pituitary, 8: 197–202.
Lin, R.C. (1995) Reactive astrocytes express substance-P
immunoreactivity in the adult forebrain after injury. Neuro-
report, 7: 310–312.
Ljungdahl, A., Hokfelt, T. and Nilsson, G. (1978) Distribution
of substance P-like immunoreactivity in the central nervous
system of the rat-I. Cell bodies and nerve terminals. Neuro-
science, 3: 861–943.

Lobato, R.D., Sarabia, R., Cordobes, F., Rivas, J.J., Adrados,
A., Cabrera, A., Gomez, P., Madera, A. and Lamas, E.
(1988) Posttraumatic cerebral hemispheric swelling. Analysis
of 55 cases studied with computerized tomography. J. Neuro-
surg., 68: 417–423.
Lowe III, J.A., Drozda, S.E., Mclean, S., Bryce, D.K., Craw-
ford, R.T., Snider, R.M., Longo, K.P., Nagahisa, A. and
Tsuchiya, M. (1994) Aza-tricyclic substance P antagonists.
J. Med. Chem., 37: 2831–2840.
Lundberg, J.M., Martling, C.R., Saria, A., Folkers, K. and
Rosell, S. (1983) Cigarette smoke-induced airway oedema
due to activation of capsaicin-sensitive vagal afferents and
substance P release. Neuroscience, 10: 1361–1368.
Macleod, A.M., Cascieri, M.A., Merchant, K.J., Sadowski, S.,
Hardwicke, S., Lewis, R.T., Macintyre, D.E., Metzger, J.M.,
Fong, T.M., Shepheard, S., et al. (1995) Synthesis and
biological evaluation of NK1 antagonists derived from
L-tryptophan. J. Med. Chem., 38: 934–941.
Macleod, A.M., Merchant, K.J., Cascieri, M.A., Sadowski, S.,
Ber, E., Swain, C.J. and Baker, R. (1993) N-acyl-L-try-
ptophan benzyl esters: potent substance P receptor antago-
nists. J. Med. Chem., 36: 2044–2045.
Maggi, C.A. and Schwartz, T.W. (1997) The dual nature of the
tachykinin NK1 receptor. Trends Pharmacol. Sci., 18:
351–355.
Malcangio, M., Ramer, M.S., Jones, M.G. and Mcmahon, S.B.
(2000) Abnormal substance P release from the spinal cord
following injury to primary sensory neurons. Eur. J. Neuro-
sci., 12: 397–399.
Mantyh, P.W., Johnson, D.J., Boehmer, C.G., Catton, M.D.,
Vinters, V.H., Maggio, J.E., Too, H.P. and Vigna, S.R.
(1989) Substance P receptor binding sites are expressed by
glia in vivo after neuronal injury. Proc. Natl. Acad. Sci.
U.S.A., 86: 5193–5197.
Marlier, L., Poulat, P., Rajaofetra, N. and Privat, A. (1991)
Modifications of serotonin-, substance P- and calcitonin
gene-related peptide-like immunoreactivities in the dorsal
horn of the spinal cord of arthritic rats: a quantitative
immunocytochemical study. Exp. Brain Res., 85: 482–490.
Marmarou, A., Fatouros, P.P., Barzo, P., Portella, G., Yoshi-
hara, M., Tsuji, O., Yamamoto, T., Laine, F., Signoretti, S.,
Ward, J.D., Bullock, M.R. and Young, H.F. (2000) Contri-
bution of edema and cerebral blood volume to traumatic
brain swelling in head-injured patients. J. Neurosurg., 93:
183–193.
Matsas, R., Kenny, A.J. and Turner, A.J. (1984) The meta-
bolism of neuropeptides. The hydrolysis of peptides, includ-
ing enkephalins, tachykinins and their analogues, by
endopeptidase-24.11. Biochem. J., 223: 433–440.
McCarson, K.E. (1999) Central and peripheral expression
of neurokinin-1 and neurokinin-3 receptor and substance
P-encoding messenger RNAs: peripheral regulation during
formalin-induced inflammation and lack of neurokinin
receptor expression in primary afferent sensory neurons.
Neuroscience, 93: 361–370.
McCulloch, J., Uddman, R., Kingman, T.A. and Edvinsson, L.
(1986) Calcitonin gene-related peptide: functional role in
107
cerebrovascular regulation. Proc. Natl. Acad. Sci. U.S.A., 83:
5731–5735.
McIntosh, T.K., Smith, D.H., Meaney, D.F., Kotapka, M.J.,
Gennarelli, T.A. and Graham, D.I. (1996) Neuropathological
sequelae of traumatic brain injury: relationship to neuro-
chemical and biomechanical mechanisms. Lab. Invest., 74:
315–342.
Nadel, J.A. (1991) Neutral endopeptidase modulates neuro-
genic inflammation. Eur. Respir. J., 4: 745–754.
Nakanishi, S. (1991) Mammalian tachykinin receptors. Annu.
Rev. Neurosci., 14: 123–136.
Nicoll, R.A., Schenker, C. and Leeman, S.E. (1980) Substance
P as a transmitter candidate. Annu. Rev. Neurosci., 3:
227–268.
Nilsson, G., Larsson, L.I., Hakanson, R., Brodin, E., Pernow,
B. and Sundler, F. (1975) Localization of substance P-like
immunoreactivity in mouse gut. Histochemistry, 43: 97–99.
Nimmo, A.J., Cernak, I., Heath, D.L., Hu, X., Bennett, C.J.
and Vink, R. (2004) Neurogenic inflammation is associated
with development of edema and functional deficits following
traumatic brain injury in rats. Neuropeptides, 38: 40–47.
Noguchi, K., Morita, Y., Kiyama, H., Ono, K. and Tohyama,
M. (1988) A noxious stimulus induces the preprotachykinin-
A gene expression in the rat dorsal root ganglion: a quanti-
tative study using in situ hybridization histochemistry. Brain
Res., 464: 31–35.
O’Connor, C.A., Cernak, I. and Vink, R. (2003) Interaction
between anesthesia, gender, and functional outcome task
following diffuse traumatic brain injury in rats. J. Neuro-
trauma, 20: 533–541.
Obrenovitch, T.P. and Urenjak, J. (1997) Is high extracellular
glutamate the key to excitotoxicity in traumatic brain injury?
J. Neurotrauma, 14: 677–698.
Oehmichen, M., Eisenmenger, W., Raff, G. and Berghaus, G.
(1986) Brain macrophages in human cortical contusions as
indicator of survival period. Forensic Sci. Int., 30: 281–301.
Oehmichen, M. and Raff, G. (1980) Timing of cortical contu-
sion. Correlation between histomorphologic alterations and
post-traumatic interval. Z. Rechtsmed., 84: 79–94.
Olgart, L., Gazelius, B., Brodin, E. and Nilsson, G. (1977) Re-
lease of substance P-like immunoreactivity from the dental
pulp. Acta Physiol. Scand., 101: 510–512.
Otsuka, M. and Konishi, S. (1976) Release of substance P-like
immunoreactivity from isolated spinal cord of newborn rat.
Nature, 264: 83–84.
Otsuka, M. and Takahashi, T. (1977) Putative peptide neuro-
transmitters. Annu. Rev. Pharmacol. Toxicol., 17: 425–439.
Otsuka, M. and Yoshioka, K. (1993) Neurotransmitter func-
tions of mammalian tachykinins. Physiol. Rev., 73: 229–308.
Penfield, W. (1932) Intracerebral vascular nerves. Arch. Neurol.
Psychiatry, 27: 30–44.
Penfield, W. (1934) A contribution to the mechanism of intra-
cranial pain. Proc. Assoc. Res. Nerv. Mental Dis., 15:
399–415.
Penfield, W. (1940) Dural headache and the innervation of the
dura mater. Arch. Neurol. Psychiatry, 44: 43–75.
Pernow, B. (1983) Substance P. Pharmacol. Rev., 35: 85–141.
108
Powell, D., Leeman, S., Tregear, G.W., Niall, H.D. and Potts
Jr., J.T. (1973) Radioimmunoassay for substance P. Nat.
New Biol., 241: 252–254.
Probert, L. and Hanley, M.R. (1987) The immunocytochemical
localisation of ‘substance-P-degrading enzyme’ within the rat
spinal cord. Neurosci. Lett., 78: 132–137.
Ranga, K. and Krishnan, R. (2002) Clinical experience with
substance P receptor (NK1) antagonists in depression. J.
Clin. Psychiatry, 63(Suppl 11): 25–29.
Regoli, D., Boudon, A. and Fauchere, J.L. (1994) Receptors
and antagonists for substance P and related peptides.
Pharmacol. Rev., 46: 551–599.
Reilly, P.L. (2001) Brain injury: the pathophysiology of the first
hours. ‘Talk and die revisited’. J. Clin. Neurosci., 8: 398–403.
Rupniak, N.M., Carlson, E.C., Harrison, T., Oates, B., Seward,
E., Owen, S., De Felipe, C., Hunt, S. and Wheeldon, A.
(2000) Pharmacological blockade or genetic deletion of sub-
stance P (NK(1)) receptors attenuates neonatal vocalisation
in guinea-pigs and mice. Neuropharmacology, 39:
1413–1421.
Rupniak, N.M. and Kramer, M.S. (1999) Discovery of the
antidepressant and anti-emetic efficacy of substance P recep-
tor (NK1) antagonists. Trends Pharmacol. Sci., 20: 485–490.
Sakurada, T., Hara, A., Matsumura, H., Yamada, H.,
Sakurada, S. and Kisara, K. (1990) A substance P analogue
reduces amino acid contents in the rat spinal cord. Pharma-
col. Toxicol., 66: 75–76.
Santarelli, L., Gobbi, G., Debs, P.C., Sibille, E.T., Blier, P.,
Hen, R. and Heath, M.J. (2001) Genetic and pharmacologi-
cal disruption of neurokinin 1 receptor function decreases
anxiety-related behaviors and increases serotonergic func-
tion. Proc. Natl. Acad. Sci. U.S.A., 98: 1912–1917.
Sarabia, R., Lobato, R.D., Rivas, J.J., Cordobes, F., Rubio, J.,
Cabrera, A., Gomez, P., Munoz, M.J. and Madera, A. (1988)
Cerebral hemisphere swelling in severe head injury patients.
Acta Neurochir. Suppl. (Wien.), 42: 40–46.
Saria, A. (1984) Substance P in sensory nerve fibres contributes
to the development of oedema in the rat hind paw after
thermal injury. Br. J. Pharmacol., 82: 217–222.
Schaible, H.G., Jarrott, B., Hope, P.J. and Duggan, A.W.
(1990) Release of immunoreactive substance P in the spinal
cord during development of acute arthritis in the knee joint
of the cat: a study with antibody microprobes. Brain Res.,
529: 214–223.
Schultzberg, M., Hokfelt, T., Olson, L., Alund, M., Nilsson,
G., Terenius, L., Elde, R., Goldstein, M. and Said, S. (1980)
Substance P, VIP, enkephalin and somatostatin immunore-
active neurons in intestinal tissue transplanted to the anterior
eye chamber. J. Auton. Nerv. Syst., 1: 291–303.
Severini, C., Improta, G., Falconieri-Erspamer, G., Salvadori,
S. and Erspamer, V. (2002) The tachykinin peptide family.
Pharmacol. Rev., 54: 285–322.
Sharma, H.S., Nyberg, F., Olsson, Y. and Dey, P.K. (1990)
Alteration of substance P after trauma to the spinal cord: an
experimental study in the rat. Neuroscience, 38: 205–212.
Shults, C.W., Quirion, R., Chronwall, B., Chase, T.N. and
O’donohue, T.L. (1984) A comparison of the anatomical
distribution of substance P and substance P receptors in the
rat central nervous system. Peptides, 5: 1097–1128.
Skidgel, R.A. and Erdos, E.G. (1987) Cleavage of peptide
bonds by angiotensin I converting enzyme. Agents Actions
Suppl., 22: 289–296.
Snider, R.M., Longo, K.P., Drozda, S.E., Lowe III, J.A. and
Leeman, S.E. (1991) Effect of CP-96,345, a nonpeptide sub-
stance P receptor antagonist, on salivation in rats. Proc. Natl.
Acad. Sci. U.S.A., 88: 10042–10044.
Souza, D.G., Mendonca, V.A., De, A.C.M.S., Poole, S. and
Teixeira, M.M. (2002) Role of tachykinin NK receptors on
the local and remote injuries following ischaemia and reper-
fusion of the superior mesenteric artery in the rat. Br. J.
Pharmacol., 135: 303–312.
Stahel, P.F., Morganti-Kossmann, M.C. and Kossmann, T.
(1998) The role of the complement system in traumatic brain
injury. Brain Res. Brain Res. Rev., 27: 243–256.
Stein, D.G. and Hoffman, S.W. (2003) Concepts of CNS plas-
ticity in the context of brain damage and repair. J. Head
Trauma Rehabil., 18: 317–341.
Sternini, C., Su, D., Gamp, P.D. and Bunnett, N.W. (1995)
Cellular sites of expression of the neurokinin-1 receptor in the
rat gastrointestinal tract. J. Comp. Neurol., 358: 531–540.
Stumm, R., Culmsee, C., Schafer, M.K., Krieglstein, J. and
Weihe, E. (2001) Adaptive plasticity in tachykinin and tachy-
kinin receptor expression after focal cerebral ischemia is
differentially linked to gabaergic and glutamatergic cerebro-
cortical circuits and cerebrovenular endothelium. J. Neuro-
sci., 21: 798–811.
Szallasi, A. and Blumberg, P.M. (1999) Vanilloid (capsaicin)
receptors and mechanisms. Pharmacol. Rev., 51: 159–212.
Szolcsanyi, J. and Mozsik, G. (1984) Effects of capsaicin on the
development of gastric mucosal damage by different necro-
tizing agents and of gastric cytoprotection by PGI2
atropine and cimetidine on rats. Acta Physiol. Hung., 64:
287–291.
Takahashi, T., Konishi, S., Powell, D., Leeman, S.E. and
Otsuka, M. (1974) Identification of the motoneuron-depo-
larizing peptide in bovine dorsal root as hypothalamic sub-
stance P. Brain Res., 73: 59–69.
Takahashi, T. and Otsuka, M. (1975) Regional distribution of
substance P in the spinal cord and nerve roots of the cat and
the effect of dorsal root section. Brain Res., 87: 1–11.
Tregear, G.W., Niall, H.D., Potts Jr, J.T., Leeman, S.E. and
Chang, M.M. (1971) Synthesis of substance P. Nat. New
Biol., 232: 87–89.
Turner, R.J., Blumbergs, P.C., Sims, N.R., Helps, S.C., Rod-
gers, K.M. and Vink, R. (2006) Increased substance P
immunoreactivity and edema formation following reversible
ischemic stroke. Acta Neurochir. Suppl., 96: 263–266.
Uddman, R., Edvinsson, L., Ekman, R., Kingman, T. and
McCulloch, J. (1985) Innervation of the feline cerebral
vasculature by nerve fibers containing calcitonin gene-related
peptide: trigeminal origin and co-existence with substance P.
Neurosci. Lett., 62: 131–136.
Vaz, R., Sarmento, A., Borges, N., Cruz, C. and Azevedo, I.
(1997) Ultrastructural study of brain microvessels in patients

with traumatic cerebral contusions. Acta Neurochir. (Wien.),
139: 215–220.
Vespa, P., Prins, M., Ronne-Engstrom, E., Caron, M., Shal-
mon, E., Hovda, D.A., Martin, N.A. and Becker, D.P. (1998)
Increase in extracellular glutamate caused by reduced cerebral
perfusion pressure and seizures after human traumatic brain
injury: a microdialysis study. J. Neurosurg., 89: 971–982.
Vink, R., McIntosh, T.K., Demediuk, P. and Faden, A.I.
(1987) Decrease in total and free magnesium concentration
following traumatic brain injury in rats. Biochem. Biophys.
Res. Commun., 149: 594–599.
Vink, R., McIntosh, T.K., Demediuk, P., Weiner, M.W. and
Faden, A.I. (1988) Decline in intracellular free Mg2+ is as-
sociated with irreversible tissue injury after brain trauma.
J. Biol. Chem., 263: 757–761.
Von Euler, U.S. and Gaddum, J.H. (1931) An unidentified de-
pressor substance in certain tissue extracts. J. Physiol.
(Lond.), 72: 74–87.
Wang, L.H., Ahmad, S., Benter, I.F., Chow, A., Mizutani, S.
and Ward, P.E. (1991) Differential processing of substance P
109
and neurokinin A by plasma dipeptidyl(amino)peptidase IV,
aminopeptidase M and angiotensin converting enzyme. Pep-
tides, 12: 1357–1364.
Watling, K.J. (1992) Nonpeptide antagonists herald new era in
tachykinin research. Trends Pharmacol. Sci., 13: 266–269.
Woie, K., Koller, M.E., Heyeraas, K.J. and Reed, R.K. (1993)
Neurogenic inflammation in rat trachea is accompanied by
increased negativity of interstitial fluid pressure. Circ. Res.,
73: 839–845.
Xiong, Y., Peterson, P.L., Muizelaar, J.P. and Lee, C.P. (1997)
Amelioration of mitochondrial function by a novel antioxi-
dant U-101033E following traumatic brain injury in rats.
J. Neurotrauma, 14: 907–917.
Yonehara, N., Shibutani, T. and Inoki, R. (1987) Contribution
of substance P to heat-induced edema in rat paw. J. Pharma-
col. Exp. Ther., 242: 1071–1076.
Yu, Z., Cheng, G., Huang, X., Li, K. and Cao, X. (1997)
Neurokinin-1 receptor antagonist SR140333: a novel type
of drug to treat cerebral ischemia. Neuroreport, 8:
2117–2119.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 8

Current concepts of cerebral oxygen transport and

energy metabolism after severe traumatic brain injury

B.H. Verweij1,, G.J. Amelink1 and J.P. Muizelaar2

1
Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical Center Utrecht,
Heidelberglaan 100, 3584 CX Utrecht, The Netherlands
2
Department of Neurosurgery, University of California at Davis Medical Center, 4860 Y Street, Suite 3740, Sacramento,
CA 95817, USA

Abstract: Before energy metabolism can take place, brain cells must be supplied with oxygen and glucose.
Only then, in combination with normal mitochondrial function, sufficient energy (adenosine tri-phosphate
(ATP)) can be produced. Glucose is virtually the sole fuel for the human brain. The brain lacks fuel stores
and requires a continuous supply of glucose and oxygen. Therefore, continuous cerebral blood flow (CBF),
cerebral oxygen tension and delivery, and normal mitochondrial function are of vital importance for the
maintenance of brain function and tissue viability. This review focuses on three main issues: (1) Cerebral
oxygen transport (CBF, and oxygen partial pressure (PO2) and delivery to the brain); (2) Energy metabo-
lism (glycolysis, mitochondrial function: citric acid cycle and oxidative phosphorylation); and (3) The role
of the above in the pathophysiology of severe head injury. Basic understanding of these issues in the normal
as well as in the traumatized brain is essential in developing new treatment strategies. These issues also play
a key role in interpreting data collected from monitoring techniques such as cerebral tissue PO2, jugular
bulb oxygen saturation (SjvO2), near infra red spectroscopy (NIRS), microdialysis, intracranial pressure
monitoring (ICP), laser Doppler flowmetry, and transcranial Doppler flowmetry — both in the experi-
mental and in the clinical setting.

Keywords: traumatic brain injury; metabolism; mitochondrial function; mitochondria; lactate; cerebral
blood flow; ischemia

Cerebral oxygen transport electro-encephalographic slowing and at 20 ml/

Cerebral blood flow
Under normal conditions, the brain has critical
thresholds for cerebral blood flow (CBF) as well as
oxygen tension (PO2). If CBF is reduced, neuronal
events will result (Astrup, 1982). Normal CBF
is
50 ml/100 g brain tissue/min (Sokoloff, 1960).
If CBF is reduced to
25 ml/100 g/min there is
Corresponding author. Tel.: +31-30-2507059;
Fax: +31-30-2542100; E-mail: bverweij@umcutrecht.nl
100 g/min loss of consciousness occurs, but may be
tolerated without long term functional conse-
quences. Below a CBF of 18 ml/100 g/min, ionic
homeostasis becomes jeopardized and neurons
convert to anaerobic metabolism (Jones et al.,
1981; Siesjo, 1992; Schroder et al., 1996). At a
CBF of 10 ml/100 g/min, membrane integrity is
lost and irreversible brain damage is inevitable.
Tissue infarction is related to CBF but is also time
dependent as shown in Fig. 1 (Jones et al., 1981).
Arterial blood contains 13 vol% of O2, while
jugular venous blood contains 6.7 vol%, for an

DOI: 10.1016/S0079-6123(06)61008-X 111

112
Fig. 1. Schematic showing the relation between CBF and ischemic duration/tissue infarction. (Adapted with permission from Jones
et al., 1981.)
arterio venous difference of oxygen (AVDO2) of
(136.7) ¼ 6.3 vol% (ml of O2/100 ml of blood).
Knowing how much blood is flowing to the brain
(
50 ml/100 g brain tissue/min) and how much
oxygen the brain extracts from this blood
(AVDO2), one can calculate cerebral metabolic
rate of oxygen (CMRO2):
CMRO2 ¼ CBF  AVDO2
which is generally 3.2 ml (of O2/100 g of brain
tissue/min) (Gibbs et al., 1942).
Cerebrovascular reactivity
There are two circumstances in which the brain
governs its own flow. In one, CBF changes
proportionally with changes in CMRO2 — meta-
bolic autoregulation. In the other, CBF remains
constant despite changes in blood pressure or ICP
(together CPP ¼ MABP  ICP, where CPP is
cerebral perfusion pressure and MABP is mean
arterial blood pressure) — pressure autoregulation
— and blood viscosity — viscosity autoregulation
(McHenry et al., 1974; Muizelaar et al., 1986;
Muizelaar, 1989).
Metabolic-, pressure-, and viscosity-autoregula-
tion have in common that AVDO2 remains essen-
tially constant (under physiologic conditions),
considering CMRO2 ¼ CBF  AVDO2.
Another important factor to be considered is
carbon dioxide (CO2) reactivity: with hyperventi-
lation (resulting in low blood CO2 levels) cerebral
vasoconstriction occurs with an ensuing lower CBF
and higher AVDO2, whereas with high blood CO2
levels the reverse occurs (Muizelaar et al., 1991).
Autoregulation is fundamentally different from
CO2 reactivity: while in both metabolic and
pressure autoregulation vessel diameter changes
are compensatory responses to maintain a constant
AVDO2, in CO2 reactivity the diameter changes
are primary and CBF and AVDO2 follow
passively. Thus, CO2 reactivity differs from any
before-mentioned type of autoregulation in that
AVDO2 changes. It appears not to be an adaptive
response of the brain to changing circumstances.
Physiology of cerebral blood flow (CBF) and
cerebral blood volume (CBV)
The factors governing CBF are expressed in the
Hagen–Poiseuille equation:
CPP  d 4
CBF ¼ k 
8lv
where k is a constant, CPP cerebral perfusion
pressure in turn defined by mean arterial blood
pressure (MABP) minus intracranial pressure
(ICP), d diameter of the blood vessel, l the length

of the blood vessel (which is practically constant),
and v blood viscosity. The most powerful factor in
this equation is vessel diameter. For instance, the
maximum constriction that can be obtained by
hyperventilation is
20% from normal baseline
(Kontos et al., 1977). This however, leads to a
decrease in CBF of
60% from a normal value of
50 ml/100 g/min to 20 ml/100 g/min. Practically all
of this diameter regulation takes place in the
microcirculation, especially in the arterioles with a
diameter of 300–15 m (Kontos et al., 1977, 1987).
The most intense changes in diameter and there-
fore, cerebral blood volume occur in the microcir-
culation. Although it is unclear how much blood is
to be found in this part of the cerebral circulation, it
is estimated to be one-third of 60 ml of total blood
volume in the brain, i.e., 20 ml under normal con-
ditions (Muizelaar, 1989). With diameter ranging
between 80 and 160% of baseline, this translates
into volume ranging between 64 and 256% of the
baseline 20 ml; 13 ml with maximum vasoconstric-
tion, and 51 ml with maximal vasodilatation.
Although many different factors are essential in
maintaining adequate CBF, it is suggested that
local metabolic factors are of primary importance
in the local tissue regulation. Under normal
circumstances, in areas of increased brain activity,
vasoactive substances are released which alter
vascular tone and local perfusion. Increased per-
fusion then creates a washout effect, which leads to
reduction of perfusion. This feedback system
allows for the modification of CBF for short
periods during times of increased metabolic
requirements. Several metabolic factors play a role
in the autoregulation of CBF under normal con-
ditions, such as CO2/H+ (pH), K+, adenosine,
prostaglandins, and nitric oxide (NO), as well as
serotonin, histamine, neuropeptide Y, vasoactive
intestinal peptide, calcitonin generated peptide,
and others.
Brain tissue oxygen tension
Under physiological conditions, a linear relation-
ship exists between arterial PO2 and brain PO2 with
arterial levels being
90 mm Hg and cerebrovenous
levels
35 mm Hg. Because oxygen is consumed in
113
the tissue, cerebral tissue PO2 is best described as a
continuum that can vary from
90 mm Hg very
close to capillaries to much less than 34 mm Hg in
more distal regions (Zauner et al., 2002).
Reductions of partial pressure of arterial oxygen
(PaO2) with normal rates of CBF also lead to func-
tional deficits. A reduction of PaO2 to 65 mm Hg
induces in humans an impaired ability to perform
complex tasks. Short-term memory is impaired at
55 mm Hg. A PaO2 of 30 mm Hg causes loss of
consciousness (Siesjö, 1978). In animal models,
PaO2 reduction to 36 mm Hg cause intracellular
acidosis, reductions in phosphocreatine (PCr) and
ATP, and increases in intracellular lactate levels
(Xiong et al., 1997). Normal human brain has a
critical tissue PO2 between 15 and 20 mm Hg, below
which infarction (depending on the duration) of
tissue may occur (Fleckenstein et al., 1990; Maas
et al., 1993; Meixenberger et al., 1993; Kiening
et al., 1996; van Santbrink et al., 1996; Wu and
Saggau, 1997; Zauner et al., 1997; Van den Brink
et al., 2000).
Neuronal mitochondria require an intracellular
PO2 of at least 1.5 mm Hg to maintain aerobic
metabolism (Chance et al., 1973; Siesjö and Siesjö,
1996). If cellular PO2 is low, the driving force to
deliver oxygen to the mitochondria is dramatically
reduced. The minimum tissue PO2 required to
provide sufficient intracellular oxygen is unknown.
In addition, it has been proposed that the diffusion
distance for oxygen from the microvasculature
may increase after TBI due to astrocytic swelling,
generalized tissue swelling, and tissue damage. In
these situations the brain might require higher
tissue oxygen tensions to maintain sufficient tissue
oxygenation (Zauner et al., 2002).
Oxygen and hemoglobin
Flow of oxygen from the alveoli to the mitochon-
dria in the brain is dependent on hemoglobin
and PO2.
Once oxygen has diffused from the alveoli into
pulmonary blood, it is transported by hemoglobin
to the cerebral tissue capillaries where it is released
for use, by mitochondria. The presence of hemo-
globin in the erythrocytes of the blood allows
114
Fig. 2. The oxygen–hemoglobin saturation curve, including the effect of hypothermia and hyperventilation. (Adapted with permission
from Guyton, 1986.)
transporting 30–100 times as much oxygen as could
be transported simply in the form of dissolved
oxygen in blood without hemoglobin. The affinity
of oxygen for hemoglobin is best expressed by the
oxygen–hemoglobin saturation curve as seen in
Fig. 2 (Guyton, 1986). This physiological system is
unique in that it favors the binding of oxygen to
hemoglobin in the lungs and the release of oxygen
in the periphery. In the lungs, the curve favors

100% hemoglobin saturation at or around
normal alveolar oxygen tensions, whereas in the
periphery the rate of oxygen delivery is propor-
tional to the difference in oxygen partial pressure
(PO2) between capillary blood and the tissue cells.
Oxygen use, by the mitochondria, is responsible for
creating the driving force for oxygen delivery.
It is interesting to note the effect of temperature
and hyperventilation on tissue oxygenation. A
decrease in temperature as well as hyperventilation
(alkalosis), shifts P50 (i.e., the oxygen tension at
which hemoglobin is 50% saturated) to the left thus
reducing tissue oxygenation due to increased Hb-O2
affinity and thus, decreased O2 unloading to tissues.
Cerebral energy metabolism
Under normal circumstances, the brain requires
large amounts of energy. Although the brain com-
prises only 2–3% of whole body weight, up to 20%
of energy generated in the whole body is used by
the brain.
Fifty percent of the energy produced by the brain
is needed for synaptic activity; 25% is used for
restoring ionic gradients across the cell membrane.
The remaining energy is spent on biosynthesis
such as maintaining membrane integrity and other
processes. If the synthesis of ATP is insufficient,
homeostatic mechanisms deteriorate, intracellular
concentration of calcium increases, and cell death is
inevitable.
Most of the energy is consumed by the neurons.
Although glial cells account for almost half of the
brain volume, they have a much lower metabolic
rate and account for less than 10% of total cerebral
energy consumption (Siesjo, 1984).
Under normal conditions, almost all energy in
our body is produced by aerobic metabolism

(Stryer, 1988). Krebs described three stages in the
generation of energy:
– Large molecules in food are broken down into
smaller units. Proteins are hydrolyzed to
amino acids, polysaccharides are hydrolyzed
to simple sugars such as glucose, and fats are
hydrolyzed to glycerol and fatty acids. No
useful energy is generated in this phase.
– These numerous small molecules are degraded
to a few simple units that play a central role in
metabolism. Most of them are converted in
the acetyl unit of acetyl co-enzyme A (acetyl
CoA). A small amount of ATP is generated at
this stage.
– Acetyl-CoA brings acetyl units into the citric
acid cycle, where they are completely oxidized
to CO2. Four pairs of electrons are transferred
to NAD+ and FAD for each acetyl group
that is oxidized. Then ATP is generated as
electrons flow from the reduced forms of these
carriers to O2 during oxidative phosphorylat-
ion. Thus, most of the energy is generated in
the third stage.
The metabolic patterns of the brain are strik-
ingly different from other organs in their use of
fuel to meet their energy needs (Guyton, 1986).
Glucose is virtually the sole fuel for the human
brain, except during prolonged starvation. The
brain lacks fuel stores and hence requires a con-
tinuous supply of glucose, which enters freely at all
times. It consumes
120 g daily, which corres-
ponds to an energy input of
420 kcal. The brain
accounts for some 60% of utilization of glucose by
the whole body in resting state. During starvation,
ketone bodies (acetoacetate and 3-hydroxybuty-
rate) partly replace glucose as fuel for the brain.
Acetoacetate is activated by the transfer of CoA
from succinyl CoA to give acetoacetyl CoA.
Cleavage by thiolase then yields two molecules
of acetyl CoA, which enter the citric acid cycle.
Fatty acids do not serve as fuel for the brain
because they are bound to albumin in plasma and
so they do not traverse the blood-brain barrier.
In essence, ketone bodies are transportable
equivalents of fatty acids.
115
Glycolysis
Figure 3b illustrates the steps in the glycolysis that
are carried out in the cytoplasm (Stryer, 1988;
Zauner et al., 2002). GLUT-1 transports glucose
across the blood-brain barrier. GLUT-1 also
mediates uptake into the astrocyte while GLUT-3
does the same for neurons. The expression of these
GLUT transporters is up-regulated in experimental
models of hypoxia. The latter results in increased
import of glucose for energy production.
Glycolysis is regulated by the enzyme phospho-
fructokinase-1. Increased ATP demand will activate
this enzyme by increasing cellular cAMP levels
and thereby increasing the rate of glycolytic ATP
generation.
If mitochondria become dysfunctional, even
after restoring blood flow, a small amount of
ATP can still be formed by glycolysis because this
process does not require oxygen. In this process
only a few percent of the total energy in the
glucose molecule is released.
The law of mass action states that as the end
products of a chemical reaction build up in the re-
acting medium the rate of the reaction approaches
zero, thus preventing further production of ATP.
The two end products in the glycolytic reactions
(pyruvate and hydrogen ions) are combined with
NAD+ to form NADH and H+. The quantities of
these end products increase and react with each
other to form lactic acid. This lactic acid can diffuse
readily into the extracellular fluids and even into
the intracellular fluids of other less active cells.
Therefore, lactic acid represents an ‘‘escape’’ into
which the glycolytic end products can be directed,
thus allowing glycolysis to proceed far longer than
would be possible if the pyruvate and hydrogen
were not removed from the reacting medium.
Glycolysis could proceed for only seconds without
this conversion. Instead, it can proceed for several
minutes, supplying the body with ‘‘considerable’’
quantities of ATP. In the human at rest,
5–10%
of the glucose consumed by the body manifests as
a net output of lactate into blood (Siesjö, 1978;
Guyton, 1986; Sokoloff, 1989; Andersen and
Marmarou, 1992).
Once pyruvate has been synthesized it can either
reversibly be converted to lactate and accumulated,
116

Fig. 3. (a) Schematic showing glycolysis and Krebs cycle. (Adapted with permission from Magistretti et al., 1999; Zauner et al., 2002.)
(b) Schematic showing mitochondrial electron transport. (Adapted with permission from Magistretti et al., 1999; Zauner et al., 2002.)
(c) Schematic showing the Magistretti model of coupled metabolism. (Adapted with permission from Magistretti et al., 1999; Zauner
et al., 2002.)

converted to amino acid alanine, or enter the
citric acid cycle to produce energy via oxidative
phosphorylation.
Citric acid cycle
The citric acid cycle that takes place in the mi-
tochondria is shown in Fig. 3a (Stryer, 1988;
Zauner et al., 2002). Under aerobic conditions, the
next step in the aerobic generation of energy from
glucose is the oxidative decarboxylation of pyruv-
ate to form acetyl CoA. This activated acetyl unit
is then completely oxidized to CO2 by the citric
acid cycle, a series of reactions that is also known
as the tricarboxylic acid cycle or the Krebs cycle.
The citric acid cycle is the final common pathway
for the oxidation of fuel molecules; it also serves as
a source of building blocks for biosynthesis.
Oxidative phosphorylation
Figure 3b shows the electron transport chain
(Stryer, 1988; Zauner et al., 2002). The NADH
and FADH2 formed in glycolysis, fatty acid oxi-
dation, and the citric acid cycle are energy-rich
molecules because each contains a pair of electrons
with a high transfer potential. These electrons are
subsequently donated to molecular oxygen, result-
ing in a large amount of free energy, which can be
used to generate ATP. Oxidative phosphorylation
is the process in which ATP is formed as electrons
are transferred from NADH or FADH2 to O2 by a
series of electron carriers. This process acts as a
major source of ATP in aerobic organisms. Some
salient features of this process are (Stryer, 1988):
– Respiratory assemblies that are located in the
inner membrane of mitochondria carry out
oxidative phosphorylation. The citric acid
cycle and the pathway of the fatty acid
oxidation, which supply most of the NADH
and FADH2, are in the adjacent mitochon-
drial matrix.
– The oxidation of NADH yields 3 ATP,
whereas the oxidation of FADH2 yields 2
ATP. Oxidation and phosphorylation are
coupled processes.
117
– Respiratory assemblies contain numerous
electron carriers, such as the cytochromes.
The step-by-step transfer of electrons from
NADH or FADH2 to O2 through these car-
riers leads to pumping of protons out of the
mitochondrial matrix. A proton-motive force
is generated consisting of a pH gradient and a
transmembrane electric potential. ATP is
synthesized when protons flow back to the
mitochondrial matrix through an F0F1 ATP
synthase complex. Thus oxidation and phos-
phorylation are coupled by a proton gradient
across the inner mitochondrial membrane.
Traditionally cerebral energy production has
been considered to consist mainly of aerobic
metabolism of glucose. Although it has long been
assumed that glia and neurons use glucose as their
sole energy source, recent information has sug-
gested otherwise; astrocytes may have the ability
to transport glucose across the blood-brain barrier
via GLUT-1 and anaerobically metabolize it to
lactate. Lactate is then released into the extracel-
lular space, where it is taken up by neurons and
consumed aerobically to generate energy as seen in
Fig. 3c (Vibulsreth et al., 1987; Walz and Mukerji,
1988; Magistretti et al., 1999).
With increasing neuronal activity, potassium and
glutamate are released into the extracellular space
and are taken up by the astrocytes in an energy-
dependent fashion causing increased astrocytic
glycolysis. In traumatic brain injury conditions,
aerobic metabolism is diminished due to reductions
in cellular oxygen, or due to mitochondrial dys-
function, causing increased lactate accumulation.
Mitochondria
Mitochondria are oval-shaped organelles, typically

2 mm in length and 0.5 mm in diameter. Techniques
for isolating mitochondria were devised in the late
1940s. Eugene Kennedy and Albert Lehninger sub-
sequently discovered that mitochondria contain the
respiratory assembly, the enzymes of both the citric
acid cycle and fatty acid oxidation. Electron micro-
scopic studies by George Palade and Fritjof
Sjöstrand revealed that mitochondria have two
membrane systems: an outer membrane and a
118
Fig. 4. Schematic of mitochondrion.
highly folded inner membrane. The inner mem-
brane is folded into a series of internal ridges called
cristae. Hence, there are two compartments in
mitochondria: the intermembrane space between
the outer and inner membranes, and the matrix,
which is bounded by the inner membrane (Fig. 4).
Oxidative phosphorylation takes place in the
inner mitochondrial membrane, in contrast with
most of the reactions of the citric acid cycle and
fatty acid oxidation, which occur in the matrix.
The outer membrane is quite permeable to most
small molecules and ions because it contains many
copies of porin, a transmembrane protein with a
large pore. In contrast, the inner membrane is
intrinsically impermeable to nearly all ions and
polar molecules. Specific protein carriers transport
molecules such as adenosine di-phosphate (ADP)
and long chain fatty acids across the inner mito-
chondrial membrane.
Pathophysiology
Traumatic brain damage injury may be divided
into primary and secondary types of injury
(Verweij and Muizelaar, 1996). Mechanical forces
acting at the moment of injury damage the blood
vessels, axons, neurons, and glia of the brain
initiating an evolving sequence of secondary
changes that result in complex cellular, inflamma-
tory, neurochemical, and metabolic alterations.
It is not within the scope of this review to
describe all the changes occurring in the brain after
severe head injury; only those directly related to
cerebral oxygen transport and energy metabolism
will be mentioned: Traumatic intracranial hema-
tomas (intraparenchymal, subdural, and epidural)
are also common after head injury. Epidural hem-
atomas (in up to 5% of all patients admitted to
hospitals for head injury, and 9% of those with
severe head injury) are often the result of skull
fractures causing rupture of underlying arteries or
veins. If arterial in origin they can enlarge very
quickly and cause rapid neurological deterioration.
If surgical intervention is prompt, and no other
brain injury is present, outcome can be favorable.
Acute subdural hematoma (ASDH) occurs in up
to 25% of all patients with severe head injury
(Richards and Hoff, 1974; Hubschmann and
Nathanson, 1985). In comatose patients with TBI,
ASDH carries the highest mortality rate: 60–90
(Jamieson and Yelland, 1972; Gennarelli et al.,
1989; Wilberger et al., 1991). In this group outcome
is strikingly unfavorable due to decreased energy
metabolism. On one hand, decreased energy
metabolism is due to ischemia caused by increased
intracranial pressure and therefore decreased per-
fusion pressure; on the other hand due to the un-
derlying damaged brain being unable to use oxygen
because of damaged mitochondria (Verweij et al.,
2000b, 2001).
Hypotension and hypoxia
The majority of the potential clinical events after
neurotrauma have been investigated with respect to
their frequency of occurrence and impact on out-
come, both for the prehospital and intensive care
unit (ICU) periods (Chesnut et al., 1993; Jones
et al., 1994). These studies have uniformly identi-
fied hypotension (SABPo90 mm Hg) and hypoxia
(PaO2o60 torr) as the most influential. These
parameters, amenable to therapeutic manipulation,
seem to be the most significant predictors of poor
outcome, independent of their etiologies and
pre-resuscitation of secondary insults.
Admission to the ICU does not eliminate
secondary brain injury. Jones et al. (1994) have
reported the results of computerized online
evaluation of 14 variables in 124 head-injured
patients of a variety of grades admitted to the

neurosurgical ICU. More than one episode of
hypotension occurred in 73% of all patients,
with median durations of 29 min (SABPo90 mm
Hg), 22 min (SABPo80 mm Hg), and 32 min
(SABPo70 mm Hg). In 40% of all cases, more
than one episode of hypoxia occurred with an av-
erage duration of 12 min (PaO2o60 torr), 19 min
(PaO2o52 torr), and 20 min (PaO2o45 torr). It
has also been demonstrated that these secondary
insults do not only occur in the ICU, but also
during patient transport in X-ray and in OR suites
(Andrews et al., 1990).
ICP and cerebral circulation
According to the Monro–Kellie doctrine, ICP is
governed by the interplay between the volumes of
brain (including cytotoxic edema), cerebrospinal
fluid (including vasogenic or extracellular edema)
and the blood within the cerebral blood vessels, all
of which is within the confines of the rigid skull
(Monro, 1783; Kellie, 1824). An increase in vol-
ume in one of these compartments (or epidural,
subdural, or intracerebral hematoma) leads to a
rise in ICP, unless this increase is compensated by
an equal decrease in one of the two remaining
compartments. The natural defense against rising
ICP with brain swelling is the displacement of CSF
from the skull: hence small compressed ventricles
and absence of basal cisterns on computer tomo-
graphy (CT) scans after severe trauma. Sometimes
this process can be palliated by drainage through a
ventricular catheter.
When one considers the Hagen–Poiseuille equa-
tion, there are two practical methods of maintaining
CBF during vasoconstriction. The first is to increase
CPP, which can be done by raising the blood pres-
sure. When pressure autoregulation is intact, this
maneuver in and of itself will cause vasoconstric-
tion, and this has occasionally been used to decrease
ICP (Muizelaar, 1989). More important, however,
is the need to avoid low blood pressure, and hence,
the effect of ‘‘perfusion pressure therapy’’ may be
due in part to the simple avoidance of arterial hy-
potension (Rosner and Daughton, 1990; Bouma
et al., 1992a; Rosner et al., 1995). The second
method to maintain CBF in the face of decreasing
119
vessel diameter is to decrease blood viscosity.
Again, this decrease itself leads to vasoconstriction
if viscosity autoregulation is intact, and it has been
argued that the viscosity lowering effect mediates a
good deal of mannitol’s effect on ICP (besides the
osmotic effect) (Muizelaar et al., 1983, 1984, 1986).
When viscosity autoregulation is not intact, lower-
ing viscosity with mannitol can maintain CBF de-
spite cerebral vasoconstriction associated with
hyperventilation (Cruz et al., 1990). As stated pre-
viously, CBV can vary between 13 and 51 ml in the
microcirculation. To comprehend what these differ-
ences in volume mean to ICP, one must consider the
pressure volume index (PVI) (Marmarou et al.,
1978). Pressure volume index is defined as the
amount of fluid (in ml) needed to add to the intra-
cranial space to make ICP rise tenfold (or withdraw
to decrease ICP tenfold):
DV
PVI ¼
Log½ICPbefore =ICPafter 
Normal PVI is 20–25 ml (Shapiro et al., 1980).
However PVI has been observed as low as 6 ml
after severe head injury, which indicates that in
going from normoventilation (PaCO2 ¼ 30 mm
Hg) to strong hyperventilation (PaCO2 ¼ 18 mm
Hg) resulting in vasoconstriction, ICP could
theoretically be decreased tenfold (Bouma et al.,
1992a). (That this is not always desirable may be
clear from the following example: PaCO2 36 mm
Hg, ICP 40 mm Hg, MABP 100 mm Hg, CBF
30 ml/100 g/min; AVDO2 6 vol% - CMRO2
1.8 ml/100 g/min; now with hyperventilation to
get ICP below the desired 20 mm Hg: PaCO2 ¼ 22,
ICP 20 mm Hg, MABP 100 mg Hg - CPP from 60
to 80 - CBF to 40 ml/100 g/min.) However, be-
cause of 20% vasoconstriction and diameter being
to the fourth power in the Hagen–Poiseuille equa-
tion, CBF drops to 16 ml — well below the thresh-
old for infarction (Jones et al., 1981), and especially
so after severe head injury. Moreover, as AVDO2
cannot rise above 10, CMRO2 will drop to 1.6 ml/
100 g/min.
If ICP is uncontrollable barbiturates are some-
times administered, lowering the cerebral meta-
bolic demands. If metabolic autoregulation is still
intact this will result in decreased blood volume
and therefore reduced ICP.
120
All of these examples are common in clinical
practice, and thus it may be obvious that a good
monitor is required to guide the management of
ICP, blood pressure, ventilatory parameters, and
blood viscosity. Although monitoring of CBF and/
or AVDO2 is ideal, there are no practical ways to
do this continuously; therefore the authors revert
to a derivate of AVDO2, i.e., SjvO2 (Cruz, 1988;
Robertson et al., 1989; Gopinath et al., 1994).
Ischemia and CMRO2
Secondary cerebral ischemia is very common after
severe head injury and is associated with an un-
favorable outcome (Graham and Adams, 1971;
Bouma et al., 1991; Siesjö and Siesjö, 1996). As
discussed earlier, CBF is closely regulated by a
number of mechanisms. However, after trauma
these mechanisms can fail resulting in ischemia.
This is especially important as the brain seems more
vulnerable to ischemia after trauma and this vul-
nerability persists for at least 24 h (Jenkins, 1989). It
has also been demonstrated that CBF is low in the
hyperacute post-traumatic period (Bouma et al.,
1991, 1992b). Increased intracranial pressure (as in
cerebral edema and subdural hematoma) and there-
fore decreased cerebral perfusion pressure appears
also to be an important cause of ischemia as well as
too vigorous hyperventilation (Muizelaar et al.,
1991; Verweij et al., 2001).
Salvant and Muizelaar suggested that a parallel
reduction in CBF and CMRO2 without an increase
in AVDO2 is consistent with diminished metabo-
lism and may be due to mitochondrial dysfunction.
Verweij et al. (2000b) demonstrated mitochondrial
dysfunction in isolated mitochondria from human
tissue. In a normal coupled relationship between
AVDO2 and CBF (AVDO2 ¼ CMRO2/CBF),
AVDO2 remains unchanged when the CMRO2
changes. If however CMRO2 remains constant,
changes in AVDO2 reflect uncoupled changes in
CBF (Robertson et al., 1989). If CBF decreases
following head injury, AVDO2 will increase as the
brain compensates by extracting a greater amount
of oxygen. A further uncompensated decline in
CBF leads to ischemia and a fall in CMRO2.
Lactate/hyperglycolysis
In animals and humans it has been repeatedly
shown that TBI induces increased brain lactate
production, which normalizes gradually after the
first few days in those who survive, but remains at
levels five to ten times normal in those who suc-
cumb. Microdialysis studies have demonstrated
that extracellular fluid glucose declines to extremely
low levels when lactate is increasing (Goodman
et al., 1999; Zauner et al., 2002). If aerobic metab-
olism fails, anaerobic metabolism remains, resulting
in hyperglycolysis and lactate production. A shift to
anaerobic metabolism will occur if ischemia is
present. If brain oxygen tension levels fall below

20 mm Hg aerobic metabolism ceases to occur
(Zauner et al., 1997). However, if mitochondrial
dysfunction occurs, aerobic metabolism will also
shift toward anaerobic metabolism inducing the
same phenomena (Verweij et al., 2000b).
In human positron emission tomography (PET)
studies, increases in glucose metabolism are demo-
nstrated especially in the zone around contusions
and in the hemisphere underlying hematomas
(Bergsneider et al., 1997). Such tissues are often
adjacent to ‘‘low-density’’ cytotoxic edema areas
on CT scan. Human PET studies at later time
points (1–4 weeks post-ictus) and mitochondrial
analyses in the acute stage have shown uniformly
decreased metabolism, both for glucose and oxy-
gen in humans (Verweij et al., 2000b, 2001).
Marmarou as well as others have shown through
animal experiments that anaerobic cerebral metab-
olism with generation of lactate appears to occur
even in the absence of blood flow limitation (De
Salles et al., 1987). It has also been found by meas-
uring brain oxygenation, CO2 generation, pH and
temperature in parallel with extracellular fluid lac-
tate, and glucose levels measured by microdialysis
that lactate generation is increased in
65% of
measurements, even in the presence of adequate
CBF and brain oxygen levels (Zauner et al., 1997).
Mitochondrial dysfunction will explain lactate
production in these circumstances (Verweij et al.,
2000b).
High levels of ECF lactate might be harmful
to the injured brain. Marked cerebral acidosis
may exacerbate calcium-mediated damage to

intracellular enzyme systems and may also inter-
fere with ion-channel function (Siesjo, 1992). High
tissue lactate levels could foster a decline in brain
pH, as has been shown in numerous post-
traumatic animal and human studies.
The restoration of circulation after ischemia is
accompanied by normalization of the tissue lactate
concentration and the lactate-to-pyruvate ratio.
Initially there is an increase in pyruvate concen-
tration as lactate is converted back to pyruvate.
Animal studies have demonstrated that if the
resuscitation interval is prolonged, tissue lactate
remains elevated during reperfusion, suggesting
that residual tissue lactic acidosis is a sign of
mitochondrial dysfunction. Since mitochondrial
dysfunction is reversible, therapeutic intervention
might be possible (Verweij et al., 1997; Xiong
et al., 1998; Berman et al., 2000).
Rapid reversal of acidosis may be unfavorable,
as mild acidosis might be beneficial (pH paradox)
during recovery from hypoxia. There is little direct
evidence to demonstrate that lactate alone, or
substantial intracellular acidosis alone, is toxic to
normal cerebral tissues. This may be attributable
to preserved high-energy phosphate concentra-
tions, which allow potential intracellular buffering
and transport of hydrogen ions from the cell.
During ischemia, however, acidosis may injure
neurons by denaturation of proteins, lead to dam-
age of astrocytes owing to failure of membrane
transport systems, and cause promotion of iron-
dependent free radical formation. These events can
cause the inhibition of glycolysis by the complete
inhibition of the glycolytic phosphofructokinase at
a pH of 6.5 or below.
Mild acidosis might be protective in vitro and in
vivo in models of ischemia/hypoxia, by slowing
enzymatic processes and reducing energy consump-
tion and free radical production. Nonetheless,
tissue acidosis is consistently associated with
worsened ischemic outcome in vivo, which may
be augmented by hyperglycemia.
Cerebral tissue PO2/SjvO2
Cerebral oxygenation is currently monitored in two
ways: (1) measurement of jugular bulb oxygen sat-
uration, which reflects the relationship between
121
oxygen delivery to the brain and the extraction of
oxygen by the brain (Cruz, 1988; Cruz et al., 1990;
Gopinath et al., 1994); (2) local brain tissue oxime-
try by a single Clark-type electrode (or multi-
electrode also measuring PCO2, pH, and tempera-
ture), which gives information of local cerebral
tissue PO2. The relationship between SjvO2, cerebral
tissue PO2, and CBF has been well documented.
An increase in the concentration of tissue lactate
in the brain indicates a shift from aerobic to an-
aerobic metabolism in an attempt to maintain ATP
production. This shift occurs in case of low cere-
bral tissue PO2 that has been shown to occur in

25–39% of patients with severe TBI in the first
12 h post- injury. Low brain tissue PO2 also closely
correlates with low regional CBF. It has also been
shown that low brain tissue PO2 is strongly corre-
lated with high levels of dialysate lactate in the
brain (Zauner et al., 2002). Increasing the concen-
tration of inspired oxygen to 100% has shown to
increase brain tissue PO2. In a study by Menzel
et al. (1999), statistically significant correlation
existed between brain tissue PO2 and CPP. Brain
lactate measured by microdialysis remained high
during the entire period rising later on. No corre-
lations were found between ICP, CPP, brain tissue
PO2, or lactate (brain tissue PO2 increases over the
first 30 h with an overshoot at 36–48 h). The latter
could be explained by mitochondrial dysfunction
(Verweij et al., 2000b).
Mitochondrial function after severe head injury
Mitochondria play a vital role in cell survival and
tissue development by virtue of their role in energy
metabolism and apoptosis (Nicholls and Budd,
2000; Friberg and Wieloch, 2002). Since neuronal
tissue stores and anaerobic glycolysis provides
ATP sufficient to maintain cellular function for
only 1–2 min, mitochondrial generation of ATP is
of vital importance (Siesjo, 1992). Neuronal mito-
chondria have a high capacity to store calcium
ions, thereby protecting neurons against transient
elevations in intracellular calcium concentrations
during neuronal hyperactivity. This calcium se-
questration requires negative mitochondrial matrix
potential and therefore full functional mito-
chondria with access to oxygen and pyruvate.
122

In the past, research efforts in patients with Abbreviations

severe head injury have focused on optimizing the
delivery of oxygen and glucose to the injured brain ADP adenosine diphosphate
in an attempt to maintain the ATP supply and to AMP adenosine monophosphate
avoid neuronal damage. However, limiting factors ASDH traumatic acute subdural hematoma
in synthesizing ATP are not only inadequate ATP adenosine triphosphate
delivery of oxygen and glucose but also impair- AVDO2 arteriovenous oxygen content
ment of mitochondrial function (Verweij et al., difference
1997, 2000b). Severe TBI with or without hypoxia CBF cerebral blood flow
and or ischemia results in a number of biochemical CBV cerebral blood volume
processes such as amino acid efflux and oxygen free CMRO2 cerebral metabolic rate of oxygen
radical production. This ultimately leads to mas- CoA co-enzyme A
sive ion shifts with increased calcium in the intra- CPP cerebral perfusion pressure
cellular compartment (Fineman et al., 1993). It has CSF cerebrospinal fluid
previously been demonstrated that experimental CT computerized tomography
TBI perturbs calcium homeostasis with an over- EEG electro-encephalogram
load of cytosolic calcium and excessive calcium FAD flavin adenine dinucleotide
adsorption by the mitochondrial membranes GCS Glasgow coma scale
(Sciamanna et al., 1992; Verweij et al., 1997; Xiong GLUT glucose transporter
et al., 1997). This inhibits mitochondrial function, ICP intracranial pressure
even when there are sufficient oxygen and substrate MABP mean arterial blood pressure
present. In rats, mitochondrial dysfunction begins NAD nicotinamide adenine dinucleotide
1 h after TBI and persists for at least 14 days, with NIRS near infra red spectroscopy
the maximum level of dysfunction occurring at NO nitric oxide
12–72 h (Verweij et al., 1997). The same phenom- PaCO2 arterial PCO2 tension
ena occur in patients (Verweij et al., 2000b). PaO2 arterial oxygen partial pressure
A number of critical mechanisms have been PET positron emission tomography
elucidated by which mitochondria are involved PO2 oxygen partial pressure
in cell death. Elevated cytosolic Ca2+ and oxida- PVI pressure volume index
tive stress both contribute to the opening of the SABP systolic arterial blood pressure
mitochondrial permeability transition pore (PTP), SjvO2 jugular venous oxygen saturation
which depolarizes the mitochondrion and leads to TBI traumatic brain injury
mitochondrial swelling and subsequent release of
cytochrome ‘‘c’’ from the intermembrane space.
Cytochrome ‘‘c’’ normally functions as part of the References
respiratory chain, but when released into the
cytosol (as a result of PTP opening) it becomes a Andersen, B.J. and Marmarou, A. (1992) Functional compart-
critical component of the apoptosis execution mentalization of energy production in neural tissue. Brain
Res., 585: 190–195.
machinery, where it activates caspases (cysteine
Andrews, P.J., Piper, I.R., Dearden, N.M., et al. (1990) Seco-
aspartate proteases) and (if ATP is available) ndary insults during intrahospital transport of head injured
causes apoptotic cell death (Young, 1992). Thus, a patients. Lancet, 335(8685): 327–330.
number of new targets for therapeutic intervention Astrup, J. (1982) Energy requiring cell functions in the ischemic
have emerged (Verweij et al., 1997, 2000a; Xiong brain: their critical supply and possible inhibition in protec-
tive therapy. J. Neurosurg., 56: 482–497.
et al., 1998; Berman et al., 2000). These pos-
Bergsneider, M., Hovda, D.A., Shalmon, E., et al. (1997) Cere-
sible interventions have recently been described bral hyperglycolysis following severe traumatic brain injury
in a well-written review by Merenda and in humans: a positron emission tomographic study. J.
Bullock (2006). Neurosurg., 86: 241–251.

Berman, R.F., Verweij, B.H. and Muizelaar, J.P. (2000)
Neurobehavioral protection by the neuronal calcium chan-
nel blocker ziconotide in a model of traumatic diffuse brain
injury in rats. J. Neurosurg., 93(5): 821–828.
Bouma, G.J., Muizelaar, J.P., Bandoh, K., et al. (1992a) Blood
pressure and intracranial pressure-volume dynamics in severe
head injury: relationship with cerebral blood flow. J. Neuro-
surg., 77(1): 15–19.
Bouma, G.J., Muizelaar, J.P., Choi, S.C., et al. (1991) Cerebral
bloodflow and metabolism after severe traumatic brain injury:
the elusive role of ischemia. J. Neurosurg., 75(5): 685–693.
Bouma, G.J., Muizelaar, J.P., Stringer, W.A., et al. (1992b)
Ultra-early evaluation of regional cerebral blood flow in
severely head-injured patients using xenon-enhanced
computirized tomography. J. Neurosurg., 77: 360–368.
van den Brink, W.A., van Santbrink, H., Steyerberg, E.W.,
et al. (2000) Brain oxygen tension in severe head injury.
Neurosurgery, 46(4): 868–876.
Chance, B., Oshimo, N., Sugamo, T. and Mayersky, A. (1973)
Basic principles of tissue oxygen determination from mito-
chondrial signals. In: Bruley D.F. (Ed.), Oxygen Transport to
Tissue. Plenum Publishing, New York, pp. 277–292.
Chesnut, R.M., Marshall, S.B., Piek, J., et al. (1993) Early and
late systemic hypotension as a frequent source of cerebral
ischemia following severe brain injury in the traumatic coma
data bank. Acta Neurochir. Suppl. (Wein.), 59: 121–125.
Cruz, J. (1988) Continuous versus serial global hemometabolic
monitoring: application in acute brain trauma. Acta
Neurochir. (Wien.), 42: 35–39.
Cruz, J., Miner, M.E., Allen, S.J., et al. (1990) Continuous
monitoring of cerebral oxygenation in acute brain injury:
injection of mannitol during hyperventilation. J. Neurosurg.,
73(5): 725–730.
De Salles, A.A.F., Muizelaar, J.P. and Young, H.F. (1987)
Hyperglycemia, cerebrospinal fluid lactic acidosis, and cere-
bral blood flow in severely head injured patients. Neurosur-
gery, 21: 45–50.
Fineman, I., Hovda, D.A. and Smith, M. (1993) Concussive brain
injury is associated with a prolonged accumulation of calcium:
a 45Ca autoradiographic study. Brain Res., 624: 94–102.
Fleckenstein, W., Maas, A.I., Nollert, G. and de Long, D.A.
(1990) Oxygen pressure in cerebrospinal fluid. In: Flecken-
stein W. and Maas A.I. (Eds.), Oxygen Transport. Blackwell,
Berlin, pp. 368–393.
Friberg, H. and Wieloch, T. (2002) Mitochondrial permeability
transition in neurodegeneration. Biochimie, 84: 241–250.
Gennarelli, T.A., Champion, H.R. and Sacco, W.J. (1989)
Mortality of patients with head injury and extracranial injury
treated in trauma centers. J. Trauma, 29: 1193–1202.
Gibbs, E.L., Lennox, W.G., Nims, L.F., et al. (1942) Arterial
and cerebral venous blood: arterial venous differences in
man. J. Biol. Chem., 144: 325–332.
Goodman, J.C., Valadka, A.B., Gopinath, S.P., Uzura, M. and
Robertson, C.S. (1999) Extracellular lactate and glucose al-
terations in the brain after head injury measured by micro-
dialysis. Crit. Care Med., 27: 1965–1973.
123
Gopinath, S.P., Robertson, C.S., Contant, C.F., et al. (1994)
Jugular venous desaturation and outcome after head injury.
J. Neurol. Neurosurg. Psychiatry, 57(6): 717–723.
Graham, D.I. and Adams, J.H. (1971) Ischemic brain damage
in fatal head injuries. Lancet, 1: 265–266.
Guyton, A.C. (1986) Textbook of Medical Physiology (Seventh
edition). W.B. Saunders Company, Philadelphia, pp. 481–525.
Hubschmann, O.R. and Nathanson, D.C. (1985) The role of
calcium and cellular membrane dysfunction in experimental
trauma and subarachnoid hemorrhage. J. Neurosurg., 6:
698–703.
Jamieson, K.G. and Yelland, J.D.N. (1972) Surgically treated
traumatic subdural hematomas. J. Neurosurg., 37: 137–149.
Jones, P.A., Andrews, P.J., Midgley, S., et al. (1994) Measuring
the burden of secondary insults in head-injured patients dur-
ing intensive care. J. Neurosurg. Anesthesiol., 6(1): 4–14.
Jones, T., Morawetz, R., Crowell, R.M., et al. (1981) Threshold
of focal cerebral ischemia in awake monkeys. J. Neurosurg.,
54: 773–782.
Kellie, G. (1824) On death from cold, and on congestions of the
brain: an account of the appearances observed in the dissec-
tion of two of three individuals presumed to have perished in
the storm of 3rd November, 1821; with some reflections on
the pathology of the brain. TRANS Med. Chir. Soc., Edin-
burgh, 1: 84–169.
Kiening, K.L., Unterberg, A.W., Bardt, T.F., Schneider, G.H.
and Lanksch, W.R. (1996) Monitoring of cerebral oxygenation
in patients with severe head injuries. J. Neurosurg., 85: 751–757.
Kontos, H.A., Raper, A.J. and Patterson, J.L. (1977) Analysis
of vasoactivity of local pH, PCO2, and bicarbonate on pial
vessels. Stroke, 8(3): 358–360.
Kontos, H.A., Wei, E.P., Navari, R.M., et al. (1987) Responses
of cerebral arteries and arterioles to acute hypotension and
hypertension. Am. J. Physiol., 234(4): H371–H383.
Maas, A.I., Fleckenstein, W., de Jong, D.A. and van Santbrink,
H. (1993) Monitoring cerebral oxygenation. Acta Neurochir.
Suppl. (Wien.), 59: 50–57.
Magistretti, P.J., Pellerin, L., Rothman, D.L. and Shulman,
R.G. (1999) Energy on demand. Science, 283: 496–497.
Marmarou, A., Shulman, K. and Rosende, R. (1978) A non-
linear analysis of the cerebrospinal fluid system and intra-
cranial pressure dynamics. J. Neurosurg., 48: 332–344.
McHenry Jr., L.C., West, J.W., Coopers, E.S., et al. (1974)
Cerebral autoregulation in man. Stroke, 5(6): 695–706.
Meixenberger, J., Dings, J., Kuhnigk, H. and Roosen, K. (1993)
Studies of tissue PO2 in normal and pathophysiological hu-
man brain cortex. Acta Neurochir. Suppl. (Wien.), 59: 58–63.
Menzel, M., Doppenberg, E.M.R., Zauner, A., et al. (1999)
Increased inspired oxygen concentration as a factor in im-
proved brain tissue oxygenation and tissue lactate levels after
severe human head injury. J. Neurosurg., 91: 1–10.
Merenda, A. and Bullock, R. (2006) Clinical treatments for
mitochondrial dysfunction after brain injury. Curr. Opin.
Crit. Care, 12: 90–96.
Monro, A. (1783) Observations on the structure and functions
of the nervous system. Creech and Johnson, Edinburgh, p. 5.
124
Muizelaar, J.P. (1989) Cerebral blood flow, cerebral blood
volume, and cerebral metabolism after severe head injury.
In: Becker D.P. and Gudeman S.K. (Eds.), Textbook of
Head Injury. W.B. Saunders Company, Philadelphia, PA,
pp. 221–240.
Muizelaar, J.P., Lutz, H.A. and Becker, D.P. (1984) Effect of
mannitol on ICP and CBF and correlation with pressure
autoregulation in severely head-injured patients. J. Neuro-
surg., 61(4): 700–706.
Muizelaar, J.P., Marmarou, A., Ward, J.D., et al. (1991)
Adverse effects of prolonged hyperventilation in patients
with severe head injury: a randomized clinical trial. J. Neuro-
surg., 75: 731–739.
Muizelaar, J.P., Wei, E.P., Kontos, H.A., et al. (1983) Mannitol
causes compensatory cerebral vasoconstriction and vasodil-
atation in response to blood viscosity changes. J. Neurosurg.,
59(5): 822–828.
Muizelaar, J.P., Wei, E.P., Kontos, H.A., et al. (1986) Cerebral
blood flow is regulated by changes in blood pressure and in
blood viscosity alike. Stroke, 17(1): 44–48.
Nicholls, D.G. and Budd, S.L. (2000) Mitochondrial and neu-
ronal survival. Physiol. Rev., 80: 315–360.
Richards, T. and Hoff, J. (1974) Factors affecting survival from
acute subdural hematoma. Surgery, 75: 253–258.
Robertson, C.S., Narayan, R.K., Gokaslan, Z.L., et al. (1989)
Cerebral arteriovenous oxygen difference as an estimate of
cerebral blood flow in comatose patients. J. Neurosurg.,
70(2): 222–230.
Rosner, M.J. and Daughton, S. (1990) Cerebral perfusion pres-
sure management in head injury. J. Trauma, 30(8): 933–940.
Rosner, M.J., Rosner, S.D. and Johnson, A.H. (1995) Cerebral
perfusion pressure: management protocol and clinical results.
J. Neurosurg., 83(6): 949–962.
Schroder, M., Muizelaar, J.P., Kuta, A., et al. (1996) Thresh-
olds for cerebral ischemia after severe head injury: relation-
ship with late CT findings and outcome. J. Neurotrauma, 13:
17–23.
van Santbrink, H., Maas, A.I. and Avezaat, C.J. (1996)
Continuous monitoring of partial pressure of brain tissue
oxygen in patients with severe head injury. Neurosurgery,
38(1): 21–31.
Sciamanna, M.A., Zinkel, J., Fabi, A.Y., et al. (1992) Ischemic
injury to rat forebrain mitochondria and cellular calcium
homeostasis. Biochim. Biophys. Acta, 1134: 223–232.
Shapiro, K., Marmarou, A. and Shulman, K. (1980) Charac-
terization of clinical CSF dynamics and neural axis compli-
ance using the pressure-volume index: I. The normal pressure
volume index. Ann. Neurol., 7(6): 508–514.
Siesjö, B.K. (1978) Brain Energy Metabolism. Wiley, Chichester.
Siesjö, B.K. (1984) Cerebral circulation and metabolism. J.
Neurosurg., 60(5): 883–908.
Siesjö, B.K. (1992) Pathofysiology and treatment of focal cerebral
ischemia. Part I: pathofysiology. J. Neurosurg., 77: 169–184.
Siesjö, B.K. and Siesjö, P. (1996) Mechanisms of secondary
brain injury. Eur. J. Anaesthesiol., 13: 247–268.
Sokoloff, L. (1960) Metabolism of the central nervous system in
vivo. In: Field J., Magoun H.W. and Hall V.E. (Eds.),
Handbook of Physiology. Section 11: Neurophysiology,
Vol. 3. American Physiological Society, Washington DC,
pp. 1843–1864.
Sokoloff, L. (1989) Circulation and Energy Metabolism of the
Brain. Raven Press, New York, pp. 541–564.
Stryer, L. (1988) Biochemistry (Third edition). W.H. Freeman
and Company, New York, pp. 313–544.
Verweij, B.H. and Muizelaar, J.P. (1996) Avoiding secondary
brain injury after severe head trauma: monitoring and man-
agement. J. Craniomaxillofac. Trauma, 2(3): 8–17.
Verweij, B.H., Muizelaar, J.P. and Vinas, F.C. (2001) Hyper-
acute measurement of intracranial pressure, cerebral perfu-
sion pressure, jugular venous oxygen saturation, and laser
Doppler flowmetry, before and during removal of traumatic
acute subdural hematoma. J. Neurosurg., 95(4): 569–572.
Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L.,
Xiong, Y. and Lee, C.P. (1997) Mitochondrial dysfunction
after experimental and human brain injury and its possible
reversal with a selective N-type calcium channel antagonist
(SNX-111). Neurol. Res., 19(3): 334–339.
Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L.,
Xiong, Y. and Lee, C.P. (2000a) Improvement in mito-
chondrial dysfunction as a new surrogate efficiency measure
for preclinical trials: dose-response and time-window profiles
for administration of the calcium channel blocker Ziconotide in
experimental brain injury. J. Neurosurg., 93(5): 829–834.
Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L.,
Xiong, Y. and Lee, C.P. (2000b) Impaired cerebral mito-
chondrial function after traumatic brain injury in humans. J.
Neurosurg., 93(5): 815–820.
Vibulsreth, S., Hefti, F., Ginsberg, M.D., Dietrich, W.D.
and Busto, R. (1987) Astrocytes protect cultured neurons
from degeneration induced by anoxia. Brain Res., 422:
303–311.
Walz, W. and Mukerji, S. (1988) Lactate production and release
in cultured astrocytes. Neurosci. Lett., 86: 296–300.
Wilberger Jr., J.E., Harris, M. and Diamond, D.L. (1991) Acute
subdural hematoma: morbidity, mortality, and operative
timing. J. Neurosurg., 74: 212–218.
Wu, L.G. and Saggau, P. (1997) Presynaptic inhibition of elic-
ited neurotransmitter release. Trends Neurosci., 20: 204–212.
Xiong, Y., Gu, Q. and Peterson, P.L. (1997) Mitochondrial
dysfunction and calcium perturbation induced by traumatic
brain injury. J. Neurotrauma, 14: 23–34.
Xiong, Y., Peterson, P.L., Verweij, B.H., et al. (1998) Mi-
tochondrial dysfunction after experimental traumatic brain
injury: combined efficacy of SNX-111 and U-101033E.
J. Neurotrauma, 15(7): 531–544.
Young, W. (1992) Role of calcium in central nervous systems
injuries. J. Neurotrauma, 9(Suppl 1): S9–S25.
Zauner, A., Daugherty, W.P., Bullock, M.R., et al. (2002) Brain
oxygenation and energy metabolism: Part 1-Biological func-
tion and pathophysiology. Neurosurgery, 51: 289–302.
Zauner, A., Doppenberg, E.M., Woodward, J.J., Choi, S.C.,
Young, H.F. and Bullock, R. (1997) Continuous monitoring
of cerebral substrate delivery and clearance. Neurosurgery,
41: 1082–1091.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 9

Progressive damage after brain and spinal cord

injury: pathomechanisms and treatment strategies

Helen M. Bramlett and W. Dalton Dietrich

Department of Neurological Surgery, Neurotrauma Research Center, The Miami Project to Cure Paralysis, University of
Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA

Abstract: The pathophysiology of brain and spinal cord injury (SCI) is complex and involves multiple
injury mechanisms that are spatially and temporally specific. It is now appreciated that many of these injury
mechanisms remain active days to weeks after a primary insult. Long-term survival studies in clinically
relevant experimental studies have documented the structural changes that continue at the level of the insult
as well as in remote brain structures. After traumatic brain injury (TBI), progressive atrophy of both gray
and white matter structures continues up to 1 year post-trauma. Progressive changes may therefore underlie
some of the long-term functional deficits observed in this patient population. After SCI, similar features of
progressive injury are observed including delayed cell death of neurons and oligodendrocytes, axonal
demyelination of intact fiber tracts and retrograde tract degeneration. SCI also leads to supraspinal changes
in cell survival and remote brain circuitry. The progressive changes in multiple structures after brain and
SCI are important because of their potential consequences on chronic or developing neurological deficits
associated with these insults. In addition, the better understanding of these injury cascades may one day
allow new treatments to be developed that can inhibit these responses to injury and hopefully promote
recovery. This chapter summarizes some of the recent data regarding progressive damage after CNS trauma
and mechanisms underlying these changes.

Keywords: traumatic brain injury; spinal cord injury; progressive damage; pathophysiology; treatment

Introduction important from the prospective of clarifying mech-

Recent studies from both experimental and clinical
investigations have emphasized the progressive
nature of central nervous system (CNS) injury.
In contrast to the initial concept that the majority
of damage occurs at the time of the primary is-
chemic or traumatic insult, new evidence empha-
sizes that acute injury can initiate a variety of
pathophysiological cascades that lead to second-
ary injury mechanisms associated with subacute as
well as progressive injury. These findings are
Corresponding author. Tel.: +1-305-243-8926; Fax: +1-305-
243-3914; E-mail: hbramlett@miami.edu
anisms underlying cell death, but more impor-
tantly provide new targets for therapeutic
intervention. Indeed, the observation that proc-
esses which potentially affect long-term outcome
may be active days or even months after injury
provides new targets to improve outcome after
CNS injury. This new way of assessing and treat-
ing acute injury is also important as we think
about how acute insults such as ischemic or trau-
matic injuries may enhance the vulnerability of the
aging brain to later occurring neurodegenerative
diseases. The main objective of this chapter is to
summarize recent date emphasizing the progres-
sive nature of lesion pathology after brain and

DOI: 10.1016/S0079-6123(06)61009-1 125

126
spinal cord injury (SCI) and highlight potential
therapeutic strategies that may be relevant to these
devastating insults.
Traumatic brain injury
Traumatic brain injury (TBI) is a leading cause of
morbidity and mortality in the United States
(Langlois et al., 2000). In 2000, half of a million
new cases of moderate and severe TBI were re-
ported. Traumatic insults commonly occur in
young adults as a consequence of traffic and
sporting accidents. More recently, there has been
an increase in the number of traumatic brain in-
sults in the elderly due to falls and other traumatic
insults. In developing countries, a significant rise
in vehicular accidents has also been observed.
Thus, the magnitude of the problem internation-
ally merits increased concern regarding the pre-
vention as well as treatment of TBI.
The pathophysiology of TBI is complicated
and involves both primary and secondary insults
(Graham et al., 2000a; Bramlett and Dietrich,
2004). Primary insults due to impact injury result
in the rupture of membranes that lead to acute
damage of neuronal, glial and vascular compo-
nents. These membrane disturbances also cause
metabolic stress that initiates a cascade of cellular
and molecular mechanisms that can initiate both
reparative as well as destructive processes. Acute
consequences of TBI include alterations in the
blood-brain barrier (BBB) as well as complex
changes in local cerebral blood flow (LCBF) and
metabolism (Cortez et al., 1989; Hovda et al.,
1991; Dietrich et al., 1994). Trauma-induced glial
swelling as a consequence of both vasogenic and
cytotoxic edema is also a common early conse-
quence to TBI (Bullock et al., 1991; Kimbelberg
and Norenberg, 1994). Early metabolic and hemo-
dynamic events on their own can lead to necrotic
cell death in specific vulnerable populations of
cells. Severe reductions in LCBF are observed in
some TBI patients and may indicate ischemic
events (Bramlett and Dietrich, 2004). In human
TBI, specific gray and white matter structures in-
cluding the corpus callosum and hippocampus are
frequently damaged after moderate and severe
TBI (Povlishock and Christman, 1995; Graham
et al., 2000b; Leclereq et al., 2001). Diffuse axonal
injury underlies a major mechanism for the mor-
bidity and mortality associated with TBI (Adams
et al., 1989, 1991).
Various animal models have been developed
that mimic some of the structural and behavioral
consequences of human TBI (Gennarelli et al.,
1982; Cernak, 2005). Although no one model of
TBI exactly represents the human condition of
brain trauma, these models are important in the
understanding of critical pathomechanisms re-
sponsible for traumatic events and the eventual
testing of novel therapeutic interventions. One
model, fluid-percussion (F-P) brain injury has been
used commonly in these preclinical TBI studies
(Dixon et al., 1987). For example, after moderate
parasagital F-P brain injury, the extravasation of
the protein tracer horseradish peroxidase (HRP) is
observed in specific brain regions associated
with acute vascular damage due to shearing forces
produced by the insult (Dietrich et al., 1994). In
other studies, enduring changes in axolemmal per-
meability has been reported in white matter tracts,
not necessarily associated with overt damage
(Povlishock and Pettus, 1996). Thus, abnormal
permeability of a variety of cellular membranes are
rendered leaky to tracers after TBI (Pettus and
Povlishock, 1996).
More recently, evidence for apoptotic cell death
has been demonstrated in models of TBI as well as
SCI (Crowe et al., 1997; Beattie et al., 2002; Lu
et al., 2003). McIntosh and colleagues (Rink et al.,
1995) first provided ultrastructural evidence indi-
cating that neurons may die by apoptotic mecha-
nisms after F-P brain injury. In that study,
neurons demonstrated the classical appearance of
apoptotic bodies within the cell nucleus. More re-
cently, molecular and biochemical approaches
have been used to shed light on the various recep-
tor and intracellular cell signaling mechanisms re-
sponsible for apoptotic cell death (Keane et al.,
2001a, b). The importance of this line of research
to the present discussion is that emerging evidence
supports the concept that apoptotic pathways may
be activated for long periods after CNS injury
and that mechanisms similar to programmed cell
death may be responsible for some of the

progressive changes seen in the brain late after
trauma (Williams et al., 2001).
In addition to apoptosis, inflammatory and im-
mune processes also appear to play an important
role in pathophysiology of TBI (Shohami et al.,
1994; Schwab et al., 2001; Morganti-Kosmann
et al., 2002). Following brain injury, indicators of
the activation of pro-inflammatory processes in-
cluding increased expression of pro-inflammatory
cytokines such as IL-1b, TNFa and IL-6 have
been described (Holmin et al., 1997; Morganti-
Kossman et al., 1997; Kinoshita et al., 2002). Also,
specific receptors that interact with these cytokines
are expressed after injury and lead to the stimu-
lation of inflammatory signaling cascades respon-
sible for the expression of inflammatory genes
(Lotocki et al., 2004). Thus, experimental studies
that have targeted inflammatory processes by ap-
plying antibodies or blockers to inflammation
have in some cases lead to improved outcome in
the subacute post-traumatic setting. For example,
in several studies, treatment with the interleukin-1
receptor antagonist has been shown to lead to
improved histopathological and behavioral out-
come after TBI and cerebral ischemia (Toulmond
and Rothwell, 1995; Rothwell and Luheshi, 2000).
In reference to the present discussion, recent
evidence for inflammatory processes remaining
activated up to 1 year after experimental TBI
(Nonaka et al., 1999) and even longer after clinical
TBI (Gentleman et al., 2004) emphasizes the po-
tential role of inflammation in progressive damage
after trauma.
Clinical evidence for progressive injury
In patients following TBI, magnetic resonance im-
aging (MRI) approaches have identified evidence
for progressive atrophy of specific brain regions at
chronic periods after trauma (Cullum and Bigler,
1986; Anderson and Bigler, 1995; van der Naalt
et al., 1999). In these clinical studies, evidence for
the enlargement of ventricle structures and atro-
phy of specific gray and white matter structures
has been demonstrated. In one study, enlargement
of the lateral ventricle was identified as a relatively
late occurrence after TBI. This delayed response
127
occurred without any evidence of a systemic seco-
ndary insult and therefore had no obvious under-
lying mechanism.
In experimental studies of TBI, the majority of
investigations have evaluated traumatic outcome
in the acute or subacute post-traumatic period.
Thus, until recently, little information was avail-
able to determine the more chronic consequences
of experimental TBI. However, in 1997, the
chronic histopathological consequences of moder-
ate F-P brain injury were first assessed at 2 months
after trauma. Bramlett et al. (1997a) reported a
significant enlargement of the lateral ventricle and
with associated atrophy of cerebral cortical areas
within the traumatized hemisphere. This study was
the first to provide histopathological evidence for
structural changes continuing to occur weeks to
months after TBI. In subsequent studies, other in-
vestigators using similar or different injury models
have complemented and extended these initial
findings by assessing even longer periods of post-
traumatic injury (Smith et al., 1997; Pierce et al.,
1998; Dixon et al., 1999; Bramlett and Dietrich,
2002). For example, Smith et al. (1997) reported
progressive injury in various forebrain areas 1 year
after moderate lateral F-P injury. In a model of
controlled cortical impact (CCI) injury, Dixon
et al. (1999) also reported progressive damage and
long lasting behavioral deficits in that trauma
model of focal damage. Thus, in various labora-
tories using different injury models, evidence of
progressive damage has been demonstrated. Most
recently, Bramlett and Dietrich (2002) assessed
alterations in white matter tracts 1 year after mod-
erate F-P brain injury (Fig. 1). In that study,
significant atrophy of various white matter struc-
tures within the traumatized hemisphere provided
direct evidence of progressive white matter pa-
thology in widespread forebrain circuits. These
newer data emphasize the important fact that
white matter as well as gray matter structures may
be highly vulnerable to progressive damage after
relatively moderate degrees of trauma. Thus, ther-
apies may have to be developed that target both
gray and white matter pathology after TBI
(Medana and Esiri, 2003). The importance of
white matter vulnerability will again be empha-
sized in the SCI discussion.
128

Fig. 1. Double-stained H&E and Luxol-fast blue sections 1 year after TBI or sham procedure. (A) TBI animal showing gross atrophy
with marked expansion of the ipsilateral lateral ventricle. (B) Sham operated animal appearing unremarkable. (C) Higher magni-
fication of external capsule thinning (arrows) after TBI. Reprinted from Bramlett and Dietrich (2002) with kind permission of Springer
Science and Business Media.

Pathomechanisms underlying progressive injury
As previously discussed, both apoptotic and in-
flammatory cascades are felt to underlie some of
the acute and subacute pathophysiological mech-
anisms that are responsible for cellular dysfunction
and death. Recent data also indicate that these
mechanisms may be active weeks and months after
trauma and also participate in the progressive na-
ture of TBI. For example, prolonged apoptotic
cell death has been demonstrated in several CNS
injury models (Emery et al., 1998; Beattie et al.,
2002). In these studies neuronal, microglial and
oligodendrocyte apoptosis has been reported days
to weeks after injury. Delayed neuronal apoptosis
may lead to the continued removal of axon pro-
jections that could ultimately lead to patterns of
deafferentation syndromes in brain regions remote
from the primary site of damage. A good example
of this type of progressive and remote damage is
the delayed thalamic pathology observed after
parasagittal or lateral F-P brain injury (Bramlett
et al., 1997a, b). Apoptotic death of oligodendro-
cytes could also have devastating effects on both
structure and function of an axon and/or circuit.
Oligodendrocyte death would lead to demyelina-
tion of projecting axons that would be expected
to result in axonal dysfunction and possible
progressive damage (Waxman, 1989). Indeed,
electrophysiological studies have demonstrated
the adverse effects of demyelination on action
potentials and axonal survival (Waxman et al.,
1994). Thus, progressive apoptotic cell death can
have a variety of adverse consequences on the
structure and function of both gray and white
matter structures.
In the area of inflammation, recent data has
also emphasized the progressive nature of that re-
sponse to TBI. Several studies have provided evi-
dence for long-term inflammatory responses to
TBI. In brain sections stained for macrophage
activation for example, evidence for inflammatory
events is sometimes also seen many weeks after a
traumatic brain insult. Gentleman et al. (2004)
reported microglial activation, weeks after TBI.
Similarly, inflammatory cells have been observed
chronically in an animal model of TBI
(Rodriguez-Paez et al., 2005). In this study,
 129

Fig. 2. Light level micrographs of toluidine blue stained thick plastic sections of control (A) and traumatized tissue, specifically the
lateral posterior thalamic nuclei, (100
) at 3 days (B), 15 days (C), 3 months (D), 9 months (E) and 12 months (F) after TBI. (A)
Control tissue shows myelinated figures oriented perpendicular and parallel to the plain of the section. Normal appearing neuronal cell
bodies (N), astrocytes (black arrow) and blood vessels (V) are also apparent. (B) At 3 days, axonal abnormalities including changes in
axoplasmic density, unraveling of the myelin sheath (black pointer) and irregular swollen myelinated profiles (arrowhead) were
observed. There is also a proliferation of microglial cells (open arrow) associated with neuronal (N) and astrocytic (black arrow)
swelling. Normal appearing oligodendrocytes (double arrows) are observed. (C) At 15 days, there is an apparent increase in overall
tissue vacuolation, which appears to be associated with axonal (arrowhead) and non-axonal (black arrow) profiles. An increase in
microglial cells (open arrows) as well as unraveled myelin figures (black pointer) are observed. (D) At 3 months, parenchymal
vacuolation is still apparent with increased numbers of microglial cells (open arrows), swollen axonal figures (arrowhead) and irregular
myelin profiles (black pointer). (E) At 9 months, a dramatic number of vacuolated profiles are observed related to inflammatory cells
scattered among tissue debris. An amorphous crystallized material (black star) was observed inside the vacuolated profiles. There
appears to be a progressive increase in the size of the vacuolated profiles as the post-injury time increases. (F) At 12 months, the
vacuolization (large black arrows) continues to progress in association with a further decrease in numbers of myelinated axons without
the presence of the amorphous material described above. Reprinted from Rodriguez-Paez et al. (2005) with kind permission of Springer
Science and Business Media.
130

macrophage/microglia infiltration and swollen ax-
ons were observed as late as 6 months using both
light and electron microscopic analysis within
several vulnerable structures (Fig. 2). Addition-
ally, a temporal decline in the number of myelin-
ated axons within the cerebral cortex (Fig. 3) and
thalamus were reported which may be due to the
prolonged inflammatory response observed in this
study. Nonaka et al. (1999) also provided evidence
for inflammatory processes being active up to 1
year after TBI. In that study, NF-kappaB was
seen in both cortical and sub-cortical brain regions
undergoing progressive atrophy. Thus, the poten-
tial for macrophage/microglia released toxic
substances leading to tissue damage is a real pos-
sibility even weeks after TBI. To support this as-
sumption, recent biomarker studies that have
measured levels of pro-inflammatory mediators
in the cerebral spinal fluid (CSF) and plasma of
TBI patients have reported elevated levels, days
after injury (Rancan et al., 2004). In the future,
surrogate biochemical markers of tissue damage
may be developed to predict and treat progressive
injury mechanisms.
Abnormal protein aggregation has been impli-
cated in the pathogenesis of a number of neuro-
logical diseases including Alzheimer’s and
Parkinson’s disease (Chaudhuri and Paul, 2006).
Recent evidence indicates that abnormal protein
aggregation also occurs in models of cerebral is-
chemia and TBI (Blumbergs et al., 1994; Graham
et al., 1995; Lewen et al., 1995; Bramlett et al.,
1997b; Hamberger et al., 2003). In models of TBI,
immunocytochemical localization of beta APP
and other proteins has been identified in gray as
well as white matter tracts (Sherriff et al., 1994;
Bramlett et al., 1997b). Whether evidence for pro-
longed periods of protein aggregation can be cor-
related with progressive tissue damage merits
continued study.

Fig. 3. Estimation of myelinated axons in the cerebral cortex of animals following TBI or Sham surgery at 3 days, 15 days, 1 month,
3 months, 6 months, 9 months and 12 months. A significant difference was found for both group (po0.001) and time (po0.001)
between TBI and Sham animals. TBI results in a decrease in the number of myelinated axons compared to Sham animals at all time
points analyzed. In addition, there was a temporal decrease in the number of myelinated axons in the Sham animals as well. The
decrease in the axonal numbers of the Sham animals may be due to a normal aging process. Reprinted from Rodriguez-Paez et al.
(2005) with kind permission of Springer Science and Business Media.

Spinal cord injury
Each year in the United States, approximately
11,000 new spinal cord injuries are recorded. Cur-
rently there are over 250,000 individuals living
with chronic SCI and its devastating consequences
in the United States. Similar to what has been de-
scribed with TBI, evidence for lesion progression
has also been demonstrated in both experimental
and clinical conditions of SCI (Wallace et al.,
1987; Bunge et al., 1993; Crowe et al., 1997; Bruce
et al., 2000; Hill et al., 2001; Guest et al., 2005;
Totoiu and Keirstead, 2005). Because SCI fre-
quently occurs in young people, it is equally im-
portant that we understand the mechanisms
underlying chronic progressive damage and de-
velop therapeutic interventions to retard cell
death, axonal degeneration and demyelination in
SCI patients. The pathophysiology of acute SCI is
multifactorial and like TBI, includes both primary
and secondary injury mechanisms (Nashmi and
Fehlings, 2001; Keane et al., 2006). Primary injury
mechanisms include acute SC compression, im-
paction, laceration, shear damage and missile in-
jury (Norenberg et al., 2004). These acute injury
mechanisms initiate a complex cascade of second-
ary injury mechanisms that may remain activated
for months or years after injury. These initial
traumatic events lead to vascular damage and
hemorrhage, alterations in spinal cord blood flow,
vascular thrombosis, vasospasm and loss of auto-
regulation. As a consequence to cellular membrane
damage, metabolic abnormalities and electrolytic
shifts in ions occur (LoPachin and Lehning, 1997;
Li et al., 2000). In addition, neurotransmitters are
released into the extracellular space leading to the
abnormal activation of various receptors and in-
tracellular signaling processes (Keane et al., 2006).
Other classical injury cascades that are activated
after SCI include free radical formation, lipid per-
oxidation and edema formation. All of these acute
processes lead to the acute destruction of gray and
white matter structures (Blight, 1985; Schwab and
Bartholdi, 1996; Rosenberg and Wrathall, 1997).
As in brain injury, inflammatory cascades are
also activated after SCI (Blight, 1992; Crowe et al.,
1997; Popovich et al., 2002; Keane et al., 2006). In
131
addition, calpain activation has also been empha-
sized as an acute injury mechanism (Ray et al.,
2003). Many of these injury cascades have been
targets for therapeutic interventions (Blight and
Zimber, 2001).
More recently, evidence for apoptotic cell death
has been reported in a number of SCI models
(Springer et al., 1999; Ozawa et al., 2002; Keane
et al., 2006). A complex integrated apoptotic path-
way involving both extrinsic and intrinsic apopto-
tic cascades has been reported. Also, the study of
pro- and anti-apoptotic molecules, which can con-
trol apoptotic cell death, is an exciting area of
current investigation. Recent work has concen-
trated on mechanisms that allow for the commu-
nication between the external environment with
intracellular processes involved in cell survival and
death. After trauma for example, specific death
receptors have been shown to accumulate within
specific areas of the plasma membrane called lipid
rafts that contain high concentrations of choles-
terol and sphingolipid (Lotocki et al., 2004). Thus,
after SCI and brain trauma, TNF1 receptors ac-
cumulate in lipid rafts and assist in the formation
of cytoplasmic platforms that allow scaffolding
proteins to accumulate; leading to the increased
interactions of proteins that lead to the activation
of intracellular cascades associated with cell sur-
vival and death (Keane et al., 2006). Thus, re-
search is currently being undertaken to understand
this relatively acute injury response to neuronal
vulnerability and survival. Because a large amount
of cell death occurs in the acute post-traumatic
period, evidence for these pathomechanisms has
been commonly reported during these periods.
However, since cell death during delayed
post-traumatic periods is spread out and more
difficult to identify, other approaches including
immunocytochemistry must be used to investigate
later occurring injury mechanisms.
In addition, experimental and clinical investiga-
tions have also demonstrated that apoptotic cell
death, demyelination, remyelination and axonal
degeneration may occur weeks to months after
experimental and clinical SCI (Bunge et al., 1961;
Blight, 1993; Quencer and Bunge, 1996; Schwab
and Bartholdi, 1996; Crowe et al., 1997; Emery
132
et al., 1998; Bruce et al., 2000; Hill et al., 2001;
Guest et al., 2005; Totoiu and Keirstead, 2005).
For example, in a study by Crowe et al. (1997),
apoptotic cells were identified from 6 h to 3 weeks
after experimental SCI. Apoptotic cell death was
specifically shown to be present in the white matter
tracts where apoptotic cells were shown to be pos-
itive for cellular markers of oligodendrocytes. This
observation is important because it indicates
oligodendrocyte cell death with resulting demyeli-
nation could be an active mechanism in the pro-
gressive injury cascades associated with human
SCI. Following this particular observation, vari-
ous laboratories have also reported apoptotic cell
death in the spinal cord including neurons,
oligodendrocytes and inflammatory cells (Emery
et al., 1998; Beattie et al., 2002). In addition to
clarifying mechanisms of injury, these results are
also important because they provide new potential
targets for therapeutic interventions directed to-
ward the acute as well as chronic post-traumatic
period (Springer et al., 1999; Ozawa et al., 2002;
Demjen et al., 2004).
Evidence for apoptotic cell death and Schwa-
nnosis has also been reported in specimens ob-
tained from patients that survived long periods
after SCI. In a study by Emery et al. (1998), apo-
ptotic cells were identified using the apoptotic
marker caspase-3. In that study, oligodendrocytes
stained positively for this indicator of apoptotic
injury in specific white matter tracts. Again, this
clinical observation is important because of the
role of the oligodendrocyte in axonal myelination.
We know from various electrophysiological stud-
ies that if an axon is demyelinated by a toxic or
traumatic insult, that axon’s ability to propagate
axon potentials is severely affected (Blight and
Young, 1989; Waxman, 1989). Thus, strategies
that would target degenerative mechanisms or ax-
onal conduction blockage could potentially im-
prove outcome in SCI subjects (Blight and Young,
1989). Hill et al. (2001) have reported on extensive
dieback of the corticospinal tract chronically after
SCI (Fig. 4). However, there was a regenerative
response of this tract along with the reticulospinal
tract evidenced by an extension of collaterals into
the lesion matrix. Recently, Totoiu and Keirstead
(2005) have assessed chronic progressive
demyelination in an animal model of SCI. Data
from that study indicated that chronic progressive
demyelination after thoracic injury in rats does
occur for days after injury (Fig. 5). Interesting, a
process of secondary demyelination was reported
around 120–450 days post injury. These studies
underscore the importance of targeting demyeli-
nation in the development of therapeutic interven-
tions and again emphasize the progressive nature
of neurodegeneration after SCI.
In human tissue, evidence for chronic demyeli-
nation has also been reported (Bunge et al., 1993;
Norenberg et al., 2004; Guest et al., 2005). In the
study by Guest et al. (2005), evidence for axonal
demyelination, even a decade after human trau-
matic SCI was presented. Although the response
was very heterogeneous among the SCI specimens
evaluated, the potential for demyelination in spe-
cific spinal cord circuits in which axon profiles re-
mained intact was observed. In this regard,
evidence for spontaneous remyelination of intact
CNS axons by invading Schwann cells has also
been observed (Guest et al., 2005; Totoiu and
Keirstead, 2005). This observation is important
because other studies have reported similar cellular
responses to injury that indicate some endogenous
reparative mechanisms including remyelination
are activated after SCI (Hagg and Oudega, 2006).
In a clinical study by Bruce et al. (2000), evi-
dence for Schwannosis was seen in 32 out of 65
cases that survived after 24 years. The incidence of
Schwannosis rose to 82% in SCI patients who
survived more than 4 months. Associated with
Schwannosis was intense chondroitin sulfate pro-
teoglycan (CSPG) staining. This observation is
important because the CSPGs are thought to be
inhibitory molecules that reduce axonal regenera-
tion after CNS injury (Snow et al., 1990). Thus the
continued clarification of both the positive and
negative consequences of Schwann cell responses
to acute and chronic SCI remains an important
area of investigation.
The use of human spinal cords donated for re-
search is also allowing other injury mechanisms to
be evaluated in human tissues (Norenberg et al.,
2004). At The Miami Project to Cure Paralysis,
approximately 115 human spinal cords have been
collected from individuals surviving from 24 h to
 133

Fig. 4. Appearance of impact site after a 12.5 mm contusion injury 1 dpi (A–C), 3 dpi (A, D, E), 8 dpi (A, F, G), 21 dpi (A, H, I) and 14
weeks (A, J, K, L) after injury. (A) The progression of cavitation. At 1 day, the impact site is hemorrhagic but no cavitation is present,
and RBCs extend up to an additional 5 mm rostral (arrow); between 3, 8 and 21 days the injury site becomes progressively more defined,
and it appears as a dark region in low magnification at 8 dpi and is clearly defined at 21 days. At 21 days and 14 weeks a cellular mass is
present within the cavity attached to the spared rim of white matter at several points via trabeculae (arrow). One dpi (B) RBCs and (C)
damaged axons at the impact site. Three dpi (D) damaged axons (small arrow) intermixed with blood cell infiltrate (large arrow) at the
impact site and (E) some macrophages (large arrowhead) have infiltrated the rim of spared white matter (small arrows). Eight dpi (F)
macrophages are densely packed at the center of the cavity while (G) open spaces (small arrow) are present between groups of
macrophages and other nonfluorescently labeled cells (large arrowhead) that appear in scattered bands within the cavity. Twenty-one
dpi (H) trabeculae, thin tissue bridges with regions of no tissue beside them, are beginning to form (small arrow), and macrophages
(large arrow head) are still present within the cavity in association with the trabeculae or as (I) macrophage rafts. Fourteen weeks after
injury (J–L) autofluorescing macrophages (large arrowhead) are present within (J) fibrous trabeculae and (K) cellular trabeculae as well
as (L) the cellular infiltrate rostral to the cavity. Bar in (A), 1 mml; all other bars, 100 mm. (C, E, F), same magnification. (B, D, G–L),
same magnification. dpi ¼ days post injury. Reprinted from Hill et al. (2001) with permission from Elsevier.
134

Fig. 5. Quantification of demyleinated and remyelinated axons at different time points post injury. The numbers of axons at the six
cranial and six caudal levels examined for all animals within a group were averaged to generate each point on the graph. The number of
demyleinated axons was highest at 1 day post injury and decreased substantially by 7 days post injury. Thereafter, demyelination was a
chronic and progressive phenomenon. Both oligodendrocyte and Schwann cell remyelination was present at all subsequent time points
post injury. Reprinted from Totoiu and Keirstead (2005) with permission from John Wiley & Sons, Inc.

24 years after injury. Corresponding MRI and
immunocytochemical techniques are used to anal-
yze the injured tissue (Bunge et al., 1993; Becerra
et al., 1995; Emery et al., 1998; Guest et al., 2005).
With this approach, investigators have assessed
the acute as well as more chronic immunohisto-
pathological consequences of traumatic human
SCI. For example, in early specimens taken 3 days
after injury H&E staining shows damage to gray
and white matter areas with a well identified cen-
tral hemorrhagic lesion (Bunge et al., 1993). If
specimens are stained with various immunocyto-
chemical approaches, evidence of gliosis in the
form of increased glial fibrillary acidic protein
(GFAP) staining is seen surrounding the contusive
site. Also, the accumulation of invading inflam-
matory cells such as polymorphonuclear leuko-
cytes (PMNL) and macrophages can be visualized
(Fleming et al., 2006). These types of studies con-
ducted on human tissues are important because
they verify some of the acute and more chronic
histopathological changes that are reported in
preclinical animal studies. This information is
currently of importance to the development of
new treatments targeting inflammatory processes,
which is an important secondary injury mechanism
after SCI (Gris et al., 2004).
Evidence for retrograde degeneration of cortical
spinal tracts (CST) has also been shown in human
specimens. Following damage to thoracic cortical
spinal tracts for example, evidence for retrograde
degeneration can be seen in remote spinal tracts
(Bunge et al., 1993; Norenberg et al., 2004). In
addition to retrograde degeneration, evidence for
Wallerian degeneration can also be obtained using
repetitive MR imaging procedures in SCI patients
(Quencer and Bunge, 1996). These studies empha-
size the widespread changes that occur as a con-
sequence to local SCI. With the development and
improvement of new imaging techniques, acute
damage as well as the progressive nature of human
SCI will be more easily investigated. In acute in-
jury settings, better evidence for local cord com-
pression will provide important information to
help guide surgical approaches to target ischemic
events. In chronic postoperative SCI cases, both
cord compression and the formation and progres-
sion of cavities can be visualized in patients expe-
riencing neurological symptoms including chronic
pain. With time, some lesions expand to form cysts

that can have a dramatic effect on progression of
neurological symptoms. These imaging techniques
will continue to be improved and allow specialized
treatments to be developed for an individual pa-
tient undergoing progressive injuries.
Supraspinal alterations after SCI
In addition to changes occurring at the level of the
injured spinal cord, it is also clear that SCI leads to
alterations in supraspinal areas of the neuroaxis
(Jain et al., 1997; Raineteau et al., 2001; Hains
et al., 2003; Hubscher and Johnson, 2006; Kim
et al., 2006). In this regard, there is a rich literature
on experimental SCI lesioning studies in rodents
and non-human primates showing evidence for
degeneration of neurons in the cerebral cortex af-
ter injury. Some data indicate that cell death oc-
curs through apoptotic mechanisms whereas more
recently, additional evidence indicates that cortical
cell bodies as a consequence of SCI may only un-
dergo severe atrophy but not actually die. In these
cases, the local addition of neurotrophic factors to
a particular brain region reverses some of these
morphological changes.
In addition to cellular changes, evidence for cir-
cuit reorganization and plasticity resulting in de-
activation and reactivation of certain brain areas
has been demonstrated after clinical and experi-
mental SCI (Jain et al., 1997; Raineteau et al.,
2001). Chronic SCI has also been shown to induce
changes in the response of thalamic neurons to
physiological activation (Hubscher and Johnson,
2006). Thus, SCI leads to alterations in brain cir-
cuits that are responsible for assessing normal and
abnormal sensation. In this regard, some research-
ers believe that this circuit plasticity in addition to
other events including chronic inflammation may
account for the late development of chronic ne-
urogenic pain in patients with SCI. A clearer un-
derstanding of local circuit changes in response to
SCI may provide important information regarding
how to treat these patients with neurogenic pain.
As a consequence of chronic SCI, the cerebral
cortex may also undergo plastic changes in terms
of shifts in cortical map architecture (Topka et al.,
1991; Jain et al., 1997; Bruehlmeier et al., 1998;
135
McDonald et al., 2002; Sabbah et al., 2002; Lotze
et al., 2006). As specific inputs to the cerebral cor-
tex are removed after SCI, other cortical areas may
take over function. Thus, electrophysiological and
metabolic studies have shown evidence for cortical
plasticity after SCI (Hoffman and Field-Fote,
2007; Kim et al., 2006). These observations not
only emphasize the potential for plasticity occur-
ring in remote brain regions after SCI, but also
may be important as we think about repairing the
nervous system in the chronically injured spinal
cord. If brain circuits and/or cortical maps were
significantly altered due to chronic SCI, would
such a consequence affect the ability to repair the
nervous system and return function? Only after
successful regenerative approaches are developed
in the chronically injured state can these questions
be answered.
Therapeutic interventions targeting progressive
injury
Based on the complexity of brain and SCI, it is
clear that the injury mechanisms are multifactorial
and may require a combinational therapeutic ap-
proach. In the area of brain and SCI, various
neuroprotective agents have been evaluated
(Narayan et al., 2002). Neuroprotective agents
such as methylprednisolone, GM1 gangliosides,
lazeroids, calcium and sodium channel blockers,
growth factors as well as blockers of excitototoxic
process have been reported to be effective in some
animal models. More recently, the use of anti-in-
flammatory strategies, calpain antagonists, anti-
apoptotic strategies as well as agents targeting
cAMP have also been investigated with various
results (Blight and Zimber, 2001).
Mild hypothermia that targets multiple injury
cascades has been tested with various degrees of
success after both SCI and TBI (Hayashi et al.,
2004; Guest and Dietrich, 2005). Although hypo-
thermia was used in the 60 s to target both brain
and SCI, profound hypothermia (271C) commonly
produced severe effects on cardiac function and
increased infection rates in patients. However, in
the mid-80 s, the importance of mild to moderate
hypothermic treatment was demonstrated in
136
ischemic and traumatic animal models. Busto et al.
(1987) first reported that a reduction of just 2 or 3
degrees in the temperature of the brain provided
dramatic protection of CA1 neurons within the
post-ischemic hippocampus. Subsequent studies
demonstrated that mild hypothermia was also
protective when initiated at various times after
cerebral ischemia or TBI. Recently, mild hypo-
thermia has been shown to have a dramatic effect
on SCI. In a study by Yu et al. (2000), the bene-
ficial effects of systemic hypothermia on locomo-
tor outcome and histopathological damage were
reported after contusion SCI in rats. In that study,
mild hypothermia (331C) initiated 30 min after SCI
significantly improved open locomotor function
and significantly reduced contusion volume weeks
after injury. Because of these exciting preclinical
findings, multiple clinical trials have been initiated
in patients with various neurological insults in-
cluding cardiac arrest, stroke, TBI and SCI. Re-
cent multicenter trials in cardiac arrest patients
have shown a dramatic benefit with mild hypo-
thermia in that patient population (Bernard et al.,
2002; Hozler, 2002). Also, individual and multi-
center trials have reported that mild hypothermia
shows promise in improving outcome in severe
TBI patients (Jiang et al., 2006). Clinical investi-
gations are currently determining the effects of
mild hypothermia involving SCI (Guest and
Dietrich, 2005). In this regard, mild hypothermia
may be used during elective surgeries or in the
early post-injury periods.
The beneficial effects of mild hypothermia in-
volve multiple injury pathomechanisms. Various
studies have shown that modifying brain or SC
temperatures significantly affects injury-induced
excitatory neurotransmitter levels, free fatty acid
formation, BBB breakdown and the formation of
edema. Mild hypothermia after SCI has been
shown to significantly reduce the acute inflamma-
tory response to injury (Chatzipanteli et al., 2000).
Because the early inflammatory response to injury
is thought to represent an important secondary in-
jury mechanism, the ability of mild hypothermia to
limit the accumulation of inflammatory cells after
SCI represents an important therapeutic target for
treatment. Free radical formation can come from
multiple sources including prostaglandin and nitric
oxide (NO) synthesis. The inducible form of nitric
oxide synthase (iNOS) produces free radicals that
can be destructive to tissue survival. Post-traumatic
hypothermia following TBI also reduces iNOS ac-
tivity and may improve outcome by affecting free
radical generation as well (Chatzipanteli et al.,
1999).
Mild hypothermia has been shown to be neuro-
protective and promote functional outcome after
TBI (Hayashi et al., 2004). Bramlett et al. (1995)
first reported that post-traumatic hypothermia im-
proved cognitive function in rats. These studies are
important because cognitive deficits are some of
the most severe consequences of mild, moderate
and severe TBI. In regards to potential treatments
for progressive injury, post-traumatic hypothermia
has also been shown to reduce the progressive na-
ture of F-P brain injury in rats (Bramlett et al.,
1997a). In that study, a period of 3 h of moderate
hypothermia significantly protected against pro-
gressive damage and enlargement of the lateral
ventricle at 2 months after TBI. Thus, in addition
to the acute benefits of therapeutic hypothermia,
this therapy may also limit the amount of pro-
gressive damage after trauma. Additional studies
will be required to determine what duration of the
hypothermic therapy is most beneficial in provid-
ing long-term benefits in terms of functional
outcome.
Finally, recent data indicate that different types
of cellular transplantation strategies can promote
repair and improve recovery in animal models of
brain and SCI (Schouten et al., 2004; Schwab
et al., 2006; Thuret et al., 2006). In some cases, cell
transplantation strategies have been initiated to
replace dysfunctional or dead neurons injured by
the insult. Thus, an exciting field of research is
currently directed toward stem cell biology and the
potential for neural stem cells to repopulate the
injured nervous system and improve function. Cell
survival, migration, control of cellular differenti-
ation and the integration of new cells into existing
circuits are some of the challenges currently being
addressed with these approaches. Alternatively,
the transplantation of specific populations of
restricted progenitor cells that differentiate into
myelin-forming cells is another important research
direction (Cao et al., 2005; Keirstead et al., 2005;

Karimi-Abdolrezaee et al., 2006). In recent studies,
grafted cells have been shown to form morpho-
logically normal-appearing myelin sheaths around
damaged axons. Importantly, experimental treat-
ments targeting neuronal or glia damage after
CNS injury leads to improvements in functional
recovery. Because progressive damage after brain
and SCI can affect a variety of different cell types,
these cellular therapies may be helpful in both ne-
uroprotective as well as reparative strategies tar-
geting the long-term consequences of CNS injury.
Continued research into the progressive nature of
brain and SCI should provide new targets for
treatment and better outcome in patients that sus-
tain these devastating injuries.
Acknowledgments
This work was supported in part by NIH grants
NS30291, NS38665 and DAMD17-02-1-0190. The
authors also thank the members of the Bramlett/
Dietrich laboratory for their important contribu-
tions to this research field.
References
Adams, J.H., Doyle, D., Ford, I., Gennarelli, T.A., Graham,
D.I. and McLellan, D.R. (1989) Diffuse axonal injury in head
injury: definition, diagnosis and grading. Histopathology, 15:
49–59.
Adams, J.H., Graham, D.I., Gennarelli, T.A. and Maxwell,
W.L. (1991) Diffuse axonal injury in non-missile head injury.
J. Neurol. Neurosurg. Psychiatry, 54: 481–483.
Anderson, C.V. and Bigler, E.D. (1995) Ventricular dilation,
cortical atrophy, and neuropsychological outcome following
traumatic brain injury. J. Neuropsychiatry Clin. Neurosci., 7:
42–48.
Beattie, M.S., Hermann, G.E., Rogers, R.C. and Bresnahan,
J.C. (2002) Cell death in models of spinal cord injury. In:
McKerracher L., Doucet G. and Rossignol S. (Eds.),
Progress in Brain Research, Vol. 137. Elsevier, Amsterdam,
pp. 37–47.
Becerra, J.L., Puckett, W.R., Hiester, E.D., Quencer, R.M.,
Marcillo, A.E., Post, M.J. and Bunge, R.P. (1995) MR-
pathologic comparisons of wallerian degeneration in spinal
cord injury. AJNR Am. J. Neuroradiol., 16: 125–133.
Bernard, S.A., Gray, T.W., Buist, M.D., Jones, B.M., Silvester,
W., Gutteridge, G. and Smith, K. (2002) Treatment of com-
atose survivors of out-of-hospital cardiac arrest with induced
hypothermia. N. Engl. J. Med., 346: 557–563.
137
Blight, A.R. (1985) Delayed demyelination and macrophage
invasion: a candidate for secondary cell damage in spinal
cord injury. Cent. Nerv. Syst. Trauma, 2: 299–315.
Blight, A.R. (1992) Macrophages and inflammatory damage in
spinal cord injury. J. Neurotrauma, 9(Suppl 1): S83–S91.
Blight, A.R. (1993) Remyelination, revascularization, and re-
covery of function in experimental spinal cord injury. Adv.
Neurol., 59: 91–104.
Blight, A.R. and Young, W. (1989) Central axons in injured cat
spinal cord recover electrophysiological function following
remyelination by Schwann cells. J. Neurol. Sci., 91: 15–34.
Blight, A.R. and Zimber, M.P. (2001) Acute spinal cord injury:
pharmacotherapy and drug development perspectives. Curr.
Opin. Investig. Drugs, 2: 801–808.
Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H., Simp-
son, D.A. and McLean, A.J. (1994) Staining of amyloid pre-
cursor protein to study axonal damage in mild head injury.
Lancet, 344: 1055–1056.
Bramlett, H.M. and Dietrich, W.D. (2002) Quantitative struc-
tural changes in white and gray matter 1 year following
traumatic brain injury in rats. Acta Neuropathol. (Berl.), 103:
607–614.
Bramlett, H.M. and Dietrich, W.D. (2004) Pathophysiology of
cerebral ischemia and brain trauma: similarities and differ-
ences. J. Cereb. Blood Flow Metab., 24: 133–150.
Bramlett, H.M., Dietrich, W.D., Green, E.J. and Busto, R.
(1997a) Chronic histopathological consequences of fluid-
percussion brain injury in rats: effects of post-traumatic
hypothermia. Acta Neuropathol. (Berl.), 93: 190–199.
Bramlett, H.M., Green, E.J., Dietrich, W.D., Busto, R., Globus,
M.Y-T. and Ginsberg, M.D. (1995) Post-traumatic brain
hypothermia provides protection from sensorimotor and
cognitive behavioral deficits. J. Neurotrauma, 12: 289–298.
Bramlett, H.M., Kraydieh, S., Green, E.J. and Dietrich, W.D.
(1997b) Temporal and regional patterns of axonal damage
following traumatic brain injury: a beta-amyloid precursor
protein immunocytochemical study in rats. J. Neuropathol.
Exp. Neurol., 56: 1132–1141.
Bruce, J.H., Norenberg, M.D., Kraydieh, S., Puckett, W.,
Marcillo, A. and Dietrich, W.D. (2000) Schwannosis: role of
gliosis and proteoglycan in human spinal cord injury. J.
Neurotrauma, 17: 781–788.
Bruehlmeier, M., Dietz, V., Leenders, K.L., Roelcke, U., Miss-
imer, J. and Curt, A. (1998) How does the human brain deal
with a spinal cord injury? Eur. J. Neurosci., 10: 3918–3922.
Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and
Adams, J.H. (1991) Glial swelling following human cerebral
contusion: an ultrastructural study. J. Neurol. Neurosurg.
Psychiatry, 54: 427–434.
Bunge, M.B., Bunge, R.P. and Ris, H. (1961) Ultrastructural
study of remyelination in an experimental lesion in adult cat
spinal cord. J. Biophys. Biochem. Cytol., 10: 67–94.
Bunge, R.P., Puckett, W.R., Becerra, J.L., Marcillo, A. and
Quencer, R.M. (1993) Observations on the pathology of hu-
man spinal cord injury. A review and classification of 22 new
cases with details from a case of chronic cord compression
with extensive focal demyelination. Adv. Neurol., 59: 75–89.
138
Busto, R., Dietrich, W.D., Globus, M.Y.-T., Valdes, I.,
Scheinberg, P. and Ginsberg, M.D. (1987) Small differences
in intraischemic brain temperature critically determine the
extent of ischemic neuronal injury. J. Cereb. Blood Flow
Metab., 7: 729–738.
Cao, Q., Xu, X.M., DeVries, W.H., Enzmann, G.U., Ping, P.,
Tsoulfas, P., Wood, P.M., Bunge, M.B. and Whittemore,
S.R. (2005) Functional recovery in traumatic spinal cord in-
jury after transplantation of multineurotrophin-expressing
glial-restricted precursor cells. J. Neurosci., 25: 6947–6957.
Cernak, I. (2005) Animal models of head trauma. NeuroRx., 2:
410–422.
Chatzipanteli, K., Wada, K., Busto, R. and Dietrich, W.D.
(1999) Effects of moderate hypothermia on constitutive and
inducible nitric oxide synthase activities after traumatic brain
injury in the rat. J. Neurochem., 72: 2047–2052.
Chatzipanteli, K., Yanagawa, Y., Marcillo, A.E., Kraydieh, S.,
Yezierski, R.P. and Dietrich, W.D. (2000) Posttraumatic
hypothermia reduces polymorphonuclear leukocyte accumu-
lation following spinal cord injury in rats. J. Neurotrauma,
17: 321–332.
Chaudhuri, T.K. and Paul, S. (2006) Protein-misfolding dis-
eases and chaperone-based therapeutic approaches. FEBS J.,
273: 1331–1349.
Cortez, S.C., McIntosh, T.K. and Noble, L.J. (1989) Experi-
mental fluid percussion brain injury: vascular disruption and
neuronal and glial alternations. Brain Res., 482: 271–282.
Crowe, M.J., Bresnahan, J.C., Shuman, S.L., Masters, J.N. and
Beattie, M.S. (1997) Apoptosis and delayed degeneration af-
ter spinal cord injury in rats and monkeys. Nat. Med., 3:
73–76.
Cullum, C.M. and Bigler, E.D. (1986) Ventricle size, cortical
atrophy and the relationship with neuropsychological status
in closed head injury: a quantitative analysis. J. Clin. Exp.
Neuropsychol., 8: 437–452.
Demjen, D., Klussmann, S., Kleber, S., Zulani, C., Stieltjes, B.,
Metzger, C., Hirt, U.A., Walczak, H., Falk, W., Essig, M.,
Edler, L., Krammer, P.H. and Martin-Villalba, A. (2004) Neu-
tralization of CD95 ligand promotes regeneration and func-
tional recovery after spinal cord injury. Nat. Med., 10: 389–395.
Dietrich, W.D., Alonso, O. and Halley, M. (1994) Early mi-
crovascular and neuronal consequences of traumatic brain
injury: a light and electron microscopic study in rats. J.
Neurotrauma, 11: 289–301.
Dixon, C.E., Kochanek, P.M., Yan, H.Q., Schiding, J.K.,
Griffith, R.G., Baum, E., Marion, D.W. and DeKosky, S.T.
(1999) One-year study of spatial memory performance, brain
morphology, and cholinergic markers after moderate con-
trolled cortical impact in rats. J. Neurotrauma, 16: 109–122.
Dixon, C.E., Lyeth, B.G., Povlioshock, J.T., Findling, R.L.,
Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L.
(1987) A fluid percussion injury model of experimental brain
injury in the rat. J. Neurosurg., 67: 110–119.
Emery, E., Aldana, P., Bunge, M.B., Puckett, W., Srinivasan,
A., Keane, R.W., Bethea, J. and Levi, A.D. (1998) Apoptosis
after traumatic human spinal cord injury. J. Neurosurg., 89:
911–920.
Fleming, J.C., Norenberg, M.D., Ramsay, D.A., Dekaban,
G.A., Marcillo, A.E., Saenz, A.D., Pasquale-Styles, M.,
Dietrich, W.D. and Weaver, L.C. (2006) The cellular inflam-
matory response in human spinal cords after injury. Brain,
129: 3259–3269.
Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I.,
Thompson, C.J. and Marcincin, R.P. (1982) Diffuse axonal
injury and traumatic coma in the primate. Ann. Neurol., 12:
564–574.
Gentleman, S.M., Leclercq, P.D., Moyes, L., Graham, D.I.,
Smith, C., Griffin, W.S. and Nicoll, J.A. (2004) Long-term
intracerebral inflammatory response after traumatic brain
injury. Forensic Sci. Int., 146: 97–104.
Graham, D.I., Gentleman, S.M., Lynch, A. and Roberts, G.W.
(1995) Distribution of beta-amyloid protein in the brain fol-
lowing severe head injury. Neuropathol. Appl. Neurobiol.,
21: 27–34.
Graham, D.I., McIntosh, T.K., Maxwell, W.L. and Nicoll, J.A.
(2000a) Recent advances in neurotrauma. J. Neuropathol.
Exp. Neurol., 59: 641–651.
Graham, D.I., Raghupathi, R., Saatman, K.E., Meaney, D.
and McIntosh, T.K. (2000b) Tissue tears in the white matter
after lateral fluid percussion brain injury in the rat: relevance
to human brain injury. Acta Neuropathol. (Berl.), 99:
117–124.
Gris, D., March, D.R., Oatway, M.A., Chen, Y., Hamilton,
E.F., Dekaban, G.A. and Weaver, L.C. (2004) Transient
blockade of the CD11d/CD18 integrin reduces secondary
damage after spinal cord injury, improving sensory, auto-
nomic, and motor function. J. Neurosci., 24: 4043–4051.
Guest, J.D. and Dietrich, W.D. (2005) Spinal cord ischemia and
trauma. In: Tisherman S.A. and Sterz F. (Eds.), Therapeutic
Hypothermia. Springer, New York, pp. 101–118.
Guest, J.D., Hiester, E.D. and Bunge, R.P. (2005) Demyelina-
tion and Schwann cell responses adjacent to injury epicenter
cavities following chronic human spinal cord injury. Exp.
Neurol., 192: 384–393.
Hagg, T. and Oudega, M. (2006) Degenerative and spontane-
ous regenerative processes after spinal cord injury. J. Neuro-
trauma, 23: 264–280.
Hains, B.C., Willis, W.D. and Hulsebosch, C.E. (2003) Tem-
poral plasticity of dorsal horn somatosensory neurons after
acute and chronic spinal cord hemisection in rat. Brain Res.,
970: 238–241.
Hamberger, A., Huang, Y.L., Zhu, H., Bao, F., Ding, M.,
Blennow, K., Olsson, A., Hansson, H.A., Viano, D. and
Haglid, K.G. (2003) Redistribution of neurofilaments and
accumulation of beta-amyloid protein after brain injury by
rotational acceleration of the head. J. Neurotrauma, 20:
169–178.
Hayashi, N., Bullock, R., Dietrich, W.D. and Maekawa, T.
(2004) Hypothermia for Acute Brain Damage. Springer, New
York.
Hill, C.E., Beattie, M.S. and Bresnahan, J.C. (2001) Degener-
ation and sprouting of identified descending supraspinal ax-
ons after contusive spinal cord injury in the rat. Exp. Neurol.,
171: 153–169.

Hoffman, L.R. and Field-Fote, E.C. (2007) Cortical reorgan-
ization following bimanual training and somatosensory
stimulation in cervical spinal cord injury: a case report.
Phys. Ther., 87: 208–223.
Holmin, S., Schalling, M., Hojeberg, B., Nordqvist, A.C.,
Skeftruna, A.K. and Mathiesen, T. (1997) Delayed cytokine
expression in rat brain following experimental contusion. J.
Neurosurg., 86: 493–504.
Hovda, D.A., Yoshino, A., Kawamata, T., Katayama, Y. and
Becker, D.P. (1991) Diffuse prolonged depression of cerebral
oxidative metabolism following concussive brain injury in the
rat: a cytochrome oxidase histochemistry study. Brain Res.,
567: 1–10.
Hozler M., for the Hypothermia after Cardiac Arrest Study
Group. (2002) Mild therapeutic hypothermia to improve the
neurological outcome after cardiac arrest. N. Engl. J. Med.,
346: 549–556.
Hubscher, C.H. and Johnson, R.D. (2006) Chronic spinal cord
injury induced changes in the responses of thalamic neurons.
Exp. Neurol., 197: 177–188.
Jain, N., Catania, K.C. and Kaas, J.H. (1997) Deactivation and
reactivation of somatosensory cortex after dorsal spinal cord
injury. Nature, 386: 495–498.
Jiang, J.Y., Xu, W., Li, W.P., Gao, G.Y., Bao, Y.H., Liang,
Y.M. and Luo, Q.Z. (2006) Effect of long-term mild hypo-
thermia or short-term mild hypothermia on outcome of pa-
tients with severe traumatic brain injury. J. Cerebral Blood
Flow Metab., 26: 771–776.
Karimi-Abdolrezaee, S., Eftekharpour, E., Wang, J., Mors-
head, C.M. and Fehlings, M.G. (2006) Delayed transplanta-
tion of adult neural precursor cells promotes remyelination
and functional neurological recovery after spinal cord injury.
J. Neurosci., 26: 3377–3389.
Keane, R.W., Davis, A.R. and Dietrich, W.D. (2006) Inflam-
matory and apoptotic signaling after spinal cord injury.
J. Neurotrauma, 23: 335–344.
Keane, R.W., Kraydieh, S., Lotocki, G., Alonso, O.F., Aldana,
P. and Dietrich, W.D. (2001a) Apoptotic and antiapoptotic
mechanisms after traumatic brain injury. J. Cereb. Blood
Flow Metab., 21: 1189–1198.
Keane, R.W., Kraydieh, S., Lotocki, G., Bethea, J.R., Krajew-
ski, S., Reed, J.C. and Dietrich, W.D. (2001b) Apoptotic and
anti-apoptotic mechanisms following spinal cord injury.
J. Neuropathol. Exp. Neurol., 60: 422–429.
Keirstead, H.S., Nistor, G., Bernal, G., Totoiu, M.O., Cloutier,
F., Sharp, K. and Steward, O. (2005) Human embryonic stem
cell-derived oligodendrocyte progenitor cell transplants
remyelinate and restore locomotion after spinal cord injury.
J. Neurosci., 25: 4694–4705.
Kim, B.G., Dai, H.N., McAtee, M., Vicini, S. and Bregman,
B.S. (2006) Remodeling of synaptic structures in the motor
cortex following spinal cord injury. Exp Neurol., 198:
401–415.
Kimbelberg, H.K. and Norenberg, M.D. (1994) Astroglial re-
sponse to CNS trauma. In: Salzman S.K. and Faden A.I.
(Eds.), The Neurobiology of Central Nervous System
Trauma. Oxford University Press, New York, pp. 193–208.
139
Kinoshita, K., Chatzipanteli, K., Vitarbo, E., Truettner, J.S.,
Alonso, O.F. and Dietrich, W.D. (2002) Interleukin-1beta
messenger ribonucleic acid and protein levels after fluid-
percussion brain injury in rats: importance of injury severity
and brain temperature. Neurosurgery, 51: 195–203.
Langlois, J.A., Rutland-Brown, W. and Thomas, K.E. (2000)
Traumatic Brain Injury in the United States: Emergency De-
partment Visits, Hospitalizations, and Deaths. Centers for
Disease Control and Prevention, National Center for Injury
Prevention and Control, Atlanta, GA.
Leclercq, P.D., McKenzie, J.E., Graham, D.I. and Gentleman,
S.M. (2001) Axonal injury is accentuated in the caudal corpus
callosum of head-injured patients. J. Neurotrauma, 18: 1–9.
Lewen, A., Li, G.L., Nilsson, P., Olsson, Y. and Hillered, L.
(1995) Traumatic brain injury in rat produces changes of
beta-amyloid precursor protein immunoreactivity. Neurore-
port, 6: 357–360.
Li, S., Jiang, Q. and Stys, P.K. (2000) Important role of reverse
Na(+)-Ca(2+) exchange in spinal cord white matter injury
at physiological temperature. J. Neurophysiol., 84:
1116–1119.
LoPachin, R.M. and Lehning, E.J. (1997) Mechanism of cal-
cium entry during axon injury and degeneration. Toxicol.
Appl. Pharmacol., 143: 233–244.
Lotocki, G., Alonso, O.F., Dietrich, W.D. and Keane, R.W.
(2004) TNF receptor 1 and its signaling intermediates are
recruited to lipid rafts in the traumatized brain. J. Neurosci.,
24: 11010–11016.
Lotze, M., Laubis-Herrmann, U. and Topka, H. (2006) Com-
bination of TMS and fMRI reveals a specific pattern of re-
organization in M1 in patients after complete spinal cord
injury. Restor. Neurol. Neurosci., 24: 97–107.
Lu, J., Moochhala, S., Shirhan, M., Ng, K.C., Tan, M.H., Teo,
A.L. and Ling, E.A. (2003) Nitric oxide induces macrophage
apoptosis following traumatic brain injury in rats. Neurosci.
Lett., 339: 147–150.
McDonald, J.W., Becker, D., Sadowsky, C.L., Jane, Sr., J.A.,
Conturo, T.E., Schultz, L.M. (2002) Late recovery following
spinal cord injury. Case report and review of the literature.
J. Neurosurg., 97(Suppl 2): 252–265
Medana, I.M. and Esiri, M.M. (2003) Axonal damage: a key
predictor of outcome in human CNS diseases. Brain, 126:
515–530.
Morganti-Kossman, M.C., Lenzlinger, P.M., Hans, V., Stahel,
P., Csuka, E., Ammann, E., Stocker, R., Trentz, O. and
Kossmann, T. (1997) Production of cytokines following brain
injury: beneficial and deleterious for the damaged tissue.
Mol. Psychiatry, 2: 133–136.
Morganti-Kossmann, M.C., Rancan, M., Stahel, P.F. and
Kossmann, T. (2002) Inflammatory response in acute trau-
matic brain injury: a double-edged sword. Curr. Opin. Crit.
Care, 8: 101–105.
van der Naalt, J., Hew, J.M., van Zomeren, A.H., Sluiter, W.J.
and Minderhoud, J.M. (1999) Computed tomography and
magnetic resonance imaging in mild to moderate head injury:
early and late imaging related to outcome. Ann. Neurol., 46:
70–78.
140
Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A.,
Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C.,
Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D.,
Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D.,
Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller,
N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W.,
Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E.,
Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P.,
Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G.,
Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich,
M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewics, L.
(2002) Clinical trials in head injury. J. Neurotrauma, 19:
503–557.
Nashmi, R. and Fehlings, M.G. (2001) Mechanisms of axonal
dysfunction after spinal cord injury: with an emphasis on the
role of voltage-gated potassium channels. Brain Res. Brain
Res. Rev., 38: 65–191.
Nonaka, M., Chen, X.H., Pierce, J.E., Leoni, M.J., McIntosh,
T.K., Wolf, J.A. and Smith, D.H. (1999) Prolonged activa-
tion of NF-kappaB following traumatic brain injury in rats.
J. Neurotrauma, 16: 1023–1034.
Norenberg, M.D., Smith, J. and Marcillo, A. (2004) The pa-
thology of human spinal cord injury: defining the problems.
J. Neurotrauma, 21: 429–440.
Ozawa, H., Keane, R.W., Marcillo, A.E., Diaz, P.H. and
Dietrich, W.D. (2002) Therapeutic strategies targeting
caspase inhibition following spinal cord injury in rats. Exp.
Neurol., 177: 306–313.
Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a
distinct set of intra-axonal ultrastructural changes associated
with traumatically induced alteration in axolemmal perme-
ability. Brain Res., 722: 1–11.
Pierce, J.E., Smith, D.H., Trojanowski, J.Q. and McIntosh,
T.K. (1998) Enduring cognitive, neurobehavioral and histo-
pathological changes persist for up to one year following
severe experimental brain injury in rats. Neuroscience, 87:
359–369.
Popovich, P.G., Guan, Z., McGaughy, V., Fisher, L., Hickey,
W.F. and Basso, D.M. (2002) The neuropathological and
behavioral consequences of intraspinal microglial/macroph-
age activation. J. Neuropathol. Exp. Neurol., 61: 623–633.
Povlishock, J.T. and Christman, C.W. (1995) The pathobiology
of traumatically induced axonal injury in animals and hu-
mans: a review of current thoughts. J. Neurotrauma, 12:
555–564.
Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced
axonal damage: evidence for enduring changes in axolemmal
permeability with associated cytoskeletal change. Acta
Neurochir. Suppl., 66: 81–86.
Quencer, R.M. and Bunge, R.P. (1996) The injured spinal cord:
imaging, histopathologic clinical correlates, and basic science
approaches to enhancing neural function after spinal cord
injury. Spine, 21: 2064–2066.
Raineteau, O. and Schwab, M.E. (2001) Plasticity of motor
systems after incomplete spinal cord injury. Nat. Rev.
Neurosci., 2: 263–273.
Raineteau, O., Fouad, K., Noth, P., Thallmair, M. and
Schwab, M.E. (2001) Functional switch between motor
tracts in the presence of the mAb IN-1 in the adult rat.
Proc. Natl. Acad. Sci. USA, 98: 6929–6934.
Rancan, M., Bye, N., Otto, V.I., Trentz, O., Kossmann, T.,
Frentzel, S. and Morganti-Kossmann, M.C. (2004) The
chemokine fractalkine in patients with severe traumatic brain
injury and a mouse model of closed head injury. J. Cereb.
Blood Flow Metab., 24: 1110–1118.
Ray, S.K., Hogan, E.L. and Banik, N.L. (2003) Calpain in the
pathophysiology of spinal cord injury: neuroprotection with
calpain inhibitors. Brain Res. Rev., 42: 169–185.
Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M.,
Neugebauer, E. and McIntosh, T.K. (1995) Evidence of
apoptotis cell death after experimental traumatic brain
injury. Am. J. Pathol., 147: 1575–1583.
Rodriguez-Paez, A.C., Brunschwig, J.P. and Bramlett, H.M.
(2005) Light and electron microscopic assessment of pro-
gressive atrophy following moderate traumatic brain injury
in the rat. Acta Neuropathol., 109: 603–616.
Rosenberg, L.J. and Wrathall, J.R. (1997) Quantitative analysis
of acute axonal pathology in experimental spinal cord con-
tusion. J. Neurotrauma, 14: 823–838.
Rothwell, N.J. and Luheshi, G.N. (2000) Interleukin 1 in the
brain: biology, pathology and therapeutic target. TINS, 23:
618–625.
Sabbah, P., De, S.S., Leveque, C., Gay, S., Pfefer, F., Nioche,
C., Sarrazin, J.L., Barouti, H., Tadie, M. and Cordoliani,
Y.S. (2002) Sensorimotor cortical activity in patients with
complete spinal cord injury: a functional magnetic resonance
imaging study. J. Neurotrauma, 19: 53–60.
Schouten, J.W., Fulp, C.T., Royo, N.C., Saatman, K.E.,
Watson, D.J., Snyder, E.Y., Trojanowski, J.Q., Prockop,
D.J., Maas, A.I. and McIntosh, T.K. (2004) A review and
rationale for the use of cellular transplantation as a thera-
peutic strategy for traumatic brain injury. J. Neurotrauma,
21: 1501–1538.
Schwab, M.E. and Bartholdi, D. (1996) Degeneration and
regeneration of axons in the lesioned spinal cord. Physiol.
Rev., 76: 319–370.
Schwab, J.M., Brechtel, K., Mueller, C.A., Failli, V., Kaps,
H.P., Tuli, S.K. and Schluesener, H.J. (2006) Experimental
strategies to promote spinal cord regeneration — an integra-
tive perspective. Prog. Neurobiol., 78: 91–116.
Schwab, J.M., Seid, K. and Schluesener, H.J. (2001) Traumatic
brain injury induces prolonged accumulation of cyclooxy-
genase-1 expressing microglia/brain macrophages in rats. J.
Neurotrauma, 18: 881–890.
Sherriff, F.E., Bridges, L.R. and Sivaloganathan, S. (1994)
Early detection of axonal injury after human head trauma
using immunocytochemistry for beta-amyloid precursor pro-
tein. Acta Neuropathol. (Berl.), 87: 55–62.
Shohami, E., Novikov, M., Bass, R., Yamin, A. and Gallily, R.
(1994) Closed head injury triggers early production of TNF
alpha and IL-6 by brain tissue. J. Cereb. Blood Flow Metab.,
14: 615–619.

Smith, D.H., Chen, X.H., Pierce, J.E., Wolf, J.A., Trojanowski,
J.Q., Graham, D.I. and McIntosh, T.K. (1997) Progressive
atrophy and neuron death for one year following brain
trauma in the rat. J. Neurotrauma, 14: 715–727.
Snow, D.M., Lemmon, V., Carrino, D.A., Caplan, A.I. and
Silver, J. (1990) Sulfated proteoglycans in astroglial barriers
inhibit neurite outgrowth in vitro. Exp. Neurol., 109: 111–130.
Springer, J.E., Azbill, R.D. and Knapp, P.E. (1999) Activation
of the caspase-3 apoptotic cascade in traumatic spinal cord
injury. Nat. Med., 5: 943–946.
Thuret, S., Moon, L.D. and Gage, F.H. (2006) Therapeutic
interventions after spinal cord injury. Nat. Rev. Neurosci., 7:
628–643.
Topka, H., Cohen, L.G., Cole, R.A. and Hallet, M. (1991)
Reorganization of corticospinal pathways following spinal
cord injury. Neurology, 41: 1276–1283.
Totoiu, M.O. and Keirstead, H.S. (2005) Spinal cord injury is
accompanied by chronic progressive demyelination. J. Comp.
Neurol., 486: 373–383.
Toulmond, S. and Rothwell, N.J. (1995) Interleukin-1 receptor
antagonist inhibits neuronal damage caused by fluid percus-
sion injury in the rat. Brain Res., 671: 261–266.
141
Wallace, M.C., Tator, C.H. and Lewis, A.J. (1987)
Chronic regenerative changes in the spinal cord after
cord compression injury in rats. Surg. Neurol., 27:
209–219.
Waxman, S.G. (1989) Demyelination in spinal cord injury. J.
Neurol. Sci., 91: 1–14.
Waxman, S.G., Utzschneider, D.A. and Kocsis, J.D. (1994)
Enhancement of action potential conduction following de-
myelination: experimental approaches to restoration of func-
tion in multiple sclerosis and spinal cord injury. Prog. Brain
Res., 100: 233–243.
Williams, S., Raghupathi, R., MacKinnon, M.A., McIntosh,
T.K., Saatman, K.E. and Graham, D.I. (2001) In situ
DNA fragmentation occurs in white matter up to 12 months
after head injury in man. Acta Neuropathol. (Berl.), 102:
581–590.
Yu, C.G., Jimenez, O., Marcillo, A.E., Weider, B., Bangerter,
K., Dietrich, W.D., Castro, S. and Yezierski, R.P. (2000)
Beneficial effects of modest systemic hypothermia on loco-
motor function and histopathological damage following con-
tusion-induced spinal cord injury in rats. J. Neurosurg.,
93(Suppl 1): 85–93.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 10

Injury-induced alterations in CNS

electrophysiology $

Akiva S. Cohen1,, Bryan J. Pfister2, Elizabeth Schwarzbach3, M. Sean Grady4,

Paulette B. Goforth5 and Leslie S. Satin5

1
Department of Pediatrics, University of Pennsylvania, School of Medicine and Division of Neurology, Children’s Hospital
of Philadelphia, Philadelphia, PA, USA
2
Department of Biomedical Engineering, New Jersey Institute of Technology, Newark, NJ, USA
3
Department of Pharmacology, University of Pennsylvania, School of Medicine, Philadelphia, PA, USA
4
Department of Neurosurgery, University of Pennsylvania, School of Medicine, Philadelphia, PA, USA
5
Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Commonwealth University, Richmond,
VA, USA

Abstract: Mild to moderate cases of traumatic brain injury (TBI) are very common, but are not always
associated with the overt pathophysiogical changes seen following severe trauma. While neuronal death has
been considered to be a major factor, the pervasive memory, cognitive and motor function deficits suffered
by many mild TBI patients do not always correlate with cell loss. Therefore, we assert that functional
impairment may result from alterations in surviving neurons. Current research has begun to explore CNS
synaptic circuits after traumatic injury. Here we review significant findings made using in vivo and in vitro
models of TBI that provide mechanistic insight into injury-induced alterations in synaptic electrophysio-
logy. In the hippocampus, research now suggests that TBI regionally alters the delicate balance between
excitatory and inhibitory neurotransmission in surviving neurons, disrupting the normal functioning of
synaptic circuits. In another approach, a simplified model of neuronal stretch injury in vitro, has been used
to directly explore how injury impacts the physiology and cell biology of neurons in the absence of
alterations in blood flow, blood brain barrier integrity, or oxygenation associated with in vivo models of
brain injury. This chapter discusses how these two models alter excitatory and inhibitory synaptic trans-
mission at the receptor, cellular and circuit levels and how these alterations contribute to cognitive
impairment and a reduction in seizure threshold associated with human concussive brain injury.

Keywords: TBI; electrophysiology; hippocampus; cortical neurons; excitation; inhibition; stretch injury;
synapse

Introduction upon healthcare systems worldwide. Many who

survive the initial traumatic incident endure conti-
Traumatic brain injury (TBI) is a major public nuing problems with memory, cognition and
health problem, which has had a significant impact motor function that are debilitating to their daily
$ lives. While TBI has been extensively studied from
Supported by NIH-NINDS RO1 NS 45975 (ASC), Common-
wealth Neurotrauma Initiative Fund CNI 04-094 (LSS). both a clinical and a basic science standpoint,
Corresponding author. Tel.: +1-215-590-1472; many of these studies are descriptive rather than
E-mail: Cohena@email.chop.edu mechanistic, often focusing on the behavioral

DOI: 10.1016/S0079-6123(06)61010-8 143

144
consequences and neuronal cell loss associated
with severe TBI in human patients or animal
models. However, the extensive and widespread
primary neuronal destruction seen in severe cases
of TBI is generally irreversible and thus, in all
likelihood, therapeutically intractable.
Mild and moderate TBI, on the other hand, is
much more commonly encountered than severe
cases. From the authors’ standpoint, the pervasive
deficits associated with mild injury do not neces-
sarily require the frank loss of neuronal or glial cells
in the CNS, but result from the disruption of the
synaptic function of surviving neurons, negatively
impacting the brain function of TBI patients. It is
our contention that understanding how the delicate
balance of excitation and inhibition in CNS synap-
tic circuits is disturbed after TBI will help explain
and ultimately treat disturbances of cognition and
memory in mildly and moderately injured patients.
In contrast to the large number of histological, be-
havioral or biochemical studies that have appeared
in the literature to date, there have been relatively
few studies of CNS electrophysiology after trauma.
The aim of the present review, is to emphasize and
discuss the significant progress made recently using
in vivo and in vitro models of TBI. Furthermore, to
understand the mechanisms of brain injury at the
level of synaptic electrophysiology, we highlight the
need for more mechanistically designed studies of
synaptic electrophysiology in the future.
The chapter is divided into two major sections.
The first part is concerned with how TBI results
in specific changes in neuronal function that dis-
rupt the function of synaptic circuits in the brain,
affecting the balance between excitatory and in-
hibitory neurotransmission. Here we focus on
studies of hippocampal slices from rodents sub-
jected to lateral fluid percussion injury (LFPI)
in vivo, a widely used experimental model of
TBI (Dixon et al., 1987; McIntosh et al., 1989;
Thompson et al., 2005). In the second part, we
discuss studies obtained using a simplified model
of TBI, where cortical pyramidal neurons, co-
cultured with glia, are subjected to a defined
degree of tensile stretch in vitro. Use of this model
has allowed perturbations to be delivered in the
absence of changes in blood flow, blood brain
barrier integrity or oxygenation, and facilitate
understanding how injury impacts the physiology
and cell biology of neurons and neural circuits. We
believe that comparing and contrasting results
obtained using these two very different models is
productive and leads to a deeper understanding of
the mechanisms evoked by physical trauma. In the
last section, we summarize what has been learned
to date, and suggest some directions for future
research.
In vivo models of TBI
Fluid percussion injury
LFPI is a reliable and reproducible experimental
model of concussive TBI implemented in rodents.
LFPI is widely used since it reproduces many
pathophysiological features of mild to moderate
human TBI including blood brain barrier break-
down, neuronal cell loss, gliosis, perturbation of
ionic homeostasis and long-term alterations in
cognitive and motor function (Dixon et al., 1987;
McIntosh et al., 1989; Thompson et al., 2005). In
this model, a pendulum strikes a piston sending a
brief (15 ms) percussion pulse of saline into the
extradural space of the closed cranial cavity caus-
ing a brief displacement and deformation of the
brain mimicking concussive injury. The severity of
injury is set by adjusting the height of the pendu-
lum. LFPI results in both focal as well as diffuse
brain injury consisting of a cortical contusion at
the impact site and cell death and axonal damage
in the hippocampus, thalamus and cortex (Dixon
et al., 1987; McIntosh et al., 1989; Carbonell et al.,
1998; Thompson et al., 2005). LFPI has been ex-
tensively characterized and widely used in rats as a
model, but has recently been adapted for use in
mice (Carbonell et al., 1998; Witgen et al., 2005).
Importantly, the use of mice to study TBI will al-
low investigators to exploit transgenic and knock-
out mouse technology to directly test the role of
specific genes or gene mutations on TBI, as well as
lowering the costs associated with the use of rats.
In this review chapter, we will include data
obtained using fluid percussion to injure mice to
demonstrate that these results are consistent with
the rat LFPI model.

Experimental LFPI and the hippocampus
In humans, TBI results in heterogeneous changes in
brain function, including cognitive impairments in
learning and memory. Damage to the hippocampus
associated with significant impairment of learning
and memory was first recognized in studies of bi-
lateral damage to the hippocampus in humans
(Rempel-Clower et al., 1996; Deweer et al., 2001).
Subsequently, a large body of knowledge has ac-
cumulated indicating that, while not exclusively re-
sponsible for memory function, the hippocampus
plays a central role in memory consolidation and
retrieval (Cave and Squire, 1991; Miller et al., 1998;
Bohbot et al., 2000). Moreover, the hippocampus is
frequently damaged in TBI patients, thus likely
contributing to observed cognitive deficits includ-
ing impaired learning and memory (Kotapka et al.,
1992, 1993, 1994; Bigler et al., 1997).
The dependence of memory acquisition on
the hippocampus is similar in humans and rats,
which both show selective loss of recently ac-
quired memory and the failure to imprint new
memories after hippocampal damage (He et al.,
2001; Liu and Bilkey, 2001). In rodents, LFPI
produces a consistent injury to the hippocampus
including neuronal loss within all hippocampal
subregions as well as physiologic, ionic and neuro-
chemical changes in the dentate hilus and
areas CA1 and CA3. This damage has been cor-
related to visuo-spatial memory deficits observed
in the Morris water maze assay (Smith et al.,
1991; Gorman et al., 1993) and to hippocampal-
dependent cognitive impairment demonstrated by
changes in the conditioned fear response (Hogg
et al., 1998a, b; Witgen et al., 2005). Accordingly,
LFPI in the rat and mouse is currently the most
widely accepted experimental model to mimic hip-
pocampal injury.
Functional electrophysiological changes in the
hippocampus after TBI
Normal hippocampal function is directly deter-
mined by a delicate balance between neuronal ex-
citation and inhibition. Perturbation of this state of
equilibrium can have catastrophic consequences,
145
including impaired cognitive function and the
development of seizures (Cave and Squire, 1991;
Annegers et al., 1996; Asikainen et al., 1999).
Within the hippocampus there are three intercon-
nected subregions referred to collectively as the
‘Trisynaptic Circuit’. This circuit consists of the
dentate gyrus (DG), area CA3, and area CA1, each
having distinct physiological roles (see Fig. 1). The
DG ‘‘filters’’ out aberrant or excessive input to
the hippocampus from propagating further along
the hippocampal circuit. Area CA3 ‘‘amplifies’’ the
activity that the DG has allowed to enter, while
area CA1 ‘‘transduces’’ processed hippocampal in-
formation outflow to the cerebral cortex.
It is believed that the DG acts as a filter because
it is itself extraordinarily resistant to the genera-
tion of synchronized bursting that is characteristic
of epileptic seizures (Heinemann et al., 1992;
Lothman et al., 1992). This dampening of synap-
tic input may thus be critical in protecting the
fragile neurons within the hippocampus from
excessive activation and seizure formation. The
filtering or gatekeeper behavior of the DG is due
to three main mechanisms: (1) the intrinsic mem-
brane properties of dentate granule cells (DGCs)
tend to resist excessive excitation (Fricke and
Prince, 1984), (2) at the circuit level, the dearth of
interconnections between DGCs impedes synchro-
nization since neurons firing inappropriately
cannot easily transmit this overexcitation to their
neighbors (Claiborne et al., 1986) and (3) the effi-
cacy of DG-mediated filtering is further enhanced
by robust surround inhibition of the DG (Sloviter,
1994).
In area CA3, pyramidal neurons are highly in-
terconnected to each other via recurrent axon col-
laterals (Dudek et al., 1986; Wong et al., 1986),
providing the anatomical substrate for efficacious
synchronization and signal amplification. Thus,
area CA3 is highly susceptible to recurrent excita-
tion and is consequently prone to seizure genera-
tion and injury-related pathology (i.e., if the DG
does not filter excessive input). This activity is in
turn transmitted via CA3 axons (Schaffer collat-
erals) to area CA1, which excite both pyramidal
neurons as well as inhibitory interneurons, thereby
regulating both inhibition and excitation in area
CA1.
146

Fig. 1. Illustration of the hippocampal ‘Trisynaptic circuit’. This circuit consists of three interconnected subregions: the dentate gyrus
(DG), area CA3 and area CA1. The DG is thought to filter aberrant or excessive input from the entorhinal cortex via the perforant
pathway. Filtered activity is relayed to area CA3 via the mossy fibers where it is amplified and sent onward to area CA1 via the Schaffer
collaterals where the information is then transduced out to the cerebral cortex.

The role of GABAergic inhibition in the
hippocampus
Pathological alterations in the effectiveness of
GABAA-mediated inhibition are implicated in
many neurological disease processes, including
TBI (Reeves et al., 1995; Toth et al., 1997;
Santhakumar et al., 2001). GABA is the principal
inhibitory neurotransmitter in the mammalian
brain and not only controls the degree of neuronal
activation but, importantly, is also critical in
initiating, synchronizing and terminating both
normal and pathological neuronal network acti-
vity in the central nervous system. A majority of
inhibitory synapses are localized on neuronal
somata and axon initial segments, since synaptic
input to these regions are highly efficacious in
regulating neuronal output. Synaptic localization,
coupled with the heterogeneous patterns of in-
terneuronal innervation of excitatory (principal)
neurons (Freund and Buzsaki, 1996), endows
individual GABAAergic interneurons with the

capability of synchronizing the activity of hun-
dreds to thousands of principal neurons within
the hippocampus (Cobb et al., 1995). The syn-
chronization of neocortical structures is known to
subserve important functions, such as the process-
ing of sensory information (Lytton and Sejnowski,
1991; Singer and Gray, 1995) and memory con-
solidation and formation (Joliot et al., 1994).
Therefore, oscillations which are synchronized by
GABAAergic inhibition can be viewed as being
critical for normal hippocampal function.
GABAA receptors are pentameric structures
composed of several related subunit families. At
present, eight different GABAA receptor subunit
families have been cloned (with many families
having multiple members), including a (1–6), b
(1–4), g (1–3), d, r(1–3), e, p and y (Barnard et al.,
1998). The subunit stoichiometry is considered to
be a primary determinant of the pharmacology of
the receptor. Although subunit composition is
modulable under different developmental and
pathological conditions, to date no injury-induced
alterations in GABAA receptor subunit composi-
tion have been reported.
Surprisingly, there have been few studies ex-
ploring functional changes in hippocampal cir-
cuitry following TBI. Instead, most studies to date
have attempted to correlate regional hippocampal
cell loss with behavioral changes induced by TBI,
despite the gap between these two divergent end-
points. While hippocampal cell loss is often hypo-
thesized to alter the balance between neuronal
excitation and inhibition which is necessary for
proper hippocampal function, iterations in func-
tion do not always correlate with the degree of
observed cell death (Lyeth et al., 1990). Recent
studies have supported the alternative hypothesis
that alterations in the behavior of surviving neu-
rons may be a key. Specifically, a few experimental
studies using LFPI indicate that TBI affects exci-
tatory and inhibitory synaptic transmission, giving
rise to dysfunctional hippocampal circuits in ro-
dents despite the absence of significant cell death
(Reeves et al., 1995; Toth et al., 1997; D’Ambrosio
et al., 1998; Witgen et al., 2005). However, the
precise cellular mechanism(s) responsible for the
TBI-induced pathological alterations in hippocam-
pal function currently remain unknown.
147
Injury-induced functional alterations in area CA1
Area CA1: injury-induced alterations in
hippocampal area CA1 excitability
To evaluate hippocampal function after injury,
several groups have specifically investigated
changes in area CA1 excitability. Electrophysio-
logical studies on hippocampal slices from injured
rodents using LFPI generally show a decrease in
excitability in hippocampal area CA1. D’Ambrosio
et al. (1998) initially found, 2 days post-LFPI, that
increased simulation intensity is required to over-
come the threshold of postsynaptic population
spikes (PS) as well as field extracellular postsynap-
tic potentials (fEPSP). In addition, the potentials
exhibit smaller amplitudes and decreased linear
slopes after injury. These results suggest a decrease
in area CA1 excitability, which was further corro-
borated and expanded upon using mice 1 week
after LFPI (Witgen et al., 2005). Input/output
(I/O) curves, relating the stimulus intensity as a
function of the linear slope of the fEPSP are shifted
downward after injury compared with curves
obtained from naı̈ve/sham mice, Fig. 2 (Witgen
et al., 2005). In addition, the threshold for evoking
PS in slices from injured animals is correspondingly
increased. The observed shifts in the I/O curve and
the PS threshold denote a decrease in excitability,
and are indicative of diminished synaptic commu-
nication between pre and postsynaptic neurons.
While these studies in hippocampal slices estab-
lish reduced excitability in area CA1 after LFPI,
earlier in vivo studies reported increased excitabil-
ity during the acute period of 2 h–2 days following
injury. Stimulation thresholds for the generation
of PS in vivo decreases within 2–3 h after LFPI
(Miyazaki et al., 1992), and the decrease persists
for 2 days, returning to control levels by 7 days
after LFPI (Reeves et al., 1995, 1997b). Although
lowered thresholds suggest increased excitability,
the amplitudes of the PS and fEPSPs are reduced
compared with controls. The discrepant nature of
these results could be explained by differences in
injury severity in the respective studies, the use of
in vivo vs. in vitro slice preparations, differences in
the recording techniques used, or the difference in
post-injury time points chosen. An additional
148
Fig. 2. Extracellular input/output curves for area CA1 evoked by afferent Schaffer collateral stimulation. LFPI in mice results in
smaller fEPSP slopes compared with those from sham animals. Reprinted from Neuroscience, Witgen et al. (2005), with permission
from Elsevier.
report recently suggested increased excitability of
area CA1 at 1 week after LFPI (Akasu et al., 2002).
Area CA1: Glutamate evoked currents
Alterations in excitatory synaptic transmission are
one potential mechanism underlying the injury-
induced shifts in area CA1 excitability seen
following TBI. Supporting the hypothesis that
excitatory neurotransmission is decreased after
injury, a variety of immunocytochemical studies
of the hippocampus indicate that the number of
N-methyl-D-aspartate (NMDA) receptors decreases
within 12 –24 h post-injury, with no associated
changes in non-NMDA receptors such as AMPA
or kainate receptors (Miller et al., 1990; Sihver
et al., 2001; Osteen et al., 2004).
The NMDA receptor is important for the induc-
tion of excitatory synaptic plasticity (see below).
Therefore, to examine whether injury causes
NMDA receptor dysfunction, extracellular record-
ing techniques were used to measure evoked
NMDA-mediated depolarizing potentials (AMPA-
and GABAergic activity were pharmacologically
blocked) (Cohen and Abraham, 1996) in brain slices
from sham vs. LFPI mice, Fig. 3 (Schwarzbach
et al., 2006). These data show that in mild to mod-
erately injured mice, the amplitude of NMDA
potentials (Figs. 3B, C) is significantly smaller than
in slices from sham animals (Figs. 3A, C). To
evaluate NMDA receptor alterations in individual
hippocampal neurons, whole-cell patch voltage-
clamp recordings were performed in visually identi-
fied pyramidal neurons in hippocampal slices
(Schwarzbach et al., 2006). Focal application of
glutamate results in significantly smaller NMDA
isolated excitatory current (sodium, AMPA and
GABA currents blocked) in slices from LFPI
(Fig. 3E) vs. sham (Fig. 3D) mice. Furthermore,
by subtracting the NMDA-mediated currents from
the total glutamate-evoked currents, the AMPA-
ergic component could be determined. This
demonstrates there is a significant reduction in
AMPA as well as NMDA currents in LFPI slices
compared with sham controls (Fig. 3G).
 149

FPI SHAM

D E

BMI BMI
TTX APV TTX APV
CNQX CNQX
glycine glycine

40 80
F G

20 40
* *

0 0
FPI SHAM FPI SHAM

Fig. 3. Injury diminishes isolated NMDA receptor potentials and EPSCs following injury. (A) Sham; (B) FPI, trace of fEPSP, isolated
NMDA potential, recording after addition of APV to ensure the waveform is NMDA dependent. Scale bar: 0.5 mV, 10 ms.
(C) Histogram showing the quantification of the reduction in the peak amplitude of the NMDA potential recorded in slices from FPI
mice compared with sham animals (n ¼ 11 and 7 for FPI and sham, respectively, po0.05, denoted by *). (D) Sham; (E) FPI,
glutamatergic currents were recorded by focal application of glutamate (100 mM, 50 ms, 40 psi), the NMDA component was isolated
with the addition of CNQX to the bath, and at the end of each experiment, APV was perfused to ensure the current is NMDA-
dependent. Scale bar: 15 pA, 100 ms. (F) The histogram shows FPI causes a significant reduction in the peak amplitude of the NMDA
current. (G) The histogram is created by subtracting the isolated NMDA-mediated current from the entire glutamate evoked current to
determine the AMPA-mediated current. This shows FPI also causes a significant reduction in the peak amplitude of the AMPA
currents (* denotes significance, po0.05).

Area CA1: changes in inhibitory synaptic
transmission
While we have presented evidence showing
decreases in excitatory synaptic transmission after
injury, alterations in overall synaptic excitation may
also reflect changes in inhibitory circuits as these
effects may be additive postsynaptically. Specifi-
cally, interneurons innervate CA1 pyramidal
neurons to provide GABAA-mediated inhibition
control over the initiation, synchronization and
termination of excitatory neuronal activity. There-
fore, augmentation of inhibitory activity could
also contribute to decreased area CA1 excitability
(Fig. 2). Furthermore, because synaptic plasticity,
especially long-term potentiation (LTP), requires a
precise functional balance of inhibition and excita-
tion, alterations in either excitation or inhibition
may interfere with functional information process-
ing. A role for GABAA alteration as a cause for
memory impairments is supported by the findings
that GABAA receptor immunolocalization in area
150

CA1 increases after LFPI (Reeves et al., 1997a, b)
and that the administration of GABAA receptor
antagonists reduces cognitive deficits in rats after
FPI (O’Dell and Hamm, 1995).
To test whether LFPI leads to augmented
GABAA-mediated inhibition, miniature inhibitory
postsynaptic currents (mIPSCs) were recorded in
slices from sham vs. injured mice 1 week post-
injury. Miniature IPSCs are spontaneously occur-
ring events, reflecting the spontaneous release of
single packets of GABA from individual presy-
naptic terminals of GABAergic interneurons.
Analysis of mIPSCs provides detailed information
about the properties of postsynaptic receptors
because mIPSCs result from the activation of
individual synapses, in contrast to stimulation
studies which involve the activation of tens to
hundreds of synapses (Mody et al., 1994). Specifi-
cally, changes in the number of postsynaptic recep-
tors are generally accepted to cause changes in
mIPSC amplitude, while shifts in receptor subunit
composition and/or GABA release kinetics alter
mIPSC kinetics. In addition, a change in the
total number of active synapses (i.e., presynaptic
release sites) is reflected by variations in mIPSC
frequency.
The median mIPSC amplitude in CA1 neurons
recorded from FPI animals is significantly greater
than in sham animals (Figs. 4A, B) (Witgen et al.,
2005). The significant increase in mIPSC ampli-
tude is further demonstrated by calculating total
mIPSC charge transfer (i.e., the amount of current
flowing during an individual mIPSC and calcu-
lated by integrating the total area associated with

Fig. 4. Whole-cell voltage clamp recordings from area CA1 pyramidal neurons from FPI animals demonstrate increases in spon-
taneous miniature (mIPSC) activity 1 week after LFPI. (A) Cumulative frequency amplitude histogram for neurons from sham and
LFPI animals. (B) The mIPSCs are larger in CA1 pyramidal neurons from LFPI mice. Histograms of mean median (C) mIPSC
amplitude and (D) net charge transfer demonstrate significant increases in inhibitory activity of neurons from LFPI mice compared
with control animals. (E) Histogram demonstrating that mIPSC frequency is not significantly different in sham compared with LFPI
slices. Reprinted from Neuroscience, Witgen et al. (2005), with permission from Elsevier.

an mIPSC) from the two populations of neurons.
Miniature IPSC charge transfer in neurons in
slices from FPI animals is 155% of that calculated
for the sham population (Fig. 4D). Interestingly,
neither the 50% decay time (T50), weighted
decay or the frequency of occurrence of mIPSCs
recorded in slices from injured animals are signifi-
cantly different from values obtained in slices from
sham animals (Fig. 4E) suggesting that concussive
brain injury does not alter the number of active
presynaptic inhibitory synapses or GABAA recep-
tor subunit composition. The enhanced action
potential-independent spontaneous inhibitory ac-
tivity in area CA1 following FPI would be expec-
ted to contribute to the reduced excitability in this
region. Qualitatively similar results were seen in
stretch injured cortical neurons where the amplitude
of GABAA-mediated currents are directly increased
following injury (see below; Kao et al., 2004).
Area CA1: excitatory synaptic plasticity
Long-term potentiation
Persistent and life-long cognitive impairment is a
hallmark of brain injury pathology (Barth et al.,
1983; Bennett-Levy, 1984; Lyeth et al., 1990;
McAllister, 1992; McAllister et al., 1999) and even
mild TBI can damage the fine structure of the
hippocampus (Kotapka et al., 1991; Lowenstein
et al., 1992; Grady et al., 2003), contributing to a
reduction in information processing (Mathias et al.,
2004) and learning and memory impairments. LTP,
a form of activity-dependent plasticity, is currently
our best physiological correlate of learning and
memory. LTP can be experimentally induced in
area CA1 and is characterized by a long-lasting
potentiation of fEPSP slope above baseline in re-
sponse to tetanic stimulation of the Schaffer col-
laterals. Several rodent LFPI models report a
consistent inability to induce and maintain LTP
in area CA1 in the ipsilateral hippocampus in
vivo (Miyazaki et al., 1992; Reeves et al., 1995,
1997b) and in vitro beginning at 2 h and persisting
to 8 weeks post-injury (D’Ambrosio et al., 1998;
Sick et al., 1998; Sanders et al., 2000; Schwarzbach
et al., 2006) and may contribute to learning and
151
memory deficits observed following TBI (Lyeth
et al., 1990; Smith et al., 1991; Gorman et al.,
1993).
The complement of LTP is long-term depression
(LTD), which is a decrease in net synaptic efficacy
seen after low frequency stimulation (Christie et
al., 1994; Bear and Abraham, 1996). Unlike LTP,
this form of synaptic plasticity can be generated in
slices from LFPI mice and rats (D’Ambrosio et al.,
1998; Schwarzbach et al., 2006), demonstrating
the ability of neurons to express some forms of
synaptic plasticity even after LFPI.
Mechanisms of LTP
The progression of LTP under physiological con-
ditions consists of pre to postsynaptic induction,
postsynaptic expression and subsequent long-term
expression and maintenance. Results reported in
the early literature were mixed and often discrep-
ant as to the mechanisms involved in injury-
induced loss of LTP. Sanders et al. (2000) suggests
that while there are major deficits in the main-
tenance phase, measurements at 5 min show no
impairment in potentiation and that TBI may not
affect the mechanisms essential to LTP induction
and expression. This is the only study, however,
that does not show a depression of the fEPSP
immediately after injury.
Post-tetanic potentiation (PTP), an initial com-
ponent of LTP also cannot be observed after LFPI
(Sick et al., 1998; Schwarzbach et al., 2006). Since,
PTP is believed to be due to residual presynaptic
calcium buildup resulting in increased neurotrans-
mitter release (Del Castillo and Katz, 1954; Martin
and Pilar, 1964; Barrett and Stevens, 1972; Hirst
et al., 1981), this led to the hypothesis that injury is
affecting LTP via action on presynaptic mecha-
nisms. To determine if there are alterations in pre-
synaptic transmitter release following TBI, paired
pulse studies (two identical stimulus pulse sepa-
rated by less than 100 ms) were compared in sham
vs. injured slices. Paired pulse facilitation (PPF) in
area CA1 results in a modest augmentation of
the second fEPSP from the excess calcium in
the presynaptic terminal, whereas impaired PPF
would denote alterations in release probability
152

have occurred. Reeves et al. (2000) demonstrated
that the second response is enhanced within 1 h
after LFPI, returning to control levels by 7 days
after injury. Schwarzbach et al. (2006), however,
using the same interpulse interval, did not observe
differences in PPF between sham vs. injured 7 days
after injury. Methodological or different recording
configurations may underlie this discrepancy.
Synaptic plasticity, like proper synaptic function
in general, requires a balance between inhibition
and excitation (Liu et al., 2004). As previously
reviewed, injury leads to a downward ampli-
tude shift in the area CA1 I/O curve (see Fig. 2),
demonstrating decreased excitability due to aug-
mented inhibition (Witgen et al., 2005). Thus, an
initial hypothesis postulates that augmented inhi-
bition is the reason why LTP cannot be induced
after TBI in area CA1. However, fully blocking
inhibition with the GABAA antagonist bicuculline
methiodide does not rescue LTP (Fig. 5C) refuting

Fig. 5. Inability to induce area CA1 LTP but not LTD following lateral LFPI in injured mice (open circles) compared with sham (filled
circles). (A) High frequency tetanic stimulation (arrows) results in facilitation of the fEPSP in hippocampal slices from sham but not
from injured mice. (B) Low frequency stimulation resulted in LTD of the fEPSP in slices derived from both injured and sham animals.
(C) Histogram demonstrating that LTP can be induced in slices from sham and naı̈ve animals using either the standard HFS or theta
burst stimulation (TBS); however, with both protocols, LTP cannot be induced in slices from LFPI animals. However, LTP can be
induced and maintained in area CA1 in the hippocampus contralateral to injury. Area CA1 LTP cannot be rescued by varying the
stimulation protocol, altering extracellular Ca2+ and Mg2+ concentrations or bath application of BMI. (D) LTD induced by a LFS
(line) in slices from FPI animals does not make subsequent LTP induction by HFS (arrows) possible. Responses were followed for an
additional 30 min post-tetanic stimulation, no potentiation (even a return to baseline) was observed. Reprinted from Hippocampus,
Schwarzbach et al. (2006), with permission from Wiley-Liss, Inc.

the hypothesis that overwhelming inhibition alone
underlies dysfunctional LTP (Schwarzbach et al.,
2006).
Data suggesting that LTD can be induced in
slices from injured animals show that plasticity
is possible in mice that have undergone FPI.
D’Ambrosio et al. (1998) suggested that it is not
alterations in the induction mechanisms which
induce LTP, but rather that injury itself potenti-
ates transmission leading to a saturation of LTP.
Therefore, if injury prepotentiates synapses, teta-
nic presynaptic stimulation will be unable to fur-
ther potentiate the fEPSP. To test this hypothesis,
slices from injured animals were first ‘‘depotenti-
ated’’ (using the same low frequency stimulation
protocol used to classically induce LTD) and
slices were then subsequently tetanized with a high
frequency, presynaptic stimulus in an attempt to
induce LTP. However, using this protocol did
not result in either LTP or, a return to baseline in
slices from FPI animals (Fig. 5D). This suggests
that the mechanism(s) specific for the induction of
LTP, and not for synaptic plasticity in general,
are in fact altered by injury (Schwarzbach et al.,
2006).
The anatomical substrate for LTP induction and
expression is the dendritic spine (Leuner et al.,
2003; Segal, 2005). Anatomical studies using
lucifer yellow illustrate that following FPI, the
mean apical and basal dendritic spine size (mm2)
significantly increases (Fig. 6; Schwarzbach et al.,
2006). This increased spine size could effect the
threshold level of calcium required for LTP
induction, which may be kept to a minimum due
to small spine size (Malenka and Nicoll, 1999;
Gazzaley et al., 2002).
It is known from many studies that the cal-
cium and calmodulin-dependent protein kinase
a-CaMKII, which is abundant postsynaptically, is
critical for the induction of LTP (Malenka et al.,
1989). Hence, a functional decrease in a-CaMKII
or a decrease in its phosphorylating activity after
injury could contribute to an inability to induce
LTP. Western blot analysis demonstrates a signifi-
cant decrease in the alpha form of CaMKII pro-
tein expression in regionally dissected area CA1
tissue from FPI injured animals (Schwarzbach
et al., 2006). However, CaMKII activity has been
153
shown in another study to increase after FPI in rat
brain (Atkins et al., 2006). Thus, Western blot
analysis of subcellular fractions from the hippo-
campus or parietal cortex regions reveal a redis-
tribution of phosphorylated a-CaMKII from the
cytosolic to membrane fractions 30 min after mode-
rate lateral FPI (Atkins et al., 2006). Presently
more studies are therefore needed to determine
whether reduced CaMKII function contributes to
injury-induced deficits in LTP expression.
The data presented above describe a consistent
inability to induce area CA1 LTP following LFPI
and explore many possible contributing mecha-
nisms. The work has presented many valuable
injury-induced alterations in the cascade leading to
increased synaptic efficacy, however the inability
to induce and maintain LTP does not seem to be
due to one finite cause.
Injury-induced functional alterations in the DG
DG: injury-induced alterations in DG excitability
The DG is thought to act as a filter to dampen
synaptic input from the perforant pathway (PP)
before it is amplified by area CA3. The lack of
interconnections between DGCs (Claiborne et al.,
1986) and robust feed-forward and feed-back
inhibition (Sloviter, 1994) function to impede syn-
chronization and endow the DG with its filtering
ability. Studies on the electrophysiological prop-
erties of the DG have consistently shown that
LFPI leads to increased DG excitability in rodents,
effectively reducing its filtering capacity. While
significant cell death has been observed in the hilar
region, the preferential survival of excitatory neu-
rons is not found (Lowenstein et al., 1992; Toth
et al., 1997; Santhakumar et al., 2000).
In injured animals, stimulation of the PP evokes
repetitive PS firing (Lowenstein et al., 1992; Santha-
kumar et al., 2000; Golarai et al., 2001), reduced
PS threshold and increased spike amplitude (Toth
et al., 1997; Santhakumar et al., 2000, 2001; Witgen
et al., 2005) compared with non-injured animals.
Studies have also shown that the amplitude and
slope of fEPSPs substantially increases in injured
animals (Coulter et al., 1996; Golarai et al., 2001;
154

Fig. 6. Increased dendritic spine size in slices from LFPI mice. (A) CA1 pyramidal neuron and its spiny dendrites from an injured
mouse loaded with Lucifer yellow. (pcl, pyramidal cell layer; sl, stratum lucidum; so, stratum oriens.) Scale bar: 10 mm. (B) Higher
magnification of area highlighted by box in section A showing fourth order basal dendrite segment spines. Scale bar: 1 mm. (C) Higher
magnification of area highlighted by box in section A showing second order apical dendrite segment spines. Scale bar: 1 mm.
(D) Neurolucida drawing of the spiny segment shown in B. Scale bar: 1 mm. (E) Comparison of the distribution of dendritic spine size
of sham (black) and FPI (gray) apical (E) and basal (F) CA1 pyramidal cells. Reprinted from Hippocampus, Schwarzbach et al. (2006),
with permission from Wiley-Liss, Inc.

Fig. 7. Extracellular input/output curves for the dentate gyrus evoked by afferent perforant path stimulation. LFPI in mice results in
increased fEPSP slopes compared with those from sham animals. Reprinted from Neuroscience, Witgen et al. (2005), with permission
from Elsevier.
Witgen et al., 2005). Consistent with rat models,
1 week after LFPI in the mouse, the I/O curve
(measuring the normalized slope of the fEPSP vs.
the stimulation intensity), is significantly shifted
upward with respect to amplitude as compared with
sham animals (Fig. 7). This injury-induced shift in
the I/O curve suggests that net synaptic efficacy
increases in the mouse DG after LFPI.
Since the discharge of DGCs is under tight in-
hibitory control by GABAergic interneurons, it is
constructive to completely antagonize inhibition,
in order to test whether injury-induced alterations
in DG excitability are due to changes in gluta-
matergic synaptic activity. One study presents data
supporting injury-induced alterations in the exci-
tatory limb of the DG (Santhakumar et al., 2000).
Using slices from injured rats 1 week post-LFPI,
afferent PP stimulation causes DG neurons to re-
spond with protracted depolarizations and aug-
ments action potential discharges in the absence of
GABAAergic inhibition. The late depolarizations
exhibited in DGCs are mediated by a polysynaptic
155
pathway and could be blocked with AMPA re-
ceptor antagonist(s) (Santhakumar et al., 2000).
The overall conclusion of this study is that exci-
tatory mossy cells become ‘‘irritable’’ and lead
to an increased excitatory drive onto both DG
excitatory and inhibitory neurons (Santhakumar
et al., 2000).
More recently studies have supported this
hypothesis showing mossy fiber sprouting at 2–3
and 14–15 weeks after a weight drop model of
injury (Golarai et al., 2001). In another study,
mossy fiber sprouting led to increased output
from the excitatory principal cells of the DG to
other granule cells and interneurons augmenting
excitation (Santhakumar et al., 2001). This was
based on findings that injury caused an increase
in the evoked PS lasting 1 month post-LFPI and
sIPSC from granule cells exhibited a higher
frequency through 6 months after LFPI. The
cause of this increased sIPSC was demonstrated
to be due to the glutamatergic excitatory drive to
interneurons.
156

DG: changes in inhibitory synaptic transmission (1–4 h) and lasting up to 4 days post-LFPI (Ross
and Soltesz, 2000). This depolarization is mediated
While it has been shown that surviving DGCs by an injury-induced reduction in the activity of
sustain TBI-induced increases in excitability, alter- the Na+/K+ pump (ATPase), which contributes
ations in inhibitory function cannot be ruled out. to maintaining the resting membrane potential.
Many studies demonstrate neuronal loss in the This 10 mV depolarization transiently raises the
hilar region, however, no preferential survival of excitability of the GABAAergic network in the
inhibitory neurons is observed, including soma- early post-traumatic DG.
tostatin, parvalbumin and cholecystokinin positive
interneurons (Lowenstein et al., 1992; Toth et al.,
TBI effects seizure threshold
1997; Santhakumar et al., 2000). While inhibitory
neuronal loss could result in a decrease of GABA-
As previously discussed, the DG’s filtering efficacy
ergic inhibitory efficacy, functional changes in
is crucial to preventing seizure activity. Thus, a
GABA receptors is also a possibility.
dysfunction in this region is a plausible expla-
Indeed, studies show injury-induced alterations
nation for the post-traumatic epilepsy observed
in the inhibitory network of the DG. By 1 week
(Annegers et al., 1996). Santhakumar et al. (2001)
post-injury in the rat, there is a decrease in mIPSCs
show that there is a decrease in the threshold to
frequency, limiting GABAergic inhibition without
generate sustained (30+ min) of seizure like acti-
altering mIPSC kinetics or amplitude (Toth et al.,
vity to tetanic stimulation, however no spontane-
1997). Similar results are found in the mouse but
ous seizure like activity is present in area CA1 at 3
the amplitudes of mIPSCs recorded in slices from
months (after recovery period).
injured animals 1 week post-LFPI are smaller than
Furthermore, slices generated from LFPI
sham animals (Figs. 8A–D), while the kinetics of
animals have a lower threshold for developing
mIPSCs are not altered. The consequence of the
self-sustained seizure activity 1 week post-injury
decrease in mIPSC amplitude is demonstrated by a
(Coulter et al., 1996) and persisting up to 3 months
significant reduction in total mIPSC net charge
(Santhakumar et al., 2001). Coulter et al. (1996)
transfer in neurons from FPI animals, which was
found greater disinhibition in the DG of slices
76% of that calculated for neurons from sham
from FPI animals 1 week after injury when com-
slices (Fig. 8E). Furthermore, the mean frequency
pared with both sham-operated and epileptic ani-
of occurrence of mIPSCs is significantly lower
mals, which could be a possible explanation for the
in neurons from FPI animals than that present
observed decreased seizure threshold. In addition,
in DG neurons in sham slices (Fig. 8F). These
using the weight drop model of TBI, Golarai et al.
alterations in GABAAergic function demonstrate
(2001) demonstrate that injured animals have a
that spontaneous inhibitory activity (known to be
lower seizure threshold to pentylenetetrazole-
mediated by basket and axo-axonic interneurons)
induced convulsions 15 weeks post-TBI (Golarai
regulating feed-forward inhibition on dentate gran-
et al., 2001).
ule neurons is significantly compromised 1 week
post-LFPI.
Interestingly, the frequency of spontaneous ac- Cellular mechanisms of abnormal electrophysiology
tion potential-mediated inhibitory postsynaptic after trauma: the effect of tensile strain on
currents is enhanced up to 5 months after injury. ionotropic (excitatory and inhibitory) receptors of
This increase in spontaneous inhibitory post- cortical neurons in an in vitro model of TBI
synaptic current (sIPSC) frequency is due to en-
hanced excitatory drive (from irritable mossy cells) As illustrated above, TBI induces regional altera-
onto inhibitory interneurons (Santhakumar et al., tions in excitatory and inhibitory synaptic function
2000). Additionally, LFPI also causes a transient in hippocampal slices from rats and mice subjected
(approximately 10 mV) depolarized shift in DG to LFPI. Elucidation of the molecular sites and
interneurons exhibited immediately after injury biochemical pathways that mediate synaptic
 157

Fig. 8. Whole-cell voltage clamp recordings from dentate granule neurons from injured mice demonstrate reductions in spontaneous
miniature (mIPSC) activity 1 week after LFPI. The mIPSCs are smaller and less frequent in dentate granule neurons from LFPI
animals. (A) Continuous sweeps of mIPSCs from sham (left panel) and LFPI (right panel) dentate granule cells at
60 mV. (B)
Cumulative frequency amplitude histogram for neurons in (A) with representative mIPSCs from sham and injured dentate granule
neurons. Histograms of mean median (D) mIPSC amplitude, (E) net charge transfer, and (F) mIPSC frequency demonstrate significant
reductions in inhibitory activity of neurons from LFPI mice compared with sham animals. Reprinted from Neuroscience, Witgen et al.
(2005), with permission from Elsevier.

dysfunction in the CNS would provide favorable
targets for putative pharmacological therapeutic
treatments directed toward improving cognitive
function and memory after TBI. Yet, presently,
little is known in detail regarding the cellular
mechanisms underlying dysfunctional neurotrans-
mission. For example, it is still unresolved whether
mechanical forces directly impact upon neurons
during traumatic impact to trigger secondary con-
sequences of injury or affect neuronal function in-
directly by the release of excitatory and other CNS
neurotransmitters during the so-called ‘depolari-
zation impact’ phase of trauma (Katayama et al.,
1990). Investigators have thus begun to examine
158
the cellular processes that result from mechanical
trauma in further detail and more directly by using
simplified in vitro cell injury models.
In the earliest in vitro models of mechanical
trauma, pyramidal neurons were injured by
scratching a cultured neuronal cell layer with a
glass sylet (Tecoma et al., 1989), transecting cells in
a manner similar to tissue damage incurred during
penetrating TBI. Alternative in vitro models have
been developed to simulate the forces (tensile,
compressional or torsional) acting on brain tissue
during blunt trauma or acceleration/deceleration
injury, in order to more fully characterize the
mechanisms by which mechanical deformation (or
tensile strain) leads to changes in neurons and glia
(Pike, 2001). Models of tensile strain (stretch) typi-
cally use pneumatic or mechanical translational
devices to deform a stretchable substrate upon
which various types of neurons or brain slices are
cultured and are thus adherent (Ellis et al., 1995;
Morrison et al., 1998, 2006; Arundine et al., 2003).
Shear strain or torque, which approximates the in-
ertial loading caused by the rapid acceleration of
brain tissue during TBI, has been modeled by forc-
ing fluid over stationary cells using a cell shearing
injury device (CSID) (LaPlaca and Thibault, 1998),
or a newly developed electro-mechanical cell shear-
ing model, which can be used to examine the effects
of torsional stress applied to three dimensional
neuronal cultures (LaPlaca et al., 2005).
As simplified systems, in vitro models allow in-
jury to be studied in the absence of ischemia,
blood/brain barrier considerations, or other
changes, and the models provide increased exper-
imental control of tissue oxygenation, the cell type
to be studied, availability of nutrients, temperature
and ionic and drug concentrations. For electro-
physiological studies, in vitro models enable rapid
drug or agonist application, patch clamping and
intracellular calcium measurements to be carried
out with relative ease. In vitro models are thus well
suited for examining the direct effects of a me-
chanical insult on ion channel or receptor function.
Using the in vitro model of stretch injury devel-
oped by Ellis et al. (1995) several studies reveal
alterations in the function of neurotransmitter re-
ceptors, including AMPA, NMDA and GABAA
receptors following mechanical strain (Zhang et al.,
1996; Goforth et al., 1999, 2004; Kao et al., 2004).
Ellis et al. (1995) developed an in vitro TBI model
in which neonatal rat cortical neurons are co-
cultured with astrocytes in 6-well plates having
deformable SILASTIC (Dow Corning) bottoms
(FlexCell Plates; Ellis et al., 1995). An injury con-
troller used in this model applies a controlled pulse
of compressed nitrogen to individual wells via an
airtight gasket (Fig. 9; Ellis et al., 1995). Displace-
ment of the SILASTIC membrane with its adher-
ent cells thereby rapidly stretches the cells within
100 ms, mimicking the deformation forces that
contribute to rapid acceleration/deceleration in-
jury (Pike, 2001). It is important to emphasize
that the forces produced by this model are likely
to be highly relevant and akin to those encoun-
tered by the brain in vivo following acceleration/
deceleration induced TBI, as evidenced by bio-
mechanical studies and simulations carried out
using computer modeling (Margulies et al., 1990;
Meaney et al., 1995).
The Ellis model produces a range of injury levels,
ranging from mild to moderate to severe. Signifi-
cantly, the model generates an injury in a very re-
producible manner. Access to the multiple wells
provides good experimental control, and facilitates
studies of pharmacological agents to probe the
mechanisms of trauma or to evaluate potential
therapeutic agents. The model also facilitates
studies of cell viability by using propidium iodide
staining (McKinney et al., 1996), mitochondrial
membrane potential measurement using rhodamine
123 (Ahmed et al., 2000, 2002), cell calcium levels
measured with fura-2 (Rzigalinski et al., 1997,
1998a, b; Weber et al., 2001, 2002), electrophysio-
logy (Tavalin et al., 1995, 1997; Zhang et al., 1996;
Goforth et al., 1999, 2004; Lea et al., 2002; Kao
et al., 2004) and biochemical assays of either the
attached cells or the bathing media (Lamb et al.,
1997; Rzigalinski et al., 1999; Hoffman et al.,
2000).
Mild injury due to a 5.7 mm displacement of the
Silastic membrane (corresponding to 31% stretch of
brain cells) releases glutamate from astrocytes, with
peak glutamate occurring within 5 min after injury,
and then declining within an hour (Rzigalinski et al.,
1998a, b). The model also produces a transient rise
in neuronal [Ca2+]i which persists for 410 min and

Fig. 9. In vitro TBI model of mechanical injury. An injury controller applies a timed pulse of nitrogen to cultured cells grown on a
deformable SILASTIC membrane to mimic stretch deformation in TBI. Reprinted from J. Neurotrauma, Ellis et al. (1995), with
permission from Mary Ann Liebert, Inc., Publishers.
is partially attenuated by the NMDA antagonist
MK-801 (Ahmed et al., 2002). In addition [Ca2+]i
responses of mildly injured neurons to exogenous
glutamate are potentiated 15 min–24 h post-injury
(Weber et al., 1999). The Ellis model reproduces
many features of whole animal TBI including
changes in cell viability (McKinney et al., 1996),
neuronal and glial [Ca2+]i homeostasis (McKinney
et al., 1996; Zhang et al., 1996; Rzigalinski et al.,
1998a, b; Weber et al., 1999, 2002), the status of
endoplasmic reticulum (ER) Ca stores (Weber et al.,
1999; Floyd et al., 2001), free radicals (McKinney
et al., 1996), phosphoinositide metabolism (Floyd
et al., 2001), mitochondrial function (Ahmed et al.,
2000, 2002) and electrophysiology (Tavalin et al.,
1995, 1997; Ahmed et al., 2000).
159
Stretch injury alters cortical glutamate receptor
function
Excitatory synaptic transmission is mediated by
AMPA and NMDA glutamate receptor subtypes.
Both NMDA and AMPA receptors are heteromul-
timers, composed of combinations of NR1 and
NR2A-D or GluR1-4 subunits, respectively. The
NMDA current of control, uninjured neurons is
well known to be characterized by its ‘‘J-shaped’’
voltage dependence, which is due to external
voltage-dependent Mg2+ block of NMDA recep-
tors (NMDARs) (Mayer et al., 1984; McBain and
Mayer, 1994; Dingledine et al., 1999). However, in
mildly injured neurons, voltage-dependent Mg2+
block of NMDARs is greatly reduced (Fig. 10;
160
Fig. 10. Stretch injury reduces Mg2+ block of NMDARs, as
evidenced by linearization of the J-shaped NMDA current
(I)–voltage (V) curve. I–V currents generated by the application
of 200 mM NMDA and 10 mM glycine to an uninjured control
and a mildly stretch-injured cortical pyramidal neuron. Re-
printed from Science, Zhang et al. (1996), with permission from
AAAS.
Zhang et al., 1996). Thus, injury shifts the IC50 of
Mg2+ inhibition of NMDA current by 100-fold,
and reduces its Bmax by 50% (Zhang et al., 1996).
Moreover, raising [Mg2+]o restores some of the
block to the channels (Zhang et al., 1996). The loss
of Mg2+ sensitivity of NMDARs results in turn in
potentiated [Ca2+]i responses to exogenously ap-
plied NMDA (Zhang et al., 1996). This enhanced
NMDA-elicited [Ca2+]i response was recently du-
plicated in a different in vitro model of stretch in-
jury (Geddes-Klein et al., 2006), and in other
studies by Lea et al. (2002, 2003, see below). Im-
portantly, pretreating neurons with the PKC in-
hibitor calphostin C, partially prevents the loss of
Mg2+ block (Zhang et al., 1996), consistent with
an earlier report that PKC modulates the Mg2+
sensitivity of NMDA receptors in nodose neurons
(Chen and Huang, 1992).
Using the stretch-injury model, Faden and co-
workers confirmed that in vitro injury indeed de-
creases external Mg2+ block of NMDARs and
increases NMDA current in cultured neurons (Lea
et al., 2002). The effect of stretch injury on NMDA
receptors was shown to be altered by the activation
of the Group I metabotropic glutamate receptors
(mGluRs), mGluR1 and mGluR5 (Lea et al., 2002,
2003). Furthermore, the effect of mGluR agonists
and antagonists on injury-induced NMDA potent-
iation is dependent on the presence or absence of
glial cells (Lea et al., 2002, 2003). In pure neuronal
cultures, inhibition of either mGluR1 or mGluR5
prevents stretch-induced NMDA receptor potent-
iation, suggesting a contributing role of Group I
mGluRs in NMDA alteration (Lea et al., 2002,
2003). However, while mGluR5 inhibition also at-
tenuates NMDA alterations in neurons grown in
mixed cultures with glia, inhibition of mGluR 1
augments injury-induced NMDA potentiation (Lea
et al., 2002, 2003). Activation of mGluR1 also
differentially modulates the effect of injury on
NMDA receptors, depending on the presence or
absence of glia (Lea et al., 2003). The above data
highlight the complexity of the biochemical and
cellular pathways leading to alterations in NMDA
receptors after mechanical injury and suggest the
possible involvement of not only neuronal but also
glial mGluRs in this process.
Reduced Mg2+ block of NMDARs has been
observed in other pathological conditions, includ-
ing crush injury of motoneuron axons (Furukawa,
2000), axotomization of spinal cord motoneurons
(Abdrachmanova et al., 2002), kainate lesioning of
hippocampal neurons (Chen et al., 1999) and peri-
pheral inflammation of dorsal horn neurons (Guo
and Huang, 2001). The neurohormone somatosta-
tin also reduces the Mg2+ block of hippocampal
NMDARs by increasing PKC activity (Pittaluga
et al., 2000). Thus, loss of Mg2+ block appears to
be a general mechanism of traumatic injury and
neuromodulation.
Stretch injury via the Ellis in vitro model also
modulates AMPARs in cortical neurons. Goforth
et al. (1999) found mild injury similar to that used
to modify NMDA current also increases whole-cell
steady-state AMPAR current density (Goforth
et al., 1999; Fig. 11A). Stretch does not appear to
modify the sensitivity of AMPARs to AMPA,
AMPAR current–voltage properties or the appar-
ent number of AMPA receptors on pyramidal neu-
rons. However, pharmacological reduction of
AMPAR desensitization using cyclothiazide or ac-
tivation of AMPARs with the weakly AMPAR
desensitizing agonist kainate, offsets the observed
differences between control and injured AMPA
 161

Fig. 11. Stretch injury alters AMPAR kinetics and currents. (A) Whole-cell AMPA-elicited currents in control and injured cultured
cortical pyramidal neurons. Stretch injury reduces fast desensitization of AMPARs, resulting in larger steady-state currents.
(B) AMPAR currents recorded from outside/out patches in response to a 100 ms pulse of 1 mM glutamate+20 mM APV. (C) Mean
20–80% activation time (top) and desensitization rate (bottom) of AMPAR currents recorded from outside/out patches (*po0.05).
Reprinted from J. Neurosci., copyright 1999 by the Society for Neuroscience, and J. Neurotrauma, Goforth et al. (2004), with
permission from Mary Ann Liebert, Inc., Publishers.

currents, suggesting that injury suppresses AMPAR
desensitization (Goforth et al., 1999). Changes in
AMPAR kinetics surprisingly even persist in cell
free membrane patches excised from injured neu-
rons vs. controls and include slower AMPAR
activation as well as desensitization (Fig. 11B).
AMPA current alterations persist for at least 24 h
post-injury and stretching neurons with the NMDA
antagonist APV present prevents the alterations
in AMPA currents (Goforth et al., 2004). This
suggests that NMDAR activation is again a pre-
requisite for AMPAR modulation by injury.
CaMKII is well known to be an important
mediator of synaptic modulation in the CNS (see
above in CA1 plasticity section). CaMKII also
appears to play an important role in mediating
changes in ionotropic receptors associated with
TBI in cortical neurons. Thus, pretreatment of
cultured cortical neurons with the CaMKII inhib-
itor KN-93 (N-[2-[[[3-(4-chlorophenyl)-2-propenyl]
methylamino]methyl]phenyl]-N-(2-hydroxyethyl)-4
methoxybenzenesulphonamide) in the Ellis TBI
model also prevents the alterations in AMPAR
kinetics observed following mild mechanical stretch
(Goforth et al., 2004). CaMKII is [Ca2+]i -activated,
is known to phosphorylate the GluR1 subunit of the
AMPA receptor and is thought to participate in the
induction of LTP in area CA1 of the hippocampus
(Lisman et al., 1997; Malenka and Nicoll, 1997,
1999; Malinow et al., 2000; Pickard et al., 2001;
Fink and Meyer, 2002; Malinow and Malenka,
2002; Song and Huganir, 2002). In support of the
involvement of CAMKII in injury, measurements
of CaMKII activity in cell lysates from control
and stretch-injured neurons show an increase in
the percent of autophosphorylated autonomous
CaMKII activity at 10 min post-injury (R. Tombes,
P. Goforth and L. Satin, unpublished data). This is
consistent with a demonstrated increase in CamKII
after FPI in vivo, which is accompanied by
increased expression of phosphorylated GluR1, an
AMPA receptor subunit and CaMKII substrate
(Atkins et al., 2006).
Although studies indicate that the effect of
stretch injury on NMDA and AMPA receptors
appear to involve the activation of the protein
162
kinases PKC and CaMKII, respectively (Zhang
et al., 1996; Goforth et al., 2004), it is not known
whether activation of these kinases after injury
leads to direct phosphorylation of ionotropic gluta-
mate receptors or associated proteins. Modulation
of AMPARs by phosphorylation reflects a com-
plex balance of kinase and phosphatase activity
that can affect not only receptor subunits but an-
choring or modulatory proteins that interact with
AMPARS. Thus, it is important to determine the
phosphorylation state of specific AMPA and
NMDA subunits after injury in a highly quanti-
tative manner. It is also possible that injury alters
the expression and/or localization of ionotropic
glutamate receptors and these changes may or may
not be dependent on phosphorylation.
Changes in glutamate receptor function such as
those described above following stretch injury
would be expected to contribute to dysfunctional
synaptic transmission in CNS circuits impacted
by mechanical trauma. In fact preliminary data
demonstrate decreased mEPSC amplitude, and
slowed mEPSC rise time and duration in cortical
pyramidal neurons observed immediately after
mild stretch injury (Satin, unpublished data).
Injury-induced reductions in AMPA rise time
may contribute to decreased mEPSC amplitudes,
while the slowing of the mEPSC time course may
be effected by reduced AMPA receptor desensiti-
zation as well increased contribution of NMDA
receptor activation after mechanical injury (Zhang
et al., 1996; Goforth et al., 1999, 2004). Sponta-
neous glutamatergic EPSCs of cortical pyramidal
neurons also exhibit reduced amplitude following
mild stretch injury (Satin, unpublished data).
These data are in general agreement with studies
that show attenuation of glutamate-evoked cur-
rents recorded from pyramidal CA1 neurons from
mice subjected to FPI (Schwarzbach et al., 2006).
The discrepant findings regarding the relative
contribution of NMDA receptors to EPSCs after
injury may be due, in part, to the examination of
different brain regions (cortex vs. hippocampus),
or the post-injury time point at which studies were
conducted (immediately vs. 7 days post-injury). To
further elucidate the role of direct glutamate re-
ceptor alterations in dysfunctional neurotransmis-
sion after TBI, it will be important to determine
whether similar alterations in AMPA and NMDA
receptors occur in different brain regions (i.e., hippo-
campal CA1, CA3 and DG vs. cortical neurons)
and determine the time course of these alterations.
Stretch injury alters cortical GABAA receptor
function
As discussed above, many studies have demon-
strated injury-induced dysfunctional inhibitory
synaptic transmission in the hippocampus (Toth
et al., 1997; Witgen et al., 2005). As with glutamate
receptors, a study of cortical GABAA receptor
function suggests that direct alterations in GABAA
receptors likely contributes to inhibitory dysfunc-
tion after trauma (Kao et al., 2004). Using the Ellis
in vitro model, Kao et al. (2004) demonstrate that
mild mechanical injury augments GABA-activated
currents, which were blocked by bicuculline,
confirming their mediation by GABAA receptors
(Fig. 12). Stretch injury does not shift the EC50 or
current–voltage relationship of these GABAA cur-
rents. However, as reported for AMPA receptors,
GABA current potentiation is prevented by injur-
ing neurons in the presence of the NMDA re-
ceptor antagonist APV or the CaMKII inhibitor
KN-93 (Kao et al., 2004). CaMKII has been
shown to phosphorylate GABAA channels (Swope
et al., 1999), and inhibiting the expression of
the CaMKII alpha subunit is known to decrease
GABAA receptor function (Churn et al., 2000).
More studies are needed to determine the mech-
anism mediating enhanced GABA current, which,
in theory, could result from either an increase in
the number of functional GABAA channels in the
neuronal plasma membrane, an increase in the
unitary current through open GABAA channels or
an increase in the probability of opening of a fixed
number of GABAA channels.
As with glutamate receptors, direct alterations
in GABAA channels would likely contribute to
dysfunctional inhibitory synaptic transmission
after TBI. In fact, miniature as well as spontane-
ous IPSCs recorded from stretch-injured cortical
neurons exhibit increased amplitude (Goforth and
Satin, unpublished data), but no change in mIPSC
frequency, such as observed in hippocampal
 163

Fig. 12. Stretch injury potentiates GABA-elicited whole-cell current. (A) Patch-clamp recordings of whole-cell currents elicited by the
application of 50 mM GABA (filled bar) for a control neuron (left) and a stretched-injured neuron (right). Neurons were voltage-
clamped to –60 mV, and the rapid application of GABA-induced inward current. GABA-elicited currents of mildly injured neurons
were increased in amplitude compared with uninjured control neurons. (B) Amplitude histograms of GABA current density shown for
control neurons (left) and stretch-injured neurons (right). Currents were elicited by 50 mM GABA, and were normalized by neuron
capacitance to yield GABA current density (pA/pF). Mean current densities obtained were 20.271.7 pA/pF (mean7SE, n ¼ 69) for
control neurons and 41.272.6 pA/pF (n ¼ 82) for injured neurons (po0.001; Student’s t test). Reprinted from J. Neurotrauma, Kao et
al. (2004), with permission from Mary Ann Liebert, Inc., Publishers.

CA1 neurons from mice subjected to FPI in vivo
(Witgen et al., 2005).
The data discussed clearly demonstrate that
there are prominent injury-induced alterations in
the function of glutamate and GABA receptors
in cortical neurons. Such changes occurring in
postsynaptic neurotransmitter receptor function
would thus be expected to contribute to dysfunc-
tional synaptic transmission in vivo or in brain
slices from TBI rodents. To underscore an
164
enduring theme of this chapter, we stress that the
changes observed in synaptic function in vitro
following injury were seen using levels of injury
unable to cause significant derangement of cell
attachment, cell morphology or outright cell
death. This emphasizes the importance of study-
ing changes in synaptic function after mild and
moderate TBI, which as a topic needs far more
emphasis by the neurotrauma field.
Thus, the detailed study of CNS synaptic mech-
anisms and their exquisite sensitivity to mechanical
deformation remains highly significant for the field
and will likely result in new insights of importance
to the development of new approaches for treating
TBI patients.
Stretch injury alters Na+/K+ ATPase activity
In addition to abnormal neurotransmitter receptor
function, neurotransmission and neuronal excita-
bility in the CNS may also be modulated by
trauma-induced alterations in ion pump and trans-
porter activity. Under normal conditions, the
resting membrane potentials of many pyramidal
neurons includes an electrogenic component medi-
ated by the basal activity of plasma membrane
Na+/K+ ATPases or ‘Na pumps’ (Tavalin et al.,
1995, 1997). This occurs because of unequal (i.e.,
electrogenic) fluxes of Na+ out of the neuron
and K+ into the neuron driven by the Na+/K+
ATPase, which is mediated energetically by ATP
hydrolysis. In vitro studies of hippocampal neurons
in brain slices from TBI rats and cortical neurons
subjected to mild stretch result in clear injury-
induced decrease in Na+/K+ ATPase activity,
which in turn leads to depolarized neuronal resting
membrane potentials recorded in both the in vitro
stretch and LFPI injury models (Tavalin et al.,
1997; Ross and Soltesz, 2000). In the mildly
stretch-injured neurons, resting membrane poten-
tial depolarization (magnitude 10–12 mV), deve-
loped 15–60 min after injury and was termed
‘‘stretch-induced delayed depolarization’’ (or
‘‘SIDD’’) (Tavalin et al., 1995). SIDD results from
a Ca2+-influx dependent loss of electrogenic Na/K
ATPase activity and is prevented by pretreatment
with the NMDA antagonist APV, the Na+ channel
blocker TTX (tetrodotoxin) or a bathing solution
lacking calcium. However, SIDD is unaffected by
the AMPA receptor (AMPAR) antagonist CNQX
(6-cyano-7-nitroquinoxaline-2,3-dione) indicating
that the requirement for glutamate receptor acti-
vation is subtype specific (Tavalin et al., 1997).
Soltesz and colleagues (Ross and Soltesz, 2000)
found that the membrane potentials of DGCs in
brain slices from TBI rats are also depolarized due
to the same ionic mechanism identified using the
Ellis model by Tavalin et al. (Ross and Soltesz,
2000). Functionally, SIDD would be expected to
contribute to the disruptions in ionic homeostasis,
and changes in neuronal excitability known to
occur after TBI.
Conclusions
Normal brain function is critically dependent on
the activation and tight regulatory control of
highly organized and precise neural circuits, the
fundamental unit of which is the synapse. It is well
documented that TBI disrupts normal brain func-
tion, resulting in behavioral, motor and cognitive
deficits such as impaired memory, attention and
executive function, as well as slowed processing
speed (Salmond and Sahakian, 2005). We postu-
late that such deficits most likely reflect abnormal
synaptic transmission after TBI.
Synapses are either excitatory or inhibitory,
defined by activation of the specific neurotransmit-
ter released. Overall neuronal activity is shaped by
the integration of inhibitory and excitatory synap-
tic potentials, and slight changes in the balance of
excitation and inhibition, or their speed of occur-
rence can produce profound changes in brain func-
tion by affecting synaptic information processing.
Synaptic inputs moreover are not static, but rather
quite plastic and are capable of being modulated in
response to physiological stimulation, resulting in
changes in the number or pattern of synaptic con-
nections, presynaptic release of neurotransmitter or
the functional properties of neuronal postsynaptic
receptors (Malenka and Bear, 2004). Pathological
alterations in the strength, timing or number of
synaptic inputs would be expected to affect overall
integration of synaptic potentials in individual

neurons, leading to abnormal cell firing and dys-
functional information processing and transmis-
sion of signals within the important neural circuits
of the cerebral cortex and hippocampus, as well as
other brain regions.
In order to understand the mechanisms underly-
ing the cognitive changes that occur after TBI, it is
thus critically important to elucidate the effects of
injury on each key component of brain function:
neuronal circuitry, cell to cell synaptic transmission
and intrinsic cellular electrical activity. Despite im-
portant advances made in this area of investigation,
most of the mechanisms involved in mediating the
effects of TBI on the electrophysiology of neurons
and synapses in the CNS still remain unknown.
These mechanisms and their molecular and genetic
underpinnings are therefore a major gap in our
knowledge of the effects of TBI on the brain and
should therefore be considered the next frontier of
brain injury research. We thus believe that they de-
serve much more attention by investigators work-
ing in this field and hope that the present review will
stimulate new investigations of synaptic function
and neuronal electrophysiology following TBI.
References
Abdrachmanova, G., Teisinger, J. and Vyklicky Jr., L. (2002)
Axotomy-induced changes in the properties of NMDA re-
ceptor channels in rat spinal cord motoneurons. J. Physiol.,
538(Pt 1): 53–63.
Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding,
H.A. and Ellis, E.F. (2000) Stretch-induced injury alters mi-
tochondrial membrane potential and cellular ATP in cultured
astrocytes and neurons. J. Neurochem., 74(5): 1951–1960.
Ahmed, S.M., Weber, J.T., Liang, S., Willoughby, K.A., Sit-
terding, H.A., Rzigalinski, B.A. and Ellis, E.F. (2002)
NMDA receptor activation contributes to a portion of the
decreased mitochondrial membrane potential and elevated
intracellular free calcium in strain-injured neurons. J. Neuro-
trauma, 19(12): 1619–1629.
Akasu, T., Muraoka, N. and Hasuo, H. (2002) Hyperexcitability
of hippocampal CA1 neurons after fluid percussion injury of
the rat cerebral cortex. Neurosci. Lett., 329(3): 305–308.
Annegers, J.F., Rocca, W.A. and Hauser, W.A. (1996) Causes
of epilepsy: contributions of the Rochester epidemiology
project. Mayo Clin. Proc., 71(6): 570–575.
Arundine, M., Chopra, G.K., Wrong, A., Lei, S., Aarts, M.M.,
MacDonald, J.F. and Tymianski, M. (2003) Enhanced vul-
nerability to NMDA toxicity in sublethal traumatic neuronal
injury in vitro. J. Neurotrauma, 20(12): 1377–1395.
165
Asikainen, I., Kaste, M. and Sarna, S. (1999) Early and late
posttraumatic seizures in traumatic brain injury rehabilita-
tion patients: brain injury factors causing late seizures and
influence of seizures on long-term outcome. Epilepsia, 40(5):
584–589.
Atkins, C.M., Chen, S., Alonso, O.F., Dietrich, W.D. and Hu,
B.R. (2006) Activation of calcium/calmodulin-dependent
protein kinases after traumatic brain injury. J. Cereb. Blood
Flow Metab., 26(12): 1507–1518.
Barnard, E.A., Skolnick, P., Olsen, R.W., Mohler, H., Sieghart,
W., Biggio, G., Braestrup, C., Bateson, A.N. and Langer,
S.Z. (1998) International Union of Pharmacology. XV. Sub-
types of gamma-aminobutyric acid a receptors: classification
on the basis of subunit structure and receptor function.
Pharmacol. Rev., 50(2): 291–313.
Barrett, E.F. and Stevens, C.F. (1972) The kinetics of trans-
mitter release at the frog neuromuscular junction. J. Physiol.,
227(3): 691–708.
Barth, J.T., Macciocchi, S.N., Giordani, B., Rimel, R., Jane,
J.A. and Boll, T.J. (1983) Neuropsychological sequelae of
minor head injury. Neurosurgery, 13(5): 529–533.
Bear, M.F. and Abraham, W.C. (1996) Long-term depression
in hippocampus. Annu. Rev. Neurosci., 19: 437–462.
Bennett-Levy, J.M. (1984) Long-term effects of severe closed
head injury on memory: evidence from a consecutive series of
young adults. Acta Neurol. Scand., 70(4): 285–298.
Bigler, E.D., Blatter, D.D., Anderson, C.V., Johnson, S.C.,
Gale, S.D., Hopkins, R.O. and Burnett, B. (1997) Hippo-
campal volume in normal aging and traumatic brain injury.
AJNR Am. J. Neuroradiol., 18(1): 11–23.
Bohbot, V.D., Allen, J.J. and Nadel, L. (2000) Memory deficits
characterized by patterns of lesions to the hippocampus and
parahippocampal cortex. Ann. N.Y. Acad. Sci., 911: 355–368.
Carbonell, W.S., Maris, D.O., McCall, T. and Grady, M.S.
(1998) Adaptation of the fluid percussion injury model to the
mouse. J. Neurotrauma, 15(3): 217–229.
Cave, C.B. and Squire, L.R. (1991) Equivalent impairment of
spatial and nonspatial memory following damage to the hu-
man hippocampus. Hippocampus, 1(3): 329–340.
Chen, L. and Huang, L.Y. (1992) Protein kinase C reduces
Mg2+ block of NMDA-receptor channels as a mechanism of
modulation. Nature, 356(6369): 521–523.
Chen, Y., Chad, J.E., Cannon, R.C. and Wheal, H.V. (1999)
Reduced Mg2+ blockade of synaptically activated N-methyl-
D-aspartate receptor-channels in CA1 pyramidal neurons in
kainic acid-lesioned rat hippocampus. Neuroscience, 88(3):
727–739.
Christie, B.R., Kerr, D.S. and Abraham, W.C. (1994) Flip side
of synaptic plasticity: long-term depression mechanisms in
the hippocampus. Hippocampus, 4(2): 127–135.
Churn, S.B., Sombati, S. and Delorenzo, R.J. (2000) Inhibition
of CaMKII alpha subunit expression results in decreased
GABAA receptor function. 2000 Abstract Viewer and Itinerary
Planner. Washington D.C. Society for Neuroscience. Online.
Claiborne, B.J., Amaral, D.G. and Cowan, W.M. (1986) A light
and electron microscopic analysis of the mossy fibers of the
rat dentate gyrus. J. Comp. Neurol., 246(4): 435–458.
166
Cobb, S.R., Buhl, E.H., Halasy, K., Paulsen, O. and Somogyi, P.
(1995) Synchronization of neuronal activity in hippocampus by
individual GABAergic interneurons. Nature, 378(6552): 75–78.
Cohen, A.S. and Abraham, W.C. (1996) Facilitation of long-
term potentiation by prior activation of metabotropic gluta-
mate receptors. J. Neurophysiol., 76(2): 953–962.
Coulter, D.A., Rafiq, A., Shumate, M., Gong, Q.Z., De-
Lorenzo, R.J. and Lyeth, B.G. (1996) Brain injury-induced
enhanced limbic epileptogenesis: anatomical and physiolog-
ical parallels to an animal model of temporal lobe epilepsy.
Epilepsy Res., 26(1): 81–91.
D’Ambrosio, R., Maris, D.O., Grady, M.S., Winn, H.R. and
Janigro, D. (1998) Selective loss of hippocampal long-term
potentiation, but not depression, following fluid percussion
injury. Brain Res., 786(1–2): 64–79.
Del Castillo, J. and Katz, B. (1954) Statistical factors involved
in neuromuscular facilitation and depression. J. Physiol.,
124(3): 574–585.
Deweer, B., Pillon, B., Pochon, J.B. and Dubois, B. (2001) Is
the HM story only a ‘‘remote memory’’? Some facts about
hippocampus and memory in humans. Behav. Brain Res.,
127(1–2): 209–224.
Dingledine, R., Borges, K., Bowie, D. and Traynelis, S.F.
(1999) The glutamate receptor ion channels. Pharmacol.
Rev., 51(1): 7–61.
Dixon, C.E., Lyeth, B.G., Povlishock, J.T., Findling, R.L.,
Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L.
(1987) A fluid percussion model of experimental brain injury
in the rat. J. Neurosurg., 67(1): 110–119.
Dudek, F.E., Snow, R.W. and Taylor, C.P. (1986) Role of
electrical interactions in synchronization of epileptiform
bursts. Adv. Neurol., 44: 593–617.
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-
induced injury of cells in culture: characterization of the
model using astrocytes. J. Neurotrauma, 12(3): 325–339.
Fink, C.C. and Meyer, T. (2002) Molecular mechanisms of
CaMKII activation in neuronal plasticity. Curr. Opin. Ne-
urobiol., 12(3): 293–299.
Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A.,
Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury
of cultured astrocytes alters inositol (1,4,5)-trisphosphate-
mediated signaling. Glia, 33(1): 12–23.
Freund, T.F. and Buzsaki, G. (1996) Interneurons of the hippo-
campus. Hippocampus, 6(4): 347–470.
Fricke, R.A. and Prince, D.A. (1984) Electrophysiology of
dentate gyrus granule cells. J. Neurophysiol., 51(2): 195–209.
Furukawa, Y. (2000) Reduction of voltage-dependent magne-
sium block of N-methyl-D-aspartate receptor-mediated cur-
rent by in vivo axonal injury. Neuroscience, 96(2): 385–392.
Gazzaley, A., Kay, S. and Benson, D.L. (2002) Dendritic spine
plasticity in hippocampus. Neuroscience, 111(4): 853–862.
Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F. (2006)
Mechanisms and consequences of neuronal stretch injury in
vitro differ with the model of trauma. J. Neurotrauma, 23(2):
193–204.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement
of AMPA-mediated current after traumatic injury in cortical
neurons. J. Neurosci., 19(17): 7367–7374.
Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical
injury modulates AMPA receptor kinetics via an NMDA
receptor-dependent pathway. J. Neurotrauma, 21(6):
719–732.
Golarai, G., Greenwood, A.C., Feeney, D.M. and Connor, J.A.
(2001) Physiological and structural evidence for hippocampal
involvement in persistent seizure susceptibility after trau-
matic brain injury. J. Neurosci., 21(21): 8523–8537.
Gorman, L.K., Shook, B.L. and Becker, D.P. (1993) Traumatic
brain injury produces impairments in long-term and recent
memory. Brain Res., 614(1–2): 29–36.
Grady, M.S., Charleston, J.S., Maris, D., Witgen, B.M. and
Lifshitz, J. (2003) Neuronal and glial cell number in the hippo-
campus after experimental traumatic brain injury: analysis
by stereological estimation. J. Neurotrauma, 20(10): 929–941.
Guo, H. and Huang, L.Y. (2001) Alteration in the voltage de-
pendence of NMDA receptor channels in rat dorsal horn
neurones following peripheral inflammation. J. Physiol.,
537(Pt 1): 115–123.
He, J., Yamada, K., Zou, L.B. and Nabeshima, T. (2001) Spa-
tial memory deficit and neurodegeneration induced by the
direct injection of okadaic acid into the hippocampus in rats.
J. Neural. Transm., 108(12): 1435–1443.
Heinemann, U., Beck, H., Dreier, J.P., Ficker, E., Stabel, J. and
Zhang, C.L. (1992) The dentate gyrus as a regulated gate for
the propagation of epileptiform activity. Epilepsy Res. Sup-
pl., 7: 273–280.
Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis,
E.F. (2000) Astrocytes generate isoprostanes in response to
trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420.
Hogg, S., Perron, C., Barneoud, P., Sanger, D.J. and Moser,
P.C. (1998a) Neuroprotective effect of eliprodil: attenuation
of a conditioned freezing deficit induced by traumatic injury
of the right parietal cortex in the rat. J. Neurotrauma, 15(7):
545–553.
Hogg, S., Sanger, D.J. and Moser, P.C. (1998b) Mild traumatic
lesion of the right parietal cortex in the rat: characterisation
of a conditioned freezing deficit and its reversal by dizocil-
pine. Behav. Brain Res., 93(1–2): 157–165.
Joliot, M., Ribary, U. and Llinas, R. (1994) Human oscillatory
brain activity near 40 Hz coexists with cognitive temporal
binding. Proc. Natl. Acad. Sci. U.S.A., 91(24): 11748–11751.
Kao, C.Q., Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004)
Potentiation of GABA(A) currents after mechanical injury of
cortical neurons. J. Neurotrauma, 21(3): 259–270.
Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A.
(1990) Massive increases in extracellular potassium and the
indiscriminate release of glutamate following concussive
brain injury. J. Neurosurg., 73(6): 889–900.
Kotapka, M.J., Gennarelli, T.A., Graham, D.I., Adams, J.H.,
Thibault, L.E., Ross, D.T. and Ford, I. (1991) Selective vul-
nerability of hippocampal neurons in acceleration-induced
experimental head injury. J. Neurotrauma, 8(4): 247–258.

Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli,
T.A. (1992) Hippocampal pathology in fatal non-missile hu-
man head injury. Acta Neuropathol. (Berl.), 83(5): 530–534.
Kotapka, M.J., Graham, D.I., Adams, J.H., Doyle, D. and
Gennarelli, T.A. (1993) Hippocampal damage in fatal pae-
diatric head injury. Neuropathol. Appl. Neurobiol., 19(2):
128–133.
Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli,
T.A. (1994) Hippocampal pathology in fatal human head
injury without high intracranial pressure. J. Neurotrauma,
11(3): 317–324.
Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A.
and Ellis, E.F. (1997) Alterations in phosphatidylcholine
metabolism of stretch-injured cultured rat astrocytes. J.
Neurochem., 68(5): 1904–1910.
LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II,
R.S. (2005) High rate shear strain of three-dimensional neu-
ral cell cultures: a new in vitro traumatic brain injury model.
J. Biomech., 38(5): 1093–1105.
LaPlaca, M.C. and Thibault, L.E. (1998) Dynamic mechanical
deformation of neurons triggers an acute calcium response
and cell injury involving the N-methyl-D-aspartate glutamate
receptor. J. Neurosci. Res., 52(2): 220–229.
Lea, P.M., Custer, S.J., Vicini, S. and Faden, A.I. (2002)
Neuronal and glial mGluR5 modulation prevents stretch-
induced enhancement of NMDA receptor current. Pharmacol.
Biochem. Behav., 73(2): 287–298.
Lea, P.M.T., Custer, S.J., Stoica, B.A. and Faden, A.I. (2003)
Modulation of stretch-induced enhancement of neuronal
NMDA receptor current by mGluR1 depends upon presence
of glia. J. Neurotrauma, 20(11): 1233–1249.
Leuner, B., Falduto, J. and Shors, T.J. (2003) Associative
memory formation increases the observation of dendritic
spines in the hippocampus. J Neurosci, 23(2): 659–665.
Lisman, J., Malenka, R.C., Nicoll, R.A. and Malinow, R.
(1997) Learning mechanisms: the case for CaM-KII. Science,
276(5321): 2001–2002.
Liu, L., Wong, T.P., Pozza, M.F., Lingenhoehl, K., Wang, Y.,
Sheng, M., Auberson, Y.P. and Wang, Y.T. (2004) Role of
NMDA receptor subtypes in governing the direction of hippo-
campal synaptic plasticity. Science, 304(5673): 1021–1024.
Liu, P. and Bilkey, D.K. (2001) The effect of excitotoxic lesions
centered on the hippocampus or perirhinal cortex in object
recognition and spatial memory tasks. Behav. Neurosci.,
115(1): 94–111.
Lothman, E.W., Stringer, J.L. and Bertram, E.H. (1992) The
dentate gyrus as a control point for seizures in the hippo-
campus and beyond. Epilepsy Res. Suppl., 7: 301–313.
Lowenstein, D.H., Thomas, M.J., Smith, D.H. and McIntosh,
T.K. (1992) Selective vulnerability of dentate hilar neurons
following traumatic brain injury: a potential mechanistic link
between head trauma and disorders of the hippocampus.
J. Neurosci., 12(12): 4846–4853.
Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips,
L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990)
Prolonged memory impairment in the absence of hippocampal
167
cell death following traumatic brain injury in the rat. Brain
Res., 526(2): 249–258.
Lytton, W.W. and Sejnowski, T.J. (1991) Simulations of
cortical pyramidal neurons synchronized by inhibitory inter-
neurons. J. Neurophysiol., 66(3): 1059–1079.
Malenka, R.C. and Bear, M.F. (2004) LTP and LTD: an
embarrassment of riches. Neuron, 44(1): 5–21.
Malenka, R.C., Kauer, J.A., Perkel, D.J. and Nicoll, R.A.
(1989) The impact of postsynaptic calcium on synaptic
transmission — its role in long-term potentiation. Trends
Neurosci., 12(11): 444–450.
Malenka, R.C. and Nicoll, R.A. (1997) Silent synapses speak
up. Neuron, 19(3): 473–476.
Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation
— a decade of progress? Science, 285(5435): 1870–1874.
Malinow, R., Mainen, Z.F. and Hayashi, Y. (2000) LTP mecha-
nisms: from silence to four-lane traffic. Curr. Opin.
Neurobiol., 10(3): 352–357.
Malinow, R. and Malenka, R.C. (2002) AMPA receptor
trafficking and synaptic plasticity. Annu. Rev. Neurosci.,
25: 103–126.
Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990)
Physical model simulations of brain injury in the primate.
J. Biomech., 23(8): 823–836.
Martin, A.R. and Pilar, G. (1964) Presynaptic and post-synap-
tic events during post-tetanic potentiation and facilitation in
the avian ciliary ganglion. J. Physiol., 175: 17–30.
Mathias, J.L., Bigler, E.D., Jones, N.R., Bowden, S.C., Barrett-
Woodbridge, M., Brown, G.C. and Taylor, D.J. (2004)
Neuropsychological and information processing performance
and its relationship to white matter changes following
moderate and severe traumatic brain injury: a preliminary
study. Appl. Neuropsychol., 11(3): 134–152.
Mayer, M.L., Westbrook, G.L. and Guthrie, P.B. (1984)
Voltage-dependent block by Mg2+ of NMDA responses in
spinal cord neurones. Nature, 309(5965): 261–263.
McAllister, T.W. (1992) Neuropsychiatric sequelae of head in-
juries. Psychiatr. Clin. North Am., 15(2): 395–413.
McAllister, T.W., Saykin, A.J., Flashman, L.A., Sparling,
M.B., Johnson, S.C., Guerin, S.J., Mamourian, A.C.,
Weaver, J.B. and Yanofsky, N. (1999) Brain activation dur-
ing working memory 1 month after mild traumatic brain in-
jury: a functional MRI study. Neurology, 53(6): 1300–1308.
McBain, C.J. and Mayer, M.L. (1994) N-methyl-D-aspartic acid
receptor structure and function. Physiol. Rev., 74(3):
723–760.
McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak,
S., Soares, H. and Faden, A.L. (1989) Traumatic brain injury
in the rat: characterization of a lateral fluid-percussion
model. Neuroscience, 28(1): 233–244.
McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F.
(1996) Stretch-induced injury of cultured neuronal, glial, and
endothelial cells. Effect of polyethylene glycol-conjugated
superoxide dismutase. Stroke, 27(5): 934–940.
Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller,
R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical
168
analysis of experimental diffuse axonal injury. J. Neuro-
trauma, 12(4): 689–694.
Miller, L.A., Lai, R. and Munoz, D.G. (1998) Contributions of
the entorhinal cortex, amygdala and hippocampus to human
memory. Neuropsychologia, 36(11): 1247–1256.
Miller, L.P., Lyeth, B.G., Jenkins, L.W., Oleniak, L., Panchision,
D., Hamm, R.J., Phillips, L.L., Dixon, C.E., Clifton, G.L.
and Hayes, R.L. (1990) Excitatory amino acid receptor sub-
type binding following traumatic brain injury. Brain Res.,
526(1): 103–107.
Miyazaki, S., Katayama, Y., Lyeth, B.G., Jenkins, L.W.,
DeWitt, D.S., Goldberg, S.J., Newlon, P.G. and Hayes, R.L.
(1992) Enduring suppression of hippocampal long-term
potentiation following traumatic brain injury in rat. Brain
Res., 585(1–2): 335–339.
Mody, I., De Koninck, Y., Otis, T.S. and Soltesz, I. (1994)
Bridging the cleft at GABA synapses in the brain. Trends
Neurosci., 17(12): 517–525.
Morrison, III, B., Cater, H.L., Benham, C.D. and Sundstrom,
L.E. (2006) An in vitro model of traumatic brain injury uti-
lising two-dimensional stretch of organotypic hippocampal
slice cultures. J. Neurosci. Methods, 150(2): 192–201.
Morrison, III, B., Meaney, D.F. and McIntosh, T.K. (1998)
Mechanical characterization of an in vitro device designed to
quantitatively injure living brain tissue. Ann. Biomed. Eng.,
26(3): 381–390.
O’Dell, D.M. and Hamm, R.J. (1995) Chronic postinjury
administration of MDL 26,479 (Suritozole), a negative
modulator at the GABAA receptor, and cognitive impair-
ment in rats following traumatic brain injury. J. Neurosurg.,
83(5): 878–883.
Osteen, C.L., Giza, C.C. and Hovda, D.A. (2004) Injury-
induced alterations in N-methyl-D-aspartate receptor subunit
composition contribute to prolonged 45 calcium accumula-
tion following lateral fluid percussion. Neuroscience, 128(2):
305–322.
Pickard, L., Noel, J., Duckworth, J.K., Fitzjohn, S.M., Henley,
J.M., Collingridge, G.L. and Molnar, E. (2001) Transient
synaptic activation of NMDA receptors leads to the insertion
of native AMPA receptors at hippocampal neuronal plasma
membranes. Neuropharmacology, 41(6): 700–713.
Pike, B.R. (2001) In vitro models of traumatic brain injury.
In: Miller L.P. and Hayes R.L. (Eds.), Head Trauma: Ba-
sic, Preclinical and Clinical Directions. Wiley, New York,
pp. 37–64.
Pittaluga, A., Bonfanti, A. and Raiteri, M. (2000) Somatostatin
potentiates NMDA receptor function via activation of
InsP(3) receptors and PKC leading to removal of the
Mg(2+) block without depolarization. Br. J. Pharmacol.,
130(3): 557–566.
Reeves, T.M., Kao, C.Q., Phillips, L.L., Bullock, M.R. and
Povlishock, J.T. (2000) Presynaptic excitability changes fol-
lowing traumatic brain injury in the rat. J. Neurosci. Res.,
60(3): 370–379.
Reeves, T.M., Lyeth, B.G. and Povlishock, J.T. (1995) Long-
term potentiation deficits and excitability changes following
traumatic brain injury. Exp. Brain Res., 106(2): 248–256.
Reeves, T.M., Lyeth, B.G., Phillips, L.L., Hamm, R.J. and
Povlishock, J.T. (1997a) The effects of traumatic brain injury
on inhibition in the hippocampus and dentate gyrus. Brain
Res., 757(1): 119–132.
Reeves, T.M., Zhu, J., Povlishock, J.T. and Phillips, L.L.
(1997b) The effect of combined fluid percussion and ento-
rhinal cortical lesions on long-term potentiation. Neurosci-
ence, 77(2): 431–444.
Rempel-Clower, N.L., Zola, S.M., Squire, L.R. and Amaral,
D.G. (1996) Three cases of enduring memory impairment
after bilateral damage limited to the hippocampal formation.
J. Neurosci., 16(16): 5233–5255.
Ross, S.T. and Soltesz, I. (2000) Selective depolarization of in-
terneurons in the early posttraumatic dentate gyrus: involve-
ment of the Na(+)/K(+)-ATPase. J. Neurophysiol., 83(5):
2916–2930.
Rzigalinski, B.A., Liang, S., McKinney, J.S., Willoughby, K.A.
and Ellis, E.F. (1997) Effect of Ca2+ on in vitro astrocyte
injury. J. Neurochem., 68(1): 289–296.
Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis,
E.F. (1998a) Intracellular free calcium dynamics in stretch-
injured astrocytes. J. Neurochem., 70(6): 2377–2385.
Rzigalinski, B.A., Weber, J.T., Willoughby, K.A., L.S., Wood-
ward, J.J. and Ellis, E.F. (1998b). Glutamate release and en-
hanced receptor sensitivity in stretch-injured neurons. J
Neurotrauma, 15: 894.
Rzigalinski, B.A., Willoughby, K.A., Hoffman, S.W., Falck,
J.R. and Ellis, E.F. (1999) Calcium influx factor, further ev-
idence it is 5,6-epoxyeicosatrienoic acid. J. Biol. Chem.,
274(1): 175–182.
Salmond, C.H. and Sahakian, B.J. (2005) Cognitive outcome in
traumatic brain injury survivors. Curr. Opin. Crit. Care, 11:
111–116.
Sanders, M.J., Sick, T.J., Perez-Pinzon, M.A., Dietrich, W.D.
and Green, E.J. (2000) Chronic failure in the maintenance of
long-term potentiation following fluid percussion injury in
the rat. Brain Res., 861(1): 69–76.
Santhakumar, V., Bender, R., Frotscher, M., Ross, S.T.,
Hollrigel, G.S., Toth, Z. and Soltesz, I. (2000) Granule cell
hyperexcitability in the early post-traumatic rat dentate
gyrus: the ‘irritable mossy cell’ hypothesis. J. Physiol.,
524(Pt 1): 117–134.
Santhakumar, V., Ratzliff, A.D., Jeng, J., Toth, Z. and Soltesz,
I. (2001) Long-term hyperexcitability in the hippocampus af-
ter experimental head trauma. Ann. Neurol., 50(6): 708–717.
Schwarzbach, E., Bonislawski, D.P., Xiong, G. and Cohen,
A.S. (2006) Mechanisms underlying the inability to induce
area CA1 LTP in the mouse after traumatic brain injury.
Hippocampus, 16(6): 541–550.
Segal, M. (2005) Dendritic spines and long-term plasticity. Nat.
Rev. Neurosci., 6(4): 277–284.
Sick, T.J., Perez-Pinzon, M.A. and Feng, Z.Z. (1998) Impaired
expression of long-term potentiation in hippocampal slices 4
and 48 h following mild fluid-percussion brain injury in vivo.
Brain Res., 785(2): 287–292.
Sihver, S., Marklund, N., Hillered, L., Langstrom, B., Wata-
nabe, Y. and Bergstrom, M. (2001) Changes in mACh,

NMDA and GABA(A) receptor binding after lateral fluid-
percussion injury: in vitro autoradiography of rat brain
frozen sections. J. Neurochem., 78(3): 417–423.
Singer, W. and Gray, C.M. (1995) Visual feature integration
and the temporal correlation hypothesis. Annu. Rev. Neuro-
sci., 18: 555–586.
Sloviter, R.S. (1994) The functional organization of the hippo-
campal dentate gyrus and its relevance to the pathogenesis of
temporal lobe epilepsy. Ann. Neurol., 35(6): 640–654.
Smith, D.H., Okiyama, K., Thomas, M.J., Claussen, B. and
McIntosh, T.K. (1991) Evaluation of memory dysfunction
following experimental brain injury using the Morris water
maze. J. Neurotrauma, 8(4): 259–269.
Song, I. and Huganir, R.L. (2002) Regulation of AMPA
receptors during synaptic plasticity. Trends Neurosci.,
25(11): 578–588.
Swope, S.L., Moss, S.J., Raymond, L.A. and Huganir, R.L.
(1999) Regulation of ligand-gated ion channels by protein
phosphorylation. Adv. Second Messenger Phosphoprotein
Res., 33: 49–78.
Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical
perturbation of cultured cortical neurons reveals a stretch-
induced delayed depolarization. J. Neurophysiol., 74(6):
2767–2773.
Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1997) Inhibition of
the electrogenic Na pump underlies delayed depolarization
of cortical neurons after mechanical injury or glutamate.
J. Neurophysiol., 77(2): 632–638.
Tecoma, E.S., Monyer, H., Goldberg, M.P. and Choi, D.W.
(1989) Traumatic neuronal injury in vitro is attenuated by
NMDA antagonists. Neuron, 2(6): 1541–1545.
169
Thompson, H.J., Lifshitz, J., Marklund, N., Grady, M.S.,
Graham, D.I., Hovda, D.A. and McIntosh, T.K. (2005) Lat-
eral fluid percussion brain injury: a 15-year review and eva-
luation. J. Neurotrauma, 22(1): 42–75.
Toth, Z., Hollrigel, G.S., Gorcs, T. and Soltesz, I. (1997)
Instantaneous perturbation of dentate interneuronal net-
works by a pressure wave-transient delivered to the neocor-
tex. J. Neurosci., 17(21): 8106–8117.
Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F.
and Ellis, E.F. (1999) Alterations in calcium-mediated signal
transduction after traumatic injury of cortical neurons. Cell
Calcium, 26(6): 289–299.
Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2001) Traumatic
injury of cortical neurons causes changes in intracellular
calcium stores and capacitative calcium influx. J. Biol.
Chem., 276(3): 1800–1807.
Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2002) Calcium
responses to caffeine and muscarinic receptor agonists are
altered in traumatically injured neurons. J. Neurotrauma,
19(11): 1433–1443.
Witgen, B.M., Lifshitz, J., Smith, M.L., Schwarzbach, E.,
Liang, S.L., Grady, M.S. and Cohen, A.S. (2005) Regional
hippocampal alteration associated with cognitive deficit
following experimental brain injury: a systems, network
and cellular evaluation. Neuroscience, 133(1): 1–15.
Wong, R.K., Traub, R.D. and Miles, R. (1986) Cellular basis of
neuronal synchrony in epilepsy. Adv. Neurol., 44: 583–592.
Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996)
Reduction of voltage-dependent Mg2+ blockade of NMDA
current in mechanically injured neurons. Science, 274(5294):
1921–1923.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 11

Traumatic injury of the spinal cord and nitric oxide

Jozef Maršala, Judita Orendáčová, Nadežda Lukáčová and Ivo Vanický

Institute of Neurobiology, Slovak Academy of Sciences, Košice, Slovak Republic

Abstract: In the current report, we summarize our findings related to the involvement of nitric oxide (NO)
in the pathology of spinal cord trauma. We initially studied the distribution of nitric oxide synthase (NOS)-
immunolabeled and/or nicotinamide adenine dinucleotide phosphate diaphorase (NADPHd; which is
highly colocalized with NOS)-stained somata and fibers in the spinal cord of the rabbit. Segmental and
laminar distribution of NADPHd-stained neurons in the rabbit revealed a large number of NADPHd-
stained neurons in the spinal cord falling into six categories, N1–N6, while others could not be classified.
Large numbers of NADPHd-stained neurons were identified in the superficial dorsal horn and around the
central canal. Four morphologically distinct kinds of NADPHd-stained axons 2.5–3.5 mm in diameter were
identified throughout the white matter in the spinal cord. Moreover, a massive occurrence of axonal
NADPHd-staining was detected in the juxtagriseal layer of the ventral funiculus along the rostrocaudal
axis. The prominent NADPHd-stained fiber bundles were identified in the mediobasal and central portion
of the ventral funiculus. The sulcomarginal fasciculus was found in the basal and medial portion of the
ventral funiculus in all cervical and thoracic segments. Since the discovery that NO may act as a neuronal
transmitter, an increasing interest has focused on its ability to modulate synaptic function. NO passes
through cell membranes without specific release or uptake mechanisms inducing changes in signal-related
functions by several means. In particular, the activation of the soluble guanylyl cyclases (sGC), the for-
mation of cyclic guanosine 30 ,50 -monophosphate (cGMP) and the action of cGMP-dependent protein
kinases has been identified as the main signal transduction pathways of NO in the nervous system including
spinal cord. It is known that the intracellular level of cGMP is strictly controlled by its rate of synthesis via
guanylyl cyclases (GC) and/or by the rate of its degradation via 30 ,50 -cyclic nucleotide phosphodiesterases
(PDE). GC can be divided into two main groups, i.e., the membrane-bound or particular guanylyl cyclase
(pGC) and the cytosolic or sGC. In the spinal cord, the activation of pGC has only been demonstrated for
natriuretic peptides, which stimulate cGMP accumulation in GABA-ergic structures in laminae I–III of the
rat cervical spinal cord. These neurons are involved in controlling the action of the locomotor circuit. In
view of the abundance of NO-responsive structures in the brain, it is proposed that NO–cGMP signaling
will be part of neuronal information processing at many levels. In relation to this, we found that surgically
induced Th7 constriction of 24 h duration stimulated both the constitutive NOS activity and cGMP level by
120 and 131%, respectively, in non-compartmentalized white matter of Th8–Th9 segments, located just
caudally to the site of injury. NO-mediated cGMP formation was only slightly increased in the dorsal
funiculus of Th5–Th9 segments. There are some other sources that may influence the NO-mediated cGMP
formation in spinal cord. A high level of glutamate produced at the site of the lesion and an excessive
accumulation of intracellular Ca2+ may stimulate NOS activity and create suitable conditions for NO
synthesis and its adverse effect on white matter. An increased interest has focused on the role of NO at the

Corresponding author. Tel.: +421 55 6785 062; Fax: +421 55

6785 074; E-mail: marsala@saske.sk

DOI: 10.1016/S0079-6123(06)61011-X 171

172

site of injury and in areas located close to the epicenter of the impact site and, in these connections an
upregulation of NOS was noted in neurons and interneurons. However, the upregulation of NOS expres-
sion was also seen in interneurons located just rostrally and caudally to the lesion. A quantitative analysis
of laminar distribution of multiple cauda equina constriction (MCEC) induced NADPHd-stained neurons
revealed a considerable increase in these neurons in laminae VIII–IX 8 h postconstriction, and a highly
statistically significant increase of such neurons in laminae VII–X 5 days postconstriction in the
lumbosacral segments. Concurrently, the number of NADPHd-stained neurons on laminae I–II in LS
segments was greatly reduced. It is concluded that a greater understanding of NO changes after spinal cord
trauma is essential for the possibility of targeting this pathway therapeutically.

Keywords: spinal cord; trauma; cauda equina constriction; nitric oxide

The occurrence of NADPH diaphorase-stained and/ containing widely branching NADPHd-stained

or bNOS immunoreactive neurons in the gray
matter of the spinal cord
Small, morphologically heterogeneous popula-
tions of neurons containing nicotinamide adenine
dinucleotide phosphate diaphorase (NADPHd),
and/or brain nitric oxide synthase (bNOS), an en-
zyme known to generate nitric oxide (NO), have
been identified at various sites in the central ner-
vous system (CNS) in a number of mammalian
species including the human brain (Kowall et al.,
1985, 1987; Kowall and Mueller, 1988; Mizukawa
et al., 1989; Bredt et al., 1990; Vincent and
Kimura, 1992; Egberongbe et al., 1994; Vincent,
1994). Immunohistochemistry of bNOS revealed
that the occurrence of this enzyme is almost com-
pletely homotopic with the localization of neurons
stained for NADPHd (Lukáčová et al., 1999).
It has long been known that NADPHd-stained
neurons are present in the spinal cord, especially in
the substantia gelatinosa, lamina X, and the inter-
mediolateral cell column (Thomas and Pearse,
1964; Mizukawa et al., 1989; Blottner and
Baumgarten, 1992; Bredt and Snyder, 1992). How-
ever, the analysis of NADPHd-stained and/or
bNOS-immunoreactive neurons in the spinal cord
of different animal species (Anderson, 1992;
Valtschanoff et al., 1992; Dun et al., 1993; Vizzard
et al., 1994a, 1995) revealed a morphologically
heterogeneous pattern of NADPHd-stained neu-
ronal pools ranging from bipolar, poorly branched
NADPHd-stained neurons in the superficial dor-
sal horn to highly differentiated neurons in the
pericentral region (lamina X), deep dorsal horn
(laminae IV–V), and dorsal gray commissure
neurons (Anderson, 1992; Valtschanoff et al.,
1992; Dun et al., 1993; Lee et al., 1993; Saito
et al., 1994; Vizzard et al., 1994b; Wetts and
Vaughn, 1994; Burnett et al., 1995; Maršala et al.,
1997, 1998).
A study from our laboratory aimed at segmental
and laminar distribution of NADPHd-stained
neurons in the rabbit revealed a large number of
NADPHd-positive neurons in the spinal cord fall-
ing into six categories, N1–N6, while others could
not be classified (Maršala et al., 1999). Large
numbers of NADPHd-stained neurons were iden-
tified in the superficial dorsal horn and around the
central canal at all spinal levels and in the inter-
mediolateral cell column at thoracic and upper
lumbar levels. NADPHd-stained somata of the
pericentral region were divided into a thin sub-
ependymal cell column containing longitudinally
arranged, small bipolar neurons with processes
penetrating deeply into the intermediolateral cell
column and/or running rostrocaudally in the sub-
ependymal layer. The second pericentral cell col-
umn located more laterally in lamina X contains
NADPHd-stained neurons with long dendrites ra-
diating in the transverse plane. In the pericentral
region (lamina X), close association of NADPHd-
stained somata and fibers, and mostly longitudi-
nally oriented blood vessels was detected. Somata
of the sacral parasympathetic nucleus seen in seg-
ments S1–S3, exhibited prominent NADPHd
staining accompanied by heavily stained fibers ex-
tending from Lissauer’s tract through lamina I
along the lateral edge of the dorsal horn to lamina
V. A massive dorsal gray commissure, highly posi-
tive in NADPHd staining, was found in segments

Sl–S3. Scattered positive cells were also found in
the deeper dorsal horn, ventral horn, and white
matter. Fiber-like NADPHd staining was found in
the superficial dorsal horn and pericentral region
in all cervical, thoracic, and lumbosacral segments.
Dense, punctate, non-somatic NADPHd staining
was detected in the superficial dorsal horn, in the
pericentral region all along the rostrocaudal axis
and in the phrenic nucleus (segments C4–C5),
dorsal nucleus (segments Th2–L2), Onuf’s nucleus
(segments S1–S3), and the dorsal part of the dorsal
gray commissure (segments S1–S3; Maršala et al.,
1999).
The distribution of NADPH diaphorase-positive
and/or NOS immunolabeled fibers in spinal cord
white matter
The funicular distribution of NADPHd-stained
axons was examined in the white matter of the
rabbit spinal cord by using horizontal, para-
sagittal, and transverse sections. Four morphologi-
cally distinct kinds of NADPHd-stained axons
2.5–3.5 mm in diameter were found in the sulco-
marginal fasciculus as a part of the ventral funi-
culus in the cervical and upper thoracic segments
and in the long propriospinal bundle of the ventral
funiculus in Th3–L3 segments. Varicose NADPHd-
stained axons of the sympathetic preganglionic
neurons, characterized by widely spaced varicosi-
ties, were found in the ventral funiculus of Th2–L3
segments. A third kind of NADPHd-stained ultra-
fine axons, 0.3–0.5 mm in diameter with numerous
varicosities mostly spherical in shape, was identified
in large number within Lissauer’s tract. The last
group of NADPHd-stained axons, 1.0–1.5 mm in
diameter, occurred in the Lissauer tract. Most of
these axons were traceable for considerable dis-
tances and generating spherical and elliptical forms
of varicosities.
The majority of NADPHd-stained axons iden-
tified in the cuneate and gracile fascicles were
concentrated in the deep portion of the
dorsal funiculus. An extremely reduced number
of NADPHd-stained axons, confirmed by a com-
puter-assisted image-processing system were found
in the dorsal half of the gracile fascicle. Axonal
173
NADPHd-staining could not be detected in the
lateral funiculus consistent with the location of the
dorsal spinocerebellar tract. Numerous, mostly
thin NADPHd-stained axonal profiles could be
detected in the dorsolateral funiculus in all seg-
ments studied and in juxtagriseal portion of the
lateral funiculus in the extent of the cervical and
lumbar enlargement. A massive occurrence of axo-
nal NADPHd-staining was detected in the
juxtagriseal layer of the ventral funiculus along
the rostrocaudal axis of the spinal cord. The
prominent NADPHd-stained bundles containing
smooth, thick, non-varicose axons were identified
in the mediobasal and central portion of the ven-
tral funiculus. First, the sulcomarginal fasciculus
was found in the basal and medial portion of the
ventral funiculus in all cervical and upper thoracic
segments. Second, more caudally, a long proprio-
spinal bundle displaying prominent NADPHd-
staining was localized in the central portion of the
ventral funiculus throughout the Th3–L3 segments
(Maršala et al., 2003).
NO/cGMP signaling in the spinal cord
For a long time it was believed that interneuronal
communication was realized exclusively via synap-
tic contacts. However, since the discovery of NO
as a neuronal transmitter (Snyder and Bredt, 1991;
Dawson et al., 1992), an increasing interest has
focused on its ability to modulate synaptic func-
tions. NO passes through cell membranes without
specific release or uptake mechanisms and acts di-
rectly on surrounding neural tissue in extended
spatial limits of approximately 300–400 mm (Wood
and Garthwaite, 1994; Vincent, 1994). After its
synthesis in the soma of neurons, NO induces
changes in signal-related functions in several ways.
In particular, the activation of soluble guanylyl
cyclases (sGC), the formation of cyclic guanosine
30 ,50 -monophosphate (cGMP), and the action of
cGMP-dependent protein kinases have been iden-
tified as the main signal transduction pathway of
NO in the nervous system (Bredt and Snyder,
1992; Wang and Robinson, 1997; Smolenski et al.,
1998).
174
It is known that the intracellular level of cGMP
is strictly controlled by its rate of synthesis via GCs
and/or by the rate of its degradation via 30 ,50 -cyclic
nucleotide phosphodiesterases (PDEs; Beavo,
1995; Lucas et al., 2000; Francis et al., 2001). GC
can be divided into two main groups, i.e., the
membrane-bound or pGC and the cytosolic or
sGC (Schmidt, 1992; Murad, 1994; Wedel and
Garbers, 1997). PDEs comprise a large group of
enzymes that hydrolyze cGMP to its inactive 50 -
derivate. PDE families (PDE1–11) and a large
number of their isoforms are characterized with
distinct localization pattern in the brain. However,
only three PDE families, i.e., PDE5, PDE9, and a
photoreceptor-specific PDE6, identified in retina,
specifically use cGMP as a substrate (Beavo, 1995;
van Staveren et al., 2003). cGMP can be degraded
by dual-substrate PDEs (PDE1, PDE2, PDE10,
and PDE11) as well.
pGC is a large transmembrane molecule that has
a receptor domain at the outside and the cata-
lytic site at the inside of the cell (Vles et al., 2000).
This enzyme is activated through interaction of
peptide hormones with the receptor domain. In the
spinal cord, the activation of pGC has only been
demonstrated for atrial natriuretic peptide (ANP),
which stimulates cGMP accumulation in GABA-
ergic structures of laminae I–III of the rat cervical
spinal cord (Vles et al., 2000); these neurons are
involved in controlling the action of the spinal loco-
motor circuit. The GABA-ergic agonist, Baclofen,
a well-known drug in the treatment of spasticity,
has been shown to express an effect on ANP-
mediated cGMP synthesis in the superficial dorsal
horn layers of the cervical spinal cord (de Louw
et al., 2002). These data propose that the clinical
effect of Baclofen on the reduction of spasticity
may be involved in the triggering of a phosphory-
lation/dephosphorylation cycle of the intracellular
part of the ANP receptor and, subsequently, a
downregulation of the pGC activity (Potter and
Hunter, 2001).
sGC is a heterodimeric enzyme composed of
four different subunits (a1, a2, b1, b2; Russwurm
et al., 1998); however, all catalytically active GC
isoforms are built of the b1 subunit and either the
a1 or the a2 subunit (Gupta et al., 1997; Koesling
and Friebe, 2000; Koglin et al., 2001). Activation
of different subunits of sGC is associated with the
binding of NO to a heme group, leading to for-
mation of a nitrosyl–heme complex and conse-
quent conformational change (Humbert et al.,
1990). The activation of sGC and the formation of
cGMP have been identified as the main NO/cGMP
signal transduction pathway in the CNS (Bredt
and Snyder, 1992; Wang and Robinson, 1997;
Smolenski et al., 1998). In view of the abundance
of NO-responsive structures in the brain, it is pro-
posed that NO-cGMP signaling will be part of
neuronal information processing at many levels.
To date, little is known about the role of NO-
cGMP signaling in the spinal cord. Morris et al.
(1994) showed that the NO-cGMP pathway is
located primarily in the superficial layers of the
dorsal horn in the neonatal rat spinal cord. The
NO-mediated cGMP response in immature spinal
cord was also confirmed by Vles et al. (2000). The
increased level of cGMP in the dorsal horn of the
spinal cord seems to be associated with hyper-
algesia through activation of cGMP dependent
protein kinase Ia (Aley et al., 1998; Tao et al.,
2000). NO-mediated cGMP was also identified in
parvalbumin-immunoreactive (GABA-ergic) neu-
rons and in the axon terminals of the ventral horn
probably apposing the motor neurons (Vles et al.,
2000).
NO-mediated cGMP synthesis in the spinal cord
after traumatic injury
Recent data from our laboratory have shown that
surgically induced Th7 constriction of 24 h dura-
tion stimulated both the constitutive NOS activity
and cGMP level by 120 and 131%, respectively, in
non-compartmentalized white matter of Th8–Th9
segments, located just caudally to the site of injury
(Lukáčová et al., 2002). NO-mediated cGMP for-
mation was only slightly increased in the dorsal
funiculus of Th5–Th9 segments, composed entirely
of long ascending axons. However, a strong in-
crease of cGMP formation was noted in the lateral
funiculus of segments located below the site of Th7
constriction. These results suggest a decrease of
the axoplasmic transport in ascending NOS-
immunoreactive (NOS-IR) axons, which clearly

prevail in the lateral funiculus in comparison with
descending axons. The NOS-IR neurons of the in-
termediolateral cell column sending many NOS-IR
dendrites into the lateral funiculus (Maršala
et al., 1998, 1999), together with the NOS-IR ax-
ons of the lateral spinothalamic tract may, upon
midthoracic constriction, provide the substrate for
NO/cGMP formation. A significant decrease in the
level of cGMP found in the ventral funiculus of
Th5–Th6 segments correlates with a significant de-
crease of both the constitutive NOS activity, at-
tributable to a slight constriction of the superficial
portion of the ventral funiculus at Th7 level, and
the axoplasmic transport of the NOS affecting
thick, highly NOS-IR axons (Lukáčová et al.,
2000). These axons form an ascending and, to a
lesser extent, a descending propriospinal bundle,
connecting NOS-IR neurons in lumbosacral en-
largement with the ventral motor nucleus at
cervicothoracic level (Sherrington and Laslett,
1903; Barilari and Kuypers, 1969; Yezierski
et al., 1980; Maršala et al., 2004; Lukáčová
et al., 2006). The detection of a considerably higher
radioassay-assessed NOS activity in the ventral
funiculus of the lumbosacral segments, in compar-
ison to cervical segments, supports the conclusions
based on immunocytochemical and histochemical
studies (Lukáčová et al., 1999; Lukáčová and
Pavel, 2000). There are some other sources that
may influence the NO-mediated cGMP formation
in spinal cord injury (SCI). A high level of gluta-
mate produced at the site of the lesion in the
spinal cord (Mc Adoo et al., 1997) and an excessive
accumulation of intracellular Ca2+ (Lehotský
et al., 1999) may stimulate NOS activity and
create suitable conditions for NO synthesis and
its adverse effect on white matter. An extreme
production of NO and superoxide, appearing
early after SCI in its close vicinity, may stimulate
the formation of peroxynitrite (ONOO
) that
in turn may break up, thus giving nitrogen
dioxide and a compound having a hydroxyl-like
reactivity leading to oxidation of protein, DNA,
and membrane-damaging lipid peroxidation
(Lukáčová et al., 1994). Furthermore, the release
of free fatty acids, particularly arachidonic acid, in
SCI (Horrocks et al., 1984; Halát et al., 1987;
Nakano et al., 1990) may lead to an overexpression
175
of NOS accompanied by an increase in cellular
cGMP (Strosznajder and Chalimoniuk, 1996;
Toborek et al., 2000). Morton and Bredt (1998)
have suggested that in primary cell cultures
taken from the cerebellum of nNOS knockout
mice, norepinephrine is able to stimulate cGMP
formation independently from sGC. In addition,
a selective inhibitor of sGC points to a possibility
that extracellular cGMP released under basal
conditions in the brain might come from both
NO-sensitive and NO-insensitive GCs (Vallebuona
and Raiteri, 1993; Luo et al., 1994; Fedele et al.,
1996). The elevation of cGMP levels could be
elicited by atrial natriuretic factor, possibly
localized in astroglial cells (de Vente et al., 1990).
These non-neuronal elements form dense perinodal
processes approaching the node of Ranvier
(Ransom et al., 1993) and, as such, are located
at sites known for a high rate of transmembrane
ion traffic between the axolemma and the extra-
cellular space (Stys et al., 1991; Stys and Steffensen,
1996). Based on biochemical and immunocyto-
chemical experiments we can conclude that the
NO/cGMP signaling pathway is implicated in the
pathology of SCI.
NOS in the spinal cord: effect of traumatic injury in
the epicenter of injury
Spinal cord neurons are susceptible to toxic effects
of excitatory amino acids because of the excitatory
amino acid glutamate, which is markedly released
from glutamatergic neurons to the extracellular
space, stimulates N-methyl-D-aspartate (NMDA)
receptors located on nitrergic neurons, and con-
secutively enhances the release of NO (Prast and
Philippu, 2001). Both NO and glutamate, when
overstimulated, can initiate a neurotoxic cascade.
An increasing interest has focused on the role of
NO at the site of SCI and in areas located close to
the epicenter of the impact site.
Acute changes in NO production after SCI have
been studied by Liu et al. (2000), demonstrating a
rapid increase of NO immediately after injury fol-
lowed by a rapid decline during approximately the
first 60 min after injury. A microdialysis study by
Nakahara et al. (2002) has also demonstrated a
176
second wave of NO production after SCI in the
rat — the second wave was observed 24 h and 3
days after injury. Long-term changes in the pro-
duction of parenchymal NO after SCI have been
studied by detecting activities of NOS at various
posttraumatic periods. Recent data obtained by
Trudrung et al. (2000) have shown the upregula-
tion of NOS expression in neurons adjacent to the
lesion on both sides of a complete transverse lesion
of the spinal cord. Yick et al. (1998) have described
the increase of nNOS expression in Clarke’s col-
umn 3 days after hemisection, reaching its maxi-
mum 20 days after injury. The upregulation of
NOS expression was also seen in interneurons lo-
cated rostrally and caudally to the lesion (Wu
et al., 1994; Sharma et al., 2000; Lukáčová et al.,
2005). Surgically induced Th7 constriction of 24 h
duration caused an increase of calcium-dependent
nitric oxide synthase (cNOS) activity in lateral
funiculus just below the site of injury and a strong
decrease in the ventral funiculus of two segments,
located above the injured site (Lukáčová et al.,
2000, 2002). Diaz-Ruiz et al. (2002) studied the
changes in constitutive and inducible NOS (iNOS)
activities after severe SCI in rats that survived
2–72 h. They observed a significant increase in
cNOS after 4 and 8 h, followed by a decline to
normal levels. In contrast, iNOS levels gradually
increased after injury, and the difference reached
statistical significance at 72 h posttrauma. Simi-
larly, time-dependent decrease in cNOS activity
and the expression of iNOS were noted in trau-
matized Th10 segment and in segments located in
close vicinity of spinal cord trauma (Chatzipanteli
et al., 2002). This study noted a significant de-
crease in cNOS activities at 3 h postinjury, and a
gradual return to control values within 24 h.
Again, iNOS activities displayed contrasting
changes. Activity was increased in all spinal cord
segments and time points studied (3 h-3 days). The
most robust increase was detected at 24 h at rostral
and caudal segments relative to injury site. Data
obtained recently from our laboratory have shown
a remarkable decrease of cNOS activity in the in-
jured site analyzed 1 day after Th9–Th10 spinal
cord hemisection (Lukáčová et al., 2006). At 7th
day of survival the enzyme activity returned to
approximately half of the control value. The
delayed enhancement in cNOS activity observed
at day 7 in the hemisected region may result from
the alterations in the neuronal circuitry (Xu et al.,
2000), subsequent proliferation of fibers contain-
ing NOS, and the expression of NOS observed in
axonal swelling after SCI (Guizar-Sahagun et al.,
1998). A strong increase of iNOS activity was
noted after thoracic spinal cord hemisection in
both postoperative times studied. These data sug-
gest that an increase of iNOS expression within
polymorphonuclear leucocyte and macrophages
(Xu et al., 2001; Chatzipanteli et al., 2002), ap-
pearing obvious from 3 h up to 14 days, is con-
sistent with the inflammatory response. In CNS
inflammation, infiltrated monocytes and macroph-
ages are the primary source of iNOS, but the
presence of endotoxins and proinflammatory
cytokines induce similar response in astrocytes
and microglia, the resident CNS immune cells
(Murphy et al., 1993). After SCI, iNOS positivity
has been observed and colocalized with macroph-
ages, neurons, astrocytes, and oligodendrocytes
(Kwak et al., 2005). Obviously, it is difficult to
specify unambiguously all the cell types that ex-
press iNOS in the damaged and hemorrhaged tis-
sue invaded by leucocytes. Chatzipanteli et al.
(2002) have identified polymorphonuclear leuco-
cytes as a significant cellular source of iNOS pro-
tein (Chatzipanteli et al., 2002). Satake et al.
(2000) has identified macrophages and perivascu-
lar cells as predominant iNOS positive cells after
SCI. A rapid formation and release of NO by
cNOS activity early after SCI (Hamada et al.,
1996) and later, increased formation of NO via
iNOS in the affected spinal cord region (Chatzi-
panteli et al., 2002) may accelerate neurodestruc-
tive events and lead to a secondary inactivation of
cNOS enzyme activity, possibly by NO binding to
the heme iron of the enzyme (Rogers and Ignarro,
1992) and/or by the phosphorylation of cNOS by
proteinkinases (Dalkara and Moskowitz, 1997).
Similarly, the reduction in blood flow and decrease
in oxygen tension caused by the structural damage
of cell bodies might be responsible for decreased
cNOS activity at the site of injury (Tator and
Fehlings, 1991; Vanderkooi et al., 1991; Kim et al.,
1993; Rengasamy and Johns, 1993). These data
provide strong evidence that intrinsic properties of

neurons contribute to the different cellular re-
sponses. The changes in the constitutive and iNOS
activities observed after spinal cord hemisection
within the site of injury support the participation of
NO in the pathology of SCI and result in a de-
creased neuronal viability. In principle, the primary
injury of the spinal cord evokes an inflammatory
response marked by generation of tumor necrosis
factor-a, the accumulation and adhesion of ne-
utrophils, and the activation of endothelial cells
and macrophages at the site of injury (Taoka et al.,
1997). These events, in turn, may contribute to
neuronal death or injury by inducing an excessive
formation of NO and oxygen free radicals (Merril
et al., 1993). Another very important mediator of
secondary CNS injury is oxidative-nitrosative stress
(Hall and Braughler, 1993), induced by antioxida-
tive depletion and/or an excess production of ox-
ygen free radicals and NO at the site of injury. NO
is able to promote oxidative damage by reacting
with superoxide anion to form a strong oxidant,
ONOO
, and by perturbing ion metabolism (Beck-
man et al., 1990; Reif and Simmons, 1990). The
rate of ONOO
formation is three times faster than
the dismutation of the superoxide anion to hydro-
gen peroxide, the reaction catalyzed by superoxide
dismutase (SOD) (Dawson and Dawson, 1996).
Thus, at appropriate concentrations NO can effec-
tively compete with SOD for superoxide anion. In
addition, NOS may generate ONOO
directly by
producing both NO and superoxide (Iadecola,
1997). NO can oxidize lipids, proteins, and nucleic
acid (Dawson and Dawson, 1994) causing spinal
cord cytotoxicity either directly through NO or its
derived species. Taken together, such processes
point to a possibility that NO represents a critical
factor in determining the extent of the functional
impairment of the spinal cord.
NADPHd-exhibiting and Fos-like immunolabeling
of intrinsic spinal cord neurons in a model of MCEC
We found considerable differences in the segmen-
tal and laminar distribution of Fos-like immuno-
reactive and NADPHd-stained neurons in the
lower lumbar and sacral segments of the dog spi-
nal cord using the model of MCEC (Orendáčová
177
et al., 2000, 2001). NADPHd histochemistry was
used as a marker of NOS-containing neurons. The
appearance and the time course of Fos-like
immunoreactive, NADPHd, and double-labeled
neurons was studied at 2 and 8 h postconstriction,
characterized as the incipient phase of cauda
equina syndrome. An increase in Fos-like
immunoreactivity in superficial laminae (I–II)
and enhanced NADPHd staining of lamina VIII
neurons were found. A statistically significant in-
crease in Fos-like immunoreactive neurons was
found in laminae I–II and VIII–X 8 h postcon-
striction, and in contrast, a prominent decrease in
Fos-like immunoreactive neurons was found in
laminae I–II, accompanied by a statistically sig-
nificant increase in Fos-like immunoreactive neu-
rons in more ventrally located laminae VII–X at 5
days postconstriction. A quantitative analysis of
laminar distribution of constriction-induced
NADPHd-stained neurons revealed a considera-
ble increase in these neurons in laminae VIII–IX
8 h postconstriction and highly statistically signif-
icant increase in NADPHd-stained neurons in
laminae VII–X 5 days postconstriction. Concur-
rently, the number of NADPHd-stained neurons
in laminae I–II was greatly reduced. While a low
number of double-labeled neurons was found
throughout the gray matter of lower lumbar and
sacral segments at 2 h postconstriction, a statisti-
cally significant number of double-labeled neurons
was found in lamina X 8 h and in laminae VII–X 5
days postconstriction (Orendáčová et al., 2000). A
prominent involvement of the spinal cord neurons
appearing in the lumbosacral segments at the be-
ginning, and seen early (8 h postconstriction) and
in fully developed cauda equina syndrome (5 days
postconstriction), results in a Fos-like immunore-
activity and strongly enhanced NADPHd staining
of some neuronal pools.
Immunolabeling and cNOS activity in the incipient
cauda equina syndrome of the dog
Remarkable changes appeared in the distribution
and appearance of NOS-IR neurons in the lower
lumbar and sacral segments analyzed 5 days post-
constriction in a dog model of cauda equina
178
constriction (Maršala et al., 2003). While NOS-
immunolabeled neurons detected in the upper
lumbar segments in the intermediate zone and
ventral horns were seen dispersed across both re-
gions in a quite small number, passing more cau-
dally, the number of NOS-immunolabeled neurons
was steadily growing, reaching in the sacral seg-
ments not only to the lateral portion of the inter-
mediate zone, but also encroaching to the lateral
and ventralmost portion of the ventral horn
(Orendáčová et al., 2001). Along with this increas-
ing density of NOS-IR neurons, several more or
less loosely dispersed dotted NOS-IR foci ap-
peared in the ventral horn and in the dorsal gray
commissure of S1–S3 segments. Highly positive,
usually small and middle-sized NOS-IR neurons
could be detected around the periphery of the
Onuf’s nucleus in S1–S3 segments. Under normal
circumstances, such NOS-IR positivity could be
seen only scarcely. It should be noted that in-
creased dotted non-somatic NOS-IR positivity
was often detected around interneurons in the
core region of the ventral horn and in the inter-
mediate zone. In some NOS-IR positive foci, the
dark small profiles truly matched the outlines of
these small neurons having, of course, no somatic
NOS-IR positivity. These findings clearly docu-
ment a high activation of NOS containing neurons
in this compression-induced cauda equina injury.
Considerable segmental and regional differences
of cNOS activity were found in the gray and white
matter regions of the lumbosacral segments of the
dog 5 days after MCEC surgery (Lukáčová et al.,
2004). The enzyme activity was assessed by a ra-
dioassay (Bredt and Snyder, 1990) in the gray
matter, divided into dorsal horn, intermediate
zone, ventral horn, and in the white matter, di-
vided into dorsal, lateral, and ventral columns, re-
spectively, comprising the upper (L1–L3) and
lower (L4–L7) lumbar segments, then in the non-
compartmentalized gray and white matter of
S1–S3 segments and cauda equina. While a statis-
tically significant decrease of cNOS activity was
found in all gray matter regions of the upper lum-
bar segments, a statistically significant increase of
cNOS activity could be detected in the ventral
horn of the lower lumbar segments. With the ex-
ception of a statistically significant increase of
cNOS activity in the lateral funiculus of the upper
lumbar segments, no changes of cNOS activity
could be detected in the white matter funiculus of
the upper and lower lumbar segments (Maršala
et al., 2003). An increase of cNOS activity was
detected in the non-compartmentalized white mat-
ter of S1–S3 segments, a finding strongly contrast-
ing with a statistically significant decrease of
cNOS activity assessed in the cauda equina.
Therapeutic interventions and NO reduction after
spinal trauma
Therapies aimed at limiting the evolution of seco-
ndary damage have a long history and are still
extensively studied both in experimental and clin-
ical use. Intravenous steroids (methylpredniso-
lone) have been registered for clinical use in
acute SCI, although there is still considerable de-
bate whether this therapy was really proved to be
safe and efficacious. Importantly, there is strong
evidence that secondary damage is associated with
invading inflammatory cells. Experimental results
have shown that depletion of hematogeneous ma-
crophages (Popovich et al., 1999) and inhibiting
diapedesis of blood borne leucocytes by antibodies
directed against leucocyte adhesion molecules
(Mabon et al., 2000; Bao et al., 2004; Gris et al.,
2004) reduces secondary damage after SCI, and
improves motor, sensitive, and autonomic func-
tion recovery in experimental animals.
However, as posttraumatic inflammation also
has a potentially beneficial role, better understand-
ing of the molecular mechanisms and their timing
in this process would allow more specific treat-
ments. These might diminish destructive processes,
while preserving the protective and regenerative
functions of inflammatory cells. Evolving under-
standing of the role of posttraumatic NO changes
represents one example of such therapeutic strat-
egy refinement.
Experiments with genetically manipulated ani-
mals demonstrated improved recovery with both
neuronal NOS-deficient (Farooque et al., 2001)
and iNOS-deficient mice (Isaksson et al., 2005).
Pharmacological inhibition of NO production by
aminoguanidine has been shown to improve func-
tional recovery in various independent studies
 179

(Chatzipanteli et al., 2002; Zhang et al., 2003; Soy References

et al., 2004). The effect of a more specific iNOS
inhibitor, 1400W has been tested in a traumatic Aley, K.O., McCarter, G. and Levine, J.D. (1998) Nitric oxide
signaling in pain and nociceptor sensitization in the rat. J.
brain injury model, where it dramatically reduced
Neurosci., 18: 7008–7014.
lesion volume, even if administered with 18 h delay Anderson, C.R. (1992) NADPH diaphorase-positive neurons in
(Jafarian-Tehrani et al., 2005). the rat spinal cord include a subpopulation of autonomic
In conclusion, SCI is a very complex pathologi- preganglionic neurons. Neurosci. Lett., 139: 280–284.
cal situation. It is believed that effective therapy Bao, F., Chen, Y., Dekaban, G.A. and Weaver, L.C. (2004)
will require combined approaches for minimizing Early anti-inflammatory treatment reduces lipid peroxidation
and protein nitration after spinal cord injury in rats. J. Neuro-
adverse reactions in the tissue and maximizing the chem., 88: 1335–1344.
regenerative capacity of the spinal cord (Ramer Barilari, M.G. and Kuypers, J.M. (1969) Propriospinal fibers
et al., 2000; Thuret et al., 2006). Precise under- interconnecting the spinal enlargements in the cat. Brain
standing of the timing of iNOS expression in spe- Res., 14: 321–330.
Beavo, J.A. (1995) Cyclic nucleotide phosphodiesterases: func-
cific models of SCI, and development of new more
tional implications of multiple isoforms. Physiol. Rev., 75:
specific inhibitors might represent one component 725–748.
of such effective therapy. Beckman, J.S., Beckman, T.W., Chen, J., Marshall, P.A. and
Freeman, B.A. (1990) Apparent hydroxyl radical production
by peroxynitrite: implications for endothelial injury from
Abbreviations nitric oxide and superoxide. Proc. Natl. Acad. Sci. U.S.A.,
87: 1620–1624.
Blottner, D. and Baumgarten, H. (1992) Nitric oxide synthase
ANP atrial natriuretic peptide (NOS) containing sympathoadrenal cholinergic neurons of
cGMP cyclic guanosine 30 ,50 -monophos- the rat IML-cell column: evidence from histochemistry,
phate immunohistochemistry and retrograde labeling. J. Comp.
CNS central nervous system Neurol., 316: 45–56.
Bredt, D.S., Hwang, P.M. and Snyder, S.H. (1990) Localization
GC guanylyl cyclase
of nitric oxide synthase indicating a neuronal role for nitric
iNOS inducible nitric oxide synthase oxide. Nature, 347: 768–770.
MCEC multiple cauda equina constric- Bredt, D.S. and Snyder, S.H. (1990) Isolation of nitric oxide
tion synthase, a calmodulin-requiring enzyme. Proc. Natl. Acad.
NADPHd nicotinamide adenine dinucleo- Sci. U.S.A., 87: 682–685.
Bredt, D.S. and Snyder, S.H. (1992) Nitric oxide: a novel neu-
tide phosphate diaphorase
ronal messenger. Neuron, 8: 3–11.
NMDA N-methyl-D-aspartate Burnett, A.L., Saito, S., Maguire, M.P., Yamaguchi, H.,
NO nitric oxide Chang, T.S.K. and Hanley, D.F. (1995) Localization of ni-
NOS nitric oxide synthase tric oxide synthase in spinal nuclei innervating pelvic ganglia.
NOS-IR NOS-immunoreactive J. Urol., 153: 212–217.
Chatzipanteli, K., Garcia, R., Marcillo, A.E., Loor, K.E.,
ONOO
peroxynitrite
Kraydieh, S. and Dietrich, W.D. (2002) Temporal and seg-
PDE phosphodiesterase mental distribution of constitutive and inducible nitric oxide
pGC particular guanylyl cyclase synthases after traumatic spinal cord injury: effect of amino-
SCI spinal cord injury guanidine treatment. J. Neurotrauma, 5: 639–651.
sGC soluble guanylyl cycle Dalkara, T. and Moskowitz, M.A. (1997) Neurotoxic and ne-
uroprotective roles of nitric oxide in cerebral ischaemia. Int.
SOD superoxide dismutase
Rev. Neurobiol., 40: 319–336.
Dawson, T.M. and Dawson, V.L. (1994) gp120 neurotoxicity in
primary cortical cultures. Adv. Neuroimmunol., 4: 167–173.
Dawson, T.M., Dawson, V.L. and Snyder, S.H. (1992) A novel
neuronal messenger molecule in brain: the free radical, nitric
Acknowledgments
oxide. Ann. Neurol., 32: 297–311.
Dawson, V.L. and Dawson, T.M. (1996) Nitric oxide actions in
This work was supported by the APVT grant: 51- neurochemistry. Neurochem. Int., 29: 97–110.
013002 and VEGA grants: 2/5134/25, 2/5135/25, Diaz-Ruiz, A., Ibarra, A., Perez-Severiano, F., Guizar-Sahagun,
and 2/6212/26. G., Grijalva, I. and Rios, C. (2002) Constitutive and inducible
180
nitric oxide synthase activities after spinal cord contusion in
rats. Neurosci. Lett., 319: 129–132.
Dun, N.J., Dun, S.L., Wu, S.Y., Förstermann, U., Schmidt,
H.H.H.W. and Tseng, L.F. (1993) Nitric oxide synthase
immunoreactivity in the rat, mouse, cat and squirrel monkey
spinal cord. Neuroscience, 54: 845–857.
Egberongbe, Y.I., Gentleman, S.M., Falkai, P., Bogerts, B.,
Polak, J.M. and Roberts, G.W. (1994) The distribution of
nitric oxide synthase immunoreactivity in the human brain.
Neuroscience, 59: 561–578.
Farooque, M., Isaksson, J. and Olsson, Y. (2001) Improved
recovery after spinal cord injury in neuronal nitric oxide
synthase-deficient mice but not in TNF-alpha-deficient mice.
J. Neurotrauma, 18: 105–114.
Fedele, E., Jin, Y., Varnier, G. and Raiteri, M. (1996) In vivo
microdialysis study of a specific inhibitor of soluble guanylyl
cyclase on the glutamate receptor/nitric oxide/cyclic GMP
pathway. Br. J. Pharmacol., 119: 590–594.
Francis, S.H., Turko, I.V. and Corbin, J.D. (2001) Cyclic nuc-
leotide phosphodiesterases: relating structure and function.
Prog. Nucleic Acid Res. Mol. Biol., 65: 1–52.
Gris, D., Marsh, D.R., Oatway, M.A., Chen, Y., Hamilton,
E.F., Dekaban, G.A. and Weaver, L.C. (2004) Transient
blockade of the CD11d/CD18 integrin reduces secondary
damage after spinal cord injury, improving sensory, auto-
nomic, and motor function. J. Neurosci., 24: 4043–4051.
Guizar-Sahagun, G., Garcia-Lopez, P., Espitia, A.L., Grijalva,
I., Franco-Bourland, R.E. and Madrazo, I. (1998) Transitory
expression of NADPH diaphorase (NOS) in axonal swellings
after spinal cord injury. NeuroReport, 9: 2899–2902.
Gupta, G., Azam, M., Yang, L. and Danziger, R.S. (1997) The
beta2 subunit inhibits stimulation of the alpha1/beta1 form
of soluble guanylyl cyclase by nitric oxide: potential relevance
to regulation of blood pressure. J. Clin. Invest., 100:
1488–1492.
Halát, G., Lukáčová, N., Chavko, M. and Maršala, J. (1987)
Effects of incomplete ischemia and subsequent recirculation
on free palmitate, stearate, oleate and arachidonate levels in
lumbar and cervical spinal cord of rabbit. Gen. Physiol.
Biophys., 6: 387–399.
Hall, E.D. and Braughler, J.M. (1993) Free radicals in CNS
injury. Res. Publ. Assoc. Res. Nerv. Ment. Dis., 71: 81–105.
Hamada, Y., Ikata, T., Katoh, S., Tsuchiya, K., Niwa, M.,
Tsutsumishita, Y. and Fukuzawa, K. (1996) Roles of nitric
oxide in compression injury of rat spinal cord. Free Radic.
Biol. Med., 20: 1–9.
Horrocks, L.A., Dorman, R.V. and Porcellati, G. (1984) Fatty
acids and phospholipids in brain during ischemia. In: Bes P.,
Braquet R., Paoletti R. and Siesjö B.K. (Eds.), Cerebral Is-
chemia Excerpta Medica. Elsevier, Amsterdam, Congress
Series 654A, pp. 211–226.
Humbert, P., Niroomand, F., Fischer, G., Mayer, B., Koesling,
D., Hinsch, K.D., Gausepohl, H., Frank, R., Schultz, G. and
Bohme, E. (1990) Purification of soluble guanylyl cyclase
from bovine lung by a new immunoaffinity chromatographic
method. Eur. J. Biochem., 190: 273–278.
Iadecola, C. (1997) Bright and dark sides of nitric oxide in
ischemic brain injury. Trends Neurosci., 20: 132–139.
Isaksson, J., Farooque, M. and Olsson, Y. (2005) Improved
functional outcome after spinal cord injury in iNOS-deficient
mice. Spinal Cord, 43: 167–170.
Jafarian-Tehrani, M., Louin, G., Royo, N.C., Besson, V.C.,
Bohme, G.A., Plotkine, M. and Marchand-Verrecchia, C.
(2005) 1400W, a potent selective inducible NOS inhibitor:
improves histopathological outcome following traumatic
brain injury in rats. Nitric Oxide, 12: 61–69.
Kim, N., Vardi, Y., Padma-Nathan, H., Daley, J., Goldstein, I.
and Saenz de Tejada, I. (1993) Oxygen tension regulates the
nitric oxide pathway: physiological role in penile erection.
J. Clin. Invest., 91: 437–442.
Koesling, D. and Friebe, A. (2000) Enzymology of soluble
guanylyl cyclase. In: Mayer B. (Ed.), Nitric Oxide. Springer,
Berlin, pp. 93–109.
Koglin, M., Vehse, K., Budaeus, L., Scholz, H. and Behrends,
S. (2001) Nitric oxide activates the beta 2 subunit of soluble
guanylyl cyclase in the absence of a second subunit. J. Biol.
Chem., 276: 30737–30743.
Kowall, N.W., Beal, M.F., Ferrante, R.J. and Martin, J.B.
(1985) Topography of nicotinamide adenine dinucleotide
phosphate-diaphorase staining neurons in rat striatum. Neuro-
sci. Lett., 59: 61–66.
Kowall, N.W., Ferrante, R.J., Beal, M.F., Richardson, Jr.,
E.P., Sofroniew, M.V., Cuello, A.C. and Martin, J.B. (1987)
Neuropeptide Y, somatostatin and reduced nicotinamide ad-
enine dinucleotide phosphate diaphorase in the human stria-
tum: a combined immunocytochemical and enzyme
histochemical study. Neuroscience, 20: 817–828.
Kowall, N.W. and Mueller, M.O. (1988) Morphology and dis-
tribution of nicotinamide adenine dinucleotide phosphate
(reduced form) diaphorase reactive neurons in human brain-
stem. Neuroscience, 26: 645–654.
Kwak, E.K., Kim, J.W., Kang, K.S., Lee, Y.H., Hua, Q.H.,
Park, T.I., Park, J.Y. and Sohn, Y.K. (2005) The role of
inducible nitric oxide synthase following spinal cord injury in
rat. J. Korean Med. Sci., 20: 663–669.
Lee, J.H., Price, R.H., Williams, F.G., Mayer, B. and Beitz,
A.J. (1993) Nitric oxide synthase is found in some spin-
othalamic neurons and in neuronal processes that appose
spinal neurons that express Fos induced by noxious stimu-
lation. Brain Res., 608: 324–333.
Lehotský, J., Kaplán, P., Račay, P., Mézešová, V. and Raey-
maekers, L. (1999) Distribution of plasma membrane Ca2+
pump (PMCA) isoforms in the gerbil brain: effect of
ischemia-reperfusion injury. Neurochem. Int., 35: 221–227.
Liu, D., Ling, X., Wen, J. and Liu, J. (2000) The role of reactive
nitrogen species in secondary spinal cord injury: formation of
nitric oxide, peroxynitrite, and nitrated protein. J. Neuro-
chem., 75: 2144–2154.
de Louw, A.J., de Vente, J., Steinbusch, H.P., Steinbusch,
H.W., Troost, J. and Vles, J.S. (2002) Baclofen inhibits ANP-
mediated cyclic GMP synthesis in the rat cervical spinal cord.
Neurosci. Lett., 321: 120–122.

Lucas, K.A., Pitari, G.M., Kazerounian, S., Ruiz-Stewart, I.,
Park, J., Schulz, S., Chepenik, K.P. and Waldman, S.A.
(2000) Guanylyl cyclases and signaling by cyclic GMP.
Pharmacol. Rev., 52: 375–414.
Lukáčová, N., Chavko, M. and Maršala, J. (1994) Ischemia/
reperfusion-induced lipid peroxidation in the central nervous
system. Biologia, 49: 871–879.
Lukáčová, N., Čı́žková, D., Maršala, M., Jalč, P. and Maršala,
J. (1999) Segmental and laminar distributions of nicotin-
amide adenine dinucleotide phosphate–diaphorase expressing
and neuronal nitric oxide synthase–immunoreactive neurons
versus radioassay detection of catalytic nitric oxide synthase
activity in the rabbit spinal cord. Neuroscience, 94: 229–237.
Lukáčová, N., Čı́žková, D., Maršala, M., Pavel, J., Jalč, P.,
Šulla, I., Kafka, J. and Maršala, J. (2000) Effect of midtho-
racic spinal cord constriction on catalytic nitric oxide synt-
hase activity in the white matter columns of rabbit.
Neurochem. Res., 25: 1139–1148.
Lukáčová, N., Kafka, J., Čı́žková, D., Maršala, M. and
Maršala, J. (2004) The effect of cauda equina constriction
on nitric oxide synthase activity. Neurochem. Res., 29:
429–439.
Lukáčová, N., Kolesárová, M., Kuchárová, K., Pavel, J.,
Kolesár, D., Radoňák, J., Maršala, M., Chalimoniuk, M.,
Langfort, J. and Maršala, J. (2006) The effect of a spinal cord
hemisection on changes in nitric oxide synthase pools in the
site of injury and in regions located far away from the injured
site. Cell. Mol. Neurobiol., 26: 1365–1383.
Lukáčová, N. and Pavel, J. (2000) Catalytic nitric oxide synt-
hase activity in the white and gray matter regions of the
spinal cord of rabbits. Physiol. Res., 49: 167–173.
Lukáčová, N., Pavel, J., Jalč, P., Čı́žková, D., Maršala, M.,
Silvašiová, M., Lukáč, I. and Maršala, J. (2002) The effect of
spinal cord constriction on NOS-cGMP signaling pathway in
the white matter of rabbit. Biologia, 57: 789–795.
Lukáčová, N., Pavel, J., Kuchárová, K. and Maršala, J. (2005)
Plasticity of the spinal neural circuitry in nitric oxide synthase
immunoreactive pathways after spinal cord hemisection. In:
Coleman R.M. (Ed.), Trends in Neurochemistry Research.
Nova Publishers, New York, pp. 189–222.
Luo, D., Leung, E. and Vincent, S.R. (1994) Nitric oxide-
dependent efflux of cGMP in rat cerebellar cortex: an in vivo
microdialysis study. J. Neurosci., 14: 263–271.
Mabon, P.J., Weaver, L.C. and Dekaban, G.A. (2000) Inhibi-
tion of monocyte/macrophage migration to a spinal cord in-
jury site by an antibody to the integrin alphaD: a potential
new anti-inflammatory treatment. Exp. Neurol., 166: 52–64.
Maršala, J., Kafka, J., Lukáčová, N., Čı́žková, D., Maršala, M.
and Katsube, N. (2003) Cauda equina syndrome and nitric
oxide synthase immunoreactivity in the spinal cord of the
dog. Physiol. Res., 52: 481–496.
Maršala, J., Kluchová, D. and Maršala, M. (1997) Spinal cord
gray matter layers rich in NADPH diaphorase positive neu-
rons are refractory to ischemia-reperfusion-induced injury: a
histochemical and silver impregnation study in rabbit. Exp.
Neurol., 145: 165–179.
181
Maršala, J., Lukáčová, N., Čı́žková, D., Lukáč, I., Kuchárová,
K. and Maršala, M. (2004) Premotor nitric oxide synthase
immunoreactive pathway connecting lumbar segments with
the ventral motor nucleus of the cervical enlargement in the
dog. J. Chem. Neuroanat., 27: 43–54.
Maršala, J., Maršala, M., Vanický, I. and Taira, Y. (1999)
Localization of NADPHd exhibiting neurons in the spinal
cord of the rabbit. J. Comp. Neurol., 406: 263–284.
Maršala, J., Vanický, I., Maršala, M., Jalč, P., Orendáčová, J.
and Taira, Y. (1998) NADPH diaphorase in the spinal cord
of dogs. Neuroscience, 85: 847–862.
Mc Adoo, J., Hughes, M.G., Xu, G.Y., Robak, G. and DeC-
astro, Jr., R., (1997) Microdialysis studies of the role of
chemical agents in secondary damage upon spinal cord in-
jury. J. Neurotrauma, 14: 507–515.
Merril, J.E., Ignarro, L., Sherman, M.P., Melinek, J. and Lane,
T.E. (1993) Microglial cell cytotoxicity of oligodendrocytes is
mediated through nitric oxide. J. Immunol., 151: 2132–2141.
Mizukawa, K., McGeer, P.L., Vincent, S.R. and McGeer, E.G.
(1989) Distribution of reduced-nicotinamide-adenine-dinuc-
leotide-phosphate diaphorase-positive cells and fibers in the
cat central nervous system. J. Comp. Neurol., 279: 281–311.
Morris, R., Southam, E., Gittins, S.R., de Vente, J. and Garth-
waite, J. (1994) The NO-cGMP pathway in neonatal rat
dorsal horn. Eur. J. Neurosci., 6: 876–879.
Morton, D.B. and Bredt, D.S. (1998) Norepinephrine increases
cyclic GMP levels in cerebellar cells from neuronal nitric
oxide synthase knockout mice. J. Neurochem., 71: 440–443.
Murad, F. (1994) Regulation of cytosolic guanylyl cyclase by
nitric oxide: the NO-cyclic GMP signal transduction system.
Adv. Pharmacol., 26: 19–33.
Murphy, S., Simmons, M.L., Agullo, L., Garcia, A., Feinstein,
D.L., Galea, E., Reis, D.J., Minc-Golomb, D. and Schwartz,
J.P. (1993) Synthesis of nitric oxide in CNS glial cells. Trends
Neurosci., 16: 323–328.
Nakahara, S., Yone, K., Setoguchi, T., Yamaura, I., Arishima,
Y., Yoshino, S. and Komiya, S. (2002) Changes in nitric
oxide and expression of nitric oxide synthase in spinal cord
after acute traumatic injury in rats. J. Neurotrauma, 19:
1467–1474.
Nakano, S., Kogure, K., Abe, K. and Yae, T. (1990) Ischemia-
induced alterations in lipid metabolism of the gerbil cerebral
cortex: I. Changes in free fatty acid liberation. J. Neuro-
chem., 54: 1911–1916.
Orendáčová, J., Čı́žková, D., Kafka, J., Lukáčová, N.,
Maršala, M., Šulla, I., Maršala, J. and Katsube, N. (2001)
Cauda equina syndrome. Prog. Neurobiol., 64: 613–637.
Orendáčová, J., Maršala, M., Šulla, I., Kafka, J., Jalč, P.,
Čı́žková, D., Taira, Y. and Maršala, J. (2000) Incipient cauda
equina syndrome as a model of somatovisceral pain in dogs:
spinal cord structures involved as revealed by the expression
of c-fos and NADPH diaphorase activity. Neuroscience, 95:
543–557.
Popovich, P.G., Guan, Z., Wei, P., Huitinga, I., van Rooijen,
N. and Stokes, B.T. (1999) Depletion of hematogenous ma-
crophages promotes partial hindlimb recovery and
182
neuroanatomical repair after experimental spinal cord injury.
Exp. Neurol., 158: 351–365.
Potter, L.R. and Hunter, T. (2001) Guanylyl cyclase-linked
natriuretic peptide receptors: structure and regulation. J.
Biol. Chem., 276: 6057–6060.
Prast, H. and Philippu, A. (2001) Nitric oxide as modulator of
neuronal function. Prog. Neurobiol., 64: 51–68.
Ramer, M.S., Harper, G.P. and Bradbury, E.J. (2000) Progress
in spinal cord research: a refined strategy for the Interna-
tional Spinal Research Trust. Spinal Cord, 38: 449–472.
Ransom, B.R., Waxman, S.G. and Stys, P.K. (1993) Anoxic
injury of central myelinated axons: ionic mechanisms and
pharmacology. Res. Publ. Assoc. Res. Nerv. Ment. Dis., 71:
121–151.
Reif, D.W. and Simmons, R.D. (1990) Nitric oxide mediates
iron release from ferritin. Arch. Biochem. Biophys., 283:
537–541.
Rengasamy, A. and Johns, R.A. (1993) Regulation of nitric
oxide synthase by nitric oxide. Mol. Pharmacol., 44: 124–128.
Rogers, N.E. and Ignarro, L.J. (1992) Constitutive nitric oxide
synthase from cerebellum is reversibly inhibited by nitric ox-
ide formed from L-arginine. Biochem. Biophys. Res. Co-
mmun., 189: 242–249.
Russwurm, M., Behrends, S., Harteneck, C. and Koesling, D.
(1998) Functional properties of a naturally occurring isoform
of soluble guanylyl cyclase. Biochem. J., 335: 125–130.
Saito, S., Kidd, G.J., Trapp, B.D., Dawson, T.M., Bredt, D.S.,
Wilson, D.A., Traystmann, R.J., Snyder, S.H. and Hanley,
D.F. (1994) Rat spinal cord neurons contain nitric oxide
synthase. Neuroscience, 59: 447–452.
Satake, K., Matsuyama, Y., Kamiya, M., Kawakami, H.,
Iwata, H., Adachi, K. and Kiuchi, K. (2000) Nitric oxide via
macrophage iNOS induces apoptosis following traumatic
spinal cord injury. Brain Res. Mol. Brain Res., 85: 114–122.
Schmidt, H.H. (1992) NO, CO and OH: endogenous soluble
guanylyl cyclase-activating factors. FEBS Lett., 307: 102–107.
Sharma, H.S., Westman, J., Gordh, T. and Nyberg, F. (2000)
Spinal cord injury induced c-fos expression is reduced by
p-CPA, a serotonin synthesis inhibitor: an experimental
study using immunohistochemistry in the rat. Acta Neuro-
chir. Suppl., 76: 297–301.
Sherrington, C.S. and Laslett, E.E. (1903) Observations on
some spinal reflexes and the inter connection of spinal seg-
ments. J. Physiol. (Lond.), 29: 58–96.
Smolenski, A., Burkhardt, A.M., Eigenthaler, M., Butt, E.,
Gambaryan, S., Lohmann, S.M. and Walter, U. (1998)
Functional analysis of cGMP-dependent protein kinases I
and II as mediators of NO/cGMP effects. Naunyn
Schmiedebergs Arch. Pharmacol., 358: 134–139.
Snyder, S.H. and Bredt, D. (1991) Nitric oxide as a neuronal
messenger. Trends Pharmacol. Sci., 12: 125–128.
Soy, O., Aslan, O., Uzun, H., Barut, S., Igdem, A.A., Belce, A.
and Colak, A. (2004) Time-level relationship for nitric oxide
and the protective effects of aminoguanidine in experimental
spinal cord injury. Acta Neurochir. (Wien.), 146: 1329–1336.
van Staveren, W.C., Steinbusch, H.W., Markerink-van Itter-
sum, M., Repaske, D.R., Goy, M.F., Kotera, J., Omori, K.,
Beavo, J.A. and de Vente, J. (2003) mRNA expression pat-
terns of the cGMP-hydrolyzing phosphodiesterases types 2,
5, and 9 during development of the rat brain. J. Comp. Ne-
urol., 467: 566–580.
Strosznajder, J. and Chalimoniuk, M. (1996) Biphasic enhance-
ment of nitric oxide synthase activity and cGMP level fol-
lowing brain ischemia in gerbils. Acta Neurobiol. Exp., 56:
71–81.
Stys, P.K. and Steffensen, I. (1996) Na(+)-Ca2+ exchange in
anoxic/ischemic injury of CNS myelinated axons. Ann. N.Y.
Acad. Sci., 779: 366–378.
Stys, P.K., Waxman, S.G. and Ransom, B.R. (1991) Na+-Ca2+
exchanger mediates Ca2+ influx during anoxia in mammalian
central nervous system white matter. Ann. Neurol., 30:
375–380.
Tao, Y.X., Hassan, A., Haddad, E. and Johns, R.A. (2000)
Expression and action of cyclic GMP-dependent protein kin-
ase I alpha in inflammatory hyperalgesia in rat spinal cord.
Neuroscience, 95: 525–533.
Taoka, Y., Okajima, K., Uchiba, M., Murakami, K., Kushi-
moto, S., Johno, M., Naruo, M., Okabe, H. and Takatsuki,
K. (1997) Role of neutrophils in spinal cord injury in the rat.
Neuroscience, 79: 1177–1182.
Tator, C.H. and Fehlings, M.G. (1991) Review of the secondary
injury theory of acute spinal cord trauma with emphasis on
vascular mechanisms. J. Neurosurg., 75: 15–26.
Thomas, E. and Pearse, A.G.E. (1964) The solitary active cells:
histochemical demonstration of damage-resistant nerve cells
with a TPN-diaphorase reaction. Acta Neuropathol., 3:
238–249.
Thuret, S., Moon, L.D. and Gage, F.H. (2006) Therapeutic
interventions after spinal cord injury. Nat. Rev. Neurosci., 7:
628–643.
Toborek, M., Garrido, R., Malecki, A., Kaiser, S., Mattson,
M.P., Hennig, B. and Young, B. (2000) Nicotine attenuates
arachidonic acid-induced overexpression of nitric oxide synt-
hase in cultured spinal cord neurons. Exp. Neurol., 161:
609–620.
Trudrung, P., Wirth, U. and Mense, S. (2000) Changes in the
number of nitric oxide-synthesizing neurons on both sides of
a chronic transection of the rat spinal cord. Neurosci. Lett.,
287: 125–128.
Vallebuona, F. and Raiteri, M. (1993) Monitoring of cyclic
GMP during cerebellar microdialysis in freely-moving rats as
an index of nitric oxide synthase activity. Neuroscience, 57:
577–585.
Valtschanoff, J.G., Weinberg, R.J. and Rustioni, A. (1992)
NADPH diaphorase in the spinal cord of rats. J. Comp.
Neurol., 321: 209–222.
Vanderkooi, J.M., Erecinska, M. and Silver, I.A. (1991) Oxygen
in mammalian tissue: methods of measurement and affinities
of various reactions. Am. J. Physiol., 260: C1131–C1150.
de Vente, J., Manshanden, C.G., Sikking, R.A., Ramaekers,
F.C. and Steinbusch, H.W. (1990) A functional parameter to
study heterogeneity of glial cells in rat brain slices: cyclic
guanosine monophosphate production in atrial natriuretic
factor (ANF)-responsive cells. Glia, 3: 43–54.

Vincent, S.R. (1994) Nitric oxide: a radical neurotransmitter in
the central nervous system. Prog. Neurobiol., 42: 129–160.
Vincent, S.R. and Kimura, H. (1992) Histochemical mapping of
nitric oxide synthase in the rat brain. Neuroscience, 46:
755–784.
Vizzard, M.A., Erdman, S.L. and de Groat, W.C. (1995) In-
creased expression of neuronal nitric oxide synthase (NOS) in
visceral neurons after nerve injury. J. Neurosci., 15:
4033–4045.
Vizzard, M.A., Erdman, S.L., Erickson, V.L., Stewart,
R.J., Roppolo, J.R. and de Groat, W.C. (1994a) Localiza-
tion of NADPH diaphorase in the lumbosacral spinal cord
and dorsal root ganglia of the cat. J. Comp. Neurol., 339:
62–75.
Vizzard, M.A., Erdman, S.L., Roppolo, J.R., Förstermann, U.
and de Groat, W.C. (1994b) Differential localization of neu-
ronal nitric oxide synthase immunoreactivity and NADPH-
diaphorase activity in the cat spinal cord. Cell. Tissue Res.,
278: 299–309.
Vles, J.S., de Louw, A.J., Steinbusch, H., Markerink-van Itter-
sum, M., Steinbusch, H.W., Blanco, C.E., Axer, H., Troost,
J. and de Vente, J. (2000) Localization and age-related
changes of nitric oxide- and ANP-mediated cyclic-GMP syn-
thesis in rat cervical spinal cord: an immunocytochemical
study. Brain Res., 857: 219–234.
Wang, X. and Robinson, P.J. (1997) Cyclic GMP-dependent
protein kinase and cellular signaling in the nervous system.
J. Neurochem., 68: 443–456.
Wedel, B.J. and Garbers, D.L. (1997) New insights on the
functions of the guanylyl cyclase receptors. FEBS Lett., 410:
29–33.
183
Wetts, R. and Vaughn, J.E. (1994) Choline acetyltransferase
and NADPH diaphorase are co-expressed in rat spinal cord
neurons. Neuroscience, 63: 1117–1124.
Wood, P.L. and Garthwaite, J. (1994) Models for the diffu-
sional spread of nitric oxide: implications for neuronal nitric
oxide signalling and its pharmacological properties. Neuro-
pharmacology, 33: 1235–1244.
Wu, W., Liuzzi, F.J., Schinco, F.P., Depto, A.S., Li, Y., Mong,
J.A., Dawson, T.M. and Snyder, S.H. (1994) Neuronal nitric
oxide synthase is induced in spinal neurons by traumatic in-
jury. Neuroscience, 61: 719–726.
Xu, J., Kim, G.M., Chen, S., Yan, P., Ahmed, S.H., Ku, G.,
Beckman, J.S., Xu, X.M. and Hsu, C.Y. (2001) iNOS and
nitrotyrosine expression after spinal cord injury. J. Neuro-
trauma, 18: 523–532.
Xu, W., Liu, L., Smith, G.C. and Charles, G. (2000) Nitric
oxide upregulates expression of DNA-PKCs to protect cells
from DNA-damaging anti-tumour agents. Nat. Cell. Biol., 2:
339–345.
Yezierski, R.P., Culberson, J.L. and Brown, P.B. (1980) Cells of
origin of propriospinal connections to cat lumbosacral gray
as determined with horseradish peroxidase. Exp. Neurol., 69:
493–512.
Yick, L.W., Wu, W., So, K.F. and Wong, S.Y. (1998) Time
course of NOS expression and neuronal death in Clarke’s
nucleus following traumatic injury in adult rat spinal cord.
Neurosci. Lett., 241: 155–158.
Zhang, X.Y., Zhou, C.S., Jin, A.M., Tian, J., Zhang, H., Yao,
W.T. and Zheng, G. (2003) Effect of aminoguanidine on the
recovery of rat hindlimb motor function after spinal cord
injury. Di Yi Jun Yi Da Xue Xue Bao, 23: 687–689.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 12

Aquaporins: role in cerebral edema and brain water

balance$

Zsolt Zador1,2, Orin Bloch1,2, Xiaoming Yao1,2 and Geoffrey T. Manley1,2,

1
Department of Neurological Surgery, University of California, San Francisco, CA, USA
2
Brain and Spinal Injury Center, University of California, San Francisco, CA 94110, USA

Abstract: The regulation of water balance in the brain is crucial. A disruption in this equilibrium causes an
increase in brain water content that significantly contributes to the pathophysiology of traumatic brain
injury, hydrocephalus, and a variety of neurological disorders. The discovery of the aquaporin (AQP)
family of membrane water channels has provided important new insights into the physiology and pathology
of brain water homeostasis. A number of recent studies are described in the review that demonstrated the
important role of AQP1 and AQP4 in brain water balance and cerebral edema. Phenotypic analyses of
AQP deficient mice have allowed us to explore the role of these membrane water channels in the mech-
anisms of cytotoxic edema, vasogenic edema, and CSF production. These studies indicate that AQP4 plays
significant role in the development of cytotoxic edema and the absorption of excess brain water resulting
from vasogenic edema. They also have demonstrated the role of AQP1 in CSF production and maintenance
of steady-state ICP. The ability to modulate water flux through AQP deletion has provided new insights
into brain water homeostasis and suggested a number of new research directions. However, these efforts
have not yet translated to the treatment human clinical diseases. These advances will require the deve-
lopment of AQP inhibitors and activators to establish the benefit modulating the function of these water
channels.

Keywords: water transport; aquaporin; edema; brain injury; stroke; epilepsy; cytotoxic; vasogenic

Introduction water transport (Verkman, 2002). To date, over 12

Water is regarded as the matrix of life: it is an
essential solvent for all living organisms. Although
it diffuses through tissues relatively freely, the
transport of water across cell membranes is facil-
itated by highly efficient transmembrane channels
called aquaporins (AQPs). The selective water flow
through AQPs is passive, moving along osmotic
gradients, which provide the driving force for
$
Supported by NIH NS050173 and the UCSF Brain and Spi-
nal Injury Center.
Corresponding author. Tel.: +415-206-4536; Fax: +415-206-
3948; E-mail: manleyg@neurosurg.ucsf.edu
different members of the AQP family have been
identified, and a number of them have been shown
to contribute to rapid water flux in various tissues
including the kidney, lung, gastrointestinal tract,
and central nervous system (CNS) (Agre et al.,
2002; Verkman, 2002).
The regulation of water balance in the brain is
crucial. Normally, water transport is tightly regu-
lated to maintain a strict homeostatic balance bet-
ween the cerebral vascular, brain tissue, and
cerebrospinal fluid (CSF) compartments. A dis-
ruption in this equilibrium causes an increase in
brain water content that significantly contributes
to the pathophysiology of traumatic brain injury,

DOI: 10.1016/S0079-6123(06)61012-1 185

186
hydrocephalus, and a variety of neurological dis-
orders (Fishman, 1975). The rigid nature of the
skull provides little capacity to buffer intracranial
volume changes, and beyond a limited threshold
the intracranial pressure (ICP) rises rapidly. The
increased ICP can ultimately lead to the impair-
ment of cerebral blood flow resulting in further
brain injury and death. In spite of the clinical im-
portance of brain water balance, the molecular
mechanisms remain poorly understood and the
therapeutic interventions have improved little over
past 80 years (Weed and McKibben, 1919). A
number of recent studies have demonstrated that
the AQP water channels play a significant role in
brain water balance and cerebral edema. Thus,
they may provide a novel molecular target for the
treatment of a number of neurological disorders.
The aim of this review is to summarize the current
understanding of AQPs in cerebral edema and
brain water balance.
Aquaporin expression in the CNS
A number of AQPs are expressed in the brain and
spinal cord. Two of these water channels, AQP1
and AQP4, are expressed in the brain at tissue–
fluid interfaces important for brain water balance.
AQP1 is expressed in the apical membrane of the
choroid plexus epithelium (Hasegawa et al., 1994;
Speake et al., 2003), where it has been shown to
facilitate CSF secretion into the cerebral ventricles
(Oshio et al., 2005). AQP4 is expressed in the
basolateral membrane of the ependymal cells lin-
ing the cerebral ventricles, where it may play a
role in water transport between the brain paren-
chyma and CSF compartments. In addition,
AQP4 is expressed in the in the astocytic foot
processes in direct contact with the cerebral
blood vessels that comprise the blood-brain
barrier. Highly polarized AQP4 expression is also
found in the dense astrocyte cell processes that
form the glia limitans, a structure that is adjacent
to the CSF-filled subarachnoid space. The expres-
sion of AQP1 and AQP4 at these tissue–fluid in-
terfaces indicate a potential role in maintaining
brain water homeostasis by facilitating water flux
between these compartments. Furthermore, their
expression pattern suggests they may contribute to
the underlying mechanism of cerebral edema.
Aquaporins and brain water balance
The brain is composed of nearly 78% water
(McIlwain and Bachelard, 1985). While the ma-
jority of the brain water is intracellular, it is also
distributed in the extracellular space and the CSF
compartments (cerebral ventricles and sub-
arachnoid space). Water enters the brain through
the blood-brain barrier or across the choroids
plexus (Fig. 1). Under normal physiological con-
ditions, the extracellular fluid in the brain is be-
lieved to flow into the cerebral ventricles or
subarachnoid space and exit the brain through
the arachnoid granulations into to the venous sys-
tem (Abbott, 2004). Recent studies of transgenic
mice lacking AQP1 and AQP4 have helped to de-
fine the role of these channels in CSF production
and absorbtion.
The choroid plexus epithelium expresses a va-
riety of ion transporters and channels that provide
the osmotic driving force for para- and transcel-
lular water flux into the cerebral ventricles to form
CSF (Wright et al., 1977). The rate of CSF
secretion by the choroid plexus is substantial,
comparable to the rate of fluid reabsorption in the
proximal tubules (Welch, 1963). Earlier studies
had demonstrated a high osmotic–diffusion ratio
of the choroids plexus, suggesting the presence of
a water pore-like pathway (Wright et al., 1977).
To explore this hypothesis, our group developed
novel methods to measure choroids plexus water
permeability and CSF production in wildtype
and AQP1-null mice. We found that deletion of
AQP1 reduced osmotically induced water trans-
port in the choroid plexus by fivefold. Interest-
ingly, CSF production was significantly reduced in
AQP1-null mice, but only by 20–25%, indicating a
substantial contribution of extrachoroidal fluid
production by the brain parenchyma. This result
supported previous studies in which removal of
the choroid plexus in rhesus monkeys (Milhorat
et al., 1971) and dogs (Bering and Sato, 1963)
resulted in only a 30–50% decrease in the CSF
production.
 187

BRAIN SURFACE
Dura
Arachnoid Sinus Arachnoid granulation
2
Pia 1
AQP1
AQP4 Ependyma

Endothelial Astrocyte Choroid

cell foot processes plexus

EXTRACELLULAR
VENTRICLE
SPACE

PARENCHYMA P
V

Fig. 1. Aquaporins and brain water balance. The production and circulation of cerebrospinal fluid (CSF) in the brain is facilitated by
the highly polarized expression of aquaporin-1 (AQP1) and aquaporin-4 (AQP4). (1) Water transport from the vasculature into the
ventricle is facilitated by AQP1 expressed in the apical membrane of the choroid plexus, and via AQP4 in the ependymal lining.
(2) Another potential pathway for water flux into the subarachnoid space via the astrocytic processes lining the pial membrane that
heavily express AQP4. Fluid from the subarachnoid space is drained through the arachnoid granulations into the low-pressure venous
sinus that exits the cranium. (3) Extracellular fluid may alternatively pass through the perivascular processes of astrocytes as another
pathway to drain fluid into the vascular compartment.

The rates of CSF production and absorption
must be equal in the steady-state. Hydrocephalus
and increased ICP can occur when the balance
between CSF production and absorption is dis-
turbed. A reversal of CSF flow can be seen in ex-
perimental models of obstructive hydrocephalus,
where outflow-obstruction drives CSF into the
brain parenchyma causing extracellular edema
(Hiratsuka et al., 1982; Braun et al., 1997). Given
the widespread expression of AQP4 at the
brain–CSF interface, we examined whether this
water channel might facilitate the transparenchy-
mal absorption of CSF when the normal ventricu-
lar outflow pathways are blocked (Bloch et al.,
2006). Using a mouse model of obstructive
hyprocephalus, we found that AQP4-null mice
had an accelerated course of ventricular enlarge-
ment and ICP elevation compared with wildtype
mice, indicating a role for AQP4 in CSF absorp-
tion. We hypothesized that AQP4 most likely par-
ticipates in clearance of excess parenchymal fluid
seen in this model.
These phenotypic studies of mice lacking AQP1
and AQP4 provide evidence for the role of these
water channels in the physiology of brain water
homeostasis under normal and abnormal condi-
tions (Oshio et al., 2003; Manley et al., 2004). A
key feature in a number of neurological disorders
is an imbalance in brain water homeostasis that
leads to cerebral edema (Fishman, 1975). The re-
mainder of the review focuses on the compensa-
tory and maladaptive mechanisms linked to these
water channels and the discussion of their poten-
tial as targets for therapeutic interventions in brain
edema.
Cerebral edema
Cerebral edema is defined as an abnormal increase
in brain water content. Klatzo (1994) described
two distinct forms of cerebral edema based their
pathogenesis (Fig. 2): vasogenic (extracellular) and
cytotoxic (intracellular) brain edema. In most
188
A Cytotoxic
Swollen foot processes
TJ
Endothelial
cell
Capillary
lumen
Fig. 2. Principles of cytotoxic and vasogenic edema. (A) The pathologic swelling of astrocytic foot processes is the main contributor to
cytotoxic edema. The tight junctions (TJ) help to establish the blood-brain barrier (BBB). (B) The disruption of the BBB allows
extravasation of the fluid from the brain capillary into the extracelullar space resulting in vasogenic edema.
human disease, both forms of edema are found,
however one type typically dominates depending
on the particular disorder. However, these distinct
edema types can be purely created in experimental
models allowing independent study of the molec-
ular mechanisms involved in their evolution.
Vasogenic (extracellular) edema results from ex-
travasation of fluid secondary to defects in the
blood-brain barrier, and is seen in experimental
models of CNS inflammation (infection), tumors
(Davies, 2002), and cold injury (Chan et al., 1991;
Oury et al., 1993). Another mechanism of extra-
cellular edema is the pressure driven accumulation
of CSF in the brain interstitium of the periven-
tricular parenchyma as previously described for
obstructive hydrocephalus. In contrast, cytotoxic
edema is characterized by acute cellular swelling in
the presence of an intact blood-brain barrier. A
classic example of this edema type is modeled with
water intoxication. In this model, rapid intraperi-
toneal water infusion causes serum hyponatremia
that creates an osmotic gradient driving water
entry into the brain, resulting in pure cytotoxic
edema. Given the highly polarized expression of
AQP4 at the blood-brain barrier, and the bi-
directional nature of water flux through this chan-
nel, it is not surprising that AQP4 participates in
the molecular mechanisms of both types of edema.
B Vasogenic
Astrocyte
Capillary foot
lumen process
Endothelial cell
Aquaporins and cytotoxic edema
Although water intoxication produces a pure cel-
lular edema, in the brain the swelling is predom-
inantly localized to the glial processes around the
capillaries with sparing of the neurons (Wasterlain
and Torack, 1968; Kimelberg, 1995). Characteris-
tic swelling of astrocytic foot processes is also
found in brain tissue from head injured patients
(Bullock et al., 1991). The selective swelling of
astrocytes following a systemic perturbation in
osmolality raised questions about the mechanism
of this phenomenon (Kimelberg, 1995). Localiza-
tion studies of AQP4 suggested that it might play a
role in glial-specific swelling. Using the water in-
toxication model, our group demonstrated a sig-
nificant reduction in astrocytic foot process
swelling in AQP4-null mice (Manley et al., 2000).
This was associated with a decrease in brain water
content and a profound improvement in survival
(Fig. 3). Using a clinically relevant model of is-
chemic stroke, with its predominantly cytotoxic
edema, the AQP4-null mice also had decreased
cerebral edema and improved outcome.
Similar studies were conducted with the dystro-
phin null mdx-bgeo transgenic mouse and the
a-syntrophin null mouse (Frigeri et al., 2001; Vajda
et al., 2002; Amiry-Moghaddam et al., 2003).
 189

A B
100
AQP4+/+
0 min
80 AQP4−/− AQP4−/−
Survival (%)

60
15 min
40

20 AQP4+/+
30 min
0
0 20 40 60 80 100 1440
1.045 1.047 1.049 1.051 1.053
Time (min)
Specific gravity (g/cm3)

C D E
ICP cmH2O at 60min % Infused Non-Infused Water content
water (µL)
80
82 30

60

81
20

40

80
10
20

79
0 0
+/+ −/− +/+ −/−

Fig. 3. AQP4 deletion in models of cytotoxic (A, B) and vasogenic edema (C, D, E). AQP4 deletion improves short-term survival
following water intoxication produced by intraperitoneal fluid injection (A). Specific gravity of brain tissue display substantially higher
values for AQP4
/
mice 30 min after water intoxication (B), indicating reduced brain water content compared with AQP4+/+
littermates. The clearance of edema fluid is impaired in AQP4 null mice following intraparenchimal fluid injection. Higher ICP (C) and
brain water content (D) is seen in AQP4-null mice compared with AQP4 controls. Brain water content of AQP4-null mice was also
higher than the wildtype controls following brain abscess implantation, a clinically relevant model of vasogenic edema.

These mice have normal levels of AQP4 protein,
but because membrane localization of AQP4 re-
quires dystrophin-associated protein complex and
a-syntrophin (Frigeri et al., 2001; Neely et al.,
2001), AQP4 expression in the astrocytic foot
processes is reduced (Vajda et al., 2002; Amiry-
Moghaddam et al., 2003). As a result, these mice
are regarded as alternative models for the AQP4-
null genotype. In general, these studies have con-
firmed and extended our observations regarding
the role of AQP4 in cytotoxic edema (Vajda et al.,
2002; Amiry-Moghaddam et al., 2003).
To further clarify the role of AQP4 in astrocyte
swelling, osmotically induced volume changes
were assessed in primary astrocyte cultures de-
rived from AQP4-null mice (Solenov et al., 2004).
The measurement of rapid changes in cell volume
was possible by adapting and validating a method
of calcein quenching (Hamann et al., 2002) to
measure water permeability in astrocytes. We
found that water permeability (Pf) was decreased
by sevenfold in the AQP4-null astrocytes com-
pared with wildtype astrocytes. Using small inter-
fering RNA (siRNA) targeted to reduce the level
190
of AQP4 expression, a 50% reduction in apparent
water permeability of wildtype astrocytes was
found (Nicchia et al., 2003), which also caused a
68% inhibition of cell growth. In contrast, there
was no alteration in cell morphology or growth
characteristics in the primary cultures of the
AQP4-null astrocytes.
Astrocytes constitute the largest pool of cells in
the human brain, outnumbering neurons by at
least 10 to 1. Their numbers and ability to rapidly
swell in response to an osmotic gradient highlights
their importance in the pathogenesis of cellular
edema. Studies of astrocytes in which AQP4 has
been deleted or knocked down, confirm that AQP4
is the principal water channel in these cells. From a
clinical perspective, the phenotypic studies of
AQP4-null mice have clearly demonstrated that
deletion or reduction of AQP4 reduces edema and
improves neurological outcome in well-established
model where cytotoxic edema is in primary patho-
physiological mechanism. Together these studies
indicate that AQP4 inhibition may be a promising
target for drug discovery for the treatment of
cytotoxic cerebral edema.
Aquaporins and vasogenic edema
The key feature of vasogenic edema is the disrup-
tion of the blood-brain barrier and subsequent
leakage of the intravascular fluid into the extra-
cellular space of the brain parenchyma. The blood-
brain barrier is essential for the normal function of
the brain. It provides a highly selective barrier be-
tween the vascular space and the brain tissue that
allows for maintenance of the microenvironment
that is crucial for overall homeostasis of the brain.
The anatomical substrates of the blood-brain bar-
rier are the capillary endothelium and the astocytic
foot processes that surround the cerebral capillar-
ies (Reese and Karnovsky, 1967).
Unlike the microvasculature in the rest of the
body, the cerebral capillaries are not fenestrated
and contain few endocytic vesicles (Sage et al.,
1998) suggesting limited diffusion and transcellu-
lar transport. Molecular tracking studies show
that only small lipophilic molecules are allowed to
diffuse passively from the vascular bed into the
brain parenchyma, while all remaining molecules
are trafficked via active transport across the blood-
brain barrier (Rowland et al., 1992).
Surrounding the endothelia are the astrocytic
foot processes, or ‘‘end-feet’’ (Janzer and Raff,
1987). These highly specialized processes have
multiple ion transporters and channels that are
specifically targeted to the foot processes indicat-
ing that they also play a key role in blood-brain
barrier function. Membranes of astrocyte foot
processes also contain numerous arrays of intra-
membrane particles in freeze-fracture electron
micrographs (FFEM). These particles, which are
regular square arrays with a characteristic cobble-
stone pattern, are referred to as square arrays or
orthogonal arrays of particles (OAPs) (Landis and
Reese, 1974). Based on the finding that AQP4 was
present in the same cell types in which OAPs were
identified, it was proposed that AQP4 is an OAP
protein (Frigeri et al., 1995). Support for this
hypothesis came from FFEM studies of brain
from AQP4-null mice that were found not to have
OAPs (Verbavatz et al., 1997). Subsequent anti-
AQP4 antibody labeling of OAPs in brain tissue
confirmed that OAPs contain AQP4 (Rash et al.,
1998). The polarized expression and co-localiza-
tion of ion and water channels in the astrocyte foot
processes suggest that these structures may facil-
itate ion and water transport across the blood-
brain barrier.
Given that the formation of vasogenic edema is
primarily due to the disruption of the blood-brain
barrier and subsequent leakage of the intravascu-
lar fluid into the extracellular space, a role for as-
trocyte water channels in the formation of this
type of edema was not anticipated. However, re-
sults from multiple independent experiments sug-
gest a definite role for AQP4 in the clearance of
extracellular fluid, and the resolution of vasogenic
edema.
As with the CSF, the amount of interstitial fluid
present in extracellular space is determined by the
balance of fluid production and clearance from the
brain parenchyma. Similarly, when the clearance
or absorption is impaired, fluid accumulates and
ICP increases. To directly examine the role of

AQP4 in clearance of extracellular fluid, artificial
CSF was continuously infused into the brain
parenchyma of wildtype and AQP4-null mice
(Fig. 3). In comparison with the wildtype mice,
AQP4-null mice had significantly more brain water
and elevated ICP (Papadopoulos et al., 2004). Al-
though the exact mechanism is yet to be defined,
these results demonstrate that AQP4 is involved in
the clearance of extracellular fluid from the brain
parenchyma.
Our group has used a number of models where
vasogenic edema is the predominant form of ed-
ema, including cortical freeze injury, tumor im-
plantation, and brain abscess. Cortical freeze
injury lead to a marked disruption of the blood-
brain barrier in both wildtype and AQP4-null
mice, however AQP4-null mice had a significantly
greater increase in brain water content and ICP
(Papadopoulos et al., 2004). Peritumoral vaso-
genic edema was generated by implanting me-
lanoma cells into the striatum of AQP4-null and
wildtype mice. In accordance with the results from
previous models, there was higher ICP and accel-
erated neurological deterioration in the AQP4 de-
ficient mice (Papadopoulos et al., 2004). A similar
result was obtained with implantation of Staphylo-
coccus aureus to create brain abscesses in the stria-
tum of mice. In addition to increased peri-abscess
edema and increased ICP, AQP4 deficient mice
demonstrated a greater mortality despite equiva-
lent abscess size and cytokine levels (Bloch et al.,
2005). In all of the mentioned experiments, the
primary lesion size and degree of BBB permeabil-
ity were similar, indicating that the forces driving
the development of vasogenic edema were equiv-
alent in both groups and that differences were due
to impaired fluid clearance from the extracellular
space in AQP4-null mice.
Fluid reabsorption from the brain extracellular
space is thought to occur by bulk flow clearance
into the CSF and by transport back into cerebral
capillaries (Reulen et al., 1978; Marmarou et al.,
1994). Although a clear understanding of brain
water homeostasis remains to be fully elucidated,
several lines of evidence point towards an impor-
tant role for AQP4. In pathological states that in-
crease extracellular fluid, AQP4 expression has
191
been shown to be upregulated (Vizuete et al.,
1999); perhaps as an adaptive response to edema.
From a clinical perspective, the phenotypic studies
of AQP4-null mice indicate that deletion of AQP4
reduces the clearance of extracellular fluid and
improves outcome in models where vasogenic
edema is the primary pathophysiological mecha-
nism. Thus, AQP4 inhibition would not be indi-
cated for the treatment of vasogenic edema.
Whether or not strategies aimed at increasing
AQP4 expression or enhancing water permeability
will improve outcome from vasogenic edema, re-
main to be investigated.
Aquaporins, ICP, and traumatic brain injury
One of the more intriguing and clinically relevant
findings has been the observation that ICP is sig-
nificantly reduced in AQP1-null mice (Oshio et al.,
2005). ICP is the dynamic result of the interaction
of the CSF, cerebral blood, and brain tissue com-
partments. The individual effect of each compart-
ment on ICP is difficult to assess due to the
complexity of the physiologic relationship between
these compartments. However, if the intracranial
compliance is assumed to be constant, the
steady-state ICP can be expressed using the fol-
lowing equation introduced by Marmarou:
ICP ¼ If  Rout+Pss, where If is CSF formation
rate, Rout is outflow resistance, and Pss is sagittal
sinus pressure (Marmarou et al., 1978). As previ-
ously described, AQP1-null mice have a reduced
CSF formation rate that accounts for some of the
50% reduction in ICP seen in these mice. The de-
letion of AQP1 also leads to polyuria and hypo-
volemia due to its role in fluid absorption in the
kidney (Schnermann et al., 1998). We suspected
that this urinary concentrating defect might also
decreases the central venous pressure, which in
turn, is in continuity with the sagittal sinus venous
pressure. This was indeed the case, indicating that
the effects of AQP1 deletion on venous pressure
accounted for the balance of the reduction in ICP.
These results demonstrate that two different
AQP1-dependent mechanisms act synergistically
to significantly decrease in ICP.
192
Fig. 4. The beneficial effect of AQP1 deletion in brain injury. (A) The absence of AQP1 improves survival in focal brain injury. (B) The
ICP peak is reduced in AQP1-null mice following cold injury compared with wildtype controls.
ICP management has been the cornerstone of
traumatic brain injury care for many years. It is
well known that when ICP is elevated and cannot
be reduced, nearly all patients with traumatic
brain injury will die (Miller et al., 1981). All cur-
rent surgical and non-surgical treatments of ele-
vated ICP are directed towards decreasing the
volume in one of the three major compartments —
brain, blood, or CSF (Brain Trauma Foundation,
American Association of Neurological Surgeons
and Joint Section on Neurotrauma and Critical
Care, 1996). The ability of AQP1 deletion to re-
duce two of the three principal determinants of
steady-state ICP, CSF production, and venous
pressure, suggested that AQP1deletion might be
protective in a model of brain trauma (Fig. 4). In a
focal model of brain injury, the peak ICP was 60%
lower in AQP1-null mice compared with wildtype
littermates. In addition, a dramatic improvement
in survival rate was seen in the AQP1 deficient
mice: 84% compared with the 25% seen in wild
type controls. Currently, there is no consensus on
the treatment of elevated ICP. These results pro-
vide evidence for the importance of CVP and CSF
production on ICP under normal and pathological
conditions and suggest a novel application of
AQP1 inhibitors for pharmacological treatment of
elevated ICP.
Conclusion
Phenotypic analyses of AQP deficient mice have
allowed us to explore the role of these membrane
water channels in the mechanisms of cytotoxic ed-
ema, vasogenic edema, and CSF production.
These studies indicate that AQP4 plays significant
role in the development of cytotoxic edema and
the absorption of excess brain water resulting from
vasogenic edema. They also have demonstrated
the role of AQP1 in CSF production and mainte-
nance of steady-state ICP. The ability to modulate
water flux through AQP deletion has provided new
insights into brain water homeostasis and sug-
gested a number of new research directions. How-
ever, these efforts have not yet translated to the
treatment human clinical diseases. These advances
will require the development of AQP inhibitors
and activators to establish the benefit modulating
the function of these water channels.

References
Abbott, N.J. (2004) Evidence for bulk flow of brain interstitial
fluid: significance for physiology and pathology. Neurochem.
Int., 45(4): 545–552.
Agre, P., King, L.S., et al. (2002) Aquaporin water channels:
from atomic structure to clinical medicine. J. Physiol., 542
(Pt 1): 3–16.
Amiry-Moghaddam, M., Otsuka, T., et al. (2003) An alpha-
syntrophin-dependent pool of AQP4 in astroglial end-feet
confers bidirectional water flow between blood and brain.
Proc. Natl. Acad. Sci. U.S.A., 100(4): 2106–2111.
Bering, E. and Sato, O. (1963) Hydrocephalus: changes in for-
mation and absorption of cerebrospinal fluid within the cere-
bral ventricles. J. Neurosurg., 20: 1050.
Bloch, O., Auguste, K.I., Manley, G.T. and Verkman, A.S.
(2006) Accelerated progression of kaolin-induced hydroceph-
alus in aquaporin-4-deficient mice. J. Cereb. Blood Flow
Metab., 26(12): 1527–1537.
Bloch, O., Papadopoulos, M.C., et al. (2005) Aquaporin-4 gene
deletion in mice increases focal edema associated with
staphylococcal brain abscess. J. Neurochem., 95(1): 254–262.
Brain Trauma Foundation, American Association of Neuro-
logical Surgeons, Joint Section on Neurotrauma and Critical
Care. (1996) Guidelines for the management of severe head
injury. J. Neurotrauma, 13(11): 641–734.
Braun, K.P., Dijkhuizen, R.M., et al. (1997) Cerebral ischemia
and white matter edema in experimental hydrocephalus: a
combined in vivo MRI and MRS study. Brain Res., 757(2):
295–298.
Bullock, R., Maxwell, W.L., et al. (1991) Glial swelling follow-
ing human cerebral contusion: an ultrastructural study. J.
Neurol. Neurosurg. Psychiatry, 54(5): 427–434.
Chan, P.H., Yang, G.Y., et al. (1991) Cold-induced brain ed-
ema and infarction are reduced in transgenic mice overex-
pressing CuZn-superoxide dismutase. Ann. Neurol., 29(5):
482–486.
Davies, D.C. (2002) Blood-brain barrier breakdown in septic
encephalopathy and brain tumours. J. Anat., 200(6): 639–646.
Fishman, R.A. (1975) Brain edema. N. Engl. J. Med., 293(14):
706–711.
Frigeri, A., Gropper, M.A., et al. (1995) Localization of MIWC
and GLIP water channel homologs in neuromuscular, epi-
thelial and glandular tissues. J. Cell Sci., 108(Pt 9):
2993–3002.
Frigeri, A., Nicchia, G.P., et al. (2001) Aquaporin-4 deficiency
in skeletal muscle and brain of dystrophic mdx mice. FASEB
J., 15(1): 90–98.
Hamann, S., Kiilgaard, J.F., et al. (2002) Measurement of cell
volume changes by fluorescence self-quenching. J. Fluoresc.,
12: 139–145.
Hasegawa, H., Ma, T., et al. (1994) Molecular cloning of a
mercurial-insensitive water channel expressed in selected
water-transporting tissues. J. Biol. Chem., 269(8): 5497–5500.
Hiratsuka, H., Tabata, H., et al. (1982) Evaluation of periven-
tricular hypodensity in experimental hydrocephalus by metriz-
amide CT ventriculography. J. Neurosurg., 56(2): 235–240.
193
Janzer, R.C. and Raff, M.C. (1987) Astrocytes induce blood-
brain barrier properties in endothelial cells. Nature,
325(6101): 253–257.
Kimelberg, H.K. (1995) Current concepts of brain edema. Re-
view of laboratory investigations. J. Neurosurg., 83(6):
1051–1059.
Klatzo, I. (1994) Evolution of brain edema concepts. Acta
Neurochir. Suppl. (Wien.), 60: 3–6.
Landis, D.M. and Reese, T.S. (1974) Arrays of particles in
freeze-fractured astrocytic membranes. J. Cell Biol., 60(1):
316–320.
Manley, G.T., Binder, D.K., et al. (2004) New insights into
water transport and edema in the central nervous system
from phenotype analysis of aquaporin-4 null mice. Neuro-
science, 129(4): 983–991.
Manley, G.T., Fujimura, M., et al. (2000) Aquaporin-4 deletion
in mice reduces brain edema after acute water intoxication
and ischemic stroke. Nat. Med., 6(2): 159–163.
Marmarou, A., Hochwald, G., et al. (1994) Brain edema reso-
lution by CSF pathways and brain vasculature in cats. Am. J.
Physiol., 267(2 Pt 2): H514–H520.
Marmarou, A., Shulman, K., et al. (1978) A nonlinear analysis
of the cerebrospinal fluid system and intracranial pressure
dynamics. J. Neurosurg., 48(3): 332–344.
McIlwain, H. and Bachelard, H.S. (1985) Biochemistry and the
Central Nervous System. Churchill Livingstone, Edinburgh.
Milhorat, T., Hammock, M., et al. (1971) Cerebrospinal fluid
production by the choroid plexus and brain. Science,
173(994): 330–332.
Miller, J.D., Butterworth, J.F., et al. (1981) Further experience
in the management of severe head injury. J. Neurosurg.,
54(3): 289–299.
Neely, J.D., Amiry-Moghaddam, M., et al. (2001) Syntrophin-
dependent expression and localization of Aquaporin-4 water
channel protein. Proc. Natl. Acad. Sci. U.S.A., 98(24):
14108–14113.
Nicchia, G.P., Frigeri, A., et al. (2003) Inhibition of aquaporin-
4 expression in astrocytes by RNAi determines alteration in
cell morphology, growth, and water transport and induces
changes in ischemia-related genes. FASEB J., 17(11):
1508–1510.
Oshio, K., Song, Y., et al. (2003) Aquaporin-1 deletion reduces
osmotic water permeability and cerebrospinal fluid produc-
tion. Acta Neurochir. Suppl., 86: 525–528.
Oshio, K., Watanabe, H., et al. (2005) Reduced cerebrospinal
fluid production and intracranial pressure in mice lacking
choroid plexus water channel Aquaporin-1. FASEB J., 19(1):
76–78.
Oury, T.D., Piantadosi, C.A., et al. (1993) Cold-induced brain
edema in mice. Involvement of extracellular superoxide dis-
mutase and nitric oxide. J. Biol. Chem., 268(21):
15394–15398.
Papadopoulos, M.C., Manley, G.T., et al. (2004) Aquaporin-4
facilitates reabsorption of excess fluid in vasogenic brain ed-
ema. FASEB J., 18(11): 1291–1293.
Rash, J.E., Yasumura, T., et al. (1998) Direct immunogold
labeling of aquaporin-4 in square arrays of astrocyte and
194
ependymocyte plasma membranes in rat brain and spinal
cord. Proc. Natl. Acad. Sci. U.S.A., 95(20): 11981–11986.
Reese, T.S. and Karnovsky, M.J. (1967) Fine structural local-
ization of a blood-brain barrier to exogenous peroxidase. J.
Cell Biol., 34(1): 207–217.
Reulen, H.J., Tsuyumu, M., et al. (1978) Clearance of edema
fluid into cerebrospinal fluid. A mechanism for resolution of
vasogenic brain edema. J. Neurosurg., 48(5): 754–764.
Rowland, L.P., Fink, M.E., et al. (1992) Cerebrospinal fluid:
blood-brain barrier, brain oedema and hydrocephalus. In:
Kandel E.R., Schwartz J.H. and Jessel T.M. (Eds.), Princi-
ples of Neuroscience. Elsevier, New York, pp. 1050–1060.
Sage, M.R., Wilson, A.J., et al. (1998) Contrast media and
the brain. The basis of CT and MR imaging enhancement.
Neuroimaging Clin. N. Am., 8(3): 695–707.
Schnermann, J., Chou, C.-L., et al. (1998) Defective proximal
tubular fluid reabsorption in transgenic aquaporin-1 null
mice. PNAS, 95(16): 9660–9664.
Solenov, E., Watanabe, H., et al. (2004) Sevenfold-reduced os-
motic water permeability in primary astrocyte cultures from
AQP-4-deficient mice, measured by a fluorescence quenching
method. Am. J. Physiol. Cell Physiol., 286(2): C426–C432.
Speake, T., Freeman, L.J., et al. (2003) Expression of aquaporin
1 and aquaporin 4 water channels in rat choroid plexus.
Biochim. Biophys. Acta, 1609(1): 80–86.
Vajda, Z., Pedersen, M., et al. (2002) Delayed onset of brain
edema and mislocalization of aquaporin-4 in dystrophin-null
transgenic mice. Proc. Natl. Acad. Sci. U.S.A., 99(20):
13131–13136.
Verbavatz, J.M., Ma, T., et al. (1997) Absence of orthogonal
arrays in kidney, brain and muscle from transgenic knockout
mice lacking water channel aquaporin-4. J. Cell Sci., 110(Pt
22): 2855–2860.
Verkman, A.S. (2002) Physiological importance of aquaporin
water channels. Ann. Med., 34(3): 192–200.
Vizuete, M.L., Venero, J.L., et al. (1999) Differential upregu-
lation of aquaporin-4 mRNA expression in reactive as-
trocytes after brain injury: potential role in brain edema.
Neurobiol. Dis., 6(4): 245–258.
Wasterlain, C.G. and Torack, R.M. (1968) Cerebral edema
in water intoxication. II. An ultrastructural study. Arch.
Neurol., 19(1): 79–87.
Weed, L.H. and McKibben, P.S. (1919) Experimental alteration
of brain bulk. Am. J. Physiol., 48: 531–558.
Welch, K. (1963) Secretion of cerebrospinal fluid by choroid
plexus of the rabbit. Am. J. Physiol., 205(3): 617–624.
Wright, E.M., Wiedner, G., et al. (1977) Fluid secretion by the
frog choroid plexus. Exp. Eye Res., 25(Suppl): 149–155.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 13

Sodium channel expression and the molecular

pathophysiology of pain after SCI

Bryan C. Hains and Stephen G. Waxman

Department of Neurology and Center for Neuroscience and Regeneration Research, Yale University School of Medicine,
New Haven, CT 06510, USA; Rehabilitation Research Center, VA Connecticut Healthcare System, West Haven,
CT 06516, USA

Abstract: The chronic pain that develops as a result of spinal cord injury (SCI) is extremely debilitating and
remains largely unmanageable by current therapeutic strategies. Voltage-gated sodium channels regulate
the biophysical properties, and thus firing characteristics, of neurons. After SCI the repertoire of sodium
channels produced by dorsal horn nociceptive neurons is altered, enabling neurons to fire at higher than
normal rates in response to unchanged peripheral stimuli as well as to generate spontaneous discharges in
the absence of stimuli, resulting in the genesis of neuropathic pain. Our results have shown increased
expression of the Nav1.3 sodium channel in the spinal cord and thalamus. Nav1.3 upregulation allows
dorsal horn neurons to generate ramp currents, enhanced persistent currents, and shifts in steady-state
activation and inactivation. Further downstream, Nav1.3 causes increased spontaneous and evoked firing
of neurons in the ventroposterior lateral (VPL) nucleus of the thalamus. Nav1.3 also underlies changes in
burst firing properties of VPL neurons. The combination of spinal and thalamic generation and ampli-
fication of pain by Nav1.3 dysregulation contributes to post-SCI chronic pain. If proven to be similar in
humans, targeting of this system after SCI may offer hope for treatment of clinical pain.

Keywords: spinal cord injury; pain; sodium channels; thalamus; dorsal horn

Introduction Turner et al., 2001), even to a greater extent than

Although less apparent than paralysis, the preva-
lence of chronic pain disability after spinal cord
injury (SCI) is widespread. Sixty to eighty percent
of persons who have sustained SCI, regardless of
the level or completeness of lesion or of the type of
injury, experience clinically significant pain after
injury (Finnerup et al., 2001; Siddall et al., 2003).
The pain, which can have a pricking, burning, or
aching quality (Cairns et al., 1996), can be so se-
vere as to produce a drastic impairment in daily
routines and quality of life (Rintala et al., 1998;
Corresponding author. Tel.: +203-785-6351; Fax: +203-785-
7826; E-mail: stephen.waxman@yale.edu
the motor impairment (Mariano, 1992; Stormer
et al., 1997). Of patients with low thoracic or
lumbrosacral lesions, 37% are willing to trade
the possibility of recovery for pain relief
(Nepomuceno et al., 1979).
Nociceptive signals are transmitted through the
dorsal horn of the spinal cord. The spinal cord
dorsal horn contains the primary synapses through
which afferent somatosensory information, related
to touch, pressure, brush, temperature, and noxi-
ous stimuli, is received from the periphery. Dorsal
horn sensory neurons receive this information pri-
marily from the skin, perform a degree of process-
ing, and transmit signals through distinct tracts
within the spinal cord to supraspinal structures

DOI: 10.1016/S0079-6123(06)61013-3 195

196
where pain is interpreted and perceived. Within
each level of the nervous system including the spi-
nal cord, nociceptive signals are subject to a degree
of modulation by circuitry that it passes through.
Experimental SCI induces electrophysiological
changes in dorsal horn and thalamic sensory
neurons that contribute to pain-like behaviors
in animals. These include shifts in proportions of
cells responding to evoked noxious stimulation,
increases and irregularity in spontaneous back-
ground activity, increased evoked activity to (for-
merly) innocuous and noxious stimuli, increases
in after discharge activity following stimulation,
and emergence of abnormal burst firing.
Since the nature of the applied stimuli does not
change, other mechanisms must account for the
alterations in stimulus processing that lead to cen-
tral pain after SCI. Central changes in expression
of molecules such as ion channels, neurotransmit-
ters, and their receptors, contribute to altered sen-
sory processing by changing the electrophysiologic
excitability, and therefore output, of spinal sen-
sory neurons. Specific molecular changes in the
expression of voltage-gated sodium channels have
recently been shown by our laboratory to contrib-
ute to the development and maintenance of central
pain following SCI.
Spinal cord sodium channels
Action potential generation and propagation by
sensory neurons relies on multiple isoforms of
voltage-gated sodium channels (termed Nav). The
selective expression of ensembles of sodium chan-
nels, unique in different types of neurons, tunes the
biophysical properties of each cell. Within the
normal nervous system, properties fundamental to
neuronal function such as activation threshold,
inactivation, refractory period, rates of repriming,
and the ability to generate and conduct high-
frequency trains of action potentials all depend on
the type(s) of sodium channels expressed within a
given neuron (Waxman, 2000). As might be ex-
pected, dysregulation of channel expression can
abnormally reconfigure neuronal function in dis-
ease states.
Ten genes encode molecularly distinct voltage-
gated sodium channels, at least seven of which are
expressed in the rat nervous system. In the adult
spinal cord dorsal horn, Nav1.1, Nav1.2, and
Nav1.6 are strongly detectable, whereas Nav1.3
expression, while present at early ontogenetic
states, decreases with development and is nearly
undetectable in adults (Felts et al., 1997).
SCI and dorsal horn ion channel dysregulation
We have studied sodium channel expression at
several levels along the pain-signaling pathway in a
model of contusive SCI in which, as in humans
with non-penetrating SCI, there is central necrosis,
surrounded by a rim of surviving ascending and
descending axons at the level of the injury (Hains
et al., 2004). In this rodent model in situ hybrid-
ization and immunocytochemistry with an iso-
form-specific Nav1.3 antibody show that
expression of Nav1.3 is upregulated within dorsal
horn neurons caudal to the level of the injury
(Hains et al., 2003). These studies show that, 4
weeks after SCI, expression of Nav1.3 is increased,
within neurons contained within laminae I–VI in
the lumbar dorsal horn (L3–L5) (Fig. 1B). Nav1.3
does not co-localize with OX-42, a marker for ac-
tivated microglia (which proliferate after SCI and
are present in all laminae within the dorsal horn
(Hains and Waxman, 2006)), or with GFAP, a
marker for astrocytes (which also proliferate in all
spinal laminae after SCI). CGRP, restricted to the
terminals of primary afferent fibers within laminae
I–II, does not co-localize with Nav1.3. The sub-
stance-P receptor NK1R, found in second-order
nociceptive neurons within laminae I–IV, does
however co-localize with Nav1.3, indicating that
Nav1.3 is upregulated within nociceptive dorsal
horn neurons.
In this model, 88% of sampled lumbar dorsal
horn units show increased evoked activity to nat-
ural peripheral stimuli (brush, press, pinch, graded
von Frey filaments, and 471C thermal stimuli). In
comparison to intact animals in which dorsal horn
neurons fire at rates of 5–21 Hz, SCI animals
demonstrate increased evoked rates of up to
55–60 Hz (Fig. 1E).
 197

Fig. 1. Nav1.3 mRNA is expressed at low levels in naı̈ve animals (A), but is upregulated in lumbar dorsal horn neurons after spinal cord
injury (SCI) (B). Treatment with antisense (AS) oligodeoxynucleotides generated against Nav1.3 reduces Nav1.3 transcripts after SCI (C).
Corresponding unit recordings show evoked activity to peripheral stimulation (BR, brush; PR, press; PI, pinch, increasing strength von
Frey filament stimulation, and noxious thermal heating (471C)), after SCI (E) compared with controls (D). After Nav1.3 AS delivery,
evoked activity of dorsal horn neurons resembles that found in naı̈ve levels (F). Adapted with permission from Hains et al. (2003).

The Nav1.3 sodium channel plays a role in
maintaining neuronal hyperresponsiveness to
peripheral stimulation, as well as pain-related
behaviors after SCI, as evidenced through selec-
tive knock-down of Nav1.3 expression via anti-
sense (AS) oligodeoxynucleotide administration
(Hains et al., 2003). Lumbar intrathecal admin-
istration of Nav1.3 AS very effectively reduces
the expression of Nav1.3 within dorsal horn
neurons after SCI (Fig. 1C), and significantly
decreases the evoked hyperresponsiveness of
dorsal horn multi-receptive neurons (Fig. 1F),
whereas administration of Nav1.3 mismatch
(MM) sequence, used as a control, has no effect.
Electrophysiological recordings show that 4 days
after initiation of Nav1.3 AS, high levels of
evoked activity of dorsal horn neurons in SCI
animals are markedly decreased in response to all
peripheral stimuli, and only 20% of sampled units
are hyperexcitable.
198

Nav1.3 and neuronal hyperresponsiveness We recently examined the biophysical properties

Expression of the Nav1.3 sodium channel after
SCI uniquely configures neurons to fire in an ab-
normal, heightened manner. Nav1.3 produces a
rapidly repriming tetrodotoxin-sensitive sodium
current that promotes neuronal firing at higher-
than-normal frequencies (Cummins et al., 2001).
Since stimulus intensity is encoded in the dorsal
horn by the rate of firing, this change in neuronal
processing of afferent sensory information serves
to amplify incoming signals, such that perceived
pain thresholds are lowered after SCI.
of sodium currents in acutely dissociated dorsal
horn neurons after chronic SCI, and found a
number of changes in sodium current characteris-
tics via whole-cell patch-clamp recordings (Fig. 2A)
(Lampert et al., 2006). First, we found shifts in the
activation and inactivation properties of the volt-
age-gated sodium currents to more positive poten-
tials after SCI (Fig. 2B). Second, persistent sodium
currents in dorsal horn neurons from SCI animals
were increased when compared with cells from in-
tact animals (Fig. 2C). Third, the ramp current
elicited in response to slow depolarizing potentials

Fig. 2. Whole-cell patch-clamp recordings of acutely dissociated lumbar dorsal horn neurons 28 days after T9 SCI reveal changes in
sodium current properties. Representative current traces from intact and SCI animals, elicited by stepwise depolarizations from a
holding potential of
120 mV to voltages ranging form
110 to +40 mV (A). Steady-state inactivation (open squares: intact, filled
squares: SCI) measured as relative available current after a 500 ms prepulse to the indicated potential. Steady-state activation shown as
relative conductance (B). Note that SCI shifts steady-state inactivation in dorsal horn neurons toward the curve for axotomized DRG
neurons, in which Nav1.3 is upregulated. Representative current traces showing enhancement of persistent current after SCI (C).
Vertical lines delimit the region in which the mean persistent current was assessed. Inset: Representative record from a human
embryonic kidney cell heterologously transfected with Nav1.3, showing persistent current. Representative current traces in response to
a 200 ms voltage ramp from
120 to +30 mV showing an increase in ramp currents in SCI but not intact animals (D). Adapted with
permission from Lampert et al. (2006).
 199

was increased after SCI (Fig. 2D). The shift in
steady-state inactivation and the increase in the
persistent and ramp current all would be expected
to contribute to dorsal horn neuron hyperrespon-
siveness, which is observed after SCI. All of these
properties, which are consistent with upregulation
of Nav1.3, can contribute to hyperresponsiveness
of dorsal horn neurons, after SCI.
Thalamic Nav1.3 dysregulation
Dorsal horn neurons project rostrally within the
spinothalamic tract to third-order neurons of the
ventroposterior lateral (VPL) nucleus of the
thalamus (Jones, 1998). VPL neurons are involved
in sensory-discriminative aspects of pain process-
ing. Thalamic changes have been associated with
pain following SCI in humans (Lenz et al., 1989;
Pattany et al., 2002), primates (Weng et al., 2000),
and rodent models (Koyama et al., 1993; Gerke
et al., 2003), but the underlying molecular mech-
anisms are still not fully understood.
After SCI, expression of the Nav1.3 sodium
channel is upregulated in VPL neurons (Fig. 3B),
and an increase in the spontaneous activity and
responses to natural stimuli is observed (Hains
et al., 2005). Lumbar administration of Nav1.3 AS

Fig. 3. Expression of Nav1.3 protein within the ventral posterolateral (VPL) nucleus of the thalamus is absent in intact animals (A),
but 28 days after SCI Nav1.3 expression is upregulated within VPL and to some degree the ventral posteromedial (VPM) nuclei.
Intrathecal Nav1.3 antisense (AS) administration (C) significantly reduced the expression of Nav1.3 within the VPL (D). Long-term
recording of a single VPL unit after SCI with a hindpaw receptive field revealed high spontaneous discharge activity and evoked
responsiveness to brush (BR) stimulation of the receptive field (E). Following topical application of lidocaine (lido) and cord tran-
section (tx), evoked responses were abolished however high spontaneous activity persisted. Adapted with permission from Hains et al.
(2005).
200
reduced the number of neurons displaying Nav1.3
upregulation within the VPL (Fig. 3C), and re-
versed thalamic electrophysiologic abnormalities
caused by SCI. Importantly, we also observed an
increased level of spontaneous background activ-
ity in VPL neurons associated with SCI which
persists after spinal cord transection (Fig. 3E) that
disconnects the site of SCI and ascending projec-
tions from the lumbar enlargement from the
thalamus, indicating a degree of autonomous
hyperexcitability of the thalamus. We cannot ex-
clude the possibility that factors other than Nav1.3
also contribute to the reconfiguration of the firing
properties of thalamic units following SCI since
the magnitudes of change of Nav1.3 expression
levels did not perfectly match the magnitude of the
electrophysiologic changes. However, our results
demonstrate changes in sodium channel expression
within the thalamus that are associated with
abnormal sensory processing and chronic neuro-
pathic pain after SCI.
The thalamus also undergoes a change in burst
firing properties after SCI. Bursts are involved in
normal and pathological perceptual processing;
however, rhythmic oscillation of burst firing is ob-
served in pathophysiological conditions, and it has
been suggested that abnormal thalamic activity
may contribute to the perception of chronic pain
(Lenz et al., 1989, 1994; Jeanmonod et al., 1993;
McCormick, 1999). Furthermore, the information
content of bursts is higher than for single-spikes in
the visual system (Reinagel et al., 1999), and the
overall probability of generating at least one
postsynaptic spike (in the pain-processing sensory
cortex) is higher for bursts than for single-spikes
(Csicsvari et al., 1998). In neurons of the ventro-
basal thalamus, increased gain or transfer ratio
induced by burst firing results in increased tha-
lamocortical efficacy, enhancing the postsynaptic
response (Swadlow and Gusev, 2001). Thus, it is
possible that pathological burst firing after SCI
may more potently activate cortical circuits in-
volved in pain perception.
We recently showed that after SCI, burst firing
intervals become more regular, spike events are
reduced within each burst, acceleration in burst
duration occurs in bursts containing higher spike
counts, and shifts occur among spike firing modes
(Hains et al., 2006). Furthermore, we observed
that Nav1.3 AS returned the number of spikes/
burst, burst duration, and interburst interval to-
ward control levels following SCI. Our data are
similar to those reported by Lenz et al. (1989,
1994) in humans with post-SCI pain showing that
thalamic neurons exhibit oscillatory burst firing
characterized by high discharge rates, and decel-
eration of firing rate throughout the burst period,
and suggest that altered sodium channel expres-
sion within thalamic neurons contributes to these
functional abnormalities.
Multi-tiered alterations in nociceptive processing
Our findings demonstrate changes in excitability
and expression of Nav1.3 within dorsal horn and
VPL neurons following SCI. We have also ob-
served that hyperexcitability of thalamic neurons
following SCI is to at least some degree autono-
mous, since it persists following spinal cord tran-
section which abolishes ascending barrages from
spinal cord neurons near or below the injury site.
Together with our earlier results on dorsal root
ganglion neurons (Cummins and Waxman, 1997;
Cummins et al., 2001), these studies provide evi-
dence for a link between pain after SCI and mo-
lecular changes in pain-signaling neurons,
suggesting that dysregulation of sodium channel
Nav1.3 expression at both spinal and supraspinal
levels contribute to altered processing of somato-
sensory information and chronic neuropathic pain
after SCI. Specifically, our results indicate that
there is upregulated expression of Nav1.3 in first-,
second-, and third-order neurons of the pain path-
way after contusive SCI, and suggest that this
leads to enhanced neuronal excitability at each of
these multiple levels.
As originally proposed by Yezierski (2001), SCI
can trigger the activity of pain generators and
amplifiers within the CNS. On the basis of our
observations, we propose an integrated model
(Fig. 4) in which increased Nav1.3 expression
within pain-signaling neurons at multiple levels of
the neuraxis causes exaggerated nociceptive signa-
ling by way of spontaneous firing, a lowered
threshold for firing in response to synaptic drive,
 201

Fig. 4. Model whereby after SCI, multi-tiered dysregulation of sodium channel expression contributes to the pathological ampli-
fication and generation of aberrant nociceptive information, and ultimately chronic pain. In comparison to intact animals (A), second-
order neurons within the dorsal horn increase the gain with which they respond to innocuous and noxious inputs (10 ) after SCI (B). The
exaggerated responses of these dorsal horn neurons in response to peripheral stimulation poise them to act as an amplifier of normal as
well as inappropriate pathological nociceptive signals (20 ). After SCI, the thalamus also acts as a pain amplifier by responding in an
increased manner to nociceptive information relayed from the dorsal horn, and in addition, thalamic circuitry can act as an au-
tonomous pain signal generator (30 ). Treatment strategies that target one or more components of this pathway may be of value in
combating chronic pain following SCI.

and increased synaptic drive due to enhanced ex-
citability of presynaptic elements. According to
this model, second-order neurons within the dorsal
horn increase the gain with which they respond to
innocuous and noxious inputs after SCI. The ex-
aggerated responses of these dorsal horn neurons
configures them to act as an amplifier of normal as
well as inappropriate pathological nociceptive sig-
nals.
Because the thalamus is involved not only in
relaying, but also processing incoming information
from the spinal cord en route to the cortex, injury-
induced changes in spinal pain generator circuitry
may feed aberrant signals into the injured thala-
mus which further processes and amplifies the sig-
nals before relaying messages that are interpreted
as signaling pain to suprathalamic structures
(Waxman and Hains, 2006). Abnormal processing
at thalamic levels would then be expected to fur-
ther exaggerate abnormal firing patterns from the
spinal cord after SCI. Our observations of in-
creased primary burst firing activity and reduced
silence in VPL neurons after SCI leads us to sug-
gest that, following SCI, an increased level of
abnormal afferent firing is being forwarded to
cortical structures involved in interpreting pain.
Conclusion
SCI induces nervous-system-wide changes in neu-
ronal firing properties which contribute to chronic
pain, some of which are governed by the patho-
logical expression of the Nav1.3 sodium channel.
202
By prevention of the upregulation of Nav1.3,
knock-down of Nav1.3 after it is upregulated, or
selective pharmacological targeting of Nav1.3
channels deployed after injury, it is possible that
the molecular amplifiers and generators of chronic
pain associated with SCI may be effectively tar-
geted and muted so that chronic pain can be
ameliorated.
Acknowledgments
The authors thank Dr. Joel Black for valuable
experimental advice. This work was supported in
part by grants from the Medical Research Service
and Rehabilitation Research Service, Department
of Veterans Affairs, and the National Multiple
Sclerosis Society. The Center for Neuroscience and
Regeneration Research is a Collaboration of the
Paralyzed Veterans of America and the United
Spinal Association. BCH was funded by The
Christopher Reeve Paralysis Foundation (HB1-
0304-2), the NIH/NINDS (1 F32 NS046919-01),
and Pfizer (Scholar’s Grant in Pain Medicine).
References
Cairns, D.M., Adkins, R.H. and Scott, M.D. (1996) Pain and
depression in acute traumatic spinal cord injury: origins of
chronic problematic pain? Arch. Phys. Med. Rehab., 77:
329–335.
Cummins, T.R., Aglieco, F., Renganathan, M., Herzog, R.I.,
Dib-Hajj, S.D. and Waxman, S.G. (2001) Nav1.3 sodium
channels: rapid repriming and slow closed-state inactivation
display quantitative differences after expression in a mam-
malian cell line and in spinal sensory neurons. J. Neurosci.,
21: 5952–5961.
Cummins, T.R. and Waxman, S.G. (1997) Downregulation of
tetrodotoxin-resistant sodium currents and upregulation of a
rapidly repriming tetrodotoxin-sensitive sodium current in
small spinal sensory neurons after nerve injury. J. Neurosci.,
17: 3503–3514.
Csicsvari, J., Hirase, H., Czurko, A. and Buzsaki, G. (1998)
Reliability and state dependence of pyramidal cell-interneu-
ron synapses in the hippocampus: an ensemble approach in
the behaving rat. Neuron, 21: 179–189.
Felts, P.A., Yokoyama, S., Dib-Hajj, S., Black, J.A. and Wax-
man, S.G. (1997) Sodium channel alpha-subunit mRNAs I,
II, III, NaG, Na6 and hNE (PN1): different expression pat-
terns in developing rat nervous system. Mol. Brain Res., 45:
71–82.
Finnerup, N.B., Johannesen, I.L., Sindrup, S.H., Bach, F.W.
and Jensen, T.S. (2001) Pain and dysesthesia in patients
with spinal cord injury: a postal survey. Spinal Cord, 39:
256–262.
Gerke, M.B., Duggan, A.W., Xu, L. and Siddall, P.J. (2003)
Thalamic neuronal activity in rats with mechanical allodynia
following contusive spinal cord injury. Neuroscience, 117:
715–722.
Hains, B.C., Klein, J.P., Saab, C.Y., Craner, M.J., Black, J.A.
and Waxman, S.G. (2003) Upregulation of sodium channel
Nav1.3 and functional involvement in neuronal hyperexcit-
ability associated with central neuropathic pain after spinal
cord injury. J. Neurosci., 23: 8881–8892.
Hains, B.C., Saab, C.Y., Lo, A.C. and Waxman, S.G. (2004)
Sodium channel blockade with phenytoin protects spinal
cord axons, enhances axonal conduction, and improves func-
tional motor recovery after contusion SCI. Exp. Neurol., 188:
365–377.
Hains, B.C., Saab, C.Y. and Waxman, S.G. (2005) Changes in
electrophysiological properties and sodium channel Nav1.3
expression in thalamic neurons after spinal cord injury.
Brain, 128: 2359–2371.
Hains, B.C., Saab, C.Y. and Waxman, S.G. (2006) Alterations
in burst firing of thalamic VPL neurons and reversal by
Na(v)1.3 antisense after spinal cord injury. J. Neurophysiol.,
95: 3343–3352.
Hains, B.C. and Waxman, S.G. (2006) Activated microglia
contribute to the maintenance of chronic pain after spinal
cord injury. J. Neurosci., 26: 4308–4317.
Jeanmonod, D., Magnin, M. and Morel, A. (1993) Thalamus
and neurogenic pain: physiological, anatomical, and clinical
data. Neuroreport, 4: 475–478.
Jones, E.G. (1998) Viewpoint: the core and matrix of thalamic
organization. Neuroscience, 85: 331–345.
Koyama, S., Katayama, Y., Maejima, S., Hirayama, T., Fujii,
M. and Tsubokawa, T. (1993) Thalamic neuronal hyperac-
tivity following transection of the spinothalamic tract in the
cat: involvement of N-methyl-D-aspartate receptor. Brain
Res., 612: 345–350.
Lampert, A., Hains, B.C. and Waxman, S.G. (2006) Upregu-
lation of persistent and ramp sodium current in dorsal horn
neurons after spinal cord injury. Exp. Brain Res., 174:
660–666.
Lenz, F.A., Kwan, H.C., Dostrovsky, J.O. and Tasker, R.R.
(1989) Characteristics of the bursting pattern of action po-
tentials that occurs in the thalamus of patients with central
pain. Brain Res., 496: 357–360.
Lenz, F.A., Kwan, H.C., Martin, R., Tasker, R., Richardson,
R.T. and Dostrovsky, J.O. (1994) Characteristics of somato-
topic organization and spontaneous neuronal activity in the
region of the thalamic principal sensory nucleus in patients
with spinal cord transection. J. Neurophysiol., 72: 1570–1587.
Mariano, A.J. (1992) Chronic pain and spinal cord injury. Clin.
J. Pain, 8: 87–92.
McCormick, D.A. (1999) Are thalamocortical rhythms the
Rosetta Stone of a subset of neurological disorders? Nat.
Med., 5: 1349–1351.

Nepomuceno, C., Fine, P.R., Richards, J.S., Gowens, H.,
Stover, S.L., Rantanuabol, U. and Houston, R. (1979) Pain
in patients with spinal cord injury. Arch. Phys. Med. Re-
habil., 60: 605–609.
Pattany, P.M., Yezierski, R.P., Widerstrom-Noga, E.G., Bowen,
B.C., Martinez-Arizala, A., Garcia, B.R. and Quencer, R.M.
(2002) Proton magnetic resonance spectroscopy of the
thalamus in patients with chronic neuropathic pain after spi-
nal cord injury. Am. J. Neuroradiol., 23: 901–905.
Reinagel, P., Godwin, D., Sherman, S.M. and Koch, C. (1999)
Encoding of visual information by LGN bursts. J. Neuro-
physiol., 81: 2558–2569.
Rintala, D.H., Loubser, P.G., Castro, J., Hart, K.A. and
Fuhrer, M.J. (1998) Chronic pain in a community-based
sample of men with spinal cord injury: prevalence, severity,
and relationship with impairment, disability, handicap, and
subjective well-being. Arch. Phys. Med. Rehabil., 79:
604–614.
Siddall, P.J., McClelland, J.M., Rutkowski, S.B. and Cousins,
M.J. (2003) A longitudinal study of the prevalence and char-
acteristics of pain in the first 5 years following spinal cord
injury. Pain, 103: 249–257.
Stormer, S., Gerner, H.J., Gruninger, W., Metzmacher, K.,
Follinger, S., Wienke, C., Aldinger, W., Walker, N., Zimmer-
mann, M. and Paeslack, V. (1997) Chronic pain/
203
dysaesthesiae in spinal cord injury patients: results of a multi-
centre study. Spinal Cord, 35: 446–455.
Swadlow, H.A. and Gusev, A.G. (2001) The impact of ‘burst-
ing’ thalamic impulses at a neocortical synapse. Nat. Neuro-
sci., 4: 402–408.
Turner, J.A., Cardenas, D.D., Warms, C.A. and McClellan,
C.B. (2001) Chronic pain associated with spinal cord injuries:
a community survey. Arch. Phys. Med. Rehabil., 82:
501–509.
Waxman, S.G. (2000) The neuron as a dynamic electrogenic
machine: modulation of sodium-channel expression as a basis
for functional plasticity in neurons. Philos. Trans. R. Soc.
Lond. B. Biol. Sci., 355: 199–213.
Waxman, S.G. and Hains, B.C. (2006) Fire and phantoms after
spinal cord injury: Na+ channels and central pain. Trends
Neurosci., 29: 207–215.
Weng, H.R., Lee, J.I., Lenz, F.A., Schwartz, A., Vierck, C.,
Rowland, L. and Dougherty, P.M. (2000) Functional plas-
ticity in primate somatosensory thalamus following chronic
lesion of the ventral lateral spinal cord. Neuroscience, 101:
393–401.
Yezierski, R.P. (2001) Pain following spinal cord injury: patho-
physiology and central mechanisms. In: Sandkühler J., et al.
(Eds.), Progress in Brain Research. Elsevier, Amsterdam,
pp. 429–449.
This page intentionally left blank
 SECTION IV

Novel Aspects of Clinical Research in CNS Injury

This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 14

Monitoring cerebral oxygenation in traumatic

brain injury

Iain K. Haitsma and Andrew I.R. Maas

Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands

Abstract: Ischemia is a common problem after traumatic brain injury (TBI) that eludes detection with
standard monitoring. In this review we will discuss four available techniques (SjVO2, PET, NIRS and
PbrO2) to monitor cerebral oxygenation. We present technical data including strengths and weaknesses of
these systems, information from clinical studies and formulate a vision for the future.

Keywords: TBI; PbrO2; brain tissue oxygen tension; SjVO2; PET; NIRS; tissue oxygen reactivity; advanced
neuromonitoring

Introduction (dis)advantages, the available clinical data with

Treatment of patients with severe traumatic brain
injury (TBI) is largely focused on the prevention of
secondary insults. Standard monitoring consists
of measuring intra cranial pressure (ICP),
(mean) arterial blood pressure (MAP) and thus
calculating the cerebral perfusion pressure
(CPP ¼ MAP
ICP). This has led to ICP and
CPP driven treatment protocols (Rosner et al.,
1995; Nordstrom, 2005). However these measure-
ments do not give further information on the
existence of ischemia which is common in post
mortem studies after TBI (Graham et al., 1978,
1989). Much effort has gone into the development
of extra monitoring of the injured brain, so called
multi modal monitoring (Unterberg et al., 1997;
Meixensberger et al., 1998; Mulvey et al., 2004;
De Georgia and Deogaonkar, 2005; Vespa, 2005).
This review will focus on monitoring oxygenation
in the injured brain. We will discuss the available
technologies with special emphasis on brain
tissue oxygen tension measurement, their
Corresponding author. Tel.: +31-10-463-4077; Fax: +31-10-
463-4075; E-mail: airmaas@erasmusmc.nl
results of interventions with CPP and respiratory
variations and will conclude summarizing our view
of the future with these new technologies.
Monitoring technologies
Global vs. focal monitoring
The different measurement modalities have differ-
ent working areas. Some work very focally (e.g.
brain tissue oxygen tension measurement) others
globally (e.g. jugular bulb oxygenation). This leads
to differences in their use and interpretation of
their values (Sarrafzadeh et al., 1998; Gopinath
et al., 1999; Bellander et al., 2004; Engstrom et al.,
2005). The advantage of a global measurement is
the information on a large part of the injured
brain, more so because most treatments in TBI are
systemic.
The more focal measurements have been typi-
cally used in two ways. Either they have been ap-
plied to non-injured areas in essence using them
as global monitors or to the penumbra of focal
lesions, for instance contusions. The second

DOI: 10.1016/S0079-6123(06)61014-5 207

208
Table 1. Information regarding measurement technologies
Technique Catheter location
Jugular bulb oximetry Intravenous
PET Extracranial
Near infra red spectroscopy Extracranial
Brain tissue oxygen tension Intraparenchymal
approach will provide earlier information on the
state of the tissue most threatened at that time.
However there are practical problems in defining
the penumbra zone, in accurately positioning the
catheter and finally in interpreting the absolute
values measured.
The data regarding the different systems com-
bined with their ability to measure continuously or
intermittently are summarized in Table 1.
Jugular bulb oximetry
The jugular bulb, the cranial tip of the internal
jugular vein, contains pure cerebrovenous blood.
Back in the early 1900s the jugular vein was
pierced to gain information on cerebrovenous
oxygenation (White and Baker, 2002). Later in
the 1940s it was used for calculating CBF by Kety
and Schmidt (1945). Since the early 1990s litera-
ture has accumulated on the value of measuring
the oxygen saturation of jugular blood (Robertson
et al., 1992; Cruz, 1993a; De Deyne, 1999; White
and Baker, 2002).
The measurement can be done with intermittent
blood gas samples or continuously by inserting
a fiberoptic catheter. The side of cannulation is
debated. Some propose to use the right side, as the
right jugular bulb is usually dominant, others test
for dominancy of a side, but alternatively the
choice can be made to measure on the side that is
most injured. Variability exists if both sides are
measured continuously (Metz et al., 1998). This of
course raises questions on whether the values can
really be interpreted as indicative of the whole
brain. Technical problems have been reported
with continuous measurements with time of
good quality ranging from 43 to 91.5% (Cruz,
Global vs. focal Continuous or intermittent
Global Continuous or intermittent
Global and focal Intermittent
Focal Continuous or intermittent
Focal Continuous
1993b; Fortune et al., 1994; Kiening et al., 1996;
Meixensberger et al., 1998; Gopinath et al., 1999).
One article is devoted to possible technical prob-
lems with jugular oximetry (Dearden and Midgley,
1993). Others claim to have little difficulties with
jugular venous oxygen saturation (SjVO2) meas-
urements (Cruz, 1997). Apart from measuring
jugular oxygen saturation further information can
be gained by calculating the arteriovenous differ-
ence in oxygen content (AVDO2) or even cerebral
extraction of oxygen (Le Roux et al., 1997; Cruz,
1998).
In conclusion, it can be stated that jugular
oximetry is a relatively invasive mode of measure-
ment that can be subject to technical difficulties,
the best side of measurement is the topic of debate.
It can give valuable information on oxygen con-
sumption either continuously or intermittent.
PET scan
Using 15O2 as an inhalant information can be
gained on oxygen consumption using positron
emission tomography (PET) scanning. This
method is non-invasive, it does however expose
the patient to radiation and more importantly de-
pends on the patient being out of his relatively safe
ICU environment for a longer period. Other
downsides include the intermittent character and
the limited availability of this technique. Informa-
tion can be gained on cerebral blood flow (CBF),
oxygen extraction fraction (OEF) and cerebral
metabolic rate of O2 (CMRO2). The calculations
are regional, but the whole brain can be measured.
Using voxel-based analysis the ischemic burden
can be quantified (Coles et al., 2004b).

Near infra red spectroscopy (NIRS)
NIRS is a non-invasive measurement of Hb oxy-
genation. As the name implies it uses near infrared
light that can cross the skull. Hb is a strong
absorber of near infrared light and the amount of
absorption depends on the degree of Hb oxygen-
ation. Information can be supplied in the percent-
age of HbO2 saturation, or absolute changes in
HbO2, deoxyhemoglobin and total hemoglobin
in micromolar concentrations, referenced to an
arbitrary baseline (Cho et al., 2000). Technical
problems have precluded wide scale use of this non-
invasive measurement. If the source and detector
are not far enough apart, measurements are influ-
enced by subcutaneous blood flow. With complete
ischemia near normal values have been found, pos-
sibly related to venous blood sequestered in the
infracted tissue (Gomersall et al., 1997).
Because of these technical problems, especially
in TBI patients, NIRS is currently rarely used in
adults. With recent modifications its use might
increase again (Al-Rawi, 2005). Technical limita-
tions are less relevant in neonates.
Brain tissue oxygen tension (PbrO2)
Following extensive pre-clinical work, the intro-
duction of clinical PbrO2 measurements was pio-
neered by the Rotterdam group (Maas et al., 1993)
and the technique formally introduced during the
first congress on cerebral oxygenation organized
by the Rotterdam, Berlin and Wurzburg groups
in 1995. Initial publications followed rapidly
(Kiening et al., 1996; van Santbrink et al., 1996;
Meixensberger et al., 1998). Many articles includ-
ing review articles are available on this subject
(Andrews, 2001; Haitsma and Maas, 2002; Bader
et al., 2003; Littlejohns et al., 2003; De Georgia
and Deogaonkar, 2005; Vespa, 2005; Rose et al.,
2006). Most reports use one of the two technolo-
gies — the Licox system (Integra lifesciences,
Plainsboro, NJ) or the Para/Neurotrend system
(Charbel et al., 1997) (Diametrics Medical Ltd.,
UK). Unfortunately the Para/Neurotrend system
is no longer available. Both techniques require
intraparenchymal insertion of the 0.5–0.8 mm
209
thick probes, usually via a bolt system. The Licox
system uses a Clark type electrode, which gener-
ates a current that is dependent on the amount of
oxygen near the catheter tip. The Paratrend also
uses a Clark electrode and later changes to an
optode system in the Neurotrend. The Para/
Neurotrend supplies information on local temper-
ature, pCO2 and pH as well. The Licox now also
incorporates a temperature sensor. Complication
rates are equally low as in the use of intraparen-
chymal ICP monitors and data are reliable with
very low catheter drift rates (Kiening et al., 1996;
Dings et al., 1998; van den Brink et al., 2000;
Hoelper et al., 2005). Time of good data quality
of up to 95% has been reported (Meixensberger
et al., 1998). Catheters are usually inserted in the
frontal white matter and give continuous informa-
tion on the local balance between oxygen supply
and uptake. After insertion, a run in time of 1–2 h
is advised before reliable data are produced (Dings
et al., 1998; van den Brink et al., 2000). We reco-
mmend performing an oxygen challenge after
catheter insertion to verify its proper placement,
as even small hematomas around the catheter tip
can lower the oxygen readings, which would result
in no or very little increase in PbrO2 with a large
increase in inspiratory oxygen fraction (FiO2) or
more appropriate arterial oxygen tension (PaO2)
(van den Brink et al., 1998).
In conclusion, PbrO2 measurement is an inva-
sive, but safe and reliable measurement that gives
continuous information on local oxygenation.
Clinical data
Jugular oximetry
Low SjVO2 values (below 50–55%) are related to
poorer outcome (Gopinath et al., 1994; Fandino
et al., 2000; Perez et al., 2003). The same holds
for values higher than 75% (Cormio et al., 1999;
Macmillan et al., 2001). Others have challenged
these classic cut-off values after measurements in
patients with Cushing syndrome (44.7–69.5%)
(Chieregato et al., 2003). A limited improvement
as opposed to clear improvement in increased
210
AVDO2 after treatment is also related to poorer
prognosis (Le Roux et al., 1997).
In two important publications the Houston
group compared ICP (CPP50 mmHg, PaCO2
25–30 mmHg) and CBF (CPP70 mmHg, PaCO2
35 mmHg) regimens to each other (Robertson
et al., 1999; Contant et al., 2001). In the CBF
group jugular desaturations were less frequent
(30% vs. 50.6%). No difference in outcome was
observed, possibly because benefits of treatment
of jugular desaturations were offset by a fivefold
increase in ARDS in the CBF group. These pub-
lications led to lowering of the recommended CPP
level in the guidelines from 70 to 60 mmHg.
Hyperventilation leads to a decrease in SjVO2
and increase in AVDO2, these effects however can
be offset by hyperoxygenation (Matta et al., 1994;
Thiagarajan et al., 1998). A treatment protocol
using the cerebral extraction of oxygen as guide for
treatment including targeted hyperventilation has
shown promising results in adults with TBI and
diffuse brain swelling and separately in children
with severe TBI and increased ICP (Cruz, 1998;
Cruz et al., 2002).
In TBI patients with decreased CBF hyperbaric
O2 therapy can increase CBF and CMRO2 at 1
and 6 h after treatment, at constant AVDO2
(Rockswold et al., 2001). Increased ICP and CSF
lactate levels were also lowered.
PET measurements
Even with PET measurements debate exists on the
frequency of regional ischemia (Coles et al., 2004a;
Vespa et al., 2005). Ischemia was defined as the
oxygen extraction fraction at a cerebral venous
oxygen (CVO2) content below 3.5 ml/100 ml. One
study reports ischemic volume ranging from 1 to
16% (Coles et al., 2004a), the other maximally 1%
of brain volume in TBI patients (Vespa et al.,
2005).
Increasing CPP from 70 to 90 mmHg produces
a significant fall in ischemic brain volume (IBV)
which is most pronounced in patients with a larger
IBV (Coles et al., 2004c).
A study published in 2000 found a decrease
in CBF but no change in CMRO2 with
hyperventilation (Diringer et al., 2000). In a study
with 13 TBI patients receiving hyperventilation,
CBF was reduced to even below 20 ml/100 g/min,
but no changes in oxygen extraction fraction were
noted. It was concluded that no energy failure
took place during hyperventilation (Diringer et al.,
2002). Another study confirmed that hyperventi-
lation lowered CBF but not to the level of ischemia
(Coles et al., 2002).
NIRS
Few clinical studies with regard to oxygenation
measurements have been published on NIRS.
Interesting is its possible use as a detector of
intracranial hematomas. In the presence of a sub-
dural hematoma or bloody contusion, or even in
the event of massive blood in the subarachnoid
space, absorption is high, thus preventing suffi-
cient light to return to the detectors. This phe-
nomenon may be used to detect delayed traumatic
hematomas, leading to better timed follow-up CTs
and operations (Gopinath et al., 1993, 1995).
PbrO2
Low PbrO2 values are observed in up to 50% of
patients during the first 24 h with depth and du-
ration of tissue hypoxia being related to outcome
and being an independent predictor of unfavorable
outcome and death (Bardt et al., 1998; Valadka
et al., 1998; van den Brink et al., 2000; Stiefel et al.,
2006). No absolute value can be given as a cut-off
point for tissue hypoxia but values ranging from
10 to 20 mmHg in uninjured frontal white matter
are mentioned based on prognostic analysis or
related to CBF values (Zauner et al., 1997; Bardt
et al., 1998; Doppenberg et al., 1998b; Valadka
et al., 1998; van den Brink et al., 2000).
A sudden fall in PbrO2 to zero, without the
development of a hematoma around the catheter,
is considered an early sign of brain death (van
Santbrink et al., 1996; Palmer and Bader, 2005).
Recently interest has increased on testing auto-
regulation (Mascia et al., 2000; Czosnyka et al.,
2001; Steiner et al., 2002; Cremer et al., 2004).
Patients with mean CPP levels close to their

optimal CPP are more likely to have a favorable
outcome (Steiner et al., 2002). The response of
PbrO2 to changes in CPP is also dependent on the
state of autoregulation (Lang et al., 2003; Cremer
et al., 2004). With increased ICP a CPP reduction
below 77 mmHg can lead to a further ICP increase
and decrease in PbrO2 (Cremer et al., 2004).
Higher CPP levels might be useful in normalizing
tissue oxygenation in ischemic areas (Stocchetti
et al., 1998). A historical case control study by the
Leipzig group from 2003 with increased CPP for
low PbrO2 values showed less frequent cerebral
hypoxia, but no difference in 6-month outcome
(Meixensberger et al., 2003).
One publication by the Pennsylvania group
claimed reduced patient death using both ICP
and PbrO2 monitoring (Stiefel et al., 2005). This
was based on a historical case control study, in
which patients with ICP and PbrO2 monitoring
first were treated according to an ICP/CPP proto-
col, later patients were entered in a stepwise pro-
tocol that also treated PbrO2 values below
25 mmHg. Mortality rate dropped from 44 to
25%. The major critique on this study concerns its
retrospective nature.
Hyperventilation usually lowers PbrO2 (Schneider
et al., 1998), and reactivity is higher in patients
with poorer prognosis, statistically significant on
day 5 after injury (Carmona Suazo et al., 2000).
Increases in PbrO2 with hyperventilation in TBI
have however also been reported (Dings et al.,
1996; Imberti et al., 2000).
PbrO2 can be increased by increasing FiO2/
PaO2.
Tissue oxygen response (TOR) can be calculated
as:
PbrO2 1
TOR ¼  et al., 2005).
PaO2 PbrO2
Higher values of TOR are related to poorer
outcome and may reflect a loss of oxygen auto-
regulation (van Santbrink et al., 1996, 2003).
Several groups have investigated the effects of
increasing FiO2/PaO2 to supranormal levels. In
1999 the Richmond group reported a fall in brain
lactate levels measured by microdialysis after
administering 100% O2 (Menzel et al., 1999a, b).
These results were not without controversy (Rossi
211
and Stocchetti, 1999; Bressack and Schiffman,
2003). A report by the Milan group showed re-
duced lactate levels, but not the lactate–pyruvate
ratio with hyperoxia which was interpreted as ab-
sence of improvement in cerebral metabolism
(Magnoni et al., 2003). A more recent report by
the Richmond and Bern groups also demonstrated
a fall in lactate–pyruvate ratio after hyperoxia
compared to a historical control group (Tolias
et al., 2004). Nevertheless, even oxygen can be
considered a drug and is not without its side effects
(Lodato, 1990). Further reports on outcome meas-
ures with hyperoxia treatment are eagerly awaited.
PbrO2 decreases with hypothermia with an extra
reduction below 351C (Gupta et al., 2002) in one
study, others report an increase in PbrO2 with
hypothermia (Zhi et al., 1999; Jia et al., 2005).
Decompressive craniectomies for intractable
ICP leads to increased PbrO2 (Jaeger et al., 2003;
Stiefel et al., 2004).
Technology comparisons
SjVO2 values correlate with IBV, defined as the
area with CVO2 content p3.5 ml/100 ml as meas-
ured by PET scan, with SjVO2 o50% occurring at
IBV at 1375% (Coles et al., 2004a).
In a study in 14 patients NIRS detected twice
as many events as SjVO2 monitoring although
the clinical significance of this is not known
(Kirkpatrick et al., 1995). Another study found
poor correlation between NIRS and SjVO2 mon-
itoring even with SjVO2 desaturations below 55%
(Lewis et al., 1996). In a study in 60 children
without TBI, SjVO2 and NIRS correlated signifi-
cantly but NIRS had lower sensitivity (Nagdyman
SjVO2 and PbrO2 correlate closely especially
when CPP falls below 60 mmHg (Kiening et al.,
1996). This correlation is less in areas with focal
pathology (Gupta et al., 1999). Correlation be-
tween SjVO2 and PbrO2 during CO2 reactivity
testing is low (Fandino et al., 1999).
In a study in 36 TBI patients moderate hyper-
ventilation reduced SjVO2, but not below 50%
with PbrO2 values below 10 mmHg (Imberti et al.,
2002), in some tests SjVO2 and PbrO2 changed in
212
different directions, supporting the concept that
these technologies are complementary.
In a 2002 study PbrO2 did not correlate with
cerebral end capillary O2 tension as measured with
PET. The change in both values during hyperven-
tilation was however significantly correlated.
PbrO2 is strongly correlated with local CBF
measurement (Doppenberg et al., 1998a; Jaeger
et al., 2005).
Summary and conclusions
ICP and CPP measurements alone are inadequate
for detecting ischemia. The above-mentioned tech-
niques can provide valuable information on oxy-
genation. An ideal technique (non-invasive,
continuous, reliable and easy to use on the ICU)
does not exist. To gain maximal insight the com-
bination of a local and global monitor should
suffice. However the advantages in outcome with
extra measurements or targeted treatment have
not yet been unequivocally proven. Any effort to
do so should be stimulated and probably be part
of a multi-center initiative, for instance Brain-IT.
Recent reports of improved survival after PbrO2
measurement and treatment (Stiefel et al., 2005)
and reports of improved metabolism with hyper-
oxia (Tolias et al., 2004) seem very promising.
The same holds true for the reports on targeted
hyperventilation with jugular bulb measurements
(Cruz, 1998; Cruz et al., 2002). These studies de-
serve verification in other centers. Because of their
broad availability SjVO2 and PbrO2 measurements
seem to be the most logical for multi center studies
with a slight preference for the Licox system be-
cause of its ease of use. After stepwise implemen-
tations of new monitoring techniques a software
system capable of integrating all values and sup-
plying recommendations for further diagnostic
steps or treatment will be crucial. Until then we
will eagerly await and participate in further studies
on autoregulation guided CPP treatment, effect
of hyperoxygenation treatment, targeted hyper-
ventilation and using multimodality monitoring
as secondary endpoints in the use of treatments
such as barbiturate coma, hypothermia and de-
compressive craniectomy.
Acknowledgments
The work of Dr. Haitsma is supported by NWO
grant: 920-03-130. The Department of Neurosur-
gery, EMC, has received in natural support of
research in neurocritical care from GMS mbh.
(Kiel, Germany) and Codman/Johnson & Johnson
(Raynham, MA, USA).
References
Al-Rawi, P.G. (2005) Near infrared spectroscopy in brain
injury: today’s perspective. Acta Neurochir. Suppl., 95:
453–457.
Andrews, R.J. (2001) Monitoring for neuroprotection. New
technologies for the new millennium. Ann. N.Y. Acad. Sci.,
939: 101–113.
Bader, M.K., Littlejohns, L.R. and March, K. (2003) Brain
tissue oxygen monitoring in severe brain injury, II. Implica-
tions for critical care teams and case study. Crit. Care Nurse,
23: 29–38, 40–42, 44.
Bardt, T.F., Unterberg, A.W., Hartl, R., Kiening, K.L.,
Schneider, G.H. and Lanksch, W.R. (1998) Monitoring of
brain tissue PO2 in traumatic brain injury: effect of cerebral
hypoxia on outcome. Acta Neurochir. Suppl., 71: 153–156.
Bellander, B.M., Cantais, E., Enblad, P., Hutchinson, P.,
Nordstrom, C.H., Robertson, C., Sahuquillo, J., Smith, M.,
Stocchetti, N., Ungerstedt, U., Unterberg, A. and Olsen,
N.V. (2004) Consensus meeting on microdialysis in neuroin-
tensive care. Intensive Care Med., 30: 2166–2169.
Bressack, M.A. and Schiffman, J. (2003) Hyperoxia: good or
bad? J. Neurosurg., 98: 223–224; author reply 224–225.
Carmona Suazo, J.A., Maas, A.I., van den Brink, W.A.,
van Santbrink, H., Steyerberg, E.W. and Avezaat, C.J. (2000)
CO2 reactivity and brain oxygen pressure monitoring in se-
vere head injury. Crit. Care Med., 28: 3268–3274.
Charbel, F.T., Hoffman, W.E., Misra, M., Hannigan, K. and
Ausman, J.I. (1997) Cerebral interstitial tissue oxygen ten-
sion, pH, HCO3, CO2. Surg. Neurol., 48: 414–417.
Chieregato, A., Calzolari, F., Trasforini, G., Targa, L. and
Latronico, N. (2003) Normal jugular bulb oxygen saturation.
J. Neurol. Neurosurg. Psychiatry, 74: 784–786.
Cho, H., Nemoto, E.M., Sanders, M., Fernandez, K. and
Yonas, H. (2000) Comparison of two commercially available
near-infrared spectroscopy instruments for cerebral oximetry.
Technical note. J. Neurosurg., 93: 351–354.
Coles, J.P., Fryer, T.D., Smielewski, P., Chatfield, D.A.,
Steiner, L.A., Johnston, A.J., Downey, S.P., Williams,
G.B., Aigbirhio, F., Hutchinson, P.J., Rice, K., Carpenter,
T.A., Clark, J.C., Pickard, J.D. and Menon, D.K. (2004a)
Incidence and mechanisms of cerebral ischemia in early clini-
cal head injury. J. Cereb. Blood Flow Metab., 24: 202–211.
Coles, J.P., Fryer, T.D., Smielewski, P., Rice, K., Clark, J.C.,
Pickard, J.D. and Menon, D.K. (2004b) Defining ischemic

burden after traumatic brain injury using 15O PET imaging
of cerebral physiology. J. Cereb. Blood Flow Metab., 24:
191–201.
Coles, J.P., Minhas, P.S., Fryer, T.D., Smielewski, P.,
Aigbirihio, F., Donovan, T., Downey, S.P., Williams, G.,
Chatfield, D., Matthews, J.C., Gupta, A.K., Carpenter, T.A.,
Clark, J.C., Pickard, J.D. and Menon, D.K. (2002) Effect of
hyperventilation on cerebral blood flow in traumatic head
injury: clinical relevance and monitoring correlates. Crit.
Care Med., 30: 1950–1959.
Coles, J.P., Steiner, L.A., Johnston, A.J., Fryer, T.D.,
Coleman, M.R., Smieleweski, P., Chatfield, D.A., Aigbirhio, F.,
Williams, G.B., Boniface, S., Rice, K., Clark, J.C., Pickard, J.D.
and Menon, D.K. (2004c) Does induced hypertension reduce
cerebral ischaemia within the traumatized human brain?
Brain, 127: 2479–2490.
Contant, C.F., Valadka, A.B., Gopinath, S.P., Hannay, H.J.
and Robertson, C.S. (2001) Adult respiratory distress syn-
drome: a complication of induced hypertension after severe
head injury. J. Neurosurg., 95: 560–568.
Cormio, M., Valadka, A.B. and Robertson, C.S. (1999) Ele-
vated jugular venous oxygen saturation after severe head in-
jury. J. Neurosurg., 90: 9–15.
Cremer, O.L., van Dijk, G.W., Amelink, G.J., de Smet, A.M.,
Moons, K.G. and Kalkman, C.J. (2004) Cerebral hemody-
namic responses to blood pressure manipulation in severely
head-injured patients in the presence or absence of intracra-
nial hypertension. Anesth. Analg., 99: 1211–1217, table of
contents.
Cruz, J. (1993a) Cerebral oxygenation. Monitoring and man-
agement. Acta Neurochir. Suppl., 59: 86–90.
Cruz, J. (1993b) On-line monitoring of global cerebral hypoxia
in acute brain injury. Relationship to intracranial hyperten-
sion. J. Neurosurg., 79: 228–233.
Cruz, J. (1997) Continuous monitoring of partial pressure
of brain tissue oxygen in patients with severe head injury.
Neurosurgery, 40: 1104.
Cruz, J. (1998) The first decade of continuous monitoring of
jugular bulb oxyhemoglobinsaturation: management strate-
gies and clinical outcome. Crit. Care Med., 26: 344–351.
Cruz, J., Nakayama, P., Imamura, J.H., Rosenfeld, K.G.,
de Souza, H.S. and Giorgetti, G.V. (2002) Cerebral extrac-
tion of oxygen and intracranial hypertension in severe, acute,
pediatric brain trauma: preliminary novel management strat-
egies. Neurosurgery, 50: 774–779; discussion 779–780.
Czosnyka, M., Smielewski, P., Piechnik, S., Steiner, L.A. and
Pickard, J.D. (2001) Cerebral autoregulation following head
injury. J. Neurosurg., 95: 756–763.
Dearden, N.M. and Midgley, S. (1993) Technical considera-
tions in continuous jugular venous oxygen saturation meas-
urement. Acta Neurochir. Suppl., 59: 91–97.
De Deyne, C. (1999) Jugular bulb oximetry: the link between
cerebral and systemic management of severe head injury.
Intensive Care Med., 25: 430–431.
De Georgia, M.A. and Deogaonkar, A. (2005) Multimodal
monitoring in the neurological intensive care unit. Neuro
logist, 11: 45–54.
213
Dings, J., Meixensberger, J., Amschler, J., Hamelbeck, B. and
Roosen, K. (1996) Brain tissue PO2 in relation to cerebral
perfusion pressure, TCD findings and TCD-CO2-reactivity
after severe head injury. Acta Neurochir., 138: 425–434.
Dings, J., Meixensberger, J., Jager, A. and Roosen, K. (1998)
Clinical experience with 118 brain tissue oxygen partial pres-
sure catheter probes. Neurosurgery, 43: 1082–1095.
Diringer, M.N., Videen, T.O., Yundt, K., Zazulia, A.R., Aiyagari,
V., Dacey Jr., R.G., Grubb, R.L. and Powers, W.J. (2002) Re-
gional cerebrovascular and metabolic effects of hyperventilation
after severe traumatic brain injury. J. Neurosurg., 96: 103–108.
Diringer, M.N., Yundt, K., Videen, T.O., Adams, R.E.,
Zazulia, A.R., Deibert, E., Aiyagari, V., Dacey Jr., R.G.,
Grubb Jr., R.L. and Powers, W.J. (2000) No reduction
in cerebral metabolism as a result of early moderate hyper-
ventilation following severe traumatic brain injury. J. Neuro-
surg., 92: 7–13.
Doppenberg, E.M., Zauner, A., Bullock, R., Ward, J.D.,
Fatouros, P.P. and Young, H.F. (1998a) Correlations bet-
ween brain tissue oxygen tension, carbon dioxide tension,
pH, and cerebral blood flow–a better way of monitoring the
severely injured brain? Surg. Neurol., 49: 650–654.
Doppenberg, E.M., Zauner, A., Watson, J.C. and Bullock, R.
(1998b) Determination of the ischemic threshold for brain
oxygen tension. Acta Neurochir. Suppl., 71: 166–169.
Engstrom, M., Polito, A., Reinstrup, P., Romner, B.,
Ryding, E., Ungerstedt, U. and Nordstrom, C.H. (2005)
Intracerebral microdialysis in severe brain trauma: the im-
portance of catheter location. J. Neurosurg., 102: 460–469.
Fandino, J., Stocker, R., Prokop, S. and Imhof, H.G. (1999)
Correlation between jugular bulb oxygen saturation and
partial pressure of brain tissue oxygen during CO2 and O2
reactivity tests in severely head-injured patients. Acta
Neurochir., 141: 825–834.
Fandino, J., Stocker, R., Prokop, S., Trentz, O. and Imhof,
H.G. (2000) Cerebral oxygenation and systemic trauma
related factors determining neurological outcome after brain
injury. J. Clin. Neurosci., 7: 226–233.
Fortune, J.B., Feustel, P.J., Weigle, C.G. and Popp, A.J. (1994)
Continuous measurement of jugular venous oxygen satura-
tion in response to transient elevations of blood pressure in
head-injured patients. J. Neurosurg., 80: 461–468.
Gomersall, C.D., Joynt, G.M., Gin, T., Freebairn, R.C. and
Stewart, I.E. (1997) Failure of the INVOS 3100 cerebral oxi-
meter to detect complete absence of cerebral blood flow. Crit.
Care Med., 25: 1252–1254.
Gopinath, S.P., Robertson, C.S., Contant, C.F., Hayes, C.,
Feldman, Z., Narayan, R.K. and Grossman, R.G. (1994)
Jugular venous desaturation and outcome after head injury.
J. Neurol. Neurosurg. Psychiatry, 57: 717–723.
Gopinath, S.P., Robertson, C.S., Contant, C.F., Narayan, R.K.,
Grossman, R.G. and Chance, B. (1995) Early detection of de-
layed traumatic intracranial hematomas using near-infrared
spectroscopy. J. Neurosurg., 83: 438–444.
Gopinath, S.P., Robertson, C.S., Grossman, R.G. and Chance, B.
(1993) Near-infrared spectroscopic localization of intracranial
hematomas. J. Neurosurg., 79: 43–47.
214
Gopinath, S.P., Valadka, A.B., Uzura, M. and Robertson, C.S.
(1999) Comparison of jugular venous oxygen saturation and
brain tissue PO2 as monitors of cerebral ischemia after head
injury. Crit. Care Med., 27: 2337–2345.
Graham, D.I., Adams, J.H. and Doyle, D. (1978) Ischaemic
brain damage in fatal non-missile head injuries. J. Neurol.
Sci., 39: 213–234.
Graham, D.I., Ford, I., Adams, J.H., Doyle, D., Teasdale, G.M.,
Lawrence, A.E. and McLellan, D.R. (1989) Ischaemic brain
damage is still common in fatal non-missile head injury.
J. Neurol. Neurosurg. Psychiatry, 52: 346–350.
Gupta, A.K., Al-Rawi, P.G., Hutchinson, P.J. and
Kirkpatrick, P.J. (2002) Effect of hypothermia on brain tis-
sue oxygenation in patients with severe head injury. Br. J.
Anaesth., 88: 188–192.
Gupta, A.K., Hutchinson, P.J., Al-Rawi, P., Gupta, S., Swart, M.,
Kirkpatrick, P.J., Menon, D.K. and Datta, A.K. (1999) Meas-
uring brain tissue oxygenation compared with jugular venous
oxygen saturation for monitoring cerebral oxygenation after
traumatic brain injury. Anesth. Analg., 88: 549–553.
Haitsma, I.K. and Maas, A.I. (2002) Advanced monitoring in
the intensive care unit: brain tissue oxygen tension. Curr.
Opin. Crit. Care, 8: 115–120.
Hoelper, B.M., Alessandri, B., Heimann, A., Behr, R. and
Kempski, O. (2005) Brain oxygen monitoring: in-vitro accu-
racy, long-term drift and response-time of Licox- and Neuro-
trend sensors. Acta Neurochir. (Wien.), 147: 767–774.
Imberti, R., Bellinzona, G. and Langer, M. (2002) Cerebral
tissue PO2 and SjvO2 changes during moderate hyper
ventilation in patients with severe traumatic brain injury.
J. Neurosurg., 96: 97–102.
Imberti, R., Ciceri, M., Bellinzona, G. and Pugliese, R. (2000)
The use of hyperventilation in the treatment of plateau
waves in two patients with severe traumatic brain injury:
contrasting effects on cerebral oxygenation. J. Neurosurg.
Anesthesiol, 12: 124–127.
Jaeger, M., Soehle, M. and Meixensberger, J. (2003) Effects of
decompressive craniectomy on brain tissue oxygen in patients
with intracranial hypertension. J. Neurol. Neurosurg. Psy-
chiatry, 74: 513–515.
Jaeger, M., Soehle, M., Schuhmann, M.U., Winkler, D. and
Meixensberger, J. (2005) Correlation of continuously moni-
tored regional cerebral blood flow and brain tissue oxygen.
Acta Neurochir. (Wien.), 147: 51–56; discussion 56.
Jia, J., Lin, Y.Q., Liu, W.F., Zhong, T.A., Zhang, J., Ye, Y.
and Xu, Y.Q. (2005) Study of the effects of mild hypothermia
on cerebral PO2, PCO2 and pH and body temperature in
patients with acute severe head injury. Chin. J. Traumatol., 8:
138–141.
Kety, S.S. and Schmidt, C.F. (1945) The determination of cere-
bral blood flow in man by the use of nitrous oxide in low
concentrations. Am. J. Physiol., 143: 53–66.
Kiening, K.L., Unterberg, A.W., Bardt, T.F., Schneider, G.H.
and Lanksch, W.R. (1996) Monitoring of cerebral oxygen-
ation in patients with severe head injuries: brain tissue PO2
versus jugular vein oxygen saturation. J. Neurosurg., 85:
751–757.
Kirkpatrick, P.J., Smielewski, P., Czosnyka, M., Menon, D.K.
and Pickard, J.D. (1995) Near-infrared spectroscopy use in
patients with head injury. J. Neurosurg., 83: 963–970.
Lang, E.W., Czosnyka, M. and Mehdorn, H.M. (2003)
Tissue oxygen reactivity and cerebral autoregulation
after severe traumatic brain injury. Crit. Care Med., 31:
267–271.
Le Roux, P.D., Newell, D.W., Lam, A.M., Grady, M.S. and
Winn, H.R. (1997) Cerebral arteriovenous oxygen difference:
a predictor of cerebral infarction and outcome in patients
with severe head injury. J. Neurosurg., 87: 1–8.
Lewis, S.B., Myburgh, J.A., Thornton, E.L. and Reilly, P.L.
(1996) Cerebral oxygenation monitoring by near-infrared
spectroscopy is not clinically useful in patients with severe
closed-head injury: a comparison with jugular venous bulb
oximetry. Crit. Care Med., 24: 1334–1338.
Littlejohns, L.R., Bader, M.K. and March, K. (2003) Brain
tissue oxygen monitoring in severe brain injury, I. Research
and usefulness in critical care. Crit. Care Nurse, 23: 17–25;
quiz 26–27.
Lodato, R.F. (1990) Oxygen toxicity. Crit. Care Clin., 6:
749–765.
Maas, A.I., Fleckenstein, W., de Jong, D.A. and van Santbrink, H.
(1993) Monitoring cerebral oxygenation: experimental studies
and preliminary clinical results of continuous monitoring of
cerebrospinal fluid and brain tissue oxygen tension. Acta Neuro-
chir. Suppl., 59: 50–57.
Macmillan, C.S., Andrews, P.J. and Easton, V.J. (2001) In-
creased jugular bulb saturation is associated with poor out-
come in traumatic brain injury. J. Neurol. Neurosurg.
Psychiatry, 70: 101–104.
Magnoni, S., Ghisoni, L., Locatelli, M., Caimi, M., Colombo, A.,
Valeriani, V. and Stocchetti, N. (2003) Lack of improvement in
cerebral metabolism after hyperoxia in severe head injury: a
microdialysis study. J. Neurosurg., 98: 952–958.
Mascia, L., Andrews, P.J., McKeating, E.G., Souter, M.J.,
Merrick, M.V. and Piper, I.R. (2000) Cerebral blood flow
and metabolism in severe brain injury: the role of pressure
autoregulation during cerebral perfusion pressure manage-
ment. Intensive Care Med., 26: 202–205.
Matta, B.F., Lam, A.M. and Mayberg, T.S. (1994) The influ-
ence of arterial oxygenation on cerebral venous oxygen
saturation during hyperventilation. Can. J. Anaesth., 41:
1041–1046.
Meixensberger, J., Jaeger, M., Vath, A., Dings, J., Kunze, E.
and Roosen, K. (2003) Brain tissue oxygen guided treatment
supplementing ICP/CPP therapy after traumatic brain injury.
J. Neurol. Neurosurg. Psychiatry, 74: 760–764.
Meixensberger, J., Jager, A., Dings, J., Baunach, S. and
Roosen, K. (1998) Multimodal hemodynamic neuromoni-
toring–quality and consequences for therapy of severely head
injured patients. Acta Neurochir. Suppl., 71: 260–262.
Menzel, M., Doppenberg, E.M., Zauner, A., Soukup, J.,
Reinert, M.M. and Bullock, R. (1999a) Increased inspired
oxygen concentration as a factor in improved brain tissue
oxygenation and tissue lactate levels after severe human head
injury. J. Neurosurg., 91: 1–10.

Menzel, M., Doppenberg, E.M., Zauner, A., Soukup, J.,
Reinert, M.M., Clausen, T., Brockenbrough, P.B. and
Bullock, R. (1999b) Cerebral oxygenation in patients after
severe head injury: monitoring and effects of arterial hype-
roxia on cerebral blood flow, metabolism and intracranial
pressure. J. Neurosurg. Anesthesiol., 11: 240–251.
Metz, C., Holzschuh, M., Bein, T., Woertgen, C., Rothoerl, R.,
Kallenbach, B., Taeger, K. and Brawanski, A. (1998) Moni-
toring of cerebral oxygen metabolism in the jugular bulb:
reliability of unilateral measurements in severe head injury.
J. Cereb. Blood Flow Metab., 18: 332–343.
Mulvey, J.M., Dorsch, N.W., Mudaliar, Y. and Lang, E.W.
(2004) Multimodality monitoring in severe traumatic brain
injury; The role of brain tissue oxygenation monitoring.
Neurocrit. Care, 391–402.
Nagdyman, N., Fleck, T., Schubert, S., Ewert, P., Peters, B.,
Lange, P.E. and Abdul-Khaliq, H. (2005) Comparison bet-
ween cerebral tissue oxygenation index measured by near-
infrared spectroscopy and venous jugular bulb saturation in
children. Intensive Care Med., 31: 846–850.
Nordstrom, C.H. (2005) Physiological and biomechanical prin-
ciples underlying volume-targeted therapy – the "Lund Con-
cept". Neurocrit. Care, 83–96.
Palmer, S. and Bader, M.K. (2005) Brain tissue oxygenation in
brain death. Neurocrit. Care, 17–22.
Perez, A., Minces, P.G., Schnitzler, E.J., Agosta, G.E., Medina, S.A.
and Ciraolo, C.A. (2003) Jugular venous oxygen saturation or
arteriovenous difference of lactate content and outcome in chil-
dren with severe traumatic brain injury. Pediatr. Crit. Care Med.,
4: 33–38.
Robertson, C.S., Contant, C.F., Gokaslan, Z.L., Narayan, R.K.
and Grossman, R.G. (1992) Cerebral blood flow, arteriove-
nous oxygen difference, and outcome in head injured patients.
J. Neurol. Neurosurg. Psychiatry, 55: 594–603.
Robertson, C.S., Valadka, A.B., Hannay, H.J., Contant, C.F.,
Gopinath, S.P., Cormio, M., Uzura, M. and Grossman, R.G.
(1999) Prevention of secondary ischemic insults after severe
head injury. Crit. Care Med., 27: 2086–2095.
Rockswold, S.B., Rockswold, G.L., Vargo, J.M., Erickson, C.A.,
Sutton, R.L., Bergman, T.A. and Biros, M.H. (2001) Effects
of hyperbaric oxygenation therapy on cerebral metabolism
and intracranial pressure in severely brain injured patients.
J. Neurosurg., 94: 403–411.
Rose, J.C., Neill, T.A. and Hemphill, III, J.C. (2006) Contin-
uous monitoring of the microcirculation in neurocritical care:
an update on brain tissue oxygenation. Curr. Opin. Crit.
Care, 12: 97–102.
Rosner, M.J., Rosner, S.D. and Johnson, A.H. (1995) Cerebral
perfusion pressure: management protocol and clinical results.
J. Neurosurg., 83: 949–962.
Rossi, S. and Stocchetti, N. (1999) Brain tissue oxygenation. J.
Neurosurg., 91: 1065–1067.
Sarrafzadeh, A.S., Kiening, K.L., Bardt, T.F., Schneider, G.H.,
Unterberg, A.W. and Lanksch, W.R. (1998) Cerebral oxy-
genation in contusioned vs. nonlesioned brain tissue: moni-
toring of PtiO2 with Licox and Paratrend. Acta Neurochir.
Suppl., 71: 186–189.
215
Schneider, G.H., Sarrafzadeh, A.S., Kiening, K.L., Bardt, T.F.,
Unterberg, A.W. and Lanksch, W.R. (1998) Influence of hy-
perventilation on brain tissue-PO2, PCO2, and pH in patients
with intracranial hypertension. Acta Neurochir. Suppl., 71:
62–65.
Steiner, L.A., Czosnyka, M., Piechnik, S.K., Smielewski, P.,
Chatfield, D., Menon, D.K. and Pickard, J.D. (2002) Con-
tinuous monitoring of cerebrovascular pressure reactivity
allows determination of optimal cerebral perfusion pressure
in patients with traumatic brain injury. Crit. Care Med., 30:
733–738.
Stiefel, M.F., Heuer, G.G., Smith, M.J., Bloom, S., Maloney-
Wilensky, E., Gracias, V.H., Grady, M.S. and LeRoux, P.D.
(2004) Cerebral oxygenation following decompressive hemi-
craniectomy for the treatment of refractory intracranial
hypertension. J. Neurosurg., 101: 241–247.
Stiefel, M.F., Spiotta, A., Gracias, V.H., Garuffe, A.M.,
Guillamondegui, O., Maloney-Wilensky, E., Bloom, S.,
Grady, M.S. and LeRoux, P.D. (2005) Reduced mortality
rate in patients with severe traumatic brain injury treated
with brain tissue oxygen monitoring. J. Neurosurg., 103:
805–811.
Stiefel, M.F., Udoetuk, J.D., Spiotta, A.M., Gracias, V.H.,
Goldberg, A., Maloney-Wilensky, E., Bloom, S. and
LeRoux, P.D. (2006) Conventional neurocritical care and
cerebral oxygenation after traumatic brain injury. J. Neuro-
surg., 105: 568–575.
Stocchetti, N., Chieregato, A., De Marchi, M., Croci, M.,
Benti, R. and Grimoldi, N. (1998) High cerebral perfusion
pressure improves low values of local brain tissue O2 tension
(PtiO2) in focal lesions. Acta Neurochir. Suppl., 71: 162–165.
Thiagarajan, A., Goverdhan, P.D., Chari, P. and Somasunde-
ram, K. (1998) The effect of hyperventilation and hyperoxia
on cerebral venous oxygen saturation in patients with trau-
matic brain injury. Anesth. Analg., 87: 850–853.
Tolias, C.M., Reinert, M., Seiler, R., Gilman, C., Scharf, A.
and Bullock, M.R. (2004) Normobaric hyperoxia–induced
improvement in cerebral metabolism and reduction in intra-
cranial pressure in patients with severe head injury: a pro-
spective historical cohort-matched study. J. Neurosurg., 101:
435–444.
Unterberg, A.W., Kiening, K.L., Hartl, R., Bardt, T.,
Sarrafzadeh, A.S. and Lanksch, W.R. (1997) Multimodal
monitoring in patients with head injury: evaluation of the
effects of treatment on cerebral oxygenation. J. Trauma, 42:
S32–S37.
Valadka, A.B., Gopinath, S.P., Contant, C.F., Uzura, M. and
Robertson, C.S. (1998) Relationship of brain tissue PO2 to
outcome after severe head injury. Crit. Care Med., 26:
1576–1581.
van den Brink, W.A., Haitsma, I.K., Avezaat, C.J.,
Houtsmuller, A.B., Kros, J.M. and Maas, A.I. (1998) Brain
parenchyma/PO2 catheter interface: a histopathological
study in the rat. J. Neurotrauma, 15: 813–824.
van den Brink, W.A., van Santbrink, H., Steyerberg, E.W.,
Avezaat, C.J., Suazo, J.A., Hogesteeger, C., Jansen, W.J.,
Kloos, L.M., Vermeulen, J. and Maas, A.I. (2000) Brain
216
oxygen tension in severe head injury. Neurosurgery, 46:
868–876; discussion 876–878.
van Santbrink, H., Maas, A.I. and Avezaat, C.J. (1996)
Continuous monitoring of partial pressure of brain tissue
oxygen in patients with severe head injury. Neurosurgery, 38:
21–31.
van Santbrink, H., vd Brink, W.A., Steyerberg, E.W.,
Carmona Suazo, J.A., Avezaat, C.J. and Maas, A.I. (2003)
Brain tissue oxygen response in severe traumatic brain
injury. Acta Neurochir. (Wien.), 145: 429–438; discussion
438.
Vespa, P., Bergsneider, M., Hattori, N., Wu, H.M., Huang, S.C.,
Martin, N.A., Glenn, T.C., McArthur, D.L. and Hovda, D.A.
(2005) Metabolic crisis without brain ischemia is common after
traumatic brain injury: a combined microdialysis and positron
emission tomography study. J. Cereb. Blood Flow Metab.,
25(6): 763–774.
Vespa, P.M. (2005) Multimodality monitoring and telemoni-
toring in neurocritical care: from microdialysis to robotic
telepresence. Curr. Opin. Crit. Care, 11: 133–138.
White, H. and Baker, A. (2002) Continuous jugular venous
oximetry in the neurointensive care unit – a brief review:
[L’oxymetrie continue de la veine jugulaire a l’unite des soins
intensifs neurologiques – une breve revue]. Can. J. Anaesth.,
49: 623–629.
Zauner, A., Doppenberg, E.M., Woodward, J.J., Choi, S.C.,
Young, H.F. and Bullock, R. (1997) Continuous monitoring
of cerebral substrate delivery and clearance: initial experience
in 24 patients with severe acute brain injuries. Neurosurgery,
41: 1082–1091; discussion 1091–1093.
Zhi, D.S., Zhang, S. and Zhou, L.G. (1999) Continuous mon-
itoring of brain tissue oxygen pressure in patients with severe
head injury during moderate hypothermia. Surg. Neurol., 52:
393–396.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 15

Update on the treatment of spinal cord injury

Darryl C. Baptiste1 and Michael G. Fehlings1,2,

1
Division of Cell and Molecular Biology, Toronto Western Research Institute and Krembil Neuroscience Centre, Toronto
Western Hospital, Toronto, ON, Canada
2
Division of Neurosurgery, Krembil Neuroscience Centre, Toronto Western Hospital, University of Toronto, Toronto,
ON, Canada

Abstract: Acute spinal cord injury (SCI) is a devastating neurological disorder that can affect any indi-
vidual at a given instance. Current treatment options for SCI include the use of high dose methyl-
prednisolone sodium succinate, a corticosteroid, surgical interventions to stabilize and decompress the
spinal cord, intensive multisystem medical management, and rehabilitative care. While utility of these
therapeutic options provides modest benefits, there is a critical need to identify novel approaches to treat or
repair the injured spinal cord in hope to, at the very least, improve upon the patient’s quality of life.
Thankfully, several discoveries at the preclinical level are now transitioning into the clinical arena. These
include the Surgical Treatment for Acute Spinal Cord Injury Study (STASCIS) Trial to evaluate the role
and timing of surgical decompression for acute SCI, neuroprotection with the semisynthetic second gene-
ration tetracycline derivative, minocycline; aiding axonal conduction with the potassium channel blockers,
neuroregenerative/neuroprotective approaches with the Rho antagonist, Cethrins; the use of anti-NOGO
monoclonal antibodies to augment plasticity and regeneration; as well as cell-mediated repair with stem
cells, bone marrow stromal cells, and olfactory ensheathing cells. This review overviews the pathobiology of
SCI and current treatment choices before focusing the rest of the discussion on the variety of promising
neuroprotective and cell-based approaches that have recently moved, or are very close, to clinical testing.

Keywords: acute spinal cord injury; pathobiology; pharmacologic therapy; neuroprotection; rehabilitation;
cell-mediated repair

Introduction
Acute or traumatic spinal cord injury (SCI) occurs
with an annual incidence of 11.5–53.4 cases per
million within developed nations, with the causes of
these injuries ranging from motor vehicle accidents
and community violence to recreational activities
and workplace-related injuries (Sekhon and
Fehlings, 2001). SCI affects fewer than 200,000
people in the United States and due to the perceived
Corresponding author. Tel.: +1 416 603 5229 (lab); +1 416
603 5800 ext. 2973 (office); Fax: +1 416 603 5745;
E-mail: michael.fehlings@uhn.on.ca
modest financial incentive for the private sector to
make and market new medications, SCI can po-
tentially be considered an ‘‘orphan disease.’’ This is
troubling since the epidemiological predominance
for this debilitating neurological disorder tends to
affect individuals aged in their 3rd decade of life; a
time typically associated with one’s prime earning
potential (Sekhon and Fehlings, 2001). Coupled
with the necessity to provide lifelong health care
support in the form of therapy and rehabilitation, a
significant economic strain is therefore placed on
society by SCI. Moreover, therapies for SCI have
a high likelihood of being translatable for other
neurological indications such as stroke and

DOI: 10.1016/S0079-6123(06)61015-7 217

218
traumatic brain injury. Thus, we would argue that
SCI presents excellent opportunities for commer-
cialization of therapies, which target central nerv-
ous system disorders characterized by cell death or
loss of key neural circuits or pathways.
Since the published results from the landmark
National Acute Spinal Cord Injury Study (NASCIS)
II trial during the early 90 s, steroid use with
methylprednisolone has become widely used in the
treatment of SCI (Bracken et al., 1990, 1992). How-
ever, with methylprednisolone only yielding modest
improvements in neurological recovery coupled with
the potential risks of impaired wound healing and
sepsis the utility of methylprednisolone has been
seriously questioned (Coleman et al., 2000; Hurlbert,
2000). Moreover, findings from the Maryland
monosialotetrahexosylganglioside sodium salt
(GM-1) clinical trial suggested that GM-1 had the
ability to enhance recovery of lower extremities,
which had provided strong rationale to further in-
vestigate the efficacy of low and high dose GM-1
following standard treatment with methylpredniso-
lone in the Sygen multicenter study (Geisler et al.,
1991). However, despite GM-1 demonstrating
improved rates of recovery, along with improved
bladder/bowel function, sacral sensation, and anal
contraction over the first 3 months post-injury the
prospectively planned analysis at 6 months was neg-
ative (Geisler et al., 2001). Hence, with no univer-
sally accepted standard therapy in place, novel
therapeutic approaches for SCI are urgently re-
quired. Surely, such a finding can only be made
through further understanding of the etiological
events following traumatic SCI. As such, we owe
much of our current state of knowledge regarding
the etiology of SCI to those involved in basic sci-
entific research utilizing clinically relevant animal
models of SCI, often involving extradural spinal
cord contusion or compression in rodents (Rivlin
and Tator, 1978; Gruner, 1992; Stokes et al., 1992;
Joshi and Fehlings, 2002a, b).
Pathobiological mechanisms of SCI
The pathobiology of SCI follows a biphasic proc-
ess, involving a ‘‘primary’’ traumatic insult to the
spinal cord that initiates a disruption in spinal
cord blood flow dynamics and local ischemia-
reperfusion. This initial trauma is then followed by
‘‘secondary’’ biochemical events that set the stage
for membrane dissolution and cellular death.
Some notable ‘‘secondary’’ etiological events
include disturbances to ionic homeostasis (i.e.,
irregular regulation of cellular calcium, sodium,
and potassium trafficking), glutamate-mediated
excitotoxicity, mitochondrial instability, reactive
oxygen species generation, elevated activation of
proteolytic enzymes such as caspases and calpains,
and inflammation.
Cell death following SCI likely occurs as a con-
tinuum of necrosis and apoptosis. While necrosis
predominates immediately following the primary
traumatic episode, researchers have observed de-
layed neuronal and oligodendroglial apoptosis, al-
beit the former occurs to a lesser degree, relative to
the latter (Casha et al., 2001). Applied neuropro-
tective approaches have been tested in SCI models
with the intention that drug-mediated protective
effects will lead to preservation of neuronal and
glial cell populations via targeting one or more of
the aforementioned secondary injury events. How-
ever, to date, few pharmacological approaches
have been successful in translating therapeutic
value in patients.
Current non-pharmacological treatment options for
SCI
Surgery
In SCI, primary mechanical forces together with
secondary injurious mechanisms collectively con-
tribute to nervous tissue damage. Experimental
evidence exists to support surgical decompression
following persistent compression of the spinal cord
following SCI (Delamarter et al., 1995). However,
despite its widespread use in patients with SCI, the
role of surgery in improving neurological recovery
remains controversial because of the absence of
randomized, controlled clinical trials. Further-
more, the presence and duration of a therapeutic
window, during which surgical decompression
could mitigate the secondary mechanisms of SCI
have yet to be clearly defined.

The timing of surgical intervention in the treat-
ment of experimental and clinical SCI was recently
reviewed (Fehlings and Perrin, 2006). The authors
conducted a MEDLINE search of the literature
from 1966 to 2005, dealing with the role of de-
compression in SCI using the medical subject
heading of ‘‘decompression’’ and ‘‘spinal cord in-
jury.’’ The overall conclusions from this extensive
search revealed that decompressive surgery per-
formed within the first 24 h of injury could be
conducted safely. It is the senior authors’ recom-
mendation that urgent decompression be per-
formed within the first 24 h of injury following
an isolated cervical SCI only if hemodynamic sta-
bility can be maintained.
The Spinal Cord Injury Committee of the Joint
Section on Neurotrauma and Critical Care of the
American Association of Neurological Surgeons
(AANS) and the Congress of Neurological Sur-
geons (CNS) formed the Surgical Treatment for
Acute Spinal Cord Injury Study (STASCIS) group
in 1992. By 1996, STASCIS was joined by the
Joint Section on Spinal Disorders and Peripheral
Nerves of the AANS/CNS. The principal aim of
the STASCIS group was to conduct a randomized
prospective controlled trial to examine the role and
timing of decompressive surgery performed on the
spinal cord and cauda equina of patients after
acute spinal injury.
A multicenter, retrospective study was per-
formed in 36 North American centers to examine
the use and timing of surgery in patients who have
sustained SCI (Tator et al., 1999). The study was
performed to obtain information required for the
planning of a randomized controlled trial in which
early and late decompressive surgery are com-
pared. A total of 585 patients aged 16–75 years
with acute SCI or cauda equina injury were ad-
mitted to participating centers, although approxi-
mately half of the patients were excluded due to
late admission, age, gunshot wound, or absence of
signs of compression on imaging. The timing of
surgery varied widely, where 23.5, 15.8, 19, and
41.7% of the patients received surgery within 24,
25–48, 48–96 h, and 5 days post-injury, respec-
tively. The findings from this study have not led to
unequivocal support for varied timing of surgical
management in SCI patients. Thus, the benefit of
219
surgical management following SCI will only be
demonstrated clearly with a carefully designed
prospective controlled trial to assess the optimum
timing of decompressive surgery in SCI. These re-
sults spurred the STASCIS to embark on a pro-
spective, multicenter study in North America to
determine the effectiveness of early surgical de-
compression and stabilization procedures for re-
ducing the possibility of further damage to the
spinal cord following acute compression of the
cervical spinal cord. This study is currently under-
way and is being converted into a collaborative
trial under the auspice of the Spine Trauma Study
Group (STSG).
Rehabilitation
Gait rehabilitation is a specific component of
physical rehabilitation for persons with subacute
or chronic motor-incomplete SCI. One novel
method of gait rehabilitation involves the use of
an overhead support point and a harness (i.e.,
body weight supported or BWS). The BWS strat-
egy has been combined with treadmill-based gait
training in recent studies with dramatic results
(Hicks et al., 2005). This combination of BWS and
treadmill rationally evolved from the idea that
man possesses a central pattern generator, a cir-
cuitry located within the spinal cord capable of
driving involuntary locomotor movements,
similar to lower mammals (Duysens and Van de
Crommert, 1998; Van de Crommert et al., 1998).
Indeed, an electrical train of stimuli applied over
the second lumbar segment with a frequency of
25–60 Hz and an amplitude of 5–9 V was shown to
be effective in inducing rhythmic, alternating
stance and swing phases of the lower limbs in
paraplegic subjects, providing proof of a central
pattern generator in humans (Dimitrijevic et al.,
1998). Thus, it is believed that the combination of
BWS and treadmill training will enhance the
output by the intrinsic central pattern generator.
The effects of long-term BWS treadmill training
on functional walking ability and perceived quality
of life were recently assessed in 14 persons with
chronic incomplete SCIs (Hicks et al., 2005). Study
subjects, classified on entrance as being ASIA
220
(American Spinal Injury Association) grade B or
C, participated in three weekly sessions of BWS
treadmill training over the course of 1 year and
were assessed for improvements in functional abil-
ity and quality of life throughout the year and up
to 8 months thereafter. The effect of this regimen
showed that all subjects required less body weight
support, and were able to increase both walking
speed and distance per session, as well as improved
over-ground walking ability. Heightened physical
ability was correlated with an enhanced perspec-
tive of life. By the 8-month follow-up subjects
demonstrated diminished performance in treadmill
sessions with stable over-ground performance.
Their quality of life reflected slight disappoint-
ment in their decreased physical function. Thus,
overall the BWS treadmill training appeared to be
a useful form of rehabilitation.
However, the efficacy of BWS treadmill training
may not confer additional benefit over other forms
of intensive rehabilitation therapies. Recently, the
efficacy of step training with BWS treadmill train-
ing was compared with over-ground practice to the
efficacy of a defined over-ground mobility therapy
in 146 patients with incomplete SCI admitted for
inpatient rehabilitation in six regional centers
(Dobkin et al., 2006a). Following receipt of 12
weeks of BWS treadmill training or controlled
over-ground mobility training, primary outcome
variables of functional independence measures for
locomotion or walking speed and distance for
ASIA grade A and B SCI patients, or ASIA grade
C and D SCI patients, respectively, obtained at the
6 month endpoint, did not support the expectation
that BWS treadmill training would be more effica-
cious than intensive over-ground mobility therapy.
Improving axonal conduction in the injured spinal
cord
Fampridines
Physical trauma to the spinal cord results in de-
myelination of preserved spinal tracts. Without
insulating sheaths of myelin, these surviving axons
become less efficient in their ability to transmit
electrical impulses to be conducted. As a result, the
demyelinated axon cannot transmit motor or
sensory impulses. When an axon is demyelinated
after injury, large numbers of potassium channels
are exposed to the extracellular space and permit
potassium ion channels to open and release
potassium ions. Thus, the result of demyelinated
axons and potassium channel exposure is a
reduced safety factor of action potential propaga-
tion across the demyelinated region of the axon
(Nashmi and Fehlings, 2001). As such, pharma-
cological antagonism of exposed potassium
channels on demyelinated axons with Fampri-
dines (4-aminopyridine or 4-AP) had been
hypothesized to restore the ability of axons to
transmit electrical impulses.
In a randomized, double-blind, crossover de-
signed trial, where each patient received 4-AP and
a vehicle placebo on different occasions by infu-
sion over 2 h, separated by 2 weeks, 4-AP treat-
ment demonstrated a significant temporary
neurologic improvement in five of six patients
with incomplete SCI. No effect was detected in
two cases of complete paraplegia and one of two
severe incomplete cases (Hansebout et al., 1993).
More recently, in a prospective, randomized,
double-blind, placebo-controlled, crossover trial
15 chronic ambulatory SCI patients were ran-
domized to receive either an initial 2 weeks of
40 mg/day, oral 4-AP medication of placebo or
immediate-release, 4-AP before being crossed over
to an alternate medication for the next 2 weeks of
the study (DeForge et al., 2004). Evaluations for
the study were conducted at baseline before
initiating treatment, 2 weeks and 4 weeks post-
treatment through measurement of dynamometer
lower limb isometric muscle force and biomechan-
ical gait assessments. Patients were also required to
give their subjective impressions of their received
therapy through a survey on exit of the study. The
results of the study demonstrated that despite
positive feelings towards the treatment no statis-
tical differences were noted between placebo and
4-AP therapies.
Information regarding the current status of Fa-
mpridines clinical testing for SCI is not yet pub-
lished, but can be found at the Acorda Therapeutics
web site. In 2004, Acorda Therapeutics completed
two Phase III clinical trials with their proprietary

Fampridines (Sustained Release) formulation. Nei-
ther of these studies demonstrated statistical sig-
nificance in their primary endpoints, reduction of
spasticity, as measured by the Ashworth score, and
improvement of participant’s Subject Global Im-
pression rating. However, in one of the studies, the
data showed a positive trend toward improvement
on the Ashworth score (http://www.acorda.com/
pipeline_fampridine_sci1.asp).
Potassium and sodium channel blockade
While the preclinical findings strongly support the
use of potassium channel blockade with 4-amino-
pyridine to improve axonal conduction following
trauma-induced demyelination, the clinical utility
of 4-aminopyridine is limited due to its associated
adverse effects, which include restlessness, confu-
sion, and infrequent generalized tonic-clonic sei-
zure (Bever, 1994).
In an effort to minimize these unwanted side
effects Sanofi-Aventis has designed HP184 (N-[N-
propyl]-N-[3-fluoro-4-pyridinyl]-1H-3-methylindole-
1-amine hydrochloride), a pharmacological blocker
capable of antagonizing both potassium and
sodium channels. Sanofi-Aventis is currently in
the process of conducting a Phase II, double-blind,
placebo-controlled, multicenter study to assess the
efficacy and safety of HP184 at 100, 200, and
400 mg doses administered orally once daily for 24
weeks in 18–65 year old patients with chronic SCIs
classed ASIA grade C or D. Sanofi-Aventis will
also assess the effect of HP184 treatment on
recovery of walking function and spasticity.
Neuroprotective/neuroregenerative approaches to
treating the injured spinal cord
Methylprednisolone sodium succinate
Preclinical research originally identified methyl-
prednisolone as a potent inhibitor of lipid peroxi-
dation following traumatic injury (Hall and
Braughler, 1981, 1982). These findings led to the
initiation of NASCIS I, which compared the effec-
tiveness of treating SCI patients with an adminis-
tered intravenous (i.v.) standard or high dose
221
methylprednisolone sodium succinate (MPSS)
consisting of 100 or 1000 mg bolus followed by
25 or 250 mg MPSS daily thereafter for 10 days,
respectively (Bracken et al., 1984). However the
6-month follow-up data for NASCIS I revealed no
difference in neurological recovery of motor
function or pinprick and light touch sensation
between the two treatment arms of the study.
Moreover, increased fatality, wound infections
of both trauma and operative sites were associ-
ated to a greater extent with high-dose MPSS
(Bracken et al., 1984).
Further animal testing later generated data to
suggest that the NASCIS I MPSS dose used was
not sufficient to elicit notable therapeutic efficacy
related to improved protection from ischemic
insults and calcium-dependent degradation of
neurofilament cytoskeletal proteins (Braughler
and Hall, 1984). Thus, the second NASCIS trial
was designed to evaluate the efficacy of high-dose
MPSS administered as a 30 mg/kg bolus over the
first hour followed by an infusion of 5.4 mg/kg/h
over the next 23 h to placebo, and to a group of
patients receiving the opioid receptor antagonist
naloxone (Bracken et al., 1990). The trial was fur-
ther designed to investigate the effects of differing
times for drug administration following SCI. An
analysis of all patients who entered into the trial
failed to demonstrate a significant difference
among the treatments. However, upon stratifica-
tion of the data on the basis of time to loading
dose (less than 8 h vs. greater than 8 h from SCI)
and adjustment for the severity of SCI (complete
vs. incomplete), analysis of the results from this
second study at the 6-week and 6-month follow-up
revealed that patients treated with MPSS within
the first 8 h of injury had significantly improved
motor and sensory function in comparison to pa-
tients receiving placebo, naloxone, or MPSS at
later times. Furthermore, these differences re-
mained significant 1 year following SCI (Bracken
et al., 1992). None of the differences in patients
taking naloxone or in patients treated with
MPSS more than 8 h after SCI were statistically
significant.
Despite the beneficial therapeutic effects demon-
strated with MPSS treatment, the results from the
NASCIS II trial have not been universally
222
accepted (Casha et al., 2001; Hurlbert, 2001;
AANS/CNS, 2002). Concerns surrounding the
small sample population size for the groups show-
ing beneficial effects, performance of medical and
surgical protocols by participating centers in a
nonstandardized manner, and the fact that there
were no functional outcome measures defined to
assess whether the improvements noted with
MPSS treatments were correlated with clinical sig-
nificance all reduced enthusiasm for the indication
of MPSS for acute SCI.
In an attempt to resolve some of the issues
raised from the second NASCIS trial, a third
NASCIS trial was organized to include a func-
tional independence measure (Bracken et al.,
1997). A total of 499 patients suffering a SCI
within 8 h were enrolled into the study and re-
ceived 30 mg/kg bolus of MPSS before random-
ization. Patients in the 24 h regimen group received
a MPSS infusion of 5.4 mg/kg/h over 24 h, those in
the 48 h regimen group received a MPSS infusion
of 5.4 mg/kg/h for 48 h and those in the tirilazad
group received a 2.5 mg/kg bolus infusion of
tirilazad mesylate (TM; a ‘‘lazaroid’’ with inhibi-
tory effects on lipid peroxidation without gluco-
corticoid side effects) every 6 h for 48 h. The results
from this 3rd study concluded that no benefit was
associated with extending MPSS treatment beyond
24 h if MPSS had been administered within the
first 3 h of SCI, while those that started MPSS
therapy between 3 and 8 h demonstrated improve-
ments in motor capabilities if drug infusion was
continued for 48 h in comparison to 24 h infusion.
It should be noted that the guidelines committee
of the AANS/CNS Joint Section on Disorders of
the Spine and Peripheral Nerves on reviewing the
evidence regarding the use of MPSS in the treat-
ment of acute SCI in adults concluded that its use
could only be supported at the level of a treatment
option (AANS/CNS, 2002). However, it is the au-
thors’ view that the intense criticism directed at the
NASCIS II and III trials must be balanced by the
current lack of alternative neuroprotective strate-
gies for acute SCI. Moreover, the modest thera-
peutic benefit demonstrated by MPSS may
potentially prove to impart a major benefit on
cervical SCI patient’s functional independence and
quality of life (Sekhon and Fehlings, 2001). Still,
despite these concerns the NASCIS II and III trials
effectively demonstrated the validity for targeting
secondary injurious pathological events in SCI,
while also emphasizing the importance of timing
with applied interventions.
Monosialotetrahexosylganglioside
Monosialotetrahexosylganglioside GM-1 sodium
salt is a naturally occurring compound, which re-
sides within cell membranes of the mammalian
CNS (Geisler et al., 2001). Several preclinical stud-
ies reporting neuroprotective and neuroregenera-
tive actions by GM-1 in experimental models of
ischemia and injury have been reported (Agnati
et al., 1983; Fass and Ramirez, 1984; Bose et al.,
1986). These findings have led to the initiation of
the Maryland GM-1 Ganglioside Study, a small
randomized, placebo-controlled, double-blind trial
designed to investigate the efficacy of GM-1 in
patients with cervical and thoracic SCIs (Geisler
et al., 1991). In this trial, 34 patients completed the
test-drug protocol that consisted of either 100 mg
GM-1 or placebo i.v. per day for 18–32 doses, with
the first dose beginning within 72 h of the onset of
SCI. The results from this study demonstrated the
effectiveness for GM-1 to improve the recovery of
neurologic function after 1 year and provided the
necessary rationale to support the recruitment of
SCI patients for a larger GM-1 study.
The randomized double-blind Sygens Multi-
center Acute Spinal Cord Injury Study was com-
pleted in the late 1990s and reported in 2001. In
this trial, all 797 patients were randomized to re-
ceive a standard 30 mg/kg bolus of methyl-
prednisolone followed by 5.4 mg/kg/h infusion of
methylprednisolone for 23 h within the first 8 h of
injury. Placebo, low-dose GM-1 (300 mg loading
dose followed by 100 mg/day for 56 days) or high-
dose GM-1 (600 mg loading dose followed by
200 mg/day for 56 days) was then randomly as-
signed following the completion of corticosteroid
administration to avoid unwanted drug interac-
tions between methylprednisolone and GM-1.
The findings from this larger trial demonstrated
that patients receiving either low- or high-dose
GM-1 had enhanced recovery of neurologic

function by 3 months post-injury regardless of
their baseline severity, and showed a trend for im-
proved bowel/bladder function, sacral sensation,
and anal contraction (Geisler et al., 2001). How-
ever, the primary outcome variable, the percentage
of marked recovery, at the designated 6 month
time-point was negative. While no future clinical
trials with Sygen are currently planned, the spon-
soring company Fidia is evaluating this option
(Corporate Communications Fidia Farmaceutici
SpA, Personal communications).
Minocycline
The broad-spectrum antibiotic, minocycline, is a
lipophilic derivative of tetracycline that has demon-
strated ability to provide neuroprotection via intra-
peritoneal (Wells et al., 2003; Teng et al., 2004) or
i.v. (Xu et al., 2004) routes of administration.
Mechanisms attributed to the protective actions
elicited by minocycline include ability to overcome
glutamate-mediated excitotoxicity (Tikka and
Koistinaho, 2001; Baptiste et al., 2004), anti-
inflammatory effects through blocking the activa-
tion of microglial cells (Dommergues et al., 2003;
Baptiste et al., 2005), inhibiting both cytochrome c
release and caspase-dependent apoptotic neuronal
death (Zhu et al., 2002), as well as a capacity to
antagonize the activities of matrix metal-
loproteinases-2 (Cho et al., 2006) and -9 (Brundula
et al., 2002). Additionally, minocycline was shown
to reduce oligodendrocyte apoptosis, corticospinal
tract dieback, and lesion size, while improving
functional outcome following a C7–C8 dorsal col-
umn transection SCI in rats (Stirling et al., 2004).
Considering, these findings together with the rel-
atively safe clinical track record of minocycline has
made this pharmacological approach a very prom-
ising candidate for the SCI.
A Phase I/II pilot study to investigate the effi-
cacy of intravenously administered minocycline in
patients with acute SCI is currently underway in
Calgary, Alberta. The study has aimed to enroll
patients above 16 years of age possessing a
SCI between C1 and T11. The major objectives
of this trial will be to assess the feasibility of
both minocycline administration and clinical and
223
biochemical monitoring in the acute setting of SCI;
to monitor the safety and tolerance of i.v. mino-
cycline administration; to assess systemic and cer-
ebral spinal fluid (CSF) absorption characteristics
of i.v. minocycline administration; and to assess
serial CSF interleukin 1 beta (IL-1b) and nitric
oxide concentration, which correspond to the ac-
tivities of caspase-1 and -3, respectively, along with
tumor necrosis factor alpha (TNF-a) levels to
reflect the levels of a major pro-inflammatory
cytokine.
Cethrins
CNS myelin inhibits axon growth due to the ex-
pression of several extracellular growth-inhibitory
proteins, including myelin-associated glycoprotein
(MAG), myelin oligodendrocyte glycoprotein
(MOG), and neurite outgrowth (Nogo). It is now
evident that the growth-inhibitory effects of these
proteins are exerted through the regulatory actions
of small intracellular GTPase-associated signaling
proteins, Rho and Rac (Niederost et al., 2002;
Winton et al., 2002). Rho and Rac can exist in one
of two possible forms — an active GTP-bound
state, and an inactive GTP-bound state. In the ac-
tive conformation, these small GTPase-associated
proteins work to inhibit neurite outgrowth. For
example, purified cerebellar granule cells grown on
Nogo-A, Nogo-66, or MAG containing substrates
have all been shown to result in inhibited neurite
outgrowth (Niederost et al., 2002). Furthermore,
increased levels of Rho-GTP were detected in
pheochromocytoma cell line cultures that had been
plated on myelin, MAG, or poly-L-lysine contain-
ing substrates by measuring the precipitate from
cell homogenates mixed with Rho binding domain
isolated from rhotekin, a GTP-interacting protein
(Dubreuil et al., 2003). Moreover, MAG, MOG,
or Nogo-mediated neurite outgrowth inhibition
can be relieved with the addition of Rho-A-
associated kinase ROCK inhibitor (Niederost
et al., 2002).
Preclinical studies now support the hypothesis
that SCI is associated with an upregulation of
GTP-bound Rho. Dubreuil et al. (2003) demon-
strated that Rho-GTP levels are robustly elevated
224
following transection or contusion cord injuries in
comparison to uninjured control animals. Rho ac-
tivation following transection-induced injury be-
gan as early as 1.5 h post-SCI and was sustained
for 7 days (Dubreuil et al., 2003). Addition of the
Rho-selective antagonist, C3–05 transferase, in-
jected in a fibrin matrix to the lesion site following
cord transection, or C3–05 alone into the contused
cord demonstrated ability to reduce RhoA activa-
tion levels back to physiological conditions
(Dubreuil et al., 2003). Also, treatment with
C3–05 reduced neuronal and glial cell apoptosis
that would pathologically occur in secondary in-
jury processes of SCI (Dubreuil et al., 2003). Thus,
this novel approach has demonstrated multifac-
eted therapeutic actions, namely neuroregenerative
and neuroprotective properties.
BioAxone Inc., a biopharmaceutical company
based in Montreal, Quebec, recently sponsored a
Phase I/IIa multinational clinical trial (lead site in
Toronto, Ontario) designed to evaluate the safety
and pharmacokinetics of rising dosages of Cethrin
(BA-210), a recombinant fusion protein composed
of C3 transferase and a transport sequence that
aids the protein’s ability to breach cellular mem-
branes together with a fibrin sealant, following a
single extradural administration during surgery for
acute thoracic and cervical SCI. The trial was
conducted in nine centers located in Canada and
the USA. In December 2005, the U.S. Food &
Drug Administration granted Orphan Drug Status
to Cethrins, which will likely hasten clinical eval-
uation of BioAxone’s lead product through reduc-
ing clinical development costs and facilitating
regulatory filings in other countries. It is antici-
pated that the results of this Phase I/IIa trial will
become available in early 2007.
Nogo
It has long been thought that the capacity for the
CNS to repair lesions through regeneration was
limited. However, hope in this context has been
invigorated since the landmark discoveries by
Richardson et al. (1980), which have provided
the first indication that CNS neurons possess an
intrinsic ability to regenerate following injury that
can be encouraged further upon manipulation of
the local tissue microenvironment.
Years later, Caroni et al. (1988) reported that
nonpermissive substrates within the CNS were
actually associated with oligodendrocytes and my-
elin. Moreover, monoclonal antibody or pharma-
cological treatment directed against white matter
nonpermissive substrates allowed neuron out-
growth (Caroni and Schwab, 1988; Savio and
Schwab, 1989). By 2000, the gene nogo, which can
be transcribed to yield Nogo-A, -B, and -C mRNA
products, was cloned (Chen et al., 2000). Nogo-A
is a high molecular weight transmembrane protein
and potent inhibitor of neurite outgrowth pro-
duced by oligodendrocytes. Inhibition of Nogo-A
with the selective monoclonal antibody IN-1 or
with the peptide antagonist NEP1-40, in spinally
injured rats demonstrates neutralization of myelin-
dependent neurite outgrowth blockade, improved
axonal regeneration, and enhanced rates of func-
tional locomotor recovery (Chen et al., 2000; Li
and Strittmatter, 2003).
Novartis Institutes for BioMedical Research has
since taken a focused interest in developing anta-
gonists for Nogo-A and its corresponding recep-
tor, Nogo-66, to promote neuroregeneration
(Wiessner et al., 2003; Walmsley et al., 2004).
Currently, Novartis is sponsoring a Phase I clinical
trial to evaluate the safety of anti-Nogo-A mono-
clonal antibody therapy.
Autologous macrophages
The CNS was once considered as an ‘‘immune-
privileged’’ environment that was neither suscepti-
ble to nor could evoke an inflammatory response
(Lucas et al., 2006). Consequently, the hypothesis
that inflammation may be harnessed to heal the
injured spinal cord has not been received without
skepticism. Although evidence exists to support a
pathological role for inflammation in the etiology
of many CNS disorders (Stoy et al., 2005; Man
et al., 2007; Miklossy et al., 2006; Noorbakhsh
et al., 2006), it is very likely that inflammatory me-
diators possess dual roles, with detrimental acute
effects together with healing effects in longer-term
repair and recovery (Lucas et al., 2006).

Preclinical investigations involving rats that had
received a single complete spinal cord transection
together with local implantation of ex vivo-
activated homologous macrophages incubated
with homologous peripheral nerves demonstrated
that stimulation of the host immune response can
promote partial recovery of motor function, as
demonstrated through hind limb movement, plan-
tar paw placement, and weight-bearing steps,
along with electrophysiological assessments via
cortically evoked hind limb muscle response
(Rapalino et al., 1998). Additionally, macrophag-
es incubated with autologous skin were found to
be just as efficacious (Rapalino et al., 1998). The
mechanism of action associated with autologous
ex vivo-activated macrophages is believed to be
both neuroprotective and neuroregenerative. Ne-
uroprotective properties likely arise due to the
ability of activated macrophages to synthesize and
secrete concurrently protective cytokines IL-1b
and IL-6, together with the trophic factor, brain
derived neurotrophic factor, and the chemokine,
IL-8, while reducing levels of the neurotoxic cyto-
kine, TNF-a. In addition, regenerative properties
may result due to macrophage-mediated phagocy-
tosis of myelin debris, an event that would effec-
tively make the environment of the injured spinal
cord less cytotoxic and more responsive to axonal
regeneration (Bomstein et al., 2003).
From 2000 to 2003, Proneuron Biotechnologies
conducted a Phase I, open-label nonrandomized
safety study in patients diagnosed with a complete
SCI (ASIA grade A) within C5–T11 within 14 days
of injury at centers located in Israel and Belgium
(Knoller et al., 2005). ASIA grade A patients
were the only SCI population admitted into the
ProCords study, in an effort to limit the con-
founding interpretation of spontaneous recovery
that would result from incomplete injuries (ASIA
grades B, C, or D).
Proneuron Biotechnologies’ proprietary proce-
dure for obtaining activated autologous macroph-
ages consists of isolating white lymphocytes from
the patient’s blood and incubating them ex vivo
with autologous dermis. A single dose of 4 million
activated autologous macrophages was adminis-
tered by four microinjections into the spinal cord
parenchyma at the caudal border of the spinal
225
cord lesion. The results from this study at 1-year
follow-up demonstrated that the therapy was safe
and warranted further investigation in a randomi-
zed Phase II, multicenter clinical trial within the
USA and Israel. However, despite the FDA grant-
ing an orphan drug status to ProCord on Septem-
ber 13, 2004, recruitment into this trial is currently
suspended due to presumed financial challenges
related to the conduct of such a large complex
trial. Clinical follow-up is currently being con-
ducted for those patients that have already been
admitted.
Cell-mediated repair of the injured spinal cord
Neural stem and progenitor cells
A variety of cell types have been investigated for
their potential use in cell-mediated tissue repair
following SCI. Theses include neural stem cells
(NSCs), neural progenitor cells (NPCs), Schwann
cells, oligodendroglial precursor cells (OPCs),
bone marrow stem cells (BMSCs), and olfactory
ensheathing cells (OECs).
NSCs and NPCs have functional differences,
where the former represents an uncommitted cell
type that can divide repeatedly while maintaining
potency to generate differentiated cell types (neu-
rons and glia), whereas the latter represents a kind
of cell that exhibits a limited number of cell divi-
sions before it differentiates into a specific cell type
(Pollard et al., 2006). Despite the fact that stem/
precursor cells may be derived from a variety of
sources, research with these cells raises both po-
litical and ethical concerns. As a result, stem cell
technology is trapped within an ethical dilemma of
balancing heightened emotions regarding embryo/
fetal research with the potentially therapeutic
benefits to patients with desperate hopes.
NSCs have been isolated from the adult human
hippocampus and lateral ventricle wall (Johansson
et al., 1999b). Additionally, postmortem analysis
of tissue derived from the hippocampus and sub-
ventricular zone of the caudate nucleus revealed
that NPCs reside in the adult human brain
(Eriksson et al., 1998). Thus, these findings show
that neurogenesis occurs throughout life and
226
supports the hypothesis that the mature CNS pos-
sesses an innate ability to repair damaged tissue. In
light of these findings researchers have attempted
to demonstrate the utility harnessing the capabil-
ities of endogenous stem/progenitor cells following
traumatic injuries (Johansson et al., 1999a; Kernie
et al., 2001). However, the results of these studies
demonstrated that the endogenous stem/progeni-
tor supply tend to move toward an astrocytic lin-
eage (Johansson et al., 1999a; Kernie et al., 2001).
Studies involving transplantation of NSCs/
NPCs have shown greater promise for traumatic
injuries than attempts to stimulate or control en-
dogenous neurogenesis. For instance Cummings
et al. (2005) demonstrated that human CNS stem
cells grown as neurospheres survive, migrate, and
express differentiation markers for neurons and
oligodendrocytes after long-term engraftment in
the injured spinal cord. Furthermore, the stem cell
engraftment was associated with improved loco-
motor recovery with diminished astrocytic differ-
entiation, remyelination in both the spinally
injured and myelin-deficient mice, along with elec-
tron microscope evidence consistent with synapse
formation between stem cells and host neurons
(Cummings et al., 2005). However, despite signifi-
cant advances in this, adult NSCs remain an elu-
sive cell for study, and researchers are facing many
challenges to the development of therapeutic ap-
plications from adult NSC research. Among these
challenges are the identification and characteriza-
tion of NSCs, the understanding of the physiology
of newly generated neuronal cells in the adult
brain, and the isolation and culture of homoge-
nous populations of NSCs/NPCs from the adult
CNS for cell-based therapy. Furthermore, meas-
ures must be taken to improve on the survival of
transplanted stem/progenitor cells in the hostile
environment of the injured spinal cord (Karimi-
Abdolrezaee et al., 2006).
Schwann cells
Schwann cells are a type of glial cell responsible
for forming myelin sheaths around peripheral
nerves. Researchers are excited about the use of
Schwann cells in neural repair because these cells
may act as a physical bridge, allowing the normal
functioning of undamaged or regenerated axons.
Moreover, these cells are capable of secreting nu-
merous trophic factors including brain derived ne-
urotrophic factor, glial derived neurotrophic
factor, nerve growth factor, and ciliary neurotro-
phic factor (Pellitteri et al., 2006), while also being
able to remyelinate (Kohama et al., 2001) and
guide regenerating CNS axons to their intended
tracts (Zheng and Kuffler, 2000). The transplant
procedure would involve removing a small amount
of the patient’s own peripheral nerve tissue, iso-
lating Schwann cells from the tissue, growing them
in plastic culture dishes in an incubator, and then
implanting the cultured Schwann cells into the site
of SCI.
A recent report has demonstrated the efficacy of
subarachnoid space transplantation of Schwann
cells into the contused epicenter in adult rats 7
days post-SCI resulting in significantly enhanced
locomotor recovery and significantly greater ax-
onal density at the lesion epicenter (Firouzi et al.,
2006). Furthermore, this route for transplantation
was without notable reduction in tissue integrity or
augmented levels of pro-inflammatory and pro-
apoptotic mediators (Firouzi et al., 2006). How-
ever, in light of these advances with Schwann cells,
researchers have still not found a solution to per-
mit regenerating axons to extend beyond the
transplant due to the inhibitory nature of the sur-
rounding glial scar (Oudega and Xu, 2006). This
problem likely calls for a combinatorial approach,
which makes the local microenvironment more re-
ceptive for regeneration.
Bone marrow stromal cells
Bone marrow stromal cells (BMSCs) provide yet
another source for obtaining pluripotent stem
cells. Bone marrow contains a population of pluri-
potent cells that have the capacity to migrate to-
wards a lesion and release various growth factors
to facilitate regeneration and lesion repair. Under
physiological settings BMSCs function to regulate
hematopoiesis; however, if these cells are grown
away from their natural environment these cells
can be readily and effectively propagated and

manipulated genetically to yield a variety of cell
types including phenotypes that closely resemble
neurons (Clark and Keating, 1995). The advantage
of using bone marrow as a source for stem cells is
that they are relatively easy to isolate, the cells
grow well in tissue culture, BMSCs may be used in
autologous transplantation protocols, and these
BMSCs have already received approval for
hematopoietic diseases (Syková et al., 2006).
Hofstetter et al. (2002) differentiated BMSCs in
vitro into neuronal-like cells and investigated their
fate following immediate or delayed injection into
the contused spinal cord of adult rats. Their find-
ings demonstrated that BMSCs administered 1-
week post-SCI had better rates of survival than
BMSCs administered at the time of injury. Fur-
thermore, the delayed treatment was associated
with significantly improved locomotor function.
Despite their success in improving neurological
recovery rates, electrophysiological and
immunohistochemical analysis of these differenti-
ated BMSCs demonstrated that the transplanted
cells were likely rather immature, which lacked
functional sodium and potassium channels. Still
the effectiveness of this approach administered
within a clinically relevant therapeutic window of
opportunity makes BMSCs a promising strategy
for SCI. To date, two clinical trials have been
preformed in patients with SCI.
In South Korea, Park et al. (2005) evaluated the
therapeutic efficacy of combining autologous
BMSC transplantation, administered directly into
the spinal cord lesion site, with granulocyte ma-
crophage-colony stimulating factor (GM-CSF),
given subcutaneously, in six patients with com-
plete SCIs. At the 6- and 18-month follow-up pe-
riods, four of the six patients showed neurological
improvements by two ASIA grade (from ASIA A
to ASIA C), while another improved from ASIA A
to ASIA B. Moreover, BMSC transplantation to-
gether with GM-CSF was not associated with in-
creased morbidity or mortality.
The second clinical trial, which was conducted
in Prague, Czech Republic, is a Phase I study that
has enrolled 20 patients with transversal SCIs to
test the safety of autologous bone marrow cell
implantation (Syková et al., 2006). The investiga-
tors compared intra-arterial versus i.v. routes of
227
administration and a group of acute SCI patients
(10–30 days post-SCI) against patients sustaining
chronic SCIs. Motor evoked potential, somato-
sensory evoked potential, magnetic resonance im-
aging, and ASIA scores were used in patient
follow-up. The results of this study demonstrated
that the BMSC-mediated repair is associated with
modest improvements in acute patients and that
BMSC transplantation is a relatively safe proce-
dure. Thus, it is anticipated that a Phase II clinical
trial designed to test the efficacy will be initiated in
the near future.
Oligodendrocyte precursor cells
Since the neurological disability in SCI is associ-
ated with oligodendrocyte death, resulting in de-
myelination and axonal degeneration, one would
anticipate that OPC-mediated repair would pro-
duce significant improvements in neural recovery.
In preclinical research OPC transplant alone or in
combination with the glycoprotein molecule Sonic
hedgehog (Shh), to further drive OPCs toward an
oligodendroglial lineage, was administered to
adult rats 5 days following a moderate thoracic
SCI (Bambakidis and Miller, 2004). The results of
this treatment regimen demonstrated that trans-
plantation of OPCs improves axonal conduction
and spinal cord function in the injured spinal cord.
Furthermore, the therapeutic benefits were more
pronounced when OPC treatment was given in
combination with Shh, while treatment of injured
rats with Shh on its own resulted in the prolifer-
ation of an endogenous population of neural pre-
cursor cells (Bambakidis and Miller, 2004).
The most frequently investigated OPC is O-2A,
which can differentiate into oligodendroglial cells
and type-2 astrocytes in vitro, but only into
oligodendrocytes following in vivo transplantation
into brain or spinal cord (Cao et al., 2002). O-2A
cell transplant demonstrates ample capacity to
migrate to regions adjacent to the site of in-
jury, differentiate into oligodendrocytes, remyelin-
ate axons, and improve on functional, behavioral,
electrophysiological, and morphological meas-
urable outcome variables (Lee et al., 2005).
Still, a major shortcoming associated with
228
O-2A-mediated repair is their relatively slow pro-
liferation rates, which effectively limits their ability
to self-renewal.
Olfactory ensheathing cells
OECs are specialized glial cells that surround the
olfactory sensory axons of the first cranial nerve
within the nose, which possess neuroregenerative
properties akin to Schwann cells of the peripheral
nervous system (Doucette, 1995). OECs are suit-
able candidates for cell-mediated repair following
SCI because this cell type can support axonal
outgrowth and regrowth through the adult CNS;
they are able to remyelinate the axons that have
undergone primary demyelination; and they can
migrate within the CNS and coexisting in an
astrocyte-rich environment and therefore fully
integrating within a lesion (Barnett, 2004). This
latter characteristic of OECs to integrate with
astrocytes sets them apart from Schwann cells.
There are now several reports documenting the
robust therapeutic efficacy of OEC transplantation
in animal models of SCI. For instance, Li et al.
(1998) have demonstrated that cultured OECs in-
jected into focal lesions within the adult rat cor-
tical spinal tract begin to regenerate within the first
week following transplantation. Moreover, the cut
corticospinal axons extended caudally and en-
sheathed by P0-positive peripheral myelin had ac-
cumulated into parallel bundles, which
appropriately extended across the full length of
the lesioned area and reentered the caudal part of
the host corticospinal tract (Li et al., 1998).
Ramon-Cueto et al. (2000) showed that OEC
transplantation can provide improved rates of
functional and structural recovery following a
complete spinal cord transection from 3 to 7
months post-transplant and corticospinal, raphe-
spinal, and coeruleospinal tracts all regenerated
for long distances within caudal cord stumps.
With these and other findings promoting the
utility for OECs to be used in neural repair follow-
ing SCI clinical trials have arisen around the world.
A Phase I clinical trial was initiated in 2001, in
Brisbane, Australia, to investigate the safety of
OEC transplantation into the injured thoracic
(T4-T10) spinal cord, 6 months to 3 years post-
injury (Feron et al., 2005). Only eight patients were
admitted into the trial (four patients to receive OEC
transplantation and four sham-operated controls).
All participants were closely monitored before the
trial was started and for 3 years following the trial
admittance. All patients underwent regular MRI
examination, along with physical, psychiatric,
neurological, and neurophysiological assessments
by medical staff and physiotherapy and occupa-
tional therapy assessments by respective therapists.
By far the largest of clinical trials testing the
efficacy of OEC transplants to treat SCI patients
comes from Beijing, China, where 171 subjects
have participated (Huang et al., 2003). Patients
aged 2–64 years, had traumatic or acute compress-
ive injuries due to epidural hematoma. Outcome
variables included motor score, light touch, and
pin prick scores according to the International
Standards for Neurological and Functional Clas-
sification of SCI by ASIA and the International
Society of Paraplegia before and 2 and 8 weeks
post-intervention. The conclusions made from this
study were that OEC transplantation can partially
preserve spinal functions regardless of patient age.
However, the recent publication by Dobkin et al.
(2006b) warns the SCI community of the proce-
dure of OEC transplantation as performed by Dr.
Huang as not meeting international standards of a
Phase I safety or Phase II efficacy trial. Consid-
ering the rather poorly defined inclusion and ex-
clusion criteria, coupled with the inconsistencies in
OEC injection sites, a lack of observed functional
benefit, and perioperative morbidity caution must
be taken when assessing these anecdotal claims.
Conclusion
In addition to methylprednisolone, GM-1 ganglio-
side, and Fampridine a number of novel pharma-
cological and cell-based approaches appear to be
on the horizon for the therapeutic management of
SCI (Table 1). Considering the complexity and
multifactorial nature of pathobiological mecha-
nisms that contribute to SCI, it is very likely that
combinatorial approaches will have the greatest
opportunity to improve on the quality of life of
 229

Table 1. Select pharmacological and cell-based therapies for spinal cord injury

Approach Proposed mechanism of action

Methylprednisolone A glucocorticosteroid that has demonstrated ability to inhibit posttraumatic lipid peroxidation and
inflammation in animal models of spinal cord injury (SCI), and has shown ability to improve
neurological recovery in spinal cord injured humans

GM-1 ganglioside A naturally occurring compound, residing within cell membranes of the mammalian CNS that has been
shown to promote neuritic sprouting, and has demonstrated ability to oppose excitotoxicity and
apoptosis
Minocycline A second generation tetracycline derivative with demonstrated ability to inhibit excitotoxicity, oxidative
stress, microglial cell activation, and caspase-dependent and caspase-independent-mediated pathways of
neuronal death
Fampridines A specific voltage-dependent potassium channel blocker shown to restore action potential conduction,
and enhance synaptic transmission in damaged demyelinated nerve fibers via blockade of exposed
potassium channels within the internodal membrane of injured axons

HP 184 A use- and frequency-dependent antagonist of sodium channels, and voltage-dependent blocker of
potassium channels designed to have less associated adverse effects than Fampridines
Cethrins A Rho antagonist, administered directly into the injured spinal cord in combination with a fibrin sealant
for improved drug delivery, to allow reversal of abnormal Rho activation in neurons and glial cells, and
promotion of neuroprotection and neuroregeneration following acute SCI
Anti-Nogo-A A monoclonal antibody targeted at Nogo-A to neutralize neurite outgrowth inhibition, and allow
monoclonal antibodies enhanced regeneration, compensatory sprouting, structural reorganization, plasticity, and functional
recovery following SCI
Procords Autologous macrophages, activated through co-incubation with regenerative autologous skin tissue, are
implanted into the injured spinal cord to elevate the secretion of protective cytokines, interleukin-1 beta
(IL-1b) and IL-6, the trophic factor, brain derived neurotrophic factor (BDNF), and the chemokine IL-
8, while reducing the secretion of the proinflammatory cytokine, tumor necrosis factor alpha (TNF-a);
all of which helps to make the injured environment of the lesioned spinal cord more feasible for axonal
regeneration, reduce cyst formation, and improve motor recovery

spinally injured individuals. Moreover, preclinical
analysis view SCI in a larger view with a thorough
examination of potentially beneficial effects in
addition to ambulation. For example, neuropro-
tective/regenerative approaches that yield modest
effects in enhancing rates of locomotor recovery
would be of bladder/bowel function and relieve
chronic pain. With regard to cell-based ap-
proaches, in addition to political and religious
conflicts, it is clear that careful planning and
design is required to obtain meaningful data.
References
AANS/CNS. (2002) Pharmacological therapy after acute cer-
vical spinal cord injury. Neurosurgery, 50: S63–S72.
Agnati, L.F., Fuxe, K., Calza, L., Benfenati, F., Cavicchioli, L.,
Toffano, G. and Goldstein, M. (1983) Gangliosides increase
the survival of lesioned nigral dopamine neurons and favour
the recovery of dopaminergic synaptic function in striatum of
rats by collateral sprouting. Acta Physiol. Scand., 119:
347–363.
Bambakidis, N.C. and Miller, R.H. (2004) Transplantation of
oligodendrocyte precursors and sonic hedgehog results in
improved function and white matter sparing in the spinal
cords of adult rats after contusion. Spine J., 4: 16–26.
Baptiste, D.C., Hartwick, A.T., Jollimore, C.A., Baldridge,
W.H., Seigel, G.M. and Kelly, M.E. (2004) An investigation
of the neuroprotective effects of tetracycline derivatives in
experimental models of retinal cell death. Mol. Pharmacol.,
66: 1113–1122.
Baptiste, D.C., Powell, K.J., Jollimore, C.A., Hamilton, C.,
LeVatte, T.L., Archibald, M.L., Chauhan, B.C., Robertson,
G.S. and Kelly, M.E. (2005) Effects of minocycline and
tetracycline on retinal ganglion cell survival after axotomy.
Neuroscience, 134: 575–582.
230
Barnett, S.C. (2004) Olfactory ensheathing cells: unique glial
cell types? J. Neurotrauma, 21: 375–382.
Bever Jr., C.T. (1994) The current status of studies of amino-
pyridines in patients with multiple sclerosis. Ann. Neurol.,
36(Suppl): S118–S121.
Bomstein, Y., Marder, J.B., Vitner, K., Smirnov, I., Lisaey, G.,
Butovsky, O., Fulga, V. and Yoles, E. (2003) Features of
skin-coincubated macrophages that promote recovery from
spinal cord injury. J. Neuroimmunol., 142: 10–16.
Bose, B., Osterholm, J.L. and Kalia, M. (1986) Ganglioside-
induced regeneration and reestablishment of axonal continu-
ity in spinal cord-transected rats. Neurosci. Lett., 63: 165–169.
Bracken, M.B., Collins, W.F., Freeman, D.F., Shepard, M.J.,
Wagner, F.W., Silten, R.M., Hellenbrand, K.G., Ransohoff,
J., Hunt, W.E., Perot Jr., P.L., et al. (1984) Efficacy of me-
thylprednisolone in acute spinal cord injury. JAMA, 251:
45–52.
Bracken, M.B., Shepard, M.J., Collins Jr., W.F., Holford, T.R.,
Baskin, D.S., Eisenberg, H.M., Flamm, E., Leo-Summers,
L., Maroon, J.C., Marshall, L.F., et al. (1992) Methyl-
prednisolone or naloxone treatment after acute spinal cord
injury: 1-year follow-up data. Results of the second national
acute spinal cord injury study. J. Neurosurg., 76: 23–31.
Bracken, M.B., Shepard, M.J., Collins, W.F., Holford, T.R.,
Young, W., Baskin, D.S., Eisenberg, H.M., Flamm, E., Leo-
Summers, L., Maroon, J., et al. (1990) A randomized, con-
trolled trial of methylprednisolone or naloxone in the treat-
ment of acute spinal-cord injury. Results of the second
national acute spinal cord injury study. N. Engl. J. Med., 322:
1405–1411.
Bracken, M.B., Shepard, M.J., Holford, T.R., Leo-Summers,
L., Aldrich, E.F., Fazl, M., Fehlings, M., Herr, D.L., Hitc-
hon, P.W., Marshall, L.F., Nockels, R.P., Pascale, V., Perot
Jr., P.L., Piepmeier, J., Sonntag, V.K., Wagner, F.,
Wilberger, J.E., Winn, H.R. and Young, W. (1997) Admini-
stration of methylprednisolone for 24 or 48 hours or tirilazad
mesylate for 48 hours in the treatment of acute spinal cord
injury. Results of the third national acute spinal cord injury
randomized controlled trial. National acute spinal cord in-
jury study. JAMA, 277: 1597–1604.
Braughler, J.M. and Hall, E.D. (1984) Effects of multi-dose
methylprednisolone sodium succinate administration on in-
jured cat spinal cord neurofilament degradation and energy
metabolism. J. Neurosurg., 61: 290–295.
Brundula, V., Rewcastle, N.B., Metz, L.M., Bernard, C.C. and
Yong, V.W. (2002) Targeting leukocyte MMPs and transmi-
gration: minocycline as a potential therapy for multiple scle-
rosis. Brain, 125: 1297–1308.
Cao, Q., Benton, R.L. and Whittemore, S.R. (2002) Stem cell
repair of central nervous system injury. J. Neurosci. Res., 68:
501–510.
Caroni, P., Savio, T. and Schwab, M.E. (1988) Central nervous
system regeneration: oligodendrocytes and myelin as non-
permissive substrates for neurite growth. Prog. Brain Res.,
78: 363–370.
Caroni, P. and Schwab, M.E. (1988) Antibody against
myelin-associated inhibitor of neurite growth neutralizes
nonpermissive substrate properties of CNS white matter.
Neuron, 1: 85–96.
Casha, S., Yu, W.R. and Fehlings, M.G. (2001) Oligoden-
droglial apoptosis occurs along degenerating axons and is
associated with FAS and p75 expression following spinal
cord injury in the rat. Neuroscience, 103: 203–218.
Chen, M.S., Huber, A.B., van der Haar, M.E., Frank, M.,
Schnell, L., Spillmann, A.A., Christ, F. and Schwab, M.E.
(2000) Nogo-A is a myelin-associated neurite outgrowth in-
hibitor and an antigen for monoclonal antibody IN-1. Na-
ture, 403: 434–439.
Cho, K.O., La, H.O., Cho, Y.J., Sung, K.W. and Kim, S.Y.
(2006) Minocycline attenuates white matter damage in a rat
model of chronic cerebral hypoperfusion. J. Neurosci. Res.,
83: 285–291.
Clark, B.R. and Keating, A. (1995) Biology of bone marrow
stroma. Ann. N.Y. Acad. Sci., 770: 70–78.
Coleman, W.P., Benzel, D., Cahill, D.W., Ducker, T., Geisler,
F., Green, B., Gropper, M.R., Goffin, J., Madsen III, P.W.,
Maiman, D.J., Ondra, S.L., Rosner, M., Sasso, R.C., Trost,
G.R. and Zeidman, S. (2000) A critical appraisal of the re-
porting of the national acute spinal cord injury studies (II
and III) of methylprednisolone in acute spinal cord injury. J.
Spinal Disord., 13: 185–199.
Cummings, B.J., Uchida, N., Tamaki, S.J., Salazar, D.L.,
Hooshmand, M., Summers, R., Gage, F.H. and Anderson,
A.J. (2005) Human neural stem cells differentiate and pro-
mote locomotor recovery in spinal cord-injured mice. Proc.
Natl. Acad. Sci., U.S.A., 102: 14069–14074.
DeForge, D., Nymark, J., Lemaire, E., Gardner, S., Hunt, M.,
Martel, L., Curran, D. and Barbeau, H. (2004) Effect of 4-
aminopyridine on gait in ambulatory spinal cord injuries: a
double-blind, placebo-controlled, crossover trial. Spinal
Cord, 42: 674–685.
Delamarter, R.B., Sherman, J. and Carr, J.B. (1995) Patho-
physiology of spinal cord injury. Recovery after immediate
and delayed decompression. J. Bone Joint Surg. Am., 77:
1042–1049.
Dimitrijevic, M.R., Gerasimenko, Y. and Pinter, M.M. (1998)
Evidence for a spinal central pattern generator in humans.
Ann. N.Y. Acad. Sci., 860: 360–376.
Dobkin, B., Apple, D., Barbeau, H., Basso, M., Behrman, A.,
DeForge, D., Ditunno, J., Dudley, G., Elashoff, R., Fugate,
L., Harkema, S., Saulino, M. and Scott, M. (2006a) Weight-
supported treadmill vs. over-ground training for walking
after acute incomplete SCI. Neurology, 66: 484–493.
Dobkin, B.H., Curt, A. and Guest, J. (2006b) Cellular trans-
plants in China: observational study from the largest human
experiment in chronic spinal cord injury. Neurorehabil.
Neural Repair, 20: 5–13.
Dommergues, M.A., Plaisant, F., Verney, C. and Gressens, P.
(2003) Early microglial activation following neonatal excito-
toxic brain damage in mice: a potential target for neuropro-
tection. Neuroscience, 121: 619–628.
Doucette, R. (1995) Olfactory ensheathing cells: potential for
glial cell transplantation into areas of CNS injury. Histol.
Histopathol., 10: 503–507.

Dubreuil, C.I., Winton, M.J. and McKerracher, L. (2003) Rho
activation patterns after spinal cord injury and the role of
activated Rho in apoptosis in the central nervous system. J.
Cell Biol., 162: 233–243.
Duysens, J. and Van de Crommert, H.W. (1998) Neural control
of locomotion: the central pattern generator from cats to
humans. Gait Posture, 7: 131–141.
Eriksson, P.S., Perfilieva, E., Bjork-Eriksson, T., Alborn, A.M.,
Nordborg, C., Peterson, D.A. and Gage, F.H. (1998) Ne-
urogenesis in the adult human hippocampus. Nat. Med., 4:
1313–1317.
Fass, B. and Ramirez, J.J. (1984) Effects of ganglioside treat-
ments on lesion-induced behavioral impairments and sprout-
ing in the CNS. J. Neurosci. Res., 12: 445–458.
Fehlings, M.G. and Perrin, R.G. (2006) The timing of surgical
intervention in the treatment of spinal cord injury: a system-
atic review of recent clinical evidence. Spine, 31: S28–S36.
Feron, F., Perry, C., Cochrane, J., Licina, P., Nowitzke, A.,
Urquhart, S., Geraghty, T. and Mackay-Sim, A. (2005) Au-
tologous olfactory ensheathing cell transplantation in human
spinal cord injury. Brain, 128: 2951–2960.
Firouzi, M., Moshayedi, P., Saberi, H., Mobasheri, H., Abol-
hassani, F., Jahanzad, I. and Raza, M. (2006) Transplanta-
tion of Schwann cells to subarachnoid space induces repair in
contused rat spinal cord. Neurosci. Lett., 402: 66–70.
Geisler, F.H., Coleman, W.P., Grieco, G. and Poonian, D.
(2001) The Sygen multicenter acute spinal cord injury study.
Spine, 26: S87–S98.
Geisler, F.H., Dorsey, F.C. and Coleman, W.P. (1991) Recov-
ery of motor function after spinal-cord injury — a randomi-
zed, placebo-controlled trial with GM-1 ganglioside. N. Engl.
J. Med., 324: 1829–1838.
Gruner, J.A. (1992) A monitored contusion model of spinal
cord injury in the rat. J. Neurotrauma, 9: 123–128.
Hall, E.D. and Braughler, J.M. (1981) Acute effects of intra-
venous glucocorticoid pretreatment on the in vitro peroxi-
dation of cat spinal cord tissue. Exp. Neurol., 73: 321–324.
Hall, E.D. and Braughler, J.M. (1982) Effects of intravenous
methylprednisolone on spinal cord lipid peroxidation and
Na++K+-ATPase activity. Dose-response analysis during
1st hour after contusion injury in the cat. J. Neurosurg., 57:
247–253.
Hansebout, R.R., Blight, A.R., Fawcett, S. and Reddy, K.
(1993) 4-Aminopyridine in chronic spinal cord injury: a con-
trolled, double-blind, crossover study in eight patients. J.
Neurotrauma, 10: 1–18.
Hicks, A.L., Adams, M.M., Martin Ginis, K., Giangregorio,
L., Latimer, A., Phillips, S.M. and McCartney, N. (2005)
Long-term body-weight-supported treadmill training and
subsequent follow-up in persons with chronic SCI: effects
on functional walking ability and measures of subjective well-
being. Spinal Cord, 43: 291–298.
Hofstetter, C.P., Schwarz, E.J., Hess, D., Widenfalk, J., El
Manira, A., Prockop, D.J. and Olson, L. (2002) Marrow
stromal cells form guiding strands in the injured spinal cord
and promote recovery. Proc. Natl. Acad. Sci. U.S.A., 99:
2199–2204.
231
Huang, H., Chen, L., Wang, H., Xiu, B., Li, B., Wang, R.,
Zhang, J., Zhang, F., Gu, Z., Li, Y., Song, Y., Hao, W.,
Pang, S. and Sun, J. (2003) Influence of patients’ age on
functional recovery after transplantation of olfactory en-
sheathing cells into injured spinal cord injury. Chin. Med. J.
(Engl.), 116: 1488–1491.
Hurlbert, R.J. (2000) Methylprednisolone for acute spinal cord
injury: an inappropriate standard of care. J. Neurosurg., 93:
1–7.
Hurlbert, R.J. (2001) The role of steroids in acute spinal cord
injury: an evidence-based analysis. Spine, 26: S39–S46.
Johansson, C.B., Momma, S., Clarke, D.L., Risling, M., Len-
dahl, U. and Frisen, J. (1999a) Identification of a neural stem
cell in the adult mammalian central nervous system. Cell, 96:
25–34.
Johansson, C.B., Svensson, M., Wallstedt, L., Janson, A.M.
and Frisen, J. (1999b) Neural stem cells in the adult human
brain. Exp. Cell Res., 253: 733–736.
Joshi, M. and Fehlings, M.G. (2002a) Development and char-
acterization of a novel, graded model of clip compressive spi-
nal cord injury in the mouse: Part 1. Clip design, behavioral
outcomes, and histopathology. J. Neurotrauma, 19: 175–190.
Joshi, M. and Fehlings, M.G. (2002b) Development and char-
acterization of a novel, graded model of clip compressive
spinal cord injury in the mouse: Part 2. Quantitative neuro-
anatomical assessment and analysis of the relationships be-
tween axonal tracts, residual tissue, and locomotor recovery.
J. Neurotrauma, 19: 191–203.
Karimi-Abdolrezaee, S., Eftekharpour, E., Wang, J., Mors-
head, C.M. and Fehlings, M.G. (2006) Delayed transplanta-
tion of adult neural precursor cells promotes remyelination
and functional neurological recovery after spinal cord injury.
J. Neurosci., 26: 3377–3389.
Kernie, S.G., Erwin, T.M. and Parada, L.F. (2001) Brain re-
modeling due to neuronal and astrocytic proliferation after
controlled cortical injury in mice. J. Neurosci. Res., 66:
317–326.
Knoller, N., Auerbach, G., Fulga, V., Zelig, G., Attias, J.,
Bakimer, R., Marder, J.B., Yoles, E., Belkin, M., Schwartz, M.
and Hadani, M. (2005) Clinical experience using incubated au-
tologous macrophages as a treatment for complete spinal cord
injury: phase I study results. J. Neurosurg. Spine, 3: 173–181.
Kohama, I., Lankford, K.L., Preiningerova, J., White, F.A.,
Vollmer, T.L. and Kocsis, J.D. (2001) Transplantation of
cryopreserved adult human Schwann cells enhances axonal
conduction in demyelinated spinal cord. J. Neurosci., 21:
944–950.
Lee, K.H., Yoon do, H., Park, Y.G. and Lee, B.H. (2005)
Effects of glial transplantation on functional recovery fol-
lowing acute spinal cord injury. J. Neurotrauma, 22:
575–589.
Li, S. and Strittmatter, S.M. (2003) Delayed systemic Nogo-66
receptor antagonist promotes recovery from spinal cord in-
jury. J. Neurosci., 23: 4219–4227.
Li, Y., Field, P.M. and Raisman, G. (1998) Regeneration of
adult rat corticospinal axons induced by transplanted olfac-
tory ensheathing cells. J. Neurosci., 18: 10514–10524.
232
Lucas, S.M., Rothwell, N.J. and Gibson, R.M. (2006) The role
of inflammation in CNS injury and disease. Br. J. Pharma-
col., 147(Suppl 1): S232–S240.
Man, S.M., Ma, Y.R., Shang, D.S., Zhao, W.D., Li, B., Guo,
D.W., Fang, W.G., Zhu, L. and Chen, Y.H. (2007) Periph-
eral T cells overexpress MIP-1 alpha to enhance its
transendothelial migration in Alzheimer’s disease. Neurobiol.
Aging, 28: 485–496.
Miklossy, J., Doudet, D.D., Schwab, C., Yu, S., McGeer, E.G.
and McGeer, P.L. (2006) Role of ICAM-1 in persisting in-
flammation in Parkinson disease and MPTP monkeys. Exp.
Neurol., 197: 275–283.
Nashmi, R. and Fehlings, M.G. (2001) Mechanisms of axonal
dysfunction after spinal cord injury: with an emphasis on the
role of voltage-gated potassium channels. Brain Res. Brain
Res. Rev., 38: 165–191.
Niederost, B., Oertle, T., Fritsche, J., McKinney, R.A. and
Bandtlow, C.E. (2002) Nogo-A and myelin-associated glyco-
protein mediate neurite growth inhibition by antagonistic
regulation of RhoA and Rac1. J. Neurosci., 22: 10368–10376.
Noorbakhsh, F., Tsutsui, S., Vergnolle, N., Boven, L.A.,
Shariat, N., Vodjgani, M., Warren, K.G., Andrade-Gordon,
P., Hollenberg, M.D. and Power, C. (2006) Proteinase-acti-
vated receptor 2 modulates neuroinflammation in experi-
mental autoimmune encephalomyelitis and multiple sclerosis.
J. Exp. Med., 203: 425–435.
Oudega, M. and Xu, X.M. (2006) Schwann cell transplanta-
tion for repair of the adult spinal cord. J. Neurotrauma, 23:
453–467.
Park, H.C., Shim, Y.S., Ha, Y., Yoon, S.H., Park, S.R., Choi,
B.H. and Park, H.S. (2005) Treatment of complete spinal
cord injury patients by autologous bone marrow cell trans-
plantation and administration of granulocyte-macrophage
colony stimulating factor. Tissue Eng., 11: 913–922.
Pellitteri, R., Russo, A. and Stanzani, S. (2006) Schwann cell: a
source of neurotrophic activity on cortical glutamatergic
neurons in culture. Brain Res., 1069: 139–144.
Pollard, S.M., Conti, L., Sun, Y., Goffredo, D. and Smith, A.
(2006) Adherent neural stem (NS) cells from fetal and adult
forebrain. Cereb. Cortex, 16: i112–i120.
Ramon-Cueto, A., Cordero, M.I., Santos-Benito, F.F. and
Avila, J. (2000) Functional recovery of paraplegic rats and
motor axon regeneration in their spinal cords by olfactory
ensheathing glia. Neuron, 25: 425–435.
Rapalino, O., Lazarov-Spiegler, O., Agranov, E., Velan, G.J.,
Yoles, E., Fraidakis, M., Solomon, A., Gepstein, R., Katz,
A., Belkin, M., Hadani, M. and Schwartz, M. (1998)
Implantation of stimulated homologous macrophages results
in partial recovery of paraplegic rats. Nat. Med., 4: 814–821.
Richardson, P.M., McGuinness, U.M. and Aguayo, A.J. (1980)
Axons from CNS neurons regenerate into PNS grafts. Na-
ture, 284: 264–265.
Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute
spinal cord compression in a new acute cord injury model in
the rat. Surg. Neurol., 10: 38–43.
Savio, T. and Schwab, M.E. (1989) Rat CNS white matter, but
not gray matter, is nonpermissive for neuronal cell adhesion
and fiber outgrowth. J. Neurosci., 9: 1126–1133.
Sekhon, L.H. and Fehlings, M.G. (2001) Epidemiology, demo-
graphics, and pathophysiology of acute spinal cord injury.
Spine, 26: S2–S12.
Stirling, D.P., Khodarahmi, K., Liu, J., McPhail, L.T.,
McBride, C.B., Steeves, J.D., Ramer, M.S. and Tetzlaff, W.
(2004) Minocycline treatment reduces delayed oligodendro-
cyte death, attenuates axonal dieback, and improves func-
tional outcome after spinal cord injury. J. Neurosci., 24:
2182–2190.
Stokes, B.T., Noyes, D.H. and Behrmann, D.L. (1992) An
electromechanical spinal injury technique with dynamic
sensitivity. J. Neurotrauma, 9: 187–195.
Stoy, N., Mackay, G.M., Forrest, C.M., Christofides, J., Eger-
ton, M., Stone, T.W. and Darlington, L.G. (2005) Try-
ptophan metabolism and oxidative stress in patients with
Huntington’s disease. J. Neurochem., 93: 611–623.
Syková, E., Jendelova, P., Urdzikova, L., Lesny, P. and Hejcl,
A. (2006) Bone marrow stem cells and polymer hydrogels:
two strategies for spinal cord injury repair. Cell. Mol.
Neurobiol., 26: 1111–1127.
Tator, C.H., Fehlings, M.G., Thorpe, K. and Taylor, W. (1999)
Current use and timing of spinal surgery for management of
acute spinal cord injury in North America: results of a
retrospective multicenter study. J. Neurosurg., 91: 12–18.
Teng, Y.D., Choi, H., Onario, R.C., Zhu, S., Desilets, F.C.,
Lan, S., Woodard, E.J., Snyder, E.Y., Eichler, M.E. and
Friedlander, R.M. (2004) Minocycline inhibits contusion-
triggered mitochondrial cytochrome c release and mitigates
functional deficits after spinal cord injury. Proc. Natl. Acad.
Sci. U.S.A., 101: 3071–3076.
Tikka, T.M. and Koistinaho, J.E. (2001) Minocycline provides
neuroprotection against N-methyl-D-aspartate neurotoxicity
by inhibiting microglia. J. Immunol., 166: 7527–7533.
Van de Crommert, H.W., Mulder, T. and Duysens, J. (1998)
Neural control of locomotion: sensory control of the central
pattern generator and its relation to treadmill training. Gait
Posture, 7: 251–263.
Walmsley, A.R., McCombie, G., Neumann, U., Marcellin, D.,
Hillenbrand, R., Mir, A.K. and Frentzel, S. (2004) Zinc
metalloproteinase-mediated cleavage of the human Nogo-66
receptor. J. Cell Sci., 117: 4591–4602.
Wells, J.E., Hurlbert, R.J., Fehlings, M.G. and Yong, V.W.
(2003) Neuroprotection by minocycline facilitates significant
recovery from spinal cord injury in mice. Brain, 126:
1628–1637.
Wiessner, C., Bareyre, F.M., Allegrini, P.R., Mir, A.K.,
Frentzel, S., Zurini, M., Schnell, L., Oertle, T. and Schwab,
M.E. (2003) Anti-Nogo-A antibody infusion 24 hours
after experimental stroke improved behavioral outcome and
corticospinal plasticity in normotensive and spontane-
ously hypertensive rats. J. Cereb. Blood Flow Metab., 23:
154–165.

Winton, M.J., Dubreuil, C.I., Lasko, D., Leclerc, N. and
McKerracher, L. (2002) Characterization of new cell perme-
able C3-like proteins that inactivate Rho and stimulate neu-
rite outgrowth on inhibitory substrates. J. Biol. Chem., 277:
32820–32829.
Xu, L., Fagan, S.C., Waller, J.L., Edwards, D., Borlongan,
C.V., Zheng, J., Hill, W.D., Feuerstein, G. and Hess, D.C.
(2004) Low dose intravenous minocycline is neuroprotective
after middle cerebral artery occlusion-reperfusion in rats.
BMC Neurol., 4: 7.
233
Zheng, M. and Kuffler, D.P. (2000) Guidance of regen-
erating motor axons in vivo by gradients of diffusible
peripheral nerve-derived factors. J. Neurobiol., 42:
212–219.
Zhu, S., Stavrovskaya, I.G., Drozda, M., Kim, B.Y., Ona, V.,
Li, M., Sarang, S., Liu, A.S., Hartley, D.M., Wu du, C.,
Gullans, S., Ferrante, R.J., Przedborski, S., Kristal, B.S. and
Friedlander, R.M. (2002) Minocycline inhibits cytochrome c
release and delays progression of amyotrophic lateral sclero-
sis in mice. Nature, 417: 74–78.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 16

Cerebral contusion: a role model for lesion

progression

Tatsuro Kawamata1,2, and Yoichi Katayama1,2

1
Department of Neurological Surgery, Nihon University School of Medicine, Tokyo, Japan
2
The Japan Neurotrauma Data Bank Committee (The Japan Society of Neurotraumatology, The Japanese Council of
Traffic Science)

Abstract: The early massive edema caused by severe cerebral contusion results in progressive intracranial
pressure (ICP) elevation and clinical deterioration within 24–72 h post-trauma. Surgical excision of the
necrotic brain tissue represents the only therapy, which can provide satisfactory control of the elevated ICP
and clinical deterioration. In this chapter, we review the results of our clinical studies regarding the
pathophysiology of contusion edema and evaluate the effects of surgical treatment, i.e. contusion necro-
tomy, by analyzing the data from the Japan Neurotrauma Data Bank.

Keywords: cerebral contusion; brain edema; diffusion-weighted image; vascular permeability; tissue
osmolality; contusion necrotomy

Introduction therapy, which can provide satisfactory control of

In patients with cerebral contusions, two types of
edema can be clinically recognized. One is the early
massive edema that occurs within the period of
24–72 h post-trauma. This type of edema creates
strong mass effect resulting in progressive elevation
of the intracranial pressure (ICP) and clinical dete-
rioration known as talk-and-deteriorate (Katayama
et al., 1990). The other is the delayed peri-contusion
edema, which is typically seen in the white matter
adjacent to the cerebral contusion, at several days
post-trauma by T2-weighted magnetic resonance
(MR) imaging. This type of edema rarely causes
ICP elevation leading to fatal deterioration.
Despite intensive medical therapy, the elevated
ICP in patients with early massive edema is often
uncontrollable and fatal. In such instances, surgi-
cal excision of the necrotic brain tissue is the only
Corresponding author. Tel.: +81-3-3972-8111 ext. 2481;
Fax: +81-3-3554-0425; E-mail: kawamata@med.nihon-u.ac.jp
the elevated ICP. The precise mechanism under-
lying such an early massive edema is not yet clearly
understood. We review the results of our clinical
studies (Katayama et al., 1990, 1992, 1998; Kushi
et al., 1994; Kawamata et al., 2000; Katayama and
Kawamata, 2003), which have provided several
lines of evidence to suggest that a large amount of
edema fluid is accumulated in the necrotic brain
tissue within the central area of contusion, and this
contributes to the early massive edema. The clini-
cal results of surgical treatment, i.e. contusion
necrotomy, are also discussed.
Histopathology of cerebral contusion
The classical histopathological study of Lindenberg
and Freytag (1957) demonstrated the presence of
two components of cerebral contusion; one is the
central (core) area in which cells undergo necrosis
as the primary consequence of mechanical injury

DOI: 10.1016/S0079-6123(06)61016-9 235

236

(contusion necrosis proper), and the other is the

peripheral (rim) area in which cellular swelling oc-
curs as a consequence of ischemia. A clear demar-
cation line separates these two components. The
area of contusion necrosis is histopathologically
evident as early as at 3 h post-trauma, as a primary
brain damage (Eriskat et al., 1994). The cellular
elements in the central area, both neuronal as well
as glial cells, uniformly undergo shrinkage, and
then disintegration, homogenation and cyst forma-
tion eventually. In contrast, cell swelling is pre-
dominant in the peripheral area (Lindenberg and
Freytag, 1957). The ischemia in the peripheral area
is largely attributable to microthrombosis, which is
a main cause of secondary brain damages. Numer-
ous clinical studies have demonstrated a decrease in
cerebral blood flow in the central as well as the
peripheral areas of contusion (e.g., Alexander
et al., 1994).

MR diffusion study (ADC mapping)

We have investigated the evolution of cerebral

contusion by diffusion-weighted MR imaging and
apparent diffusion co-efficient (ADC) mapping in
head trauma patients (Kawamata et al., 2000).
Following conventional T1- and T2-weighted MR
imaging, diffusion-weighted MR images are ob-
tained, and ADC mapping is computed from var-
ious b factor diffusion images.
The diffusion-weighted MR images demonstrate
a low intensity core in the central area, beginning
at approximately 24 h post-trauma. The ADC
value within this central area clearly increases
during the period of 24–72 h post-trauma (Fig. 1;
Kawamata et al., 2000). This elevated ADC value
appears to represent the contusion necrosis proper,
since cellular disintegration and homogenization
in this area would result in an expansion of the
extracellular space.
The diffusion-weighted MR images demonstrate
a high intensity rim in the peripheral area of contu-
sion beginning at approximately 24 h post-trauma,
which corresponds with the timing of the elevated
ICP and neurological deterioration. The combina-
tion of a low intensity core and a high intensity rim
is a consistent finding in the cerebral contusion
Fig. 1. A representative case of cerebral contusion in the acute
phase post-trauma. Upper: A CT scan revealed low-density
area in bilateral frontal lobe. Lower: Diffusion-weighted image
demonstrated low intensity core with surrounding high inten-
sity rim, which was typical pattern of diffusion-weighted image
in the acute phase (o48 h post-trauma) of cerebral contusion
without massive hemorrhage.
during this period, which can be termed a halo ap-
pearance (Fig. 2). The ADC value decreases in the
peripheral area during the period of 24–72 h post-
trauma (Kawamata et al., 2000). This decreased
ADC value appears to represent the cellular swell-
ing, which would result in shrinkage of the extra-
cellular space. The ADC values between the central
and peripheral areas are maximally dissociated
during the period of 24–72 h post-trauma.
The ADC value in the peripheral area shifts
from a decrease to an increase after 72 h post-
trauma. At the same time, an increase in ADC
value becomes evident in the adjacent white mat-
ter. This increase in ADC value in the adjacent
white matter appears to represent the delayed

Fig. 2. Bilateral contusion demonstrating a halo appearance with a combination of a low intensity core and a high intensity rim on
diffusion-weighted image (left). On ADC mapping (right), the peripheral rim showed low ADC value representing the cellular swelling
resulting from shrinkage of the extracellular space.
peri-contusion edema, which can commonly be seen
on T2-weighted MR images as vasogenic edema.
Increased cerebrovascular permeability
Since gadolinium (Gd)-DTPA, like plasma pro-
tein, does not normally cross the blood-brain bar-
rier, enhancement with Gd-DTPA, if observed in
association with cerebral contusion, implies an in-
creased cerebrovascular permeability. In a previ-
ous study (Lang et al., 1991), Gd-DTPA was
administered intravenously by bolus injection and
MR imaging was undertaken soon after the ad-
ministration. Such a procedure failed to detect any
increase in cerebrovascular permeability associ-
ated with cerebral contusions during the initial few
days post-trauma. Enhancement with Gd-DTPA
has been reported to become detectable at 6–9
days post-trauma.
Furthermore, an immunohistochemical study of
the post-mortem brain in such patients failed to
reveal any plasma protein leakage around the con-
tused brain areas during this early period (Todd
and Graham, 1990). Evaluations of the vascular
permeability by 99mTc pertechnetate single pho-
ton emission tomography failed to reveal any evi-
dence of an increased vascular permeability within
the area of contusion during the initial few days
post-trauma (Bullock et al., 1990). Such findings
237
appear to contradict the widely held view that an
increased cerebrovascular permeability is respon-
sible for the development of contusion edema
(Marmarou, 2003).
We examined the changes in cerebrovascular per-
meability employing intravenous slow infusion of
Gd-DTPA and delayed MR imaging (Kushi et al.,
1994). In general, increases in cerebrovascular per-
meability can be detected more clearly by intrave-
nous slow infusion of high dose contrast medium
and delayed neuroimaging (Ito et al., 1988). This
technique revealed that cerebral contusions can be
enhanced at as early as 24–48 h post-trauma. In MR
imaging undertaken at 2 h after Gd-DTPA admini-
stration, enhancement on T1-weighted images was
observed in either the central area or the peripheral
area, or both. It is evident that water supply from
the blood vessels is not completely interrupted even
in the central area of contusion.
The delayed peri-contusion edema has commonly
been attributed to an increased cerebrovascular
permeability. As mentioned above, enhancement
with Gd-DTPA has been reported to become de-
tectable at 6–9 days post-trauma. It is also possible,
however, that resolution of the cellular swelling
in the peripheral area of contusion might permit
propagation of the edema fluid accumulated within
the central area to the adjacent white matter. This
would lead to edema appearance on CT and MR
images which resembles that induced by an
238

increased cerebrovascular permeability, i.e. vaso-
genic edema.
Edema fluid accumulation in the central area
In approximately 50% of patients with cerebral
contusions, a crescent-shaped zone of very high
ADC value develops at the border between the
central and peripheral areas beginning at approxi-
mately 24 h post-trauma (Fig. 3; Kawamata et al.,
2000). This crescent-shaped zone was always lo-
cated within the central area of contusion. The
crescent-shaped zone still existed at 2 weeks post-
trauma in some cases (Fig. 4). The very high ADC
value appears to represent edema fluid accumula-
tion within the central area of contusion.
Edema fluid accumulation within the necrotic
brain tissue is also suggested by the formation of a
fluid–blood interface within cerebral contusions
(Fig. 5). The fluid–blood interface can be formed
without a fluid cavity. When we carry out surgery,
we find no fluid cavity but softened and water-rich
necrotic brain tissue is present (Katayama et al.,
1992). Within the contusion necrosis, hemorrh-
age undergoes enlargement within the initial 6 h

Fig. 3. Cerebral contusion in the temporal tip at 45 h post-trauma. Left: CT scan showed small low-density area in the temporal tip.
Middle: T2-weighted MRI revealed mixed intensity area in the core of contusion. Right: ADC mapping demonstrated a crescent-
shaped zone (arrows) of very high ADC value, as high as that of CSF, representing edema fluid accumulation within the central area of
contusion.

Fig. 4. Cerebral contusion in the subacute phase, 2 weeks post-trauma. Left: T2-weighted image. Middle: Diffusion-weighted image.
Right: ADC mapping. A crescent shaped zone still existed (arrows). ADC value in the peripheral area of contusion increased,
indicating that vasogenic edema turned to be predominant.

Fig. 5. CT scan and its scheme of cerebral contusion in the temporal tip (10 h post-trauma), showing fluid–blood interface in the area
of contusion necrosis. (A) edema fluid; (B) hematoma; (C) fluid–blood interface; (D) layering of red blood cells.
post-trauma (Katayama et al., 1990), and red
blood cells diffusely permeate the softened necrotic
brain tissue (Lindenberg and Freytag, 1957). The
fluid–blood interfaces observed within the central
area of contusion may represent layering of red
blood cells in the softened necrotic brain tissue,
which has accumulated voluminous edema fluid.
Osmotic potential of the contusion necrosis
We have reported that necrotic brain tissue sam-
pled from the central area of contusion during
surgery shows a very high osmolality, reaching
350–400 mOsm (Katayama et al., 1998). It is un-
certain whether or not such a marked increase in
osmolality is osmotically active and causes edema
fluid accumulation. It appears that expansion of
the extracellular space in the central area increases
the capacitance for edema fluid accumulation. In
contrast, shrinkage of the extracellular space in the
peripheral area increases the resistance for edema
fluid propagation or resolution. We hypothesize
that the barrier formed by swollen cells in the peri-
pheral area may prevent edema fluid propagation
and also help to generate osmotic potentials across
the central and peripheral areas. Since blood flow
is greatly reduced but is not completely interrupted
in the contused brain tissue, water is supplied from
the blood vessels into the central area. We suggest
that a combination of these events may facilitate
239
edema fluid accumulation in the central area and
contribute to the early massive edema of cerebral
contusion.
Surgical treatment
We investigated the effects of surgical excision of
the necrotic brain tissue in patients with severe cere-
bral contusion. The data from the Japan Neuro-
trauma Data Bank (JNTDB) in which a total of
1002 patients suffering severe traumatic brain
injury registered during the period between 1998
and 2001 were analyzed (Nakamura et al., 2002).
Among these patients, 182 (18%) demonstrated
severe cerebral contusions as the major lesions
contributing to their clinical status.
Results
Among the 182 patients with severe cerebral con-
tusion, 121 (66%; Group I) were treated conserv-
atively and the remaining 61 (34%; Group II)
underwent surgery. The ratio of selecting surgical
management was far lower in patients with cere-
bral contusion (34719%), as compared with those
with acute epidural hematoma (88711%) or acute
subdural hematoma (68718%). There was a huge
variation in the ratio of selecting surgical mana-
gement (9–77%) among the contributing centers
(n ¼ 10) of this data bank. While older patients,
240

especially those between 40 and 60 years old, mortality between the two groups was observed
tended to undergo surgery more frequently, even when the analysis was restricted to patients
younger patients tended to be treated conserva- who definitely demonstrated ‘‘talk-and-deteriorate’’
tively. There was, however, no significant differ- (64% vs. 22%; p ¼ 0.026; n ¼ 29; Table 2).
ence in age between Groups I and II (47.8723.8 The indications for surgical intervention in case
vs. 54.4719.5 years). The surgical management of severe cerebral contusion remain controversial
involved internal decompression (complete exci- (Bullock et al., 1989). The huge variation in ratio
sion of the necrotic brain tissue and evacuation of of selecting surgical management among the
clots) with or without external decompression in present centers reflects a diversity in management
most patients (90%) of Group II. The remaining policy and an absence of consensus regarding the
patients underwent external decompression alone. indications for surgery.
Surgery was performed at 1.8
86.1 (19.5724.2) h, The higher mortality in Group I, as compared
mostly (73%) within 24 h post-trauma. with Group II, suggests that the surgery performed
Group I demonstrated a clearly poorer outcome in the Group II patients helped to prevent their
on the Glasgow outcome scale (GOS) at 6 months clinical deterioration and death. Such an effect of
post-trauma (Table 1). The mortality was clearly surgery was most striking in those patients who
higher in Group I, as compared with Group II were scored at 9 or better at the time of admission.
(48% vs. 23%; p ¼ 0.0001; n ¼ 182). A difference It is by no means certain whether patients in
in mortality between the two groups was noted in Group II who were scored at 9 or better before
patients who were scored at 8 or less on the GCS at surgery would have deteriorated or not if surgery
the time of admission, but did not reach a statis- had not been carried out. An effect of surgery on
tically significant level (Table 1). The most striking mortality is evident, however, since a difference in
difference was observed in patients who were scored mortality was clearly observed even when the
at 9 or better on the GCS at the time of admission analysis was restricted to patients who definitely
(Table 1). The mortality was clearly higher in demonstrated ‘‘talk-and-deteriorate’’. This finding
Group I, as compared with Group II (56% vs. strongly suggests that the surgery itself was the
17%; p ¼ 0.017; n ¼ 45). A clear difference in major reason for the improved mortality in Group
II. In other words, death was probably prevented
by the surgery in many patients of Group II.
Table 1. Outcome (6 months post-trauma)
The present findings support our hypothesis that
GOS (%) early massive edema is caused by cerebral contu-
sion through the presence of necrotic brain tissue,
n GR MD SD VS D and indicate that surgical excision of the necrotic
GCS on admission: 3–5 brain tissue is the only therapy which can provide
Conservative 47 7 2 11 11 70 satisfactory control of the progressive elevation of
Surgical 11 9 9 27 0 55 the ICP and clinical deterioration in many cases.
GCS on admission: 6–8 Surgical intervention should be considered in
Conservative 58 29 21 10 10 29
Surgical 21 24 29 24 10 14
Table 2. Outcome (6 months post-trauma) in patients demon-
GCS on admission: 9–15 strating ‘‘talk-and-deteriorate’’
Conservative 16 19 13 13 0 56
Surgical 29 28 28 17 10 17 GOS (%)

Total n GR MD SD VS D
Conservative 121 19 12 12 9 48
Surgical 61 23 25 21 8 23 Conservative 11 18 9 9 0 64
Surgical 18 11 22 39 6 22
Note: GCS, Glasgow coma scale; GOS, Glasgow outcome scale.
 p ¼ 0.017. Note: GCS, Glasgow coma scale; GOS, Glasgow outcome scale.
 p ¼ 0.0001. p ¼ 0.026.

patients with severe cerebral contusion who demon-
strate ‘‘talk-and-deteriorate’’.
Conclusion
There is at present no established medical treat-
ment which can effectively inhibit edema fluid ac-
cumulation within cerebral contusions. The most
effective therapy for ameliorating the potentially
fatal edema is surgical excision of the necrotic
brain tissue. The effects of surgical excision of ne-
crotic brain tissue have commonly been accounted
for on the basis of an increased space compensa-
tion for mass lesions. It is possible, however, that
excision of the necrotic brain tissue does mean
elimination of the cause of edema fluid accumu-
lation. If the ICP is elevated by early massive ed-
ema due to cerebral contusion and the elevated
ICP is medically uncontrollable, surgical excision
of the necrotic brain tissue would appear to rep-
resent the therapy of choice, regardless of the size
of the associated hemorrhages. Surgery should be
considered in patients who are scored at 9 or better
at the time of admission, as soon as they show
clinical deterioration.
References
Alexander, M.J., Martin, N.A., Khanna, M., Caron, M. and
Becker, D.P. (1994) Regional cerebral blood flow trends in
head injured patients with focal contusions and cerebral
edema. Acta Neurochir., s60: 479–481.
Bullock, R., Golek, J. and Blake, G. (1989) Traumatic intra-
cerebral hematoma Which patients should undergo surgical
evacuation? CT scan features and ICP monitoring as a basis
for decision making. Surg. Neurol., 32: 181–187.
Bullock, R., Statham, J., Patterson, D., Wyper, D., Hadley, D.
and Teasdale, E. (1990) The time course of vasogenic oedema
after focal human head injury: evidence from SPECT mapping
of blood brain barrier defects. Acta Neurochir., s51: 286–288.
241
Eriskat, J., Schurer, L., Kempski, O. and Baethmann, A. (1994)
Growth kinetics of a primary brain tissue necrosis from focal
lesion. Acta Neurochir., s60: 425–427.
Ito, U., Reulen, H.-J., Tomita, H., Ikeda, J., Saito, J. and
Maehara, T. (1988) Formation and propagation of brain
oedema fluid around human brain metastasis: a CT study.
Acta Neurochir., 90: 35–41.
Katayama, Y. and Kawamata, T. (2003) Edema fluid accumu-
lation within necrotic brain tissue as a cause of the mass
effect of cerebral contusion in head trauma patient. Acta
Neurochir., s86: 323–327.
Katayama, Y., Mori, T., Maeda, T. and Kawamata, T. (1998)
Pathogenesis of the mass effect of cerebral contusions: a
rapid increase in osmolality within the contusion necrosis.
Acta Neurochir., s71: 289–292.
Katayama, Y., Tsubokawa, T., Kinoshita, K. and Himi, K.
(1992) Intra-parenchymal fluid-blood levels in traumatic in-
tracerebral hematomas. Neuroradiology, 34: 381–383.
Katayama, Y., Tsubokawa, T., Miyazaki, S., Kawamata, T.
and Yoshino, A. (1990) Oedema fluid formation within con-
tused brain tissue as a cause of medically uncontrollable el-
evation of intracranial pressure in head trauma patients. Acta
Neurochir., s51: 308–310.
Kawamata, T., Katayama, Y., Aoyama, N. and Mori, T. (2000)
Heterogeneous mechanisms of early edema formation in cere-
bral contusion: diffusion MRI and ADC mapping study.
Acta Neurochir., s76: 9–12.
Kushi, H., Katayama, Y., Shibuya, T., Tsubokawa, T. and
Kuroha, T. (1994) Gd-DTPA enhanced magnetic resonance
imaging of cerebral contusions. Acta Neurochir., s60:
472–474.
Lang, D.A., Hadley, D.M., Teasdale, G.T., Macpherson, P.
and Teasdale, E. (1991) Gadolinium DTPA enhanced
magnetic resonance imaging in acute head injury. Acta
Neurochir., 109: 5–11.
Lindenberg, R. and Freytag, E. (1957) Morphology of cortical
contusions. AMA Arch. Pathol., 63: 23–42.
Marmarou, A. (2003) Pathophysiology of traumatic brain
edema: current concepts. Acta Neurochir., s86: 7–10.
Nakamura, N., Yamaura, A., Shigemori, M., Ono, J., Kawamata,
T., Sakamoto, T. and Japanese Data Bank Committee for
Traumatic Brain Injury. (2002) Epidemiology, prevention and
countermeasures against severe traumatic brain injury in Japan
and abroad. Neurol. Res., 24: 45–53.
Todd, N.V. and Graham, D.I. (1990) Blood-brain barrier
damage in traumatic brain contusion. Acta Neurochir., s51:
296–299.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 17

Ethical implications of time frames in a randomized

controlled trial in acute severe traumatic brain injury

Erwin J.O. Kompanje1,2,, Andrew I.R. Maas2, Franc- ois J.A. Slieker2 and
Nino Stocchetti3

1
Department of Intensive Care, Erasmus MC University Medical Center Rotterdam, P.O. Box 2040, 3000 CA Rotterdam,
The Netherlands
2
Department of Neurosurgery, Erasmus MC University Medical Center Rotterdam, P.O. Box 2040, 3000 CA Rotterdam,
The Netherlands
3
Ospedale Policlinico IRCCS, Milan University, Milan, Italy

Abstract: Objectives: To analyze factors determining the time between injury and study drug administra-
tion (SDA) in a randomized controlled trial (RCT) of acute severe traumatic brain injury (TBI) and to
discuss the ethical implications. Methods: Time frames prior to SDA, differentiated per country, were
analyzed in a recently conducted RCT in severe TBI. Per protocol, the time window for SDA was 6 h after
injury. We selected patients for whom written proxy consent (PC) was obtained prior to SDA (n ¼ 631).
Results: The time between injury and admission to the neurotrauma center (NTC) varied per country from
1.16 to 2.35 h, but CT scan was obtained on average within 1 h of admission. The median time between
injury and CT scan was within 3 h in all but one country. The broadest time window was observed between
CT scan and obtaining required PC (1.71–2.74 h). The median time between injury and PC varied between
countries from 3.75 to 5.00 h. After consent had been obtained, almost all patients subsequently received
study drug within 1 h. In 85.3% of all cases time between injury and SDA exceeded 4 h, in 60% 5 h.
Conclusions: The requirement of written PC causes a significant delay in SDA in TBI. With deferred
consent, the first dose of an investigational drug could potentially be administered directly after completion
of the admission CT scan, which reduce the time to SDA by 50%. We argue that randomization under
deferred consent is ethically defendable for emergency research in severe TBI. Recommendations for
patient protection are proposed.

Keywords: traumatic brain injury; informed consent; emergency trial; deferred consent

Introduction TBI is about 30% and a significant disability per-

Severe traumatic brain injury (TBI) remains a
major cause of death and disability afflicting
mostly young adult males and elderly people,
and results in high economic costs to society
(McGarry et al. 2002). The fatality rate for severe
Corresponding author. Tel.: +31-6-53837655; Fax: +31-10-
4634075; E-mail: e.j.o.kompanje@erasmusmc.nl
sists in a further 35–40%. These data signify an
ethical imperative to develop and test neuropro-
tective agents and other new therapeutic strategies.
Various neuroprotective agents, mainly targeting
specific pathophysiologic mechanisms, have been
tested in TBI, but convincing benefit has not been
shown (Maas et al., 1999). Randomized controlled
trials (RCTs) in emergency and intensive care
medicine pose complex ethical and methodological

DOI: 10.1016/S0079-6123(06)61017-0 243

244
challenges (Maas et al., 2004; Kompanje et al.,
2005).
Specific ethical issues pertaining to clinical
testing of neuroprotective agents in TBI include
the emergency nature of the research, the incapac-
ity of the patients to informed consent before in-
clusion, short therapeutic time windows, and a
risk–benefit ratio based on concept that in relation
to the severity of the trauma, significant adverse
side effects may be acceptable for treatments
with proven benefit. The relevance of these issues
and implications for trial design are, however,
not fully recognized in- and outside the expert
field. Legislation in countries in the European
Union is being amended to comply with the
European Union Directive 2001/20/EC (European
Union, 2001). In this new European legislation,
emergency research under deferred or waiver of
consent is not permitted. This will impede or
even obviate emergency research phase III trials in
TBI in the European countries (Kompanje and
Maas, 2004; Silverman et al., 2004; Liddell et al.,
2006a, b).
We aimed to analyze the implications of critical
time frames (time between injury and admission to
a neurotrauma center (NTC), between admission
and the first cranial CT scan, between CT scan and
Time window as main reason for exclusion
5000
4000
3000
2000 671
1000
0
time window as time window +
only reason for other reason for
exclusion
patients 671
Fig. 1. Reasons for exclusion (n ¼ 6303).
proxy consent (PC), and between PC and study
drug administration (SDA)) in a recently
completed RCT in TBI.
Materials and methods
We analyzed critical time frames in a multi-center
placebo-controlled phase III trial, investigating the
efficacy and safety of a single dose of dexanabinol
in severe TBI. This RCT was conducted from
January 2001 to March 2004 in Europe, Israel,
Australia, and the United States. Dexanabinol is a
synthetic cannabinoid analogue with strong
neuroprotective potential and devoid of psycho-
tropic activity. The protocol stipulated SDA
within 6 h after injury. The study recruited 861
patients with severe TBI. No beneficial effect of
dexanabinol was found. Full details and results of
the study have been published (Maas et al., 2006).
In total, 7164 patients were screened for partic-
ipation in the trial. Enrollment criteria were not met
in 6303 patients. Of these, the sole reason for
exclusion was the time window in 671 patients and
in a further 944 patients inability to give study
medication within 6 h was one of the reasons for
exclusion (Fig. 1).
4688
944
other reason for
exclusion
exclusion
944 4688
 245

In view of the severity of the brain injury, in-
formed consent could not be obtained from the
patients themselves. PC was accepted in all coun-
tries. Deferred consent was allowed in Australia,
Austria, Finland, and in some centers in France
and Germany. Consent by an independent physi-
cian was allowed in Israel, Italy, Spain, and the
United Kingdom. In all cases of deferred consent,
subsequent written assent by patient or proxy was
obtained.
For the analysis of time frames we selected pa-
tients enrolled in Europe and Israel, in whom PC
was obtained before SDA (n ¼ 631). We excluded
patients from the United States for reason of
incomplete screening logs due to HIPAA regula-
tions (Maas et al., 2005; Kompanje and Maas,
2006), and patients from Australia seeing the
different infrastructure of this country in compar-
ison to European countries.
Results
The median age of our study population was 32
years (IQR 23–44); 520 (82.4%) were male, 111
(17.6%) female. The time between injury and ad-
mission to the NTC varied between countries from
1.16 to 2.35 h (Table 1, Fig. 2). A total of 501

Table 1. Time windows per country (median+IQR)

Country (N) Hours between injury Hours between injury Hours between injury Hours between injury
and admission NTC and CT scan median and obtained consent and SDA median
median (IQR) (IQR) median (IQR) (IQR)

Belgium (23) 0.93 (0.65–1.27) 1.80 (1.28–2.27) 3.75 (2.75–4.75) 4.60 (3.98–5.42)
Netherlands (73) 1.00 (0.75–1.33) 1.65 (1.32–2.00) 4.53 (3.95–5.05) 5.53 (5.07–5.75)
Israel (116) 0.93 (0.72–1.40) 1.91 (1.58–2.47) 4.01 (3.20–4.83) 4.67 (4.00–5.33)
Spain (75) 1.33 (0.97–1.67) 2.07 (1.65–2.53) 4.17 (3.33–5.00) 5.17 (4.30–5.58)
Germany (109) 1.20 (0.88–2.00) 1.65 (1.30–2.13) 4.08 (3.42–4.98) 5.25 (4.25–5.67)
Italy (146) 1.25 (0.83–2.60) 1.77 (1.40–2.35) 4.92 (4.08–5.28) 5.50 (4.98–5.75)
France (34) 2.17 (1.42–3.00) 3.08 (1.97–3.53) 5.00 (4.50–5.38) 5.75 (5.17–5.83)
Other countriesa (55) 1.47 (1.00–2.67) 1.82 (1.33–2.50) 4.00 (3.08–4.75) 5.25 (4.33–5.75)
a
Countries with small patient populations (United Kingdom, Denmark, Austria, Poland, and Turkey) were combined.

other

france

italy

germany

spain

israel

netherlands

belgium

injury 1h 2h 3h 4h 5h 6h
time between injury and admission NTC
time between admission and ct-scan
time between ct-scan and obtained consent
time between consent and SDA

Fig. 2. Time between injury and admission neurotrauma center, time between admission and first CT scan, time between first CT scan
and informed consent for inclusion in trial and time between consent and start study drug admission.
246
379
400
350
300
number of patients
250
200 168
150
100 70
50 13
1
0 given the high morbidity and mortality. Neuro-
1-2hours 2-3hours 3-4hours 4-5hours >5hours
time between injury and study drug admission
Fig. 3. Distribution of number of patients. Time between injury
and study drug administration (n ¼ 631).
(79.4%) patients were directly admitted to the
NTC, and 130 (20.6%) concerned secondary refer-
rals. In 71 secondarily referred patients the CT
scan was made before admission to the NTC. The
time between admission and the first diagnostic
CT scan was within 1 h in all patients. With the
exception of France, in all countries the median
time between injury and CT scan was within 3 h
(Table 1). The broadest time window was found
between the admission CT scan and obtaining the
required PC (between 1.71 and 2.74 h). The median
time between injury and obtained PC varied be-
tween 3.75 and 5.00 h (IQR 2.75–5.38 h) (Table 1).
After consent had been given, almost all patients
subsequently received the study drug within 1 h
(Table 1, Fig. 2). In 85.3% of all cases the time
between injury and SDA exceeded 4 h, in 60% of
the cases it even exceeded 5 h (Fig. 3). In total, 139
patients were randomized with deferred consent.
The median time between injury and SDA in this
group was 4.75 h (IQR 3.90–5.67). These 139 pa-
tients are not included in our analysis.
Discussion
Specific ethical issues pertaining to clinical evalu-
ation of neuroprotective agents in TBI include the
emergency nature of the research, short therapeu-
tic windows, the incapacity of the patients to
informed consent before inclusion, and a
risk–benefit ratio based on the concept that in
relation to the severity of the trauma, significant
adverse side effects may be acceptable for treat-
ments with proven benefit (Kompanje et al., 2005).
The emergency nature of research and short
therapeutic time windows in TBI
Severe TBI is an emergent and life-threatening
condition and existing therapy is unsatisfactory
logical damage does not only occur at the moment
of impact, but can evolve over the following hours
and days. Deleterious effects of this progressive
damage are determined at clinical and biochemical
levels. This has led to the development of new
pharmaceuticals with promising potential to limit
secondary damage and to improve outcome. One
of these new pharmaceuticals was dexanabinol,
but efficacy in the treatment of severe TBI was not
demonstrated (Maas et al., 2006). Recruitment in
this study was relatively slow and the majority of
patients enrolled toward the end of the 6 h time
window stipulated in the protocol. We hypothe-
sized that the relatively slow recruitment and lack
of demonstrable efficacy in the clinical situation
may have been influenced by informed consent
procedures and resulting delays in SDA. Early in-
tervention would appear crucial to the effect of
neuroprotective agents. Experimental data have
consistently shown better protection the sooner an
agent is administered after TBI (Hoff, 1986). For
dexanabinol, experimental studies have shown
beneficial effects with administration within 3 h
after injury, demonstrating protection against
breakdown of the blood-brain barrier, reduction
of edema formation, and improved outcome
(Shohami, 1995). No significant reduction of cer-
ebral edema was noted if the drug was administered
between 4 and 6 h after injury, but some improve-
ment in neurological symptoms was found. Based
on these findings, it may be concluded that in the
experimental model the pathophysiologic time win-
dow can be determined at 3 h. Whether this exper-
imental time window may be extrapolated to the
clinical situation in patients with TBI remains

uncertain. Prespecified subgroup (patients who
received the SD within 4 h and patients who re-
ceived the SD after 4 h) analysis showed no signifi-
cant differential treatment effect (Maas et al.,
2006). Nonetheless, it may be concluded in general
that chances of efficacy are greater if treatment is
provided earlier. We found that in almost all of the
studied cases the time between injury and comple-
tion of the primary diagnostic CT scan remained
within 3 h post injury, which corresponds to the
therapeutic time window in the animal model. In
60% of the cases, however, the time between injury
and SDA was more than 5 h, and in 85.3% of all
cases more than 4 h (Fig. 3).
We have shown that the main determinant of
the time to SDA is formed by the time required to
obtain informed (proxy) consent, and these results
indicate that delays in SDA could be reduced by
50% with the adoption of deferred consent or
waiver of consent. We could, however, not
demonstrate shorter inclusion times in the cohort
of patients randomized with deferred consent. The
reason for the delay in this cohort was that most
investigators waited for PC before randomization,
and only used deferred consent when at the end of
the inclusion boundary. These observations favor
adopting deferred consent procedures in trials in
acute TBI as primary approach, rather than as
‘‘ultimum refugium’’, only to be undertaken if PC
cannot be obtained. Other studies have demon-
strated advantages in using the deferred consent
and waiver of consent in emergency research. In
the National Acute Brain Injury Study: Hypother-
mia (NABIS-H) the adoption of waiver of consent
resulted in higher enrollment and reduced the time
between injury and treatment by approximately
45 min (Clifton et al., 2002). In this study, relatives
of only 11 out of 113 patients arrived within 6 h
after the injury. In a septic shock trial the inves-
tigators could not contact the proxies within the
inclusion time in 74% of the cases, and these were
included under waiver of consent (Annane et al.,
2004). In the CRASH trial, mean time to random-
ization was significantly longer in those hospitals
where consent was required compared with those
it was not (4.4 h [SE ¼ 0.21] versus 3.2 h
[SE ¼ 0.16]), the difference in the mean time to
randomization was 1.2 h [95% CI 0.7–1.8 h]
247
(CRASH trial management group, 2004). The ob-
servation in the cohort of patients enrolled in the
dexanabinol trial with deferred or waiver of con-
sent, that (proxy) assent was obtained later in all
cases supports the concept of accepting deferred or
waiver of consent for emergency research in TBI
trials.
The acute incapacity of the patients and informed
consent
All patients with severe TBI are unconscious, and
consequently informed consent can never be ob-
tained from the patients themselves. As substitute
for informed consent, a legal representative must
give consent for inclusion in research, or in the
absence thereof, by proxies; alternatively consent
may be deferred or waived. We have argued that
deferred consent is preferable. However, the new
European Union Directive (2001/20/EC) (Euro-
pean Union, 2001; Liddell et al., 2006a, b) stipu-
lates a requirement for informed (proxy) consent.
This is motivated by respect for the autonomy of
patients, and to ensure that that patient’s wishes
are guaranteed as far as possible. The requirement
for PC assumes that relatives are available in emer-
gency situations, and that these relatives can be fully
informed and given sufficient time to make a bal-
anced, ethically valid decision in a relatively short
time period under emotional distress. Even when
proxies are available, many are not aware of the
patient’s wishes (Luce, 2003). Surrogate decision
makers for critical care research resulted in false-
positive consent rates of 16–20.3% (Coppolino
and Ackerson, 2001). The emotional nature of an
emergency situation limits the reliability of PC
for clinical research (Mason and Allmark, 2000;
Coppolino and Ackerson, 2001; Hsieh et al.,
2001). Under emergency circumstances, PC does
not seem to secure proper patient/subject protec-
tion. In our experience, the validity of informed
consent and PC given in an emergency situation is
at least troubling. When consent for clinical re-
search is sought during an emergency situation,
comprehension is generally less than optimal (Cut-
tini, 2000; Sugarman, 2000; Williams et al., 2003).
A small minority realizes that pharmacological
248
trials are designed to assess not only efficacy but
safety as well (Harth and Thong, 1995). Are pa-
tients willing to be represented by their close rel-
atives? Roupie et al. (2000) found that only 40.6%
of 1089 patients wanted their spouse/partner to be
their surrogate, 28% wanted to be represented by
the physician in charge of their care.
One study searching for public views on emer-
gency exception to informed consent found that
most of 530 people (88%) believed that research
subjects should be informed prior to being en-
rolled, while 49% believed enrolling patients with-
out prior consent would be acceptable in an
emergency situation and 70% (369) would not ob-
ject to be entered into such a study without pro-
viding prospective informed consent (McClure
et al., 2003). In another study 11 of 12 stroke
patients stated that, if the patient or family was
not able to consent, the treating physician should
make the decision for inclusion in an emergency
trial (Blixen and Agich, 2005).
Furthermore, the requirement for all patients to
give written informed consent before enrollment
can result in major selection biases, such that en-
rolled patients may not be representative of the
typical patient (Tu et al., 2004).
Although respect for a patient’s autonomy is a
guiding principle in human rights, we doubt very
much whether this can be guaranteed by manda-
tory PC in acute TBI research. A balance should
be sought between optimal patient protection and
research to advance the standards of clinical care
in emergency situations to the benefit of society.
Risk– benefit ratio and patient protection
In our opinion the balance between risk and ben-
efit should be the guiding principle in emergency
research in severe TBI. This also applies to the
nature and the type of consent procedures. The
ethical principle of respect for the autonomy of the
patient underpinning the informed consent proce-
dures is not valid for acutely incapacitated patients
as TBI victims (Kompanje et al., 2005). Significant
concerns have been raised on the validity and eth-
ics of PC in acute emergency situations, and the
required written consent causes a significant delay
in treatment initiation, as we have shown by our
analysis. The possible therapeutic benefit, with
short therapeutic time windows in experimental
models, forms the moral justification for random-
izing patients under deferred consent or waiver of
consent within a sufficient period of time. The
risks should however be acceptable in relation to
the severity of the disease or injury.
For an effective agent in life-threatening condi-
tion adverse effects may be expected and accepted
in some patients. Careful monitoring and follow-
up of such adverse events is mandatory.
For trials under deferred consent or waiver of
consent in acute emergency situations we propose
to institute an independent safety committee, un-
der the auspices of regulatory authorities. Such an
independent safety committee, without (financial)
ties to industry or investigators, offers the best
safeguard for patient protection. The obligation
to such a committee is based on the experience of
a dramatically harmful outcome in some trials
under waiver of consent in other fields of medicine
(Freeman, 2001; Lewis et al., 2001).
Clinical research in emergency situations with-
out prospective informed or PC is ethically chal-
lenging. Severe TBI is without doubt an emergent
and life-threatening condition and existing therapy
is unsatisfactory. This should qualify severe TBI as
emergency exception to informed consent or de-
ferred consent for randomized clinical controlled
trials involving pharmacological agents with
promising therapeutic benefit facing short thera-
peutic time windows. Randomized placebo-con-
trolled investigations are necessary to determine
the safety and efficacy of new developed agents
under these circumstances. The requirement for
prior written PC causes significant delays in SDA.
With deferred consent or waiver of consent the
first dose of the experimental drug can be admin-
istered directly after completion of the first diag-
nostic CT scan, which is very close to the
experimental therapeutic time window. Randomi-
zed controlled phase III trials investigating the
safety and efficacy of agents with promising ben-
efit, conducted in acute emergency situations with
short therapeutic time windows, should allow ran-
domization under deferred consent or waiver of
consent. This is ethically defendable, off course

with proper safeguard from an independent safety
committee. Making progress in knowledge of
treatment in acute neurological and other inten-
sive care conditions is only possible if national
regulations and legislations allow waiver of con-
sent or deferred consent for clinical trials (Lemaire,
2005). As two of us have said before: ‘treat first,
ask later’ is ethically defendable in emergency and
intensive care medicine research (Kompanje and
Maas, 2004).
References
Annane, D., Outlin, H., Fisch, C. and Bellissant, E. (2004) The
effect of waiving consent on enrolment in a sepsis trial. Inten-
sive Care Med., 30: 632–637.
Blixen, C.E. and Agich, G.J. (2005) Stroke patients preferences
and values about emergency research. J. Med. Ethics, 31:
608–611.
Clifton, G.L., Knudson, P. and McDonald, M. (2002) Waiver
of consent in studies of acute brain injury. J. Neurotrauma,
19: 1121–1126.
Coppolino, M. and Ackerson, L. (2001) Do surrogate decision
makers provide accurate consent for intensive care research?
Chest, 119: 603–612.
CRASH trial management group. (2004) Research in emer-
gency situations: with or without relatives consent. Emerg.
Med. J., 21: 703.
Cuttini, M. (2000) Proxy informed consent in pediatric re-
search: a review. Early Hum. Dev., 60: 89–100.
European Union. (2001) Directive 2001/20/EC of the European
Parliament and the Council of 4 April 2001 on the approxi-
mation of the laws, regulations and administrative provi-
sions of the Member States relating to the implementation
of good clinical practice in the conduct of clinical trials on
medicinal products for human use. Off. J. Eur. Comm.,
L121/34.
Freeman, D.B. (2001) Safeguarding patients in clinical trials
with high mortality rates. Am. J. Crit. Care Med., 164:
190–192.
Harth, S.C. and Thong, Y.H. (1995) Parental perceptions and
attitudes about informed consent in clinical research involv-
ing children. Soc. Sci. Med., 40: 1573–1577.
Hoff, J.T. (1986) Cerebral protection. J. Neurosurg., 65:
579–591.
Hsieh, M., Dailey, M.W. and Callaway, C.W. (2001) Surrogate
consent by family members for out-of-hospital cardiac arrest
research. Acad. Emerg. Med., 8: 851–853.
Kompanje, E.J.O., Maas, A.I.R., Hilhorst, M.T., Slieker,
F.J.A. and Teasdale, G.M. (2005) Ethical considerations on
consent procedures for emergency research in severe and
moderate traumatic brain injury. Acta Neurochir. (Wien.),
147: 633–640.
249
Kompanje, E.J.O. and Maas, A.I.R. (2004) ‘Treat first, ask
later?’ Emergency research in acute neurology and neurotrau-
matology in the European Union. Intensive Care Med., 30:
168–169.
Kompanje, E.J.O. and Maas, A.I.R. (2006) Is the Glasgow
Coma Scale score protected health information? The effect of
new United States regulations (HIPAA) on completion of
screening logs in emergency research trials. Intensive Care
Med., 32: 313–314.
Lemaire, F. (2005) Waiving consent for emergency research.
Eur. J. Clin. Invest., 35: 287–289.
Lewis, R.J., Berry, D.A., Cryer, H., Fost, N., Krome, R.,
Washington, G.R., Houghton, J., Blue, J.W., Bechhofer, R.,
Cook, T. and Fisher, M. (2001) Monitoring a clinical trial
conducted under the Food and Drug Administration regu-
lations allowing a waiver of prospective informed consent:
the diaspirin cross-linked hemoglobin traumatic hemorrhagic
shock efficacy trial. Ann. Emerg. Med., 38: 397–404.
Liddell, K., Chamberlain, D., Menon, D.K., Bion, J., Kom-
panje, E.J.O., Lemaire, F., Druml, C., Vrhovac, B.,
Wiedermann, C.J. and Sterz, F. (2006a) The European Clin-
ical Trial Directive revisited: the VISEAR recommendations.
Resuscitation, 69: 9–14.
Liddell, K., Kompanje, E.J.O., Lemaire, F., Vrhovac, B., Menon,
D.K., Bion, J., Chamberlain, D., Wiedermann, C.J. and
Druml, C. (2006b) Recommendations in relation to the EU
Clinical Trials Directive and medical research involving inca-
pacitated adults. Wien. Klin. Wochenschr., 118: 183–191.
Luce, J.M. (2003) Is the concept of informed consent applicable
to clinical research involving critically ill patients? Crit. Care
Med., 31(Suppl): S153–S160.
Maas, A.I.R., Kompanje, E.J.O., Slieker, F.J.A. and Stocchetti,
N. (2005) Differences in completion of screening logs bet-
ween Europe and the United States in an Emergency phase
III trial resulting from HIPAA requirements. Ann. Surg.,
241(2): 382–383.
Maas, A.I.R., Marmarou, A., Murray, G.D. and Steyerberg,
E.W. (2004) Clinical trials in traumatic brain injury: current
problems and future solutions. Acta Neurochir., 89: 113–118.
Maas, A.I.R., Murray, G., Henney, H., Kassem, N., Legrand,
V., Mangelus, M., Muizelaar, J.P., Stocchetti, N., Knoller,
N. and on behalf of the Pharmos TBI Investigators. (2006)
Efficacy and safety of dexanabinol in sever traumatic brain
injury: results of a phase III randomized placebo controlled
clinical trial. Lancet Neurol., 5: 38–45.
Maas, A.I.R., Steyerberg, E.W. and Murray, G.D. (1999) Why
have recent trials of neuroprotective agents in head injury
failed to show convincing efficacy? A pragmatic analysis and
theoretical considerations. Neurosurgery, 44: 1286–1298.
Mason, S.A. and Allmark, P.J. (2000) Obtaining informed
consent to neonatal randomised controlled trials: interviews
with parents and clinicians in the Euricon study. Lancet, 356:
2045–2051.
McClure, K.B., Delorio, N.M., Gunnels, M.D., Ochsner, M.J.,
Biros, M.H. and Schmidt, T.A. (2003) Attitudes of emer-
gency department patients and visitors regarding emergency
exception from informed consent in resuscitation research,
250
community consultation and public notification. Acad. Em-
erg. Med., 10: 352–359.
McGarry, L.J., Thompson, D., Millham, F.H., Cowell, L.,
Snyder, P.J., Lenderking, W.R. and Weinstein, M.C. (2002)
Outcomes and costs of acute treatment of traumatic brain
injury. J. Trauma, 53: 1152–1159.
Roupie, E., Santin, A., Boulme, R., Wartel, J.S., Lepage, E.,
Lemaire, F., Lejonc, J.L. and Montagne, O. (2000) Patients
preferences concerning medical information and surrogacy:
results of a prospective study in a French emergency depart-
ment. Intensive Care Med., 26: 52–56.
Shohami, E. (1995) Long-term effect of HU-211, a novel
competitive NMDA antagonist, on motor and memory
functions after closed head injury in the rat. Brain Res.,
674: 55–62.
Silverman, H.J., Druml, C., Lemaire, F. and Nelson, R. (2004)
The European Union Directive and the protection of inca-
pacitated subjects in research: an ethical analysis. Intensive
Care Med., 30: 1723–1729.
Sugarman, J. (2000) Is the emperor really wearing new clothes?
Informed consent for acute coronary syndromes. Am. Heart
J., 140: 2–3.
Tu, J.V., Willison, D.J., Silver, F.L., Fang, J., Richards, J.A.,
Laupacis, A. and Kapral, M.K. (2004) Impracticability of
informed consent in the registry of the Canadian stroke net-
work. NEJM, 350: 1414–1421.
Williams, B.F., French, J.K. and White, H.D. (2003) Informed
consent during the clinical emergency of acute myocardial
infarction (HERO-2 consent substudy): a prospective obser-
vational study. Lancet, 361: 918–922.
 SECTION V

Emerging Topics in CNS Trauma

This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 18

Experimental models of repetitive brain injuries

John T. Weber

Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands

Abstract: Repetitive traumatic brain injury (TBI) occurs in a significant portion of trauma patients, es-
pecially in specific populations, such as child abuse victims or athletes involved in contact sports (e.g.
boxing, football, hockey, and soccer). A continually emerging hypothesis is that repeated mild injuries may
cause cumulative damage to the brain, resulting in long-term cognitive dysfunction. The growing attention
to this hypothesis is reflected in several recent experimental studies of repeated mild TBI in vivo. These
reports generally demonstrate cellular and cognitive dysfunction after repetitive injury using rodent TBI
models. In some cases, data suggests that the effects of a second mild TBI may be synergistic, rather than
additive. In addition, some studies have found increases in cellular markers associated with Alzheimer’s
disease after repeated mild injuries, which demonstrates a direct experimental link between repetitive TBI
and neurodegenerative disease. To complement the findings from humans and in vivo experimentation, my
laboratory group has investigated the effects of repeated trauma in cultured brain cells using a model of
stretch-induced mechanical injury in vitro. In these studies, hippocampal cells exhibited cumulative damage
when mild stretch injuries were repeated at either 1-h or 24-h intervals. Interestingly, the extent of damage
to the cells was dependent on the time between repeated injuries. Also, a very low level of stretch, which
produced no cell damage on its own, induced cell damage when it was repeated several times at a short
interval (every 2 min). Although direct comparisons to the clinical situation are difficult, these types of
repetitive, low-level, mechanical stresses may be similar to the insults received by certain athletes, such as
boxers, or hockey and soccer players. This type of in vitro model could provide a reliable system in which to
study the mechanisms underlying cellular dysfunction following repeated injuries. As this area of TBI
research continues to evolve, it will be imperative that models of repetitive injury replicate injuries in
humans as closely as possible. For example, it will be important to model appropriately concussive episodes
versus even lower level injuries (such as those that might occur during boxing matches). Suitable inter-
injury intervals will also be important parameters to incorporate into models. Additionally, it will be crucial
to design and utilize proper controls, which can be more challenging than experimental approaches to single
mild TBI. It will also be essential to combine, and compare, data derived from in vitro experiments with
those conducted with animals in vivo. These issues, as well as a summary of findings from repeated TBI
research, are discussed in this review.

Keywords: hippocampus; in vivo models; in vitro; mild injury; repeated injury; TBI; trauma

Corresponding author. Present address: School of Pharmacy, Introduction

Memorial University of Newfoundland, Health Sciences
Centre, St. John’s, NL, Canada. Tel.: +1-709-777-7022; Repetitive injuries occur in a considerable portion
Fax: +1-709-777-7044; E-mail: jweber@mun.ca of individuals who experience a traumatic brain

DOI: 10.1016/S0079-6123(06)61018-2 253

254
injury (TBI). Child abuse victims, and in particu-
lar, victims of shaken baby syndrome, are often
subjected to multiple injuries to the head (Shannon
et al., 1998). Spousal abuse also contributes to the
number of repeated brain injuries (Roberts et al.,
1990). However, many injuries of these types go
unreported, and it is difficult to assess how many
insults a patient may have suffered. Arguably,
athletes represent the largest group of patients that
are at risk for experiencing repeated brain injuries
(Kelly and Rosenberg, 1997; Kelly, 1999; Powell
and Barber-Foss, 1999). In fact, sports and recre-
ational activities contribute to approximately 10%
of all mild injuries, and these individuals are at
high risk of experiencing a second TBI (Woo and
Thoidis, 2000). Also, in comparison to child or
spousal abuse victims, there is generally better
documentation of how many brain injuries an in-
dividual has sustained due to recreational or sports
related activities, making this population easier to
study.
The idea that multiple head injuries in athletes
could lead to clinical problems has long been sug-
gested. For example, the development of dementia
pugilistica in professional boxers is believed to be
caused by the multiple hits to the head that boxers
endure over the course of their career (Jordan,
2000). Also, recent evidence has shown that the
number of concussions is inversely related to per-
formance on several neuropsychological tests in
soccer players (Matser et al., 1999, 2001) and
jockeys who have experienced multiple concuss-
ions generally display more cognitive dysfunctions
than those who have had a single injury (Wall
et al., 2006).
When studying repetitive brain trauma in ath-
letes, we can gain much information about the
pathology and progress of such injuries from the
injured athletes themselves, for example by meas-
uring changes in cognitive and motor perform-
ance. However, these injuries are generally at a
mild level, and therefore, except in rare cases when
athletes die as a result of the insult, we cannot
assess the changes that have occurred in the brain
at the cellular and sub-cellular levels. In order to
compile this type of information, we must turn to
experimental models of TBI.
Experimental studies of repeated mild TBI in vivo
It should be made clear that when discussing ex-
perimental studies of repetitive TBI in vivo, this
does not include studies of secondary insults, such
as a mechanical insult to the head followed by a
duration of ischemia or glutamate exposure. Re-
peated TBI experimentation can be defined as an
initial mechanical injury to the head followed by
another mechanical insult to the head of the same
or different degree. Based on these criteria, little
attention was given to these types of experiments
before the year 2000, with only a few repetitive
injury studies being published (Olsson et al., 1971;
Weitbrecht and Noetzel, 1976; Kanayama et al.,
1996). Several additional in vivo studies of repeated
injuries in rodents have now been conducted in the
past decade (Allen et al., 2000; Laurer et al., 2001;
DeFord et al., 2002; Uryu et al., 2002; Conte et al.,
2004; Creeley et al., 2004; Raghupathi et al., 2004;
Dobrowolska and Gibson, 2005; Gibson and
Schalles, 2005; Longhi et al., 2005; Yoshiyama et
al., 2005). All of these repeated mild injury studies
were conducted using rodent models of TBI with
the exception of the study by Raghupathi et al.
(2004), which reported a pediatric model of re-
peated injury conducted in pigs.
Repetitive TBI generally occurs at a mild level,
therefore experimental models have been used
which are minimally invasive and do not require a
craniotomy, such as weight drop models. The
models must also be administered at a level that
produces minimal, or preferably, no fatality. Indi-
viduals who have suffered from a mild TBI often
complain of cognitive difficulties post-injury.
Therefore, repeated injury studies usually evalu-
ate cognitive function, for example using the
Morris water maze (MWM) test, as well as the
extent of cell abnormalities in the cortex and hip-
pocampus. The hippocampus in particular has re-
ceived significant attention in the study of repeated
mild TBI, because it plays a critical role in certain
types of learning and aspects of memory storage.
Experimental and clinical data have demonstrated
not only the importance of this brain region in
learning and memory, but also that the hippo-
campus is uniquely vulnerable to injury, even after

mild brain trauma (Lyeth et al., 1990; Lowenstein
et al., 1992). In a study by DeFord et al. (2002),
repeated mild injuries were administered to mice
(four times every 24 h), followed by MWM testing
and histological analysis. Significant learning defi-
cits were found after repeated injuries, which were
not evident after a single injury. These deficits oc-
curred even in absence of cell death within the
cortex and hippocampus. These studies were ex-
tended by Gibson and Schalles (2005) using the
same injury model; they demonstrated impaired
function in the Barnes circular maze after multiple,
but not single injuries. Cognitive deficits after
multiple mild TBIs (using MWM analysis) have
also been demonstrated in a similar study using a
weight drop model (Creeley et al., 2004).
In order to most closely mimic the type of insult
that athletes may receive, Laurer et al. (2001) de-
veloped an injury regimen that they described as
‘‘concussive’’; this model was used for subsequent
studies as well (Uryu et al., 2002; Conte et al., 2004;
Longhi et al., 2005). In an assessment of cognitive
and motor function after repeated injury in mice,
Laurer et al. (2001) found that the brain was more
vulnerable to a second insult if the second injury
occurred 24 h after the first. Even though no cog-
nitive deficits were demonstrated in mice receiving
repeated injuries, there was a decrease in motor
function and some neuronal loss. The authors also
stated that the effects of a second mild TBI may be
synergistic, rather than additive. To analyze further
the effects of lengthening the inter-injury interval,
Longhi et al. (2005) investigated repetitive injuries
three, five, and seven days apart. Animals that re-
ceived repeated injuries three or five days apart ex-
hibited cognitive dysfunction not evident in sham
animals or those injured only once. However, no
deficits were observed when the injury interval was
extended to seven days. This experimental evidence
— that the brain can recover from a first injury,
given a sufficient amount of time — is appealing,
especially in relation to establishing ‘‘return-to-
play’’ guidelines for athletes. Overall, the evidence
from in vivo experimental models suggests that re-
petitive mild TBI causes more cognitive and cellular
dysfunction than a single injury, if the brain is not
given a sufficient amount of time to recover.
255
Repetitive injury and neurodegenerative disease
It has been noted for some time that there is a
correlation between the occurrence of TBI and
the further development of neurodegenerative dis-
ease later in life. In fact, TBI is considered to be
one of the most robust risk factors for developing
Alzheimer’s disease (AD) (Szczygielski et al.,
2005). There is also evidence that genetic predis-
position may increase one’s risk of developing
AD, such as possession of the apolipoprotein E e4
allele. A phenomenon known as chronic TBI oc-
curs in a significant amount of professional boxers
(Jordan, 2000), with the most serious form,
dementia pugilistica, resulting in severe motor
and cognitive dysfunctions. It is now known that
the pathology of AD and dementia pugilistica are
quite similar (Geddes et al., 1999; Schmidt et al.,
2001). Although the epidemiological data are
quite strong, very little has been done to make
the mechanistic link between repeated mild TBI
and the development of either AD or dementia
pugilistica.
A few studies have found increases in cellular
markers associated with AD after repeated mild
injuries (Kanayama et al., 1996; Uryu et al., 2002;
Conte et al., 2004; Dobrowolska and Gibson,
2005). For example, Kanayama et al. (1996)
showed an increase in tau immunoreactivity in
neurons and Dobrowolska and Gibson (2005) re-
ported increased amyloid precursor protein fol-
lowing multiple injuries. Uryu et al. (2002) and
Conte et al. (2004) used a transgenic mouse model
of AD-like amyloidosis, and demonstrated that
amyloid levels and deposits increased in the brain
after repeated injuries, but not after a single insult.
Yoshiyama et al. (2005) utilized the same trans-
genic mouse model and analyzed the effects of four
injuries given per day, once a week, for a period of
four weeks, in order to simulate dementia pugi-
listica. After nine months, only one mouse showed
pathology consistent with this syndrome. Al-
though the findings were not robust, this is per-
haps the only study to attempt to model dementia
pugilistica, and more studies are warranted.
The overall findings of these studies are quite
important, because they can demonstrate a direct
256
experimental link between repeated mild TBI and
the development of AD-like pathology. Generally,
it takes years before the onset of symptoms of ne-
urodegenerative disorders after an individual has
experienced a TBI. Therefore, it requires an ex-
tremely long amount of time to gather this type of
epidemiological data from the human population.
This area of research, in particular, is where ex-
perimental models could truly help decipher the
mechanisms by which neurodegenerative disease
may be triggered by repetitive brain injury.
Studying repeated injuries in vitro
Several in vitro approaches have now been devel-
oped to study traumatic injury, which utilize cells
grown in culture (Morrison et al., 1998; Weber,
2004; LaPlaca et al., 2005; also see Chapter 3 by
Spaethling et al., in this volume). My laboratory
group utilizes a model of stretch-induced mechan-
ical injury in vitro originally developed by Ellis
et al. (1995). We have characterized this stretch
injury model in cell cultures composed of neurons
and glia from murine hippocampus (Slemmer
et al., 2002; Slemmer and Weber, 2005), cortex
(Engel et al., 2005), and cerebellum (Slemmer
et al., 2004).
To complement the findings from humans and
from in vivo experimentation, we have recently
conducted a series of studies investigating the
effects of repeated trauma on hippocampal cells
(Slemmer et al., 2002; Slemmer and Weber, 2005).
In these studies, we utilized a mild level of stretch
injury that produces some measurable damage to
cells when administered a single time. When mild
stretch injuries were repeated at either 1-h or 24-h
intervals, cells exhibited cumulative damage. For
example, cultures that received a second insult
displayed a significant loss of neurons not evident
in cultures that received only one injury. Addi-
tionally, cultures injured twice released a signifi-
cant level of neuron specific enolase (NSE), which
was not observed in cultures injured a single time.
Interestingly, the extent of damage to the cells was
dependent on the time between repeated injuries.
For example, cultures that received a second insult
1 h after the first injury released more S-100B
protein (a biomarker of injury commonly em-
ployed in the clinic) than cultures that received a
second injury at 24 h. Cultures injured 24 h apart
also exhibited less staining with the intravital dye,
propidium iodide, than those injured 1 h apart. As
shown in some in vivo studies, these findings sug-
gest that a level of injury, which produces meas-
urable damage or dysfunction on its own, may
cause cumulative damage if repeated within
a certain time frame (Laurer et al., 2001; Longhi
et al., 2005).
We also investigated the effects of a very low
level of stretch, which produced no overt cell dam-
age (Slemmer and Weber, 2005). This ‘‘subthresh-
old’’ level of stretch did not cause significant
damage or death, even when it was repeated at a
1-h interval. However, this low level of stretch did
induce cell damage when it was repeated several
times at a short interval (every 2 min), indicated by
increased propidium iodide staining, neuronal
loss, and an increase in NSE release. Although
direct comparisons to the clinical situation are
difficult, these types of repetitive, low-level, me-
chanical stresses may be similar to the insults re-
ceived by certain athletes, such as boxers and
hockey and soccer players (Matser et al., 1998,
1999; Jordan, 2000; Webbe and Ochs, 2003;
Wennberg and Tator, 2003). This type of in vitro
model could comprise a reliable system in which to
study the mechanisms underlying cellular dysfunc-
tion following repeated injuries. In addition, this
approach could provide a means for relatively
rapid screening of potential therapeutic strategies
for both single and repeated mild TBI.
The phenomenon of preconditioning
Several studies have indicated that an initial, very
mild insult to either cultured cells or to the brain
itself, may provide some protection from a second,
more severe insult, a finding that has been termed
‘‘preconditioning’’. Ischemic preconditioning, in
which a brief exposure to ischemia renders the
brain more resistant to subsequent longer periods
of ischemia, has been well described (for review,
see Schaller and Graf, 2002). There is also evidence
of preconditioning cross-tolerance. For example,

brief ischemia lessens damage following TBI in
vivo (Pérez-Pinzón et al., 1999). Another interest-
ing phenomenon is that heat acclimation (chronic
exposure to moderate heat) can also provide re-
sistance to TBI (see Chapter 25 by Shein et al., in
this volume).
In our in vitro studies, we observed a novel form
of mechanical preconditioning. When hippocam-
pal cultures were administered a subthreshold level
of stretch 24 h prior to a mild stretch, there was a
significant decrease in released S-100B protein
compared to cultures that were injured at a mild
level alone (Slemmer and Weber, 2005). This ob-
servation suggests some form of protection initi-
ated by this low level of stretch. A similar finding
in vivo was reported by Allen et al. (2000). In their
study, rats received a series of mild injuries spaced
three days apart using a weight drop model. Some
of these animals received a severe injury after the
repetitive mild injuries. Motor function deficits
were evident in severely injured animals, but not in
animals that received repeated mild injuries or re-
peated mild injuries followed by a severe injury.
This last observation suggests a preconditioning
effect.
The most important question now is how do we
utilize this information? One can imagine the ethi-
cal implications of suggesting to people that a mild
insult to their brains may in fact protect them from
worse insults in the future. We still have much to
learn about preconditioning. For example, what is
the threshold for mechanical insults between ini-
tiating protective versus damaging mechanisms in
the brain? A clearer understanding of the mecha-
nisms by which this protection is elicited holds
potential for the management of mild TBI. The
fact that different stressors can protect the brain
from TBI (i.e. cross-tolerance) suggests that the
same or similar mechanisms are responsible for the
endogenous protection. Increasing the expression
of these protective systems could not only be a
reliable way for managing mild TBI, but also
could provide resistance in individuals who may be
at risk of sustaining an additional head injury,
such as athletes. Both in vivo and in vitro models
could provide reliable systems in which to study
the mechanisms underlying the preconditioning
phenomenon.
257
Future directions
Experimental evidence suggests that animals can
demonstrate cognitive deficits and cellular dys-
function after repetitive mild TBI, even though
the injury may not necessarily lead to cell death
(Kanayama et al., 1996; DeFord et al., 2002).
Therefore, rather than trying to prevent cells from
dying after repeated injuries, it is probably more
important to learn how to restore normal cellular
physiology after a traumatic episode. Combining
studies at the cellular and behavioral levels is es-
sential, and one particular area of interest is the
evaluation of the effects of repeated TBI on synap-
tic plasticity in the hippocampus. This ability of
neurons to undergo changes in synaptic strength,
such as long-term potentiation (LTP), is postu-
lated to be a cellular correlate of learning and
memory (Bliss and Collingridge, 1993; Malenka
and Nicoll, 1999). Several studies have reported
impaired hippocampal LTP after TBI in vivo
(Albensi, 2001; Weber, 2004). One area of future
research could focus on restoring mechanisms of
synaptic plasticity after injury (such as LTP), as
well as correlated hippocampal-mediated behavi-
oral tasks.
The hippocampus shares neuronal projections
with areas of the cerebral cortex, which undoubt-
edly also contributes to memory formation and
storage. Indeed, alterations in synaptic plasticity
may also occur directly in the cortex after repeated
mild TBI. Therefore, although the hippocampus
may play a central role in the cognitive dysfunc-
tion observed after mild TBI, it is important not to
overlook contributions from other brain areas as
well. Along these lines, since some repeated injury
studies demonstrate motor dysfunction, it may
also be appropriate to investigate cellular physiol-
ogy and synaptic plasticity in the cerebellum
(Hansel et al., 2001; Weber et al., 2003; Slemmer
et al., 2005) after repetitive TBI.
The utilization of proper parameters for re-
peated injury studies may be even more crucial
than deciding on appropriate research directions.
For example, what are the best inter-injury inter-
val, or intervals, to use? Although 24 h between
injuries is the most common (and perhaps practi-
cal) interval in the laboratory (Weitbrecht and
258

Noetzel, 1976; Laurer et al., 2001; DeFord et al.,
2002; Uryu et al., 2002; Conte et al., 2004; Creeley
et al., 2004; Dobrowolska and Gibson, 2005;
Gibson and Schalles, 2005; Yoshiyama et al.,
2005), is it the most appropriate in mimicking
what occurs in humans? Also, how many injuries
should a researcher administer? If one is attempt-
ing to model concussive episodes, then two, three,
or four may be enough, as this may closely mimic a
true situation, especially with athletes. However,
when attempting to recreate dementia pugilistica
(Yoshiyama et al., 2005), the number of injuries
should certainly be increased.
What are the proper controls and endpoints to
use for repeated injury studies? For in vivo studies
analyzing the effects of a single TBI, the issue of
controls is fairly straightforward. Sham animals
are treated at an equivalent time as injured ani-
mals, and the analysis, cellular or behavioral, is
also performed at the same timepoint. However,
when comparing uninjured animals to animals
that have received more than one injury, what is

Fig. 1. Issues for consideration when designing repeated TBI experiments (i.e. choosing proper timepoints for controls and behavioral/
tissue analysis). (A) Hypothetical timeline for two injuries 24 h apart. (B) Timeline for greater than two injuries. T ¼ time.

the proper comparison? For example, if an animal
receives an injury on day one, an additional injury
on day two, and analysis takes place on day three,
does one compare the data with sham animals
from day one, or from day two (or both, see
Fig. 1A). The issue is further complicated when
comparing repeatedly injured animals to animals
that have received a single TBI. If the comparison
concerns animals that undergo four injuries or a
single injury, are the single insult animals injured
at the same time as injury one in the repeated
group, or at the same time as the fourth injury (see
Fig. 1B)? This decision will also affect the endpoint
as well. For example, if animals or tissue are
analyzed one day after the fourth injury, then four
days will have passed for the single injury group if
those animals were injured on day one. This
difference in time could affect the observations.
We have struggled with these issues when design-
ing our in vitro experiments. One could argue that
if a long enough period of time passes after the
injuries, such as weeks or months, then the effect
of when the single insult animals were injured will
be negligible. Admittedly, this would be more
proper for comparison to the human situation in
which the effects of mild TBI can be manifested for
weeks, months, or even years. However, this is of-
ten not practical for many laboratories, as the
costs of housing animals for months can at times
be prohibitive. Also, conducting long-term experi-
ments in vitro is limited, since the cells generally
remain viable for only a few weeks. This raises a
critical point as to the relevance of repeated injury
studies in vitro. I strongly believe that in vitro ex-
periments can deliver information about the cel-
lular mechanisms of repeated injury that are
difficult to obtain in vivo, and that it is essential
to combine data derived from in vitro experiments
with those conducted with animals in vivo. How-
ever, I am unsure how to directly compare the
data. For example, is a 24 h injury interval in vitro
equivalent to 24 h in vivo? The more consensus
that exists on these issues with individuals who
conduct repeated injury studies, the easier it will be
to compare the data, and the stronger a case can
be made for showing unequivocally, that repeated
mild TBI could lead to long-term dysfunction in
humans.
259
Abbreviations
AD Alzheimer’s disease
LTP long-term potentiation
MWM Morris water maze
NSE neuron specific enolase
TBI traumatic brain injury
Acknowledgments
The in vitro work discussed in this manuscript was
supported by grants from the Research Council
for Earth and Life Sciences (ALW) and Research
Council for Medical Sciences (MW) with financial
aid from the Netherlands Organization for Scien-
tific Research (NWO), a Breedtestrategie subsidie
from Erasmus Medical Centre, and funding from
Hersenstichting Nederland. I would also like to
thank Jennifer E. Slemmer for editorial assistance.
References
Albensi, B.C. (2001) Models of brain injury and alterations in
synaptic plasticity. J. Neurosci. Res., 65: 279–283.
Allen, G.V., Gerami, D. and Esser, M.J. (2000) Conditioning
effects of repetitive mild neurotrauma on motor function in
an animal model of focal brain injury. Neuroscience, 99:
93–105.
Bliss, T.V. and Collingridge, G.L. (1993) A synaptic model of
memory: long-term potentiation in the hippocampus. Nature,
361: 31–39.
Conte, V., Uryu, K., Fujimoto, S., Yao, Y., Rokach, J., Long-
hi, L., Trojanowski, J.Q., Lee, V.M.-Y., McIntosh, T.K. and
Pratico, D. (2004) Vitamin E reduces amyloidosis and im-
proves cognitive function in Tg2576 mice following repetitive
concussive brain injury. J. Neurochem., 90: 758–764.
Creeley, C.E., Wozniak, D.F., Bayly, P.V., Olney, J.W. and
Lewis, L.M. (2004) Multiple episodes of mild traumatic brain
injury result in impaired cognitive performance in mice.
Acad. Emerg. Med., 11: 809–819.
DeFord, S.M., Wilson, M.S., Rice, A.C., Clausen, T., Rice,
L.K., Barabnova, A., Bullock, R. and Hamm, R.J. (2002)
Repeated mild brain injuries result in cognitive impairment in
B6C3F1 mice. J. Neurotrauma, 19: 427–438.
Dobrowolska, J.A. and Gibson, C.J. (2005) Immunohisto-
chemical analysis of amyloid precursor protein in the hippo-
campus and cortex of mice following repeated mild traumatic
brain injuries. Soc. Neurosci. Abst., 31: Program Number
334.7.
260
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-
induced injury of cells in culture: characterization of the
model using astrocytes. J. Neurotrauma, 12: 325–339.
Engel, D.C., Slemmer, J.E., Vlug, A.S., Maas, A.I. and Weber,
J.T. (2005) Combined effects of mechanical and ischemic in-
jury to cortical cells: secondary ischemia increases damage
and decreases effects of neuroprotective agents. Neurophar-
macology, 49: 985–995.
Geddes, J.F., Vowles, G.H., Nicoll, J.A. and Revesz, T. (1999)
Neuronal cytoskeletal changes are an early consequence of
repetitive head injury. Acta Neuropathol. (Berl.), 98:
171–178.
Gibson, C.J. and Schalles, M.D. (2005) Barnes circular maze
deficits in repeated mildly brain injured mice. Soc. Neurosci.
Abst., 31: Program Number 334.6.
Hansel, C., Linden, D.J. and D’Angelo, E. (2001) Beyond par-
allel fiber LTD: the diversity of synaptic and non-synaptic
plasticity in the cerebellum. Nat. Neurosci., 4: 467–475.
Jordan, B.D. (2000) Chronic traumatic brain injury associated
with boxing. Semin. Neurol., 20: 179–185.
Kanayama, G., Takeda, M., Niigawa, H., Ikura, Y., Tamii, H.,
Taniguchi, N., Kudo, T., Miyamae, Y., Morihara, T. and
Nishimura, T. (1996) The effects of repetitive mild brain in-
jury on cytoskeletal protein and behavior. Methods Find.
Exp. Clin. Pharmacol., 18: 105–115.
Kelly, J.P. (1999) Traumatic brain injury and concussion in
sports. JAMA, 282: 989–991.
Kelly, J.P. and Rosenberg, J.H. (1997) Diagnosis and manage-
ment of concussion in sports. Neurology, 48: 575–580.
LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II,
R.S. (2005) High rate shear strain of three-dimensional neu-
ral cell cultures: a new in vitro traumatic brain injury model.
J. Biomech., 38: 1093–1105.
Laurer, H.L., Bareyre, F.M., Lee, V.M., Trojanowski, J.Q.,
Longhi, L., Hoover, R., Saatman, K.E., Raghupathi, R.,
Hoshino, S., Grady, S.M. and McIntosh, T.K. (2001) Mild
head injury increasing the brain’s vulnerability to a second
concussive impact. J. Neurosurg., 95: 859–870.
Longhi, L., Saatman, K.E., Fujimoto, S., Raghupathi, R.,
Meaney, D.F., Davis, J., McMillan, B.S.A., Conte, V.,
Laurer, H.L., Stein, S., Stocchetti, N. and McIntosh, T.K.
(2005) Temporal window of vulnerability to repetitive
experimental concussive brain injury. Neurosurgery, 56:
364–374.
Lowenstein, D.H., Thomas, M.J., Smith, D.H. and McIntosh,
T.K. (1992) Selective vulnerability of dentate hilar neurons
following traumatic brain injury: a potential mechanistic link
between head trauma and disorders of the hippocampus.
J. Neurosci., 12: 4846–4853.
Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips,
L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990)
Prolonged memory impairment in the absence of hippocam-
pal cell death following traumatic brain injury in the rat.
Brain Res., 526: 249–258.
Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation
— a decade of progress? Science, 285: 1870–1874.
Matser, E.J.T., Kessels, A.G.H., Lezak, M.D., Jordan, B.D.
and Troost, J. (1999) Neuropsychological impairment in am-
ateur soccer players. JAMA, 282: 971–973.
Matser, J.T., Kessels, A.G.H., Jordan, B.D., Lezak, M.D. and
Troost, J. (1998) Chronic traumatic brain injury in profes-
sional soccer players. Neurology, 51: 791–796.
Matser, J.T., Kessels, A.G.H., Lezak, M.D. and Troost, J.
(2001) A dose-response relation of headers and concussions
with cognitive impairment in professional soccer players.
J. Clin. Exp. Neuropsychol., 23: 770–774.
Morrison III, B., Saatman, K.E., Meaney, D.F. and McIntosh,
T.K. (1998) In vitro central nervous system models of
mechanically induced trauma: a review. J. Neurotrauma, 15:
911–928.
Olsson, Y., Rinder, L., Lindgren, S. and Stalhammar, D. (1971)
Studies on vascular permeability changes in experimental
brain concussion. 3. A comparison between the effects of
single and repeated sudden mechanical loading of the brain.
Acta Neuropathol. (Berl.), 19: 225–233.
Pérez-Pinzón, M.A., Alonso, O., Kraydieh, S. and Dietrich,
W.D. (1999) Induction of tolerance against traumatic
brain injury by ischemic preconditioning. Neuroreport, 10:
2951–2954.
Powell, J.W. and Barber-Foss, K.D. (1999) Traumatic brain
injury in high school athletes. JAMA, 282: 958–963.
Raghupathi, R., Mehr, M.F., Helfaer, M.A. and Margulies,
S.S. (2004) Traumatic axonal injury is exacerbated following
repetitive closed head injury in the neonatal pig. J. Neuro-
trauma, 21: 307–316.
Roberts, G.W., Whitwell, H.L., Acland, P.R. and Bruton, C.J.
(1990) Dementia in a punch-drunk wife. Lancet, 335:
918–919.
Schaller, B. and Graf, R. (2002) Cerebral ischemic precondi-
tioning. An experimental phenomenon or a clinical impor-
tant entity of stroke prevention? J. Neurol., 249: 1503–1511.
Schmidt, M.L., Zhukareva, V., Newell, K.L., Lee, V.M.-Y. and
Trojanowski, J.Q. (2001) Tau isoform profile and phosphoryl-
ation state in dementia pugilistica recapitulate Alzheimer’s
disease. Acta Neuropathol. (Berl.), 101: 518–524.
Shannon, P., Smith, C.R., Deck, J., Ang, L.C., Ho, M. and
Becker, L. (1998) Axonal injury and the neuropathology of
shaken baby syndrome. Acta Neuropathol. (Berl.), 95:
625–631.
Slemmer, J.E., De Zeeuw, C.I. and Weber, J.T. (2005) Don’t get
too excited: mechanisms of glutamate-mediated Purkinje cell
death. Prog. Brain Res., 148: 367–390.
Slemmer, J.E., Matser, E.J.T., De Zeeuw, C.I. and Weber, J.T.
(2002) Repeated mild injury causes cumulative damage to
hippocampal cells. Brain, 125: 2699–2709.
Slemmer, J.E. and Weber, J.T. (2005) The extent of damage
following repeated injury to hippocampal cells is dependent
on the severity of insult and inter-injury interval. Neurobiol.
Dis., 18: 421–431.
Slemmer, J.E., Weber, J.T. and De Zeeuw, C.I. (2004) Cell
death, glial protein alterations and elevated S-100b protein in
cerebellar cell cultures following mechanically induced
trauma. Neurobiol. Dis., 15: 563–572.

Szczygielski, J., Mautes, A., Steudel, W.I., Falkai, P., Bayer,
T.A. and Wirths, O. (2005) Traumatic brain injury: cause or
risk of Alzheimer’s disease? A review of experimental studies.
J. Neural Transm., 112: 1547–1564.
Uryu, K., Laurer, H., McIntosh, T., Pratico, D., Martinez, D.,
Leight, S., Lee, V.M.-Y. and Trojanowski, J.Q. (2002) Re-
petitive mild brain trauma accelerates Ab deposition, lipid
peroxidation, and cognitive impairment in a transgenic mouse
model of Alzheimer amyloidosis. J. Neurosci., 22: 446–454.
Wall, S.E., Williams, W.H., Cartwright-Hatton, S., Kelly, T.P.,
Murray, J., Murray, M., Owen, A. and Turner, M. (2006)
Neuropsychological dysfunction following repeat concussions
in jockeys. J. Neurol. Neurosurg. Psychiatry, 77: 518–520.
Webbe, F.M. and Ochs, S.R. (2003) Recency and frequency of
soccer heading interact to decrease neurocognitive perform-
ance. App. Neurol., 10: 31–41.
Weber, J.T. (2004) Calcium homeostasis following traumatic
neuronal injury. Curr. Neurovasc. Res., 1: 151–171.
Weber, J.T., De Zeeuw, C.I., Linden, D.J. and Hansel, C.
(2003) Long-term depression of climbing fiber-evoked
261
calcium transients in Purkinje cell dendrites. Proc. Natl.
Acad. Sci. U.S.A., 100: 2878–2883.
Weitbrecht, W.U. and Noetzel, H. (1976) Autoradiographic
investigations in repeated experimental brain concussion
(author’s transl.) [Article in German]. Arch. Psychiatr.
Nervenkr., 223: 59–68.
Wennberg, R.A. and Tator, C.H. (2003) National Hockey
League reported concussions, 1986–87 and 2001–02. Can. J.
Neurol. Sci., 30: 206–209.
Woo, B.H. and Thoidis, G. (2000) Epidemiology of traumatic
brain injury. In: Woo B.H. and Nesathurai S. (Eds.), The
Rehabilitation of People with Traumatic Brain Injury. Black-
well Science, Malden, pp. 13–17.
Yoshiyama, Y., Uryu, K., Higuchi, M., Longhi, L., Hoover, R.,
Fujimoto, S., McIntosh, T., Lee, V.M.-Y. and Trojanowski,
J.Q. (2005) Enhanced neurofibrillary tangle formation, cer-
ebral atrophy, and cognitive deficits induced by repetitive
mild brain injury in a transgenic tauopathy mouse model.
J. Neurotrauma, 22: 1134–1141.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 19

Minor traumatic brain injury in sports: a review in

order to prevent neurological sequelae

Nicola Biasca1, and William L. Maxwell2

1
Clinic of Orthopaedic, Sports Medicine and Traumatology, Department of Surgery, Spital Oberengadin, CH-7503
Samedan/St. Moritz, Switzerland
2
Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK

Abstract: Minor traumatic brain injury (mTBI) is caused by inertial effects, which induce sudden rotation
and acceleration forces to and within the brain. At less severe levels of injury, for example in mTBI, there is
probably only transient disturbance of ionic homeostasis with short-term, temporary disturbance of brain
function. With increased levels of severity, however, studies in animal models of TBI and in humans have
demonstrated focal intra-axonal alterations within the subaxolemmal, neurofilament and microtubular
cytoskeletal network together with impairment of axoplasmic transport. These changes have, until very
recently, been thought to lead to progressive axonal swelling, axonal detachment or even cell death over a
period of hours or days, the so-called process of ‘‘secondary axotomy’’. However, recent evidence has
suggested that there may be two discrete pathologies that may develop in injured nerve fibers. In the TBI
scenario, disturbances of ionic homeostasis, acute metabolic changes and alterations in cerebral blood flow
compromise the ability of neurons to function and render cells of the brain increasingly vulnerable to the
development of pathology. In ice hockey, current return-to-play guidelines do not take into account these
new findings appropriately, for example allow returning to play in the same game. It has recently been
hypothesized that the processes summarized above may predispose brain cells to assume a vulnerable state
for an unknown period after mild injury (mTBI). Therefore, we recommend that any confused player with
or without amnesia should be taken off the ice and not be permitted to play again for at least 72 h.

Keywords: minor traumatic brain injury, mTBI; neuromechanics; neuropathology; neurobiology and
associated neurometabolic changes of mTBI; process of delayed axotomy; vulnerability of cells in the brain;
return to play guidelines; ice hockey

Introduction nervous system. This is especially true in ice

Ice hockey is one of the speediest, competitive
contact team sports involving both rapid changes
of direction of the whole body and controlled ag-
gression. Athletes are therefore exposed to an in-
creased risk of a high-velocity collision that may
result in a sports-related injury to the central
Corresponding author. Tel.: +41-81-851-85-15; Fax: +41-81-
851-85-16; E-mail: Biasca@medicmotion.com
hockey, which has many inherent features that
may predispose players to injury, including high
acceleration–deceleration, rapid changes of direc-
tion, shooting, body checking and a low friction
ice surface. Before the introduction of full-face
masks and helmets, head injuries, including inju-
ries to the face, scalp and brain, accounted for at
least 50% of all serious injuries in ice hockey
(Pashby, 1993). Indeed, before the use of any
type of helmet and before the mandatory use of

DOI: 10.1016/S0079-6123(06)61019-4 263

264
standardized helmets, in the early 1960s, fatalities
secondary to head injuries have been reported in
Sweden, Canada and the United States (Pashby,
1993). Since 1963, however, there have been no
fatalities from head injury reported in Sweden
(Odelgard, 1989).
However, an alarming and persistently high rate
of minor traumatic brain injuries or mTBI, often
termed cerebral concussion worldwide over the
last decade and a half has been reported (Clayton,
1997; Dick, 1997; Biasca, 2000; Tegner, 2000;
www.hockeyinjuries). Although the majority of
cases of traumatic brain injury (TBI) do occur af-
ter Motor Vehicle Accidents and falls, it is notable
that sport activities are responsible for approxi-
mately 15–25% of all cases (Gennarelli, 1991;
McGehee, 1996). Recent evidence, obtained over
the last 15 years, has suggested that mTBI may be
more common and more serious than previously
thought worldwide (Denny and Russel, 1941;
Gennarelli, 1991; Castaldi, 1993; Tegner and Lore-
ntzon, 1996; Clayton, 1997; Dick, 1997; Gennarelli
et al., 1998; Biasca et al., 1999; Roberts et al.,
1999; Biasca, 2000; Ranalli, 2000; Tegner, 2000,
2001; www.hockeyinjuries; Burke, 2001; Laprade,
2001; Pashby et al., 2001). According to the Center
for Disease Control in the United States, there are
more than 300,000 sports-related mTBI each year
and certainly this number underestimates the true
level of incidence, because many of these injuries
are not reported (MMWR, 1997).
There has long been controversy concerning the
structural basis of mTBIs, like concussion. How-
ever, the importance of acceleration forces was al-
ready established by the early 1940s (Denny and
Russel, 1941; Holbourn, 1943; Gennarelli, 1991;
Elson and Ward, 1994). Historically, many inves-
tigators of the early 1990s were of the belief that
mTBI of the brain could occur without evidence of
cellular or vascular lesions of the cerebrum, for
example a number of studies have been cited by
Denny and Russel (1941), Gennarelli (1991) and
Elson and Ward (1994). Further, areas of pet-
echiae, in the absence of other lesions, have been
reported after fatal brain injury (Elson and Ward,
1994).
Holbourn (1943) stated that the relative motion
between brain constituents is insignificant when
the brain is exposed to a linear acceleration, but
not so in angular acceleration. This set the stage
for further inquiry into the effects of acceleration
versus impact trauma to the head (Denny and
Russel, 1941; Holbourn, 1943; Gennarelli, 1991;
Elson and Ward, 1994). It is now generally ac-
cepted that in patients who experience blunt head
injury, nerve fibers/axons are particularly suscep-
tible to damage. It is also now well recognized that
diffuse axonal injury (DAI) is a consistent feature
of severe human TBI, particularly those injuries
involving rapid acceleration/deceleration of the
brain (Denny and Russel, 1941; Gennarelli, 1991;
Gennarelli et al., 1998; Kelly, 1999; Graham et al.,
2000). Recently, there have been considerable ad-
vances in our understanding of the nature and time
course of axonal injury (AI) after TBI (Graham
et al., 2000). There is also increasing evidence
that an mTBI may represent the mildest form
of a diffuse brain injury. The descriptor ‘‘minor
Traumatic Brain Injury’’ (mTBI) has been used to
refer to such an injury (Gennarelli et al., 1998;
Kelly, 1999; Graham et al., 2000).
The purpose of this manuscript is to review re-
cent studies in the epidemiology, neuromechanics,
neuropathology, neurobiology and associated ne-
urometabolic changes in mTBI in order to under-
stand why, following a ‘‘mild’’ insult to the head,
brain cells are more susceptible and are more
‘‘vulnerable’’ to a second and perhaps more dam-
aging accident.
Epidemiology of mTBI
Injuries to the head constitute a significant pro-
portion of the total number of injuries in many,
different contact sports (Kelly et al., 1991; Cantu,
1992; Biasca et al., 1995; Tegner and Lorentzon,
1996; Clayton, 1997; Dick, 1997; Roberts et al.,
1999; Biasca, 2000; Tegner, 2000; Tegner et al.,
2000; Laprade, 2001; Pashby et al., 2001; Sports-
Related Recurrent Brain Injuries–United States,
1997). National and international analyses report
that the proportion of mTBI to the overall number
of injuries fluctuates in American football between
3 and 24%, and in soccer between 4 and 22%
(Castaldi, 1993; Tegner and Lorentzon, 1996;

Clayton, 1997; Dick, 1997; Gennarelli et al., 1998;
Biasca et al., 1999; Roberts et al., 1999; Biasca,
2000; Ranalli, 2000; Tegner, 2000, 2001; Tegner
et al., 2000; Burke, 2001; Laprade, 2001; Pashby
et al., 2001).
In ice hockey, more specifically, the proportion
of mTBI to the overall number of injuries fluctu-
ates between 2 and 20%. More detailed analysis
has also shown that the frequency of mTBI has
risen markedly over the last 15 years. In the Ca-
nadian Hockey League (CHL) Clayton (1997) re-
ported that the proportion of mTBI rose from 4%
of all injuries in the period 1991–1996, to 8% in
1997, 14% in 1998 and 1999 and 17% in the period
1999–2000. The National Collegiate Athletic
Association (NCAA) Injury Surveillance System
recorded a similar increase in North America Jun-
ior Hockey where the incidence of mTBI almost
doubled between 1986 (0.7 injuries per 1000 AE in
1986) and 1996 (1.5 injuries per 1000 AE in 1996)
(Dick, 1997). Similarly, in a study in Scandinavia,
one Swedish ice hockey team reported a rising in-
cidence of mTBI over the last 15 years from 2% of
all injuries in the season 1986–1987 to 18% in the
season 2000–2001 (Tegner, 2000; Tegner et al.,
2000). In parallel, in the professional Ice Hockey
Fig. 1. Percent of mTBI, i.e. cerebral concussion, to all injuries in Sweden, Canada, North America and Switzerland from 1986 to 2001
(Clayton, 1997; Biasca et al., 2000; Tegner et al., 2000; Tegner, 2000; www.hockeyinjuries).
265
League (NHL), John Powell (2001) reported an
increased incidence of mTBI from 2% in the sea-
son 1989–1990, to 4.9% in 1995–1996, to 8% in
each of the last two Seasons 1999–2001 for which
data is available (www.hockeyinjuries). Lastly,
mTBI represented 20% of all head injuries in the
Swiss National Ice Hockey Leagues A and B dur-
ing the seasons 1996–1998 (Fig. 1) (Biasca, 2000).
However, the diagnosis of an mTBI is subject to
widespread controversy both in definition and
the degree of severity of injury. This controversy
has contributed to the current paucity of reliable
epidemiological data. It has recently been sug-
gested that it would be helpful to have a world-
wide-standardized assessment for mTBI and a
centralized data pool using the International
Sports Injury System ‘‘ISIS’’. Such would greatly
facilitate international comparison through use of
an Internet-based database system (Biasca and
Tegner, 2001).
Neuromechanics of mTBI
The most common mechanical input to the head is
dynamic loading of either impact or impulsive type
266
(Elson and Ward, 1994; Gennarelli, 1991; Bailes
and Cantu, 2001; Graham et al., 2000).
Impact loading requires a direct blow to the head
occurring in a period of less than 200 ms (Elson
and Ward, 1994; Gennarelli, 1991) and, in most
cases, in less than 20 ms (Graham and Gennarelli,
1997). It occurs when a blunt object strikes the
head and usually initiates a combination of con-
tact and inertial forces that result in a series of
events. The forces vary with the size of the im-
pacting object and the magnitude of the force deli-
vered to the contact point, the so-called contact
phenomena (Elson and Ward, 1994; Graham et al.,
2000). These contact phenomena are a complex
group of mechanical events that occur both locally
and distant from the point of the impact. They
may cause laceration of the scalp, skull deforma-
tions, fractures, cerebral contusions and propaga-
tion of shock waves that travel through the skull
and brain. Protective headgear is designed to
reduce the severity of these injuries (i.e., skull de-
formation, fracture and hematoma).
Impulsive loading, in contrast, occurs when the
head is put into motion and accelerated as a result
of either an impact to another part of the body, or
as a result of a secondary response to a direct im-
pact. Impulsive loading occurs also when the mov-
ing head is stopped without it striking anything or
is arrested by impact (Elson and Ward, 1994;
Gennarelli, 1991; Graham et al., 2000). Under
these conditions the resulting head injuries are
solely caused by inertia resulting from the manner
(translation, rotation, angulation) in which the
head is moved. In biomechanical terms this results
in shearing stresses within the brain substance, re-
sulting in the incidence of foci of strain. Strain is
the amount of deformation of brain or other tissue
as a result of the application of mechanical force.
There are three types of strain: compression, ten-
sion and shear. The brain is highly incompressible
but will deform in shear most easily when sub-
jected to rotational loads (reviewed by Margulies
and Thibault, 1989). Individual axons can sustain
only tensile strain, which is the amount of elon-
gation that occurs when a material is stretched.
Biological materials withstand strain if they are
deformed slowly. But in head-injury impulsive and
inertial loading, giving rise to dynamic strain, is
both more frequent and more damaging to the
substance of the brain because the substance of the
brain is exposed to mechanical loading over a very
short time span — 2–20 ms — and has a very low
tolerance of tensile and shear strain. Thus, the
highest levels of damage to the brain result from
exposure to shearing stresses where individual ax-
ons are stretched.
Inertial forces are responsible for the two most
important types of damage encountered in blunt
head injury:
First, acute subdural hematoma (SDH) resulting
from tearing of subdural bridging veins. Such in-
jury occurs when the rise time and duration of
inertial loading to the head is relatively short over
2–12 ms (Gennarelli and Thibault, 1982). SDH
occurs most frequently in patients that have
experienced a fall or been exposed to an assault
(Graham and Gennarelli, 1997).
Second, widespread damage in the white matter
resulting in diffuse axonal injury DAI (Elson and
Ward, 1994; Gennarelli, 1991; Graham et al.,
2000) where inertial injury to the head occurs over
a longer time span — 11–22 ms (Gennarelli et al.,
1982). Adams et al. (1989) defined DAI and dis-
tinguished three levels or grades of severity of in-
jury to white matter — grade I in which there is
widespread damage to axons within the cerebral
hemispheres; grade II in which there is, in addition
to the above, the occurrence of focal lesions in the
corpus callosum often related to small hemorrh-
ages (petechia) or tissue tears and grade III in
which there is also damage to axons within the
rostral brainstem.
The brain, which is housed in the protective
bony cranium and bathed by a layer of cerebro-
spinal fluid (CSF), has freedom of movement be-
fore it abuts the cranium (i.e., smooth intracranial
surfaces, CSF interfaces). The CSF provides a nat-
ural shock-absorbing system, converting focally
applied external stress to compressive stress, fol-
lowing the contours of the sulci and gyri, and dis-
tributing the force in a uniform fashion. The CSF,
however, does not provide complete protection
against shearing forces being induced in the brain,
especially when rotational forces are applied to the
head and shearing forces occur at those sites in the
brain where rotational gliding is hindered — for

example by the relatively rigid membranes forming
the falx cerebri and the tentorium cerebelli. The
incidence of DAI is greatest when the brain is ex-
posed to rotational forces arising during a hori-
zontal/coronal displacement of the head. Inertial
effects result in the skull moving laterally before
the brain does and the brain may continue to move
after the head has stopped moving. The falx and
tentorium are firmly adherent to the internal sur-
face of the skull. The brain, to the contrary, floats
in CSF. The mid-line falx will therefore move
within a different time frame to the brain. The
medial side of one cerebral hemisphere may contact
the surface of the falx while the other hemisphere
continues moving. Structures that cross the mid-
line, for example nerve fibers in the corpus callo-
sum will therefore be exposed to tensile strain as
one hemisphere moves away from the other. Thus
nerve fibers passing through the corpus callosum or
through the cerebral peduncles have a greater risk
of being exposed to tensile strain and thus to in-
jury. In the last several years a widely held con-
sensus has developed that there is a spectrum of
levels of injury to nerve fibers when they are ex-
posed to tensile strain. When an athlete or patient
experiences either mild concussion — (disturbance
of neurological function without loss of conscious-
ness), classical cerebral concussion — (a tempo-
rary, reversible loss of neurological function with
temporary loss of consciousness for up to 6 h) or
more severe head injury when the patient is un-
conscious for more than 6 h — is now thought that
nerve fibers are injured in all of the above clinical
scenarios. The distinction between the above levels
of injury is thought to reflect the number and dis-
tribution of injured fibers within the white matter
of the brain. The less severe forms of injury in
which only a small number of fibers are injured are
now thought to be the principal causes of an mTBI
(Elson and Ward, 1994; Gennarelli et al., 1998;
Kelly, 1999; Graham et al., 2000). The situation is
further complicated in that an inertial loading of
the brain and head, whether caused by an impact
or by an impulsive loading is influenced by the di-
rection of movement of the head during the injury
episode. If the acceleration causes a straight move-
ment of the center of the head, then we have a
linear acceleration; if the center of gravity of the
267
head travels in an arc then we have a rotational
acceleration (Bishop, 1997, 2000; Bishop and
Arnold, 1993; Bailes and Cantu, 2001). Focal
brain injuries are more common after a linear
acceleration and diffuse brain injuries after a
rotational acceleration (Bishop, 1997, 2000; Bishop
and Arnold, 1993). However, in humans, almost
any injury event involves a combination of these
two and separation of brain responses due to
rotation and linear acceleration is almost impossi-
ble (Bishop, 1997, 2000; Bishop and Arnold, 1993).
Often the mechanism of a head injury is complex
involving a combination of focal and diffuse com-
ponents. Direct head contact may induce both a
rotational acceleration to cause both diffuse and
focal injury (Bishop, 1997, 2000; Bishop and
Arnold, 1993; Gennarelli et al., 1998; Kelly, 1999;
Graham et al., 2000). Further, the direction in
which the head moves plays an important role in
determining the amount and distribution of axonal
damage in a given situation (Gennarelli et al.,
1998).
mTBI in ice hockey and correlation with the type of
blow to the head
Specifically in relation to ice hockey, following
examination of 11 cases of mTBI in the profes-
sional NHL, in the Canadian Amateur Hockey
League and in the Swiss Ice Hockey League, it was
proposed that three different possible mechanisms
may be responsible for causing mTBI in ice hockey
(Bishop, 2001):
1. a direct eccentric blow to the head,
2. a direct blow to the face and
3. a blow directed to the chin.
A direct eccentric blow to the head, not passing
through the center of mass of the head, may ex-
pose brain tissue to many forces. These may be
simplified as a force Ft with translatory compo-
nents (with transversal and axial forces) and a
force Fr with rotatory components in the coronal,
sagittal and transverse (horizontal) planes (Fig. 2).
The rotatory component in the coronal plane, af-
ter a blow to the side of the head, causes the most
severe injuries within the brain while a blow in the
268
Fig. 2. Example of a direct eccentric blow to the head: a pro-
tective headgear, with its padding, helps to reduce the effect of
the final force F and therefore may reduce the risk of a focal
head injury. (Mt: Moving force in the transversal plane; Ms:
moving force in the sagittal plane; Mf: moving force in the
frontal plane; Ft1: force 1 in the transversal plane; Ft2: force 2
in the transversal plane; Fa: force in the axial plane.) Reprinted
with permission and courtesy of the player.
sagittal plane (flexion/extension of the cervical
vertebral joints) is tolerated best (Gennarelli et al.,
1998). Protective headgear including padding,
helps to reduce the overall effect of the final force
F, and also reduces the contributory effects of
contact and rotational phenomenon. Hockey hel-
mets were originally developed to reduce the risk
of serious head injuries, precipitated by direct
blunt traumas and provided protection against
brain injury death and/or intracranial hematoma
(Odelgard, 1989; Bishop, 1997, 2000; Bishop and
Arnold, 1993; Dixon and Brodie, 1993; Pashby,
1993; Stoner and Kreating, 1993; Pashby et al.,
2001; Tegner, 2001). The two most critical func-
tions of helmets are to reduce impact energy and
load distribution to the head. These properties re-
duce the magnitude of the forces applied to the
head, reduce the stress and strain to which the
skull and brain are exposed and thus reduce the
severity of mTBI and head injury. Thus, protective
headgear first of all reduces the risk of a focal head
Fig. 3. Example of a direct blow to the face or jaw: There is no
protective equipment, such as helmets and mouth guards, which
may be effective in reducing any rotatory component. (Mt:
Moving force in the transversal plane; Mf: moving force in the
frontal plane; Ft: force in the transversal plane.) Reprinted with
permission and courtesy of the player.
injury, but although development of padding ma-
terials has resulted in improvement of the energy
absorption characteristics of helmets, the crucial
factor causing mTBI is widely acknowledged to be
the effect of rotational acceleration of the brain.
A direct blow to the face may cause both a force
Ft with translatory and force Fr with rotatory
components. In ice hockey contacts to the jaw and
face, a similar mechanism to the knockout punch
in boxing, are numerous. As a result there is often
a significant rotational acceleration of the head
and brain (Fig. 3).
Gennarelli et al. (1982) found that the severity
of experimental TBI in a non-human primate
appeared to be related to the plane of rotational
acceleration. For equivalent levels of angular ac-
celeration, nerve fibers in the white matter of the
brain were most vulnerable to injury if the head
was rotated laterally (vide supra). When the brain
moved laterally tensile strain occurred in a coronal
plane parallel to the long axis of nerve fibers
crossing the sagittal plane. Lesser degrees of injury
were obtained when the brain moved either in
the sagittal or horizontal planes (Gennarelli
et al., 1982). Recently Smith et al. (2000) have

suggested that the severity of experimental TBI
may be more of a reflection of the severity of ax-
onal damage in specific regions of the brain, most
notably the brainstem, rather than the total sum of
AI distributed throughout the organ (Smith et al.,
2000). Following rotation of the head in the axial
plane, substantially more axonal damage was
found in the brainstem, compared with that in-
curred when the head was rotated in the coronal
plane (Smith et al., 2000). But care needs to be
taken in the interpretation of these results and
their comparability with injury in the human. In
the pig the brain stem lies in the same approximate
longitudinal axial plane as the cerebrum. In the
human, however, the brain stem lies at an angle of
some 1201 to the cerebrum. Thus, in the pig an
axial injury which occurs in the horizontal plane of
the head (Fig. 1 in Smith et al., 2000) will provide
the greatest mechanical strain to nerve fibers at the
lateral limits of the brain stem. Indeed, Smith et al.
(2000) illustrate such data in Fig. 4 of their paper.
In humans, however, because of the different
planes of orientation of the longitudinal axes of
the cerebrum and brain stem mechanical strain will
be placed largely on those nerve fibers lying par-
allel to the direction of displacement of the head.
That is in coronal and parasagittal fibers within
the cerebral hemispheres. Nonetheless, Smith et al.
(2000) are the first to suggest (1) a relationship
between the plane of head rotational acceleration
and the distribution of AI and (2) that injury to
axons in the brainstem plays a major role in the
induction of an immediate post-traumatic coma.
With specific regard to sports-related injuries, no
protective equipment, which may reduce the ef-
fects of this rotatory force (Fr), is presently avail-
able. For this reason it is necessary to eliminate the
practice by players of checking against the head
and neck of other players and which may thereby
expose the brain to rotatory acceleration/deceler-
ation. The Rules Committee of the International
Ice Hockey Federation IIHF, under the chairman-
ship of Mr. Philippe Lacarrière, has therefore in-
troduced in 2002 new ‘‘Head-Checking Rules’’
(IIHF Rule No. 540: http://www.iihf.com//
hockey/rules/rules.htm) in order to eliminate
every intentional or unintentional check or blow
to the head and neck. This rule has been made
269
Fig. 4. Example of a direct blow to the chin: mouth guards may
have the ability to absorb the impact loading from the chin and
to distribute the remaining energy, throughout the resilience of
its material combined with its design, over a much larger sur-
face area, thereby reducing the final force on the brain. (Ft:
Force in the transversal plane.) Reprinted with permission and
courtesy of the player.
mandatory, worldwide since June 2002. This
means that every player making contact to the
head and neck with the body, elbow, shoulder,
knee or stick of an opposing player will be penal-
ized, at the discretion of the referee, with either a
Minor (2 min) or Major (5 min) Penalty plus au-
tomatic Game Misconduct or even Match Penalty.
Prevention strategies, such as the introduction of
‘‘Checking from behind’’ Rules in 1994, have be-
come effective during the last years in decreasing
the number of severe spinal injuries worldwide.
Therefore, it is hoped and expected that these
new ‘‘Head-Checking’’ rules should reduce the in-
cidence and severity of mTBI.
A blow directed to the chin may cause a trans-
latory force Ft, which is transmitted from the chin
through the lower jaw, through the temporoman-
dibular joint at the base of the skull and then to
the brain. These forces may also propagate shock
waves through the brain, which may induce mTBIs
(Fig. 4).
Mouth guards alone or in conjunction with ad-
ditional face protection, have been demonstrated
270
to be effective in preventing and/or reducing the
incidence of dental and orofacial injuries (Piccin-
inni, 2001). Numerous anecdotal reports have in-
dicated a correlation between the use of mouth
guards and prevention of mTBI. Piccininni (2001)
focused on two primary mechanisms to explain
this correlation. The first mechanism was dissipa-
tion of forces delivered to the maxilla, skull and
temporomandibular joint complex when the man-
dible received a blow from the front whereby the
force applied to the mandible is transmitted to
both temporomandibular joints, absorbed in part
by the articular disc and fibrous capsule of those
joints before being transmitted to the temporal
bone and the base of the skull. The second mech-
anism was the stabilization of the skull through the
increased contraction of the muscles of mastica-
tion, principally the massetur and temporalis mus-
cles, when clenching the lower jaw. This may be
enhanced with the presence of a mouth guard (Pi-
ccininni, 2001). For this reason Piccininni (2001)
stated that mouth guards can and should be pro-
moted as effective devices for the prevention of
dental and orofacial injuries and that properly de-
signed mouth guards may also play a role in the
reduction of incidence or severity of mTBI. How-
ever, evidence of any injury protection using
mouth guards is derived from case series and ret-
rospective injury surveys. Further epidemiological
research using the ISIS will surely contribute to an
increased understanding of the incidence and
mechanism of any injury and allow a reduction
of the risk of injury. The use of properly fitted
mouth guards would no doubt reap numerous
benefits in term of reducing medical, financial,
cognitive, psychological and social consequences
of any dental and cerebral injuries, like mTBI, at a
minimal cost. As a result, the Rules Committee of
the IIHF has made in 2002 some rule changes.
This rule will be mandatory, as a first step, for all
junior ice hockey players younger than 20 years
and not wearing full-face masks. As a next step,
this rule should be mandatory for all players born
after December 31, 1975 and for all players wear-
ing full-face masks as well, because a full-face
mask alone can not reduce the forces transmitted
to the brain from any blow to the jaw or chin
(Biasca et al., 2002).
Neuropathology of mTBI
The pathology of human TBI is complex and het-
erogeneous, comprising either single or a combi-
nation of focal and diffuse lesions (Gennarelli,
1991; Elson and Ward, 1994; Graham et al., 2000;
Tegner, 2001). Trauma to the skull and brain
manifests in a variety of pathological hallmarks
reflecting the distribution of the impact energy.
Sharp objects at high velocity tend to cause per-
forating focal cortical contusions (FCC). Colli-
sions against sharp edges lead to non-penetrating,
mostly focal head injuries (FHI). Blunt trauma —
where the head does not necessarily hit anything
but is exposed to rapid acceleration and deceler-
ation — often leads to diffuse trauma, that is DAI.
FCC, FHI and DAI often occur together at any
combination of lesion severity (Elson and Ward,
1994; Graham et al., 2000).
FHI are caused by forces of contact and head
acceleration from direct blunt trauma, such as be-
ing struck with a hockey stick or a puck, or falling
on the ice surface and striking the head. These
types of injuries may produce a focal cortical con-
tusion with well-demarcated damage to the paren-
chyma of the brain and associated disruption of
arterial and venous vessels in the region of the
contusion. Such gross lesions may readily be ob-
served in CT scans, MRI or macroscopically at
autopsy. FHI often result in fracture of the skull,
with or without an associated epidural hematoma,
in cortical contusions and in intracranial hem-
orrhage (Gennarelli, 1991; Bishop, 1997; Bishop
and Arnold, 1993; Elson and Ward, 1994; Kelly,
1999; Graham et al., 2000; Tegner, 2001). Tissue
damage occurs either by direct mechanical de-
struction of brain parenchyma, compression
through interstitial bleeding or by secondary inf-
arction following disrupted circulation. Wearing
an internationally approved, standardized helmet
will clearly reduce the number and severity of these
focal injuries (Bishop, 1997; Bishop and Arnold,
1993).
Diffuse head injuries, on the other hand, are
caused by the inertial effect of either a mechanical
blow to the head or rapid rotation acceleration/
deceleration phenomena as a result of a blow to
another part of the body or a fall (Gennarelli,

1991; Elson and Ward, 1994; Graham et al., 2000;
Tegner, 2001). In ice hockey, such injuries may
involve axonal damage due to excessive angular
acceleration (e.g., after an abrupt body check)
caused by either direct or indirect inertial loading
of the head and are characterized by a global dis-
ruption of neurological function (Gennarelli, 1991;
Elson and Ward, 1994; Graham et al., 2000;
Tegner, 2001). It is now well recognized that (1)
traumatic axonal injury (TAI) is a consistent
feature of human TBI, particularly those injuries
involving rapid acceleration/deceleration of the
brain (Elson and Ward, 1994; Graham et al., 2000)
and (2) that these injuries represent a continuous
spectrum of the same pathology: a progressive
widespread but heterogeneous damage to axons
(Gennarelli, 1991; Elson and Ward, 1994; Graham
et al., 2000; Tegner, 2001).
The majority of studies of TAI have long been
focused upon human DAI and animal models
thereof. Here TAI has been linked to the morbid-
ity associated with that condition (Strich, 1961,
1970; Nevin, 1967; Peerless and Rewcastle, 1967;
Oppenheimer, 1968; Clark, 1974; Adams et al.,
1982; Gennarelli et al., 1982; Povlishock et al.,
1983; Pettus et al., 1994; Maxwell et al., 1997).
Historically, it was suggested that forces acting
during TBI and TAI tear or shear axons, causing
them to retract from the site of injury and expel a
ball of axoplasm. This forms a reactive swelling or
retraction ball, the histological structural feature
then used to diagnose DAI (Adams et al., 1982).
More recent studies, however, have not provided
support for the concept that large numbers of
axons are sheared at the time of injury. Rather
experimental, animal models have demonstrated
that traumatic injuries elicit a delayed or second-
ary axotomy wherein the trauma initially injures
an axon, which, however, remains in continuity,
causing it, over a period of at least several hours,
to fail at one or more foci, and disconnect. A re-
active swelling of classic description then forms
(Povlishock et al., 1983; Povlishock, 1986, 1992,
1993; Gennarelli, 1991; Pettus et al., 1994; Povli-
shock and Christman, 1995; Povlishock and Jenk-
ins, 1995; Povlishock and Pettus, 1996; Maxwell et
al., 1997; Gennarelli et al., 1998; Graham et al.,
2000). Further, studies within the last decade have
271
provided considerable advances in the understand-
ing of the nature and time course of TAI in and
following TBI (Povlishock et al., 1983; Grady et
al., 1993; Blumbergs et al., 1994, 1995; Christman
et al., 1994; Elson and Ward, 1994; Pettus et al.,
1994; Sherriff et al., 1994a, b; Blumbergs, 1997;
Maxwell et al., 1997; Gennarelli et al., 1998; Gra-
ham et al., 2000).
Graham et al. (2000) reported that over the last
several years a consensus has arisen that the prime
site of injury in an injured axon in TBI is the axo-
lemma. There is now a consensus that pathology in
the injured axon then occurs because of loss of
homeostatic mechanisms that maintain differential
ionic gradients necessary for the normal electrical
activity of the axon.
TAI has been shown in a variety (fluid percus-
sion, cortical impact, rotational acceleration and
stretch-injury) of experimental animal models of
human head injury (Povlishock et al., 1983; Pettus
et al., 1994; Graham et al., 2000). Povlishock et al.
(1983) reported TBI after fluid percussion injury
and controlled cortical impact in rats, cats and
swine. Gennarelli et al. (1982) reported TBI after
rotational head acceleration in a non-human pri-
mate, and Ross et al. (1995) in miniature swine.
Lastly, Gennarelli et al. (1989) reported TAI after
controlled stretch-injury to the guinea pig optic
nerve.
A consensus concerning the sequence of path-
ologies within injured axons has developed from
the above studies of moderate-to-severe TAI gen-
erated in numerous, different laboratories. This
sequence may be summarized very briefly as fol-
lows. First, injury to the axolemma; second, swell-
ing of axonal mitochondria as a result of specific
damage to the internal mitochondrial membrane;
third, loss of axonal microtubules and alterations
of the intra-axonal relationship of neurofilaments;
fourth, the occurrence of so-called ‘‘axonal swell-
ings’’ (foci of increased axonal diameter at which
there is still continuity of the axon on both prox-
imal and distal sides of the region of increased
axonal diameter) and finally axonal disconnection
to form so-called degeneration bulbs or ‘‘axonal
bulbs’’ (Povlishock et al., 1983, 1997; Pettus et al.,
1994; Povlishock and Pettus, 1996; Maxwell et al.,
1997, 1999; Okonkwo and Povlishock, 1999).
272

Povlishock and colleagues initiated such study

in traumatically induced axons using the antero-
grade tracer horseradish peroxidase (HRP) (Po-
vlishock et al., 1983; Erb and Povlishock, 1988).
Their approach was based on the premise that ax-
ons tear at the moment of injury. At injury, axons
laden with anterogradely transported protein
would rupture and expel a mass of protein (per-
oxidase)-containing axoplasm that could be easily
visualized using histochemical methods at by both
light and electron microscopes (Povlishock et al.,
1983). But evidence for tearing or shearing — that
is loss of the integrity of the axolemma and the
presence of label free within the periaxonal space
— was not obtained (Erb and Povlishock, 1988).
Rather at 1 h and later after injury a local, lobular
pooling of intra-axonal peroxidase still in conti-
nuity with both proximal and distal parts of the
peroxidase containing axon occurred. Moreover,
the diameter of the focal pools of HRP increased
in diameter — from
10 mm at 1 h to 18 mm at 6 h
after injury (Erb and Povlishock, 1988). In addi-
tion, multilobulated profiles interconnected by Fig. 5. Light micrograph (A) illustrates the changes in intra-
thin HRP labeled bands were observed. Evidence axonal anterograde horseradish peroxidise passage following
for loss of continuity of labeled axons was only fluid percussion injury. Note that within several hours of the
traumatic insult the axon shows a lobulated swelling with a site
described at 2 h and later after injury. The pooling
of axonal narrowing (red arrow). Reprinted with permission
of peroxidase was suggested to occur because of a and courtesy from Maxwell et al. (1997). Light micrograph (B)
focal impairment of axonal transport. Continued illustrates a lobulated terminal swelling after axonal disconnec-
anterograde tracer delivery to the site, resulted in tion to form a conspicuous proximal bulb. Human material
focal expansion of diameter of the axon, swelling labelled with antibodies for beta amyloid precursor protein.
and ultimately detachment from its distal coun-
terpart (Figs. 5A, B) (Povlishock et al., 1983; loss of axonal microtubules and a reduced spacing
Povlishock, 1992; Pettus et al., 1994; Maxwell between, or compaction, of neurofilaments. How-
et al., 1997). ever, such alterations were not representative of all
A second approach using HRP, as a marker of axons demonstrating pathology. In some axons,
AI (Pettus et al., 1994; Pettus and Povlishock, despite the fact that peroxidase had not entered the
1996), was to infuse peroxidase into the CSF of axoplasm, the components of the cytoskeleton no
anaesthetized animals prior to TBI (Pettus and longer ran parallel to the long axis of the axon but
Povlishock, 1996). The intact axolemma normally adopted a spiral course through the axoplasm
excludes molecules as large as HRP from the axo- (Pettus and Povlishock, 1996). Thus in mild ex-
plasm. But, within 5 min after TBI, axons were perimental TBI neurofilament misalignment and
obtained that contained peroxidase and such labe- axonal swelling were not associated with the pas-
led axons were obtained up to 6 h, the end of the sage of peroxidase from the extracellular to the
experimental period examined, after injury. Thus intra-axonal compartment (Pettus et al., 1994,
TBI resulted in damage to the axolemma of some 1996). The latter findings provided support for the
axons such that peroxidase was able to enter the concept that there is a spectrum of AI in TAI.
axoplasm (Pettus et al., 1994; Pettus and Povli- More recently, this concept has been strengthened
shock, 1996). In the aforesaid axons there was also by the findings of other investigators who have

shown that TAI does not affect only large axons,
as had been suggested in the mid 1980s, but also
injures small, myelinated fibers that occur, for ex-
ample in the optic nerve (Jafari et al., 1997, 1998).
Indeed, these latter studies showed that alterations
for numbers of and spacing between microtubules
and neurofilaments differed between nerve fibers
of different axonal diameter. Most recently, a fur-
ther complication of the study of developing pa-
thology in TAI has been indicated in that
immunocytochemical labeling of injured axons
for either b-amyloid precursor protein, a normal
component of axoplasm, or compacted neurofila-
ments (Stone et al., 2001) has suggested that one
subtype of injured axons undergo a progressive
focal swelling resulting in disconnection or sec-
ondary axotomy between 2 and 6 h while others
show compaction of neurofilaments within min-
utes of injury but do not undergo any further
change over the following several hours.
Over the last decade use of antibodies against
amyloid precursor protein (APP) (Sherriff et al.,
1994a, b; Gentleman et al., 1995; McKenzie et al.,
1996) has been adopted as the marker of choice for
detecting TAI in the diagnostic laboratory. APP is
a membrane-spanning glycoprotein originating
from a gene on chromosome 21 and of unknown
function (Hartmann, 2001). It is transported by
fast axoplasmatic transport, is located at both the
pre- and post-synaptic sites and accumulates
within axonal swellings in TBI very quickly in hu-
mans (Blumbergs et al., 1994; Blumbergs, 1997) as
a result of loss of fast axonal transport. However,
the recent findings by Stone et al. (2001) now
question the sensitivity of a-APP immunohisto-
chemistry alone as a tool for demonstrating TAI.
Rather, a range of immonocytochermical markers
should be used because use of APP labeling prob-
ably underestimates the true number of axons un-
dergoing pathology. Indeed, use of the antibody
Ab38 that labels sites of calpain-mediated spectrin
proteolysis (Buki et al., 1999), shows increased
density of axons/mm2 labeled for Ab38 (calpain-
mediated spectrin proteolysis) in the pyramidal
tracts and medial lemniscus of rats after impact
acceleration injury at 60–120 compared with
15–30 min after injury. In addition, a-APP labe-
ling techniques have also shown accumulation of
273
b-APP in relatively normal-looking axons (Gen-
tleman et al., 1995) and may simply represent
changes in axoplasmic transport of a temporary
nature in axons which, hypothetically, have a po-
tential for recovery (Gennarelli, 1991; Gentleman
et al., 1995; Gennarelli et al., 1998). The time
course of recovery is unclear, but it is interesting to
note that Blumbergs et al. (1994) reported that b-
APP accumulation can be shown in humans as
long as 99 days after the initial injury (Blumbergs
et al., 1994, 1995; Blumbergs, 1997).
Overall, these immunocytochemical alterations
have provided further confidence that, after TAI,
there is a focal impairment of axoplasmic trans-
port, leading to progressive axonal swelling and
detachment/secondary axotomy over a period of
several hours or days after injury (Grady et al.,
1993; Christman et al., 1994; Pettus et al., 1994;
Povlishock and Christman, 1995; Maxwell et al.,
1997; Graham et al., 2000).
However, patients with mTBI rarely come to
neuropathological examination. Nonetheless,
there is increasing evidence that some permanent
damage may be present (Christman et al., 1994;
Blumbergs, 1997; Gennarelli et al., 1998; Graham
et al., 2000). Initially it was thought that mTBI
might produce only temporary disturbances of
brain function without gross structural changes.
Now there is increasing evidence that mTBI, in
both animals and humans, can lead to an impair-
ment of axoplasmic transport, which results in a
progressive pathology comparable with that de-
scribed earlier in moderate or severe TBI (vide
supra). Thus axonal swellings form in mTBI, fol-
lowed by axonal detachment in ‘‘delayed axo-
tomy’’ (Blumbergs et al., 1995; Povlishock and
Christman, 1995; Maxwell et al., 1997; Gennarelli
et al., 1998; Graham et al., 2000). This process of
delayed axotomy is a complex network of inter-
acting functional, structural, cellular and molecu-
lar changes (Gennarelli et al., 1998). Fairly recent
evidence (Maxwell et al., 1999) has provided fur-
ther confidence for the concept that tensile strain
to an axon leads to injury to the axolemma. Cyto-
chemical techniques which label for activity of
Ca2+-ATPase and Na+/K+-ATPase (pNNP-ase)
show loss of activity of these two membrane
pumps after injury. Loss of such membrane pump
274
activity would result in a loss of transmembrane
homeostasis for ions, and in particular for Ca2+.
The resulting influx of calcium down the normal
diffusion gradient, which maintains Ca2+ levels
10,000 times higher in the extra-axonal compart-
ment in the normal axon, will result in elevation of
intra-axonal concentrations of Ca2+ and allow
activation of calpain-mediated spectrin proteloysis
(CMSP). Spectrin is an integral component of the
subaxolemma cytoskeleton. Buki et al. (1999)
showed that CMSP breakdown products appear
first within the subaxolemma compartments of in-
jured axons at 15 min and only extend throughout
the whole diameter of the axoplasm at 1 h after
injury. Unfortunately CMSP immunocytochemis-
try was not examined beyond 2 h after injury. But,
quantitative freeze-fracture studies provided mor-
phological evidence of damage to cell membranes
(Maxwell et al., 1999), both the axolemma and
myelin sheath, up to 24 h after TAI. In parallel,
excess axoplasmic calcium leads to a dysfunction
and a breakdown of neurofilaments and microtu-
bules. Further, release of acetylcholine contributes
not only to the depolarization of neurons but also
to calcium-associated excitoxic tissue damage.
Thus there is increasing evidence that the initial
site of damage is the axolemma. Such damage may
then allow abnormal ionic or molecular influx into
the injured nerve fibers. Over time, these changes
lead to foci of disruption of axonal structure and
impaired axonal transport culminates in secondary
or delayed axotomy.
Thus, experimental studies of TAI, over the last
decade, have lead to the now widely held consen-
sus that there is a spectrum of pathological axonal
changes reflecting both the severity of injury (Po-
vlishock et al., 1983; Pettus and Povlishock, 1996)
and size of axons (Maxwell et al., 1997, 2003). The
situation is complicated because there is increasing
evidence that the spectrum of axonal response to
TAI may reflect either or a combination of (1) the
level of tensile strain applied to individual axons at
the time of injury, (2) the degree of damage to the
axolemma leading to the injured axon entering the
‘‘pathological cascade’’ which results in secondary
axotomy, (3) the size of the axon before injury in
that large and small axons respond differently to
TAI and (4) whether the axonal response is
manifested as either long-term compaction of ne-
urofilaments or loss of fast axonal transport lead-
ing to secondary axotomy.
APP as a marker for TAI, is of particular im-
portance with regard to mTBI following the very
important observation of multifocal AI within the
fornices in human beings with a recorded loss of
consciousness of as little as 60 s (Blumbergs et al.,
1994). Thus TAI has been documented after rel-
atively mild head injury and in milder grades of
DAI in humans (Christman et al., 1994; Blumb-
ergs et al., 1995; Povlishock and Christman, 1995;
Povlishock and Jenkins, 1995). There is also a
growing consensus that mTBI may represent the
mildest form of TAI (Povlishock et al., 1983; Elson
and Ward, 1994; Pettus et al., 1994; Gennarelli
et al., 1998). At mild levels of AI, a primary mis-
alignment of the components of the axonal cyto-
skeleton occurs in which neurofilaments and
microtubules assume a spiral orientation within
the axoplasm independent of damage to or an in-
crease in permeability of the axolemma as may be
indicated through use of HRP tracers within the
CSF (Sherriff et al., 1994b; Gentleman et al., 1995;
Pettus and Povlishock, 1996; Maxwell et al., 1997).
In damaged axons regions of increased axonal di-
ameter are termed axonal swellings. Here conti-
nuity of the axon is maintained on both the
proximal and distal sides of the swelling (Fig. 6)
(Povlishock et al., 1983; Pettus et al., 1994; Sherriff
et al., 1994b; Gentleman et al., 1995; Maxwell
et al., 1997).
Fig. 6. This light micrograph of human material labelled for
beta amyloid precursor protein, illustrates, early in the post-
traumatic course, the occurrence of two regions of increased
axonal caliber where the two swellings are still linked.

In parallel with moderate or severe TAI, in
mTBI in sports-related injuries the magnitude, du-
ration and rate of onset of angular acceleration as
well as the direction of head motion during the
injury episode will determine the amount of axonal
damage, the severity and the reversibility of the
clinical syndromes and the resulting neurological
deficits (Gennarelli et al., 1998; Graham et al.,
2000). Although, hypothetically, both the number
and overall anatomical distribution of injured ax-
ons is lower and less widespread compared with
the severe traumatic state, the very presence of
immunocytochemically labeled, damaged axons is
indicative of AI and thought to reflect the lower
end of the spectrum of TAI (Povlishock, 1992,
1993). It is now becoming accepted that there are a
range of categories of diffuse brain injury, which
have the same pathophysiology within the indi-
vidual, injured axons; but a different clinical out-
come. The latter, hypothetically, relates to the
severity of axonal elongation resulting from me-
chanical strain placed upon axons at the time of
injury (Elson and Ward, 1994; Graham et al.,
2000).
Current thinking with regard to the spectrum of
axonal responses between momentary periods of
unconsciousness on the playing field, mild concus-
sion and concussion may be summarized as fol-
lows in terms of the degree of mechanical strain
placed upon axons at the time of injury and their
physiological/pathological responses.
(1) At the lowest level of strain (5% or less
elongation) (Stage I) (Fig. 7), ionic influx of
various species occurs and causes temporary
failure of the generation and propagation of
action potentials (Gennarelli et al., 1998).
Influx of Ca2+, Na+ and Cl and the efflux
of K+ at this level of injury are fully restored
in a matter of minutes and the damaged
axon recovers its function completely.
With increasing strain applied to the axon, a
series of increasingly severe mechanisms and path-
ological changes occur. If the level of axonal
stretch is in the 5–10% range (Stage II) (Fig. 7)
then additional ionic perturbations are associated
with either fluid fluxes in an attempt to maintain
osmotic balance (Gennarelli et al., 1998) or
275
saturation of the mechanism of mitochondrial se-
questration of excess cytoplasmic calcium. This
causes damage to and swelling of mitochondria
(Okonkwo and Povlishock, 1999), together with
enlargement of the injured axon. Hypothetically,
this mild impairment of axoplasmic transport is
reflected, ultrastructurally, as a non-linear align-
ment of the components of the axonal cytoskele-
ton. It is currently hypothesized that most of these
axons recover and thus restitution of structure and
function may be expected.
But, if the level of stretch is between 10 and 20%
(Stages III and IV respectively) (Fig. 7), then the
axon cannot fully restore ionic and osmotic home-
ostasis and, as a consequence, numerous patho-
logical changes are initiated (Gennarelli et al.,
1998). Recent evidence (Stone et al., 2001) suggests
that there may be two separate pathological path-
ways followed by different axons (Fig. 7).
As the level of clinical injury progresses from
less to more severe, greater amounts of damage to
individual axons and greater numbers of axons are
thought to be involved. In injuries from which the
patient apparently recovers, for example mTBI,
there is presently no evidence for Stages III and IV
damage. This is due to the fact that the patient
resumes normal activities with no detectable
change in their performance and that Stages III
and IV may only be determined post-mortem.
Therefore, it is suggested that Stages I/II represent
the great majority of axonal injuries experienced,
for example by ice hockey and other contact sports
players.
(2) In the context of mTBI, the influence of the
degree of damage to the axolemma upon
subsequent axonal responses has not been
investigated in any systematic or controlled,
experimental manner. But there is now good
evidence that the integrity of and the func-
tional state of the axolemma are related to
axonal responses after TAI. The structural
and physiological integrity of the axolemma
is essential for the generation and propaga-
tion of action potentials. It is suggested that
perhaps electrophysiological assessment of
membrane function would possibly provide
the best means of analysis of the functional
276

WALLERIAN DEGERATION

(24 hours - 7 days)
PROXIMAL REMNANT OF AXON
reduced calibre of axon
neurofilament compaction over
30-50 µm of axonal length
(1 to 12 weeks)
Proteolysis of axonal components
but myelin figures remain
DISTAL REMNANT OF AXON

SECONDARY AXOTOMY
(minimum of 4 hours - 7 days)

(up to 12 weeks)
15 - 20% strain (4 hours - 99 days)
axonal disconnection Compaction of neurofilaments
reduced axonal cross-sectional area
PRIMARY AXOTOMY disruption of the myelin sheath increased g ratio
(less than 1 hour) enlarged periaxonal space
focal compaction of neurofilaments
disruption of the axolemma
proteolysis/destruction of neurofilaments

% strain currently unknown (30 min-7 days)

more than 20% strain multiple foci of neurofilament compaction
membrane renting/fragmentation swelting of/damage to mitochondria
proteolysis of axonal cytoskeleton neurofilament compaction
in zone of axotomy marked infolding of the exolemma
10-15% strain (1.5 - 6 hours)
axonal swelling
loss of axonal transport
loss of microtubules
µM calpain/phosphatase/kinase activation
collapse/loss of neurofilament sidearms
compaction of neurofilaments

5 - 10% strain (15 min - 24 hours)

functional and structural damage to the axolemma
depolarisation and loss of ionic homeostasis
loss of and/or de-activation of ion channels and ATP
dependent membrane pumps, mitochondrial damage
KEY
5% or less strain (seconds) Effects of injury
translent depolarisation Mechanical strain
(Time scale)
Recovery
mechanism and ultrastructural changes
INJURY DESCRIPTOR

Fig. 7. Schematic overview of current thinking with regards to AI in human DAI and animal diffuse traumatic brain injury. (Modified
from Maxwell et al., 1977.)

integrity of an injured axon. The authors are
aware of only two experimental, electro-
physiological studies concerning axonal re-
sponses to the application of tensile strain
(Tomei et al., 1990; Galbraith et al., 1993)
and only the former involved experiments on
mammalian axons. With increasing loading
from 0 to
20% elongation there is rapid
depolarization following by a slow repolar-
ization. But with an increased loading, the
time required for repolarization increases in
a non-linear manner. At loadings resulting
in more than 20% elongation the axolemma
is irreparably damaged (Galbraith et al.,
1993). There is thus an apparent, critical
threshold at or around 20% elongation at
which the axolemma is irreparably damaged.
In mammalian myelinated axons, analysis of
visual evoked potentials in the visual cortex
after TAI to optic nerve showed (1) a

reduction in signal amplitude to 33% of
control values immediately after injury, to
66% at 24 h and 77% at 1 week. (2) An in-
creased signal latency to 126% of control
values after injury, 119% at 24 h and 97% at
1 week in an animal model (guinea pig optic
nerve stretch-injury) in which the current
literature documents that only 20% of fibers
provide morphological evidence for injury
(Gennarelli et al., 1989). Thus, at the func-
tional level, axons are rapidly depolarized at
TAI (Stage I, Fig. 7). However, in this ex-
perimental paradigm, recovery of activity
occurs relatively slowly, will vary with the
initial loading, and total recovery is not ob-
tained even at 1 week after injury (Tomei et
al., 1990). Two research groups have pro-
vided morphological evidence for damage to
the axolemma. First, after fluid percussion
injury, influx of HRP from the CSF into
damaged axons was shown (Pettus and Po-
vlishock, 1996). But, in molecular terms
HRP, is a large molecule and thus influx
into the axoplasm of an injured axon would
necessitate the appearance of relatively large
‘‘holes’’ in the membrane. Hypothetically,
such damage would probably result from
elongation of axons by more that 20% of
their pre-injury length. It is noteworthy that
in the same study the authors noted axons in
which there was axonal pathology but that
HRP had not entered the axoplasm. It is
currently hypothesized that such damage
represents a mild level of injury to those ax-
ons. Second, using the same model as Tomei
et al. (1990), Maxwell et al. (1999) provided
both cytochemical evidence for loss of mem-
brane pump activity — vide supra — and
quantitative data for an alteration in the
number and distribution of intramembra-
nous particles (IMPs) visualized through
freeze-fracture techniques. It is currently be-
lieved that the altered distribution of IMPs
reflects changes in the distribution and
number of transmembrane proteins some
of which are membrane pumps and ion
channels. Hypothetically, such changes may
also reflect damage to the axolemma in fibers
277
which are not damaged to the degree that
large molecules such as HRP may enter the
axoplasm.
(3) Evidence obtained over the last several years
has shown that neurofilaments and micro-
tubules demonstrate a spectrum or range of
responses after TAI. Current results suggest
that at lower or mild levels of injury, there is
focal disarray of neurofilament and micro-
tubule alignment within the axoplasm, but
no loss of microtubules (Povlishock et al.,
1983; Pettus and Povlishock, 1996; Maxwell
et al., 1997, 2003). But, in moderate or se-
vere TAI there is (1) rapid loss of microtu-
bules (Maxwell and Graham, 1997) and (2) a
reduction in the spacing between adjacent
neurofilaments, resulting in both an in-
creased density of neurofilaments per unit
area of the axoplasm (Pettus and Povli-
shock, 1996; Povlishock et al., 1997) and
reduced interneurofilament spacing (Jafari
et al., 1997, 1998; Maxwell et al., 2003).
The term applied to this latter change in the
organization of the axonal cytoskeleton is
that the neurofilaments are said to be ‘‘com-
pacted’’.
More recently, quantitative analysis has shown
that alterations in the organization of the axonal
cytoskeleton after TAI differ (1) in the time course
of the response by microtubules and neurofila-
ments and (2) in different sized axons. After TAI
in the guinea pig optic nerve compaction of ne-
urofilaments is most marked in smaller axons. In-
deed, the reduced spacing between neurofilaments
has been correlated with a reduction in the calibre
of axons (Jafari et al., 1997; Maxwell et al., 2003).
But in larger axons of the optic nerve there is an
increased spacing between components of the
cytoskeleton throughout the length of the nerve.
In these larger fibers, nonetheless, there are focal
sites of compacted neurofilaments. But these sites
are associated with loss of the axolemma and pos-
sibly are sites at which the axon is undergoing the
final stages of secondary axotomy (Jafari et al.,
1998; Maxwell et al., 2003). In other animal
models, however, for example cortical impact
and fluid percussion injury, there is compaction
278
of neurofilaments in relatively large fibers (Pettus
and Povlishock, 1996; Povlishock et al., 1997;
Stone et al., 2001).
(4) Recent, novel work has suggested that com-
paction of neurofilaments may occur in ax-
ons other than those which progress to the
formation of axonal swellings and secondary
axotomy (Stone et al., 2001). However, only
one research group has so far described such
a finding and thus there is presently no other
material that would allow discussion. But, it
is included here as a pointer to the fact that
attention should be paid to this phenomenon
in later investigations of TAI.
Clearly, much further work needs to be done to
fully understand the significance of differing cyto-
skeletal responses to (1) axonal loading, (2) dif-
ferent responses by axons within different CNS
tracts and (3) between axons that are undergoing
secondary axotomy and those that are not. How-
ever, in summary, three different types of evidence
support the hypothesis that damage to the axo-
lemma is key to the development of pathology in
injured axons after TAI. Overall, there is variation
in the sites of axonal damage within the brain of
each and any patient and axonal cytoskeletal re-
sponses may differ between axons in different CNS
tracts. It is intriguing as to whether these detailed
differences in axonal response to TAI are impor-
tant in consideration of the question as to why
some fibers within a particular field are injured
while other, spatially closely related fibers are not
in cases of TAI. It is suggested that the above
should be rigorously considered in the formulation
of any future experimental studies of TAI.
Neurobiology and associated neurometabolic
changes of mTBI
In recent years extensive knowledge concerning
pathophysiological processes of mTBI has been
gleaned from in vivo and in vitro experiments
(Yaghmai and Povlishock, 1992; Hovda, 1996;
Pettus and Povlishock, 1996; Cecil et al., 1998;
Hovda et al., 1994, 1999; Giza and Hovda, 2000;
Maroon et al., 2000). Alterations in ionic fluxes,
cerebral blood flow (CBF) and cerebral glucose
metabolism have all been documented in brain-
injured patients (Giza and Hovda, 2000). Utilizing
animal models, investigators have been able to
better characterize trauma-induced ionic flux and
provide quantitative data in support of the corre-
sponding metabolic dysfunction that has become
the clinically, accepted diagnostic descriptor of
mTBI. Microdialysis probes have detected ionic
disturbances (Katayama et al., 1990; Kawamata et
al., 1992; Hovda et al., 1999; Cantu, 2000; Reinert
et al., 2000; Klatky et al., 2002) and nuclear mag-
netic resonance spectroscopy (MRS) has recently
been used to demonstrate neurochemical altera-
tions associated with pathological conditions after
TBI (McIntosh et al., 1987, 1988; Pettus and Po-
vlishock, 1996; Hovda, 1996; Cecil et al., 1998;
Hovda et al., 1999; Bigler, 1999; Garnett et al.,
2000; Giza and Hovda, 2000; Brooks et al., 2001;
Sinson et al., 2001). Positron emission tomography
(PET) has provided quantitative data for altera-
tions in cerebral glucose metabolism (Nedd et al.,
1993; Ruff et al., 1994; Gross et al., 1996; Abdel-
Dayem et al., 1998; Bigler, 1999; Ricker et al.,
2001; Levine et al., 2002; Umile et al., 2002) while
both radioisotope studies (Hovda et al., 1999;
Ricker et al., 2001) and transcranial Doppler im-
aging (Ng et al., 2000; Muller et al., 2001) have
been used to follow changes in CBF. Fairly re-
cently, application of the above investigative tech-
niques have been applied to human head-injured
patients. These techniques have confirmed and ex-
tended the experimental data derived from in vivo
animal models (Pettus and Povlishock, 1996; Cecil
et al., 1998; Hovda et al., 1999) by providing good
evidence that exactly comparable pathophysiolog-
ical responses occur in head-injured patients.
Thus, there is increasingly convincing evidence
that mTBI triggers a complex and interwoven se-
quence of ionic and metabolic events from which
damaged cells may eventually either recover or de-
generate and die. The metabolic cascade is a mul-
tidimensional process (Hovda, 1996; Giza and
Hovda, 2000). To briefly summarize, after mTBI,
there is a significant K+ efflux from injured cells,
owing to mechanical deformation of the axolemma
and opening of voltage-dependent K+ channels.
These lead to neuronal firing, depolarization

and release of excitatory neurotransmitters. In the
normal brain, excess extracellular K+ is subject to
reuptake by surrounding glial cells. This compen-
satory mechanism can maintain physiologic extra-
cellular K+ levels even after mild TBI or ongoing
seizure activity (Hovda et al., 1999; Matsushita et
al., 2000; Reinert et al., 2000). But more severe TBI,
where greater numbers of axons and neurons are
injured however, will overcome this glial safety
valve. Initially there is a slow rise in extracellular
K+. But as the physiological ceiling for K+ bal-
ance is breached, an abrupt increase occurs, leading
to depolarization, release of excitatory amino acids
(EAAs) and a further massive K+ flux though
EAA/ligand-gated ion channels (Hovda et al.,
1999; Yamamoto et al., 1999). Following this wave
of excitation neurons are hyperpolarized and there
is consequent suppression of neuronal activity, re-
ferred to as ‘‘spreading depression’’, which is
thought responsible for the neurologic dysfunction
across the cortical surface noted in the clinical set-
ting of seizure propagation and migraine aura.
When the extracellular potassium concentration in-
creases beyond the normal, physiological upper
limit (4–5 mmol/L) to levels of 20–50 mmol/L and
greater, then inhibition of the action potential and
loss of consciousness may occur (Maroon et al.,
2000). The EAAs, especially glutamate, activated
kainate and a-amino-3-hydroxy-5-methylisoxazole-
4-propionic acid (AMPA) receptors located on the
cell body which open channels permeable to both
Na+ and K+. EAAs also activate N-methyl-D-as-
partate (NDMA) receptors permeable to Ca2+ as
well as Na+ and K+. The release of glutamate
permits a rapid and sustained influx of Ca2+, which
leads to dysfunction of the oxidative metabolism as
a result of damage to mitochondria (Fineman et al.,
1993; Maxwell et al., 1999) and which is amelio-
rated by therapy using cyclosporin A (Okonkwo et
al., 1999). The cell therefore becomes increasingly
dependent on ATP generated through glycolysis.
Alterations in brain Ca2+ homeostasis through
loss/inactivation of membrane pump Ca2+ ATPase
and receptors/channels associated with Ca2+ entry
(voltage sensitive channels or ionophore-associated
glutamate receptors, such as NMDA receptors)
have been also associated with regional cerebral
edema, vasospasm and delayed cell death (Graham
279
et al., 2000). Traumatic, ischemic or anoxic injury
to axons is associated with widespread neuronal
depolarization and release of EAA neurotransmit-
ters such as glutamate, leading to the opening of
NMDA receptor-associated ion channels and influx
of Ca2+ (Graham et al., 2000). The resulting post-
traumatic Ca2+ storm has been documented using
45
Ca-autoradiography after a lateral fluid percus-
sion injury (Fineman et al., 1993), and indirectly,
through use of cytochemical techniques, as a result
of redistribution of membrane pump calcium-AT-
Pase and Ca2+ influx into myelinated nerve fibers
of the guinea pig optic nerve after stretch-injury
(Maxwell et al., 1995).
Finally, the injured axon may lose the capacity
to buffer, extrude or to bind the intracellular cal-
cium. When the calcium load exceeds the capacity
of axoplamic buffers, there is a consensus that ac-
tivation of calpains and phospholipases initiate
processes of cytoskeletal and membrane degener-
ation. Several recent studies have documented
both acute calpain activation and regional calpain-
induced cytoskeletal proteolysis (Buki et al., 1999;
Graham et al., 2000; Saatman et al., 2003). This
initiates a cascade of structural changes in injured
axons beginning with breakdown of the subaxo-
plasmic cytoskeleton (Buki et al., 1999), parts of
the cytoskeleton related to swollen, damaged mi-
tochondria, and later (between 30 and 120 min af-
ter injury) compaction of cytoskeletal elements
throughout the remnants of the axoplasm. This
pathology results in focal impairment of axoplas-
mic transport leading to focal accumulation of
transported materials and loss of delivery of mol-
ecules required for repair of the damaged axolem-
ma. This leads to more widespread damage to the
axolemma (Buki et al., 1999; Maxwell et al., 1999),
generation of numerous injurious compounds, in-
cluding free radicals and various inflammatory
mediators, for example cytokines, prostaglandins
and leukotrienes. At this stage, since the cell body
is now also damaged, the potential for axonal re-
covery is lost and pathological changes in the in-
jured axon continue until the axon undergoes
secondary axotomy (Maxwell et al., 1997; Gen-
narelli et al., 1998).
Calcium influx may also activate breakdown/
loss of microtubules (Maxwell and Graham, 1997;
280
Maxwell et al., 1999) since calcium both prevents
assembly of microtubules and causes rapid disas-
sembly of pre-existing polymers (Matsumura and
Hayashi, 1976; Langford, 1978). In addition to
calcium activating calpains (Buki et al., 1999), ac-
cumulation above normal axoplasmic levels may
activate, cystein proteases, spectrin proteases, cal-
cineurin and phospholipase A2 proteases. Further,
post-traumatic depolarization will activate
phospholipase C. Collectively these act to degrade
a wide range of cytoskeletal and membrane com-
ponents and eventually lead to cell damage or
death. More specifically, in axons after TAI, post-
traumatic elevated concentrations of Ca2+ within
the axoplasm leads to compaction (i.e., a reduction
in the spacing between neighboring) of neurofila-
ments. This has recently become a major patho-
logical marker of TAI (Pettus and Povlishock,
1996; Jafari et al., 1997, 1998; Povlishock et al.,
1997; Buki et al., 1999; Stone et al., 2001; Maxwell
et al., 2003). Experimental studies following a lat-
eral fluid percussion brain injury using 45Ca auto-
radiography have shown that Ca2+ accumulation
continues for several days, only returning to con-
trol levels by the 4th day after the injury (Hovda et
al., 1999; Samii et al., 1999). In certain situations,
it may require several seconds or longer for ion
concentrations in injured axons to rise above the
threshold required for initiation of the develop-
ment of pathology. In mTBI, the presence of such
a time course may explain why some athletes can
walk from the field and then collapse unconscious
on the sideline (Maroon et al., 2000). An injured
axon will, probably, attempt to restore membrane
potential and recover homeostasis. Membrane
pumps and transporters (e.g., NPP-ase, Na+/
K+-ATPase) and membrane pump Ca2+ATPase
may work overtime, consuming increased amounts
of ATP. To meet these elevated ATP requirements,
there is a marked up-regulation of cellular glycol-
ysis, which occurs within minutes after TBI. Cor-
related with this period of hyperglycolysis, there is
an increased synthesis of lactate and other toxic
products (i.e., free fatty acids) in cells of the brain
and CSF which leads to further damage to neu-
rons. In parallel with this, CBF is reduced to
oligaemic levels. The resultant mismatch between
oxygen, glucose delivery and consumption and the
physiological stress to which cells in the brain are
thereby exposed may make those cells much more
susceptible to secondary injury. Indeed, CBF may
remain depressed for several days after TBI,
thereby limiting the ability of the brain to respond
adequately to subsequent perturbations in energy
demand (Cantu, 2000).
Bergsneider et al. (2000) suggested that there are
three distinct periods of altered tissue/brain me-
tabolism following almost all types of animal ex-
perimental TBI — fluid percussion, cortical impact
and SDH — and head-injury in humans. In sum-
mary, the initial response is a total-body increase
in glucose utilization governed by pathophysio-
logic processes, which lead to the endogenous re-
lease of EAAs and adrenergic neurotransmitters
(Bergsneider et al., 2000), which are independent
of the functional state of neuronal activation and,
we suggest, may reflect responses to release of the
above by traumatized neurons. This total-body
hypermetabolic state, reflects the activation of en-
ergy-consuming, ATP-dependent, ion pumps
which attempt to redress abnormal ion fluxes (Ho-
vda, 1996; Hovda et al., 1999; Bergsneider et al.,
2000; Giza and Hovda, 2000). Secondly, related
regions of the brain undergo a period of reduced
glucose utilization, which, over the course of 4
days, return to normal values (Fineman et al.,
1993). Little is known of the processes affecting
glucose utilization during this second phase of
metabolic depression although it has been sug-
gested that Ca2+ sequestration by mitochondria
leads to opening of the permeability transition
pore of the inner mitochondrial membrane (Ok-
inkwo and Povlishock, 1999) and inhibition of
oxidative phosphorylation and energy transduct-
ion (Fineman et al., 1993; Bergsneider et al., 2000).
This leads to cells utilizing the anaerobic, glycoly-
tic pathway to synthesize high energy compounds
needed to fuel membrane pumps. Indeed, one ex-
perimental model utilizing cytochemical labeling
for activity of two ATP-dependent ionic pumps
has shown that there is a time course of at least 1 h
after injury for the loss of pump activity of both
p-NPPase and membrane pump Ca2+ -ATPase
and that this loss of activity extends for at least
24 h after injury (Maxwell et al., 1999). The cor-
related influx of Ca2+ into injured axons has been

suggested to stimulate both synthesis of glycogen
and reduce axonal transport of phosphofructoki-
nase leading to increased content of glycogen in
injured axons with 15 min of injury. Moreover,
there is loss of glycogen deposits in injured axons
between 7 and 14 days after injury. This loss,
it is suggested, may correlate with the reported
return to normal glucose utilization by 4 days
(Fineman et al., 1993) and recovery of metabolic
function (Pappius, 1981; Fineman et al., 1993;
Hovda, 1996) which showed that neurological
deficits persist both as long as there is metabolic
depression, and that the rate of recovery of be-
havioral function parallels the recovery of meta-
bolic function in the relevant regions of the brain.
However, at present, the mechanisms leading to
the third phase (recovery) are not well understood
although they do appear to be spontaneous
(Bergsneider et al., 2000).
Overall, these studies have demonstrated that
the brain is not metabolically quiescent or stag-
nant after TBI, but instead exhibits a metabolic
response that is both dynamic and stereotypic
(Fineman et al., 1993). Recent studies in head-in-
jured humans (Bergsneider et al., 2000) further
raise the possibility that the extent of the cerebral
metabolic rate of glucose (CMRglc) and oxygen
(CMRO2) depression after TBI is governed by the
severity of injury rather than the level of con-
sciousness. Hypothetically, it is possible that the
relative demand placed upon the glycolytic path-
way will vary with the severity of the injury, as a
function of CMRO2. At this time there is too little
evidence to be able to suggest that in mTBI pa-
tients the CMRglc may be a better indicator of
neurological status. For example, Bergsneider et
al. (2000) cite two head-injured patients with com-
parable CMRglc where one patient was ambula-
tory with a Glasgow Coma Score (GCS) of 15
while the other patient was comatose and had to
be mechanically ventilated. Nonetheless, it is sug-
gested that, with data obtained from a larger al-
iquot of patients, the CMRglc may prove to be a
more sensitive indicator of neurological status in
the mTBI patient.
Furthermore, evidence that alteration of glucose
metabolism is not restricted to the first hours fol-
lowing mTBI (Hovda, 1996; Hovda et al., 1999;
281
Giza and Hovda, 2000), does not normalize until
between 5 and 10 days after injury (Kawamata
et al., 1992) and may result in reduction of both
protein synthesis and oxidative phosphorylation.
The term ‘‘energy crisis’’ has been coined to refer
to this state of metabolic disequilibrium. The mag-
nitude of this ‘‘energy crisis’’ has important impli-
cations for cellular recovery and for increased
vulnerability to secondary insults. As a result cells
in the brain are more susceptible and are more
‘‘vulnerable’’ to damage or even death should the
patient experience a second insult of, even, lesser
intensity (Kawamata et al., 1992; Hovda, 1996;
Buki et al., 1999; Hovda et al., 1999; Giza and
Hovda, 2000).
Overall, the above findings suggest that loss of
consciousness after head trauma, the development
of secondary brain damage and the enhanced vul-
nerability of cells in the brain to a second insult
can be explained largely on the basis of abnormal
ionic fluxes, acute metabolic changes and loss of
autoregulation of CBF immediately after mTBI
(Maroon et al., 2000). However, following a trau-
matic insult to the head, other cells, perhaps the
most important of which are smooth muscle and
endothelial cells in the walls of the micro-
vasculature of the brain, are also damaged. This
results in changes in the extracellular milieu and
compromise of the ability of neurons to function
and initiate a number of processes that might re-
establish the normal pericellular environment (Ho-
vda et al., 1999). It is now evident that after mTBI
the stoichiometric relationship of metabolism and
the coupling to CBF is fundamentally compro-
mised. There is also evidence that such breakdown
is thought to be responsible for the degree and
extent of vulnerability following mTBI (Hovda et
al., 1999). However, the concept of injury-induced
vulnerability to even minor changes in CBF, in-
creased intracranial pressure and apnea (Hovda,
1996; Hovda et al., 1999; Giza and Hovda, 2000)
has risen from studies of patients severely enough
injured to have been admitted to hospital and,
most probably, in intensive care. That is a patient
that has most likely experienced moderate or se-
vere head-injury. A basic tenet of this review is
that the concept of injury-induced vulnerability
should also be a major concern in the management
282
of head-injured athletes who have experienced
mTBI but are not hospitalized. However, there is
now an increasing acceptance that injured axons
(1) undergo or demonstrate a wide spectrum of
physiopathological responses with a time course
extending to at least several hours, (2) may be
physiologically and functionally stressed after in-
jury and (3) as a result, may be more susceptible to
any second, even a milder, insult for at least sev-
eral tens of hours. Therefore, after any type of
head-injury, which may potentiate the develop-
ment of pathology within axons and their cell
bodies of the injured sportsman, should be advised
not to resume play in the same game. But rather,
those sportsmen should be encouraged to rest.
This rest could, we suggest, provide mildly injured
neurons and other cells the opportunity to respond
to post-traumatic increased demands for energy
and would allow re-establishment of normal chem-
ical and ionic cellular peri- and intracellular envi-
ronments.
Biochemical markers of mTBI
During the past decade neurobiochemical markers
of brain damage have gained increasing interest in
experimental and clinical neurotraumatology. Pro-
teins synthesized in astrogial cells or neurons have
been proposed as markers of cell damage in the
central nervous system. The most common include
fibrillary acid protein, myelin basic protein, crea-
tine kinase isoenzyme BB, neuron specific enolase
(NSE) and protein serum-100b. Increased levels of
these proteins in the CSF and serum have been
observed in patients with various neurological dis-
eases (Skogseid et al., 1992; Ingebrigsten, 1995;
Ingebrigsten et al., 1995, 1997; Yamazaki et al.,
1995; Ross et al., 1996; Woertgen et al., 1997;
McKeating et al., 1998; Raabe et al., 1998; Her-
mann et al., 1999, 2000, 2001; De Kruijk et al.,
2001; Ringger et al., 2004). Structural proteins of
astroglial (glial fibrillary acidic protein) or neu-
ronal (neurofilament protein) brain tissue have
mostly been used in experimental settings. Alter-
natively, and with a much greater applicability in
the clinical environment, protein serum S-100b and
NSE are two markers which, due to present high
sensitivity of commercially available detection kits
(Sangtec 100s, AB Sangtec Medical, S-161 02
Bromma, Sweden) to analyze NSE and S-100b
concentrations in serum samples, have recently at-
tracted growing attention in clinical research and
diagnostic practice. Both proteins are now ac-
cepted as specific neurobiochemical markers of
brain damage in clinical and experimental ischemic
brain infarction and TBI (Moore, 1965; Skogseid
et al., 1992; Ingebrigsten, 1995; Ingebrigsten
et al., 1995, 1997, 1999, 2000; Yamazaki et al.,
1995; Ross et al., 1996; Woertgen et al., 1997;
McKeating et al., 1998; Raabe et al., 1998;
Hermann et al., 1999, 2000, 2001; Romner et al.,
2000; Biberthaler et al., 2001; De Kruijk et al.,
2001; Pleines et al., 2001).
The serum S-100b protein (S-100b) is a member
of a large family of Ca2+-binding proteins, the
cellular synthesis of which has been localized pre-
dominantly in astroglia and Schwann cells. When
Moore discovered the S-100 protein, he suggested
that it was located only in brain tissue (Moore,
1965). Later studies, however, demonstrated its
presence in the body in different forms, and that
the b form predominates in the brain (Moore,
1965; Ingebrigsten et al., 1995, 1997, 1999, 2000;
Woertgen et al., 1997; Ingebrigsten, 1998; McK-
eating et al., 1998; Raabe et al., 1998; Hermann et
al., 1999, 2001; Romner et al., 2000; Biberthaler et
al., 2001; Pleines et al., 2001). Increased serum S-
100b protein concentrations in peripheral blood
are considered as a marker for dysfunction of the
blood-brain barrier and therefore, protein S-100b
release into peripheral blood may indicate func-
tional brain dysfunction without the necessary
presence of a pathology visible through use of CT
imaging (Moore, 1965; Ingebrigsten et al., 1995,
1997, 1999, 2000; Woertgen et al., 1997; Ingebrigs-
ten, 1998; McKeating et al., 1998; Raabe et al.,
1998; Hermann et al., 1999, 2001; Romner et al.,
2000; Biberthaler et al., 2001; Pleines et al., 2001).
Little is known regarding the mechanism by which
S-100b passes through the blood-brain barrier and
enters the blood. However, it may be posited that
where there is loss of autoregulation and stripping
of endothelium, as has, for example been noted in
ischemia or severe TBI (Maxwell, 1999; Maxwell
et al., 1992) that damage to the microvasculature

of the brain that has been described in DAI
may provide a pathway for passage of S-100b from
injured glial cells into blood. The detection of
protein S-100b in peripheral blood may therefore
reflect either an ongoing pathophysiological
cascade that leads to secondary brain damage or
result from necrosis (apoptosis) of previously
damaged cells (Hermann et al., 2000).
However, a recent study showed that the sensi-
tivity of S-100b measurement for the identification
of patients with a pathological CT scan was high
(90%), but the specificity was only 65% with a
cutoff level of 0.5 ng/ml (Ingebrigtsen et al., 2000).
Biberthaler et al. (2001) found an even higher sen-
sitivity of S-100b measurement (100%) than that
of Ingebrigtsen et al., but that the specificity was
reduced (40.5%). Therefore, measurement of this
parameter may be helpful as an additional screen-
ing tool to identify risk groups in a cohort of
mTBI patients. Biberthaler et al. (2001) suggested
using a cutoff point at a concentration of 0.1 ng/ml
plasma since this concentration was the lowest
obtained in the serum of patients with mTBI but
lacking signs of intracerebral injury on cranial
computed tomography.
Further, it has recently been demonstrated that
the early post-traumatic elevated serum levels of S-
100b declined rapidly during the first 6 h after the
trauma. Long lasting post-traumatic elevated S-
100b values probably indicate either an ongoing
pathophysiological cascade that leads to second-
ary brain damage or the egress of S-100b either as
a result of necrosis or apoptosis of previously
damaged cells (Romner et al., 2000) or focal dam-
age to the microvasculature. Further, it has re-
cently been suggested that lack of detectable
concentrations of S-100b protein in serum pre-
dicts a normal CT scan provided the blood sample
is collected within the first 2 or 3 h after injury
(Ingebrigtsen et al., 2000). A comparative analysis
of the predictive value of the neurological status,
CT data, and NSE and S-100b serum concentra-
tions has shown that the initial (i.e., within 2–3 h of
injury) serum protein concentration of S-100b is
the best predictor of outcome for long-term ne-
uropsychological disorders (Herrmann et al.,
2001). Advanced MRI techniques, such as quan-
titative MRI analysis or diffusion weighted or
283
magnetization transfer imaging, allow for the de-
tection of diffuse white matter brain damage or
axonal brain injury with a greater sensitivity than
conventional CT procedures. Correspondingly
these probably result in a better prediction of out-
come to the patient (Ingebrigsten et al., 1999).
NSE is an isoenzyme of enolase and is predom-
inantly found in the cytoplasm of neurons, in ne-
uroendocrine cells, smooth muscle fibers and
adipose tissue (Skogseid et al., 1992; Yamazaki et
al., 1995; Ross et al., 1996; McKeating et al., 1998;
Hermann et al., 1999, 2000, 2001; De Kruijk et al.,
2001; Pleines et al., 2001). The cytoplasmic enzyme
is liberated by cell destruction. It has been sug-
gested, therefore, that increased levels of NSE in
CSF and peripheral blood indicate neuronal dam-
age. Herrmann et al. (1999, 2000) demonstrated an
association between the release of both NSE and
S-100b protein and the severity of a TBI. Notably,
patients with mTBI showed a different temporal
profile of NSE in their serum to moderate or se-
vere TBI patients (Hermann et al., 2000). A sig-
nificant difference for the concentration of NSE in
serum from mild and moderate to severe head in-
jury patients was only obtained up to 24 h after
injury. Thereafter serum values for either mTBI
or moderate to severe injury patients were not el-
evated and did not differ from control values
(Hermann et al., 2000). The release patterns of
NSE and S-100b is thought to reflect different
types and dynamics of underlying pathophysiolo-
gy between moderate to severe and mild TBI pa-
tients. Overall, patients with moderate to severe
head injury show a significantly higher and longer
release of both markers (Herrmann et al., 2000).
In addition, recent studies have shown a corre-
lation between the clinical outcome and CSF or
serum concentrations of protein S-100b and NSE
in patients with severe TBI (Skogseid et al., 1992;
Ross et al., 1996; Woertgen et al., 1997). Raabe
et al. (1998) showed that excessive secondary in-
crease in S-100b serum concentration relates
to severe brain damage associated with a fatal out-
come. The same group reported a significant cor-
relation between early S-100b values and
the volume of cerebral contusions in patients with
severe TBI but no association between NSE con-
centration and contusion volume. However, there
284
is still some controversy concerning the clinical
significance of such data. For example, Skogseid
et al. (1992) found a significant correlation between
maximum NSE concentration and contusion vol-
ume. Alternatively, Woertgen et al. (1997) found
no correlation of either NSE or S-100b and the
severity of intracranial pathology as defined by the
Traumatic Coma Data Bank (TCDB) CT classifi-
cation. Further, Hermann et al. (1999, 2000)
showed that the severity of TBI is associated (1)
with early post-traumatic release of both protein S-
100b and NSE and (2) that the early kinetics of
neurobiochemical markers of brain damage after
TBI do reflect intracranial pathology as demon-
strated in cranial CT (Hermann et al., 1999, 2000).
More recently, Pleines et al. have indicated that the
elevation of S-100b and NSE in CSF depends on
the extent of the injury and that S-100bmay be a
predictor of outcome after TBI, whereas NSE is a
better indicator of inflammatory responses (Pleines
et al., 2001). In this latter study, release patterns of
both proteins did not differ between patients with
or without signs of focal neurological deficits.
Herrmann et al. (2001) found that post-trau-
matic serum concentrations of NSE and S-100b
not only reflected the overall severity of brain
trauma, as defined by GCS scores, but also subtle
neuropsychological dysfunction. Neuropsycholog-
ical disorders, identified 2 weeks after injury, were
correlated with a significantly higher and longer
lasting release of both brain proteins. Patients with
neuropsychological disorders at 6 months after
head trauma, however, also exhibited significantly
higher NSE and S-100b serum concentrations dur-
ing the first 3 days after TBI (Herrmann et al.,
2001).
De Kruijk et al. (2001) investigated whether se-
rum concentrations of NSE and S-100b in a group
of patients with clearly defined mTBI were higher
than in serum of healthy controls. The mean serum
concentrations of NSE did not provide a sensitive
marker between mTBI patients and controls. On
the contrary, the median serum S-100b concentra-
tion from 104 mTBI patients was significantly
elevated compared with control, healthy patients
(n ¼ 92). Neither did trauma to other parts of the
body induce an elevation of either S-100b or NSE.
Thus, current opinion is that elevated serum
marker concentrations of S-100b are a conse-
quence of brain injury including mTBI (De Kruijk
et al., 2001).
A major conclusion drawn by this review is that
determination of S-100b serum levels within 2–3 h
of head-injury may prove to be a useful indicator
for damage to the brain after mTBI. But, impor-
tantly, serum levels of S-100b also rise after frac-
ture of bones, thoracic contusion, burns and even
minor bruising (Anderson et al., 2001). On the
other hand, increased serum NSE levels do not
seem to be a specific sensitive marker for mTBI
(De Kruijk et al., 2001). Given that S-100b is a
marker of damage to glial and/or Schwann cells
and NSE indicates damage to neurons, elevated
plasma levels of S-100b after mTBI is a good in-
dicator that damage which results in release of S-
100b is more likely to be localized in the white than
the grey matter of the brain. This conclusion is
supported by the literature related to the effects of
rotational and acceleration/deceleration injuries to
the brain and allows development of the hypoth-
esis that TAI is an important pathophysiological
cellular response during and following mTBI.
However, since S-100b is released by glia than di-
rectly from injured neurons it is implicit that the
time course of release of detectable levels of S-100b
in CSF is longer than that of a direct marker of
neuronal injury. Siman et al. (2004) have recently
reported, both in vitro and in vivo, elevated levels
of a-spectrin and calpain/caspase mediated amino-
and carboxy-terminal breakdown products
thereof. Moreover, 150-kDa a-spectrin fragments
and proteins 14-3-z and 14-3-b were released by 6 h
after both mild, moderate TBI and transient global
ischemia in rat (Siman et al., 2004). Although a-
spectrin is widely distributed in many types of cell,
spectrin breakdown products have been reported
early after TAI in the subaxolemma region (Buki
et al., 1999), throughout the axoplasm within 2 h
of injury, and has a biphasic presentation with
a second peak at 4–5 days after TBI in mouse
(Saatman et al., 2003). In humans peak levels oc-
cur in CSF at 2–3 days after severe TBI (Farkas et
al., 2005) and may fall over 4 days in patients with
a good outcome (D’Avella et al., 2004). Thus,
there is still a major requirement to define a
marker of injury to neurons in mTBI.

Conclusions
In mTBI
1. Primary axotomy, or shearing of axons,
does not occur.
2. There is no support for the hypothesis that
there is a rapid post-traumatic proteolysis of
the entire axonal cytoskeleton in a signifi-
cant number of injured axons except in the
most severe cases of diffuse head injury.
3. mTBI may produce only a temporary elect-
rochemical disturbance of brain function
caused by neuronal, chemical, and neuro-
electrical changes and without development
of overt pathology.
4. The great majority of injured axons in cases
of mTBI, concussion and more severe levels
of injury enter a ‘‘pathological cascade’’ of
events that culminate in ‘‘secondary axo-
tomy’’ at least 4 h and extend over many
days after the initial trauma.
5. The use of modern, sensitive, techniques for
analysis of plasma components (e.g., plasma
levels of S-100b and NSE) allows demon-
stration of injury to axons in humans even
after mTBI and detection of damaged axons
at earlier post-traumatic intervals than have
been possible until very recently.
6. The time course of formation of axonal
swellings is longer in humans than in the
majority of animal experimental models —
there is no evidence for recruitment at present
— it is just that reactive axons may be dem-
onstrated up until 99 days after injury.
7. There is now a clear consensus that axonal
changes leading to the development of pa-
thology are initiated as a result of damage to
the axolemma. This damage leads to disrup-
tion of ionic homeostatic mechanisms and/or
a changed permeability of the axolemma.
8. Responses by microtubules and neurofila-
ments differ in their time course both within
a single axon, between axons in different
tracts within the CNS and between species.
The time course for loss of microtubules,
compaction of neurofilaments and second-
ary axotomy is longest in humans.
285
9. Recent evidence has suggested that not all
injured axons proceed to secondary axo-
tomy but that some demonstrate long-term
compaction of neurofilaments within their
cytoskeleton.
10. There is increasing evidence, from both
experimental, animal and clinical human
studies that a mild or severe sudden accel-
eration–deceleration head trauma may re-
sult in cells of the brain assuming a
vulnerable state for an unknown period.
The fact that cells of the brain are in such a
vulnerable state may predispose the affected
athlete to an increased risk of a more serious
and dangerous head injury if the athlete is
exposed to a second head injury, for exam-
ple a Second Impact Syndrome, if allowed
to return to the field of play during the same
game.
There are many Return-to-play Guidelines to
help sport-physicians decide upon the initial man-
agement and determine when an injured athlete
should return to play. However, considerable con-
troversy exists and none of these guidelines has
been developed on the basis of specific, scientific
knowledge regarding the processes of either white
matter injury or the recovery thereof. These guide-
lines rely first of all upon self-reported symptoms.
If the athlete is confused, the validity of the former
has to be questioned. Generally, most current
guidelines may allow the athlete to return to play
after mTBI, even by the deepest grade I, in the
same competition, if his/her clinical examination
results are apparently normal at rest and with ex-
ertion. It is acknowledged that experts agree that
no athlete who is still symptomatic should be al-
lowed to return to competition and that ongoing
and repeated examination should be conducted
after the injury.
However, the present review provides strong
evidence that cells of the brain after mTBI may
remain alive but in a vulnerable state for an, as yet,
undetermined period. It is a major conclusion of
the present study that the brain of the injured
athlete may be in a vulnerable state, which may
predispose the athlete to an increased risk of se-
rious injury to the brain from a second head injury
286
(i.e., Second Impact Syndrome). In addition, since
brain responses may be blunted resulting in a
longer reaction time to changes in the environ-
ment, the injured athlete may be at increased risk
of serious injury both to the brain and other parts
of the body.
Therefore it is recommended that any con-
fused player, with or without amnesia, after
receiving either a blow or any significant ac-
celeration–deceleration force to the head, a
grade I mTBI lesion, should be taken off the
ice and should not be permitted to return to
play for at least 72 h.
Acknowledgements
We would like to express our deep appreciation to
Prof. Patrick Bishop, Department of Kinesiology,
University of Waterloo, Canada, for helping anal-
yzing the video mechanisms of mTBI by the ice
hockey players. We would also like to thank Mr.
Yelverton Tegner, MD, Associated Professor at
the Institution of Health Sciences at the Luleå
University of Technology, SE 961 36 Boden, Swe-
den, and Mr. Reto Agosti, MD, Swiss Specialist of
Neurology, Director of the Headache Center
Clinic Hirslanden, Zurich, Switzerland, who as-
sisted us with suggestions and input to further im-
prove the quality of this manuscript. Special
thanks go to my resident Mr. Stephan Wirth,
MD, Orthopedic University Hospital Zurich-Bal-
grist, Zurich, Switzerland, who helped me editing
the text and the figures.
References
Abdel-Dayem, H.M., Abu-Judeh, H., Kumar, M., et al. (1998)
SPECT brain perfusion abnormalities in mild or moderate
traumatic brain injury. Clin. Nucl. Med., 23(5): 309–317.
Adams, J.H., Doyle, D., Ford, I., et al. (1989) Diffuse axonal
injury in head injury: definition, diagnosis and grading. His-
topathology, 15: 49–59.
Adams, J.H., Graham, D.I., Murray, L.S., et al. (1982) Diffuse
axonal injury due to non-missile head in humans: an analysis
of 45 cases. Ann. Neurol., 12: 557–563.
Anderson, R.E., Hansson, L.O., Nilsson, O., et al. (2001) High
serum S100B levels for trauma patients without head injuries.
Neurosurgery, 48: 1255–1260.
Bailes, J.E. and Cantu, R.C. (2001) Head injury in athletes.
Neurosurgery, 48(1): 26–42.
Bergsneider, M., Hovda, D.A., Lee, S.M., et al. (2000) Disso-
ciation of cerebral glucose metabolism and level of con-
sciousness during the period of metabolic depression
following human traumatic brain injury. J. Neurotrauma,
17(5): 389–401.
Biasca, N. (2000) Review and consequences of the typical ice
hockey related injuries in the Swiss Ice Hockey Association.
(SIHA) 1996–1998. In: Biasca N., Montag W.D. and Gerber
Ch. (Eds.), Safety in Ice Hockey, Eighth International Sym-
posium of the International Ice Hockey federation IIHF. In-
ternational Ice Hockey Federation IIHF, Zurich, pp. 36–52.
Biasca, N., Battaglia, H. and Simmen, H.P. (1995) Analyse
typischer Eishockey-Verletzungen. SUVA Monographie Nr.
6, Schweizerische Unfallversicherungsanstalt SUVA, Abtei-
lung Unfallmedizin, CH-6002 Luzern, Switzerland.
Biasca, N., Battaglia, H. and Tegner, Y. (1999) Diagnose und
Behandlungsvorschläge der Commotio cerebri bei Kontakt-
und Kampf-sportarten: beispiel im Eishockey. Schweiz.
Ärtzezeit., 80(10): 595–598.
Biasca, N. and Tegner, Y. (2001) International Sports Injury
System — ISIS. Sports Medicine and Hockey; A summit for
the NHL and Beyond, American Orthopedic Society for
Sports Medicine (AOSSM) and National Hockey League
Team Physician Society (NHLTPS), Hilton Toronto, Can-
ada, August 24–26, pp. 15–36.
Biasca, N., Wirth, S. and Tegner, Y. (2002) The avoidability of
head and neck injuries in ice hockey: a historical review. Br.
J. Sports Med., 36: 410–427.
Biberthaler, P., Mussack, T., Wiedemann, E., et al. (2001)
Evaluation of S-100b as a specific marker for neuronal dam-
age due to minor head trauma. World J. Surg., 25: 93–97.
Bigler, E.D. (1999) Neuroimaging in pediatric traumatic head
injury: diagnostic considerations and relationships to ne-
urobehavioral outcome. J. Head Trauma Rehabil., 14(4):
406–423.
Bishop, P.J. (1997) Impact performance characteristics of
hockey helmets with liners of differing thickness. In: Ashare
A. (Ed.), Third International Symposium on Safety in Ice
Hockey. American Society for Testing and Materials, Phil-
adelphia, pp. 112–117.
Bishop, P.J. (2000) Biomechanics of sports protective equip-
ment in ice hockey to prevent injuries. In: Biasca N., Montag
W.D. and Gerber Ch. (Eds.), Safety in Ice Hockey, Eight
International Symposium of the International Ice Hockey
federation IIHF. International Ice Hockey Federation IIHF,
Zurich, pp. 36–52.
Bishop, P.J. (2001) Protective equipment standards as a means
of reducing injury risk. Sports Medicine and Hockey; A
summit for the NHL and Beyond, American Orthopedic

Society for Sports Medicine (AOSSM) and National Hockey
League Team Physician Society (NHLTPS), Hilton Toronto,
Canada, August 24–26, p. 57.
Bishop, P.J. and Arnold, J. (1993) The effectiveness of hockey
helmets in limiting localized loading on the head. In: Castaldi
C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice
Hockey: Second Volume, ASTM STP 1212. American Soci-
ety for Testing and Materials, Philadelphia, pp. 175–182.
Blumbergs, P.C. (1997) Pathology, Head Injury. In: Reilly P.
and Bullock R. (Eds.). Chapman and Hall, London, pp.
39–70.
Blumbergs, P.C., Scott, G., Manavis, J., et al. (1994) Staining of
amyloid precursor protein to study axonal damage in mild
head injury. Lancet, 344: 1055–1056.
Blumbergs, P.C., Scott, G., Manavis, J., et al. (1995) Topog-
raphy of axonal injury as defined by amyloid precursor pro-
tein and the sector scoring method in mild and severe closed
head injury. J. Neurotrauma, 12: 565–572.
Brooks, W.M., Frieman, S.D. and Gasparovic, C. (2001) Mag-
netic resonance spectroscopy in traumatic brain injury. J.
Head Trauma Rehabil., 16(2): 149–164.
Buki, A., Siman, R., Trojanowski, J.Q., et al. (1999) The role
of calpain-mediated spectrin proteolysis in traumatically in-
duced axonal injury. J. Neuropathol. Exp. Neurol., 58:
365–375.
Burke, C.J. (2001) Definition, incidence and initial assessment
of concussion. Sports Medicine and Hockey; A summit for
the NHL and Beyond, American Orthopedic Society for
Sports Medicine (August 24–26) (AOSSM) and National
Hockey League Team Physician Society (NHLTPS), Hilton
Toronto, Canada, pp. 107–108.
Cantu, R. (1992) Cerebral concussion in sport: management
and prevention. Sports Med., 14: 64–74.
Castaldi, C.R. (1993) The sports mouthguard: its use and mis-
use in ice hockey. In: Castaldi C.R., Bishop P.J. and Hoerner
E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM
STP 1212. American Society for Testing and Materials, Phil-
adelphia, pp. 164–174.
Cecil, K.M., Hills, E.C., Sandel, M.E., et al. (1998) Proton
magnetic resonance spectroscopy for detection of axonal in-
jury in the splenium of the corpus callosum of brain-injured
patients. J. Neurosurg., 88: 795–801.
Christman, C.W., Grady, M.S., Walker, S.A., et al. (1994) Ul-
trastructural studies of diffuse axonal injury in humans. J.
Neurotrauma, 11: 173–186.
Clark, J.M. (1974) Distribution of microglial clusters in the
brain after head trauma injury. J. Neurol. Neurosurg. Psy-
chiatry, 37: 463–474.
Clayton, P. (1997) A review of head injury data in Canadian
hockey league with an emphasis on the incidence of concus-
sion. In: Ashare A. (Ed.), Third International Symposium on
Safety in Ice Hockey. American Society for Testing and Ma-
terials, Philadelphia Abstract.
D’Avella, D., Aguennouz, M., Angileri, F.F., et al. (2004) De-
tection of a-II spectrin and breakdown products in humans
after severe traumatic brain injury. Acad. Emerg. Med., 11:
515–520.
287
De Kruijk, J.R., Leffers, P., Menheere, P.P., et al. (2001) S-
100B and neuron-specific enolase in serum of mild traumatic
brain injury patients. Acta Neurol. Scand., 103: 175–179.
Denny, B. and Russel, W.R. (1941) Experimental cerebral con-
cussion. Brain, 64: 93–164.
Dick, R.W. (1997) Injury in collegiate ice hockey with an em-
phasis on injuries to the head. In: Ashare A. (Ed.), Third
International Symposium on Safety in Ice Hockey. American
Society for Testing and Materials, Philadelphia Abstract.
Dixon, J.L. and Brodie, I.K.R. (1993) The new ISO Standards
for ice hockey helmets and face protectors: moving toward
International Standards harmonization and conformity. In:
Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in
Ice Hockey: Second Volume, ASTM STP 1212. American
Society for Testing and Materials, Philadelphia, pp. 192–213.
Elson, L.M. and Ward, C.C. (1994) Mechanisms and patho-
physiology of mild head injury. Semin. Neurol., 14(1): 8–18.
Erb, D.E. and Povlishock, J.T. (1988) Axonal damage in severe
traumatic brain injury: an experimental study in the cat. Acta
Neuropathol., 76: 347–358.
Farkas, O., Polgar, B., Szekeres-Bartho, J., et al. (2005) Spec-
trin breakdown products in the cerebrospinal fluid in severe
head injury — preliminary observations. Acta Neurochir.,
147: 855–861.
Fineman, I., Hovda, D.A., Smith, M., et al. (1993) Concussive
brain injury is associated with a prolonged accumulation of
calcium: a 45Ca autoradiographic study. Brain Res., 624:
94–102.
Galbraith, J.A., Thibault, L.E. and Matheson, R.A. (1993)
Mechanical and electrical responses of the squid giant axon
to simple elongation. J. Biomech. Eng., 115: 13–22.
Garnett, M.R., Blamire, A.M., Corkill, R.G., et al. (2000) Early
proton magnetic resonance spectroscopy in normal-appear-
ing brain correlates with outcome in patients following trau-
matic brain injury. Brain, 123: 2046–2054.
Gennarelli, T.A. (1991) Cerebral concussion and diffuse brain
injuries. In: Torg J.S. (Ed.), Athletic Injuries to the Head,
Neck and Face (2nd ed.). Mosby Year Book, St. Louis, MO,
pp. 270–282.
Gennarelli, T.A. and Thibault, L.E. (1982) Biomechanics of
acute subdural hematoma. J. Trauma, 22: 680–686.
Gennarelli, T.A., Thibault, L.E., Adams, J.H., et al. (1982)
Diffuse axonal injury and traumatic coma in the primate.
Ann. Neurol., 12: 564–574.
Gennarelli, T.A., Thibault, L.E. and Graham, D.I. (1998) Dif-
fuse axonal injury: an important form of traumatic brain
damage. Neuroscientist, 4(3): 202–215.
Gennarelli, T.A., Thibault, L.E., Tipperman, R., et al. (1989)
Axonal injury in the optic nerve: a model simulating diffuse
axonal injury in man. J. Neurosurg., 71: 244–253.
Gentleman, S.M., Roberts, G.W., Gennarelli, T.A., et al. (1995)
Axonal injury: a universal consequence of fatal closed head
injury? Acta Neuropathol., 89: 537–543.
Giza, C.C. and Hovda, D.A. (2000) Ionic and metabolic con-
sequences of concussion. In: Cantu R.C. (Ed.), Neurologic
Athletic Head and Spine Injuries. WB Saunders Company,
New York, pp. 80–100.
288
Grady, M.S., Mclaughlin, M.R., Christman, C.W., et al. (1993)
The use of antibodies targeted against the neurofilaments
subunits for the detection of diffuse axonal injury in humans.
J. Neuropathol. Exp. Neurol., 52: 143–152.
Graham, D.I. and Gennarelli, T.A. (1997) Trauma. In: Graham
D.I. and Lantos P.L. (Eds.), Chapter 5 in Greenfield’s Ne-
uropathology (6th ed.). Arnold, London, pp. 197–262.
Graham, D.I., Mcintosh, T.K., Maxwell, W.L., et al. (2000)
Recent advances in neurotrauma. J. Neuropathol. Exp. Ne-
urol., 59(8): 641–651.
Gross, H., Kling, A., Henry, G., et al. (1996) Local cerebral
glucose metabolism in patients with long-term behavioral
and cognitive deficits following mild traumatic brain injury.
J. Neuropsychiatry Clin. Neurosci., 8(3): 324–334.
Hartmann, T. (2001) Cholesterol, Ab and Alzheimer’s disease.
TINS, 24: S45–S48.
Hermann, M., Curio, N., Jost, S.T., et al. (1999) Protein S-100B
and neuron specific enolase as early neurobiochemical mark-
ers of the severity of traumatic brain injury. Restor. Neurol.
Neurosci., 14: 109–114.
Herrmann, M., Curio, N., Jost, S., et al. (2001) Release of bi-
ochemical markers of damage to neuronal and glial brain
tissue is associated with short and long term neuropsycho-
logical outcome after traumatic brain injury. J. Neurol. Ne-
urosurg. Psychiatry, 70(1): 95–100.
Hermann, M., Jost, S.T., Kutz, S., et al. (2000) Temporal pro-
file of release of neurobiochemical markers of brain damage
after traumatic brain injury is associated with intracranial
pathology as demonstrated in cranial computerized tomo-
graphy. J. Neurotrauma, 17(2): 113–122.
Holbourn, A.H.S. (1943) Mechanics of head injury. Lancet, 2:
438–441.
Hovda, D.A. (1996) Metabolic dysfunction. In: Narayan R.K.,
Wiberger J.E. and Povlishock J.T. (Eds.), Neurotrauma.
McGraw-Hill, Inc, New York, pp. 1459–1478.
Hovda, D.A., et al. (1994) Long-term changes in metabolic
rates for glucose following mild, moderate and severe con-
cussive head injuries in adult rats. Soc. Neurosci. Abstract,
20: 845.
Hovda, D.A., Prins, M., Becker, D.P., et al. (1999) Neurobi-
ology of concussion. In: Bailes J.E., Lovell M.R. and Ma-
roon J.C. (Eds.), Sports-related Concussion. Quality Medical
Publishing, Inc., Pittsburgh, PA, pp. 13–51.
Ingebrigsten, T. (1998) Aspect on the management of minor
head injury. Habilitation Work, Department of neurosur-
gery, Institute of clinical medicine. University of Tromso,
Norway.
Ingebrigsten, T., Kongstad, B.R.P. and Langbakk, B. (1995)
Increased serum concentrations of protein S-100 after minor
head injury: a biochemical serum marker with prognostic
value? J. Neurol. Neurosurg. Psychiatry, 59: 103–104.
Ingebrigtsen, T., Romner, B., Marup-Jensen, S., et al. (2000)
The clinical value of serum S-100 protein measurements in
minor head injury: a Scandinavian multicentre study. Brain
Inj., 14(12): 1047–1055.
Ingebrigsten, T., Rommer, B. and Trumpy, J.H. (1997) Man-
agement of minor head injury: the value of early computed
tomography and serum proteion S-100 measurements. J.
Clin. Neurosci., 4: 29–33.
Ingebrigtsen, T., Waterloo, K., Jacobsen, E.A., et al. (1999)
Traumatic brain damage in minor head injury: relation of
serum S-100 protein measurements to magnetic resonance
imaging and neurobehavioral outcome. Neurosurgery, 45:
468–476.
Jafari, S.S., Maxwell, W.L., Neilson, M., et al. (1997) Axonal
cytoskeletal changes after non-disruptive axonal injury. J.
Neurocytol., 26: 207–221.
Jafari, S.S., Neilson, M., Graham, D.I., et al. (1998) Axonal
cytoskeletal changes after non-disruptive axonal injury. II.
Intermediate sized axons. J. Neurotrauma, 15: 955–966.
Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A.
(1990) Massive increases in extracellular potassium and the
indiscriminate release of glutamate following concussive
brain injury. J. Neurosurg., 73: 889–900.
Kawamata, T., Katayama, Y., Hovda, D.A., et al. (1992) Ad-
ministration of excitatory amino acid antagonists via micro-
dialysis attenuated the increased in glucose utilization seen
following concussive brain injury. J. Cereb. Blood Flow Me-
tab., 12: 12–24.
Kelly, J.P. (1999) Traumatic brain injury and concussion in
sports. JAMA, 282: 989–991.
Kelly, J.P., Nichols, J.S., et al. (1991) Concussions in sports:
guidelines for the prevention of catastrophic outcome.
JAMA, 266: 2867–2869.
Klatky, R., Furuya, Y., Valadka, A.B., Goodman, J.C. and
Robertson, C.S. (2002) Comparison of microdialysate argi-
nine and glutamate levels in severely head-injured patient.
Acta Neurochir. Suppl., 81: 347–349.
Langford, G.M. (1978) In vitro assembly of dogfish brain tub-
ulin and induction of coiled ribbon polymers by calcium.
Exp. Cell Res., 111: 139–151.
Laprade, R.F. (2001) Ice hockey injuries at the collegiate level.
Sports Medicine and Hockey; A summit for the NHL and
Beyond, American Orthopedic Society for Sports Medicine
(AOSSM) and National Hockey League Team Physician So-
ciety (NHLTPS), Hilton Toronto, Canada, August 24–26,
pp. 37–45.
Levine, B., Cabezy, R., Mcintosh, A.R., et al. (2002) Functional
reorganization of memory after traumatic brain injury: a
study with H(2)(15)O positron emission tomography. J. Ne-
urol. Neurosurg. Psychiatry, 73(2): 173–181.
Margulies, S.S. and Thibault, L.E. (1989) An analytical model
of traumatic diffuse brain injury. J. Biomech. Eng., 111:
241–249.
Maroon, J.C., Lovell, M.R., Norwig, J., et al. (2000) Cerebral
concussion in athletes: evaluation and neuropsychological
testing. Neurosurgery, 47(3): 659–672.
Matsumura, F. and Hayashi, M. (1976) Polymorphism of tub-
ulin assembly. In vitro formation of sheet, twisted ribbon and
microtubule. Biochim. Biophys. Acta, 453: 162–175.
Matsushita, Y., Shima, K., Katoh, H. and Nawashiro, H.
(2000) The synergistic effect of high K+ and hypoxia on
extracellular concentrations of neuroactive amino acid in
hippocampus. Acta Neurochir. Suppl., 76: 247–249.

Maxwell, W.L. (1999) Cellular responses to ischaemic CNS in-
jury. In: Berry M. and Logan A. (Eds.), Pharmacological
strategies in the treatment of CNS injuries. CRC Press, Baton
Rouge, pp. 19–41.
Maxwell, W.L., Domleo, A., McColl, G., Jafari, S.S. and Gra-
ham, D.I. (2003) Post-acute alterations in the axonal cyto-
skeleton after traumatic axonal injury. J. Neurotrauma, 20:
151–168.
Maxwell, W.L. and Graham, D.I. (1997) Loss of axonal mi-
crotubules and neurofilaments after stretch-injury to guinea
pig optic nerve fibers. J. Neurotrauma, 14: 603–614.
Maxwell, W.L., Kosanlavit, R., McCreath, B.J., et al. (1999)
Freeze-fracture and cytochemical evidence for structural and
functional alteration in the axolemma and myelin sheath of
adult guinea pig optic nerve fibers after stretch injury. J. Ne-
urotrauma, 16: 273–284.
Maxwell, W.L., Mccreath, B.J., Graham, D.I., et al. (1995)
Cytochemical evidence for redistribution of membrane pump
calcium-ATPase and ecto-Ca-ATPase activity, and calcium
influx in myelinated nerve fibers of the optic nerve after
stretch injury. J. Neurocytol., 24: 925–942.
Maxwell, W.L., Povlishock, J.T. and Graham, D.I. (1997) A
mechanistic analysis of nondisruptive axonal injury: a review.
J. Neurotrauma, 14(7): 419–440.
Maxwell, W.L., Whitfield, P., Suzen, B., et al. (1992) The
cerebrovascular response to experimental lateral head accel-
eration. Acta Neuropathol., 84: 289–296.
McGehee, D.V. (1996) Head injury in motor vehicle crashes:
human factors, effects, and prevention. In: Rizzo M. and
Tranel D. (Eds.), Head Injury and Postconcussive Syndrome.
Churchill Livingstone, London and New York, pp. 57–69.
McIntosh, T.K., Faden, A.I., Bendall, M.R. and Vink, R.
(1987) Traumatic brain injury in the rat: alterations in brain
lactate and ph as characterized by H and P nuclear magnetic
resonance. J. Neurochem., 49: 1530–1540.
McIntosh, T.K., Faden, A.I., Yamakami, I. and Vink, R.
(1988) Magnesium deficiency exacerbates and pretreatment
improves outcome following traumatic brain injury in the rat.
J. Neurotrauma, 5: 17–31.
McKeating, E.G., Andrew, P.J. and Mascia, L. (1998) Rela-
tionship of neuron specific enolase and protein S-100 con-
centrations in systemic and jugular venous serum to injury
severity and outcome after traumatic brain injury. Acta Ne-
urochir., 71: 117–119.
McKenzie, K.J., Mclellan, D.R., Gentleman, S.M., et al. (1996)
Is b-APP a marker of axonal damage in short-surviving head
injury? Acta Neuropathol., 92: 608–613.
Moore, B.W. (1965) A soluble protein characteristic of the
nervous system. Biochem. Biophys. Res. Commun., 19:
739–744.
Muller, M., Schwerdtfegler, K., Maier, B., et al. (2001) Cerebral
blood flow velocity and inflammatory response after severe
traumatic brain injury. Eur. J. Ultrasound, 12(3): 203–208.
Nedd, K., Sfakianakis, G., Ganz, W., et al. (1993) 99mTc-
HMPAO SPECT of the brain in mild to moderate traumatic
brain injury patients: compared with CT, a prospective study.
Brain Inj., 7(6): 469–479.
289
Nevin, N.C. (1967) Neurophathological changes in the white
matter following head injury. J. Neuropathol. Exp. Neurol.,
26: 77–84.
Ng, S.C., Poon, W.S., Chan, M.T., et al. (2000) Transcranial
Doppler ultrasonography (TCD) in ventilated head injure
patients: correlation with stable xenon-enhanced CT. Acta
Neurochir. Suppl., 76: 479–482.
Odelgard, B. (1989) The development of head, face, and neck
protectors for ice hockey players. In: Castaldi C.R. and Ho-
erner E.F. (Eds.), Safety in Ice Hockey: Second Volume,
ASTM STP 1050. American Society for Testing and Mate-
rials, Philadelphia, pp. 220–234.
Okonkwo, D.O., Buki, A., Siman, R. and Povlishock,
J.T. (1999) Cyclosporin A limits calcium-induced axonal
damage following traumatic brain injury. Neuroreport, 10:
353–358.
Okonkwo, D.O. and Povlishock, J.T. (1999) An intrathecal
bolus of cyclosporin a before injury preserves mitochondrial
integrity and attenuates axonal disruption in traumatic brain
injury. J. Cereb. Blood Flow Metab., 19: 443–451.
Oppenheimer, D.R. (1968) Microscopic lesions in the brain
following head injury. J. Neurol. Neurosurg. Psychiatry, 31:
299–306.
Pappius, H.M. (1981) Local cerebral glucose utilization
in thermically traumatized rat brain. Ann. Neurol., 9:
484–491.
Pashby, T. (1993) Eye protection in ice hockey: an historical
review. In: Castaldi C.R., Bishop P.J. and Hoerner E.F.
(Eds.), Safety in Ice Hockey: Second Volume, ASTM STP
1212. American Society for Testing and Materials, Philadel-
phia, pp. 159–163.
Pashby, T., Carson, J.D., Ordogh, D., et al. (2001) Eliminate head-
checking in ice hockey. Clin. J. Sport Med., 11(4): 211–213.
Peerless, S.J. and Rewcastle, N.B. (1967) Shear injuries of the
brain. J. Can. Med. Assoc., 96: 577–582.
Pettus, E.H., Christman, C.W., Giebel, M.L., et al. (1994)
Traumatically induced altered membrane permeability: its
relationship traumatically induced reactive axonal changes. J.
Neurotrauma, 11(5): 507–522.
Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a
distinct set of intra-axonal ultrastructural changes associated
with traumatically induced alteration in axolemma permea-
bility. Brain Res., 772: 1–11.
Piccininni, P. (2001) Mouthguards and concussion prevention,
International Symposium on Concussion in Sport. Br. J.
Sports Med., 35(5): 375.
Pleines, U.E., Morganti-Kossmann, M.C., Rancan, M., et al.
(2001) S-100B reflects the extent of injury and outcome,
whereas NSE is a better indicator of neuroinflammation in
patients with severe traumatic brain injury. J. Neurotrauma,
18(5): 491–498.
Povlishock, J.T. (1986) Traumatically induced axonal damage
without concomitant change in focally related neuronal so-
mata and dendrites. Acta Neuropathol., 70: 53–59.
Povlishock, J.T. (1992) Traumatically induced axonal
injury: pathogenesis and pathobiological implications. Brain
Pathol., 2: 1–12.
290
Povlishock, J.T. (1993) Pathobiology of traumatically induced
axonal injury in animals and man. Ann. Emerg. Med., 22(6):
980–986.
Povlishock, J.T., Becker, D.P., Cheng, C.L.Z., et al. (1983)
Axonal changes in minor head injury. J. Neuropathol. Exp.
Neurol., 42(3): 225–242.
Povlishock, J.T. and Christman, C.W. (1995) The pathobiology
of traumatically induced axonal injury in animals and hu-
mans: a review of current thoughts. J. Neurotrauma, 12(4):
555–564.
Povlishock, J.T. and Jenkins, L.W. (1995) Are the pathobio-
logical changes evoked by traumatic brain injury immediate
and irreversible? Brain Pathol., 5: 415–426.
Povlishock, J.T., Marmarou, A., McIntosh, T., et al. (1997)
Impact acceleration injury in the rat: evidence for focal axo-
lemmal and related neurofilament sidearm alteration. J. Ne-
uropathol. Exp. Neurol., 56: 347–359.
Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced
axonal damage: evidence for enduring changes in axolemmal
permeability with associated cytoskeletal changes. Acta Ne-
urochir., 66: 81–86.
Raabe, A., Grolms, C., Keller, M., et al. (1998) Correlations of
computed tomography findings and serum brain damage
markers following severe head injury. Acta Neurochir., 140:
787–792.
Ranalli, D.N. (2000) Prevention of sports-related traumatic
dental injuries. Advances in Sports dentistry. Dent. Clin.
North Am., 44(1): 35–51.
Reinert, M., Hoelper, B., Doppenberg, E., Zauner, A. and
Bullock, R. (2000) Substrate delivery and ionic balance dis-
turbance after severe human head injury. Acta Neurochir.
Suppl., 76: 439–444.
Ricker, J.H., Muller, R.A., Zafonte, R.D., et al. (2001) Verbal
recall and recognition following traumatic brain injury: a
(0–15)-water positron emission tomography study. J. Clin.
Exp. Neuropsychol., 23(2): 196–206.
Ringger, N.C., O’Steen, B.E., Brabham, J.G., et al. (2004) A
novel marker for traumatic brain injury: CSF aII-spectrin
breakdown product levels. J. Neurotrauma, 21: 1443–1456.
Roberts, W.O., Brust, J.D. and Leonard, B. (1999) Youth ice
hockey tournament injuries: rates and patterns compared to
season play. Med. Sci. Sports Exerc., 31(1): 46–51.
Romner, B., Ingebrigtsen, T., Kongstad, P., et al. (2000) Trau-
matic brain damage: serum S-100 protein measurements re-
lated to neuroradiological findings. J. Neurotrauma, 17(8):
641–647.
Ross, S.A., Cunningham, R.T., Johnston, C.F., et al. (1996)
Neuron-specific enolase as an aid to outcome prediction in
head injury. Br. J. Neurosurg., 10: 471–476.
Ross, D.T., Meaney, D.F., Sabol, M.K., et al. (1995) Distri-
bution of forebrain diffuse axonal injury following inertial
closed head injury in miniature swine. Exp. Neurol., 126:
291–299.
Ruff, R.M., Crouch, J.A., Troster, A.I., et al. (1994) Selected
cases of poor outcome following a minor brain trauma:
comparing neuropsychological and positron emission tomo-
graphy assessment. Brain Inj., 8(4): 297–308.
Samii, A., Badie, H., Fu, K., et al. (1999) Effects of an N-Type
calcium channel antagonist on calcium-45 accumulation fol-
lowing fluid-percussion injury. J. Neurotrauma, 16(10):
879–892.
Sherriff, F.E., Bridges, L.R., Gentleman, S.M., et al. (1994b)
Markers of axonal injury in post mortem human brain. Acta
Neuropathol., 88: 433–439.
Sherriff, F.E., Bridges, L.R. and Sivologanathan, S. (1994a)
Early detection of axonal injury after human head trauma
using immunocytochemistry for b-amyloid precursor protein.
Acta Neuropathol., 87: 55–62.
Siman, R., McIntosh, T.K., Soltesz, K.M., et al. (2004) Proteins
released from degenerating neurons are surrogate markers
for acute brain damage. Neurobiol. Dis., 16: 311–320.
Sinson, G., Bagley, L.J., Cecil, K.M., et al. (2001) Magnetiza-
tion transfer imaging and proton MR spectroscopy in the
evaluation of axonal injury: correlation with clinical outcome
after traumatic brain injury. Am. J. Neuroradiol., 22(1):
143–151.
Skogseid, H.M., Nordby, H.K., Urdal, P., et al. (1992) In-
creased serum creatine kinase BB and neuron-specific enolase
following head injury indicates brain damage. Acta Ne-
urochir., 115: 106–111.
Smith, D.H., Nonaka, M., Miller, R., et al. (2000) Immediate
coma following inertial brain injury dependent on axonal
damage in the brainstem. J. Neurosurg., 93: 315–322.
Sports-Related Recurrent Brain Injuries-United States. (1997)
MMWR Morb. Mortal Wkly. Rep., 46: 224–227.
Stone, J.R., Singleton, R.H. and Povlishock, J.T. (2001) Intra-
axonal neurofilament compaction does not evoke local ax-
onal swelling in all traumatically injured axons. Exp. Neurol.,
172: 320–331.
Stoner, L.J. and Kreating, M. (1993) Hockey equipment: safety
or an illusion? In: Castaldi C.R., Bishop P.J. and Hoerner
E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM
STP 1212. American Society for Testing and Materials, Phil-
adelphia, pp. 183–191.
Strich, S.J. (1961) Shearing of nerves fibers as a cause of brain
damage due to head injury: a pathological study of twenty
cases. Lancet, 2: 443–448.
Strich, S.J. (1970) Lesions in the cerebral hemispheres after
blunt head injury. J. Clin. Pathol., 23(4): 166–171.
Tegner, Y. (2000) Brain concussion in ice hockey: a suggestion
for uniform treatment. In: Biasca N., Montag W.D. and
Gerber Ch. (Eds.), Safety in Ice Hockey, Eight International
Symposium of the International Ice Hockey federation IIHF.
International Ice Hockey Federation IIHF, Zurich, pp.
86–89.
Tegner, Y. (2001) Concussion experiences: Swedish elite ice
hockey league, International Symposium on Concussion in
Sport. Br. J. Sports Med., 35(5): 376.
Tegner, Y., Back, B., Gilenstam, K. and Holm, A. (2000)
Injury pattern in Swedish Elite Ice Hockey. Hygiae, 109(2):
204.
Tegner, Y. and Lorentzon, R. (1996) Concussion among Swed-
ish elite ice hockey players. Br. J. Sports Med., 30(3):
252–255.

Tomei, G., Spagnoli, D., Ducati, A., et al. (1990) Morphology
and neurophysiology of focal axonal injury experimentally
induced in the guinea pig optic nerve. Acta Neuropathol., 80:
506–513.
Woertgen, C.H., Rothoerl, R.D., Holzschuh, M., et al. (1997)
Comparison of serial S-100 and NSE serum measurements
after severe head injury. Acta Neurochir., 139: 1161–1165.
Yaghmai, A. and Povlishock, J.T. (1992) Traumatically in-
duced reactive change as visualized through the use of mono-
clonal antibodies targeted to the neurofilament subunits. J.
Neuropathol. Exp. Neurol., 51: 158–176.
291
Yamamoto, T., Rossi, S., Stiefel, M., et al. (1999) CSF and
ECF glutamate concentrations in head injured patients. Acta
Neurochir. Suppl., 75: 17–19.
Yamazaki, Y., Yada, K., Morii, S., et al. (1995) Diagnostic
significance of serum neuron-specific enolase and myelin ba-
sic protein assay in patients with acute head injury. Surg.
Neurol., 43: 267–270.
Umile, E.M., Sandel, M.E., Alavi, A., et al. (2002) Dynamic
imaging in mild traumatic brain injury: support for the the-
ory of medial temporal vulnerability. Arch. Phys. Med. Re-
habil., 83(11): 1506–1513.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 20

Traumatic brain injury in infants: the phenomenon of

subdural hemorrhage with hemispheric hypodensity
(‘‘Big Black Brain’’)

Ann-Christine Duhaime and Susan Durham

Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medical Center, Lebanon, NH 03756,
USA

Abstract: Clinical and experimental studies of traumatic brain injury during immaturity have been far less
numerous than those involving adults, and many questions remain about differences in injury responses
among patients of different ages. This chapter reviews a distinctive injury pattern common in infants, the
so-called ‘‘big black brain’’ response to acute subdural hematoma. The pathophysiology of this injury
remains incompletely understood. Insights from both clinical observation and experimental studies have
helped to clarify the probable causes of this injury pattern, which appears to require a combination of
stressors during a particular period of maturation.

Keywords: subdural hematoma; infant; black brain; child abuse; pathophysiology

Introduction how clinicians and scientists work synergistically

There are a number of unique phenomena seen in
clinical practice that distinguish infant traumatic
brain injury from that occurring in older children
or adults. One of the most dramatic of these is the
so-called ‘‘big black brain,’’ a pattern of tissue loss
affecting the entire supratentorial hemisphere in
association with acute subdural hematoma. The
circumstances required to cause this injury pattern
and the pathophysiology of the extensive paren-
chymal destruction remain incompletely under-
stood. This chapter will address insights gained
into this injury by both clinical observations and
experimental strategies. Both approaches have
contributed to advancing understanding of the re-
sponse of the infant to brain trauma, and illustrate
Corresponding author. Tel.: +1 603-653-9880;
Fax: +1 603-650-0908; E-mail: acd@hitchcock.org
in head injury research.
Early clinical observations
It has long been recognized that acute subdural
hematoma in infancy could be associated with pro-
found brain injury. The pioneering observations of
Guthkelch (1971) and Caffey (1972, 1974) linked
infantile subdural with a range of neurologic defi-
cits including chronic disability, coma, and death.
These injuries were hypothesized to occur most of-
ten from violent manual shaking. The hypothesis
that shaking caused these injuries was based on
statements from some perpetrators and witnesses
who described shaking, the frequent paucity of vis-
ible external cranial injury in affected infants, an
association with distal metaphyseal and rib frac-
tures, and contemporaneous experimental findings
on the role of rotational forces in subdural

DOI: 10.1016/S0079-6123(06)61020-0 293

294

hematoma and diffuse brain injuries (Ommaya and
Yarnell, 1962; Ommaya et al., 1968, 1973). The
forces and mechanisms necessary to cause acute
subdural hematoma in infancy remain uncertain,
and many affected infants with accidental or in-
flicted events do, in fact, have evidence of head im-
pact (Hahn et al., 1983; Duhaime et al., 1987, 1998;
Alexander et al., 1990; Ghahreman et al., 2005). In
this chapter we will focus on a specific pattern of
immature brain response to acute subdural hema-
toma, regardless of mechanism of injury.
Not all infants with acute subdural hematoma
develop severe brain injury. The spectrum of clini-
cal outcomes ranges from no deficits to death
(Ludwig and Warman, 1984; Duhaime et al., 1987,
1996; Gilles and Nelson, 1998; Ewing-Cobbs et al.,
1999; Ghahreman et al., 2005). As CT scanning
became widely available, patterns of brain damage
emerged in the premorbid state and in survivors.
Areas of hypodensity often were seen which might
be patchy in distribution or could involve the entire
supratentorial compartment unilaterally or bilater-
ally (Zimmerman et al., 1979; Cohen et al., 1986;
Dias et al., 1998; Gilles and Nelson, 1998). Because
of the typical dark-appearing, homogeneous, and
extensively distributed low density seen on the CT
scan of such infants, the term ‘‘big black brain’’ has
been used to describe those with hypodensity in-
volving the entirety of one or both hemispheres
(Duhaime et al., 1993; Graupman and Winston,
2006). This nomenclature evolved from the use of
the term ‘‘black brain’’ to describe the CT appear-
ance of diffuse bilateral hypodensity and loss of
gray-white differentiation, with or without relative
sparing of the thalami (‘‘reversal sign’’), seen most
often in cases of severe or fatal pediatric head in-
jury (Cohen et al., 1986; Whyte and Pascoe, 1989).
In a subset of infant subdurals, hypodensity is seen
which extends from the frontal pole to the occipital
pole, crosses multiple vascular territories, and stops
abruptly at the tentorium. This pattern may be
present at presentation to the hospital or evolve
over several days (Dias et al., 1998). In unilateral
cases, a contralateral frontal wedge-shaped region
of low density also typically is seen, but the rest of
the supratentorial compartment does not exhibit
hypodensity or loss of gray-white differentiation
(Fig. 1). The hypodense hemisphere(s) typically go
on to sustain rapidly progressive atrophic changes
indicative of severe, diffuse brain destruction

Fig. 1. Two-year-old boy with a fall from a bunk bed resulting in acute left subdural hematoma. (A) Acute CT scan, with midline shift.
Hemispheres appear to have similar density at this time-point. The patient underwent emergency clot evacuation. (B) Four days post-
injury. The child has developed hypodensity and swelling of the entire left hemisphere. He underwent aggressive management of
intracranial pressure and survived with a right hemiparesis and hemianopsia.
 295

Fig. 2. Twenty-three month old boy who sustained a witnessed accidental injury with an acute right subdural hematoma. He stood up
initially but then had a generalized seizure within seconds after impact. He received medical attention within minutes, and there was no
apnea or hypoventilation noted. (A) Acute CT scan. (B) CT scan after clot evacuation. There was no brain swelling at surgery. Note
lack of hypodensity, and bifrontal air. The patient regained consciousness promptly after surgery. Despite the symmetric radiologic
appearance of the hemispheres, the child was noted to have a left hemiparesis postoperatively. (C) MRI scan done on postoperative
day 2, FLAIR sequence. Note the high signal in the basal ganglia. (D) MRI, T2 weighted sequence, several months post-injury. Note
the widespread signal change and atrophy in the right hemisphere. (E) MRI, T1 sequence with contrast, 10 years after injury. There is
marked atrophy of the right hemisphere. The patient had a persistent hemiparesis and hemianopsia, developmental delays, and a
seizure disorder. He became seizure-free with no new deficits after hemispherectomy.

(Fig. 2). In children with subdural hematomas, the
bilateral pattern is seen about twice as commonly
as the unilateral pattern, and occurs more com-
monly in very young infants, while older infants
and toddlers more often develop the unilateral
form of ‘‘big black brain’’ (Duhaime et al., 1993;
Gilles and Nelson, 1998). The extent of the
hypodensity correlates with worse acute clinical
status and with worse prognosis (Duhaime et al.,
1996; Ewing-Cobbs et al., 1998; Gilles and Nelson,
1998). Mortality in children with unilateral or bi-
lateral hemispheric involvement is 67% (Duhaime
et al., 1993). Even many years after injury, survi-
vors of bilateral ‘‘big black brain’’ remain blind,
296
non-ambulatory, nonverbal, and profoundly deve-
lopmentally delayed. Those with the predominantly
unilateral form may make a better functional re-
covery but remain severely impaired (Duhaime
et al., 1996).
While many workers have described the areas of
hypodensity seen on CT as ‘‘edema’’ or ‘‘infarct-
ion,’’ these terms should be used with caution.
This is because the pathophysiologic processes
these terms most often connote, in other ages and
with adult disorders, may not apply in this context.
For instance, excitotoxic lesions, to which the im-
mature brain may be particularly vulnerable, can
occur in the setting of normal vascular perfusion.
The pathology and radiology of this type of proc-
ess may look similar to that occurring from vas-
cular occlusion, since both represent a mismatch
between metabolic demand and substrate delivery.
If such a pathophysiology were in play, it might
not be necessary to postulate a mechanism requir-
ing large vessel occlusion, such as strangulation, as
some authors have done (Bird et al., 1987). Thus,
care should be taken to describe findings rather
than ascribing mechanisms that might be associ-
ated with these radiologic findings in other ages
and clinical contexts (Gilles and Nelson, 1998).
Older children and adults with acute subdural
hematoma may suffer extensive concomitant brain
damage, presumably due to the effects of primary
mechanical brain injury and/or to secondary insults
including hypoxia, ischemia, or elevated intracra-
nial pressure. However, the rapid appearance of
extensive, diffuse supratentorial hypodensity with
loss of gray-white differentiation in the unilateral
or bilateral patterns described is seen routinely only
in infants and toddler-aged children. The vast ma-
jority of children with the specific injury type re-
viewed in this paper are under 3 years of age. For
the sake of brevity we most often use the term
‘‘infants’’ in this chapter, but it should be under-
stood that this phenomenon occurs in toddlers as
well.
More recent studies using magnetic resonance
imaging (MRI) have shown changes on diffusion-
weighted imaging which corroborate the idea that
some form of hypoxic-ischemic injury or perfusion-
demand mismatch is particularly common in in-
flicted injury (Ichord et al., 2007). Early involvement
of the basal ganglia can also been seen on MRI
(Figs. 2 and 3). Therefore, the question has arisen as
to whether this striking injury pattern reflects a
unique mechanism of injury, the presence of a par-
ticular type of secondary injury, or a unique re-
sponse of the immature brain to trauma. Since
bilateral ‘‘black brain’’ can occur from diffuse hypoxic-
ischemic insults as well as in the setting of subdural
hematomas, we will focus on the unilateral pattern
as the traumatic entity requiring a unique path-
ophysiologic explanation. Both experimental and
clinical investigations offer insights, as reviewed
below.
Experimental models
The rodent subdural hematoma model most
widely in use presently was first described by
Miller et al. (1990). The model created a subdural
hematoma by injecting autologous blood into the
subdural space of the cerebral convexity through a
small burr hole. Histopathology demonstrates a
bowl-shaped volume of brain damage underlying
the clot. By measuring brain metabolism and cere-
bral blood flow, investigators have demonstrated
that perfusion in the area of the clot is decreased,
while metabolism is increased, creating a relative
mismatch between substrate delivery and demand.
The perfusion abnormality appears to be mediated
by microvessel vasospasm rather than by large
vessel abnormalities (Inglis et al., 1990; Kuroda
and Bullock, 1992). Microdialysis studies with the
model have shown an increase in extracellular
glutamate, and treatment with glutamate receptor
blockers limits the volume of damage, suggesting
that the increased metabolism seen under the
hematoma is related to excitotoxic stress (Bullock
et al., 1990, 1991; Inglis et al., 1992). Similar dam-
age was not seen if a comparable volume of blood
was simply layered over the exposed cortical sur-
face in rodents and a clot left in place for several
days (Duhaime et al., 1992). This demonstrates
that the presence of blood alone is insufficient to
cause damage, but that the blood must be injected
into a closed space to cause injury.
While these studies are intriguing, the pathology
using the rodent subdural hematoma model does
 297

Fig. 3. Two-year-old girl with injury, suspicious for inflicted trauma with acute right subdural hematoma. She had a seizure witnessed
in an outside emergency department, and was unresponsive with a dilated right pupil on presentation. The clot was evacuated and the
brain was noted to be slightly full at surgery, so a hemicraniectomy was performed. Intracranial pressure was measured and easily
controlled, and excellent brain perfusion was maintained. The patient was intubated but responsive and was able to follow commands.
(A) MRI, T2 sequence, done 1 day after surgery. Note symmetric appearance of brain parenchyma. (B) Diffusion-weighted sequence.
Note bright signal in basal ganglia and temporal and occipital cortex. (C) CT scan, postoperative day 3. Note development of
hypodensity of entire right hemisphere, with increased swelling. (D) MRI, diffusion sequence, one week postoperatively. Note bright
signal in entire hemisphere, as well as contralateral basal ganglia. (E) MR angiogram. All arteries appear patent. (F) MRI, fast spin-
echo T2 weighted sequence, 2 months post-injury. Note widespread damage to right hemisphere. The patient has a hemiparesis and
hemianopsia.

not reproduce the pathology of the ‘‘big black
brain.’’ Injection of blood causes a localized lesion
rather than one involving the entire hemisphere.
To learn whether the infantile pattern could be
reproduced by adapting the rodent model to an
immature, gyrencephalic brain, Shaver et al.
(1996) injected blood into the subdural space of
three-week-old piglets. Their model was modified
by use of a ‘‘cranial window’’ through the bone
that allowed confirmation of a thin but extensive
subdural clot overlying the hemisphere, compara-
ble to that seen most often in human infants. This
model also failed to reproduce the ‘‘big black
brain’’ phenomenon, as it too created a localized
injury. However, unlike the rodent lesion, which
was largely cortical, the piglet lesion preferentially
affected the underlying white matter. The authors
hypothesized that the white matter was affected
because of its increased vulnerability during early
life, when the metabolic rate is relatively high in
white matter but the vascular supply is immature
and may respond inadequately to increased de-
mand. It is for this reason that the white matter in
infants is thought to be preferentially vulnerable to
298
damage from perinatal hypoxic-ischemic insults
and prematurity (Rorke, 1992).
Both the rodent and the piglet subdural injec-
tion models fail to mimic the mechanical trauma
that occurs during the injury event in patients.
Since the presence of blood in the subdural space
was insufficient to cause the hemispheric damage
seen in infant subdural injuries, Duhaime et al.
(2000) performed a series of studies using piglets to
isolate the effect of mechanical trauma on the
immature brain. This was done to test the hypo-
thesis that the infant brain might be particularly
vulnerable to damage from mechanical trauma
relative to other ages, thus explaining the extensive
hemispheric injuries seen with acute subdural hema-
toma. The group developed a model of scaled cor-
tical impact that delivered a mechanically compa-
rable strain load to the same brain structures
at different ages. These studies showed that rather
than being selectively vulnerable to mechanical
trauma, the infant brain was relatively resistant
to injury from mechanical deformation (Duhaime
et al., 2003). Therefore, another explanation for
the unique pattern of extensive brain damage seen
with subdurals in infancy was needed.
Other workers have attempted to more closely
reproduce the mechanical forces that might be re-
sponsible for infant subdural hemorrhage using ani-
mal models. Because of the long-held concept that
the injury might be caused by manual shaking of an
infant, Smith et al. (1998) subjected rats to pro-
longed mechanical shaking. While subarachnoid
hemorrhage and surface contusions were found,
hemispheric tissue loss was not produced in this
model. Anesthetized, immature sheep subjected to
30 min of intermittent, violent manual shaking
exhibited similar finding to the shaken rats, with-
out hemispheric loss (Finnie et al., 2006). Piglets
subjected to a single or double high-velocity head
rotation showed diffuse axonal injury and focal
frontal subdural hemorrhage, but have not to date
demonstrated the ‘‘big black brain’’ phenomenon
(Raghupathi and Margulies, 2002; Raghupathi
et al., 2004).
It can be seen that in all these experimental
models, elements of injury occur in isolation,
whereas in the clinical setting, multiple patho-
physiologic stressors may occur in combination.
We will now return to the clinical arena, where
additional observations relevant to this issue have
been made.
The role of apnea
It has long been recognized that infants are sus-
ceptible to apnea from a variety of causes, includ-
ing head injury. That children with subdurals often
present with hypoventilation or frank apnea is well
known (Johnson et al., 1995; Kemp et al., 2003;
Ichord et al., 2007). Therefore, it has been hypoth-
esized that hypoxia/ischemia, rather than trauma
itself, is the cause of the diffuse hypodensity pat-
tern seen in the most severe of these injuries. The
fact that many of these infants have been the vic-
tims of inflicted trauma has been thought to
contribute to this association. Specifically, it has
been speculated that after the infant is injured, the
perpetrator does not seek immediate medical atten-
tion, hoping the child will recover spontaneously.
This leads to delay in care. If the child hypoven-
tilates during this delay period, diffuse hypoxic
brain damage may result.
This scenario matches the histories obtained in
many cases, and likely contributes to the damage
seen in many infants. However, it is not a sufficient
explanation for all aspects of the ‘‘black brain’’
phenomenon. This is because there are well-
documented accidental cases of subdural hema-
toma in infants and toddlers that were witnessed
and in which children got immediate medical
attention with no apparent apnea or hypoventila-
tion, but in which the phenomenon of hemispheric
hypodensity nonetheless appeared (Figs. 1 and 2).
The other limitation of apnea as the sole explana-
tion for the phenomenon is that it fails to account
for the fact that in one-third of the cases, the
hypodensity is unilateral. One would generally
expect a diffuse insult like hypoxia to affect the
brain symmetrically. The observation that one
hemisphere can be totally destroyed while the
other remains relatively preserved suggests that
more than one factor is in play. Gilles and others
have noted that in unilateral cases, the hypoden-
sity occurs on the side of the subdural hematoma
(Duhaime et al., 1998; Gilles and Nelson, 1998;
 299

Gilles et al., 2003). Therefore, it appears that some injury as determined by both immunohistochemi-
synergistic effect of the hemorrhage (or the forces stry and light microscopic findings. However,
that caused it) and a second insult may be needed. patients frequently did exhibit ‘‘ischemic’’ findings
(Geddes et al., 2001a, b). Additionally, damage at
the cervicomedullary junction was often encoun-
The role of seizures
tered, also raising the question whether apnea due
to damage in this location might contribute to the
Infants with head injury and other neurologic in-
findings; this has been hypothesized by others as
sults often present with seizures, which also may
well (Hadley et al., 1989; Johnson et al., 1995).
occur subclinically, especially in very young chil-
However, Geddes and colleagues did not address
dren. The incidence of clinical seizures in inflicted
the specific phenomenon of unilateral ‘‘big black
head injuries has been reported in 40–79% of pa-
brain’’ in their series.
tients (Ludwig and Warman, 1984; Johnson et al.,
Kohanek and colleagues have measured a wide
1995; Gilles and Nelson, 1998; Ghahreman et al.,
variety of cerebrospinal fluid markers from ventri-
2005; Bechtel et al., 2006). Clancy et al. (1988) have
culostomy samples from infants with traumatic
noted that up to 79% of seizures in neonates are
brain injury compared to those found in lumbar
subclinical, that is, are only detected on EEG re-
puncture samples from uninjured infants. They
cordings. This would suggest the possibility that
have reported higher levels of excitatory amino
even more seizure events may occur in infants with
acids and other markers of cellular damage in fluid
inflicted injury than are apparent clinically, espe-
from infants with inflicted injuries and in younger
cially in the youngest patients. Pharmacologic
infants compared to older children and those with
agents used for sedation and/or paralysis during
accidental mechanisms (Ruppell et al., 2001;
head trauma management in the intensive care unit
Berger et al., 2006). Whether these differences
could influence clinical epileptic events, with some
reflect an age effect or more extensive damage in
agents (such as midazolam) potentially protecting
the inflicted injury patients (primary or secondary)
against seizures, and others (narcotics, paralytics)
remains unclear.
potentially obscuring their detection. Seizures are
known to increase excitotoxic stress in animal
models and to be associated with increased meta-
Brain swelling, variability, and decompressive
bolic demand and worse outcome in human pa-
craniectomy
tients. Therefore, some workers have hypothesized
that the frequent occurrence of seizures in infants
Most babies with bilateral or unilateral hemi-
with traumatic brain injury may contribute to the
spheric hypodensity develop brain swelling and
pathophysiology of an infarct-like picture, such as
increased intracranial pressure. Survival is better
occurs in hemispheric hypodensity (Duhaime et al.,
in the youngest infants, probably because of their
1998). However, the exact contribution of seizures,
ability to split the sutures to relieve some of the
clinical or subclinical, to this phenomenon remains
pressure. Interestingly, in some cases brain swell-
unknown.
ing is not problematic, and why some children do
not develop significant swelling remains unknown
Neuropathology and clinical neurophysiology (see Figs. 1–3). It may be that acute injury factors
or genetic differences determine this variability.
As immunohistochemical techniques have become In recent years, hemicraniectomy has gained in-
available, additional detail about the neuropatho- creasing acceptance as a means to alleviate the
logy of fatal infant inflicted injuries has been brainstem compressive effects of brain swelling in
determined. Geddes reported that contrary to infants and young children (Cho et al., 1995). It is
older studies in which light microscopy was used not yet clear in which children and at what time-
to assess neuropathology, in her series most pa- point this should be undertaken, and the effect of
tients exhibited little in the way of diffuse axonal this practice on outcome has been difficult to
300
assess, although it does appear to be effective in
lowering intracranial pressure (Taylor et al., 2001).
At present it appears that even in those cases in
which the hypodensity appears over several days,
the processes involved have already begun at prese-
ntation and do not appear reversible with current
therapies (Duhaime et al., 1996; Dias et al., 1998;
Graupman and Winston, 2006). Therefore, treat-
ment strategies are aimed at protecting the un-
damaged hemisphere. It has been hypothesized
that the nearly universal finding of a wedge of
damage in the contralateral medial frontal lobe
results from subfalcine herniation with trapping of
the callosomarginal branch of the anterior cerebral
artery against the falx, resulting in focal infarction
(Gilles and Nelson, 1998). In the setting of hemi-
craniectomy, this internal herniation is amelio-
rated. However, the authors have observed that
some contralateral progression can occur despite
early relief of pressure, raising the possibility of a
neurochemically-mediated process rather than a
purely vascular occlusive one (Fig. 3).
Pathophysiology of the ‘‘Big Black Brain’’:
a hypothesis
Both experimental and clinical evidences support
the idea that no single insult is sufficient to explain
all cases of unilateral hemispheric hypodensity in
association with subdural hematoma in infants and
toddlers. Not all children with subdural hemorrh-
age develop this finding; thus subdural blood itself
appears to be an insufficient cause. Global physio-
logic insults cannot fully explain the distribution in
unilateral cases. Hemispheric hypodensity can be
seen in nonaccidental and accidental injuries, with
apnea and with no apparent apnea, with clinical
seizures and without. Brain swelling usually but
not always occurs. Atrophy is rapid and profound.
The phenomenon of subdural blood with unilateral
hemispheric hypodensity has not yet been repro-
duced in an experimental model.
In many situations, infants and children demon-
strate a remarkable ability to compensate for sys-
temic physiologic stresses. A common example is
the child’s ability to maintain blood pressure in the
face of hypovolemia. However, the immature
organism has a limit to its ability to compensate,
and when this limit is reached, the resulting
decompensation is often precipitous.
It has been demonstrated in a number of con-
texts that the immature brain’s response to insult
varies from the adult response. Its response to glo-
bal hypoxia-ischemia varies, with some ages and in
some outcome measures having increased vulner-
ability to cell death and in others, increased resist-
ance compared to the mature brain (Painter, 1995;
Johnston et al., 2001; Vanucci and Hagberg, 2004).
Resistance to other insults seen in the context of
trauma has been demonstrated by a number of ex-
perimental models cited above, including isolated
focal mechanical trauma and injected subdural
hematoma with increased intracranial pressure; in
these contexts, immaturity confers protection.
In cases of infantile subdural hematoma with
hemispheric hypodensity, it appears likely that the
brain is subjected to a combination of stresses that
exceed its capacity to compensate. Most clinical
scenarios would point to a combination of local
perfusion decrement over the surface of the cere-
bral hemisphere related to the presence of subdural
blood, in combination with a second, more global
insult. This could include apnea/hypoxia, hyper-
carbia, hypotension, or increased metabolic de-
mand from seizures. There may be additional in-
sults that are at present difficult to discern or
measure. Why the process involves the entire hemi-
sphere, even beyond the extent of visible subdural
blood, remains incompletely understood. It ap-
pears to represent a diffuse decompensation re-
sulting in widespread parenchymal death that
under certain circumstances may reach threshold
only on the side of the larger hemorrhage. Indi-
vidual differences with respect to cerebrovascular
responses, acute injury cascades, inflammation,
apoptotic pathways, or other neurochemical proc-
esses likely influence why some patients have pro-
found swelling and others do not.
Conclusion
The unique infant pattern of unilateral hemi-
spheric hypodensity associated with accidental
or inflicted acute subdural hematoma provides

insight into age-related differences in brain injury
response. This pattern of extensive and homoge-
neous tissue destruction appears to require a com-
bination of insults that overwhelm the immature
brain’s ability to compensate and affects the entire
hemisphere. Because the specific physiologic stres-
ses vary from case to case, including hypoxia,
hypotension, hypercarbia, and seizure activity,
different factors that decrease substrate delivery
or increase metabolic demand may be at play for
different patients. The observation that this pat-
tern occurs uniquely in infants and toddlers more
likely reflects their age-dependent brain response
to combined stresses rather than reflecting a single,
specific mechanism or circumstance of injury.
References
Alexander, R., Sato, Y., Smith, W. and Bennett, T. (1990)
Incidence of impact trauma with cranial injuries ascribed to
shaking. Am. J. Dis. Child, 144: 724–726.
Bechtel, K., Stoessel, K., Leventhal, J.M., Ogle, E., Teague, B.,
Lavietes, S., Banyas, B., Allen, K., Dziura, J. and Duncan, C.
(2006) Characteristics that distinguish accidental from abu-
sive injury in hospitalized young children with head trauma.
Pediatrics, 114: 165–168.
Berger, P.R., Adelson, P., Richichi, R. and Kochanek, P.M.
(2006) Serum biomarkers after traumatic and hypoxemic
brain injuries: insight into the biochemical response of the
pediatric brain to inflected brain injury. Dev. Neurosci., 28:
327–335.
Bird, C.R., McMahan, J.R., Gilles, F.H., Senac, M.O. and
Apthorp, J.S. (1987) Strangulation in child abuse: CT diag-
nosis. Radiology, 163: 373–375.
Bullock, R., Butcher, S. and Mcculloch, J. (1990) Changes in
extracellular glutamate concentration after acute subdural
haematoma in the rat: evidence for an ‘‘excitotoxic’’ mech-
anism? Acta Neurochir. Suppl., 51: 274–276.
Bullock, R., Butcher, S.P., Chen, M., Kendall, L. and Mccul-
loch, J. (1991) Correlation of the extracellular glutamate
concentration with extent of blood flow reduction after sub-
dural hematoma in the rat. J. Neurosurg., 74: 794–802.
Caffey, J. (1972) On the theory and practice of shaking infants:
its potential residual effects of permanent brain damage and
mental retardation. AJDC, 124: 161–169.
Caffey, J. (1974) The whiplash shaken infant syndrome: manual
shaking by the extremities with whiplash-induced intracranial
and intraocular bleedings, linked with residual permanent
brain damage and mental retardation. Pediatrics, 54:
396–403.
Cho, D., Wang, Y. and Chi, C. (1995) Decompressive cranio-
tomy for acute shaken/impact baby syndrome. Pediatr.
Neurosurg., 23: 192–198.
301
Clancy, R.R., Legido, A. and Lewis, D. (1988) Occult neonatal
seizures. Epilepsia, 29: 256–261.
Cohen, R.A., Kaufman, R.A., Myers, P.A. and Towbin, R.B.
(1986) Cranial computed tomography in the abused child
with head injury. AJR, 146: 97–102.
Dias, M.S., Backstrom, J., Falk, M. and Li, V. (1998) Serial
radiography in the infant shaken impact syndrome. Pediatr.
Neurosurg., 29: 77–85.
Duhaime, A.C., Bilaniuk, L. and Zimmerman, R. (1993) The
‘‘big black brain’’: radiographic changes after severe inflicted
head injury in infancy. J. Neurotrauma, 10: S59.
Duhaime, A.C., Christian, C., Moss, E. and Seidl, T. (1996)
Long-term outcome in children with the shaking-impact
syndrome. Pediatr. Neurosurg., 24: 292–298.
Duhaime, A.C., Christian, C.W., Rorke, L.B. and Zimmerman,
R.A. (1998) Nonaccidental head injury in infants: the
‘‘Shaken Baby Syndrome.’’ New Engl. J. Med., 338:
1822–1829.
Duhaime, A.C., Gennarelli, L., Yachnis, A.T. and Ross, D.T.
(1992) Acute subdural hematoma: is the blood itself toxic?
Pediatric Section, American Association of Neurological
Surgeons, Annual Meeting. Vancouver, B.C.
Duhaime, A.C., Gennarelli, T.G., Thibault, L.E., Bruce, D.A.,
Margulies, S.S. and Wiser, R. (1987) The shaken baby
syndrome: a clinical, pathological, and biomechanical study.
J. Neurosurg., 66: 409–415.
Duhaime, A.C., Hunter, J.V., Grate, L.L., Kim, A., Demidenko,
E., Golden, J.A. and Harris, C. (2003) Magnetic resonance
imaging studies of age-dependent responses to scaled focal
brain injury in the piglet. J. Neurosurg., 99: 542–548.
Duhaime, A.C., Margulies, S.S., Durham, S.R., O’Rourke,
M.M., Golden, J.A., Marwaha, S. and Raghupathi, R. (2000)
Maturation-dependent response of the piglet brain to scaled
cortical impact. J. Neurosurg., 93: 455–462.
Ewing-Cobbs, L., Kramer, L., Prasad, M., Canales, D.N.,
Louis, P.T., Fletcher, J.M., Vollero, H., Landry, S.H. and
Cheung, K. (1998) Neuroimaging, physical, and develop-
mental findings after inflicted and noninflicted traumatic
brain injury in young children. Pediatrics, 102: 300–307.
Ewing-Cobbs, L., Prasad, M., Kraer, L. and Landry, S. (1999)
Inflicted traumatic brain injury: relationship of developmental
outcome to severity of injury. Pediatr. Neurosurg., 31: 251–258.
Finnie, J.W., Manavis, J. and Blumbergs, P.C. (2006) An ovine
model of the shaken baby syndrome. International Neuro-
trauma Society 8th Annual Neurotrauma Symposium. Rotter-
dam, The Netherlands, International Neurotrauma Society.
Geddes, J.F., Hackshaw, A.K., Vowles, G.H., Nickols, C.D.
and Whitwell, H.L. (2001a) Neuropathology of inflicted head
injury in children: I. Patterns of brain damage. Brain, 124:
1290–1298.
Geddes, J.F., Hackshaw, A.K., Vowles, G.H., Nickols, C.D.
and Whitwell, H.L. (2001b) Neuropathology of inflicted head
injury in children: II. Microscopic brain injury in infants.
Brain, 124: 1299–1306.
Ghahreman, A., Bhasin, V., Chaseling, R., Andrews, B. and
Lang, E.W. (2005) Nonaccidental head injuries in children: a
Sydney experience. J. Neurosurg., 103: 213–218.
302
Gilles, E.E., Mcgregor, M.L. and Levy-Clark, G. (2003) Retinal
hemorrhage asymmetry in inflicted head injury: a clue to
pathogenesis? J. Pediatr., 143: 494–499.
Gilles, E.E. and Nelson Jr., M.D. (1998) Cerebral complica-
tions of nonaccidental head injury in childhood. Pediatr.
Neurol., 19: 119–128.
Graupman, P. and Winston, K.R. (2006) Nonaccidental head
trauma as a cause of childhood death. J. Neurosurg. (4 Suppl.
Pediatr.), 104: 245–250.
Guthkelch, A.K. (1971) Infantile subdural hematoma and its
relationship to whiplash injuries. Br. Med. J., 2: 430–431.
Hadley, M.N., Sonntag, V.K.H., Rekate, H.L. and Murphy, A.
(1989) The infant whiplash-shake syndrome: a clinical and
pathological study. Neurosurgery, 24: 536–540.
Hahn, Y.S., Raimondi, A.J., Mclone, D.G. and Yamanouchi,
Y. (1983) Traumatic mechanisms of head injury in child
abuse. Childs Brain, 10: 229–241.
Ichord, R.N., Naim, M., Pollock, A.N., Nance, M.L., Margu-
lies, S.S. and Christian, C.W. (2007) Hypoxic-ischemic injury
complicates inflicted and accidental traumatic brain injury in
young children: the role of diffusion-weighted imaging. J.
Neurotrauma, 24(1): 106–118.
Inglis, F., Kuroda, Y. and Bullock, R. (1992) Glucose hyper-
metabolism after acute subdural hematoma is ameliorated by
a competitive NMDA antagonist. J. Neurotrauma, 9: 75–84.
Inglis, F.M., Bullock, R., Chen, D.I., Graham, D.I., Miller,
J.D. and Mcculloch, J. (1990) Ischaemic brain damage asso-
ciated with tissue hypermetabolism in acute subdural haem-
atoma: reduction by a glutamate antagonist. Acta Neurochir.
Suppl., 51: 277–279.
Johnson, D.L., Boal, D. and Baule, R. (1995) Role of apnea in
nonaccidental head injury. J. Ped. Neurosurg., 23: 305–310.
Johnston, M.V., Trescher, W.H., Ishida, A. and Nakajima, W.
(2001) Neurobiology of hypoxic-ischemic injury in the de-
veloping brain. Pediatr. Res., 49: 735–741.
Kemp, A.M., Stoodley, N., Cobley, C., Coles, L. and Kemp,
K.W. (2003) Apnoea and brain swelling in non-accidental
head injury. Arch. Dis. Child., 88: 472–476.
Kuroda, Y. and Bullock, R. (1992) Local cerebral blood flow
mapping before and after removal of acute subdural hema-
toma in the rat. Neurosurgery, 30: 687–691.
Ludwig, S. and Warman, M. (1984) Shaken baby syndrome: a
review of 20 cases. Ann. Emerg. Med., 13: 104–107.
Miller, J.D., Bullock, R., Graham, D.I., Chen, M.H. and Teas-
dale, G.M. (1990) Ischemic brain damage in a model of acute
subdural hematoma. Neurosurgery, 27: 433–439.
Ommaya, A.K., Corrao, P. and Letcher, F.S. (1973) Head in-
jury in the chimpanzee. Part 1: biodynamics of traumatic
unconsciousness. J. Neurosurg., 39: 152–166.
Ommaya, A.K., Faas, F. and Yarnell, P. (1968) Whiplash in-
jury and brain damage: an experimental study. JAMA, 204:
285–289.
Ommaya, A.K. and Yarnell, P. (1962) Subdural hematoma af-
ter whiplash injury. Lancet, 2: 237–239.
Painter, M.J. (1995) Animal models of perinatal asphyxia:
contributions, contradictions, and clinical relevance. Semin.
Pediatr. Neurol., 2: 37–56.
Raghupathi, R. and Margulies, S.S. (2002) Traumatic axonal
injury after closed head injury in the neonatal pig. J. Neuro-
trauma, 19: 843–853.
Raghupathi, R., Mehr, M.F., Helfaer, M.A. and Margulies,
S.S. (2004) Traumatic axonal injury is exacerbated following
repetitive closed head injury in the neonatal pig. J. Neuro-
trauma, 21: 307–316.
Rorke, L.B. (1992) Anatomical features of the developing brain
implicated in pathogenesis of hypoxic-ischemic injury. Brain
Pathol., 2: 211–221.
Ruppell, R.A., Kochanek, P.M., Adelson, P.D., Rose, M.E.,
Wisniewski, S.R., Bell, M.J., Clark, R.S., Marion, D.W. and
Graham, S.H. (2001) Excitatory amino acid concentrations
in ventricular cerebrospinal fluid after severe traumatic brain
injury in infants and children: the role of child abuse. J.
Pediatr., 138: 18–25.
Shaver, E., Duhaime, A.C., Curtis, M., Gennarelli, L.M. and
Barrett, R. (1996) Experimental acute subdural hematoma in
infant piglets. Pediatr. Neurosurg., 25: 123–129.
Smith, S.L., Andrus, P.K., Gleason, D.D. and Hall, E.D. (1998)
Infant rat model of the shaken baby syndrome: preliminary
characterization and evidence for the role of free radicals in
cortical hemorrhaging and progressive neuronal degenera-
tion. J. Neurotrauma, 15: 693–705.
Taylor, A., Butt, W., Rosenfeld, J., Shann, F., Ditchfield, M.,
Lewis, E., Klug, G., Wallace, D., Henning, R. and Tibbals,
J. (2001) A randomized trial of very early decompressive
craniectomy in children with traumatic brain injury and
sustained intracranial hypertension. Child Nerv. Syst., 17:
154–162.
Vanucci, S.J. and Hagberg, H. (2004) Hypoxia-ischemia in the
immature brain. J. Exp. Biol., 207(Pt 18): 3149–3154.
Whyte, K.M. and Pascoe, M. (1989) Does ‘‘black’’ brain
mean doom? Computed tomography in the prediction
of outcome in children with severe head injuries: ‘‘benign’’
vs. ‘‘malignant’’ brain swelling. Australas. Radiol., 33:
344–347.
Zimmerman, R.A., Bilaniuk, L.T., Bruce, D., Schut, L., Uzzell,
B. and Goldberg, H.I. (1979) Computed tomography of cra-
niocerebral injury in the abused child. Radiology, 130:
687–690.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 21

Traumatic brain injury and Alzheimer’s disease:

a review

Corinna Van Den Heuvel1,, Emma Thornton1 and Robert Vink1,2

1
Discipline of Pathology, University of Adelaide, Adelaide, Australia
2
Centre for Neurological Diseases, The Hanson Institute, Adelaide, Australia

Abstract: In an effort to identify the factors that are involved in the pathogenesis of Alzheimer’s disease
(AD), epidemiological studies have featured prominently in contemporary research. Of those epidemio-
logical factors, accumulating evidence implicates traumatic brain injury (TBI) as a possible predisposing
factor in AD development. Exactly how TBI triggers the neurodegenerative cascade of events in AD
remains controversial. There has been extensive research directed towards understanding the potential
relationship between TBI and AD and the putative influence that apolipoprotein E (APOE) genotype has
on this relationship. The aim of the current paper is to provide a critical summary of the experimental and
human studies regarding the association between TBI, AD and APOE genotype. It will be shown that
despite significant discrepancies in the literature, there still appears to be an increasing trend to support the
hypothesis that TBI is a potential risk factor for AD. Furthermore, although it is known that APOE
genotype plays an important role in AD, its link to a deleterious outcome following TBI remains incon-
clusive and ambiguous.

Keywords: amyloid precursor protein; traumatic brain injury; apolipoprotein E; Alzheimer’s disease

Epidemiology of TBI and AD (AD) is the most common neurodegenerative dis-

Traumatic brain injury (TBI) is the leading cause
of death and disability in people under the age of
45 years in industrialized countries. Studies in
Australia, the United States, France and Spain
indicate that the incidence of death from TBI is
20–30 per 100,000 (Finfer and Cohen, 2001) with
motor vehicle accidents accounting for the major-
ity of fatal head injuries (Kraus, 1993). Those in-
dividuals who survive TBI are often left with
permanent neurological deficits, which adversely
affect their quality of life, and contributes to the
enormous social and economic costs of TBI that
are borne by communities. Alzheimer’s disease
Corresponding author. Tel.: +61 8 8222 3370; Fax: +61 8
8222 3392; E-mail: corinna.vandenheuvel@adelaide.edu.au
order of modern societies accounting for 50–60%
of all age-related dementia (Andersen et al., 2006).
Specifically, AD afflicts 8–10% of the population
over the age of 65 and almost 50% over the age of
85 (Mattson, 1997). Over 24 million people world-
wide are estimated to be currently suffering from
AD, and with the projected rise in the elderly
population over the coming decades, this has been
estimated to rise to 81 million by 2040 (Miller
et al., 2006). The possibility that TBI may predis-
pose a person to developing AD in later life has
significant social and medical implications, and
reinforces the need for preventative efforts and
health service planning to cope with the potential
large increase in the number of AD patients (Lye
and Shores, 2000). It is therefore critical to estab-
lish whether any link between TBI and AD does

DOI: 10.1016/S0079-6123(06)61021-2 303

304
exist, and whether the mechanisms associated
with this link can be exploited to develop an
interventional therapy.
The ‘Amyloid Cascade Hypothesis’
Neuropathologically, AD is characterized by the
presence of neurofibrillary tangles (NFTs) and
neuritic plaques. NFTs in the brain consist prima-
rily of hyperphosphorylated tau. Neuritic amyloid
plaques consist primarily of aggregated amyloid-b
peptides (Ab), a peptide of 40–42 amino acids,
which are surrounded by dystrophic neurites, mi-
croglia and reactive astrocytes (Selkoe, 2001). It is
now generally accepted that the development of
these two pathologies is central to the pathogenesis
of AD and the leading ‘amyloid cascade hypoth-
esis’ suggests that it is the accumulation of Ab in
the brain which is the primary influence in AD
(Hardy and Selkoe, 2002). Ab has been hypothe-
sized to cause the pathologic and behavioural
manifestations of AD, including neurofibrillary
tangle formation, neuronal degeneration, synaptic
dysfunction and loss as well as impaired memory
(Janus et al., 2000). According to this theory, the
accompanying NFT formation and neuronal loss
are downstream events resulting from an imbal-
ance between Ab production from its precursor
Amyloid Precursor Protein (APP). Evidence in
support of the hypothesis comes from the know-
ledge that genetic defects such as mutations in the
APP gene, and the presenilin 1 and 2 genes, results
in aberrant amyloidogenic processing of APP
leading to Ab deposition and rare autosomal
dominant forms of familial AD (Selkoe, 1997).
Also, active immunization against Ab is associated
with reduced numbers of amyloid deposits in
human AD brains (Hardy and Selkoe, 2002),
while passive immunization with antibodies
against Ab results in neutralization of the patho-
logic Ab resulting in rapid improvement in spatial
learning and memory, and restoration of blood-
brain barrier integrity in transgenic mice (Lee
et al., 2006). Degenerating neurones in AD display
impaired energy production, increased oxidative
damage and disruption of calcium homeostasis,
and importantly Ab has been shown to be involved
in all of these cell malfunctions (Mattson, 2004).
Moreover, Ab causes apoptotic cell death of neu-
ronal cells in culture by the induction of caspases,
known instigators of apoptotic cell death (Awasthi
et al., 2005). Thus, when Ab is injected into the
hippocampal region of rats, an almost complete
loss of CA1 neurones is observed (Miguel-Hidalgo
and Cacabelos, 1998).
APP processing
Neurotoxic Ab peptides are derived from enzymatic
processing of the ubiquitously expressed, highly
conserved type-1 transmembrane glycoprotein
called APP. Understanding the normal structure,
molecular function and processing of APP is there-
fore critical to unravelling the molecular basis of
AD. APP is found in all neurons and some glial cells
in the central and peripheral nervous systems as well
as being expressed in platelets, endothelial cells and
fibroblasts. In the brain, where it is constitutively
expressed, APP serves a synaptic function and is
located predominantly in neuronal cell bodies, con-
centrated at axosomatic and other synaptic sites
(Beyreuther et al., 1993). After translation of APP,
which occurs within the endoplasmic reticulum, the
protein is transported through the secretory path-
way becoming N- and O-glycosylated and tyrosyl-
sulphated while moving through the trans-Golgi
network (Tomita et al., 1998). Once mature, APP
can be processed by two mutually exclusive complex
pathways, either the non-amyloidogenic or the amy-
loidogenic pathway (Fig. 1). The non-amyloidogenic
pathway accounts for the majority of APP process-
ing (Suh and Checler, 2002), and results in a secreted
a form of APP (sAPPa) via a-secretase cleavage.
The b- and g-secretase pathway is responsible for
producing secreted APPb (sAPPb) and the toxic Ab
which is found within amyloid plaques in AD (Mills
and Reiner, 1999). Thus, APP can result in both a
beneficial or detrimental product depending on the
way in which it is post-translationally processed
within cells.
The secretases which play a crucial role in APP
processing and cleavage have now been well chara-
cterised. The a-secretase includes members of the A
disintegrin and metalloprotease (ADAM) family of

Fig. 1. Schematically illustrating the cleavage pathways available to amyloid precursor protein and the mutually exclusive nature of
the a- and b-secretase pathways.
proteases (Lammich et al., 1999), b-secretase is a
type-1 membrane-spanning aspartic protease termed
b-site APP-cleavage enzyme 1 (BACE1) (Vassar
et al., 1999), and g-secretase comprises a complex of
at least four proteins — presenilin 1 or 2, anterior
pharynx defective homologue 1 (APH1), presenilin
enhancer 2 homologue (PEN2) and nicastrin
(reviewed by Wilquet and De Strooper (2004)).
Secretases are obvious targets for drug development
for the prevention and treatment of AD. Therapies
either aim to increase a-secretase activity, thereby
increasing non-amyloidogenic cleavage of APP, or
inhibit the b- and g-secretase (Dewachter and Van
Leuven, 2002; Arbel et al., 2005). g-Secretase inhib-
itors have also been produced (Dewachter and Van
Leuven, 2002) but these proteases have a wide-
spread tissue distribution and their use in humans
may be compromised by side effects resulting from
blockade of g-secretase cleavage of Notch and other
protein substrates (King et al., 2004). In contrast,
BACE inhibitors may prove beneficial in reducing
Ab production without major side effects. Indeed
BACE deficient mice have reduced Ab production
and do not exhibit any overt abnormal phenotypes
(Roberds et al., 2001). To date, drugs that target
specific sites of this Ab cascade have only been
employed in cell culture and animal models of AD,
such that their potential in the clinic remains
unclear.
As indicated above, a-secretase processing not
only precludes the formation of Ab (Mattson, 1997)
305
but it also gives rise to sAPPa, which has been re-
ported to have many neuroprotective and neurotro-
phic functions within the central nervous system (as
reviewed by Mattson et al. (1993)). The discovery
that sAPPa has many beneficial functions, and the
knowledge that production of sAPPa precludes the
formation of the toxic Ab, has led to an increased
awareness of the factors that govern its production.
These factors are the focus of many investigations
that attempt to modulate APP processing toward
production of sAPPa, thus slowing or inhibiting Ab
production (Racchi et al., 1999). Stimulation of
glutamatergic G-protein-coupled receptors linked
to the phospholipase C/protein kinase C (PKC)
signalling system regulate and increase the release
of sAPPa (Lee et al., 1997; Lee and Wurtman,
2000). Recently it has been shown that neuronal
adaptor proteins X 11a and X 11b interact with
APP processing, possibly by modulating BACE
cleavage of APP, and resulting in inhibition of
Ab production (Miller et al., 2006). Furthermore,
SorLA/LR11, which is a sorting receptor, regulates
the intracellular transport and processing of APP in
neurons thus inhibiting Ab production (Andersen
et al., 2006).
APP and Ab within damaged axons
It is feasible that a relationship may exist between
TBI and AD since TBI leads to overexpression of
306
APP within neuronal cell bodies (McKenzie et al.,
1994; Bramlett et al., 1997; Van Den Heuvel et al.,
1999) and accumulation of APP within traumat-
ically injured axons (Blumbergs et al., 1995).
Theoretically, the accumulation of APP results in
an increase in the substrate that could potentially
be processed to form amyloidogenic Ab. It has
therefore been hypothesized by a number of
groups that the overexpression of APP may
exceed the limit of normal processing capacity,
resulting in mismetabolization of APP into poten-
tially amyloidogenic fragments (Graham et al.,
1995, 1996). In support of this hypothesis, both
axonal APP accumulation and long-term accumu-
lation of Ab has been reported in axons following
TBI (Iwata et al., 2002) suggesting that damaged
axons may provide a key source of Ab following
TBI (Uryu et al., 2004). Co-accumulation of Ab
with APP in swollen axons and neuronal cell
bodies has been found within days following
trauma in a pig model of diffuse axonal injury
(Smith et al., 1999), and the same group also found
widespread axonal Ab accumulation associated
with Ab plaques in brain-injured humans with
diffuse axonal injury (Smith et al., 2003). It has
also been demonstrated that BACE and presenilin-1
co-localize with APP and Ab within the damaged
axons and that BACE and presenilin 1 mediate the
Ab production from APP within the axonal
membrane (Kamal et al., 2001). This was later
confirmed by Chen et al. (2004) in a pig model who
demonstrated that the Ab accumulation in the
swollen axons persisted for 6 months following
injury in the subcortical white matter and basal
ganglia. They concluded that TBI leading to
impaired axonal transport induces a long-term
pathological co-accumulation of APP with BACE,
presenilin 1 and activated cysteine-dependent
aspartate-specific proteases (caspases), thereby
providing a possible mechanism for APP cleavage
and production of Ab within axons following TBI.
Following neuronal cell injury, activation of
caspase-3 (Raghupathi et al., 2000) occurs in par-
allel with increased APP accumulation and Ab
production in several different experimental para-
digms (Stone et al., 2002), suggesting a potential
association between caspase cleavage of APP,
caspase-mediated cell death, increased production
of Ab and subsequent AD. Similarly, excito-
toxic neuronal injury and global ischaemia result
in accumulation and colocalization of activated
caspase-3, caspase-cleaved APP fragments and
apoptotic (TUNEL) labelling in affected hippo-
campal neurons and plaque-associated neurites in
AD brains (Gervais et al., 1999). Therefore, the
observations in TBI that excitotoxic neuronal
injury and global ischaemia resemble changes seen
in dying neurons and plaque-associated neurites in
AD brains (Masliah et al., 1998), supports the
hypothesis that TBI induced caspase activation
may potentially result in increased APP proteolysis
and may directly or indirectly influence Ab levels.
Evidence for the role of caspase-3 in APP cleavage
and Ab production has come from recent studies
examining the effects of caspase inhibition follow-
ing TBI in mice with the human Ab coding
sequence ‘knocked in’ to their endogenous APP
gene. Caspase inhibitors prevented the elevated level
of Ab normally seen in these animals as well as re-
ducing hippocampal cell death, although attenua-
tion of cell death could not be completely attributed
to less Ab formation (Abrahamson et al., 2006).
Further human autopsy and experimental stud-
ies are required to confirm these findings and to
ascertain the relevance of this co-accumulation
within neuronal cell bodies and axons and to
better elucidate the exact mechanisms involved in
APP cleavage and subsequent deposition of Ab
plaques.
Alzheimer’s pathology following fatal head injury in
humans
Preliminary evidence implicating a possible role
for TBI in the development of AD came from an
early case report documenting early-onset classic
AD pathology in a 38-year-old man who had
suffered a previous single episode of severe head
trauma 16 years ago (Rudelli et al., 1982). Since
then, studies of the brains of boxers suffering from
dementia pugilistica have also demonstrated
AD-like pathology with diffuse Ab plaque depo-
sition (Roberts et al., 1990). We can speculate that
such Ab deposition resulted from repeated blows
to the head over a long period of time, and that

such events may also occur in the brains of head-
injured individuals. Histopathological studies of
such individuals who died after suffering a single
severe TBI demonstrate widespread cerebral Ab
deposition in short (Roberts et al., 1991, 1994;
Gentleman et al., 1997; Ikonomovic et al., 2004)
and long term (Clinton et al., 1991) survivors ir-
respective of age. In contrast to these studies,
Adle-Biassette et al. (1996) were unable to detect
any Ab deposits in head injury cases of individuals
below the age of 63 years, in a similar age range
and survival time to those studied by Roberts et al.
(1991). Also, the extensive non-selective autopsy
investigation by Braak and Braak (1997) rarely
detected Ab plaques in the younger subjects (be-
low the age of 40 years). None of the 58 cases
examined in the 31–35 years age range, and only
two of the 83 subjects in the 36–40 years age range
had some Ab plaques. Furthermore, Ab deposits
were detected in only 20 of the 207 people in their
fifth decade of life (Braak and Braak, 1997).
Therefore, significant discrepancies exist in neuro-
pathological studies.
There also appears to be several morphological
similarities, as well as differences, between the Ab
plaques in AD and those found deposited in the
brain following TBI. In AD, there are numerous,
compact and hard core Ab deposits as well as
neurofibrillary tangles and neuropil threads,
whereas following TBI there appears to be a
higher incidence of diffuse Ab plaques (Clinton
et al., 1991; Roberts et al., 1991, 1994; Ikonomovic
et al., 2004). Exactly what this morphological
difference in plaques means is unknown. One sim-
ilarity between the deposits seen in TBI and AD is
that they are both composed primarily of Ab1–42.
These deposits appear much more abundantly
than Ab1–40 deposits seen in TBI (Gentleman
et al., 1997; Horsburgh et al., 2000a; Ikonomovic
et al., 2004) or in AD (Iwatsubo et al., 1994).
Another consequence of APP gene mutations is
elevated Ab levels in the cerebrospinal fluid (CSF)
in AD (Nakamura et al., 1994). Studies have since
aimed to determine if CSF levels of Ab are simi-
larly increased following TBI. Raby et al. (1998)
reported significantly elevated levels of Ab1–42 in
the CSF in six patients (age 19–51 years) following
severe TBI, an observation later confirmed by
307
Olsson et al. (2004) who reported elevated levels of
Ab1–42 and APP in ventricular CSF of severely
injured TBI patients. These groups postulate that
the increased levels of Ab1–42 in the CSF are
directly related to the elevated Ab levels in the
brain (Raby et al., 1998). It has also been specu-
lated that the rise in Ab1–42 in both brain and
CSF may be a direct result of the increased APP
levels in neuronal cell bodies and axons after TBI
(McKenzie et al., 1996). In direct contrast, a recent
study of 29 TBI patients found a reduction in
Ab1–42 levels in CSF, which appeared to correlate
with poor outcome after TBI (Franz et al., 2003).
Future studies are required to validate these CSF
findings and to resolve the contradictions. It is
important to determine how relevant the CSF
concentration of Ab1–42 is as a marker of poten-
tial AD, and whether monitoring CSF levels might
prove helpful in assessing the degree of neuronal
injury resulting from TBI. There is also the
potential of identifying those individuals poten-
tially at risk of developing AD later in life (Raby
et al., 1998).
Alzheimer’s pathology following experimental head
injury
Studies using different forms of experimental TBI
have given some insight into how brain injury may
lead to AD, although results remain somewhat in-
conclusive. In support of the argument that in-
creased APP levels following TBI may potentiate
AD pathology, experimental TBI in rats induced
overexpression and accumulation of APP in the
cerebral cortex and hippocampus, which subse-
quently led to neuronal degeneration in the CA3
region of the hippocampus as early as 3 days post-
injury (Murakami et al., 1998). Also, a pig model
of rotational head injury developed rapid Ab
accumulation, manifested in axonal bulbs and
diffuse plaques 3 days after trauma (Smith et al.,
1999). However, post-traumatic Ab deposition
has not been observed in the majority of non-
transgenic animal studies where most failed
to identify any Ab staining (Pierce et al., 1996,
1998; Masumura et al., 2000; Laurer et al., 2001;
Ciallella et al., 2002; Hamberger et al., 2003).
308
Experimental results from APP transgenic mice
also remain inconclusive. Transgenic mice with
AD-like pathology (platelet-derived-APP; PD-APP),
which overexpress human APP by 10-fold and
form amyloid plaques by 6 months of age, enable
the interaction between TBI and AD to be studied.
TBI in these animals resulted in a transient rise in
Ab levels, particularly in the hippocampus. This
elevation in Ab correlated with a much greater loss
of CA3 hippocampal neurones than injured wild
type mice, which also had a significant loss of
hippocampal neurones compared to uninjured
controls (Smith et al., 1998). However, TBI did
not lead to premature amyloid plaque formation
in these transgenic mice, and on long term exam-
ination there was a reported reduction in amyloid
plaques ipsilateral to the injury (Smith et al., 1998;
Nakagawa et al., 1999). Paradoxically, TBI in-
duced a regression of previously existing plaques
in aged PD-APP-transgenic mice (Nakagawa
et al., 2000). Thus, elevated Ab levels, which are
expected to be seen in the APP-transgenic mice,
resulted in hippocampal neuronal death but the
TBI did not accelerate Ab deposition. Interest-
ingly, in transgenic mice that overexpress human
APP by only twofold, Ab deposition did not occur
following TBI and similar cognitive and motor
deficits to wild-type mice were observed (Murai
et al., 1998). In contrast, repetitive TBI in a
Tg2576 APP-transgenic mice model did result in
greater Ab deposition and an increase in the pro-
duction of both soluble and insoluble cortical
Ab40 and Ab42, which may be accounted for by
the higher levels of oxidative stress in repetitive
TBI (Uryu et al., 2002). Normally, repetitive TBI
in animals does not lead to the formation of ne-
urofibrillary tangles. Yet in a study by Yoshiyama
et al. (2005), repetitive TBI in one of 12 mice
resulted in the development of neurofibrillary
tangles, which correlated with a remarkably poor
neurobehavioural score. Consequently the pres-
ence of iron deposits, which also occurs in AD,
was thought to contribute to the formation of
these neurofibrillary tangles (Yoshiyama et al.,
2005).
The results of TBI studies in transgenic models
of AD are also in conflict with that seen in human
TBI. Firstly, rapid Ab deposition has not been
demonstrated in any of the described transgenic
models, unlike human studies (Roberts et al., 1991,
1994; Gentleman et al., 1997; Ikonomovic et al.,
2004). Secondly, an increased severity of injury
does not result in increased Ab deposition; if
anything it actually seems to correlate with re-
duced Ab deposition or even resolution of already
established plaques, as reviewed by Szczygielski
et al. (2005). This trend seems to be reversed in
humans, since both the risk (Plassman et al., 2000)
and the post-traumatic Ab deposition increased
with TBI severity (Roberts et al., 1994).
Epidemiological studies
The epidemiological studies that have reported on
the relationship between TBI and AD are also
contradictory, and accordingly, the association
still remains inconclusive. Some reports suggest a
positive association with AD (French et al., 1985;
Mortimer et al., 1985, 1991; Graves et al., 1990;
Van Duijn et al., 1992; Rasmusson et al., 1995;
Salib and Hillier, 1997) while other studies were
unable to confirm head trauma as a risk factor
for AD (Katzman et al., 1989; Li et al., 1992;
Fratiglioni et al., 1993). With respect to the case-
control studies, one of the earliest showed that
patients with dementia of the Alzheimer type had a
significantly greater incidence of antecedent head
trauma (French et al., 1985). Subsequent pooled
re-analysis of some of the case-control studies
has shown positive associations between the de-
velopment of AD and a history of head trauma,
particularly amongst males (Mortimer et al., 1991;
Fleminger et al., 2003). Specifically, the meta-
analysis of seven case-control studies by Mortimer
et al. (1991) allowed for a more powerful statistical
investigation of the association between TBI and
AD and provided the first convincing evidence in
support of a strong association between these two
pathologies. They achieved a high level of statis-
tical power (0.92) as opposed to the low mean
statistical power of the individual studies (0.22).
The findings of the EURODERM re-analysis
highlighted the importance of designing studies
with adequate statistical power and suggested that
many of the early negative findings stem from the

fact that they suffered from insufficient statistical
power (see review by Lye and Shores (2000)). This
is particularly important when the risk factor
under investigation, in these cases TBI, occurs in-
frequently in the general population. Recent case-
control studies have refined their designs to avoid
the methodological flaws found in earlier studies
and have shown significant associations between
AD and TBI (Graves et al., 1990; Van Duijn et al.,
1992; Rasmusson et al., 1995; Salib and Hillier,
1997). An additional strength of the study by
Rasmusson et al. (1995) was that it examined the
association between AD and TBI of any severity,
concluding that even mild head injury may serve
as a predisposing factor for some cases of AD.
More recently, a detailed systematic review of
case-control studies conducted over the past 10
years sought to replicate those findings, and sup-
ported an association between head injury and
AD, but only in males (Fleminger et al., 2003). The
possible explanation for the gender differences in
the risk of AD following TBI may be attributed to
the neuroprotective and neuroregenerative effects
of the female hormones oestrogen and progester-
one (Stein, 2001), or it could be merely the fact
that men typically suffer more severe injuries than
women. The association between TBI and early
onset of dementia was also demonstrated in a
recent population-based study of fall-related TBI,
concluding that fall-related TBI may hasten,
increase or even trigger the onset of AD (Luukinen
et al., 2005).
As with the case-control studies, there have been
conflicting reports in cohort studies. For example,
no significant association was found between head
injury and the risk of developing AD in the early
work (Katzman et al., 1989; Williams et al., 1991).
Also, more recently a history of head trauma with
unconsciousness was excluded as a risk factor for
AD in both the Rotterdam Study (Mehta et al.,
1999) and the large EURODERM study (Launer
et al., 1999). The EURODERM study was a
pooled analysis of four European population-
based studies of individuals 65 years and older.
Plassman et al. (2000) examined the association
between early adult head injury, as documented
by military hospital records, and dementia in late
life. In the study of 548 world war II (US Navy)
309
brain-injured veterans and 1228 age-matched non-
injured controls, follow up after 50 years revealed
that moderate to severe head injuries in young men
may be associated with increased risk of AD and
other dementias in late life. These findings were
supported by the MIRAGE study, which is the
largest study to date (Guo et al., 2000). This study
analysed 2233 definite and probable AD patients,
and 14,668 first-degree relatives and showed that
head injury with loss of consciousness significantly
increased the AD risk. The magnitude of risk was
proportional to injury severity and was heightened
amongst first-degree relative of AD patients (Guo
et al., 2000).
Several studies have shown that there is a
greater risk of developing AD when head injury
has occurred in later life (within 10 years of onset
of AD) as opposed to head injury occurring earlier
in life (beyond 10 years of onset of AD) (Graves
et al., 1990; Mortimer et al., 1991; Van Duijn et al.,
1992). In contrast, other studies have shown that
head trauma occurring mainly in younger child-
hood is associated with an increased risk of AD in
later life (Schofield et al., 1997; Plassman et al.,
2000). After examining the incidence of AD pa-
thology in 58 consecutive patients with residual
closed TBI lesions and the frequency of TBI
residuals in 57 age-matched autopsy controls,
Jellinger et al. (2001) concluded that severe TBI
may have some influence on the development of
AD irrespective of age when the TBI occurred. For
an excellent detailed review of case-control studies
from 1984 to 1997 see Lye and Shores (2000).
A related hypothesis concerns the relationship
between TBI and AD focussing on the time delay
between TBI and the onset of AD. These studies
have highlighted the importance of considering
more specific factors such as duration of loss of
consciousness, severity of injury and the time of
onset of AD symptoms. Cohort studies have sug-
gested that TBI may interact with other risk fac-
tors to hasten the onset of AD in persons
susceptible to the disease (Sullivan et al., 1987;
Gedye et al., 1989; Schofield et al., 1997; Nemetz
et al., 1999). Of 17 AD cases reported, there was a
younger mean age of onset for cases with a history
of head injury than for those cases without it
(Sullivan et al., 1987). In a larger study of 148
310
referral patients with confirmed probable AD,
Gedye et al. (1989) found that the mean age of AD
onset was significantly younger for those with a
history of TBI. Also, in a slightly larger commu-
nity based longitudinal study of aging in 271 par-
ticipants in North Manhattan, participants who
suffered loss of consciousness exceeding 5 min fol-
lowing a head trauma were at a significantly in-
creased risk of earlier onset AD (Schofield et al.,
1997). Furthermore, among 1283 TBI cases in
Olmsted County, 31 patients developed AD. Given
the incidence of AD in the community, this
number was not unexpected, however the time be-
tween TBI and onset of AD was less than expected
suggesting that TBI reduces the time of onset
of AD among persons at risk of developing AD
(Nemetz et al., 1999). In contrast, Rasmusson
et al. (1995) did not find any effect of head injury
on age of onset of AD in a cohort of 68 AD cases,
nor did the MIRAGE study (Guo et al., 2000).
The influence of APOE genotype on TBI outcome
Following reports that Ab is not found in all head-
injured brains, it was speculated that it may only
be deposited in the brains of people who are ge-
netically susceptible to developing AD, for exam-
ple, those individuals possessing an apolipoprotein
E4 (APOE e4) allele. In 1993, it was demonstrated
that the APOE e4 allele was a risk factor for AD
(Strittmatter et al., 1993) and the association bet-
ween APOE e4 with late onset familial and spo-
radic AD has now been confirmed worldwide
(Corder et al., 1993; Saunders et al., 1993; Mayeux
et al., 1995; Nicoll et al., 1995; Horsburgh et al.,
2000b). APOE protein has been co-localized with
the neuropathological hallmarks of AD, including
neurofibrillary tangles, Ab plaques and amyloid
angiopathy (Namba et al., 1991; Rebeck et al.,
1993). It is now known to be a major susceptibility
factor associated with approximately 40–50% of
sporadic and familial AD compared with 30% of
the normal population (Roses, 1996).
While it is known that APOE is a key factor
involved in lipid transport within the human cen-
tral nervous system (CNS), the three alleles (2, 3
and 4) have significantly varied effects within the
CNS (Nicoll, 1996). The e4 allele has been linked
to impaired branching and growth of neurites and
it is therefore suggested to exacerbate neurological
disturbances (Kerr and Kraus, 1998). Numerous
studies have since confirmed this association by
demonstrating that Ab deposition following head
injury occurs more prominently in those who pos-
sess an APOE e4 allele. Also, clinical studies of
TBI have shown that possession of APOE e4 is
associated with a relatively poor outcome (Sorbi
et al., 1995; Teasdale et al., 1997). In the study by
Teasdale et al. (1997), 57% of APOE e4 carriers
had an unfavourable outcome (defined as dead,
vegetative state or severe disability) compared with
27% of non-carriers of APOE e4. The relationship
between functional recovery in patients with head
injury and APOE e4 genotype was later confirmed
in an independent study (Lichtman et al., 2000),
with rehabilitation outcome for TBI survivors be-
ing adversely affected in the presence of the APOE
allele. It was subsequently shown that APOEe4
individuals were 10 times more likely to develop
AD after TBI than those who did not posses the
allele (Mayeux et al., 1995). A recent case study
has demonstrated that possession of APOE e4 is
associated with a greater incidence of moderate/
severe contusional injury and severe ischaemic
damage in fatal cases of TBI (Smith et al., 2006).
These authors postulate that APOE has a role in
cerebrovascular and haematological mechanisms.
Animal studies, using APOE knockout and
transgenic mice, have also provided further infor-
mation about APOE mechanisms and the response
of the brain to injury (as reviewed by Horsburgh
et al. (2000b)). Some studies have shown that the
APOE molecule may have a direct neurotoxic role
(Neve and Robakis, 1998), whereas others have
speculated that APOE e4 may directly interact
with Ab and impact on the metabolism of APP
(Growdon, 1998). Transgenic mice that express
human APOE e4 have poorer outcome and greater
mortality following TBI than mice expressing
human APOE e3 (Sabo et al., 1999). TBI mice
expressing APOE e4 also had a decrease in neuro-
protective sAPPa within the hippocampus,
whereas mice expressing APOE e3 had an increase
in sAPPa in the cortex. These studies concluded
that the alteration in APP metabolism contributed

to the increased mortality and poorer outcome in
APOE e4 transgenic mice, and the greater recovery
and less cortical damage in APOE e3 transgenic
mice (Ezra et al., 2003). In addition, transgenic
APOE e4 mice, which also overexpress APP, had
accelerated deposition of Ab following injury,
suggesting APOE e4 may reduce the clearance of
Ab, thereby favouring its deposition (Hartman
et al., 2002).
In contrast to the studies demonstrating an
association between APOE e4 and an unfavoura-
ble outcome following TBI, a number of recent
studies examining both severe and mild TBI have
failed to support such findings (Guo et al., 2000;
Plassman et al., 2000; Jellinger et al., 2001;
Chamelian et al., 2004). Therefore there is no con-
clusive evidence linking APOE genotype with the
development of AD following TBI (Jellinger,
2004). Further experimental studies and larger
clinicopathological and prospective human studies
are warranted to clarify the relationship between
TBI, APOE genotype and AD. Additional studies
with APOE transgenic mice are also expected to
contribute to the greater understanding of the
pathophysiological role of APOE in the brain.
Conclusion
Although findings to date have been contradic-
tory, there have been major advances from epide-
miological studies and both experimental and
human autopsy studies regarding the relationship
between TBI and AD, and the influence that the
APOE e4 genotype has on this putative associa-
tion. Human studies have certainly supported the
hypothesis that either repeated mild to moderate
TBI, or moderate to severe single TBI with loss of
consciousness, are risk factors for the subsequent
development of AD. Moreover, those patients who
have the APOE e4 are at greater risk of a poorer
outcome following TBI. However, the exact mech-
anism by which APOE genotype influences out-
come and increases risk of AD following TBI is
unclear. Experimental and human studies have
also found accumulation of Ab and APP within
damaged axons following TBI, but the exact im-
plications of this and exactly how this translates
311
into release of Ab from axons, Ab plaque forma-
tion and toxicity remains unclear. Despite some
experimental and human studies suggesting that
APP overexpression following TBI may lead to Ab
formation, there are numerous studies which have
failed to demonstrate any such relationship.
Accordingly, further investigation is required,
particularly with larger prospective cohort studies
of TBI survivors and larger autopsy series before
an association between TBI and onset of AD can
be definitively supported.
References
Abrahamson, E.E., Ikonomovic, M.D., Ciallella, J.R., Hope,
C.E., Paljug, W.R., Isanski, B.A., Flood, D.G., Clark, R.S.
and Dekosky, S.T. (2006) Caspase inhibition therapy abol-
ishes brain trauma-induced increases in Abeta peptide: im-
plications for clinical outcome. Exp. Neurol., 197: 437–450.
Adle-Biassette, H., Duyckaerts, C., Wasowicz, M., He, Y.,
Fornes, P., Foncin, J.F., Lecomte, D. and Hauw, J.J. (1996)
BAP deposition and head trauma. Neurobiol. Aging, 17:
415–419.
Andersen, O.M., Schmidt, V., Spoelgen, R., Gliemann, J.,
Behlke, J., Galatis, D., Mckinstry, W.J., Parker, M.W.,
Masters, C.L., Hyman, B.T., Cappai, R. and Willnow, T.E.
(2006) Molecular dissection of the interaction between
amyloid precursor protein and its neuronal trafficking
receptor SorLA/LR11. Biochemistry, 45: 2618–2628.
Arbel, M., Yacoby, I. and Solomon, B. (2005) Inhibition of
amyloid precursor protein processing by b-secretase through
site-directed antibodies. PNAS, 102: 7718–7723.
Awasthi, A., Matsunaga, Y. and Yamada, T. (2005) Amyloid-
beta causes apoptosis of neuronal cells via caspase cascade,
which can be prevented by amyloid-beta-derived short
peptides. Exp. Neurol., 196: 282–289.
Beyreuther, K., Pollwein, P., Multhaup, G., Monning, U.,
Konig, G., Dyrks, T., Schubert, D. and Masters, C.L. (1993)
Regulation and expression of the Alzheimer’s B/A4 amyloid
protein precursor in health, disease and Down’s Syndrome.
Ann. N.Y. Acad. Sci., 695: 91–102.
Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H.,
Simpson, D.A. and Mclean, A.J. (1995) Topography of
axonal injury as defined by amyloid precursor protein and
the sector scoring method in mild and severe closed head
injury. J. Neurotrauma, 12: 565–572.
Braak, H. and Braak, E. (1997) Frequency of stages of
Alzheimer-related lesions in different age categories.
Neurobiol. Aging, 18: 351–357.
Bramlett, H.M., Kraydieh, S., Green, E.J. and Dietrich, W.D.
(1997) Temporal and regional patterns of axonal damage
following traumatic brain injury: a beta-amyloid precursor
protein immunocytochemical study in rats. J. Neuropathol.
Exp. Neurol., 56: 1132–1141.
312
Chamelian, L., Reis, M. and Feinstein, A. (2004) Six-month
recovery from mild to moderate traumatic brain injury: the
role of APOE-4 allele. Brain, 127: 2621–2628.
Chen, X.H., Siman, R., Iwata, A., Meaney, D.F., Trojanowski,
J.Q. and Smith, D.H. (2004) Long-term accumulation of
amyloid-beta, beta-secretase, presenilin-1, and caspase-3 in
damaged axons following brain trauma. Am. J. Pathol., 165:
357–371.
Ciallella, J.R., Ikonomovic, M.D., Paljug, W.R., Wilbur, Y.I.,
Dixon, C.E., Kochanek, P.M., Marion, D.W. and Dekosky,
S.T. (2002) Changes in expression of amyloid precursor
protein and interleukin-1beta after experimental traumatic
brain injury in rats. J. Neurotrauma, 19: 1555–1567.
Clinton, J., Ambler, M.W. and Roberts, G.W. (1991) Post
traumatic Alzheimer’s disease: preponderance of a single
plaque type. Neuropathol. Appl. Neurobiol., 17: 69–74.
Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechal,
D.E., Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L.
and Pericak Vance, M.A. (1993) Gene dosage of apolipo-
protein E type allele and the risk of Alzheimer’s disease in
late onset families. Science, 261: 921–923.
Dewachter, I. and Van Leuven, F. (2002) Secretases as targets
for the treatment of Alzheimer’s disease: the prospects.
Lancet Neurol., 1: 409–416.
Ezra, Y., Oron, L., Moskovich, L., Roses, A.D., Beni, S.M.,
Shohami, E. and Michaelson, D.M. (2003) Apolipoprotein
E4 decreases whereas apolipoprotein E3 increases the level of
secreted amyloid precursor protein after closed head injury.
Neuroscience, 121: 315–325.
Finfer, S.R. and Cohen, J. (2001) Severe traumatic brain injury.
Resuscitation, 48: 77–90.
Fleminger, S., Oliver, D.L., Lovestone, S., Rabe-Hesketh, S.
and Giora, A. (2003) Head injury as a risk factor for
Alzheimer’s disease: the evidence 10 years on; a partial
replication. J. Neurol. Neurosurg. Psychiatry, 74: 857–862.
Franz, G., Beer, R., Kampfl, A., Engelhardt, K., Schmutzhard,
E., Ulmer, H. and Deisenhammer, F. (2003) Amyloid beta
1–42 and tau in cerebrospinal fluid after severe traumatic
brain injury. Neurology, 60: 1457–1461.
Fratiglioni, L., Ahlbom, A., Viitanen, M. and Winblad, B.
(1993) Risk factors for late-onset alzheimer’s disease: a pop-
ulation-based, case-control study. Ann. Neurol., 33: 258–266.
French, L.R., Schuman, L.M., Mortimer, J.A., Hutton, J.T.,
Boatman, R.A. and Christians, B. (1985) A case-control
study of dementia of the Alzheimer type. Am. J. Epidemiol.,
121: 414–421.
Gedye, A., Beattie, B.L., Tuokko, H., Horton, A. and Korsarek,
E. (1989) Severe head injury hastens age of onset of
Alzheimer’s disease. J. Am. Geriatr. Soc., 37: 970–973.
Gentleman, S.M., Greenberg, B.D., Savage, M.J., Noori, M.,
Newman, S.J., Roberts, G.W., Griffin, W.S. and Graham, D.I.
(1997) AB42 is the predominant form of amyloid-b-protein in
the brains of short-term survivors of head injury. Neuroreport,
8: 1519–1522.
Gervais, F.G., Xu, D., Robertson, G.S., Vaillancourt, J.P., Zhu,
Y., Huang, J., Leblanc, A., Smith, D., Rigby, M., Shearman,
M.S., Clarke, E.E., Zheng, H., Van Der Ploeg, L.H.,
Ruffolo, S.C., Thornberry, N.A., Xanthoudakis, S., Zam-
boni, R.J., Roy, S. and Nicholson, D.W. (1999) Involvement
of caspases in proteolytic cleavage of Alzheimer’s amyloid-
beta precursor protein and amyloidogenic A beta peptide
formation. Cell, 97: 395–406.
Graham, D.I., Gentleman, S.M., Lynch, A. and Roberts, G.W.
(1995) Distribution of beta-amyloid protein in the brain
following severe head injury. Neuropathol. Appl. Neurobiol.,
21: 27–34.
Graham, D.I., Gentleman, S.M., Nicoll, J.A., Royston, M.C.,
McKenzie, J.E., Roberts, G.W. and Griffin, W.S. (1996)
Altered beta-APP metabolism after head injury and its
relationship to the aetiology of Alzheimer’s disease. Acta
Neurochir. Suppl. (Wien.), 66: 96–102.
Graves, A.B., White, E., Koepsell, T.D., Reifler, B.V., Van
Belle, G., Larson, E.B. and Raskind, M. (1990) The associ-
ation between head trauma and Alzheimer’s disease. Am. J.
Epidemiol., 131: 491–501.
Growdon, J.H. (1998) Apolipoprotein E and Alzheimer’s
disease. Arch. Neurol., 44: 1053–1054.
Guo, Z., Cuppels, L.A., Kurz, A., Auerbach, S.H., Volicer, L.,
Chui, H., Green, R.C., Sadovnick, A.D., Durara, R.,
Decarli, C., Johnson, K., Go, R.C., Growdon, J.H., Haines,
J.L., Kukull, W.A. and Farrer, L.A. (2000) Head injury
and risk of AD in the MIRAGE study. Neurology, 54:
1316–1323.
Hamberger, A., Huang, Y.L., Zhu, H., Bao, F., Ding, M.,
Blennow, K., Olsson, A., Hansson, H.A., Viano, D. and
Haglid, K.G. (2003) Redistribution of neurofilaments and
accumulation of beta-amyloid protein after brain injury by
rotational acceleration of the head. J. Neurotrauma, 20:
169–178.
Hardy, J.A. and Selkoe, D.J. (2002) The amyloid hypothesis of
Alzheimer’s disease; progress and problems on the road to
therapeutics. Science, 297: 353–356.
Hartman, R.E., Laurer, H., Longhi, L., Bales, K.R., Paul,
S.M., McIntosh, T.K. and Holtzman, D.M. (2002) Apolipo-
protein E4 influences amyloid deposition but not cell loss
after traumatic brain injury in a mouse model of Alzheimer’s
disease. J. Neurosci., 22: 10083–10087.
Horsburgh, K., Cole, G.M., Yang, F., Savage, M.J.,
Greenberg, B.D., Gentleman, S.M., Graham, D.I. and
Nicoll, J.A.R. (2000a) b-Amyloid (Ab)42(43), Ab42, Ab40
and apoE immunostaining of plaques in fatal head injury.
Neuropath. Appl. Neurobiol., 26: 124–132.
Horsburgh, K., Mccarron, M.O., White, F. and Nicoll, J.A.
(2000b) The role of apolipoprotein E in Alzheimer’s disease,
acute brain injury and cerebrovascular disease: evidence of
common mechanisms and utility of animal models.
Neurobiol. Aging, 21: 245–255.
Ikonomovic, M.D., Uryu, K., Abrahamson, E.E., Ciallella,
J.R., Trojanowski, J.Q., Lee, V.M.-Y., Clark, R.S., Marion,
D.W., Wisniewski, S.R. and Dekosky, S.T. (2004)
Alzheimer’s pathology in human temporal cortex surgically
excised after severe brain injury. Exp. Neurol., 190: 192–203.
Iwata, A., Chen, X.H., McIntosh, T.K., Browne, K.D. and
Smith, D.H. (2002) Long-term accumulation of amyloid-beta

in axons following brain trauma without persistent upregu-
lation of amyloid precursor protein genes. J. Neuropathol.
Exp. Neurol., 61: 1056–1068.
Iwatsubo, T., Odaka, A., Suzuki, N., Mizusawa, H., Nukina, N.
and Ihara, Y. (1994) Visualization of AB42 and AB40 in
senile plaques with end specific AB monoclonals: evidence
that an initially deposited species is AB42. Neuron, 13: 45–53.
Janus, C., Pearson, J., Mclaurin, J., Mathews, P.M., Jiang, Y.,
Schmidt, S.D., Chishti, M.A., Horne, P., Heslin, D., French,
J., Mount, H.T., Nixon, R.A., Mercken, M., Bergeron, C.,
Fraser, P.E., St George-Hyslop, P. and Westaway, D. (2000)
A beta peptide immunization reduces behavioural impair-
ment and plaques in a model of Alzheimer’s disease. Nature,
408: 979–982.
Jellinger, K.A. (2004) Head injury and dementia. Curr. Opin.
Neurol., 17: 719–723.
Jellinger, K.A., Paulus, W., Wrocklage, C. and Litvan, I. (2001)
Effects of closed traumatic brain injury and genetic factors
on the development of Alzheimer’s disease. Eur. J. Neurol., 8:
707–710.
Kamal, A., Almenar-Queralt, A., Leblanc, J.F., Roberts, E.A.
and Goldstein, L.S. (2001) Kinesin-mediated axonal trans-
port of a membrane compartment containing b-secretase and
presenilin-1 requires APP. Nature, 414: 643–648.
Katzman, R., Aronson, M., Fuld, P., Kawas, C., Brown, T.,
Morgenstern, H., Frishman, W., Gidez, L., Eder, H. and
Ooi, W.L. (1989) Development of dementing illness in an
80-year-old volunteer cohort. Ann. Neurol., 25: 317–324.
Kerr, M.E. and Kraus, M. (1998) Genetics and the central
nervous system: apolipoprotein E and brain injury. AACN
Clin. Issues, 9: 524–530.
King, G.D., Cherian, K. and Turner, S.R. (2004) X11a impairs
g- but not b-cleavage of amyloid precursor protein.
J. Neurochem., 88: 971–982.
Kraus, J.F. (1993) Epidemiology of head injury. In: Cooper
P.R. (Ed.), Head Injury (3rd ed). Williams and Wilkins,
Baltimore.
Lammich, S., Kojro, E., Postina, R., Gilbert, S., Pfeiffer, R.,
Jasionowski, M., Haass, C. and Fahrenholz, F. (1999) Con-
stitutive and regulated a-secretase cleavage of Alzheimer’s
amyloid precursor protein by a disintegrin metalloprotease.
Proc. Natl. Acad. Sci. U.S.A., 96: 3922–3927.
Launer, L.J., Andersen, K., Dewey, M.E., Letenneur, L.,
Ott, A., Amaducci, L.A., Brayne, C., Copeland, J.R., Dart-
igues, J.F., Kragh-Sorensen, P., Lobo, A., Martinez-Lage,
J.M., Stijnen, T. and Hofman, A. (1999) Rates and risk
factors for dementia and Alzheimer’s disease: results from
EURODEM pooled analyses. EURODEM Incidence
Research Group and Work Groups. European Studies of
Dementia. Neurology, 52: 78–84.
Laurer, H.L., Bareyre, F.M., Lee, V.M., Trojanowski, J.Q.,
Longhi, L., Hoover, R., Saatman, K.E., Raghupathi, R.,
Hoshino, S., Grady, M.S. and McIntosh, T.K. (2001) Mild
head injury increasing the brain’s vulnerability to a second
concussive impact. J. Neurosurg., 95: 859–870.
Lee, E.B., Leng, L.Z., Zhang, B., Kwong, L., Trojanowski,
J.Q., Abel, T. and Lee, V.M.Y. (2006) Targeting amyloid-b
313
peptide (Ab) oligomers by passive immunization with a con-
formation-selective monoclonal antibody improves learning
and memory in Ab precurosr protein (APP) transgenic mice.
J. Biol. Chem., 281: 4292–4299.
Lee, J. and Wurtman, R.J. (2000) Regulation of APP synthesis
and secretion by neuroimmunophilin ligands and cycloox-
ygenase inhibitors. Ann. N.Y. Acad. Sci., 920: 261–268.
Lee, R.K.K., Araki, W. and Wurtman, R.J. (1997) Stimulation
of amyloid precursor protein synthesis by adrenergic recep-
tors coupled to cAMP formation. Proc. Natl. Acad. Sci.
U.S.A., 94: 5422–5426.
Li, G., Shen, Y.C., Li, Y.T., Chen, C.H., Zhau, Y.W. and
Silverman, J.M. (1992) A case-control study of Alzheimer’s
disease in China. Neurology, 42: 1481–1488.
Lichtman, S.W., Seliger, G., Tycko, B. and Marder, K. (2000)
Apolipoprotein E and functional recovery from brain injury
following post acute rehabilitation. Neurology, 55: 1536–1539.
Luukinen, H., Viramo, P., Herala, M., Kervinen, K.,
Kesaniemi, Y.A., Savola, O., Winqvist, S., Jokelainen, J.
and Hillbom, M. (2005) Fall-related injuries and the risk of
dementia in elderly people: a population-based study. Eur. J.
Neurol., 12: 86–92.
Lye, T.C. and Shores, A.E. (2000) Traumatic brain injury as a
risk factor for Alzheimer’s disease: a review. Neuropsychol.
Rev., 10: 115–129.
Masliah, E., Mallory, M., Alford, M. and Hansen, L.A. (1998)
Caspase dependent DNA fragmentation might be associated
with excitotoxicity in Alzheimer’s disease. J. Neuropathol.
Exp. Neurol., 57: 1041–1052.
Masumura, M., Hata, R., Uramoto, H., Murayama, N.,
Ohno, T. and Sawada, T. (2000) Altered expression of
amyloid precursor proteins after traumatic brain injury in
rats: in situ hybridization and immunochemical study.
J. Neurotrauma, 17: 123–134.
Mattson, M.P. (1997) Cellular actions of b-amyloid precursor
protein and it’s soluble and fibrillogenic derivatives. Physiol.
Rev., 77: 1081–1132.
Mattson, M.P. (2004) Pathways towards and away from
Alzheimer’s disease. Nature, 430: 631–639.
Mattson, M.P., Cheng, B., Culwell, A.R., Esch, F.S.,
Lieberburg, I. and Rydel, R.E. (1993) Evidence for excito-
protective and intraneuronal calcium-regulating roles for
secreted forms of the beta-amyloid precursor protein.
Neuron, 10: 243–254.
Mayeux, R., Ottman, R., Maestre, G., Ngai, C., Tang, M.X.,
Ginsberg, H., Chun, M., Tycko, B. and Shelanski, M. (1995)
Synergistic effects of traumatic head injury and apolipopro-
tein E4 in patients with Alzheimer’s disease. Neurology, 45:
555–557.
McKenzie, J.E., Gentleman, S.M., Roberts, G.W., Graham,
D.I. and Royston, M.C. (1994) Increased numbers of beta
APP-immunoreactive neurons in the entorhinal cortex after
head injury. Neuroreport, 6: 161–164.
McKenzie, K.J., Mclellan, D.R., Gentleman, S.M., Maxwell,
W.L., Gennarelli, T.A. and Graham, D.I. (1996) Is X-APP a
marker of axonal damage in short-surviving head injury?
Acta Neuropathol. (Berl.), 92: 608–613.
314
Mehta, K.M., Ott, A., Kalmijn, S., Slooter, A.J., Van Duijn,
C.M., Hofman, A. and Breteler, M. (1999) Head trauma and
risk of dementia and Alzheimer’s disease: the Rotterdam
study. Neurology, 53: 1959–1962.
Miguel-Hidalgo, J.J. and Cacabelos, R. (1998) Beta-amyloid-
(1–40)-induced neurodegeneration in the rat hippocampal
neurons of the CA1 subfield. Acta Neuropathol. (Berl.), 95:
455–465.
Miller, C.C., Mcloughlin, D.M., Lau, K.F., Tennant, M.E. and
Rogelj, B. (2006) The X11 proteins, A beta production and
Alzheimer’s disease. Trends Neurosci., 29: 280–285.
Mills, J. and Reiner, P.B. (1999) Regulation of amyloid
precursor protein cleavage. J. Neurochem., 72: 443–460.
Mortimer, J.A., French, L.R., Hutton, J.T. and Schuman, L.M.
(1985) Head injury as a risk factor for Alzheimer’s disease.
Neurology, 35: 264–267.
Mortimer, J.A., Van Duijn, C.M., Chandra, V., Fratiglioni, L.,
Graves, A.B., Heyman, A., Jorm, A.F., Kokmen, E., Kondo,
K. and Rocca, W.A. (1991) Head trauma as a risk factor
for Alzheimer’s disease: a collaborative re-analysis of case-
control studies. EURODEM Risk Factors Research Group.
Int. J. Epidemiol., 20(Suppl 2): S28–S35.
Murai, H., Pierce, J.E.S., Raghupathi, R., Smith, D.H.,
Saatman, K.E., Trojanowski, J.Q., Lee, V.M.Y., Loring,
J.F., Eckman, C.B., Younkin, S.G. and McIntosh, T.K. (1998)
Twofold overexpression of human b-amyloid precursor
proteins in transgenic mice does not affect the neuromotor,
cognitive or neurodegenerative sequelae following experimen-
tal brain injury. J. Comp. Neurol., 392: 428–438.
Murakami, N., Yamaki, T., Iwamoto, Y., Sakakibara, T.,
Kobori, N., Fushiki, S. and Ueda, S. (1998) Experimental
brain injury induces expression of amyloid precursor protein,
which may be related to neuronal loss in the hippocampus.
J. Neurotrauma, 15: 993–1003.
Nakagawa, Y., Nakamura, M., McIntosh, T.K., Rodriguez, A.,
Berlin, J.A., Smith, D.H., Saatman, K.E., Raghupathi, R.,
Clemens, J., Saido, T.C., Schmidt, M.L., Lee, V.M. and
Trojanowski, J.Q. (1999) Traumatic brain injury in young,
amyloid-beta peptide overexpressing transgenic mice induces
marked ipsilateral hippocampal atrophy and diminished
Abeta deposition during aging. J. Comp. Neurol., 411:
390–398.
Nakagawa, Y., Reed, L., Nakamura, M., McIntosh, T.K.,
Smith, D.H., Saatman, K.E., Raghupathi, R., Clemens, J.,
Saido, T.C., Lee, V.M. and Trojanowski, J.Q. (2000) Brain
trauma in aged transgenic mice induces regression of
established abeta deposits. Exp. Neurol., 163: 244–252.
Nakamura, T., Shoji, M., Harigaya, Y., Watanabe, M.,
Hosoda, K., Cheung, T.T., Shaffer, L.M., Golde, T.E.,
Younkin, L.H. and Younkin, S.G. (1994) Amyloid B levels in
cerebrospinal fluid are elevated in early-onset Alzheimer’s
disease. Ann. Neurol., 36: 903–911.
Namba, Y., Tomonaga, M., Kawasaki, H., Otomo, E. and
Ikeda, K. (1991) Apolipoprotein E immunoreactivity in cere-
bral amyloid deposits and neurofibrillary tangles in
Alzheimer’s disease and kuru plaque amyloid in Creutz-
feldt-Jakob disease. Brain Res., 541: 163–166.
Nemetz, P.N., Leibson, C., Naessens, J.M., Beard, M.,
Kokmen, E., Annegers, J.F. and Kurland, L.T. (1999) Trau-
matic brain injury and time to onset of Alzheimer’s disease: a
population-based study. Am. J. Epidemiol., 149: 32–40.
Neve, R.L. and Robakis, N.K. (1998) Alzheimer’s disease: a
re-examination of the amyloid hypothesis. Trends Neurosci.,
21: 15–19.
Nicoll, J.A. (1996) Genetics and head injury. Neuropathol.
Appl. Neurobiol., 22: 515–517.
Nicoll, J.A.R., Roberts, G.W. and Graham, D.I. (1995) Apol-
ipoprotein E e4 allele is associated with deposition of amyloid
b-protein following head injury. Nat. Med., 1: 135–137.
Olsson, A., Scsajbok, L., Ost, M., Hoglund, K., Nylen, K.,
Rosengre, L., Nellgard, B. and Blennow, K. (2004) Marked
increase of beta-amyloid (1–42) and amyloid precursor
protein in ventricular cerebrospinal fluid after severe trau-
matic brain injury. J. Neurol., 251: 870–876.
Pierce, J.E., Smith, D.H., Trojanowski, J.Q. and McIntosh,
T.K. (1998) Enduring cognitive, neurobehavioral and histo-
pathological changes persist for up to one year following
severe experimental brain injury in rats. Neuroscience, 87:
359–369.
Pierce, J.E., Trojanowski, J.Q., Graham, D.I., Smith, D.H. and
McIntosh, T.K. (1996) Immunohistochemical characteriza-
tion of alterations in the distribution of amyloid precursor
proteins and beta-amyloid peptide after experimental brain
injury in the rat. J. Neurosci., 16: 1083–1090.
Plassman, B.L., Havlik, R.J., Steffens, D.C., Helms, M.J.,
Newman, T.N., Drosdick, D., Phillips, C., Gau, B.A., Welsh-
Bohmer, K.A., Burke, J.R., Guralnik, J.M. and Breitner,
J.C.S. (2000) Documented head injury in early adulthood and
risk of Alzheimer’s disease and other dementias. Neurology,
55: 1158–1166.
Raby, C.A., Morganti-Kossman, M.C., Kossman, T., Stahel,
P.F., Watson, M.D., Evans, L.M., Mehta, P.D., Spiegel, K.,
Kuo, Y., Roher, A.E. and Emmerling, M.R. (1998) Trau-
matic brain injury increases b-amyloid peptide 1–42 in
cerebrospinal fluid. J. Neurochem., 71: 2505–2509.
Racchi, M., Solano, C.D., Sironi, M. and Govoni, S. (1999)
Activity of a-secretase as the common final effector of protein
kinase c-dependent and independent modulation of amyloid
precursor protein metabolism. J. Neurochem., 72: 2464–2470.
Raghupathi, R., Graham, D.I. and McIntosh, T.K. (2000)
Apoptosis after traumatic brain injury. J. Neurotrauma, 17:
927–938.
Rasmusson, D.X., Brandt, J., Martin, D.B. and Folstein, M.F.
(1995) Head injury as a risk factor in Alzheimer’s disease.
Brain Inj., 9: 213–219.
Rebeck, G.W., Reiter, J.S., Strickland, D.K. and Hyman, B.T.
(1993) Apolipoprotein E in sporadic Alzheimer’s disease: all-
elic variation and receptor interactions. Neuron, 11: 575–580.
Roberds, S.L., Anderson, J., Basi, G., Bienkowski, M.J.,
Branstetter, D.G., Chen, K.S., Freedman, S.B., Frigon,
N.L., Games, N., Hu, G., Johnson-Wood, K., Kappenman,
K.E., Kawabe, T.T., Kola, I., Kuehn, R., Lee, M., Liu, W.,
Motter, R., Nichols, N.F., Power, M., Robertson, D.W.,
Schenk, D., Schoor, M., Shopp, G.M., Shuck, M.E.,

Sinha, S., Svensson, K.A., Tatsuno, G., Tintrup, H., Wijs-
man, J., Wright, S. and Mcconlogue, L. (2001) BACE
knockout mice are healthy despite lacking the primary b-
secretase activity in brain: implications for Alzheimer’s dis-
ease therapeutics. Hum. Mol. Genet., 10: 1317–1324.
Roberts, G.W., Allsop, D. and Bruton, C.J. (1990) The occult
aftermath of boxing. J. Neurol. Neurosurg. Psychiatry, 53:
373–378.
Roberts, G.W., Gentleman, S.M., Lynch, A. and Graham, D.I.
(1991) beta A 4 amyloid protein deposition in brain after
head trauma. Lancet, 338: 1422–1423.
Roberts, G.W., Gentleman, S.M., Lynch, A., Landon, M. and
Graham, D.I. (1994) Beta amyloid protein deposition in the
brain after severe head injury: implications for the patho-
genesis of Alzheimer’s disease. J. Neurol. Neurosurg.
Psychiatry, 57: 419–425.
Roses, A.D. (1996) Apolipoprotein E alleles as risk factors in
Alzheimer’s disease. Ann. Rev. Med., 47: 387–400.
Rudelli, R., Strom, J.O., Welch, P.T. and Ambler, M.W. (1982)
Post-traumatic premature Alzheimer’s disease. Arch.
Neurol., 39: 570–575.
Sabo, S.L., Lanier, L.M., Ikin, A.F., Khorkova, O.,
Sahasrabudhe, S., Greengard, P. and Buxbaum, J.D. (1999)
Regulation of X-amyloid secretion by FE65, an amyloid
protein precursor-binding protein. J. Biol. Chem., 274:
7952–7957.
Salib, E. and Hillier, V. (1997) Head injury and the risk
of Alzheimer’s disease: a case control study. Int. J. Geriat.
Psychiatry, 12: 363–368.
Saunders, A.M., Strittmatter, W.J. and Schmechel, D. (1993)
Association of apolioprotein E allele e4 with late-onset
familial and sporadic Alzheimer’s disease. Proc. Natl. Acad.
Sci. U.S.A., 90: 8098–8102.
Schofield, P.W., Tang, M.X., Marder, K., Bell, K., Dooneief,
G., Chun, M., Sano, M., Stern, Y. and Mayeux, R. (1997)
Alzheimer’s disease after remote head injury: an incidence
study. J. Neurol. Neurosurg. Psychiatry, 62: 119–124.
Selkoe, D.J. (1997) Alzheimer’s disease: genotypes, phenotypes,
and treatments. Science, 275: 630–631.
Selkoe, D.J. (2001) Alzheimer’s disease: genes, proteins, and
therapy. Physiol. Rev., 81: 741–766.
Smith, C., Graham, D.I., Murray, L.S., Stewart, J. and Nicoll,
J.A.R. (2006) Association of APOe4 and cerebrovascular
pathology in traumatic brain injury. J. Neurol. Neurosurg.
Psychiatry, 77: 363–366.
Smith, D.H., Chen, X.H., Iwata, A. and Graham, D.I. (2003)
Amyloid-X accumulation in axons after traumatic brain
injury in humans. J. Neurosurg., 98: 1072–1077.
Smith, D.H., Chen, X.H., Nonaka, M., Trojanowski, J.Q., Lee,
V.M., Saatman, K.E., Leoni, M.J., Xu, B.N., Wolf, J.A. and
Meaney, D.F. (1999) Accumulation of amyloid beta and tau
and the formation of neurofilament inclusions following
diffuse brain injury in the pig. J. Neuropathol. Exp. Neurol.,
58: 982–992.
Smith, D.H., Nakamura, M., McIntosh, T.K., Wang, J.,
Rodriguez, A., Chen, X.H., Raghupathi, R., Saatman, K.E.,
Clemens, J., Schmidt, M.L., Lee, V.M. and Trojanowski, J.Q.
315
(1998) Brain trauma induces massive hippocampal neuron
death linked to a surge in beta-amyloid levels in mice overex-
pressing mutant amyloid precursor protein. Am. J. Pathol.,
153: 1005–1010.
Sorbi, S., Nacmias, N., Piacentini, S., Repice, A., Latorraca, S.,
Forleo, P. and Amaducci, L. (1995) ApoE as a prognostic
factor for post-traumatic coma. Nat. Med., 1: 852.
Stein, D.G. (2001) Brain damage, sex hormones and recovery: a
new role for progesterone and estrogen? Trends Neurosci.,
24: 386–391.
Stone, J.R., Okonkwo, D.O., Singleton, R.H., Mutlu, L.K.,
Helm, G.A. and Povlishock, J.T. (2002) Caspase-3-mediated
cleavage of amyloid precursor protein and formation of am-
yloid b peptide in traumatic axonal injury. J. Neurotrauma,
19: 601–614.
Strittmatter, W.J., Saunders, A.M. and Schmechel, D. (1993)
Apolioprotein E high avidity binding to B-amyloid and
increased frequency of the E4 allele in late onset familial
Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A., 90:
1977–1981.
Suh, Y.H. and Checler, F. (2002) Amyloid precursor protein,
presenilins, and a-synuclein: molecular pathogenesis and
pharmacological applications in Alzheimer’s disease. Pharm.
Rev., 54: 469–525.
Sullivan, P., Petitti, D. and Barbaccia, J. (1987) Head trauma
and age of onset of dementia of the Alzheimer type (Letter).
JAMA, 257: 2289–2290.
Szczygielski, J., Mautes, A., Steudel, W.I., Falkai, P., Bayer, T.A.
and Wirths, O. (2005) Traumatic brain injury: cause or risk
of Alzheimer’s disease? A review of experimental studies.
J. Neural. Transm., 112: 1547–1564.
Teasdale, G.M., Nicoll, J.A.R., Murray, G. and Fiddes, M.
(1997) Association of apolipoprotein E polymorphism with
outcome after head injury. Lancet, 350: 1069–1071.
Tomita, S., Kirino, Y. and Suzuki, T. (1998) Cleavage of
Alzheimer’s amyloid precursor protein (APP) by secretases
occurs after O-glycosylation of APP in the protein secretory
pathway. J. Biol. Chem., 273: 6277–6284.
Uryu, K., Chen, X.H. and Graham, D.J. (2004) Short-term
accumulation of beta-amyloid in axonal pathology following
traumatic brain injury in humans. Neurobiol. Aging, 25(Sup-
pl 1): P2–P250.
Uryu, K., Laurer, H., McIntosh, T.K., Pratico, D., Martinez,
D., Leight, S., Lee, V.M.-Y. and Trojanowski, J.Q. (2002)
Repetitive mild brain trauma accelerates Ab deposition, lipid
peroxidation, and cognitive impairment in a transgenic
mouse model of Alzheimer amyloidosis. J. Neurosci., 22:
446–454.
Van Den Heuvel, C., Blumbergs, P.C., Finnie, J.W., Manavis,
J., Jones, N.R., Reilly, P.L. and Pereira, R.A. (1999) Upreg-
ulation of amyloid precursor protein messenger RNA in
response to traumatic brain injury: an ovine head impact
model. Exp. Neurol., 159: 441–450.
Van Duijn, C.M., Tanja, T.A., Haaxma, R., Schulte, W., Saan,
R.J., Lameris, A.J., Antonides-Hendriks, G. and Hofman, A.
(1992) Head trauma and the risk of Alzheimer’s disease. Am.
J. Epidemiol., 135: 775–782.
316
Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz,
E.A., Denis, P., Teplow, D.B., Ross, S., Amarante, P.,
Loeloff, R., Luo, Y., Fisher, S., Fuller, J., Edenson, S.,
Lile, J., Jarosinski, M.A., Biere, A.L., Curran, E., Burgess, T.,
Louis, J., Collins, F., Treanor, J., Rogers, G. and Citron, M.
(1999) X-Secretase cleavage of Alzheimer’s amyloid precursor
protein by the transmembrane aspartic protease BACE.
Science, 286: 735–740.
Williams, D.B., Annegers, J.F., Kokmen, E., O’brien, P.C. and
Kurland, L. (1991) Brain injury and neurologic sequelae: a
cohort study of dementia, parkinsonism, amyotrophic lateral
sclerosis. Neurology, 41: 1554–1557.
Wilquet, V. and De Strooper, B. (2004) Amyloid-X precursor
protein processing in neurodegeneration. Curr. Opin.
Neurobiol., 14: 582–588.
Yoshiyama, Y., Uryu, K., Higuchi, M., Longhi, L., Hoover, R.,
Fujimoto, S., McIntosh, T., Lee, V.M. and Trojanowski, J.Q.
(2005) Enhanced neurofibrillary tangle formation, cerebral
atrophy, and cognitive deficits induced by repetitive mild
brain injury in a transgenic tauopathy mouse model.
J. Neurotrauma, 22: 1134–1141.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 22

The neurotrophic protein S100B: value as a marker of

brain damage and possible therapeutic implications

Andrea Kleindienst1,, Felicitas Hesse1, M. Ross Bullock2 and Michael Buchfelder1

1
Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Germany
2
Department of Neurosurgery, Virginia Commonwealth University Medical Center, Richmond, VA 23298-0631, USA

Abstract: We provide a critical analysis of the value of S100B as a marker of brain damage and possible
therapeutic implications. The early assessment of the injury severity and the consequent prognosis are of
major concern for physicians treating patients suffering from traumatic brain injury (TBI). A reliable
indicator to accurately determine the extent of the brain damage has to meet certain requirements: (i) to
originate in the central nervous system (CNS) with no contribution from extracerebral sources; (ii) a passive
release from damaged neurons and/or glial cells without any stimulated active release; (iii) a lack of specific
effects on neurons and/or glial cells interfering with the initial injury; (iv) an unlimited passage through the
blood-brain barrier (BBB). The measurement of putative biochemical markers, such as the S100B protein,
has been proposed in this role. Over the past decade, numerous studies have reported a positive correlation
of S100B serum levels with a poor outcome following TBI. However, some studies raise doubt whether the
serum measurement of S100B is a valid biochemical marker of brain damage. We summarize the specific
properties of S100B and analyze whether they support or counteract the necessary requirements to desig-
nate this protein as an indicator of brain damage. Finally, we report recent experimental findings suggesting
a possible therapeutic potential of S100B.

Keywords: S100B; TBI; neuroregeneration; brain damage; biomarker; neurotrophic factor

General features and origin of the S100B protein
As early as 1965, a neurotrophic factor was puri-
fied from bovine brain for the first time (Moore,
1965), and in 1978 this factor was demonstrated to
consist of two distinct proteins, S100b and S100a
(Isobe et al., 1978). After the identification of the
chromosomal localization of S100 proteins in
1995, the nomenclature was changed from S100b
to S100B and from S100a to S100A1 (Schafer
et al., 1995). In the following decade, experimental
research was focused on identifying the specific
Corresponding author. Tel.: +49-(0)-9131-85-34577;
Fax: +49-(0)-9131-85-34551;
E-mail: kleindienst@nch.imed.uni-erlangen.de
role of S100B while in the clinical setting the
measurement of S100B became more frequently
performed.
The S100B protein belongs to a multigenic fam-
ily of low molecular weight (9–13 kDa) calcium-
binding S100 proteins (Donato, 2001; Heizmann
et al., 2002). S100B is most abundant in glial cells
of the central nervous system (CNS), predomi-
nately in astrocytes (Donato, 1986), and consti-
tutes 1–1.5 mg/mg of soluble protein (Matsutani
et al., 1985). S100 proteins contain no detectable
carbohydrate, lipid, nucleic acid or phosphate, and
are dimers consisting of at least two types of sub-
units of either identical or different amino acid
composition. S100 proteins are highly conserved in
amino acid composition among vertebrate species

DOI: 10.1016/S0079-6123(06)61022-4 317

318
(Donato, 1986), and comparison of the human and
bovine S100A1 and the human and rat S100B
subunit reveals an almost complete homology
(Isobe et al., 1984).
The extracerebral origin of S100B has also been
recognized in several studies and this data conflicts
with the interpretation of S100B serum levels fol-
lowing traumatic brain injury (TBI). The S100B
tissue distribution has been assessed using
immunocytochemical approaches (Donato, 1991).
Quantification in rat tissue found a S100B con-
centration in brain tissue of 3000 ng per milligram
of soluble protein, in adipose tissue of 1000 ng, in
the skin of 80 ng, in the testes of 40 ng, and in all
other tissue of around 1–2 ng (Zimmer and Van
Eldik, 1987). Thus, elevated S100B levels following
injury may also come from either damaged skeletal
muscle (Arcuri et al., 2002) or adipose tissue
(Suzuki et al., 1984).
Release mechanism of S100B
Experimental findings demonstrate that S100B is
actively secreted from astroglia although the exact
mechanism has yet not been identified. Under
physiological circumstances, S100B release can be
stimulated by 5 HT 1a agonists in cultured rat as-
trocytes (Shashoua et al., 1984; Van Eldik and
Zimmer, 1987; Whitaker-Azmitia et al., 1990), and
is observed within a few minutes after activation
of A1 adenosine or mGlu3 metabotropic gluta-
mate receptors and release may last up to 10 h
(Ciccarelli et al., 1999). In neuronal plus glial cul-
tures, the level of S100B in the growth medium was
measurable (Willoughby et al., 2004). Following
experimental TBI, applying a 50 ms strain (stretch)
on these cell cultures grown on deformable silicone
membranes (Ellis et al., 1995), there was a dra-
matic rise in S100B 15 s after injury (Willoughby
et al., 2004), and thereafter a continued release of
S100B until at least 48 h post-injury (Slemmer
et al., 2002; Willoughby et al., 2004).
The injury-induced S100B release is likely to be
the result of at least two possible processes. Firstly,
electron microscopic studies of stretch-injured as-
trocyte cultures show that immediately after
stretch-injury structural and membrane integrity
is severely compromised (Ellis et al., 1995) allow-
ing the cell release of preformed S100B. Secondly,
since an injury-induced ATP and glutamate release
has been shown also, the release of S100B might be
in part due to an astrocyte receptor activation by
ATP and glutamate (Ciccarelli et al., 1999). The
finding of an increased S100B release at 24 and
48 h post-injury may thus imply that this is a long-
term metabolic response to brain damage, which
promotes a ‘‘healing process’’ in injured neurons
or astrocytes.
Cellular action of S100B
Intracellularly, S100B is involved in signal trans-
duction via the inhibition of protein phoshorylat-
ion, regulation of enzyme activity and by affecting
the calcium homeostasis as well as the regulation
of cell morphology by interaction with elements of
the cytoplasmatic cytoskeleton (Zimmer and Van
Eldik, 1986; Rustandi et al., 1998; Wilder et al.,
1998). After release into the extracellular fluid,
S100B acts in an autocrine and paracrine manner.
In vitro studies demonstrate mitogenic properties
of S100B, such as an increased proliferative acti-
vity on melanoma cell lines at concentrations be-
low 50 nM and above 5 mM (Klein et al., 1989), as
well as on rat C6 glioma cells at 50 pM–1.5 nM
(Selinfreund et al., 1991), preferably in its oxidized
state (Scotto et al., 1998). Further investigations
also confirmed a dose-dependent action of S100B
exerting a neuroprotective and neurotrophic influ-
ence at nanomolar concentrations (Rickmann
et al., 1995; Li et al., 1998; Huttunen et al.,
2000), but at micromolar concentrations, an acti-
vation of the inducible nitric oxide (NO) synthase
was seen with subsequent NO generation, poten-
tially leading to astrocytic death (Hu et al., 1997;
Petrova et al., 2000).
Different studies provide evidence for neurotro-
phic properties of S100B. In cultured mesencephal-
ic rat neurons (Azmitia et al., 1990) and dorsal
root ganglia (Van Eldik et al., 1991), S100B
has been shown to promote neurite outgrowth
(Kligman and Marshak, 1985; Winningham-
Major et al., 1989). In primary rat spinal cord
culture, S100B rapidly promoted reassembly

and/or stabilization of the cytoskeletal system
(Nishi et al., 1997), and prevented apoptosis in a
neuroblastoma clonal cell line (Brewton et al.,
2001). Most importantly, however, after sciatic
nerve section in newborn rats, S100B rescued mo-
tor neuron death and preserved neuron diameter
(Iwasaki et al., 1997).
In addition to its trophic properties, a neuro-
protective effect of S100B was found in vitro.
S100B decreased neuronal cell death and mito-
chondrial dysfunction in rat hippocampal neurons
after both glucose deprivation, and increased in-
tracellular free calcium (Barger et al., 1995). S100B
is, thus, thought to be involved in regulation of
energy metabolism by stimulation of the fructose-
1,6-biphosphate aldolase (Zimmer and Van Eldik,
1986) and phosphoglucomutase (Landar et al.,
1996). S100B has also been implicated in cytosolic
Ca2+ buffering (Xiong et al., 2000). In cultures of
embryonal chick and neonatal rat neurons, S100B
was found to protect against glutamate- and
staurosporin-induced damage (Ahlemeyer et al.,
2000). The above-mentioned neuroprotective and
neurotrophic properties of S100B have not been
elucidated after TBI.
Passage of S100B through the blood-brain barrier
While in cell cultures, the injury-induced S100B
release continues to increase up to 48 h (Slemmer
et al., 2002; Willoughby et al., 2004), S100B serum
levels in patients are highest directly after injury,
and become normalized within 24 h in a high per-
centage of cases, even in those patients with a bad
outcome (Jackson et al., 2000). The underlying
mechanism of the passage of S100B through the
blood-brain barrier (BBB) has not been clarified
yet, nor do data exist about cerebral S100B levels
and their correlation to serum S100B levels. Since
one of the purposes of the BBB is to prevent pro-
teins to enter the brain, there is reasonable doubt
whether S100B of cerebral origin may be able to
cross the intact BBB in the opposite direction.
The principle of spectroscopy is widely applied
in chemistry for the analysis of molecules in solu-
tion, and MR spectroscopy can be used to identify
important molecules in living tissue. Recently, MR
319
proton spectroscopy has been shown to identify
brain metabolites like N-acteylaspartate (NAA),
creatine, choline and lactate (Bruhn et al., 1989;
Frahm et al., 1989a–c). We demonstrated MR
proton spectroscopy to detect different concentra-
tions of aqueous solutions of S100B in vitro, with
a strong correlation between the S100B concen-
tration and the area under the curve of the re-
spective S100B MR peak at 4.5 ppm (Kleindienst
et al., 2005b).
Increased cerebral S100B levels following an in-
traventricular S100B infusion in normal rats were
confirmed by MR proton spectroscopy while the
S100B serum concentration did not increase ac-
cording to the notion that proteins do not cross an
intact BBB (Kleindienst et al., 2005b). Following
controlled cortical impact injury in the rat, S100B
serum levels were elevated up to 48 h post-injury
(Rothoerl et al., 2000). In fluid percussion injury in
the rat, serum S100B levels increased immediately
after injury, peaked at 3 h and returned to normal
values at 24 h after injury (Kleindienst et al., 2004).
This time profile of serum S100B levels parallels
the BBB disruption found after experimental
TBI (Povlishock et al., 1978). In contrast, cere-
bral S100B levels as quantified by MR proton
spectroscopy were raised slightly 3 h after injury,
and increased significantly thereafter to more than
twice these values by day 5 (Kleindienst et al.,
2005b). Thus, no clear relationship exists between
the cerebral S100B dynamics and the serum S100B
levels.
Clinical studies
In 2003, a thorough review of the role of S100B as
a marker of brain damage was published summa-
rizing the results of 18 clinical studies in a total of
1085 patients (Rothermundt et al., 2003). In 2004
and 2005, another six studies comprising more
than 600 adult patients were performed supporting
the correlation of elevated serum levels of S100B
with a poor outcome after brain injury (Pelinka
et al., 2004; Savola et al., 2004; Vos et al., 2004;
Wunderlich et al., 2004; Berger et al., 2005; Chen
and Zhu, 2005). The time profile of S100B re-
ported in the serum was variable, and even a
320
delayed increase of S100B serum levels on the 6th
day after injury has been reported and speculated
to occur due to secondary brain cell damage
(Raabe and Seifert, 2000). Measurements of
cerebrospinal fluid S100B levels allow a true as-
sessment of the cerebral S100B release following
brain insults (Kleine et al., 2003; Petzold et al.,
2003; Hayakata et al., 2004; Shore et al., 2004), but
require for repetitive measurements either lumbar
or ventricular cerebrospinal fluid drainage.
The importance of S100B in the acutely injured
brain is not well known yet. In view of a substan-
tial body of evidence demonstrating an association
between S100B and bad outcome after TBI, it is
important to be aware that proof of association
is not a proof of causation, in science, and this is
especially true of S100B. In contrast to TBI, fol-
lowing electroconvulsive therapy in patients, in-
creased serum S100B levels were associated with
an improved cognitive performance (Agelink et al.,
2001). Furthermore, strong evidence exists that in
traumatized patients S100B serum levels are in-
creased without concomitant brain damage
(Ashraf et al., 1999; Anderson et al., 2001a, b;
Svenmarker et al., 2002; Kleine et al., 2003;
Pelinka et al., 2003; Nygren De Boussard et al.,
2004). Finally, the exact function and effects of
increased cerebral levels of S100B after TBI are
not thoroughly understood, yet.
Effect of S100B on recovery following experimental
TBI
Besides this experimental evidence for the benefi-
cial effect of S100B on neuronal maintenance, a
specific role of S100B has also been proposed at a
higher level in developmental plasticity (Marshak,
1990), and in cell processes thought to be involved
in learning and memory, such as long-term po-
tentiation (Fazeli et al., 1990). This has been con-
firmed by injection of S100B anti-serum into the
hemisphere of chicks causing amnesia for a passive
avoidance task (O’Dowd et al., 1997). Vice versa,
S100B infused into the rat hippocampus, has been
shown to facilitate long-term memory for an in-
hibitory avoidance task (Mello e Souza et al.,
2000). Finally, light- and electron microscopic
studies are consistent with the hypothesis that
S100B plays a role in lesion-induced collateral
sprouting and reactive synaptogenesis (Rickmann
et al., 1995; Li et al., 1998; McAdory et al., 1998;
Huttunen et al., 2000), and that repair may occur
by interaction with growth factors (Gomide and
Chadi, 1999).
The hippocampus is a region critical for learning
and memory. It displays increased susceptibility to
injury, and cognitive impairment following injury
has been linked to hippocampal dysfunction
(Hicks et al., 1993). Recent findings of neurogen-
esis within the subgranular zone of the dentate
gyrus of the hippocampus in adult mammals pro-
vides a possible source to replace lost neurons
(Eriksson et al., 1998; Kornack and Rakic, 1999;
Gould and Gross, 2002). Indeed, neuronal pro-
genitor cells within the hippocampus present in-
creased cellular proliferation and subsequent
neuron formation following experimental TBI
(Dash et al., 2001; Chirumamilla et al., 2002).
This adult neurogenesis has been found to be ac-
tively regulated by hippocampal astrocytes, which
are a cellular source for mitogenic/neurotrophic
factors and thus promote proliferation of adult
neural stem cells and instruct the fate commitment
of developing neurons (Song et al., 2002). Conse-
quently, the astrocytic neurotrophic protein S100B
is a potential candidate to increase this progenitor
cell proliferation and subsequent neuron forma-
tion following TBI, thereby enhancing hippocam-
pal network repair and improving cognitive
recovery.
This hypothesis of a neurogenic effect of S100B
has been examined by an intraventricular infusion
of S100B following fluid percussion injury in the rat
(Kleindienst et al., 2004). Functional recovery was
assessed by the Morris water maze 5 weeks post-
injury and revealed a significantly improved cogni-
tive performance following intraventricular S100B
infusion (Kleindienst et al., 2004). Within the
hippocampus, S100B increased the proliferative re-
sponse significantly by more than 40% on day 5
post-injury (Fig. 1; Kleindienst et al., 2005a). S100B
promoted the survival of the injury-induced pro-
genitor cell proliferation as well as the subsequent
differentiation into neurons (Fig. 2; Kleindienst
et al., 2005a).
 321

Fig. 1. Enhancement of progenitor cell proliferation in the hippocampus following TBI and S100B treatment. The photomicrograph
shows the distribution of BromodeoxyUridine (BrdU) immunoreactive cells within the ipsilateral dentate gyrus following lateral fluid
percussion injury and an intraventricular S100B infusion. BrdU immunoreactive cells on day 5 post-injury are predominately identified
in the area of their origin, the subgranular zone. The arrows point to typical cluster of cells.

Thus, S100B has been demonstrated to increase
hippocampal stem/progenitor cell proliferation
and neuronal differentiation following TBI, and
this enhanced neurogenesis is correlated with an
improved cognitive recovery (Kleindienst et al.,
2005a). These findings stress the importance of
astrocytic factors for neurogenesis and provide
compelling evidence for a therapeutic potential of
S100B improving functional recovery following
TBI. Further studies are needed to elucidate the
beneficial role of S100B on hippocampal network
repair, thus offering a potential therapy to aug-
ment important innate repair mechanisms of the
brain in order to promote memory consolidation
in patients with brain insults.
Conclusion
Taken together, although the desire for a marker of
brain damage is reasonable designating S100B for
this role warrants considerable simplification, and
interpretation of serum S100B levels following TBI
requires thorough knowledge of factors influencing
the measured concentration. Firstly, one has to be
aware of a contribution to serum S100B levels
originating from extracerebral sources. Secondly,
the injury-induced cerebral S100B release may be
assumed to be a combination of a passive release by
damaged astrocytes and an active release by stimu-
lated astrocytes initiating repair mechanisms, with
both release patterns varying over time and in the
presence of secondary insults. Thirdly, the passage
of cerebral S100B is modulated by the BBB on its
way into the extracerebral compartment, and se-
rum S100B levels do not reflect the corresponding
cerebral S100B release. Finally, a beneficial effect
of S100B on neuronal maintenance, neurogenesis
and cognitive performance has been demonstrated
promoting repair mechanisms and functional re-
covery following TBI.
Supplementing measurements of S100B serum
levels, we established MR proton spectroscopy to
more accurately reflect cerebral S100B levels and
322

Fig. 2. Neuronal differentiation of progenitor cells in the ipsilateral dentate gyrus following TBI and S100B treatment. The confocal
photomicrograph shows the staining of the neuronal Marker NeuN (red) and BrdU immunoreactivity (green) in the dentate gyrus at 5
weeks following lateral fluid percussion injury and an intraventricular S100B infusion. The arrows point at progenitor cells demon-
strating a co-localization of NeuN and BrdU (orange) thereby indicating neuronal differentiation.

thus provide a clinical method to non-invasively
monitor whole brain S100B levels repeatedly in
patients with brain insults. The determination of
cerebral S100B levels may allow a more accurate
estimation of the prognosis in head-injured pa-
tients. Moreover, we propose that S100B far from
being a negative determinant of outcome, as sug-
gested previously in the human TBI and ischemia
literature, may improve neurogenesis and func-
tional recovery following acute brain injury and
may be in fact potentially a new treatment option,
which might improve outcome of patients with
many forms of acute brain damage.
References
Agelink, M.W., Andrich, J., Postert, T., Wurzinger, U., Zeit,
T., Klotz, P. and Przuntek, H. (2001) Relation between
electroconvulsive therapy, cognitive side effects, neuron spe-
cific enolase, and protein S-100. J. Neurol. Neurosurg. Psy-
chiatry, 71: 394–396.
Ahlemeyer, B., Beier, H., Semkova, I., Schaper, C. and Kriegl-
stein, J. (2000) S-100beta protects cultured neurons against
glutamate- and staurosporine-induced damage and is in-
volved in the antiapoptotic action of the 5 HT(1A)-receptor
agonist, Bay  3702. Brain Res., 858: 121–128.
Anderson, R.E., Hansson, L.O., Nilsson, O., Dijlai-Merzoug,
R. and Settergren, G. (2001a) High serum S100B levels for
trauma patients without head injuries. Neurosurgery, 48:
1255–1258; discussion 1258–1260.
Anderson, R.E., Hansson, L.O., Nilsson, O., Liska, J., Setter-
gren, G. and Vaage, J. (2001b) Increase in serum S100A1-B
and S100BB during cardiac surgery arises from extracerebral
sources. Ann. Thorac. Surg., 71: 1512–1517.
Arcuri, C., Giambanco, I., Bianchi, R. and Donato, R. (2002)
Annexin V, annexin VI, S100A1 and S100B in develop-
ing and adult avian skeletal muscles. Neuroscience, 109:
371–388.
Ashraf, S., Bhattacharya, K., Tian, Y. and Watterson, K.
(1999) Cytokine and S100B levels in paediatric patients

undergoing corrective cardiac surgery with or without total
circulatory arrest. Eur. J. Cardiothorac. Surg., 16: 32–37.
Azmitia, E.C., Dolan, K. and Whitaker-Azmitia, P.M. (1990)
S-100B but not NGF, EGF, insulin or calmodulin is a CNS
serotonergic growth factor. Brain Res., 516: 354–356.
Barger, S.W., Van Eldik, L.J. and Mattson, M.P. (1995) S100
beta protects hippocampal neurons from damage induced by
glucose deprivation. Brain Res., 677: 167–170.
Berger, R.P., Adelson, P.D., Pierce, M.C., Dulani, T., Cassidy,
L.D. and Kochanek, P.M. (2005) Serum neuron-specific
enolase, S100B, and myelin basic protein concentrations after
inflicted and noninflicted traumatic brain injury in children.
J. Neurosurg., 103: 61–68.
Brewton, L.S., Haddad, L. and Azmitia, E.C. (2001) Col-
chicine-induced cytoskeletal collapse and apoptosis in N-18
neuroblastoma cultures is rapidly reversed by applied
S-100beta. Brain Res., 912: 9–16.
Bruhn, H., Frahm, J., Gyngell, M.L., Merboldt, K.D.,
Hanicke, W. and Sauter, R. (1989) Cerebral metabolism in
man after acute stroke: new observations using localized
proton NMR spectroscopy. Magn. Reson. Med., 9: 126–131.
Chen, D.Q. and Zhu, L.L. (2005) Dynamic change of serum
protein S100b and its clinical significance in patients with
traumatic brain injury. Chin. J. Traumatol., 8: 245–248.
Chirumamilla, S., Sun, D., Bullock, M.R. and Colello, R.J.
(2002) Traumatic brain injury induced cell proliferation in
the adult mammalian central nervous system. J. Neurotrau-
ma, 19: 693–703.
Ciccarelli, R., Di Iorio, P., Bruno, V., Battaglia, G., D’Alimonte,
I., D’Onofrio, M., Nicoletti, F. and Caciagli, F. (1999) Acti-
vation of A(1) adenosine or mGlu3 metabotropic glutamate
receptors enhances the release of nerve growth factor and
S-100beta protein from cultured astrocytes. Glia, 27: 275–281.
Dash, P.K., Mach, S.A. and Moore, A.N. (2001) Enhanced
neurogenesis in the rodent hippocampus following traumatic
brain injury. J. Neurosci. Res., 63: 313–319.
Donato, R. (1986) S-100 proteins. Cell Calcium, 7: 123–145.
Donato, R. (1991) Perspectives in S-100 protein biology. Re-
view article. Cell Calcium, 12: 713–726.
Donato, R. (2001) S100: a multigenic family of calcium-
modulated proteins of the EF-hand type with intracellular
and extracellular functional roles. Int. J. Biochem. Cell Biol.,
33: 637–668.
Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and
Povlishock, J.T. (1995) A new model for rapid stretch-
induced injury of cells in culture: characterization of the
model using astrocytes. J. Neurotrauma, 12: 325–339.
Eriksson, P.S., Perfilieva, E., Bjork-Eriksson, T., Alborn, A.M.,
Nordborg, C., Peterson, D.A. and Gage, F.H. (1998) Neuro-
genesis in the adult human hippocampus. Nat. Med., 4:
1313–1317.
Fazeli, M.S., Errington, M.L., Dolphin, A.C. and Bliss, T.V.
(1990) Extracellular proteases and S100 protein in long-term
potentiation in the dentate gyrus of the anaesthetized rat.
Adv. Exp. Med. Biol., 268: 369–375.
Frahm, J., Bruhn, H., Gyngell, M.L., Merboldt, K.D., Hanicke,
W. and Sauter, R. (1989a) Localized high-resolution proton
323
NMR spectroscopy using stimulated echoes: initial applica-
tions to human brain in vivo. Magn. Reson. Med., 9: 79–93.
Frahm, J., Bruhn, H., Gyngell, M.L., Merboldt, K.D., Ha-
nicke, W. and Sauter, R. (1989b) Localized proton NMR
spectroscopy in different regions of the human brain in vivo.
Relaxation times and concentrations of cerebral metabolites.
Magn. Reson. Med., 11: 47–63.
Frahm, J., Michaelis, T., Merboldt, K.D., Hanicke, W., Gyng-
ell, M.L., Chien, D. and Bruhn, H. (1989c) Localized NMR
spectroscopy in vivo. Progress and problems. NMR Biomed.,
2: 188–195.
Gomide, V.C. and Chadi, G. (1999) The trophic factors S-
100beta and basic fibroblast growth factor are increased in
the forebrain reactive astrocytes of adult callosotomized rat.
Brain Res., 835: 162–174.
Gould, E. and Gross, C.G. (2002) Neurogenesis in adult mam-
mals: some progress and problems. J. Neurosci., 22: 619–623.
Hayakata, T., Shiozaki, T., Tasaki, O., Ikegawa, H., Inoue, Y.,
Toshiyuki, F., Hosotubo, H., Kieko, F., Yamashita, T.,
Tanaka, H., Shimazu, T. and Sugimoto, H. (2004) Changes
in CSF S100B and cytokine concentrations in early-phase
severe traumatic brain injury. Shock, 22: 102–107.
Heizmann, C.W., Fritz, G. and Schafer, B.W. (2002) S100 pro-
teins: structure, functions and pathology. Front. Biosci., 7:
d1356–d1368.
Hicks, R.R., Smith, D.H., Lowenstein, D.H., Saint Marie, R.
and McIntosh, T.K. (1993) Mild experimental brain injury in
the rat induces cognitive deficits associated with regional
neuronal loss in the hippocampus. J. Neurotrauma, 10:
405–414.
Hu, J., Ferreira, A. and Van Eldik, L.J. (1997) S100beta in-
duces neuronal cell death through nitric oxide release from
astrocytes. J. Neurochem., 69: 2294–2301.
Huttunen, H.J., Kuja-Panula, J., Sorci, G., Agneletti, A.L.,
Donato, R. and Rauvala, H. (2000) Coregulation of neurite
outgrowth and cell survival by amphoterin and S100 proteins
through receptor for advanced glycation end products
(RAGE) activation. J. Biol. Chem., 275: 40096–40105.
Isobe, T., Takahashi, K. and Okuyama, T. (1984) S100a0 (al-
pha alpha) protein is present in neurons of the central and
peripheral nervous system. J. Neurochem., 43: 1494–1496.
Isobe, T., Tsugita, A. and Okuyama, T. (1978) The amino acid
sequence and the subunit structure of bovine brain S-100
protein (PAP I-b). J. Neurochem., 30: 921–923.
Iwasaki, Y., Shiojima, T. and Kinoshita, M. (1997) S100 beta
prevents the death of motor neurons in newborn rats after
sciatic nerve section. J. Neurol. Sci., 151: 7–12.
Jackson, R.G., Samra, G.S., Radcliffe, J., Clark, G.H. and
Price, C.P. (2000) The early fall in levels of S-100 beta in
traumatic brain injury. Clin. Chem. Lab. Med., 38:
1165–1167.
Klein, J.R., Hoon, D.S., Nangauyan, J., Okun, E. and Coch-
ran, A.J. (1989) S-100 protein stimulates cellular prolifera-
tion. Cancer Immunol. Immunother., 29: 133–138.
Kleindienst, A., Harvey, H.B., Rice, A.C., Muller, C., Hamm,
R.J., Gaab, M.R. and Bullock, M.R. (2004) Intraventricular
infusion of the neurotrophic protein S100B improves
324
cognitive recovery after fluid percussion injury in the rat.
J. Neurotrauma, 21: 541–547.
Kleindienst, A., McGinn, M.J., Harvey, H.B., Colello, R.J.,
Hamm, R.J. and Bullock, M.R. (2005a) Enhanced hippo-
campal neurogenesis by intraventricular S100B infusion is
associated with improved cognitive recovery after traumatic
brain injury. J. Neurotrauma, 22: 645–655.
Kleindienst, A., Tolias, C.M., Corwin, F.D., Muller, C., Mar-
marou, A., Fatouros, P. and Bullock, M.R. (2005b) Assess-
ment of cerebral S100B levels by proton magnetic resonance
spectroscopy after lateral fluid-percussion injury in the rat.
J. Neurosurg., 102: 1115–1121.
Kleine, T.O., Benes, L. and Zofel, P. (2003) Studies of the brain
specificity of S100B and neuron-specific enolase (NSE) in
blood serum of acute care patients. Brain Res. Bull., 61:
265–279.
Kligman, D. and Marshak, D.R. (1985) Purification and chara-
cterization of a neurite extension factor from bovine brain.
Proc. Natl. Acad. Sci. U.S.A., 82: 7136–7139.
Kornack, D.R. and Rakic, P. (1999) Continuation of neuro-
genesis in the hippocampus of the adult macaque monkey.
Proc. Natl. Acad. Sci. U.S.A., 96: 5768–5773.
Landar, A., Caddell, G., Chessher, J. and Zimmer, D.B. (1996)
Identification of an S100A1/S100B target protein: phospho-
glucomutase. Cell Calcium, 20: 279–285.
Li, Y., Wang, J., Sheng, J.G., Liu, L., Barger, S.W., Jones,
R.A., Van Eldik, L.J., Mrak, R.E. and Griffin, W.S. (1998)
S100 beta increases levels of beta-amyloid precursor protein
and its encoding mRNA in rat neuronal cultures. J. Neuro-
chem., 71: 1421–1428.
Marshak, D.R. (1990) S100 beta as a neurotrophic factor. Prog.
Brain Res., 86: 169–181.
Matsutani, T., Nagayoshi, M., Tamaru, M., Hirata, Y. and
Kato, K. (1985) Changes in the levels of neural cell specific
proteins in the developing rat brain. Neurochem. Res., 10:
1155–1172.
McAdory, B.S., Van Eldik, L.J. and Norden, J.J. (1998) S100B,
a neurotropic protein that modulates neuronal protein phos-
phorylation, is upregulated during lesion-induced collateral
sprouting and reactive synaptogenesis. Brain Res., 813:
211–217.
Mello e Souza, T., Rohden, A., Meinhardt, M., Goncalves,
C.A. and Quillfeldt, J.A. (2000) S100B infusion into the rat
hippocampus facilitates memory for the inhibitory avoidance
task but not for the open-field habituation. Physiol. Behav.,
71: 29–33.
Moore, B.W. (1965) A soluble protein characteristic of the
nervous system. Biochem. Biophys. Res. Commun., 19:
739–744.
Nishi, M., Kawata, M. and Azmitia, E.C. (1997) S100beta
promotes the extension of microtubule associated protein2
(MAP2)-immunoreactive neurites retracted after colchicine
treatment in rat spinal cord culture. Neurosci. Lett., 229:
212–214.
Nygren De Boussard, C., Fredman, P., Lundin, A., Andersson,
K., Edman, G. and Borg, J. (2004) S100 in mild traumatic
brain injury. Brain Inj., 18: 671–683.
O’Dowd, B.S., Zhao, W.Q., Ng, K.T. and Robinson, S.R.
(1997) Chicks injected with antisera to either S-100 alpha or
S-100 beta protein develop amnesia for a passive avoidance
task. Neurobiol. Learn Mem., 67: 197–206.
Pelinka, L.E., Kroepfl, A., Leixnering, M., Buchinger, W.,
Raabe, A. and Redl, H. (2004) GFAP versus S100B in serum
after traumatic brain injury: relationship to brain damage
and outcome. J. Neurotrauma, 21: 1553–1561.
Pelinka, L.E., Szalay, L., Jafarmadar, M., Schmidhammer, R.,
Redl, H. and Bahrami, S. (2003) Circulating S100B is in-
creased after bilateral femur fracture without brain injury in
the rat. Br. J. Anaesth., 91: 595–597.
Petrova, T.V., Hu, J. and Van Eldik, L.J. (2000) Modulation of
glial activation by astrocyte-derived protein S100B: differen-
tial responses of astrocyte and microglial cultures. Brain
Res., 853: 74–80.
Petzold, A., Keir, G., Lim, D., Smith, M. and Thompson, E.J.
(2003) Cerebrospinal fluid (CSF) and serum S100B: release
and wash-out pattern. Brain Res. Bull., 61: 281–285.
Povlishock, J.T., Becker, D.P., Sullivan, H.G. and Miller, J.G.
(1978) Vascular permeability alterations to horseradish per-
oxidase in experimental brain injury. Brain Res., 153: 223–239.
Raabe, A. and Seifert, V. (2000) Protein S-100B as a serum
marker of brain damage in severe head injury: preliminary
results. Neurosurg. Rev., 23: 136–138.
Rickmann, M., Wolff, J.R. and Meyer, D.L. (1995) Expression
of S100 protein in the vestibular nuclei during compensation
of unilateral labyrinthectomy symptoms. Brain Res., 688:
8–14.
Rothermundt, M., Peters, M., Prehn, J.H. and Arolt, V. (2003)
S100B in brain damage and neurodegeneration. Microsc.
Res. Tech., 60: 614–632.
Rothoerl, R.D., Brawanski, A. and Woertgen, C. (2000) S-100B
protein serum levels after controlled cortical impact injury in
the rat. Acta Neurochir. (Wien.), 142: 199–203.
Rustandi, R.R., Drohat, A.C., Baldisseri, D.M., Wilder, P.T.
and Weber, D.J. (1998) The Ca(2+)-dependent interaction
of S100B(beta beta) with a peptide derived from p53. Bio-
chemistry, 37: 1951–1960.
Savola, O., Pyhtinen, J., Leino, T.K., Siitonen, S., Niemela, O.
and Hillbom, M. (2004) Effects of head and extracranial in-
juries on serum protein S100B levels in trauma patients.
J. Trauma, 56: 1229–1234; discussion 1234.
Schafer, B.W., Wicki, R., Engelkamp, D., Mattei, M.G. and
Heizmann, C.W. (1995) Isolation of a YAC clone covering a
cluster of nine S100 genes on human chromosome 1q21: ra-
tionale for a new nomenclature of the S100 calcium-binding
protein family. Genomics, 25: 638–643.
Scotto, C., Mely, Y., Ohshima, H., Garin, J., Cochet, C.,
Chambaz, E. and Baudier, J. (1998) Cysteine oxidation in the
mitogenic S100B protein leads to changes in phosphorylation
by catalytic CKII-alpha subunit. J. Biol. Chem., 273:
3901–3908.
Selinfreund, R.H., Barger, S.W., Pledger, W.J. and Van Eldik,
L.J. (1991) Neurotrophic protein S100 beta stimulates
glial cell proliferation. Proc. Natl. Acad. Sci. U.S.A., 88:
3554–3558.

Shashoua, V.E., Hesse, G.W. and Moore, B.W. (1984) Proteins
of the brain extracellular fluid: evidence for release of S-100
protein. J. Neurochem., 42: 1536–1541.
Shore, P.M., Thomas, N.J., Clark, R.S., Adelson, P.D.,
Wisniewski, S.R., Janesko, K.L., Bayir, H., Jackson, E.K.
and Kochanek, P.M. (2004) Continuous versus intermittent
cerebrospinal fluid drainage after severe traumatic brain in-
jury in children: effect on biochemical markers. J. Neuro-
trauma, 21: 1113–1122.
Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T.
(2002) Repeated mild injury causes cumulative damage to
hippocampal cells. Brain, 125: 2699–2709.
Song, H., Stevens, C.F. and Gage, F.H. (2002) Astroglia induce
neurogenesis from adult neural stem cells. Nature, 417:
39–44.
Suzuki, F., Kato, K. and Nakajima, T. (1984) Hormonal reg-
ulation of adipose S-100 protein release. J. Neurochem., 43:
1336–1341.
Svenmarker, S., Sandstrom, E., Karlsson, T. and Aberg, T.
(2002) Is there an association between release of protein
S100B during cardiopulmonary bypass and memory distur-
bances? Scand. Cardiovasc. J., 36: 117–122.
Van Eldik, L.J., Christie-Pope, B., Bolin, L.M., Shooter, E.M.
and Whetsell Jr., W.O. (1991) Neurotrophic activity of S-100
beta in cultures of dorsal root ganglia from embryonic chick
and fetal rat. Brain Res., 542: 280–285.
Van Eldik, L.J. and Zimmer, D.B. (1987) Secretion of S-100
from rat C6 glioma cells. Brain Res., 436: 367–370.
Vos, P.E., Lamers, K.J., Hendriks, J.C., van Haaren, M.,
Beems, T., Zimmerman, C., van Geel, W., de Reus, H., Biert,
J. and Verbeek, M.M. (2004) Glial and neuronal proteins in
serum predict outcome after severe traumatic brain injury.
Neurology, 62: 1303–1310.
325
Whitaker-Azmitia, P.M., Murphy, R. and Azmitia, E.C. (1990)
Stimulation of astroglial 5-HT1A receptors releases the se-
rotonergic growth factor, protein S-100, and alters astroglial
morphology. Brain Res., 528: 155–158.
Wilder, P.T., Rustandi, R.R., Drohat, A.C. and Weber, D.J.
(1998) S100B(betabeta) inhibits the protein kinase C-depend-
ent phosphorylation of a peptide derived from p53 in a
Ca2+dependent manner. Protein Sci., 7: 794–798.
Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir,
J.K. and Ellis, E.F. (2004) S100B protein is released by in
vitro trauma and reduces delayed neuronal injury. J. Neuro-
chem., 91: 1284–1291.
Winningham-Major, F., Staecker, J.L., Barger, S.W., Coats, S.
and Van Eldik, L.J. (1989) Neurite extension and neuronal
survival activities of recombinant S100 beta proteins that
differ in the content and position of cysteine residues. J. Cell
Biol., 109: 3063–3071.
Wunderlich, M.T., Wallesch, C.W. and Goertler, M. (2004)
Release of neurobiochemical markers of brain damage is re-
lated to the neurovascular status on admission and the site of
arterial occlusion in acute ischemic stroke. J. Neurol. Sci.,
227: 49–53.
Xiong, Z., O’Hanlon, D., Becker, L.E., Roder, J., MacDonald,
J.F. and Marks, A. (2000) Enhanced calcium transients in
glial cells in neonatal cerebellar cultures derived from S100B
null mice. Exp. Cell Res., 257: 281–289.
Zimmer, D.B. and Van Eldik, L.J. (1986) Identification of a
molecular target for the calcium-modulated protein S100.
Fructose-1,6-bisphosphate aldolase. J. Biol. Chem., 261:
11424–11428.
Zimmer, D.B. and Van Eldik, L.J. (1987) Tissue distribution of
rat S100 alpha and S100 beta and S100-binding proteins. Am.
J. Physiol., 252: C285–C289.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 23

Cerebellar injury: clinical relevance and potential in

traumatic brain injury research

Eugene Park1,2, Jinglu Ai1 and Andrew J. Baker1,2,

1
St. Michael’s Hospital, Trauma Research, Toronto, ON, M5B 1W8, Canada
2
University of Toronto, Institute of Medical Sciences, Toronto, ON, M5S 1A, Canada

Abstract: A treatment for traumatic brain injury (TBI) remains elusive despite compelling evidence from
animal models for a variety of therapeutic targets. Numerous animal models have been developed to
address the wide spectrum of mechanisms involved in the progression of secondary injury after TBI.
Evidence from well-established models such as the fluid percussion injury (FPI) device, cortical impact
model, and the impact acceleration model has demonstrated diffuse pathophysiological mechanisms
throughout various brain structures. More specifically, we have recently extended characterization of the
FPI model to include pathophysiological changes in the cerebellum following unilateral fluid percussion.
Data suggest that the cerebellum is susceptible to selective Purkinje cell loss as well as white matter
dysfunction. Despite the cerebellum’s low profile in TBI research, there is evidence to warrant further study
of the cerebellum to examine mechanisms of neuronal death and traumatic axonal injury. Furthermore,
evidence from clinical literature and basic science suggests that some components of TBI pathophysiology
have a basis in cerebellar dysfunction. This review highlights some of the recent findings in cerebellar
trauma and builds an argument for including the cerebellum as a model to assess mechanisms of secondary
injury and its potential contribution to the pathology of TBI.

Keywords: animal models; cerebellum; electrophysiology; FPI; TAI; TBI

Introduction injury mechanisms exists in variety of injured axon

Intense scrutiny of neuronal cell death and trau-
matic axonal injury (TAI) has clarified numerous
molecular and cellular events contributing to the
progression of secondary injury cascades. As our
understanding of molecular events continues to
increase, our description of these events have be-
come increasingly complex. For example, all trau-
matically injured axons were previously thought to
undergo a series of pathophysiological events lead-
ing to disconnection and bulb formation. How-
ever, it has become evident that a spectrum of
Corresponding author. Tel.: +1-416-864-5510; Fax: +1-416-
864-5512; E-mail: bakera@smh.toronto.on.ca
subpopulations (reviewed by Buki and Povlishock
(2006)). These data support a need for in-depth
scrutiny of mechanisms encompassing all regions
susceptible to traumatic brain injury (TBI) in or-
der to formulate an accurate and thorough de-
scription of pathophysiologic processes.
The cerebellum’s role in TBI has received rela-
tively limited scrutiny and characterization as
other structures including the brain stem, hippo-
campus, and cerebral cortex. It is becoming in-
creasingly evident that general mechanistic
descriptions of cell death and axonal injury are
overly simplistic and do not accurately reflect the
full spectrum of events occurring after TBI. Clini-
cal cases of TBI and parallel research in animal

DOI: 10.1016/S0079-6123(06)61023-6 327

328
models have confirmed that the mechanisms of
injury extend to the cerebellum thus building a
case for its relevance in neurotrauma. A thorough
characterization of all affected structures through-
out the brain, including the cerebellum, would
provide a better framework with which to optimize
treatments for TBI. Furthermore, the potential for
discovering or elucidating mechanisms of injury
not yet described would also add to our existing
knowledge of secondary injury mechanisms.
It may be argued that cerebellar trauma is of
limited clinical importance; however, recent studies
have begun to elucidate roles in higher order cog-
nitive function not previously appreciated in cere-
bellar processing (Petrosini et al., 1998; Ramnani,
2006). These higher order functions as well as other
neurological processing functions not previ-
ously known may account for some of the ob-
served behavioral abnormalities in people afflicted
by TBI. In the present discussion we propose seve-
ral lines of evidence to support the role of cere-
bellar injury as an important component of TBI,
and also several unique features that render the
cerebellum a potentially useful structure in which
to assess mechanisms of secondary injury after
trauma.
Clinical evidence of cerebellar trauma
Clinical manifestation of cerebellar dysfunction
include perturbations to stance and gait, character-
ized by a loss of equilibrium, a wide-based stance,
irregular steps, and lateral veering (Mariotti
et al., 2005). Cerebellar deficits can also give rise
to tremors (e.g., Parkinsonian), alterations to
speech resulting in slowed or slurred vocalization,
as well as cerebellar mutism (Gordon, 1996).
Recent studies have also indicated an important
contribution of cerebellar function to visuo-motor
control (Glickstein, 2000). Patients with cerebellar
ataxias exhibit ocular motor defects in target
fixation and coordinating head movements
(Mariotti et al., 2005). Although these deficits are
generally not described in the context of trauma,
it is not unreasonable to draw inference to trau-
matic cerebellar injury which could result in simi-
lar pathological impairments, particularly since
many of the motor, cognitive, and speech impair-
ments described above are also observed in TBI
patients.
There are many reviews on diagnoses, mecha-
nisms, outcome, epidemiology, cost analysis,
rehabilitation, etc., and specific case studies in
TBI involving cerebral injury. However there is
limited clinical literature with direct emphasis on
cerebellar trauma. In 1917, Gordon Holmes
described observations of World War I soldiers
who had sustained gunshot wounds to the cere-
bellum. He described a drunken-sailor type gait,
as well as the postural disturbances commonly
associated with cerebellar dysfunction (Holmes,
1917). A recent report of 26 patients who had
sustained infratentorial injuries is the only clinical
study to date to specifically address the issue of
direct trauma to the cerebellum (Nathoo et al.,
2002). This particular type of injury, although rare
in the civilian population, has been documented
predominantly in military literature (Rish et al.,
1983; Brandvold et al., 1990). Moreover, the au-
thors of this study note that cerebellar injury had
no effect on outcome. It should be noted however,
that the five point Glasgow Outcome Scale (GOS)
used in this particular study might not be an ad-
equately sensitive measure of functional and be-
havioral deficits to detect cerebellar dysfunction.
Furthermore, the GOS is primarily a measure of
global outcome and does not factor in the conse-
quences of motor, speech, and cognitive impacts
on daily living.
Despite limited literature on direct cerebellar
trauma there is a varied collection of literature in
which the cerebellum has been referenced as ex-
hibiting pathological changes including selective
cell loss, altered metabolism, and white matter in-
jury after focal and diffuse TBI. For example,
Purkinje cell loss has been noted in the brains of
boxers and implicated as a significant contributor
to ataxias and Parkinsonian tremors (Guterman
and Smith, 1987; Unterharnscheidt, 1995a, b).
Another report cites hypertrophic olivary and
Purkinje cell degeneration following chronic TBI
(Anderson and Treip, 1973). Post-traumatic intra-
cerebellar hemorrhage, though an unusual clinical
occurrence, has been observed after head trauma
and is an important clinical entity with respect to

neurological outcome (Bostrom et al., 1992;
D’Avella et al., 2001). There have also been sev-
eral reports of cerebellar atrophy following TBI
(Krauss et al., 1995; Soto-Ares et al., 2001; Gale et
al., 2005). Furthermore, a subgroup of ataxic TBI
patients may have etiologic basis in cerebellar
dysfunction after injury (Chester and Reznick,
1987; Mysiw et al., 1990). Crossed cerebellar
diaschisis (CCD), a well-documented phenome-
non of altered or depressed metabolic flow and
activity in the hemisphere of the cerebellum cont-
ralateral to the side of cortical lesion or injury, has
also been documented after TBI (Alavi et al.,
1997). Ipsilateral metabolic changes as well as
hypermetabolic changes have also been docu-
mented (Shamoto and Chugani, 1997; Niimura
et al., 1999). A loss of afferent inputs from the
cortico-ponto-cerebellar pathway is believed to
result in deactivation of the targets in the cerebel-
lum, although the exact physiological mechanism
is not entirely understood (Feeney and Baron,
1986). The occurrence of CCD underscores the
importance of synaptic connectivity between
the cerebrum and the cerebellum. This would sug-
gest TBI resulting in a loss of cortical synaptic
input to the cerebellum that results in an unknown
volume of information processing that is not
taking place.
The relationship between ischemia and cere-
bellar injury is also of significance as Purkinje
cells are highly susceptible to ischemic injury
(Bhatia et al., 1995; Welsh et al., 2002). Ischemia
is a common component of non-penetrating head
injuries and an important contributor to second-
ary injury mechanisms (Graham et al., 1978,
1989). Examination of 151 cases of fatal head
injury revealed that the majority of ischemic
damage occurred in the hippocampus (81% of
patients); however, there was also a significant
proportion of patients with evidence of ischemic
injury in the cerebellum (44%; Graham et al.,
1978). Furthermore, TBI, including multisystem
trauma patients, often present with hypoxic-
hypotensive or hemorrhagic complications
(Miller et al., 1978, 1981; The Brain Trauma
Foundation, 2000) increasing the likelihood of
global brain ischemia and subsequent cerebellar
injury.
329
Cerebellar injury in animal models
Purkinje cell vulnerability
Several studies in animal models of TBI have docu-
mented vulnerability of Purkinje cells following
forebrain injury. Both midline and unilateral fore-
brain FPI have been shown to result in significant
and delayed cell death of Purkinje neurons accom-
panied by the presence of activated microglia
(Fukuda et al., 1996; Mautes et al., 1996). Other
studies have supported these findings by demon-
strating the presence of FluorojadeB positive cells
in the cerebellum, consistent with the morphology
of Purkinje neurons (Sato et al., 2001; Hallam
et al., 2004). Using double label fluorescent
immunohistochemistry, we recently demonstrated
selective vulnerability of Purkinje neurons in the
posterior regions of the cerebellum following uni-
lateral FPI in the cerebrum (Park et al., 2006b). In
this study we demonstrated that the majority of
Purkinje neurons were lost in the acute phase of
injury (24 h) while delayed cell loss occurred in the
middle and anterior regions of the cerebellum at 7
and 14 days post-injury at higher grades of injury
severity (Fig. 1).
Examination of the coronal plane cerebellar
sections following forebrain FPI indicates no
evidence of medio-lateral banding (unpublished
observations). Others, however, have reported
parasagittal banding patterns of Purkinje cell
loss using the CCI model of trauma (Weber, J.,
personal communication), perhaps indicative of
variations in injury biomechanics to the cerebel-
lum. Although the mechanisms of Purkinje cell
loss following trauma are still unknown, there
are several potential explanations to consider
with relevance to neurotrauma research. Presy-
naptic hyperexcitability, differential gene expres-
sion, pre- and post-synaptic cell survival, and
neuronal–glial interactions are several possible
avenues to be explored. The cerebellum offers
several unique features to address these mecha-
nisms as contributors to secondary injury. Fur-
thermore, these areas have a broad range of
application to the study of TBI in general and
are not limited to analysis of cerebellum specific
effects.
330

Fig. 1. Double label immunohistochemistry indicates selective Purkinje cell death in the cerebellum following forebrain fluid per-
cussion trauma: (a) Sham, (b) 1 day post-injury, (c) 14 days post-injury. Scale bar ¼ 50 mm. (d) Quantification of surviving Purkinje
cells from the posterior cerebellum following four grades of fluid percussion trauma indicates a dose–response effect at 1 day post-
injury (red asterisk). A significant decline in Purkinje cell numbers relative to sham animals is observed as early as 1 day post-injury in
the 2–2.5 atm injury groups (*). The cartoon (left) indicates the area of fluid percussion trauma (red arrow) and the region of Purkinje
cell quantification (red circle). (Adapted with permission from Park et al., 2006b.)

Presynaptic hyperexcitability presynaptic hyperexcitability in parallel fiber-

The synaptic architecture of the Purkinje cells,
consisting of glutamatergic inputs from climbing
and parallel fibers, creates an environment in
which the potential for synaptic mediated excito-
toxicity is high (for review see Slemmer et al.,
2005). Our laboratory has demonstrated that
direct cerebellar trauma results in delayed
Purkinje cell synaptic connections (Ai and Baker,
2002). Given that a single Purkinje cell receives
input from
200,000 parallel fibers (Fox and
Barnard, 1957), the potential for a presynaptic
excitotoxic event is high. In addition, a single
climbing fiber, originating from the inferior olive,
forms up to 1500 synaptic connections on the
proximal dendrite of a Purkinje cell (Strata and

Rossi, 1998). Excitation of this afferent pathway
with ibogaine administration results in excitotoxic
Purkinje cell death while ablation with 3-AP is
protective to Purkinje cells (O’Hearn and Molliver,
1997). The pattern of cell death from climbing fiber
depolarization results in very distinct medio-lateral
banding pattern of cell death likely corresponding
to the terminal afferent projections of the climbing
fibers. However, there are factors beyond simple
afferent targets to consider in this discussion of
Purkinje cell vulnerability.
Gene expression
Despite a seemingly homogeneous and redundant
arrangement of cellular architecture throughout
the cerebellum (Voogd and Glickstein, 1998;
Ramnani, 2006), there exist subtle differences
within cell populations that can give rise to vastly
different pathologies. In particular, differential
expression of genes involved in metabolism and
cell signaling have been demonstrated within sub-
sets of Purkinje neurons with both medio-lateral
and anterior–posterior patterns of expression
(reviewed by Herrup and Kuemerle (1997) and
Sarna and Hawkes (2003)) (Fig. 2). Furthermore
patterned gene expression in Purkinje neurons
occurs independently of afferent synaptic organi-
zation. The effectiveness of neuroprotective strat-
egies targeting these differentially expressed gene
products can be readily ascertained through histo-
logical examination for the presence or absence of
banding patterns (see O’Hearn and Molliver,
1997). There are numerous candidate proteins
that are expressed in distinct patterns within the
cerebellum that are of interest to neurotrauma
research including heat shock proteins, zebrin
expression, calcium binding proteins, and amino
acid transporters (Herrup and Kuemerle, 1997).
There may also be others that have not yet been
characterized to date. Also of importance is the
maintenance of the compartmentalized expression
of these gene products across mammalian species
permitting for a degree of consistency in cross-
species comparisons of mechanism and role in
TBI.
331
Pre- and post-synaptic cell survival
The cerebellum is a potentially useful structure in
which to assess the effects of neuronal cell loss in
pre- and post-synaptic targets following TBI. For
example, the lurcher mouse phenotype resulting
from a mutation of the Grid2 gene for the
d2-glutamate receptor (d2-GluR) results in chronic
depolarization and loss of Purkinje cells (Zuo
et al., 1997). Subsequent to Purkinje neuron death
is the loss of granule cells as well as retrograde loss
of inferior olive input (Heckroth and Eisenman,
1991; Heckroth, 1992; Zanjani et al., 1998).
Anterograde effects manifest as a loss of deep
cerebellar nuclei (Heckroth, 1994). Whether these
effects are specific to the mutation of the d2-GluR
or are a common response to Purkinje cell loss has
not been elucidated but may be a valuable area of
research to pursue. The extensive literature
describing afferent tracing to the cerebellum as
well as its relatively simple synaptic organization
makes it a suitable model in which to evaluate
retrograde and anterograde fates of injured or
dying neurons. These studies could add to our
understanding of how focal injuries affect distal
targets through retrograde or anterograde signa-
ling dysfunction or cell loss.
Neuronal– glial cell communication
The role of neuronal–glial communication follow-
ing neurotrauma remains a controversial issue
with evidence to supporting glial scarring as
inhibitor of endogenous repair mechanisms along
with compelling evidence to indicate neuroprotec-
tive roles as well (reviewed by Sofroniew (2005)).
In the cerebellum, Bergmann glia residing in the
molecular layer form a complex functional and
structural relationship with Purkinje cells
(reviewed by Bellamy (2006)). These specialized
astrocytes form a sheath around the Purkinje
neuron soma and synapses. This physical interac-
tion has functional implications as stimulation of
climbing and parallel fibers have been shown to
activate inward currents in Bergmann glia (Bergles
et al., 1997; Clark and Barbour, 1997; Bellamy
and Ogden, 2005). These currents are in part
332

Fig. 2. Cartoon representation of cerebellar compartmentation in the mouse, as revealed by the expression of zebrin II in subsets of
Purkinje cells (Adapted with permission from Sillitoe and Hawkes, 2002): anterior, dorsal and posterior views are shown. The
cerebellar vermis is divided into 10 lobules (I–X). However, a more fundamental parcellation is into four transverse zones — anterior
(AZ), central (CZ), posterior (PZ), and nodular (NZ). The AZ and PZ are striped — all Purkinje cells in the CZ and NZ express zebrin
II uniformly (although stripes can be revealed here by using other markers, such as the small heat shock protein, HSP25 (Armstrong et
al., 2000)). Complementary views of a cerebellum whole mount immunoperoxidase stained for zebrin II are shown on the right.
(Adapted with permission from Sarna and Hawkes, 2003.)

contributed to by glial-expressed alpha-amino-3- represent a specialized form of astrocyte, they

hydroxy-5-methyl-4-isoxazole propionic acid maintain the characteristic reactive astrocyte res-
(AMPA) receptors, GluR1 and GluR4, as well as ponse to injury following trauma. We have demon-
glial glutamate transporters, GLAST and GLT-1 strated the presence of reactive astrocytes in
(Bellamy, 2006). Although the Bergmann glia regions of Purkinje cell loss following forebrain

Fig. 3. (a) Confocal image of sham tissue expression of GFAP in Bergmann glia astrocytes, in red, in close proximity to Purkinje cells,
labeled in green. (b) Following forebrain fluid percussion trauma there is a marked increased in GFAP immunoreactivity in regions of
cerebellar injury indicated by the loss of Purkinje cells and increase in GFAP immunoreactivity. Scale bar ¼ 50 mm. (Adapted with
permission from Park et al., 2006b.)
trauma as indicated by increased GFAP expression
(Fig. 3). The close anatomical coupling and func-
tional relationship between Purkinje cells and Berg-
man glia presents an excellent opportunity to exa-
mine the changes in neuronal and glial
communication following TBI. Specifically, the
cerebellum is an ideal system in which to perform
patch-clamping experiments in brain slices, which
permits maintenance of cellular architecture
and synaptic connectivity. These features may be
helpful in elucidating the changes in communica-
tion and synaptic plasticity between glial cells and
neurons in pathophysiological states.
Cerebellar white matter injury
Numerous animal models of TBI have reported
evidence of TAI in the cerebellum (Lighthall et al.,
1990; Shima and Marmarou, 1991; Foda and
Marmarou, 1994; Hoshino et al., 2003). Informa-
tion regarding the functional consequences of
such injuries, however, is limited. We recently
characterized functional and structural changes in
cerebellar white matter following forebrain FPI
and demonstrated significant and persistent deficits
and pathological changes in the cerebellum (Park
et al., 2006a). Early neurofilament degradation and
333
parallel accumulation in a subset of axons was ob-
served at 1 day post-injury while persistent calpain-
mediated degradation of aII-spectrin was observed
up to 14 days post-injury. The acute degradation of
heavy neurofilament chain (NF200) and prolonged
degradation of calpain-mediated aII-spectrin
would suggest temporally preferential targeting of
calpain substrates, NF200 and aII-spectrin, in the
cerebellum. Another possibility is that proteolysis
of these substrates occurred in subsets of axons
with different modes of axonal injury progression.
The observations of NF200 and aII-spectrin
degradation are not novel concepts in the study
of TAI. However, the results reveal a further level
of complexity in TAI with respect to calpain’s
temporal substrate specificity.
Compound action potential recordings from cere-
bellar white matter have also demonstrated the use-
fulness of electrophysiological techniques in assess-
ment of white matter function after TBI. The results
indicate a consistent CAP response within selected
regions. Interestingly, recordings from the middle
cerebellar lobe produce a primarily fast conducting
myelinated response whereas the posterior and ante-
rior cerebellar lobes exhibit a two-peaked response
including both fast and slow myelinated and un-
myelinated responses, respectively (Fig. 4). This
334

Fig. 4. (a) Compound action potentials were recorded from three regions of cerebellar white matter as indicated by the black-boxed
regions. (b) CAP waveforms were dependent on location of recording. Anterior and posterior waveforms consisted of distinct fast and
slow components. The middle cerebellar CAP response was primarily a fast myelinated signal. (c) CAP values from cerebellar white
matter following forebrain fluid percussion trauma indicated a significant decline in electrophysiological function (*) in the posterior
and middle regions of the cerebellum which persisted at 14 days post-injury. (Adapted with permission from Park et al., 2006a.)

presents an opportunity to examine the effects of
TBI and potential therapeutic interventions on both
myelinated and demyelinated axon populations. A
recent study demonstrated varying electrophysio-
logical susceptibilities of these two axonal popu-
lations in the corpus callosum (Reeves et al., 2005).
Given the relative lack of electrophysiological data
available for TBI white matter injury studies, a
comparison of results between the cerebellum and
the corpus callosum may reveal whether these sus-
ceptibilities are structure specific or are a general
principle of injured axons throughout the brain.

The cerebellum as a model of neurotrauma
A key question in the discussion of cerebellar
injury and its application in neurotrauma research
is ultimately whether it is clinically relevant. We
would argue ‘yes’ for the following reasons. There
is sufficient evidence of clinical manifestation of
cerebellar injury following TBI. In addition to the
role in motor coordination, the role of the cere-
bellum in multiple higher order function is becom-
ing more appreciated and hence a recognition of its
potential importance if injured. The traumatized
cerebellum offers numerous technical opportuni-
ties to examine relevant areas of research related to
mechanisms of cellular injury, neuronal–glial
pathophysiological interactions, TAI, post-injury
synaptic plasticity, selective neuronal vulnerability,
and mechanisms of diffuse injury remote from the
location of initial trauma. From an anatomical
perspective, there are notable differences in human
and rodent cerebellar placement that is likely to
result in differences on the production of injury
biomechanics between these two species. However,
there are structural similarities in the cerebellum,
such as the cellular organization and foliated seg-
ments that are maintained throughout all mam-
malian species (Voogd and Glickstein, 1998). This
gyrencephalic property is not maintained in the
cerebral cortex between species and represents an
area in which the biomechanics of injury produc-
tion may differ significantly between higher and
lower order species. Despite the pros and cons of
anatomical similarity between species, the intent of
this discussion is to provide a rationale for inclu-
sion of the cerebellum not only for its clinical and
functional importance in TBI pathophysiology,
but also as a structure in which to examine the
mechanistic phenomena of TBI mechanisms in
general. We propose that comparison and com-
plement of studies across various injury models
will not only validate but also elucidate novel
mechanism of secondary injury.
Conclusion
Traumatic brain injury continues to be a health
care issue of epidemic proportions as a leading
335
cause of morbidity and mortality in young adults.
Despite the prevalence of this silent epidemic,
there is little therapeutic benefit that has translated
from the neuroprotective strategies developed in
animal models of TBI (Bullock et al., 1999; Faden,
2001; Narayan et al., 2002; Tolias and Bullock,
2004). The failure of clinical trials highlights a lack
of complete understanding of the complexity of
TBI pathophysiology. This includes optimizing
therapeutic time windows and clarifying the mul-
tiple pathways leading to cell death. Addressing
these issues will require multifaceted approaches
and balancing the inhibition of secondary injury
mechanisms while not adversely affecting normal
physiologic function (Faden, 2002; Ikonomidou
and Turski, 2002). We believe that this will be best
achieved through complementation and compari-
sons between existing and novel injury paradigms.
The cerebellum is one such novel area that remains
to be fully appreciated and described in the context
of TBI.
Abbreviations
AMPA alpha-amino-3-hydroxy-5-
methyl-4-isoxazole propionic acid
CCD crossed cerebellar diaschisis
FPI fluid percussion injury
GluR glutamate receptor
GOS Glasgow outcome scale
NF200 heavy neurofilament chain
TAI traumatic axonal injury
TBI traumatic brain injury
Acknowledgment
We thank the Ontario Neurotrauma Foundation
for its funding support.
References
Ai, J. and Baker, A. (2002) Presynaptic hyperexcitability at
cerebellar synapses in traumatic injury rat. Neurosci. Lett.,
332: 155–158.
Alavi, A., Mirot, A., Newberg, A., Alves, W., Gosfield, T.,
Berlin, J., Reivich, M. and Gennarelli, T. (1997) Fluorine-18-
FDG evaluation of crossed cerebellar diaschisis in head
injury. J. Nucl. Med., 38: 1717–1720.
336
Anderson, J.R. and Treip, C.S. (1973) Hypertrophic olivary
degeneration and Purkinje cell degeneration in a case of long-
standing head injury. J. Neurol. Neurosurg. Psychiatry, 36:
826–832.
Armstrong, C.L., Krueger-Naug, A.M., Currie, R.W. and
Hawkes, R. (2000) Constitutive expression of the 25-kDa
heat shock protein Hsp25 reveals novel parasagittal bands of
purkinje cells in the adult mouse cerebellar cortex. J. Comp.
Neurol., 416: 383–397.
Bellamy, T.C. (2006) Interactions between Purkinje neurones
and Bergmann glia. Cerebellum, 5: 116–126.
Bellamy, T.C. and Ogden, D. (2005) Short-term plasticity of
Bergmann glial cell extrasynaptic currents during parallel
fiber stimulation in rat cerebellum. Glia, 52: 325–335.
Bergles, D.E., Dzubay, J.A. and Jahr, C.E. (1997) Glutamate
transporter currents in bergmann glial cells follow the time
course of extrasynaptic glutamate. Proc. Natl. Acad. Sci.
U.S.A., 94: 14821–14825.
Bhatia, K.P., Brown, P., Gregory, R., Lennox, G.G., Manji,
H., Thompson, P.D., Ellison, D.W. and Marsden, C.D.
(1995) Progressive myoclonic ataxia associated with coeliac
disease. The myoclonus is of cortical origin, but the patho-
logy is in the cerebellum. Brain, 118(Pt 5): 1087–1093.
Bostrom, K., Helander, C.G. and Lindgren, S. (1992) Blunt
basal head trauma: aspects of unconsciousness. Acta Neuro-
chir. Suppl. (Wien.), 55: 25–28.
Brandvold, B., Levi, L., Feinsod, M. and George, E.D. (1990)
Penetrating craniocerebral injuries in the Israeli involvement
in the Lebanese conflict, 1982–1985: analysis of a less
aggressive surgical approach. J. Neurosurg., 72: 15–21.
Buki, A. and Povlishock, J.T. (2006) All roads lead to discon-
nection? Traumatic axonal injury revisited. Acta Neurochir.
(Wien.), 148: 181–194.
Bullock, M.R., Lyeth, B.G. and Muizelaar, J.P. (1999) Current
status of neuroprotection trials for traumatic brain injury:
lessons from animal models and clinical studies. Neurosur-
gery, 45: 207–220.
Chester, C.S. and Reznick, B.R. (1987) Ataxia after severe head
injury: the pathological substrate. Ann. Neurol., 22: 77–79.
Clark, B.A. and Barbour, B. (1997) Currents evoked in Berg-
mann glial cells by parallel fibre stimulation in rat cerebellar
slices. J. Physiol., 502(Pt 2): 335–350.
D’avella, D., Cacciola, F., Angileri, F.F., Cardali, S., La Rosa,
G., Germano, A. and Tomasello, F. (2001) Traumatic intra-
cerebellar hemorrhagic contusions and hematomas. J. Neuro-
surg. Sci., 45: 29–37.
Faden, A.I. (2001) Neuroprotection and traumatic brain injury:
the search continues. Arch. Neurol., 58: 1553–1555.
Faden, A.I. (2002) Neuroprotection and traumatic brain injury:
theoretical option or realistic proposition. Curr. Opin. Neurol.,
15: 707–712.
Feeney, D.M. and Baron, J.C. (1986) Diaschisis. Stroke, 17:
817–830.
Foda, M.A. and Marmarou, A. (1994) A new model of diffuse
brain injury in rats. Part II: morphological characterization.
J. Neurosurg., 80: 301–313.
Fox, C.A. and Barnard, J.W. (1957) A quantitative study of the
Purkinje cell dendritic branchlets and their relationship to
afferent fibres. J. Anat., 91: 299–313.
Fukuda, K., Aihara, N., Sagar, S.M., Sharp, F.R., Pitts, L.H.,
Honkaniemi, J. and Noble, L.J. (1996) Purkinje cell vulner-
ability to mild traumatic brain injury. J. Neurotrauma, 13:
255–266.
Gale, S.D., Baxter, L., Roundy, N. and Johnson, S.C. (2005)
Traumatic brain injury and grey matter concentration: a
preliminary voxel based morphometry study. J. Neurol.
Neurosurg. Psychiatry, 76: 984–988.
Glickstein, M. (2000) How are visual areas of the brain con-
nected to motor areas for the sensory guidance of movement?
Trends Neurosci., 23: 613–617.
Gordon, N. (1996) Speech, language, and the cerebellum. Eur.
J. Disord. Commun., 31: 359–367.
Graham, D.I., Adams, J.H. and Doyle, D. (1978) Ischaemic
brain damage in fatal non-missile head injuries. J. Neurol.
Sci., 39: 213–234.
Graham, D.I., Ford, I., Adams, J.H., Doyle, D., Teasdale,
G.M., Lawrence, A.E. and Mclellan, D.R. (1989) Ischaemic
brain damage is still common in fatal non-missile head injury.
J. Neurol. Neurosurg. Psychiatry, 52: 346–350.
Guterman, A. and Smith, R.W. (1987) Neurological sequelae of
boxing. Sports Med., 4: 194–210.
Hallam, T.M., Floyd, C.L., Folkerts, M.M., Lee, L.L.,
Gong, Q.Z., Lyeth, B.G., Muizelaar, J.P. and Berman,
R.F. (2004) Comparison of behavioral deficits and acute
neuronal degeneration in rat lateral fluid percussion and
weight-drop brain injury models. J. Neurotrauma, 21:
521–539.
Heckroth, J.A. (1992) Development of glutamic acid dec-
arboxylase-immunoreactive elements in the cerebellar cortex
of normal and lurcher mutant mice. J. Comp. Neurol., 315:
85–97.
Heckroth, J.A. (1994) Quantitative morphological analysis of
the cerebellar nuclei in normal and lurcher mutant mice. I:
morphology and cell number. J. Comp. Neurol., 343:
173–182.
Heckroth, J.A. and Eisenman, L.M. (1991) Olivary morphol-
ogy and olivocerebellar topography in adult lurcher mutant
mice. J. Comp. Neurol., 312: 641–651.
Herrup, K. and Kuemerle, B. (1997) The compartmentalization
of the cerebellum. Annu. Rev. Neurosci., 20: 61–90.
Holmes, G. (1917) The symptoms of acute cerebellar injuries
due to gunshot wounds. Brain, 40: 461–535.
Hoshino, S., Kobayashi, S., Furukawa, T., Asakura, T. and
Teramoto, A. (2003) Multiple immunostaining methods to
detect traumatic axonal injury in the rat fluid-percussion
brain injury model. Neurol. Med. Chir. (Tokyo), 43:
165–174.
Ikonomidou, C. and Turski, L. (2002) Why did NMDA recep-
tor antagonists fail clinical trials for stroke and traumatic
brain injury? Lancet Neurol., 1: 383–386.
Krauss, J.K., Wakhloo, A.K., Nobbe, F., Trankle, R.,
Mundinger, F. and Seeger, W. (1995) Lesion of

dentatothalamic pathways in severe post-traumatic tremor.
Neurol. Res., 17: 409–416.
Lighthall, J.W., Goshgarian, H.G. and Pinderski, C.R. (1990)
Characterization of axonal injury produced by controlled
cortical impact. J. Neurotrauma, 7: 65–76.
Mariotti, C., Fancellu, R. and Di Donato, S. (2005) An
overview of the patient with ataxia. J. Neurol., 252:
511–518.
Mautes, A.E., Fukuda, K. and Noble, L.J. (1996) Cellular re-
sponse in the cerebellum after midline traumatic brain injury
in the rat. Neurosci. Lett., 214: 95–98.
Miller, J.D., Butterworth, J.F., Gudeman, S.K., Faulkner, J.E.,
Choi, S.C., Selhorst, J.B., Harbison, J.W., Lutz, H.A.,
Young, H.F. and Becker, D.P. (1981) Further experience in
the management of severe head injury. J. Neurosurg., 54:
289–299.
Miller, J.D., Sweet, R.C., Narayan, R. and Becker, D.P. (1978)
Early insults to the injured brain. JAMA, 240: 439–442.
Mysiw, W.J., Corrigan, J.D. and Gribble, M.W. (1990) The
ataxic subgroup: a discrete outcome after traumatic brain
injury. Brain Inj., 4: 247–255.
Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A.,
Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C.,
Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D.,
Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D.,
Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller,
N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W.,
Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E.,
Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P.,
Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G.,
Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich,
M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L.
(2002) Clinical trials in head injury. J. Neurotrauma, 19:
503–557.
Nathoo, N., Chite, S.H., Edwards, P.J. and van Dellen, J.R.
(2002) Civilian infratentorial gunshot injuries: outcome anal-
ysis of 26 patients. Surg. Neurol., 58: 225–233.
Niimura, K., Chugani, D.C., Muzik, O. and Chugani, H.T.
(1999) Cerebellar reorganization following cortical injury in
humans: effects of lesion size and age. Neurology, 52:
792–797.
O’Hearn, E. and Molliver, M.E. (1997) The olivocerebellar
projection mediates ibogaine-induced degeneration of Pur-
kinje cells: a model of indirect, trans-synaptic excitotoxicity.
J. Neurosci., 17: 8828–8841.
Park, E., Liu, E., Shek, M., Park, A. and Baker, A. (2006a)
Heavy neurofilament accumulation and alpha-spectrin
degradation accompanies cerebellar white matter functional
deficits following forebrain fluid percussion injury. Exp.
Neurol., in press, Oct. 26, Epub ahead of print.
Park, E., McKnight, S., Ai, J. and Baker, A.J. (2006b) Purkinje
cell vulnerability to mild and severe forebrain head trauma.
J. Neuropathol. Exp. Neurol., 65: 226–234.
Petrosini, L., Leggio, M.G. and Molinari, M. (1998) The cere-
bellum in the spatial problem solving: a co-star or a guest
star? Prog. Neurobiol., 56: 191–210.
337
Ramnani, N. (2006) The primate cortico-cerebellar system:
anatomy and function. Nat. Rev. Neurosci., 7: 511–522.
Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005) Mye-
linated and unmyelinated axons of the corpus callosum differ
in vulnerability and functional recovery following traumatic
brain injury. Exp. Neurol., 196: 126–137.
Rish, B.L., Dillon, J.D. and Weiss, G.H. (1983) Mortality fol-
lowing penetrating craniocerebral injuries: an analysis of the
deaths in the Vietnam Head Injury Registry population.
J. Neurosurg., 59: 775–780.
Sarna, J.R. and Hawkes, R. (2003) Patterned Purkinje cell
death in the cerebellum. Prog. Neurobiol., 70: 473–507.
Sato, M., Chang, E., Igarashi, T. and Noble, L.J. (2001) Neu-
ronal injury and loss after traumatic brain injury: time course
and regional variability. Brain Res., 917: 45–54.
Shamoto, H. and Chugani, H.T. (1997) Glucose metabolism in
the human cerebellum: an analysis of crossed cerebellar di-
aschisis in children with unilateral cerebral injury. J. Child.
Neurol., 12: 407–414.
Shima, K. and Marmarou, A. (1991) Evaluation of brain-stem
dysfunction following severe fluid-percussion head injury to
the cat. J. Neurosurg., 74: 270–277.
Sillitoe, R.V. and Hawkes, R. (2002) Whole-mount
immunohistochemistry: a high-throughput screen for pat-
terning defects in the mouse cerebellum. J. Histochem.
Cytochem., 50: 235–244.
Slemmer, J.E., De Zeeuw, C.I. and Weber, J.T. (2005) Don’t get
too excited: mechanisms of glutamate-mediated Purkinje cell
death. Prog. Brain Res., 148: 367–390.
Sofroniew, M.V. (2005) Reactive astrocytes in neural repair and
protection. Neuroscientist, 11: 400–407.
Soto-Ares, G., Vinchon, M., Delmaire, C., Abecidan, E.,
Dhellemes, P. and Pruvo, J.P. (2001) Cerebellar atrophy after
severe traumatic head injury in children. Child’s Nerv. Syst.,
17: 263–269.
Strata, P. and Rossi, F. (1998) Plasticity of the olivocerebellar
pathway. Trends Neurosci., 21: 407–413.
The Brain Trauma Foundation. The American Association of
Neurological Surgeons. The Joint Section on Neurotrauma
and Critical Care. (2000) Resuscitation of blood pressure and
oxygenation. J. Neurotrauma, 17: 471–478.
Tolias, C.M. and Bullock, M.R. (2004) Critical appraisal of
neuroprotection trials in head injury: what have we learned?
NeuroRx, 1: 71–79.
Unterharnscheidt, F. (1995a) A neurologist’s reflections
on boxing. I: impact mechanics in boxing and injuries other
than central nervous system damage. Rev. Neurol., 23:
661–674.
Unterharnscheidt, F. (1995b) A neurologist’s reflections on
boxing. V: Concluding remarks. Rev. Neurol., 23:
1027–1032.
Voogd, J. and Glickstein, M. (1998) The anatomy of the cere-
bellum. Trends Neurosci., 21: 370–375.
Welsh, J.P., Yuen, G., Placantonakis, D.G., Vu, T.Q., Haiss,
F., O’Hearn, E., Molliver, M.E. and Aicher, S.A. (2002)
Why do Purkinje cells die so easily after global brain
338
ischemia? Aldolase C, EAAT4, and the cerebellar contri-
bution to posthypoxic myoclonus. Adv. Neurol., 89:
331–359.
Zanjani, H.S., Vogel, M.W., Martinou, J.C., Delhaye-Bou-
chaud, N. and Mariani, J. (1998) Postnatal expression of
Hu-bcl-2 gene in Lurcher mutant mice fails to rescue Purkinje
cells but protects inferior olivary neurons from target-related
cell death. J. Neurosci., 18: 319–327.
Zuo, J., De Jager, P.L., Takahashi, K.A., Jiang, W., Linden,
D.J. and Heintz, N. (1997) Neurodegeneration in Lurcher
mice caused by mutation in delta2 glutamate receptor gene.
Nature, 388: 769–773.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 24

Sex differences in brain damage and recovery of

function: experimental and clinical findings

Donald G. Stein

Department of Emergency Medicine, Emory University School of Medicine, Atlanta, GA, USA

Abstract: Until the last decade or so, there was very little systematic examination of sex differences in
recovery from brain injury — most of the work was anecdotal or based on very small studies comparing
males to females. This chapter reviews some of the physiological, morphological, and functional evidence
for sex differences in response to brain injury across the spectrum of development. It also examines more
recent data showing that fluctuations in hormonal status during the menstrual and estrous cycle can play a
determining role in functional outcome in both normal and brain-injured females, and that these hormonal
influences can be measured at both the cellular and behavioral levels.

Keywords: sex differences; brain damage; recovery; hormones; neurosteroids

Introduction animal studies females were excluded because of

Over the last 30 years, development of some 40
different compounds, and industry and government-
supported clinical trials for those compounds, have
failed to find any pharmacologic agent that safely
and effectively led to better functional outcomes
after traumatic brain injury (TBI). With the qual-
ified exception of tissue plasminogen activator
(TPA), the search for an effective treatment for
ischemic stroke yielded the same disappointing
results. Throughout all these trials, no substantive,
systematic attention was paid to whether there
were any differences in brain injury severity and
outcomes between males and females. Many of
the clinical trials did not include females and most of
the laboratory research has focused closely on male
animals. In the human trials, women were often ex-
cluded because investigators were worried about
possible side effects of new drugs on fecundity or
hormonal cycling. Indeed, in both human and
Corresponding author. Tel: +1-404-712-9704; Fax: +1-404-
727-2388; E-mail: DSTEI04@emory.edu
concerns that their hormonal cycling would interfere
with or influence drug metabolism and therefore
‘confound’ the interpretation of results. However,
there have been a number of highly publicized in-
itiatives calling for the systematic study
of gender differences, as well as several edited
volumes on sex differences and brain function
(McIntosh, 1996; Kimura, 1999; Morrison, 2000),
and the number of empirical studies examining
sex differences in brain injury has increased sub-
stantially in the last few years.
Traumatic brain injury is not a unitary event
As most of the articles in this volume will show,
TBI is not a simple event in which a sudden impact
or penetration of the brain parenchyma kills a
certain number of cells. Rather it is a complex
cascade of events that continues to unfold over
time and involves a relatively large number of
genomic, metabolic, cellular and anatomic changes
that can take months to years before complete

DOI: 10.1016/S0079-6123(06)61024-8 339

340
stabilization. To promote functional and adaptive
central nervous system (CNS) repair in the acute
stage of injury, a number of steps are required.
First, the initial progression of events causing
cytotoxicity, inflammation, swelling and short-
and long-term degeneration of nerve and glial
cells must be reduced. This process of repair en-
gages not just the damaged neurons, but the entire
organism’s adaptive capacity. The prevention of
further loss of vulnerable or injured nerve cells can
play an important role in rebuilding damaged cir-
cuitry and in providing a physiological matrix for
later, successful rehabilitation training and ther-
apy. Second, once the cascade of cytotoxic events
is controlled, endogenous, growth-promoting fac-
tors as well as the administration of exogenous,
trophic agents may then stimulate the formation
of new axons, dendrites and synapses or even en-
hance neurogenesis in the damaged adult brain
(Brinton and Wang, 2006; Kobayashi et al., 2006;
Sailor et al., 2006; Shankaran et al., 2006; Shivraj
Sohur et al.; 2006, Taupin, 2006). Taken together,
all these changes require considerable time and all
occur in a complex genomic and proteomic milieu
that can be dramatically affected by a number of
contextual parameters. We now know that such
parameters include the subject’s health, sex and
genomic and hormonal status during development
and at the time of injury. It is important to note
here that there is no empirically based ‘principle’
of neuroscience claiming that unless recovery of
function occurs within the first few weeks after an
injury, there will never be any recovery at all (see
Dancause et al. (2005) for review and discussion
on this issue).
For the purposes of this chapter, one of the
key questions to be asked is, ‘‘Are there, in fact,
sex differences in the outcome of brain injury, and
could they be traced to differences in hormonal
status or factors related to steroid synthesis in the
brain before, during, or after traumatic damage?’’
It has typically been assumed that sex differences
in the cognitive, sensory, or motor performance of
males and females are due to the divergences in the
anatomical structure and neuronal connectivity of
the male and female brain. This is often referred
to in the experimental literature as sexual dimor-
phism. In an excellent recent review, ‘‘Why Sex
Matters for Neuroscience,’’ Larry Cahill (2006)
points out that there are ‘copious’ sex influences
on brain function and anatomy that are often dis-
regarded in neuroscience research. He lists five
misconceptions concerning sex influences on brain:
(1) sex differences are small and unreliable (there-
fore not worth considering?); (2) average differ-
ences between the sexes are due to a few extreme
cases in a distribution — this was the same argu-
ment presented in the early days of work on
recovery of function — i.e., extremely rare phe-
nomena are not worthy of serious study; (3) differ-
ences within sex are more substantial than between
sex and therefore can be dismissed as trivial; (4)
any sex differences are likely to be explained by the
effects of estrogen, primarily during development;
and (5) if no sex differences in behavior exist, it
must mean that the neural mechanisms underlying
behavior(s) are identical for males and females,
although a number of CNS recording and imaging
studies now show that this is not the case (Gur and
Gur, 2004; Rider and Abdou, 2004; Stein, 2004;
Cahill, 2006).
Are there structural differences between the male
and female brain that can affect plasticity and
outcome of injury?
As Cahill has noted, much of the earlier literature
on sex differences in vertebrate brain structure
emphasized the locus of control for mating, vocali-
zation (in songbirds), reproduction, and aggres-
sion and territorial behaviors in the male and
female nervous systems (Arnold et al., 2004;
Collaer et al., 2004). Following on the rapid de-
velopments in non-invasive imaging techniques, a
substantial number of recent experiments have
studied sex differences in metabolic activity in a
variety of brain areas as men and women go about
performing sensory, motor and cognitive tasks.
This is perhaps the one area of clinical neuro-
science showing the most interest in studying
whether sexual dimorphisms underlie the differ-
ences in male and female behaviors (see Gur and
Gur (2004) and Becker (2002) for comprehensive
reviews). For the present chapter and volume the
goal is to determine whether sexual dimorphisms

in the anatomy of the brain could account for
any beneficial or detrimental differences in the re-
sponse of the nervous system to traumatic injuries.
Sex differences in the gross and cellular anatomy of
the brain
There are now many anatomical and physiological
studies describing sex differences in structure/
function relationships in the brain (Kimura,
1992; Shaywitz et al., 1995; Kimura, 1999; Cahill,
2006). Recently, Rabinowicz et al. (2002) exami-
ned 86 different areas in the brains of 6 males
and 5 females, 12–24 years old. These investiga-
tors measured cortical thickness, neuronal and
glial packing densities (the volume of cells in a
geometrically defined area), and neuronal and as-
trocyte sizes (the size of the ‘neuropil’). The
authors reported that females had more overall
neuropil than males, but the males had higher
neuronal densities (more synaptic contacts /mm3).
The men had smaller but more numerous units,
while women had larger cells, especially in the
left hemisphere (women are more frequently right-
handed and have better language abilities). In
another recent report, Goldstein et al. (2001)
used MRI to obtain T1-weighted three-dimen-
sional images from 48 normal adults of similar
education, age, socioeconomic status, intelligence,
and handedness. These investigators found that, in
general, women had larger cortical brain volumes
than men; however, the frontomedial cortex, the
amygdala and the hypothalamus were larger in
men. In particular, there was greater sexual dimor-
phism in those brain regions which, compared to
homologous animal studies, show the greatest num-
bers of sex steroid receptors during development.
One area of the cortex that can be affected by
sex, age, and hormonal milieu of the subject is the
medial frontal cortex, implicated by numerous
lesion studies in spatial learning and cognition
(Markham and Juraska, 2002). The frontal cortex
and its related structures have also received
considerable attention in the functional recovery
literature in both humans and laboratory animals.
Markham et al. (2005) examined both young and
aged male and female rats (the latter in various
341
stages of the estrous cycle) and measured dendritic
spine density and arborization (the major site of
excitatory activity) in the anterior cingulate cortex
to determine whether age, sex, and ovarian hor-
monal status would change the morphology of
these cells. In general, young males had greater
spine densities and arborization than females,
although both sexes showed an expected decline
in the medial frontal cortex with age. However,
the aged females showed less age-related drop-off
of spine density and arborization than the males.
This may be because the aged female rats, unlike
human females, continue to produce ovarian hor-
mones, which could prevent some of the cell loss.
It is also worth noting that the Juraska laboratory
claims that males also show an age-related loss of
dendritic spines in the motor cortex while females
do not (unpublished meeting abstract).
Other investigators have reported sex differences
in the long-term synaptic potentiation (LTP) of
dentate granule cells of the hippocampus. This
activity is thought to subserve hippocampally
mediated learning and memory. For example,
Maren (1995) recorded LTP in young male and
female rats after single-pulse stimulation of the
perforant path. In this study, stimulation led to
a greater magnitude of LTP in the male rats
compared to females and was attributed to the
males having higher levels of NMDA receptor
activation than the females. In this context, since
males have higher levels of LTP activation, and
since hippocampus and temporal cortex are
particularly susceptible to seizures, it would be in-
teresting to learn whether male or female patients
with brain injuries have a higher incidence of
post-traumatic epilepsy. There are two additional
points to note in this context. First, Goldstein
et al. (2001) showed that when relative difference
in overall brain size is taken into consideration,
while the female hippocampus is larger than
that of the male, males have a bigger volume of
the CA1 field and larger neurons than females.
Second, females with very high levels of estrogen
during their menstrual cycle are more likely to
suffer from catamenial epilepsy; this could be
due to the greater sensitivity of the female hippo-
campal cells to neurotransmitters, glucocorticoids,
and NMDA receptor binding (see Herzog et al.
342
(2004) and Reddy (2004) for more discussion on
this issue).
Although there are currently no direct studies
on humans bearing directly on whether women
might have a higher incidence of post-traumatic
epilepsy, there are some interesting clues. For
example, Smith et al. (2002), using transcranial
magnetic stimulation to activate motor-evoked
potentials by paired-pulse technique in the motor
cortex, evaluated the effects of ovarian hormones
on cortical excitability in healthy human females
(average age of 34 years) during the stages of their
menstrual cycle. The study sought to determine
whether levels of estrogen and progesterone could
affect cortical excitability in a way similar to what
glutamate agonists or benzodiazepines might do at
the glutamate or GABAA receptor(s), respectively.
When estrogen levels were high during the late
follicular stage, the motor-evoked potentials were
significantly higher as well. During the luteal phase
of the cycle when progesterone was at its peak,
the motor-evoked potentials were significantly
lower. These data were taken to show that sex
hormones could induce either inhibition or acti-
vation of cerebrocortical activity as well as blood
flow, thus affecting drug metabolism and possibly
CNS mechanisms involved in cognition, per-
ception and the CNS response to injury. These
data can be taken to suggest that, at least for
females, any dynamic brain imaging studies should
take into consideration the patient’s hormonal
state during testing and evaluation.
Sex differences in the brain may be present in glial
as well as neuronal morphology
Recently, Conejo et al. (2003) used GFAP-
immunoreactive labeling and unbiased stereologi-
cal counting to determine the numbers of reactive
astrocytes in the CA and CA3 areas of the adult
rat hippocampus in both males and females
(whose measurements were taken during the pro-
estrous phase of their cycle). Males had higher
numbers of astrocytes than females in the CA3
area of the hippocampus, but the females had
higher numbers of those cells in the CA1 region.
It is interesting to note that females in proestrus
are purportedly less susceptible to ischemic dam-
age of the CA1 region (He et al., 2002) than males.
Conejo et al. suggest that the neuroprotection
may be conferred by the higher number of glial
cells secreting neurotrophic factors, an idea which
is receiving growing experimental support. The
higher numbers of glial cells present could also
play a role in more efficient scavenging of cyto-
toxic agents such as excitatory amino acids and
free radicals. Of the various types of glial cells
found in the brain, astrocytes are the most active
in producing neurosteroids such as pregnenalone,
progesterone, dehydroepiandrosterone (DHEA),
testosterone, estradiol, and others (Zwain and
Yen, 1999). Recently, Cerghet et al. (2006) re-
ported that the density of oligodendrocytes and
the expression of myelin proteins in the corpus
callosum, fornix and spinal cord are 20–40%
greater in male compared to female rats. But in
the corpus callosum, the generation of new glia
and their apoptotic loss is twice as great in females
as in males. The authors took this to mean that
the lifespan and turnover of oligodendrocytes in
females is shorter, and this substantial sex dif-
ference suggests that hormones are playing a criti-
cal role in this ‘turnover’ process which would
include the synthesis of new myelin.
To know whether the amounts of such neuro-
steroids differ between the sexes or whether the
levels of neurosteroids or their receptors vary as a
function of the estrous cycle in human females
will require more research. A better understanding
of the role these factors play in mediating CNS
plasticity could help to explain why females
might have more advantage in recovery from
TBI than males, especially if they have the capacity
to generate more myelin following a brain injury
(Cerghet et al., 2006). In support of this hypothe-
sis, it is interesting to note that after brain or spinal
cord injury, the synthesis of pregnenalone, pro-
gesterone, and allopregnanolone increases two to
four times in the areas immediately adjacent to
the injury sites. Concomitant with this rise there
is a significant increase in glial hyperplasia and
the expression of myelin basic protein and re-
myelination in the areas surrounding the injury
(di Michele et al., 2000). It is also worthwhile to
note that the expression of the myelin proteins can

be directly affected by treatment with progesterone
and some of its metabolites — with progesterone
having a direct stimulatory effect (Schumacher
et al., 2004; De Nicola et al., 2006; Labombarda
et al., 2006; Magnaghi et al., 2006).
In the brain itself, McCarthy et al. (2002), using
both light and electron microscopy, looked at
the effects of the estrous cycle on changes in the
activity of the TrkA receptor, a major receptor
for neurotrophins in the brain which plays a part
in neurite outgrowth in the dendritic fields of the
hippocampus. When estrogen levels were high,
there was a 16-fold increase in the number of
TrkA-labeled astrocytes, but this number dropped
off precipitously when the animals were ovariec-
tomized. After estrogen replacement, there was
again a 12-fold increase. At the electron micro-
scope level the authors observed that the TrkA-
labeled cells were situated closest to dendrites and
to unmyelinated axons. McCarthy and colleagues
attributed the beneficial effects to estrogen-induced
activity helping the cells play a role in axonal
guidance and synaptic regeneration. This notion
would appear to support the more recent findings
described above. (For more detailed reviews of
the role of hormones and dendritic activity, see
McEwen (1998) and Arnold et al. (2004).)
Much of contemporary neuroscience involves
the use of cell culture preparations, but researchers
rarely, if ever, report the sex of the donor cells,
probably thinking that the sex origin of such cells
would have no consequence for the measures they
are making. A more recent study (Zhang et al.,
2003) now questions this assumption. Zhang et al.
compared cell survival and the levels of phospho-
kinases ERK1, ERK2, Akt and the ‘survival’ pro-
tein Bcl-2 in vitro in embryonic (E19) rat cortical
neurons (cortical plate and ventricular zone) taken
from male and female donors. To analyze cell
survival, the cells were double-labeled with MAP-2
(microtubule-associated protein) and anti-GABA
antibodies on day 14 in culture. The phospho-
kinases were evaluated with Western blots and
optical density. Cells taken from females had
a highly significant increase in survival compared
to cells taken from male donors, although there
did not appear to be any differences in the number
of GABA-ergic neurons. The authors suggest that
343
the female cells underwent less apoptosis than
male cells and that this may have been due to the
significantly higher levels of ERK1, ERK2, Bcl-2
and Akt, which helped to preserve the stability
of mitochondrial membranes and produced less
degradation over time. The authors state that,
‘‘these data are the first demonstration of sexual
dimorphism in neuronal cell signaling,’’ and fur-
ther, that gender may provide a context in which
the same biological stimuli can lead to different
outcomes in disease and behavior. For us, the fact
that hardly any molecular biological investiga-
tors consider sex as a variable in the in vitro phar-
macokinetic studies leads us to wonder how valid
are many of the reports when it comes to defining
the proper parameters for drug design, develop-
ment and application to the treatment of brain
injury and other organic disorders.
To emphasize the point that sex differences are
important in determining outcome of brain injury,
three other studies should be mentioned. First, it
is now known that small, sublethal ischemic insults
can protect from additional ischemic damage in
the adult brain, a phenomenon known as ‘ischemic
preconditioning’ (IPC) (Truettner et al., 2002).
Truettner and colleagues created bilateral cere-
bral artery occlusions to cause mild temporary
ischemia in rats. They then took tissue samples
from the hippocampus, cortex, and striatum to
examine the expression of genes thought to be
implicated in neuroprotection after injury (e.g.,
brain-derived neurotrophic factor (BDNF), nerve
growth factor (NGF), c-jun, and c-fos), and found
significant increases. The authors speculated that
the increase in neurotrophic factors following the
pre-conditioning injury could help with protecting
neurons from a second attack. What is particularly
interesting about this finding is that others have
shown that the extent of tolerance to hypoxia and
ischemia in females varies according to the estrous
cycle and relative levels of estrogen or proges-
terone at the time of the insult.
Second, recent investigations (Kasischke et al.,
1999; von Arnim et al., 2002) using mouse hippo-
campal slice preparation recordings show that
short-term hypoxia will suppress normal electrical
activity of neurons. The return of electrical activity
is measured by population spike amplitudes (PSA)
344
and varies according to stage of estrus. In the
Kasischke et al. experiment, the best overall re-
covery of the PSA was found in the males, but for
females the worst recovery was observed during
proestrus when estrogen was high, and the best
recovery was seen during diestrus when progester-
one was high. This means that if hormonal status
and cycling is not considered when females are
included in brain injury studies, the reported
findings must be considered inconclusive or in-
complete. As a case in point, in the later study by
von Arnim et al. (2002), progesterone receptors
were down-regulated by preconditioning lesions
(seen when progesterone levels are high and asso-
ciated with neuroprotection). The authors observe
that, ‘‘Not only is net hypoxic tolerance gender
dependent but underlying mechanisms are gender
specific. Both the net difference in primary and
induced hypoxic tolerance and the specificity
of the mechanisms conferring the hypoxic toler-
ance should be considered in the development of
future gender-specific therapeutic strategies’’ (p. 87
[italics added]).
The influence of sex differences and hormonal status
on cerebral functions in normal and brain-damaged
subjects
There is now a very substantial literature on the
role of sex hormones in the development and locus
of reproductive, maternal, social, and aggressive
behaviors, to which is now being added a growing
number of studies examining sex differences in
more complex cognitive and sensory functions. In
this area, especially in animal research, lesion ex-
periments combined with behavioral testing have
long been used to exemplify dimorphic cognitive
responses to brain injury. One of the first to use
this technique in studying the development of sex
differences in cognition and recovery after brain
injury was Goldman (1974). Goldman created
orbitofrontal cortical lesions in 50-day-old male
and female rhesus monkeys and then tested them
on an object discrimination learning task. At
about two-and-a-half months of age, the males
with orbitofrontal cortical lesions were impaired in
learning but the females with the same surgery
performed the task as well as intact controls.
In another series of projects Goldman created
the same injuries in males and females, tested the
animals soon after surgery and then again over a
year later for their learning abilities. At first test-
ing, the very young toddler females performed
much better than their male counterparts, but
the differences between sexes disappeared by 18
months of age. For Goldman, this suggested that
males were initially worse because their orbito-
frontal cortex and its expected ‘functions’ matured
more rapidly, so the loss of this structure produced
the early deficits. We have already seen that even
in the embryonic stage of development, female
brain cells appear to have higher titers of neuro-
trophic activity than males and this may confer
some potential protection (and better behavioral
performance) against neuronal loss following in-
jury during infancy (Zhang et al., 2003). Sub-
sequent studies showed that there are neurosteroid
receptors in the developing and adult cortex which
can play a role in mediating injury-induced neural
plasticity (Clark et al., 1988).
In a follow-up primate study, Clark and Gold-
man-Rakic (1989) gave male and female rhesus
monkeys orbitofrontal cortex lesions at about 50
days of age followed by testosterone injections
in five of the females. The injections were given
from birth to 46 days of age. These animals were
then compared to groups of intact monkeys given
testosterone prenatally or at 540 and 675 days
after birth. The animals were then tested behavi-
orally on several cognitive discrimination learning
tasks. Interestingly, intact males learned the tasks
somewhat better than the intact females but the
males with brain injury were more impaired on
object reversal learning (the task required being
able to give up a previously learned strategy) than
the lesion females. When the brain-injured females
were given testosterone, the advantage disap-
peared and the lesion females performed as badly
as the injured males, again purportedly demon-
strating either the neuroprotective effects of the
female sex hormones in a brain injury condition
or, quite possibly, the detrimental effects of giving
testosterone to females.
In adult laboratory rats, sex differences in spar-
ing and recovery of function have also been found

after unilateral lesions of the entorhinal cortex, an
area of the brain thought to be implicated in
working and reference memory processing (Roof
et al., 1993). These investigators looking at spatial
learning performance in both males and females
found that intact animals of both sexes performed
about the same on a task which required them to
swim to a submerged platform that could be lo-
cated only by using room cues as a spatial refer-
ence. After brain injury, the male rats had much
more difficulty remembering the location of the
platform than did the females with the same dam-
age; the latter did not differ from intact males or
females on performance of this task, even though
the entorhinal cortex lesions were very large. As
we learn more about sex differences in the out-
come of brain injury, it is becoming apparent that
there are no simple answers. Some studies show
females have an advantage while others report that
males do better, especially when the lesions are
inflicted early in life while hormonal parameters
are still being developed (Kolb and Cioe, 1996;
Forgie and Kolb, 1998).
Extremely low birth weight (ELBW) children
(o1 kg) also show sex differences in cognitive
abilities as early as 2 years of age (Hindmarsh
et al., 2000). The authors followed almost 400
ELBW children after discharge and tested them
for cognitive abilities at 2 years of age using the
Griffiths Mental Development Scales (locomotor,
personal social, hearing and speech, eye and hand
coordination, and ‘practical reasoning’). For the
most part, the investigators found that, except for
locomotion, the female children were ‘significantly
superior’ in all areas of testing and regardless
of the specific impairments manifested by the
boys and girls at time of testing. These data can be
taken to support other findings showing that, in
general, very young females perform better in
language and social development and that this
difference could be due to exposure to differential
gonadal hormones early in life, which then leads to
more rapid maturation and enhanced rate of func-
tion in cerebral structures of females. Although
other explanations may be offered, this may be one
of the reasons why males tend to have a higher
prevalence of developmental lags and learning dis-
abilities (Arnold, 1996; Berk, 1997).
345
Raz et al. (1995) examined 58 children (34 boys,
24 girls) 3 years or older who had suffered an in-
tracranial hemorrhage (ICH) and who were of
37 weeks of gestation or less at birth. The children
were equated for a variety of factors such as age,
race, mother’s IQ, parents’ education, and socio-
economic status. These children were compared
on the Revised Wechsler Intelligence Test to high
birth risk infants without ICH. The girls signi-
ficantly outperformed the boys when the extent
of insult was controlled for. The authors suggest
that even after the various controls for lesion size,
males seem to be more functionally affected by
early brain injury than females because their brains
are generally more vulnerable to insult. They state
that, ‘‘it is possible that regardless of developmen-
tal factors, the male brain may possess structural
or physiological properties that could not only
affect the extent of cerebral damage, but might
also compromise functional recovery’’ (p. 964).
A subsequent paper by Donders and Hoffman
(2002) also reported that girls with blunt TBI had
better verbal outcome scores on the California
Verbal Learning Test than boys, even after they
were fully equated on a variety of demographic,
premorbid measures, including basic intelligence
pre- and post-injury, and severity and type of in-
jury. After the injury, the boys could not remember
as many words, used fewer adaptive learning strat-
egies and needed more time than girls to work on
test problems. Following their review of the liter-
ature, Donders and Hoffman suggest that the out-
come of sex differences in pediatric TBI could be
due to the neuroprotective effects of sex steroids,
especially progesterone, although there are few, if
any, studies examining the role of hormonal treat-
ments in survivors of pediatric TBI. Although
there are reports of females doing better, some
studies draw the opposite conclusions. For exam-
ple, Morrison et al. (2004) found no statistically
significant differences in outcome between boys
and girls with brain injury and there was even some
trend toward the females being somewhat worse.
Sex differences in the brain-injured adult
The influence of hormonal status (e.g., estrus) on
cognitive performance in adult humans has been
346
more controversial than the findings in children,
perhaps because of the ways in which hormonal
assays are conducted and performance is measured
over the menstrual cycle (see Epting and Overman
(1998) and Mumenthaler et al. (2001) for detailed
discussions on this issue). Data from laboratory
animals may be more precise and consistent be-
cause, among other things, it is easier to obtain a
variety of carefully timed measures of hormonal
fluctuations (see DeGraba and Pettigrew (2000)
for a discussion on some of the issues). In rats
(Kinsley et al., 1999), females that had a number of
litters previously were compared to age-matched
females that never had a litter and to foster
mothers who never delivered any litters but raised
pups that were given to them. The multiparous
females and foster mothers, with higher circula-
ting levels of progesterone, performed much better
on a complex maze task than did the nulliparous
animals.
In rhesus monkeys (Lacreuse et al., 2001), spa-
tial recognition memory performance also varies
across the menstrual cycle. Performance of the ani-
mals was measured on a number of learning tasks
and found to be better during the follicular and
luteal phases than during the peri-ovulatory phase,
when estrogen levels were at their highest. This
finding fits with other data showing that high levels
of estradiol can disrupt learning and memory in
female rats (Galea et al., 2001). Taken together,
these data suggest that either reduced levels of est-
rogen, or relatively higher levels of progesterone,
or possibly both, are beneficial to learning. The
role of estrogen and related hormones in mediating
recovery after brain injury is becoming an increas-
ingly studied issue, especially in light of the re-
cently failed trials with estrogen in the treatment of
stroke and Alzheimer’s disease in women. With
respect to CNS trauma and possibly late-life de-
generative disorders such as Alzheimer’s, it may
also be important to determine first whether the
extent and severity of the diseases vary in relation
to the normal hormonal status of the female. If
natural or induced hormonal cycling does play a
role in the progression of injury or disease, then the
timing of drug therapy to the appropriate stage of
estrus will need to be considered in the treatment
of adolescent learning and attention disorders.
In an animal model of transient forebrain is-
chemia, b-estradiol has been shown to exacerbate
the loss of neurons in the CA1 pyramidal cell
field. When rats were ovariectomized the ischemia
caused a loss of 32% of pyramidal cells, but with
estradiol replacement in intact and ovariectomi-
zed rats the mean cell loss was 54% and 49%,
respectively (Harukuni et al., 2001). Would ad-
ministration of progesterone alone or in combina-
tion with the estradiol have prevented this loss?
Estrogen is a proexcitatory agent and could
have caused neuronal loss by increasing excitotoxi-
city and subsequent inflammation in neurons ren-
dered particularly vulnerable by the ischemic
event. The hippocampus is also particularly sensi-
tive to seizure activity, which could have triggered
the additional loss of the injured neurons. Proges-
terone, in contrast, has been shown to inhibit sei-
zure activity, especially in women with catemenial
epilepsy, and inhibition of progesterone metabo-
lism increases seizure activity (Herzog and Frye,
2003).
Do females generally have better outcomes after
brain damage than males?
There is considerable controversy in the literature
over the issue of whether females recover better
from TBI and stroke than males, though anecdotal
reports claim an advantage for females. Some
clinical case-study reports suggest that women re-
cover better from stroke than men, while others
claim the opposite or no differences between the
sexes. For example, a paper by Kertesz and
McCabe (1977) examined 23 males and 13 females
and found no differences in recovery between the
sexes, whereas Basso et al. (1982) examined males
and females and found long-term aphasia severity
to be much more marked in males than in females,
even though initial impairments were judged to be
the same. Groswasser et al. (1998) looked at the
outcome of ‘severe’ brain injuries in males and
females (72 females, 262 males) and controlled for
age range and severity. Females apparently had
better overall responses to rehabilitation therapy
than males when return-to-work was used as an
outcome measure.

Some reports contest the Groswasser and col-
leagues findings. Kraus et al. (2000) looked at gen-
der differences in outcome after TBI in two trauma
centers and found that mortality was substantially
higher in females than in males with the same ex-
tent of injury. The females who survived had
poorer outcomes compared to males. Supporting
this finding, Farace and Alves (2000) performed a
meta-analysis of previously published TBI studies
to determine whether women did better than men
in TBI outcome on 20 different measures. These
investigators found that fatality rates for women
with TBI were much higher than for men (because
of body size women may suffer more severe inju-
ries; there are also some data to suggest that they
are more often hit by motor vehicles than injured
while driving them). Poorer outcomes were also
reported for women on measures of disability,
seizures, coma, etc. Farace and Alves mention
that sex differences in drug metabolism and other
variables could contribute to the negative findings.
The report makes no mention of attempting to de-
termine the women’s hormonal status before, dur-
ing or after injury to see if these factors could have
contributed to the more detrimental outcomes.
Farace and Alves list a number of reasons why
females may have poorer outcomes than males,
one of them being sex hormones. Recently, Farin
et al. (2003) reported that women have signi-
ficantly greater frequencies of brain swelling than
men and worse outcomes than men if they are un-
der 50 years of age at the time of their injury.
These results could have been due to higher levels
of estrogen relative to progesterone in the younger
women at the time of injury, but this variable
was not studied (i.e., the authors did not mention
whether some of the older women were on hor-
mone replacement therapy (HRT) and whether the
younger women were on contraceptives, both of
which could affect outcomes). It was interesting to
note that after 50 years of age, when both estrogen
and progesterone are diminished, the amount of
brain swelling in older females dropped and was
almost equivalent to that of the males. Another
retrospective study by Coimbra et al. (2003) also
found no evidence of any sex differences in TBI
outcomes between males and females on any of
their measures of recovery or mortality.
347
To add to this controversy, Ratcliffe et al.
(2003) looked at the relationship between gender
and cognitive recovery one year after blunt TBI.
They selected 325 records from the TBI Model
Systems National Database and used multivariate
analyses to evaluate the association between sex
and six cognitive measures of attention — working
memory, verbal memory, language, visual analytic
skills, problem solving, and motor functioning.
Females represented 31% of the total number of
patients tested. The women showed better per-
formance on five of the six cognitive outcome
measures. The authors point out that pre-injury
differences were not a factor in this design because
men did not perform better than women after in-
jury on tasks like visual memory, on which in
the normal condition they typically outperform
females.
The discrepancies among the different outcome
studies can be due to a number of factors, not the
least of which is the way in which patients are
selected for studies. For instance, Obler et al.
(1995) reported in ‘early’ (i.e., 1970s) studies that
lesion site rather than gender was the determining
factor in selecting patients and that this could have
biased outcome results. We now know that the loci
of brain injuries related to aphasia severity and
disorders in linguistic abilities differ between men
and women. According to Kimura (1999), the in-
cidence and severity of speech disorders is much
higher in women if damage is to the anterior cortex
in the left hemisphere, while the reverse is true for
men. Women also have higher incidence of manual
apraxias after lateral frontal cortical injuries, while
men show greater deficits if the damage is in the
posterior cortex. Could the outcome of these
studies be influenced by the women’s hormonal
status at the time of their injury? There is some
tantalizing but indirect evidence bearing on this
question. Hausmann et al. (2002) used radio-
immunoassays taken every three days to evaluate
the hormonal status of healthy women volun-
teers participating in a perceptual asymmetry
study using a variety of behavioral tasks that are
taken to measure cerebral asymmetries. When
the women were high in serum progesterone, there
was less task-dependent cerebral asymmetry on
some of the behavioral tests, leading to what the
348
investigators called ‘interhemispheric decoupling,’
and this was interpreted to mean that progesterone
mediates the extent of cerebral lateralization.
Therefore, as noted earlier, when functional MRIs
are used to evaluate structure–function/metabolic
activities in the brains of women, it may be im-
portant to determine the subjects’ specific hormo-
nal status at the time of testing in order to obtain
more reliable information concerning CNS organi-
zation and whether there is, indeed, contextual
and/or sex related cerebral asymmetry.
With respect to a number of earlier studies (and
in some current ones), Obler et al. (1995) note that
many of the variables that could determine apha-
sia severity were not published (e.g., gender distri-
bution, time since onset of the injury and testing,
specific locus of the injury, prior health status,
hearing status, medication history, etc.), and this
lack of information could certainly have contri-
buted to the variability in the findings. For pur-
poses of the present discussion, especially for
female subjects, hormonal status at the time of
injury is never mentioned. Moreover, it is highly
unlikely that the patients were queried as to
whether they were taking any kind of birth con-
trol medication or HRT prior to or at the time of
their TBI. Such a chronic history of hormone
regulation could affect the outcome of the brain
injury, if, in fact, hormone levels were playing
any role in mediating inflammatory processes that
accompany TBI (Stein, 2005). In general, and es-
pecially when people suffer from a severe brain
injury, there is also accompanying severe bodily
injury and shock. There is evidence that females,
especially when progesterone levels are high rela-
tive to estrogen, maintain better immune function
activity and response to physiologic shock than
males (Wichmann et al., 1996; Zellweger et al.,
1997; Kuebler et al., 2003). These better levels of
immunodepression may also benefit the response
to brain injury.
Especially in older women with stroke or injury,
whether they are/were on HRT prior to or after
the damage is never noted. This could be parti-
cularly important if the older women were or were
not on HRT at the time of their injury. All these
factors could contribute to differences in brain in-
jury outcome, and if there are, as some have in fact
claimed, no differences between the sexes, then this
would need to be determined in systematic experi-
ments controlling for at least some of these vari-
ables before any firm conclusions can be drawn.
This highlights the validity of doing animal experi-
ments on brain injury outcomes; most of the vari-
ables relevant to gender-specific issues can be
controlled better in a laboratory setting, thereby
reducing the variability typical of uncontrolled
clinical assessments.
Clearly, the issues surrounding sex differences
in brain injury are complex and more research is
needed to tease out whether sex differences can
play a role in normal cognitive functions as well as
in TBI prognoses. Given the growing literature
showing that hormones like progesterone, estrogen
or even testosterone can serve as treatments for
TBI (see for example Wright et al. (2006)), it is
likely that the study of sex differences will begin to
receive more attention if for no other reason than
to improve the quality and success of clinical trials
seeking treatments for the victims of acquired
brain injuries.
Acknowledgment
Thanks to Leslie McCann for editorial assistance.
References
von Arnim, C.A., Etrich, S.M., Timmler, M. and Riepe, M.W.
(2002) Gender-dependent hypoxic tolerance mediated via
gender-specific mechanisms. J. Neurosci. Res., 68: 84–88.
Arnold, A.P. (1996) Genetically triggered sexual differentiation
of brain and behavior. Horm. Behav., 30: 495–505.
Arnold, A.P., Agate, R.J. and Carruth, L.C. (2004) Hormonal
and non-hormonal mechanisms of sexual differentiation of
the brain. In: Legato M.J. (Ed.), Principles of Gender-
Specific Medicine. Academic Press, New York, pp. 84–95.
Basso, A., Capitani, E. and Moraschini, S. (1982) Sex differ-
ences in recovery from aphasia. Cortex, 18: 469–475.
Becker, J.B. (2002) Behavioral Endocrinology. MIT Press,
Cambridge, MA.
Berk, L.E. (1997) Child Development. Allyn & Bacon, Boston.
Brinton, R.D. and Wang, J.M. (2006) Therapeutic potential of
neurogenesis for prevention and recovery from Alzheimer’s
disease: allopregnanolone as a proof of concept neurogenic
agent. Curr. Alzheimer Res., 3: 185–190.
Cahill, L. (2006) Why sex matters for neuroscience. Nat. Rev.
Neurosci., 7: 477–484.

Cerghet, M., Skoff, R.P., Bessert, D., Zhang, Z., Mullins, C.
and Ghandour, M.S. (2006) Proliferation and death of
oligodendrocytes and myelin proteins are differentially regu-
lated in male and female rodents. J. Neurosci., 26: 1439–1447.
Clark, A.S. and Goldman-Rakic, P.S. (1989) Gonadal hor-
mones influence the emergence of cortical function in non-
human primates. Behav. Neurosci., 103: 1287–1295.
Clark, A.S., Maclusky, N.J. and Goldman-Rakic, P.S. (1988)
Androgen binding and metabolism in the cerebral cortex of
the developing rhesus monkey. Endocrinology, 123: 932–940.
Coimbra, R., Hoyt, D.B., Potenza, B.M., Fortlage, D.,
Hollingsworth-Fridlund, P. (2003) Does sexual dimorphism
influence outcome of traumatic brain injury patients? The
answer is no! J. Trauma, 54: 689–700
Collaer, M.L., Tory, H.O. and Valkenburgh, M.C. (2004) Do
sex steroid hormones contribute to sexual differentiation of
the human brain? In: Legato M.J. (Ed.), Principles of Gen-
der-Specific Medicine, Vol. 1. Academic Press, New York,
pp. 71–83.
Conejo, N.M., Gonzalez-Pardo, H., Pedraza, C., Navarro,
F.F., Vallejo, G. and Arias, J.L. (2003) Evidence for sexual
difference in astrocytes of adult rat hippocampus. Neurosci.
Lett., 339: 119–122.
Dancause, N., Barbay, S., Frost, S.B., Plautz, E.J., Chen, D.,
Zoubina, E.V., Stowe, A.M. and Nudo, R.J. (2005) Extensive
cortical rewiring after brain injury. J. Neurosci., 25:
10167–10179.
DeGraba, T.J. and Pettigrew, L.C. (2000) Why do neuro-
protective drugs work in animals but not humans? Neurol.
Clin., 18: 475–493.
De Nicola, A.F., Gonzalez, S.L., Labombarda, F., Deniselle,
M.C., Garay, L., Guennoun, R. and Schumacher, M. (2006)
Progesterone treatment of spinal cord injury: effects on re-
ceptors, neurotrophins, and myelination. J. Mol. Neurosci.,
28: 3–15.
Donders, J. and Hoffman, N.M. (2002) Gender differences in
learning and memory after pediatric traumatic brain injury.
Neuropsychology, 16: 491–499.
Epting, L.K. and Overman, W.H. (1998) Sex-sensitive tasks in
men and women: a search for performance fluctuations
across the menstrual cycle. Behav. Neurosci., 112: 1304–1317.
Farace, E. and Alves, W.M. (2000) Do women fare worse: a
metaanalysis of gender differences in traumatic brain injury
outcome. J. Neurosurg., 93: 539–545.
Farin, A., Deutsch, R., Biegon, A. and Marshall, L.F. (2003)
Sex-related differences in patients with severe head injury:
greater susceptibility to brain swelling in female patients 50
years of age and younger. J. Neurosurg., 98: 32–36.
Forgie, M.L. and Kolb, B. (1998) Sex differences in the
effects of frontal cortex injury: role of differential hormonal
experience in early development. Behav. Neurosci., 112:
141–153.
Galea, L.A., Wide, J.K., Paine, T.A., Holmes, M.M., Ormerod,
B.K. and Floresco, S.B. (2001) High levels of estradiol dis-
rupt conditioned place preference learning, stimulus response
learning and reference memory but have limited effects on
working memory. Behav. Brain Res., 126: 115–126.
349
Goldman, P.L. (1974) An alternative to developmental plas-
ticity: heterology of CNS structures in infants and adults.
In: Stein D.G., Rosen J.J. and Butters N. (Eds.), Plasticity
and Recovery of Function in the Central Nervous System.
Academic Press, New York, pp. 149–174.
Goldstein, J.M., Seidman, L.J., Horton, N.J., Makris, N.,
Kennedy, D.N., Caviness Jr., V.S., Faraone, S.V. and
Tsuang, M.T. (2001) Normal sexual dimorphism of the adult
human brain assessed by in vivo magnetic resonance imaging.
Cereb. Cortex, 11: 490–497.
Groswasser, Z., Cohen, M. and Keren, O. (1998) Female TBI
patients recover better than males. Brain Inj., 12: 805–808.
Gur, R.G. and Gur, R.E. (2004) Gender differences in the
functional organization of the brain. In: Legato M.J. (Ed.),
Principles of Gender-Specific Medicine, Vol. 1. Academic
Press, New York, pp. 63–70.
Harukuni, I., Hurn, P.D. and Crain, B.J. (2001) Deleterious
effect of beta-estradiol in a rat model of transient forebrain
ischemia. Brain Res., 900: 137–142.
Hausmann, M., Becker, C., Gather, U. and Gunturkun, O.
(2002) Functional cerebral asymmetries during the menstrual
cycle: a cross-sectional and longitudinal analysis. Neuro-
psychologia, 40: 808–816.
He, Z., He, Y.J., Day, A.L. and Simpkins, J.W. (2002)
Proestrus levels of estradiol during transient global cerebral
ischemia improves the histological outcome of the hippo-
campal CA1 region: perfusion-dependent and -independent
mechanisms. J. Neurol. Sci., 193: 79–87.
Herzog, A.G. and Frye, C.A. (2003) Seizure exacerbation as-
sociated with inhibition of progesterone metabolism. Ann.
Neurol., 53: 390–391.
Herzog, A.G., Harden, C.L., Liporace, J., Pennell, P., Schomer,
D.L., Sperling, M., Fowler, K., Nikolov, B., Shuman, S. and
Newman, M. (2004) Frequency of catamenial seizure exacer-
bation in women with localization-related epilepsy. Ann.
Neurol., 56: 431–434.
Hindmarsh, G.J., O’Callaghan, M.J., Mohay, H.A. and
Rogers, Y.M. (2000) Gender differences in cognitive abili-
ties at 2 years in ELBW infants. Extremely low birth weight.
Early Hum. Dev., 60: 115–122.
Kasischke, K., Huber, R., Li, H., Timmler, M. and Riepe, M.W.
(1999) Primary hypoxic tolerance and chemical precondition-
ing during estrus cycle in mice. Stroke, 30: 1256–1262.
Kertesz, A. and McCabe, P. (1977) Recovery patterns and
prognosis in aphasia. Brain, 100(Pt 1): 1–18.
Kimura, D. (1992) Sex differences in the brain. Sci. Am., 267:
118–125.
Kimura, D. (1999) Sex and Cognition. MIT University Press,
Cambridge, MA.
Kinsley, C.H., Madonia, L., Gifford, G.W., Tureski, K.,
Griffin, G.R., Lowry, C., Williams, J., Collins, J., McLearie,
H. and Lambert, K.G. (1999) Motherhood improves learning
and memory. Nature, 402: 137–138.
Kobayashi, T., Ahlenius, H., Thored, P., Kobayashi, R.,
Kokaia, Z. and Lindvall, O. (2006) Intracerebral infusion
of glial cell line-derived neurotrophic factor promotes striatal
neurogenesis after stroke in adult rats. Stroke, 37: 2361–2367.
350
Kolb, B. and Cioe, J. (1996) Sex-related differences in cortical
function after medial frontal lesions in rats. Behav. Neuro-
sci., 110: 1271–1281.
Kraus, J.F., Peek-Asa, C. and McArthur, D. (2000) The in-
dependent effect of gender on outcomes following traumatic
brain injury: a preliminary investigation. Neurosurg. Focus,
8: e5.
Kuebler, J.F., Yokoyama, Y., Janar, D., Toth, B., Rue III,
L.W., Bland, K.I., Wang, P. and Chaudry, I.H. (2003)
Administration of progesterone after trauma and hemorr-
hagic shock prevents hepalocellular injury. Arch. Surg., 138:
727–734.
Labombarda, F., Gonzalez, S., Gonzalez Deniselle, M.C.,
Garay, L., Guennoun, R., Schumacher, M. and De Nicola,
A.F. (2006) Progesterone increases the expression of myelin
basic protein and the number of cells showing NG2 immuno-
staining in the lesioned spinal cord. J. Neurotrauma, 23:
181–192.
Lacreuse, A., Verreault, M. and Herndon, J.G. (2001) Fluctu-
ations in spatial recognition memory across the menstrual
cycle in female rhesus monkeys. Psychoneuroendocrinology,
26: 623–639.
Magnaghi, V., Veiga, S., Ballabio, M., Gonzalez, L.C., Garcia-
Segura, L.M. and Melcangi, R.C. (2006) Sex-dimorphic
effects of progesterone and its reduced metabolites on gene
expression of myelin proteins by rat Schwann cells. J. Peri-
pher. Nerv. Syst., 11: 111–118.
Maren, S. (1995) Sexually dimorphic perforant path long-term
potentiation (LTP) in urethane-anesthetized rats. Neurosci.
Lett., 196: 177–180.
Markham, J.A. and Juraska, J.M. (2002) Aging and sex influ-
ence the anatomy of the rat anterior cingulate cortex. Neuro-
biol. Aging, 23: 579–588.
Markham, J.A., McKian, K.P., Stroup, T.S. and Juraska, J.M.
(2005) Sexually dimorphic aging of dendritic morphology in
CA1 of hippocampus. Hippocampus, 15: 97–103.
McCarthy, J.B., Barker-Gibb, A.L., Alves, S.E. and Milner, T.A.
(2002) TrkA immunoreactive astrocytes in dendritic fields of
the hippocampal formation across estrous. Glia, 38: 36–44.
McEwen, B.S. (1998) Multiple ovarian hormone effects on
brain structure and function. J. Gend. Specif. Med., 1: 33–41.
McIntosh, T.K. (1996) Neuropathological sequelae of trau-
matic brain injury: relationship to neurochemical and bio-
mechanical mechanisms. Lab. Invest., 74: 315–342.
di Michele, F., Lekieffre, D., Pasini, A., Bernardi, G., Benav-
ides, J. and Romeo, E. (2000) Increased neurosteroids syn-
thesis after brain and spinal cord injury in rats. Neurosci.
Lett., 284: 65–68.
Morrison, M.F. (2000) Hormones, Gender, and the Aging
Brain: The Endocrine Basis of Geriatric Psychiatry. Cam-
bridge University Press, New York, NY.
Morrison, W.E., Arbelaez, J.J., Fackler, J.C., de Maio, A. and
Paidas, C.N. (2004) Gender and age effects on outcome after
pediatric traumatic brain injury. Pediatr. Crit. Care. Med., 5:
145–151.
Mumenthaler, M.S., O’Hara, R., Taylor, J.L., Friedman, L.
and Yesavage, J.A. (2001) Relationship between variations in
estradiol and progesterone levels across the menstrual cycle
and human performance. Psychopharmacology (Berl.), 155:
198–203.
Obler, L.K., Goral, M. and Albert, M.L. (1995) Variability in
aphasia research: aphasia subject selection in group studies.
Brain Lang., 48: 341–350.
Rabinowicz, T., Petetot, J.M., Gartside, P.S., Sheyn, D., Sheyn,
T. and De, C.M. (2002) Structure of the cerebral cortex in
men and women. J. Neuropathol. Exp. Neurol., 61: 46–57.
Ratcliffe, J.J., Greenspan, A., Goldstein, F.C., Wright, D.W.,
Stringer, A., Bell, D., Bushnik, T., Englander, J., Hammond,
F., Novack, T. and Whyte, J. (2003) Gender and traumatic
brain injury: do the sexes fare differently? Society for Experi-
mental Medicine Abstract. Acad. Emerg. Med., 10: 491.
Raz, S., Lauterbach, M.D., Hopkins, T.L., Glogowski, B.K.,
Porter, C.L., Riggs, W.W. and Sander, C.J. (1995) A female
advantage in cognitive recovery from early cerebral insult.
Dev. Psychol., 31: 958–966.
Reddy, D.S. (2004) Role of neurosteroids in catamenial epi-
lepsy. Epilepsy Res., 62: 99–118.
Rider, V.A. and Abdou, N.I. (2004) Sexual dimorphism and the
immune system. In: Legato M.J. (Ed.), Principles of Gender-
Specific Medicine, Vol. 2. Academic Press, New York,
pp. 1071–1081.
Roof, R.L., Duvdevani, R. and Stein, D.G. (1993) Gender in-
fluences outcome of brain injury: progesterone plays a pro-
tective role. Brain Res., 607: 333–336.
Sailor, K.A., Ming, G.L. and Song, H. (2006) Neurogenesis as a
potential therapeutic strategy for neurodegenerative diseases.
Expert Opin. Biol. Ther., 6: 879–890.
Schumacher, M., Guennoun, R., Robert, F., Carelli, C., Gago,
N., Ghoumari, A., Gonzalez Deniselle, M.C., Gonzalez, S.L.,
Ibanez, C., Labombarda, F., Coirini, H., Baulieu, E.E.
and De Nicola, A.F. (2004) Local synthesis and dual actions
of progesterone in the nervous system: neuroprotection
and myelination. Growth Horm. IGF Res., 14(Suppl A):
S18–S33.
Shankaran, M., King, C., Lee, J., Busch, R., Wolff, M. and
Hellerstein, M.K. (2006) Discovery of novel hippocampal
neurogenic agents by using an in vivo stable isotope labeling
technique. J. Pharmacol. Exp. Ther., 319: 1172–1181.
Shaywitz, B.A., Shaywitz, S.E., Pugh, K.R., Constable, R.T.,
Skudlarski, P., Fulbright, R.K., Bronen, R.A., Fletcher,
J.M., Shankweiler, D.P., Katz, L., et al. (1995) Sex differ-
ences in the functional organization of the brain for language
[see comments]. Nature, 373: 607–609.
Shivraj Sohur, U., Emsley, J.G., Mitchell, B.D. and Macklis,
J.D. (2006) Adult neurogenesis and cellular brain repair with
neural progenitors, precursors and stem cells. Philos. Trans.
R. Soc. Lond. B. Biol. Sci., 361: 1477–1497.
Smith, M.J., Adams, L.F., Schmidt, P.J., Rubinow, D.R.
and Wassermann, E.M. (2002) Effects of ovarian hormones
on human cortical excitability. Ann. Neurol., 51: 599–603.
Stein, D.G. (2004) Brain trauma, sex hormones, neuronal sur-
vival and recovery of function. In: Legato M. (Ed.), Princi-
ples of Gender-Specific Medicine, Vol. 1. Academic Press,
New York, pp. 104–115.

Stein, D. G. (2005) The case for progesterone. In: Singh M. and
Simpkins J.W. (Eds.), The Future of Estrogen and Hormone
Theraphy in Postmenopausal Women: What Basic Science
and Clinical Studies Teach Us, Vol. 1052. Annals of the New
York Academy of Science, pp. 152–169.
Taupin, P. (2006) Adult neurogenesis in mammals. Curr. Opin.
Mol. Ther., 8: 345–351.
Truettner, J., Busto, R., Zhao, W., Ginsberg, M.D. and Perez-
Pinzon, M.A. (2002) Effect of ischemic preconditioning on
the expression of putative neuroprotective genes in the rat
brain. Brain Res. Mol. Brain Res., 103: 106–115.
Wichmann, M.W., Zellweger, R., Demaso, C.M., Ayala, A.
and Chaudry, I.H. (1996) Enhanced immune responses in
females, as opposed to decreased responses in males follow-
ing haemorrhagic shock and resuscitation. Cytokine, 8:
853–863.
351
Wright, D.W., Kellermann, A.L., Hertzberg, V.S., Clark, P.L.,
Frankel, M., Goldstein, F.C., Salomone, J.P., Dent, L.L.,
Harris, O.A., Ander, D.S., Lowery, D.W., Patel, M.M.,
Denson, D.D., Gordon, A.B., Wald, M.M., Gupta, S.,
Hoffman, S.W., Stein, D.G. (2006) Progesterone for acute
traumatic brain injury. Ann. Emerg. Med.
Zellweger, R., Wichmann, M.W., Ayala, A., Stein, S., Demaso,
C.M. and Chaudry, I.H. (1997) Females in proestrus state
maintain splenic immune functions and tolerate sepsis better
than males. Crit. Care Med., 25: 106–110.
Zhang, L., Li, P.P., Feng, X., Barker, J.L., Smith, S.V. and
Rubinow, D.R. (2003) Sex-related differences in neuronal cell
survival and signaling in rats. Neurosci. Lett., 337: 65–68.
Zwain, I.H. and Yen, S.S. (1999) Neurosteroidogenesis in as-
trocytes, oligodendrocytes, and neurons of cerebral cortex of
rat brain. Endocrinology, 140: 3843–3852.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 25

Heat acclimation: a unique model of physiologically

mediated global preconditioning against traumatic
brain injury

Na’ama A. Shein1,2, Michal Horowitz2 and Esther Shohami1,

1
Department of Pharmacology, Hebrew University, Jerusalem, Israel
2
Department of Physiology, Hebrew University, Jerusalem, Israel

Abstract: Sub-lethal exposure to practically any harmful stimulus has been shown to induce consequent
protection against more severe stress. This preconditioning (PC) effect may be achieved by exposure to
different stressors, indicating that the induction of tolerance involves activation of common protective
pathways. Chronic exposure to moderate heat (heat acclimation, HA) is a unique PC model, since this
global physiological adaptation, as opposed to discrete organ PC, has been shown to induce cross-tolerance
against other stressors, including closed head injury (CHI). HA animals show accelerated functional re-
covery after injury which is accompanied by reduced secondary brain damage. However, the precise
mechanisms underlying this phenomenon have not been thoroughly studied until recently. Here we will
address the concept of PC, highlighting the unique properties of HA as a model which can be used for the
study of endogenous protective pathways triggered by PC procedures. Several molecular mechanisms which
are suggested to mediate HA-induced neuroprotection will also be discussed, bringing to light their pote-
ntial contribution to the development of traumatic brain injury treatment strategies utilizing therapeutic
augmentation of endogenous defense mechanisms.

Keywords: closed head injury; erythropoietin; heat acclimation; hypoxia inducible factor-1; inflammation;
neuroprotection; preconditioning

The phenomenon of preconditioning which enhance intrinsic defensive ability. Conceiv-

Throughout history, several scholars have appro-
ached the notion that harmful substances or
stressors may also engulf beneficial effects. These
apparently conflicting observations were linked
by early-time savants such as Hippocrates and
Paracelsus to the extent of exposure or the given
dose, suggesting that sub-lethal subjection to a
noxious insult may bring about the activation of
endogenous protective response mechanisms,
Corresponding author. Tel.: +972-2-675-7513; Fax: +972-2-
675-8741; E-mail: esty@cc.huji.ac.il
ably, the coping ability of pre-exposed organisms
will consequently be augmented in the setting of
recurring challenges (Fig. 1).
Janoff (1964), first introduced the terms ‘pre-
conditioning’ (PC) and ‘tolerance’ to describe the
phenomenon, setting the stage for numerous later
studies which described many different PC models
and demonstrated the beneficial effects of this
protocol in providing consequent protection.
Traditionally, PC protocols were aimed at pro-
viding enhanced defensive ability to a specific or-
gan, with cardiac PC being the most extensively
studied model. In the mid 1980s, it was initially

DOI: 10.1016/S0079-6123(06)61025-X 353

354
Non-preconditioned
Endogenous protection activation
Preconditioned
Attenuated
stress / harmful
stimuli exposure
Fig. 1. General model of preconditioning. Pre-exposure to attenuated stress/harmful stimuli enhances endogenous protective mecha-
nisms, leading to improved overall outcome following subsequent more severe insult. Note that the beneficial effect brought about by
preconditioning results from augmented defensive ability rather than a direct effect on the extent of deleterious processes.
reported that pre-exposure to an attenuated is-
chemic insult affords subsequent myocardial pro-
tection (Murry et al., 1986). The beneficial effects
of myocardial ischemic PC were later shown to
include the reduction of infarct size, a decline in
the occurrence of cardiac arrhythmias and accel-
erated functional recovery following ischemia
(Shiki and Hearse, 1987; Cleveland et al., 1996).
In vitro experiments also indicated the protective
potential of myocardial PC (Ikonomidis et al.,
1994).
Since then, ischemic PC-induced cardio-protection
has been reproduced by numerous groups in
various species including humans (Vohra and
Galinanes, 2006). In addition, myocardial PC has
shown to be clinically relevant in the setting of
myocardial infarction and in patients receiving
repeated cycles of aortic clamping and reperfu-
sion prior to cardiopulmonary bypass (CPB)
(Yellon and Dana, 2000). Subsequently, the ben-
eficial effects of ischemic PC on lung preservation
in patients undergoing open heart surgery with
CBP were also reported (see Luh and Yang, 2006
for review).
Similarly, the neuroprotective potential of is-
chemic/hypoxic PC against ischemic injury within
Basal defensive ability
Overall
outcome
Deleterious
processes
time
Time interval Severe Outcome
Hours-weeks stress/harmful measurement
stimuli induction
the central nervous system has been the subject of
extensive research. Ischemic PC, as well as PC by
exposure to hyperbaric oxygen provided protection
in various models of spinal cord injury (Toumpoulis
and Anagnostopolus, 2003; Nie et al., 2006). Nu-
merous studies have revealed the profound effect of
PC by brief ischemic exposure on the outcome of
brain ischemic injury. The beneficial consequences
of such procedures have been demonstrated in
different in vitro (Reshef et al., 1996; Liu et al.,
2000) as well as in vivo models (Sharp et al., 2004;
Glantz et al., 2005, and see Dirnagl et al., 2003;
Blanco et al., 2006 for review). The substantiation
of PC-afforded neuroprotection in animal models
consequently instigated a wide-scale research effort
aimed at exploring methods of employing ischemic
PC in stroke therapy (Schaller, 2005).
Interestingly, over the years, a body of evidence
accumulated, indicating that the initial definition
of PC as the induction of tolerance toward a spe-
cific stressor by attenuated pre-exposure to the
same type of stress might have been too narrow.
It has been shown that the effects of PC may
be mimicked by exogenous treatments. PC-like
effects have been achieved in various models by
pharmacological interventions utilizing low doses

of otherwise harmful substances. Boeck et al.
(2004) have shown that exposure to low doses of
NMDA induces protection against the hippocam-
pal neurotoxicity of quinolic acid. Analogous ob-
servations were reported while using PC with
thrombin (Hua et al., 2003; Cannon et al., 2005).
Importantly, Dahl and Balfour (1964) demon-
strated that in rats, exposure to short periods of
anoxia induces a ‘whole organism’ protective
effect, indicating that PC can be applied toward
conferring protection to organisms and not only to
discrete organs.
Taken together, these findings suggest that PC is
a global and well-preserved adaptation mechanism
which is aimed at enabling dealing with stress. The
fact that this endogenous mechanism can be
prompted, not only by exposure to stress but also
via administration of exogenous agents, is of great
importance since it suggests that PC-like effects
may be induced at will in a relevant clinical setting.
In addition, since global exposure to external am-
bient anoxia can cause PC of the entire organism,
it is conceivable that activation of the protective
mechanisms which underlie PC-induced tolerance
could be achieved by exposure to other types of
environmental stress.
Long-term heat acclimation: physiologically
mediated global preconditioning
Heat acclimation (HA) is a conserved adaptive re-
sponse to exposure to moderately high ambient
temperature. Conceptually, the process may be
delineated as a transition from an early, transient,
‘‘inefficient’’ state to an ‘‘efficient’’ acclimated
homeostatic state (Horowitz, 2003, 2007).
Long-term HA is achieved by a 3–4 week expo-
sure to mild environmental heat. The acclimation
process is biphasic, consisting of two phases which
differ in terms of the controlling mechanisms for
heat dissipation. While increased excitability of the
autonomic nervous system compensates for im-
paired cellular performance during short-term
acclimation, the features of HA include enhanced
cellular function accompanied by decreased excit-
ability (Horowitz and Meiri, 1985; Horowitz, 1994
and for review, see Horowitz, 2002, 2007).
355
Individual thermoregulatory effectors display a
similar pattern of biphasic response during HA.
Salivation, which is utilized by several rodent spe-
cies as the major pathway of evaporative cooling,
clearly demonstrates this phenomenon. While sa-
liva production is inadequate for sufficient heat
dissipation at the onset of acclimation due to im-
paired responsiveness of muscarinic receptors,
once the new acclimated homeostasis has been
achieved, 3 weeks later, glandular efficiency is
regained due to upregulation and a return of these
receptors to pre-acclimation affinity (Kloog et al.,
1985; Horowitz, 2002). Similarly, enhanced auto-
nomic sympathetic activity is utilized in order to
compensate for the desensitization of cardiac
adrenergic receptors during short-term HA,
whereas the HA heart maintains the ability to
produce adequate contractile force, to enhance
work efficiency and to enhance pressure in the
face of reduced ATP utilization (Levi et al., 1993;
Horowitz, 1998, 2002).
At the organism level, the HA phenotype has
been characterized by reduced metabolic and heart
rates, decreased core body temperature, lower tem-
perature thresholds for the activation of heat dissi-
pation mechanisms and an elevated temperature
threshold for the development of thermal injury.
Evidence now indicates that this acclimated pheno-
type results from both post-translational modifica-
tions as well as genomic responses. Maloyan et al.
(1999) showed that heat acclimated rats display
about three times more inducible heat shock protein
72 kDa than do non-acclimated counterparts, thus
showing better coping with heat stress without the
need for de novo HSP synthesis (which character-
izes the heat shock response of non-acclimated
animals). Alterations in genomic responses during
the course of HA have been shown to include
changes in the expression pattern of genes which are
involved in anti-apoptosis, antioxidation and heat
shock response (Horowitz et al., 2004).
Altogether, these responses contribute to an over-
all expansion of the dynamic thermoregulatory
range. Consequently, HA results in enhanced abil-
ity of coping with exposure to environmental heat.
Hence, in accordance with the terminology intro-
duced by Janoff (1964) HA may be viewed as a
long-term physiologically mediated heat-PC.
356
An inseparable outcome of HA is that adjusting
to one stressor can, in addition to evolving primary
adaptations, add to the amount of adjustment to
additional stressors. Such cross-reinforcement
raises the possibility of inducing adaptation to a
stressor without previous exposure to that particu-
lar one.
Indeed, the ability of HA to induce ‘‘cross-
tolerance’’ against a variety of stressors has been
reported. HA has been shown to convey improved
cardiac mechanical and metabolic performance
and reduced injury upon ischemic-reperfusion in-
sult to the heart. Levi et al. (1993) showed that
following total ischemia, acclimated hearts dis-
played enhanced preservation of the cardiac ATP
pool and a delayed decline in intracellular pH.
HA has also been shown to induce cross-tolerance
toward exposure to hyperbaric oxygen. Arieli et al.
(2003) reported a delay in the onset of central
nervous system oxygen toxicity following HA,
demonstrating that the latency to the appear-
ance of evident toxicity was twice as long in heat
acclimated rats.
Taken together, these findings indicate a wide
scale protective effect which is induced by HA.
This observation may be of particular importance
in the setting of traumatic brain injury. A neuro-
protective effect elicited by HA will potentially
have a fundamental effect on functional recovery
upon subsequent injury. Additionally, it could
provide valuable insight into common protective
pathways which are shared by HA and post-injury
recovery mechanisms.
HA-induced neuroprotection in closed head injury
The effects of HA on the outcome of closed head
injury (CHI) have been examined using a well-
established CHI model (described by Shapira et al.,
1988 and Chen et al., 1996). The model reproduces
the post-traumatic sequence of events observed in
humans and consistently recreates several patho-
physiological features such as edema formation,
blood-brain barrier disruption and functional dys-
function.
Shohami et al. (1994b) demonstrated that when
CHI is induced in HA rats, they display faster and
better functional recovery as well as reduced
secondary tissue damage. The motor ability of
the rats was evaluated according to a neurological
severity score (NSS) which is based on the pres-
ence of some reflexes and the ability to perform
motor and behavioral tasks such as beam walking,
beam balance and spontaneous locomotion. HA
rats showed significantly better motor function
48 h following injury as compared with normo-
thermic controls. Additionally, brain edema and
blood-brain barrier disruption were reduced in the
acclimated group. These findings were subse-
quently reinforced in mice (Shein et al., 2005a),
demonstrating significantly better motor function
24 h post-injury accompanied with a reduction in
brain water accumulation in HA mice. These mice
also perform better in a memory test when exam-
ined by the object recognition test (Ennaceur and
Delacour, 1988). When tested 3 days following
CHI, HA mice spent most of their exploration
time at the new object while normothermic coun-
terparts failed to prefer the ‘novel’ object over the
familiar one.
Interestingly, it has also been shown that
neuroprotection may be achieved in vitro by ex-
posing cell cultures to serum derived from HA
animals. Beit Yannai et al. (1998) demonstrated
that when exposing cultures of PC12 cells (a sym-
pathetic pheochromocytoma cell line widely used
as a neuronal cell model) to 1% serum from HA
rats, a significant reduction in cell death occurs as
compared with cultures exposed to serum taken
from normothermic ones. Additionally, the HA
serum treated cultures showed increased neuro-
protection upon exposure to a free radical gener-
ator causing oxidative stress.
Taken together, current data establishes that HA
elicits a continuum of processes resulting in en-
hanced neuroprotection. The fact that the beneficial
effect can be demonstrated in different species as
well as in varying experimental models is in agree-
ment with the observations pointing to the wide-
scale nature of PC-induced tolerance. Although
HA-induced neuroprotection was hypothesized to
be facilitated via common protective pathways
shared by different PC paradigms (Horowitz,
2004), the underlying mechanisms remained poorly
defined for some time. However, recent work has

provided new insight into several molecular proc-
esses which may contribute to the overall functional
improvement.
Mechanisms of HA-induced neuroprotection
The research of HA-induced neuroprotection is
generally aimed at identifying molecular effectors
and signaling cascades which are involved in the
development of the protected phenotype. In the
following sections we describe our own studies
which examined the effects of HA on pathways
which are known to either convey a well-established
protective effect or foster deleterious consequences
in CHI, in an effort to examine whether these
pathways may also play a role in the beneficial
outcome induced by HA. These include low mole-
cular weight antioxidants and anti-inflammatory
capacity, hypoxia inducible factor-1, erythropoietin
receptor expression and signaling as well as acute
post-CHI inflammation and the expression of
neurotrophic factors.
Effect of HA on the ability to cope with oxidative
stress: endogenous and post-CHI low-molecular
weight antioxidant profile
Reactive oxygen species (ROS) have been exten-
sively studied and proposed as candidates for the
elicitation of pathological responses in the patho-
genesis of ischemia and trauma (e.g., Chan,
2001; Kontos, 2001). A number of therapeutic ap-
proaches, based on intervention by scavenging
ROS, have been attempted both in experimental
models and in the clinical setting (for review, see
Vink and Van Den Heuvel, 2004).
Generally, several lines of defense have been
developed by living cells in order to cope with
oxidative stress. These include preventive and re-
pair mechanisms as well as an antioxidant defense
system, consisting of antioxidative enzymes and
low molecular weight antioxidants (LMWA). The
main direct-acting LMWA found within the brain
include the tripeptide glutathione (glu-cys-gly), toco-
pherols (vitamin E), ascorbic acid (vitamin C),
histidine-related compounds, melatonin, uric acid
and lipoic acid. Evaluation of individual LMWA
357
is difficult given the large number of scavengers
and the fact that they interact with each other,
such that increase of one may lead to a decrease of
the other. However, this limitation can be over-
come by the cyclic voltammetry technique. Cyclic
voltammetry was traditionally used in chemistry to
study electron transfer between molecules and was
shown initially by Kissinger et al. (1973) to be
applicable for evaluating antioxidant levels in
vivo. A new method was then developed, enabling
the measurement of overall antioxidant activity
based on the determination of the total reducing
power of the biological tissue or fluid (Kohen
et al., 1999; Kohen and Nyska, 2002).
In our model of CHI, we have previously shown
the activation of the arachidonic acid metabolic
cascade which produces free radicals (Shohami
et al., 1987, 1989) as well as enhanced cytokine
production (Shohami et al., 1994a), a typical in-
flammatory response involving ROS. In addition,
we have demonstrated that a marked decrease in
LMWA concentrations occurs within minutes
following injury, indicating their immediate utili-
zation in an effort to overcome acute oxidative
stress. The detrimental role of ROS in early-CHI
pathology was also supported by the observation
that ROS-neutralizing compounds are protective
when given soon after CHI (Beit-Yannai et al.,
1996).
We have examined the LMWA in the brains of
HA rats, in an effort to correlate the observed
protection with the ability of the brain to neutral-
ize ROS. A cyclic voltammetry study performed
on the water-soluble fraction of LMWA showed
that the chemical nature of LMWA is not altered
by HA. Interestingly, the concentrations of the
reducing equivalents were significantly lower in the
HA group, i.e. the basal LMWA levels were found
to be lower in HA rats (Beit Yannai et al., 1997).
However, while the relative changes in anodic
currents during the post-CHI period revealed
a decrease from basal (sham) levels in the normo-
thermic animals, this did not occur in the HA
group which sustained levels higher than those
measured in sham controls, for up to a week
following trauma.
In light of these results we suggested that simi-
larly to observations in the ischemic heat
358
acclimated heart (Horowitz et al., 2004), in the
setting of CHI long-term exposure to high ambient
temperature provides a more efficient mechanism
of dealing with ROS displayed by better ability of
adjusting to the subsequent acute stress.
Involvement of hypoxia inducible factor-1 and
erythropoietin signaling
Oxygen deprivation is known to evoke a sequele of
adaptive responses. These responses are aimed at
compensating for decreased aerobic ATP produc-
tion and are regulated by the transcriptional acti-
vator hypoxia inducible factor-1 (HIF-1). HIF-1
regulates the expression of over 70 known target
genes, leading to increased tissue oxygen delivery
and ATP production (Semenza, 2004). The func-
tional factor is a heterodimer, consisting of a
constitutively expressed b subunit and an inducible
a subunit, which undergo dimerization by means
of basic helix-loop-helix PAS domains. Interest-
ingly, the accumulation of HIF-1a and the tran-
scriptional activation of HIF-1 have recently been
shown to also be triggered by non-oxygen-dependent
pathways including ROS, cytokines and several
growth factors and hormones (Chandel et al.,
2000; Haddad, 2002). The fact that many HIF-1
regulated genes are related to oxygen and energy
homeostasis, taken together with its importance in
environmental adaptive processes led to the hypo-
thesis that this factor may also be involved in
the metabolic responses during HA. Furthermore,
preliminary data from Maloyan et al. (2001) and
Bromberg and Horowitz (2004) indicated that the
levels of HIF-1a are increased following HA.
However, the role of HIF-1 in HA remained
unclear and it was not known until 2 years later
whether the increase in the levels of this factor are
of functional importance or only an epipheno-
menon.
The question of the involvement of HIF-1 in
HA was assessed by Treinin et al. (2003). Their
study elegantly demonstrated that HIF-1 is essen-
tial to the development of HA, using a ‘hif-1 loss of
function’ mutant strain of the C. elegans nema-
tode. As opposed to wild type nematodes, which
demonstrated an acclimated phenotype following
an appropriate acclimation protocol, the loss of
function strain was unable to acclimate to heat.
The contribution of HIF-1 targeted pathways
to HA-induced cardiac cross-tolerance has also
recently been established. Maloyan et al. (2005)
reported an increase in HIF-1a protein levels and
HIF-1 DNA binding activity as well as enhanced
target gene mRNA expression in the hearts of HA
rats.
HIF-1 regulates the expression of erythropoietin
(Epo). Epo, traditionally recognized as the main
erythropoietic cytokine, has been shown to be ex-
tensively expressed within the brain (Buemi et al.,
2002). The cytokine as well as its specific receptor
(EpoR) are expressed by neuronal, glial and brain
capillary endothelial cells and are upregulated by
ischemic as well as metabolic stress (Bernaudin
et al., 1999, 2000). Importantly, Epo has been
shown to engulf neuroprotection in a large variety
of in vitro and in vivo models of brain injury as
well as in human stroke patients (Morishita et al.,
1997; Agnello et al., 2002; Catania et al., 2002;
Ehrenreich et al., 2002; Grasso et al., 2002).
We have previously established the neuropro-
tective effect of Epo in our model of CHI (Yatsiv
et al., 2005). When treated with recombinant
human Epo (rhEpo) 1 and 24 h following CHI,
mice displayed lower motor deficits and acceler-
ated restoration of cognitive function. In addition,
a reduction in the number of apoptotic neurons as
well as caspase-3 expression was found in the
rhEpo-treated group. This was also accompanied
by better preservation of axons at the trauma area
and reduced activation of glial cells in the injured
hemisphere.
In light of this, we hypothesized that HIF-1 and
in turn, Epo signaling, may also play a part in the
benefits brought about by HA. We proceeded to
initially examine the basal and post-injury levels of
HIF-1a in HA mice and normothermic counter-
parts. The levels of HIF-1a were found to be much
higher following HA and this upregulation was
sustained 4 h following injury (Shein et al., 2005a).
Subsequently, the expression of Epo and EpoR
were evaluated. An increase in both basal and
post-injury receptor levels was found, but surpris-
ingly this was not accompanied by upregulation of
Epo itself. This was unexpected, given the large
 359

number of studies which highlighted the associa- When we examined the levels of total and phos-
tion of HIF-1 upregulation and increased expres- phorylated Akt in HA mice and compared them
sion of Epo. However, several other reports have with the levels found in non-acclimated controls,
demonstrated that the behavior of HIF-1 target we observed an increase in post-injury phosphory-
genes varies among different genes and is also lated Akt in the HA group (Shein et al., 2005b).
greatly dependent on the experimental model used We may therefore conclude that HA has both
(Jones and Bergeron, 2001). In addition, HIF-1 constitutive and dynamic effects; namely estab-
activation which is not followed by increased lishing higher basal levels and rapidly elevated
Epo expression has previously been described post-injury levels of both HIF-1a and EpoR and
in the presence of pro-inflammatory cytokines that this in turn leads to increased post-CHI
(Hellwig-Burgel et al., 1999). enhancement of Akt phosphorylation.
The levels of EpoR are known to be upregulated
by ischemia and hypoxia (Siren et al., 2001; Genc
Effect of HA anti-inflammatory capacity and acute
et al., 2004). The established involvement of is-
post-CHI inflammation
chemia in CHI pathophysiology may therefore
explain the post-injury upregulation of this recep-
The process of acute inflammation, initiated follow-
tor. Since HA has been shown to mediate enhanced
ing CHI has been shown to involve the upregulation
heat tolerance as well as cross-tolerance via in-
of pro-inflammatory cytokines (Shohami et al.,
creased gene responsiveness (Horowitz et al., 2004),
1994a). These cytokines, namely tumor necrosis
we propose that it is likely that the marked increase
factor alpha (TNF-a) and interleukin-1b (IL-1b),
in EpoR in HA mice following injury is due to an
when present during the initial time period follow-
additive effect of both injury-induced ischemia and
ing CHI, have been implicated of involvement in
increased responsiveness resulting from HA.
blood-brain barrier disruption and associated with
Since the levels of EpoR were consistently
subsequent edema formation (Shohami et al., 1997;
higher following HA only, as well as in HA mice
Stahel et al., 2000). Knoblach and Faden (1998)
subjected to CHI, we calculated the overall recep-
have previously demonstrated that the post-injury
tor/ligand Epo ratio as a measure of overall Epo
expression of these deleterious effectors may be at-
signaling ability. This ratio was significantly higher
tenuated by pre-injury treatment with the anti-
in the HA group and led us to further examine
inflammatory cytokine interleukin-10 (IL-10). We
downstream intracellular signaling which is medi-
therefore conceived that elevated pre-injury levels
ated by this receptor.
of anti-inflammatory mediators may lead to
We chose to focus on the extent of phosphory-
decreased post-CHI acute inflammation in HA ani-
lation of Akt as an indicator of intracellular anti-
mals. Our current data indicates that HA mice
apoptotic Epo/EpoR signaling. Following EpoR
do indeed display both an increase in basal,
activation, Akt is phophorylated via the phospha-
pre-injury anti-inflammatory capacity as well as a
tidylinositol-3 kinase pathway. When phosphory-
subsequent decrease in post-injury expression of
lated, Akt subsequently acts as a kinase and
pro-inflammatory cytokine mRNA (Shein et al.,
phosphorylates pro-apoptotic mediators including
2006). Immunohistochemical findings now indicate
Bad, caspase-9, the forkhead transcription factor
that HA also enhances the presence of brain-derived
and glycogen synthase kinase 3b. Each of the
neurotrophic factor-positive microglia (Shein et al.,
cellular systems targeted by Akt is inactivated by
2006).
phosphorylation, resulting in a blockade of apo-
ptotic cell death. Additionally, Akt phosphorylat-
ion has been associated with neuroprotection and Summary
has been suggested as one of the molecular path-
ways mediating Epo-induced neuroprotection Current data suggests that the neuroprotective
(Digicaylioglu et al., 2001; Chong et al., 2002; effect of HA is likely to be mediated by a network
Kilic et al., 2006). of pathways which work in concert to yield overall
360

functional improvement. In light of the findings to
date, it can be suggested that HA leads to both the
potentiation of endogenous protective pathways as
well as to the attenuation of detrimental post-
injury processes. Although some changes are con-
stitutive and observed following the acclimation
period itself while others are dynamic and only
apparent after injury, they all eventually confer
modifications of the post-injury state (Fig. 2). The
fact that the effects of HA can be demonstrated
individually for different molecular mediators sub-
stantiates the importance of this physiological
adaptive mechanism as a model for enabling
better definition of the role of discrete endogenous
neuroprotective pathways in the development of
tolerance.
HA is a unique model in terms of providing in-
sight into physiologically induced neuroprotec-
tion. Unraveling the physiological mechanisms of
tolerance development could have profound
implications for treatment in the pathophysiologi-
cal setting of brain trauma, since future inter-
vention strategies may be aimed at enhancing
endogenous protective ability in a physiological-
mimetic manner.

Higher levels in HA Lower levels in HA

Abbreviations
HIF-1α
CHI closed head injury
Pre-CHI EpoR CPB cardiopulmonary bypass
Epo erythropoietin
IL-10/IL-4 EpoR erythropoietin receptor
HA long-term heat acclimation
HIF-1 hypoxia inducible factor-1
IL-1b interleukin-1beta
LMWA IL-10 interleukin-10
TNF-α LMWA low molecular weight antioxi-
HIF-1α
Post-CHI IL-1β dants
EpoR PC preconditioning
p-Akt ROS reactive oxygen species
TNFa tumor necrosis factor alpha.
Note, factors which are higher in HA (either pre or
post CHI) are associated with neuroprotection.
Factors which are lower in HA after CHI are
associated with damage

Fig. 2. A schematic overview summarizing current experimen- References

tal data for mechanisms suggested as being involved in HA-
induced neuroprotection in closed head injury (CHI). Follow-
ing long-term heat acclimation (HA), the pre-injury levels of
hypoxia inducible factor-1alpha (HIF-1a), erythropoietin re-
ceptor (EpoR) and anti-inflammatory cytokines are increased.
Higher levels of HIF-1a and EpoR are sustained in HA animals
following injury as well and are complemented by an increase in
post-CHI low molecular weight antioxidants (LMWA) and
phosphorylated Akt (p-Akt) as well as a reduction in early post-
injury levels of pro-inflammatory cytokines. Abbreviations:
CHI, closed head injury; HA, long-term heat acclimation;
LMWA, low molecular weight antioxidants; IL-10, interleukin-
10; IL-4, interleukin 4; TNF-a, tumor necrosis factor alpha;
IL-1b, interleukin-1beta; HIF-1a, hypoxia inducible factor-1alpha;
EpoR erythropoietin receptor; p-Akt, phosphorylated Akt.
Agnello, D., Bigini, P., Villa, P., Mennini, T., Cerami, A.,
Brines, M.L. and Ghezzi, P. (2002) Erythropoietin exerts
anti-inflammatory effect on the CNS in a model of experi-
mental autoimmune encephalomyelitis. Brain Res., 952:
128–134.
Arieli, Y., Eynan, M., Gancz, H., Arieli, R. and Kashi, Y.
(2003) Heat acclimation prolongs the time to central nervous
system oxygen toxicity in the rat: possible involvement of
HSP72. Brain Res., 962: 15–20.
Beit-Yannai, E., Kohen, R., Horowitz, M., Trembovler, V. and
Shohami, E. (1997) Changes in biological reducing activity in
rat brain following closed head injury: a cyclic voltammetry
study in normal and acclimated rats. J. Cereb. Blood Flow
Metab., 17: 273–279.

Beit-Yannai, E., Trembovler, V., Horowitz, M., Lazarovici, P.,
Kohen, R. and Shohami, E. (1998) Neuroprotection against
oxidative stress by serum from heat acclimated rats. Neuro-
sci. Lett., 254: 89–92.
Beit-Yannai, E., Zhang, R., Trembovler, V., Samuni, A. and
Shohami, E. (1996) Cerbroprotective effect of stable nitrox-
ide radicals in closed head injury in the rat. Brain Res., 717:
22–28.
Bernadin, M., Bellail, A., Marti, H.H., Yvon, A., Vivien, D.,
Duchatelle, I., Mackenzie, E.T. and Petit, E. (2000) Neurons
and astrocytes express Epo mRNA: oxygen-sensing mecha-
nisms that involve the redox-state of the brain. Glia, 30:
271–278.
Bernaudin, M., Marti, H.H., Roussel, S., Divoux, D., Nouve-
lot, A., MacKenzie, E.T. and Petit, E. (1999) A potential role
for erythropoietin in focal permanent cerebral ischemia in
mice. J. Cereb. Blood Flow Metab., 19: 643–651.
Blanco, M., Lizasoain, I., Sobrino, T., Vivancos, J. and
Castillo, J. (2006) Ischemic preconditioning: a novel target
for neuroprotective therapy. Cerebrovasc. Dis., 21(Suppl 2):
38–47.
Boeck, C.R., Ganzella, M., Lottermann, A. and Vendite, D.
(2004) NMDA preconditioning protects against seizures and
hippocampal neurotoxicity induced by quinolinic acid in
mice. Epilepsia, 45: 745–750.
Bromberg, Z. and Horowitz, M. (2004) Mus musculus: a model
for studying the dynamics of heat acclimation and cross tol-
erance between heat acclimation and oxygen deprivation.
Proceedings of the 1st Integrated meeting on Physiology and
Pharmacology of Thermoregulation (PPTR 2004), October
2004, Rhodes, Greece, p. 198.
Buemi, M., Cavallaro, E., Floccari, F., Sturiale, A., Aloisi, C.,
Trimarchi, M., Grasso, G., Corica, F. and Frisina, N. (2002)
Erythropoietin and the brain: from neurodevelopment to
neuroprotection. Clin. Sci., 103: 275–282.
Cannon, J.R., Keep, R.F., Hua, Y., Richardson, R.J.,
Schallert, T. and Xi, G. (2005) Thrombin preconditioning
provides protection in a 6-hydroxydopamine Parkinson’s
disease model. Neurosci. Lett., 373: 189–194.
Catania, M.A., Marciano, M.C., Parisi, A., Sturiale, A., Buemi,
M., Grasso, G., Squadrito, F., Caputi, A.P. and Calapai, G.
(2002) Erythropoietin prevents cognition impairment in-
duced by transient brain ischemia in gerbils. Eur. J. Pharma-
col., 437: 147–150.
Chan, P.H. (2001) Reactive oxygen radicals in signaling and
damage in the ischemic brain. J. Cereb. Blood Flow Metab.,
21: 2–14.
Chandel, N.S., McClintock, D.S., Feliciano, C.E., Wood, T.M.,
Melendez, J.A., Rodriguez, A.M. and Schumacker, P.T.
(2000) Reactive oxygen species generated at mitochondrial
complex III stabilize hypoxia inducible factor-1a during
hypoxia: a mechanism of oxygen sensing. J. Biol. Chem., 275:
25130–25138.
Chen, Y., Constantini, S., Trembovler, V., Weinstock, M. and
Shohami, E. (1996) An experimental model of closed head
injury in mice: pathophysiology, histopathology, and cogni-
tive deficits. J. Neurotrauma, 13: 557–568.
361
Chong, Z.Z., Kang, J.Q. and Maiese, K. (2002) Hematopoietic
factor erythropoietin fosters neuroprotection through novel
signal transduction cascades. J. Cereb. Blood Flow Metab.,
22: 503–514.
Cleveland Jr., J.C., Wollmering, M.M., Meidrum, D.R., Row-
land, R.T., Rehring, T.F., Sheridan, B.C., Harken, A.H. and
Banerjee, A. (1996) Ischemic preconditioning in human and
rat ventricle. Am. J. Physiol., 271: H178–H179.
Dahl, N.A. and Balfour, W.M. (1964) Prolonged anoxic sur-
vival due to anoxia pre-exposure: brain ATP, lactate, and
pyruvate. Am. J. Physiol., 207: 452–456.
Digicaylioglu, M., Bichet, S. and Marti, H.H. (2001) Erythro-
poietin-mediated neuroprotection involves cross-talk
between Jak-2 and NF-kappaB signaling cascades. Nature,
412: 641–647.
Dirnagl, U., Simon, R.P. and Hallenbeck, J.M. (2003) Ischemic
tolerance and endogenous neuroprotection. Trends Neuro-
sci., 26: 248–254.
Ehrenreich, H., Hasselblatt, M., Dembowski, C., Cepek, L.,
Lewczuk, P., Stiefel, M., Rustenbeck, H.H., Breiter, N.,
Jacob, S., Knerlich, F., Bohn, M., Poser, W., Ruther, E.,
Kochen, M., Gefeller, O., Gleiter, C., Wessel, T.C., De Ryck,
M., Itri, L., Prange, H., Cerami, A., Brines, M. and Siren,
A.L. (2002) Erythropoietin therapy for acute stroke is both
safe and beneficial. Mol. Med., 8: 495–505.
Ennaceur, A. and Delacour, J. (1988) A new one-trial test for
neurobiological studies of memory in rats. 1: behavioral data.
Behav. Brain Res., 31: 47–59.
Genc, S., Koroglu, F. and Genc, K. (2004) Erythropoietin and
the nervous system. Brain Res., 1000: 19–31.
Glantz, L., Avramovich, A., Trembovler, V., Gurvitz, V., Kohen,
R., Eidelman, L.A. and Shohami, E. (2005) Ischemic precon-
ditioning increases antioxidants in the brain and peripheral
organs after cerebral ischemia. Exp. Neurol., 192: 117–124.
Grasso, G., Buemi, M., Alafaci, C., Sfacteria, A., Passalacqua,
M., Sturiale, A., Calapai, G., De Vico, G., Piedimonte, G.,
Salpietro, F.M. and Tomasello, F. (2002) Beneficial effects of
systemic administration of recombinant human erythropoie-
tin in rabbits subjected to subarachnoid hemorrhage. Proc.
Natl. Acad. Sci. U.S.A., 99: 5627–5631.
Haddad, J.J. (2002) Recombinant human interleukin (IL)-1b-
mediated regulation of hypoxia inducible factor 1a (HIF-1a)
stabilization, nuclear translocation and activation requires an
antioxidant/reactive oxygen species (ROS)-sensitive mecha-
nism. Eur. Cytokine Netw., 13: 250–260.
Hellwig-Burgel, T., Rutkowski, K., Metzen, E., Fandrey, J. and
Jelkmann, W. (1999) Interleukin-1beta and tumor necrosis
factor-alpha stimulate DNA binding of hypoxia-inducible
factor-1. Blood, 94: 1561–1567.
Horowitz, M. (1994) Heat stress and heat acclimation: the cel-
lular response-modifier of autonomic control. Integrative and
cellular aspects of autonomic function. In: Pleschka K. and
Gerstberger R. (Eds.), Temperature and Osmoregulation.
Libbey Eurotext, Paris, pp. 87–95.
Horowitz, M. (1998) Do cellular heat acclimation responses
modulate central thermoregulatory activity? News Physiol.
Sci., 13: 218–225.
362
Horowitz, M. (2002) From molecular and cellular to integrative
heat defense during exposure to chronic heat. Comp. Bio-
chem. Physiol., 131: 475–483.
Horowitz, M. (2003) Matching the heart to heat-induced cir-
culatory load: heat-acclimatory responses. NIPS: News
Physiol. Sci., 18: 215–221.
Horowitz, M. (2007) Heat acclimation and cross-tolerance
against novel stressors: genomic-physiological linkage. In:
Sharma H.S. (Ed.), Progress in Brain Research: The Neuro-
biology of Hyperthermia, Elsevier, Amsterdam, The Nether-
lands (in press).
Horowitz, M., Eli-Berchoer, L., Wapinski, I., Friedman, N. and
Kodesh, E. (2004) Stress-related genomic responses during
the course of heat acclimation and its association with
ischemic-reperfusion cross-tolerance. J. Appl. Physiol., 97(4):
1496–1507.
Horowitz, M. and Meiri, U. (1985) Altered responsiveness to
parasympathetic activation of submaxillary salivary gland in
the heat acclimated rat. Comp. Biochem. Physiol., 80A:
57–60.
Hua, Y., Keep, R.F., Hoff, J.T. and Xi, G. (2003) Thrombin
preconditioning attenuates brain edema induced by erythro-
cytes and iron. J. Cereb. Blood Flow Metab., 23: 1448–1454.
Ikonomidis, J.S., Tumiati, L.C., Weisel, R.D., Mickel, D.A.G.
and Li, R.K. (1994) Preconditioning human ventricular cardio-
myocytes with brief periods of simulated ischemia. Cardiovasc.
Res., 28: 623–634.
Janoff, A. (1964) Alterations in lysosomes (intracellular en-
zymes) during shock; effects of preconditioning (tolerance)
and protective drugs. Int. Anesthesiol. Clin., 2: 251–259.
Jones, N.M. and Bergeron, M. (2001) Hypoxic preconditioning
induces changes in HIF-1 target genes in neonatal rat brain.
J. Cereb. Blood Flow Metab., 21: 1105–1114.
Kilic, E., Kilic, U., Wang, Y., Bassetti, C.L., Marti, H.H. and
Hermann, D.M. (2006) The phosphatidylinositol-3-kinase/
Akt pathway mediates VEGF’s neuroprotective activity and
induces blood brain barrier permeability after focal cerebral
ischemia. FASEB J., 20(8): 1185–1187.
Kissinger, P.T., Jonathan, B.H. and Adams, R.W. (1973) Voltam-
metry in brain tissue: a new neurophysiological measurement.
Brain Res., 55: 209–213.
Kloog, Y., Horowitz, M., Meiri, U., Galron, R. and Avron, A.
(1985) Regulation of submaxillary gland muscarinic recep-
tors during heat acclimation. Biochim. Biophys. Acta, 845:
428–435.
Kohen, R., Beit Yannai, E., Berry, E.M. and Tirosh, O. (1999)
Overall low molecular antioxidant activity of biological fluids
and tissues by cyclic voltammetry. Methods Enzymol., 300:
285–296.
Kohen, R. and Nyska, A. (2002) Oxidation of biological sys-
tems: oxidative stress phenomena, antioxidants, redox reac-
tions, and methods for their quantification. Toxicol. Pathol.,
30: 620–650.
Kontos, H.A. (2001) Oxygen radicals in cerebral ischemia.
Stroke, 11: 2712–2716.
Knoblach, S.M. and Faden, A.I. (1998) Interleukin-10 im-
proves outcome and alters proinflammatory cytokine
expression after experimental traumatic brain injury. Exp.
Neurol., 153: 143–151.
Levi, E., Vivi, A., Hasin, Y., Tassini, M., Navon, G. and
Horowitz, M. (1993) Heat acclimation improves cardiac
mechanics and metabolic performance during ischemia and
reperfusion. J. Appl. Physiol., 75: 833–839.
Liu, J., Ginis, I., Spatz, M. and Hallenbeck, J.M. (2000) Hypo-
xic preconditioning protects cultured neurons against hypo-
xic stress via TNF-alpha and ceramide. Am. J. Physiol. Cell
Physiol., 278: C144–C153.
Luh, S. and Yang, P. (2006) Organ preconditioning: the past,
current status and related lung studies. J. Zhejiang Univ. Sci.
B, 7: 331–341.
Maloyan, A., Eli-Berchoer, L., Semenza, G.L., Gerstenblith,
G., Stern, M.D. and Horowitz, M. (2005) HIF-1alpha-
targeted pathways are activated by heat acclimation and
contribute to acclimation-ischemic cross-tolerance in the
heart. Physiol. Genomics, 23(1): 79–88.
Maloyan, A., Palmon, A. and Horowitz, M. (1999) Heat accli-
mation increases the basal HSP72 level and alters its pro-
duction dynamics during heat stress. Am. J. Physiol., 276:
R1506–R1515.
Maloyan, A., Semenza, G.L., Gerstenblith, G., Stern, M. and
Horowitz, M. (2001) Heat-acclimation-ischemia cross tole-
rance: does HIF-1 play a role? J. Mol. Cell Cardiol., 33: A72.
Morishita, E., Masuda, S., Nagao, M., Yasuda, Y. and Sasaki,
R. (1997) Erythropoietin receptor is expressed in rat hippo-
campal and cerebral cortical neurons, and erythropoietin
prevents in vitro glutamate-induced neuronal death. Neuro-
science, 76: 105–116.
Murry, C.E., Jennings, R.B. and Reimer, K.A. (1986) Precon-
ditioning with ischemia: a delay of lethal cell injury in is-
chemic myocardium. Circulation, 74: 1124–1136.
Nie, H., Xiong, L., Lao, N., Chen, S., Xu, N. and Zhu, Z. (2006)
Hyperbaric oxygen preconditioning induces tolerance against
spinal cord ischemia by upregulation of antioxidant enzymes
in rabbits. J. Cereb. Blood Flow Metab., 26: 666–674.
Reshef, A., Sperling, O. and Zoref-Shani, E. (1996) Precondi-
tioning of primary rat neuronal cultures against ischemic in-
jury: characterization of the ‘‘time window of protection.’’
Brain Res., 741: 252–257.
Schaller, B. (2005) Ischemic preconditioning as induction of
ischemic tolerance after transient ischemic attacks in human
brain: its clinical relevance. Neurosci. Lett., 377: 206–211.
Semenza, G.L. (2004) Hydroxylation of HIF-1: oxygen sensing
in the molecular level. Physiology, 19: 176–182.
Shapira, Y., Shohami, E., Sidi, A., Soffer, D., Freeman, S. and
Cotev, S. (1988) Experimental closed head injury in rats:
mechanical, pathophysiologic, and neurologic properties.
Crit. Care Med., 16: 258–265.
Sharp, F.R., Ran, R., Lu, A., Tang, Y., Strauss, K.I., Glass, T.,
Ardizzone, T. and Bernaudin, M. (2004) Hypoxic precondi-
tioning protects against ischemic brain injury. NeuroRx, 1:
26–35.
Shein, N.A., Doron, H., Alexandrovich, A.G., Horowitz, M.
and Shohami, E. (2006) Heat acclimation increases pre-injury
anti-inflammatory capacity and leads to attenuation of

pro-inflammatory mRNA expression after closed head injury
in mice, 5th Forum of European Neuroscience, Vienna,
Austria.
Shein, N.A., Horowitz, M., Alexandrovich, A.G., Tsenter, J.
and Shohami, E. (2005a) Heat acclimation increases hypoxia-
inducible factor 1alpha and erythropoietin receptor expres-
sion: implication for neuroprotection after closed head injury
in mice. J. Cereb. Blood Flow Metab., 25: 1456–1465.
Shein, N.A., Horowitz, M., Alexandrovich, A.G., Tsenter, J.
and Shohami, E. (2005b) Akt-1 phosphorylation is increased
and tumor necrosis factor is decreased after closed head in-
jury in heat acclimated mice. J. Neurotrauma, 22(10): 1227.
Shiki, K. and Hearse, D.J. (1987) Preconditioning of ischemic
myocardium: reperfusion induced arrhythmia. Am. J. Phy-
siol., 253: H1470–H1476.
Shohami, E., Gallily, R., Mechoulam, R., Bass, R. and
Ben-Hur, T. (1997) Cytokine production in the brain follow-
ing closed head injury: dexanabinol (HU-211) is a novel
TNF-alpha inhibitor and an effective neuroprotectant. J.
Neuroimmunol., 72: 169–177.
Shohami, E., Novikov, M., Bass, R., Yamin, A. and Gallily, R.
(1994a) Closed head injury triggers early production of TNF
alpha and IL-6 by brain tissue. J. Cereb. Blood Flow Metab.,
14(4): 615–619.
Shohami, E., Novikov, M. and Horowitz, M. (1994b) Long
term exposure to heat reduces edema formation after closed
head injury in the rat. Acta Neurochir. Suppl. (Wien.), 60:
443–445.
Shohami, E., Shapira, Y., Sidi, A. and Cotev, S. (1987) Head
injury induces increased prostaglandin synthesis in rat brain.
J. Cereb. Blood Flow Metab., 7: 58–63.
Shohami, E., Shapira, Y., Yadid, G., Reisfeld, N. and Yedgar,
S. (1989) Brain phospholipase A2 is activated after experi-
mental closed head injury in the rat. J. Neurochem., 53:
1541–1546.
363
Siren, A.L., Knerlich, F., Poser, W., Gleiter, C.H., Bruck, W.
and Ehrenreich, H. (2001) Erythropoietin and erythropoietin
receptor in human ischemic/hypoxic brain. Acta Neuoro-
pathol. (Berl.), 101: 271–276.
Stahel, P.F., Shohami, E., Younis, F.M., Kariya, K., Otto, V.I.,
Lenzlinger, P.M., Grosjean, M.B., Eugster, H.P., Trentz, O.,
Kossmann, T. and Morganti-Kossmann, M.C. (2000) Ex-
perimental closed head injury: analysis of neurological out-
come, blood-brain barrier dysfunction, intracranial
neutrophil infiltration, and neuronal cell death in mice
deficient in genes for pro-inflammatory cytokines. J. Cereb.
Blood Flow Metab., 20(2): 369–380.
Toumpoulis, I.K. and Anagnostopoulos, C.E. (2003) Early
ischemic preconditioning for spinal cord protection. Ann.
Thorac. Surg., 76: 1340–1341.
Treinin, M., Shliar, J., Jiang, H., Powell-Coffman, J.A.,
Bromberg, Z. and Horowitz, M. (2003) HIF-1 is required
for heat acclimation in the nematode C. elegans. Physiol.
Genomics, 14: 17–24.
Vink, R. and Van Den Heuvel, C. (2004) Recent advances in
the development of multifactorial therapies for the treatment
of traumatic brain injury. Expert Opin. Investig. Drugs,
13(10): 1263–1274.
Vohra, H.A. and Galinanes, M. (2006) Myocardial precondi-
tioning against ischemia-reperfusion apoptosis and necrosis
in man. J. Surg. Res., 134(1): 138–144.
Yatsiv, I., Grigoriadis, N., Simeonidou, C., Stahel, P.F.,
Schmidt, O.I., Alexandrovitch, A.G., Tsenter, J. and
Shohami, E. (2005) Erythropoietin is neuroprotective, im-
proves functional recovery, and reduces neuronal apoptosis
and inflammation in a rodent model of experimental closed
head injury. FASEB J., 19: 1701–1703.
Yellon, D.M. and Dana, A. (2000) The preconditioning
phenomenon: a tool for the scientist or a clinical reality.
Circ. Res., 87: 543–550.
This page intentionally left blank
 SECTION VI

The Future of Neurotrauma: Developing Novel

Treatment Strategies
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 26

In vivo tracking of stem cells in brain and spinal cord

injury

Eva Sykova and Pavla Jendelova

Institute of Experimental Medicine ASCR, EU Centre of Excellence, Prague, Czech Republic; Center for Cell Therapy
and Tissue Repair, Charles University, Second Medical Faculty, Prague, Czech Republic; Department of Neuroscience,
Charles University, Second Medical Faculty, Prague, Czech Republic

Abstract: Cellular magnetic resonance (MR) imaging is a rapidly growing field that aims to visualize and
track cells in living organisms. Superparamagnetic iron oxide (SPIO) nanoparticles offer a sufficient signal
for T2 weighted MR images. We followed the fate of embryonic stem cells (ESCs) and bone marrow
mesenchymal stem cells (MSCs) labeled with iron oxide nanoparticles (Endorems) and human CD34+
cells labeled with magnetic MicroBeads (Miltenyi) in rats with a cortical or spinal cord lesion, models of
stroke and spinal cord injury (SCI), respectively. Cells were either grafted intracerebrally, contralaterally to
a cortical photochemical lesion, or injected intravenously. During the first post-transplantation week,
grafted MSCs or ESCs migrated to the lesion site in the cortex as well as in the spinal cord and were visible
in the lesion on MR images as a hypointensive signal, persisting for more than 30 days. In rats with an SCI,
we found an increase in functional recovery after the implantation of MSCs or a freshly prepared mono-
nuclear fraction of bone marrow cells (BMCs) or after an injection of granulocyte colony stimulating factor
(G-CSF). Morphometric measurements in the center of the lesions showed an increase in white matter
volume in cell-treated animals. Prussian blue staining confirmed a large number of iron-positive cells, and
the lesions were considerably smaller than in control animals. Additionally, we implanted hydrogels based
on poly-hydroxypropylmethacrylamide (HPMA) seeded with nanoparticle-labeled MSCs into hemisected
rat spinal cords. Hydrogels seeded with MSCs were visible on MR images as hypointense areas, and
subsequent Prussian blue histological staining confirmed positively stained cells within the hydrogels. To
obtain better results with cell labeling, new polycation-bound iron oxide superparamagnetic nanoparticles
(PC-SPIO) were developed. In comparison with Endorem, PC-SPIO demonstrated a more efficient intra-
cellular uptake into MSCs, with no decrease in cell viability. Our studies demonstrate that magnetic
resonance imaging (MRI) of grafted adult as well as ESCs labeled with iron oxide nanoparticles is a useful
method for evaluating cellular migration toward a lesion site.

Keywords: cell transplantation; iron oxide nanoparticles; magnetic resonance; MicroBeads; photochemical
lesion; scaffold; spinal cord injury

Introduction disorders of the central nervous system (CNS;

Stem cells and progenitor cells are being explored
in regenerative medicine for cell therapy in
Corresponding author. Tel.: +420-241062230;
Fax:+420-241062782; E-mail: sykova@biomed.cas.cz
Park et al., 2002; McKay, 2004; Newman et al.,
2004; Roitberg, 2004; Emsley et al., 2005; Fairless
and Barnett, 2005; Pluchino et al., 2005; Zhao
et al., 2005; Uccelli et al., 2006). Crucial to the
future success of cell transplantation in the clinical
setting is the ability of transplanted cells to migrate

DOI: 10.1016/S0079-6123(06)61026-1 367

368
from the site of transplantation to the lesioned
area and to survive, differentiate, and/or produce
growth factors and cytokines for the prolonged
periods of time necessary for the patient to benefit
from their regenerative properties. Visualizing
transplanted cells in vivo is essential for preclini-
cal studies in rodents, and the magnetic tracking of
cells appears to be a valuable tool for such studies.
Magnetic resonance (MR) imaging may also serve
to study cell migration to lesions, its timecourse,
and how long the cells persist in the target region.
Such information could help to elucidate the time
window during which the transplantation of ther-
apeutic cells may be clinically effective, the number
of cells needed, and the optimal method of their
administration. For magnetic resonance imaging
(MRI) detection, cells can be labeled with MR
contrast agents in order to make them visible in
vivo (Bulte et al., 2002; Jendelova et al., 2003,
2004; Sykova and Jendelova, 2005, 2006; Sykova
et al., 2006).
Superparamagnetic cell labels
Superparamagnetic iron oxide (SPIO) nanoparti-
cles were introduced as contrast agents shortly af-
ter the use of gadolinium-chelates (Mendonca Dias
and Lauterbur, 1986; Renshaw et al., 1986;
Pachernik et al., 2005). They are currently pre-
ferred as contrast agents primarily due to the fol-
lowing properties: (a) they provide the greatest
signal contrast change, in particular on T2 and
T2* weighted images; (b) they are composed of
biodegradable iron; (c) their surface coating makes
them soluble and stable and allows for the chemi-
cal linkage of functional groups and ligands; and
(d) they can be easily detected by both light and
electron microscopy (EM; Jendelova et al., 2003;
Bulte and Kraitchman, 2004).
Iron oxide nanoparticle stabilization in order to
prevent aggregation is most commonly accom-
plished by a surface coating of dextran. Dextran-
coated SPIO nanoparticles include the products
Feridexs and Endorems, the ultrasmall super-
paramagnetic iron oxide (USPIO) nanoparticles
Combidexs and Sinerems, monocrystalline iron
oxide nanoparticles, and cross-linked iron oxide
nanoparticles. Another approach for stabilizing
iron oxide nanoparticles is the use of carboxylated
polyamidoamine dendrimers (Bulte et al., 2001).
These highly soluble nanocomposites of iron ox-
ides and dendrimers are stable under a wide range
of temperatures and pH and have an overall size of
20–30 nm.
Iron oxide nanoparticles also require an appro-
priate outer surface layer that induces the internal-
ization of the particles into the cytoplasm,
so that they can be nonspecifically taken up by a
variety of cultured mammalian cells, regardless of
cell origin or animal species. In magnetodendrim-
ers, the highly charged carboxylated dendrimers
bind to multiple sites on the cell membrane, in-
ducing membrane bending followed by endocytosis
(Zhang and Smith, 2000). In the case of anionic
magnetic nanoparticles, the negative surface charge
induces an uptake three orders of magnitude
greater than that of conventional dextran-coated
SPIO nanoparticles (Billotey et al., 2003). A
disadvantage of these particles, as well as of
the magnetodendrimers, is that they have not yet
been commercially developed. Another method is
based on the mixing of a commercially available
(U)SPIO formulation, Feridexs (Frank et al.,
2003) or Sinerems, and a commercially available
transfection agent, for example poly-L-lysine,
Lipofectamin, or FugeneTM (Frank et al., 2003).
We have developed new polycation-bound iron
oxide superparamagnetic nanoparticles (PC-SPIO)
that can be used for intracellular labeling (Horak
et al., 2006).
Cell labeling with dextran-coated SPIO
nanoparticles
MR tracking of stem and progenitor cells in the
lesioned CNS has been performed by several
groups that utilized different methods of adminis-
tration. The first studies of imaging cell transplants
labeled by superparamagnetic contrast agents in
rat brains were reported in 1992 (Norman et al.,
1992; Hawrylak et al., 1993). Rats received grafts
of fetal rat tissue prepared as cell suspensions and

labeled by incubation with reconstituted Sendai
viral envelopes containing iron oxide particles.
When magnetically labeled neurospheres were
transplanted into the ventricles of experimental
allergic encephalomyelitis (EAE) rats at the peak
of their disease, the migration of glial precursors
into white matter structures was observed on MR
images (Bulte et al., 2003). Ferromagnetic-labeled
neural progenitor cells were also transplanted into
the cisterna magna of rats that underwent experi-
mental MCAO (middle cerebral artery occlusion).
MR images showed the migration of cells through-
out the ventricular system toward the ischemic
brain parenchyma (Zhang et al., 2003). Olfactory
ensheathing cells (OECs) labeled with magneto-
dendrimers were implanted into the transected rat
spinal cord, and their distribution was followed in
vivo using MR imaging (Lee et al., 2004). In our
experiments (Jendelova et al., 2003, 2004), we have
shown that a suitable contrast agent for me-
senchymal stem cells (MSCs), embryonic stem
cells (ESCs), and OECs is a commercially avail-
able contrast agent based on dextran-coated SPIO
nanoparticles (Fig. 1A), Endorems (Guerbet,
France), which has also been approved as a blood
pool agent for human use. The contrast agent
Endorem can be easily incorporated by end-
ocytosis, and all the cells survive and further di-
vide in vitro. Therefore, Endorem uptake does not
need to be facilitated by a transfection agent,
which can damage large numbers of cells (Arbab
et al., 2004).
Bone marrow-derived MSCs are adult stem cells
that reside within the bone marrow compartment.
Recent data have presented evidence for their
multilineage differentiation potential (Pittenger et
al., 1999). More recently, MSCs have been re-
ported to have the ability to stimulate or partic-
ipate in the regeneration of diverse tissues and
organs, including the liver, myocardium, endothe-
lium, and CNS (Takahashi et al., 1999; Orlic et al.,
2001a, b; Jiang et al., 2002; Toma et al., 2002). In
addition, they can be genetically modified, and,
due to their migratory properties, they can serve as
a carrier for drug delivery in tumor therapies
(Anderson et al., 2005). In cell therapy, bone mar-
row cells (BMCs) have some advantages over
369
other sources of cells: (a) they are relatively easy to
isolate; (b) they can be used in autologous trans-
plantation protocols; and (c) they have already
been approved for the treatment of hematopoietic
diseases.
In our experiments, cell cultures of human MSCs
or rat MSCs were incubated 2–3 days in media
containing Endorem. Nanoparticles were detected
by staining for iron (Fig. 1B). Transmission elec-
tron microscopy confirmed the presence of iron
oxide particles inside the cells, observed as large
membrane-bound clusters scattered within the cell
cytoplasm (Fig. 1C). On the day that nanoparticles
were withdrawn, the efficiency of MSC labeling
(i.e., how many cells of the total number of analy-
zed cells were labeled) was 50–70%.
Tracking mesenchymal stem cells in an
experimental model of stroke
Endorem-labeled MSCs were co-labeled with
bromdeoxyuridine (BrdU) and grafted into rats
with a cortical photochemical lesion (Jendelova
et al., 2003). Rats were examined weekly for a
period of 3–7 weeks post-transplantation using a
4.7 T Bruker spectrometer. Single sagittal, coronal,
and transversal images were obtained by a fast
gradient echo sequence for localizing subsequent
T2 weighted transversal images measured by a
standard turbospin echo sequence. The lesion was
clearly visible on MR images 2 h after lesioning
as a hyperintense signal and remained visible
during the entire measurement period. No recog-
nizable hypointense signal in the lesion was
detected during the first 2 days after implantation.
A decrease in the MR signal was found only at the
injection site in animals with cells injected contra-
laterally to the lesion. One week after grafting,
we observed a hypointense signal in the lesion,
which intensified during the second and third
weeks (Fig. 1D). Histology confirmed that a large
number of Prussian blue-positive cells had entered
the lesion. No hypointense signal was found in
other brain regions. The hypointense signal oc-
curred only in damaged areas populated with
MSCs, and its intensity corresponded to Prussian
370

Fig. 1. (A) Scheme of an iron oxide nanoparticle. The contrast agent Endorem consists of a superparamagnetic Fe3O4 core coated by a
dextran shell. (B) Rat mesenchymal stem cells (MSCs) in culture labeled with superparamagnetic nanoparticles (blue dots). The cell
nuclei are counterstained with hematoxylin. (C) Transmission electron microphotograph of a cluster of iron nanoparticles surrounded
by a cell membrane. (D) An implant of Endorem-BrdU co-labeled MSCs (in the hemisphere contralateral to the lesion, arrow) and the
lesion (arrowhead) itself are both hypointense in an MR image taken 2 weeks after implantation. (E), (F) A hypointense signal (black
arrowhead) was observed in the lesion 6 days after the intravenous (i.v.) injection of Endorem-BrdU co-labeled MSCs, becoming more
hypointense and persisting for 47 days (F). (G, H) Massive invasion of rat MSCs (Prussian blue staining counterstained with
hematoxylin) into a photochemical lesion 7 weeks after i.v. injection into a rat with a photochemical lesion. (I) Serial section stained for
BrdU, 7 weeks after i.v. injection. (Adapted with permission from Jendelova et al. (2003).)

blue or BrdU staining. Only a few (less than 3%)
of the MSCs that migrated into the lesion ex-
pressed the neuronal marker NeuN when tested 28
days post-implantation. No GFAP-positive cells
were found in the lesion.
After the intravenous injection of MSCs, we
found a similar hypointense MR signal in the le-
sion site. The signal was observed 6 days after cell
infusion and persisted for 7 weeks (Figs. 1E, F).
Prussian blue and anti-BrdU staining confirmed
the presence of iron oxide-BrdU co-labeled cells in
the lesion, which densely populated the borders of
the lesion (Figs. 1G–I).
Only a few cells weakly stained for Prussian blue
were found in photochemical lesions without any
implanted cells. The staining represents iron,
which most likely originated in hemorrhages and
iron degradation products released from iron-
containingproteins (such as hemoglobin, ferritin,
and hemosiderin) and phagocytized by microglia/
macrophages. We did not observe any BrdU-
positive cells in the brains of nongrafted animals.

Tracking embryonic stem cells in an experimental
model of stroke
ESCs are pluripotent cell lines with a capacity for
self-renewal and broad differentiation plasticity.
They are derived from embryos and can be prop-
agated as a homogeneous, uncommitted cell pop-
ulation for an almost unlimited period of time
without losing their pluripotency and their stable
karyotype. Mouse ES cell-derived glial precursors,
transplanted into rats with myelin disease, inter-
acted with the host neurons to produce myelin in
the brain and spinal cord (Brustle et al., 1999). Re-
tinoic acid-treated embryoid bodies from
mouse ESCs, when transplanted into rat spinal
cord 9 days after traumatic injury, differentiated
into astrocytes, oligodendrocytes, and neurons and
promoted recovery (McDonald et al., 1999).
Bjorklund et al. (2002) reported that undifferenti-
ated mouse ESCs can become dopamine-producing
neurons in the brain in a rat model of Parkinson’s
disease and can lead to partial functional recovery.
We used mouse ESCs transfected with the pEG-
FP-C1 vector (Pachernik et al., 2005) and labeled
with SPIO nanoparticles (Jendelova et al., 2004).
Since undifferentiated ESCs may form embryonic
tumors when grafted into a host animal, we in-
duced neural differentiation by culturing enhanced
green fluorescent protein (eGFP) ESCs in serum
containing DMEM (Dulbecco’s modified Eagle
medium)/F12 without leukemia inhibitory factor
(LIF) for 2 days and then transferred the cells into
serum-free media supplemented with insulin, trans-
ferrin, selenium, and fibronectin for further culture
(Pachernik et al., 2005). Feeder-free eGFP ESCs
were labeled with Endorem (112.4 mg/ml) during
three passages (Fig. 2A). Cell counting of Prussian
blue stained cells in suspension revealed that the
labeling efficiency was 80%. Cells were trans-
planted intracerebrally or intravenously on the 8th
day of differentiation into adult Wistar rats with a
cortical photochemical lesion 1 week after lesioning
(Jendelova et al., 2004), and were detected by
staining for iron and by GFP fluorescence. When
we implanted the ESCs 7 days post-lesion, we
found a massive migration of Endorem-labeled
GFP-positive cells into the lesion site regardless of
the route of administration (Fig. 2D). We observed
371
very similar MR images to those obtained after the
implantation of MSCs. In rats with a photochemi-
cal lesion and contralaterally injected cells, the cell
implants were visible as a hypointense area at the
injection site. Two weeks after grafting, a hypoin-
tense signal was also observed in the corpus callo-
sum and in the lesion (Fig. 2B). At the same time,
histology showed that a large number of Prussian
blue-positive cells had entered the lesion. Many
labeled cells were also detected in the corpus callo-
sum, suggesting a migration from the contralateral
hemisphere toward the lesion (Fig. 2C). When the
ESCs were injected intravenously into lesioned
rats, we found a hypointense MR signal only at the
site of the lesion. However, the extent of the dif-
ferentiation of eGFP ESCs labeled with nanopar-
ticles, and found in the lesion, was much greater
than with grafted MSCs. Of all the eGFP ESCs
containing nanoparticles found in the lesion, 70%
of them proved to be astrocytes, very few (less than
1%) were oligodendrocytes, and 5% of nanopar-
ticle-labeled eGFP ESCs had differentiated into
neurons (Jendelova et al., 2004). Electron micro-
scopy revealed mature astrocytes, neurons, and
oligodendrocytes in the lesion site containing iron
oxide nanoparticles, providing strong supporting
evidence that these cells differentiated in the lesion
from implanted embryonic cells (Figs. 2E, F).
Tracking mesenchymal stem cells in a rat model of
spinal cord injury
Evaluating the effect of different BMC populations
on morphological and functional recovery after spi-
nal cord injury (SCI) is an important aspect of our
preclinical research. In rats with a balloon-induced
spinal cord compression lesion, we studied the effect
of an intravenous injection of Endorem-labeled
nonhematopoietic MSCs (Urdzikova et al., 2006).
The results obtained were compared with those
following the implantation of a freshly isolated
mononuclear fraction of bone marrow containing
stromal cells, hematopoietic and nonhemato-
poietic stem and precursor cells, and lymphocytes
(BMCs). Furthermore, we studied the effect of
granulocyte colony stimulating factor (G-CSF) mo-
bilization of endogenous BMCs containing mainly
372

Fig. 2. (A) Trypsinized nanoparticle-labeled suspensions of ESCs after the third passage. (B) The cell implant (in the hemisphere
contralateral to the lesion) and the lesion are hypointense in MR images 2 weeks after implantation. A hypointensive signal is also
found in the corpus callosum. (C) Dense Prussian blue staining of the injection site in the contralateral hemisphere, the corpus
callosum, and the photochemical lesion, 4 weeks after grafting. (D) Invasion of GFP-labeled cells showing GFP-positive ESCs in the
lesion 4 weeks after the i.v. injection of eGFP ESCs (serial section to the slice shown in Fig. 2C). (E) Astrocyte from the brain tissue of
a rat with nanoparticle-labeled eGFP ESCs implanted contralaterally to the lesion; the nanoparticles are visible in small dense clusters
in the cell cytoplasm (arrowhead). (F) Neuron with nanoparticles marginated to the nuclear membrane. (Adapted with permission
from Jendelova et al. (2004).)

hematopoietic stem cells, along with progenitor
cells and lymphocytes (Sasaki et al., 2001; Akiyama
et al., 2002).
MSCs labeled with Endorem were injected in-
travenously into the femoral vein 1 week after le-
sioning. MR images were taken ex vivo 4 weeks
after implantation using a standard whole body
resonator. Functional status was assessed weekly
for 5 weeks after spinal cord lesioning, using the
Basso-Beattie-Bresnehan (BBB) locomotor rating
score and the plantar test. Our data indicated that
lesioned animals with grafted MSCs had higher
 373

locomotor scores as indicated by their BBB scores
and showed better responses in sensitivity testing
using the plantar test than did control animals.
Particularly, the plantar test showed the recovery
of sensitivity in the hind limbs. On MR images we
observed the lesion as an inhomogeneity in the
tissue texture with a hyperintense signal only in the
area of the SCI (Fig. 3A). Images of longitudinal
spinal cord sections from animals grafted with
nanoparticle-labeled MSCs showed the lesion as a
dark hypointense area (Fig. 3B). Prussian blue
staining confirmed a large number of positive
cells present in the lesion site (Fig. 3C). On serial
sections, Prussian blue-positive cells corresponded to
cells labeled with PKH 26, as seen in (Figs. 3D, E).
To exclude the possibility that free nanopar-
ticles were taken up by macrophages, we per-
formed immunohistochemical and Prussian blue
staining to co-localize iron in activated microglia/
macrophages. Although we found strong positive
staining for macrophages in the lesions using
ED1 antibody, Prussian blue staining did not,

Fig. 3. (A) Longitudinal MR image of a spinal cord lesion 5 weeks after induction. The formation of the lesion cavity is visible as a
strong hyperintensive signal (arrow). (B) Longitudinal MR image 4 weeks after MSC grafting. The lesion with nanoparticle-labeled
cells is visible as a dark hypointensive area (arrow). (C) Prussian blue staining of a lesion populated with nanoparticle-labeled MSCs
(blue dots). (D) Magnified region from Fig. C. (E) Intravenously injected MSCs were also detectable in the spinal cord lesion using the
membrane florescent dye PKH 26. Serial section to D. (F) Immunostaining with ED-1 antibody revealed a number of macrophages in
the lesion (white arrows). The majority of Prussian blue-positive cells did not co-localize with ED-1 staining; however, a few macro-
phages were Prussian blue-positive (black arrowheads). (Adapted with permission from Urdzikova et al. (2006).)
374

for the most part, co-localize with the ED1 staining
(Fig. 3F).
Morphometric measurements of the spared
white and gray matter were performed in the cen-
ter of the lesions. The spared cross-sectional area
of the white matter, as well as the volume of spared
white matter, was significantly greater in MSC-
treated animals. The spared cross-sectional area of
the gray matter was also significantly larger in
MSC-treated animals. In further studies, lesioned
animals grafted with BMCs or injected with G-
SCF also had higher scores in BBB testing than
did control animals and also showed a faster
recovery of sensitivity in their hind limbs using the
plantar test. However, the functional improvement
was more pronounced in MSC-treated rats
(Figs. 4A, B). Morphological measurements of
the spared cross-sectional area of the white matter
showed a statistically significant increase in groups
treated with BMCs or G-SCF, compared to con-
trols, cranially to the lesion center; a statistically
significant increase in the volume of spared white

Fig. 4. (A) Behavioral open field BBB motor scores of MSC, BMC, and G-CSF treated rats were significantly higher than those of
saline-injected (control) animals, 14, 21, 28, and 35 days after SCI (po0.05). The scores of the MSC, BMC, and G-CSF treated animals
were not significantly different from one another at any time point. (B) Time course of the animals’ response to radiant heat measured
with the plantar test in treated and saline-injected rats. In all treated rats, the latency times decreased as their recovery progressed. The
most pronounced effect was seen in MSC treated rats. The latency time in the saline-injected (control) rats did not change during the 35
days survival period. Data are averaged between right and left hind limbs and expressed as mean7SEM. *po0.05 compared to control
group. (Adapted with permission from Urdzikova et al. (2006).)
 375

Fig. 5. (A) The total volume of the white matter in 11 mm long segments of the spinal lesion in MSCs, BMCs, and G-CSF-treated and
saline-injected animals. (B) The total volume of gray matter in 11 mm long segments of the spinal lesion in MSCs, BMCs, and G-CSF-
treated and saline-injected animals. No statistically significant differences were observed between treated and saline-injected animals.
All data are presented as means7SEM. *po0.05 compared to saline-injected (control) rats. (Adapted with permission from Urdzikova
et al. (2006).)

matter was observed in the BMC-treated group.
The spared white matter volume in the G-CSF-
treated rats was also increased, but the increase did
not reach statistical significance (Fig. 5).
Stem cell labeling with polycation-bound
nanoparticles
In all the above-described experiments, dextran-
coated SPIO nanoparticles (the contrast agent
Endorem) were used as an intracellular label.
However, the efficiency of the cell labeling was
maximally 70%. In addition, on the day of particle
withdrawal a decrease in cell viability (using the
WST1 colorimetric assay) was observed. We
therefore developed new polycation-bound SPIO
nanoparticles (Fig. 6A; Horak et al., 2006). We
compared the influence of both types of nanopar-
ticles on cell viability and labeling efficiency in rat
and human MSCs. The PC-SPIO nanoparticle sus-
pension was used at a much lower iron concentra-
tion per milliliter of culture media (15.4 mg/ml), and
the cells were incubated with PC-SPIO for 3 days.
The results were compared with Endorem labeling
(Tables 1 and 2). The measurements were per-
formed on the day of withdrawal of the nanopar-
ticles (day 3) and 4 days later (day 7).
Labeling cells with PC-SPIO nanoparticles was
more efficient than labeling with Endorem, i.e.,
more cells from the total number of analyzed cells
were labeled with PC-SPIO nanoparticles than
376

Fig. 6. (A) Schematic illustration of a polycation-bound superparamagnetic iron oxide nanoparticle. (B) Transmission electron
photomicrograph of MSCs labeled with PC-SPIO showing clusters of iron nanoparticles (arrows) in the cell cytoplasm, which are not
surrounded by a cell membrane. (C) Prussian blue staining of MSCs in culture labeled with PC-SPIO. (D, E) Axial and coronal (E) MR
images of a rat brain with 1000 cells labeled with PC-SPIO implanted to the left hemisphere and 1000 Endorem-labeled cells implanted
to the right hemisphere. MR images were taken 3 days after implantation.

Table 1. The percentage of cells labeled with nanoparticles af- In Endorem-labeled human MSCs, on day 3 the cell
ter 3 days incubation viability dropped to 50%; however, within 4 days it
Endorem-labeled PC-SPIO-labeled recovered to 87%. This transient decrease in cell
cells (%) cells (%) viability might be due to the relatively high con-
centration of iron in the culture media (112.4 g/ml),
Rat MSCs 59.3 92.2
Human MSCs 65.2 87.5
to which some cell cultures might be sensitive. The
average amount of iron present in rat MSCs
was determined by spectrophotometry after min-
with Endorem (Table 1; Fig. 6C). In Table 2 the eralization of iron-labeled cell suspensions. In PC-
cell viability of labeled cells is expressed as a per- SPIO-labeled cells, even though the concentration
centage of the number of viable unlabeled cells of iron in the culture media was 10 times lower
(taken as the control value and defined as 100%). (15.4 mg/ml culture media), the average amount of

Table 2. Cell viability (in %) after incubation with nanopar-
ticles (day 3) and 4 days after withdrawal (day 7)
Endorem-labeled PC-SPIO-labeled
cells (%) cells (%)
Day 3 Day 7 Day 3 Day 7
Rat MSCs 68.9 72.9 92.8 89.7
Human MSCs 49.0 87.0 97.2 98.9
iron was 38 pg/cell, while in Endorem-labeled cells
only 17 pg/cell. Transmission electron photomicro-
graphs confirmed that more PC-SPIO nanoparticles
were taken up by the rat MSCs than were Endo-
rem nanoparticles (Fig. 6B). Moreover, Endorem
was observed as membrane-bound clusters within
the cell cytoplasm, indicating an endocytotic proc-
ess of nanoparticle uptake. The higher cellular
uptake of PC-SPIO nanoparticles is possible due
to the interaction of the surface coating with the
negatively charged cell surface and subsequent
endosomolytic uptake. Nanoparticles are trans-
ported in endosomes and finally fused with lyso-
somes, a process during which the vesicle
membranes disappear.
To check the sensitivity of the MRI technique
and to mimic signal behavior in CNS tissue, sus-
pensions of unlabeled cells and cells labeled with
PC-SPIO were imaged in vitro. MR images of
1.7% gelatin phantoms containing iron-labeled
MSCs were obtained using a 4.7 T Bruker spectro-
meter equipped with a standard resonator coil.
Even the sample containing the lowest concentra-
tion (200 cells/ml, which corresponds on average to
2 cells per image voxel) provided visible contrast
compared to a control phantom containing the
same number of unlabeled cells. A similar set of
experiments was performed in earlier work
(Jendelova et al., 2003), in which MR images of
gelatin phantoms showed a hypointense signal at
concentrations above 625 labeled cells/ml.
For in vivo imaging, rats were examined 3 days
post-transplantation in an MR imager. Figure 6
shows that PC-SPIO labeled cells are also clearly
recognizable in vivo. The iron oxide-labeled cell
implants are visible as a hypointense area at the
injection site. Cells labeled by PC-SPIO provide
377
stronger contrast change in the signal than do
Endorem-labeled cells (Figs. 6D, E). Labeling with
PC-SPIO thus enables the detection of a lower
number of cells in the tissue.
Cell labeling with magnetic MicroBeads
The disadvantage of an intracellular label is that it
can affect cell metabolism and subsequently cell
viability. In addition, these labels are rather non-
specific; they can be loaded by virtually any cell
present in the medium. The possibility of labeling
only selected cell types would therefore be very
useful. In addition, cells labeled and separated by
means of immunomagnetic selection would not
require in vitro culturing, since the label is at-
tached during separation. This would also allow
the immediate clinical use of labeled cells. The first
experiments using contrast agents bound to an
antibody that can specifically bind to a single cell
type were performed by Bulte et al. (1992). They
described experiments with human lymphocytes
labeled with biotinylated anti-lymphocyte-directed
monoclonal antibodies (mabs), to which streptavi-
din and subsequently biotinylated dextran-mag-
netite particles were coupled. We tested a new
type of specific cell labeling using commercially
available cell isolation kits for the magnetic sep-
aration of CD34+ cells. CD34+ cells are known as
hematopoietic progenitor cells. The cells were sep-
arated by means of immunomagnetic selection
with anti-CD34 antibodies. For sorting, a SPIO
core coated with a polysaccharide that is linked to
an antibody is bound to the respective cell
(Fig. 7A). The size of the label is comparable to
commonly used superparamagnetic MR contrast
agents; thus it can provide sufficient contrast on
MR images.
Human CD34+ cells from peripheral blood
were selected by CliniMACS CD34 Selection
Technology (Miltenyi). On EM images, we deter-
mined that after cryopreservation, the nanoparti-
cles remained bound to the cell surface, and we
observed several iron labels attached to the cell
surface (Fig. 7B). Purified CD34+ cells were im-
planted intracerebrally into rats with a cortical
378

Fig. 7. (A) Scheme of MicroBeads. A superparamagnetic iron oxide core coated with a polysaccharide is linked to an antibody, which
in turn is bound to a cell via an antigen–antibody complex. (B) Transmission electron microphotograph of MicroBeads binding to the
cell surface. (C) Human cells (positive staining for human nuclei) found in the lesion 4 weeks after the grafting of CD34+ cells. (D)
Staining with anti-CD34 revealed CD34+ cells in the subventricular zone. (Adapted with permission from Jendelova et al. (2005).)

photochemical lesion, contralaterally to the lesion
(Jendelova et al., 2005; Sykova and Jendelova,
2006). The average iron content per cell, deter-
mined by spectrometry, was 0.275 pg. This value is
lower by two orders of magnitude than in the
case of cell labeling using Endorem or PC-SPIO,
which enter the cell (17 or 38 pg, respectively);
nevertheless, it still provides sufficient MR con-
trast (Jendelova et al., 2005).
The cells were detected as a hypointense spot
on T2 weighted images 24 h after grafting. The
hypointensity of the implant slightly decreased
during the first week and then remained without
significant changes for the entire measurement pe-
riod (4 weeks). During the second week, we ob-
served a weak hypointense signal in the lesion that
persisted for the next 2 weeks. Prussian blue and
anti-human nuclei staining (Fig. 7C) confirmed
the presence of magnetically labeled human cells
in the corpus callosum and in the lesion. CD34+
cells were detected not only in the corpus callosum
and in the lesion, but also in the subventricular

zone (Fig. 7D). Finally, human DNA was detected
in the lesioned brain tissue using polymerase
chain reaction (PCR), confirming the presence of
human cells.
Labeling of stem cells in polymer hydrogels as stem
cell carriers
In the case of large lesions, cells alone are not able
to repair the injury. It is necessary to bridge the
gap left by the lost cell population in order to
provide support for tissue restoration, reduce the
glial scar, and create a permissive environment for
cellular ingrowth and for the diffusion of neuro-
active molecules. To bridge such a lesion, we have
used biocompatible polymer hydrogels based on
poly-hydroxypropylmethacrylamide (HPMA) in
combination with stem cell grafting. Before trans-
plantation into a lesion, the hydrogels were seeded
in vitro with MSCs. In this case the hydrogels form
an inert environment, allowing for the free diffu-
sion of intrinsic growth factors, in which the cells
start to differentiate and migrate. The inert and
biocompatible environment of the hydrogels also
provides an adequate standard background for
MR imaging of the cells (Sykova and Jendelova,
2005; Sykova et al., 2006). We employed these
cell-polymer constructs in order to facilitate the
regeneration of injured spinal cord (Lesny et al.,
2006). The right half of a spinal cord segment was
removed by hemisection, and a block of HPMA
hydrogel seeded with Endorem-labeled MSCs was
inserted (Fig. 8A; Sykova and Jendelova, 2005).
Six weeks after implantation, the hydrogel had
formed a continuous bridge between the hem-
isected spinal segments, reestablishing the anatom-
ical continuity of the tissue. The hydrogel
was visible on MR images as a hypointense area
(Fig. 8B), and Prussian blue staining confirmed
positively stained cells within the hydrogel
(Fig. 8C). Immunostaining for blood vessels
(Reca-1, Abcam, UK) confirmed neovascularizat-
ion of the hydrogel implant (Fig. 8D). Staining for
neurofilaments (NF160, Sigma, St. Louis, USA)
showed axonal ingrowth into the hydrogel (Fig.
8E; Sykova and Jendelova, 2005; Sykova et al.,
2006). Hydrogels seeded with stem cells may
379
therefore serve as an alternative to the conven-
tional grafting of dissociated cells, benefiting from
advances in surface chemistry and the cell–cell or
cell–matrix interactions that occur during deve-
lopment or regeneration.
Discussion and conclusion
For clinical studies it is obvious that traditional
histopathological methods for cell detection are not
sufficient to inform us about the migration and fate
of the grafted cells in the host tissue. Because of its
high spatial resolution, MR imaging is suitable for
imaging the distribution of magnetically labeled
cells. Labeling methods that use the internalization
of iron oxide nanoparticles have some limitations.
To label cells with commercially available contrast
agents, such as Endorem, a relatively high concen-
tration of iron in the culture media is necessary
(Jendelova et al., 2003, 2004). This may cause a
transient drop in cell viability, which can be de-
pendent on the type of labeled cell. For improved
uptake of magnetic nanoparticles, the surface of
the contrast agent needs to be optimized, so that it
can induce the internalization of the particles into
the cytoplasm. One approach is to use internalizing
mabs. The mouse anti-Tfr mab OX-26 induces the
internalization of Tft upon binding. Nanoparticles
have been conjugated to OX-26 to deliver them to
cells by receptor-mediated endocytosis (Bulte et al.,
1999). A disadvantage of the use of internalizing
mabs, however, is that they are species-specific, and
a newly synthesized antibody is required when per-
forming studies in a different animal. In addition,
for eventual clinical use, there will be regulatory
issues regarding the use of a xenogeneic (i.e.,
mouse) protein. Recently, a combination of dext-
ran-coated SPIO nanoparticles with a transfection
agent has been used (Kalish et al., 2003). When the
complexes are added to a cell culture, the transfect-
ion agent effectively transports the nanoparticles
into the cells through electrostatic interactions.
However, each combination of transfection agent
and iron oxide nanoparticle has to be carefully tit-
rated and optimized for different cell cultures, since
lower concentrations may result in insufficient cel-
lular uptake, whereas higher concentrations may
380

Fig. 8. (A) Endorem-labeled cells seeded into a hydrogel. (B) On MR images 6 weeks after implantation, the hydrogel was visible as a
hypointensive area. (C) Prussian blue staining confirmed the presence of Endorem-labeled cells in the hydrogel. (D) Neovascularization
of the implant (Reca-1). (E) Ingrowth of NF160-positive axons into the gel.
 381

induce the precipitation of complexes or may be agents would enable us to follow the migration of
toxic to the cells. In addition, although it has been such cells when transplanted into humans, establish
reported that a poly-L-lysin-Feridex labeling com- the optimal number of transplanted cells, define
plex does not appear to affect the viability or pro- therapeutic windows, and monitor cell growth and
liferation of human MSCs, it was found that their possible side effects (malignancies). Currently, the
differentiation into chondrocytes was markedly in- described immunolabeling of specific cell types with
hibited (Kostura et al., 2004), while adipogenic and clinically approved MicroBeads may help to eluci-
osteogenic differentiation were not affected. PC- date the fate of implanted stem cells and evaluate
SPIO combine a low concentration of iron in the the effect of cell therapy in patients with various
cell culture media with the high efficiency of diseases of the brain or SCI.
transfection agents and thus have a broad poten-
tial for in vivo studies. However, for any intracel-
lular label, detailed studies examining the possible Abbreviations
biological side effects, which may vary among dif-
ferent cell types, are necessary. BBB Basso-Beattie-Bresnehan
On the other hand, particles that do not inter- BMCs bone marrow cells
nalize do not cause any metabolic alternations and BrdU bromdeoxyuridine
do not have any influence on cell viability or cell CNS central nervous system
proliferation (Jendelova et al., 2005). Their disad- DMEM Dulbecco’s modified Eagle medium
vantage is that particles that do not internalize and eGFP enhanced green fluorescent protein
thus stay attached to the outer cell membrane are EM electron microscopy
more likely to interfere with cell surface interac- ESCs embryonic stem cells
tions (including cell homing into tissues), may de- G-CSF granulocyte colony stimulating fac-
tach easily from the membrane, or may be tor
transferred to other cells. HPMA hydroxypropylmethacrylamide
MR tracking may serve in experimental models LIF leukemia inhibitory factor
to study how certain lesions target cell migration, Mab monoclonal antibody
at what speed the cells migrate, and for how long MCAO middle cerebral artery occlusion
they persist in the target organ. High contrast ef- MR magnetic resonance
fects on MR images are easily detected within an MRI magnetic resonance imaging
experimental time frame of 1–2 h per animal, MSCs mesenchymal stem cells
which is ideal for short and repetitive in vivo MRI. OECs olfactory ensheathing cells
In lesioned tissue, hemorrhage products give rise PCR polymerase chain reaction
to Prussian blue-positive deposits that are difficult PC-SPIO polycation-bound iron oxide super-
to distinguish from iron-containing nanoparticles paramagnetic nanoparticles
(Urdzikova et al., 2006). The hemorrhage degra- SCI spinal cord injury
dation products may also be partially localized to SPIO superparamagnetic iron oxide
macrophages because macrophages constitute the USPIO ultrasmall superparamagnetic iron
major cellular pathway for the redistribution of oxide
iron in mammals. Furthermore, hemorrhage con-
tributes to T2 weighted hypointensity, thus inter-
fering with the detection of labeled cells and
complicating the interpretation of MR images. Acknowledgment
Therefore, it is important to determine whether
cell labels remain co-localized with cell trans- This work was supported by grants from the
plants, especially under pathological conditions. Academy of Sciences of the Czech Republic
With proper attention to the limitations desc- AV0Z50390512, the Ministry of Education,
ribed above, labeling cells with superparamagnetic Youth, and Sports of the Czech Republic
382
1M0021620803, the National Grant Agency of the
Czech Republic GACR 309/06/1594, and the EC-
FP6 project DiMI: LSHB-CT-2005-512146.
References
Akiyama, Y., Radtke, C., Honmou, O. and Kocsis, J.D. (2002)
Remyelination of the spinal cord following intravenous de-
livery of bone marrow cells. Glia, 39(3): 229–236.
Anderson, S.A., Glod, J., Arbab, A.S., Noel, M., Ashari, P.,
Fine, H.A. and Frank, J.A. (2005) Noninvasive MR imaging
of magnetically labeled stem cells to directly identify ne-
ovasculature in a glioma model. Blood, 105(1): 420–425.
Arbab, A.S., Yocum, G.T., Wilson, L.B., Parwana, A., Jordan,
E.K., Kalish, H. and Frank, J.A. (2004) Comparison of
transfection agents in forming complexes with ferumoxides,
cell labeling efficiency, and cellular viability. Mol. Imaging,
3(1): 24–32.
Billotey, C., Wilhelm, C., Devaud, M., Bacri, J.C., Bittoun, J.
and Gazeau, F. (2003) Cell internalization of anionic ma-
ghemite nanoparticles: quantitative effect on magnetic reso-
nance imaging. Magn. Reson. Med., 49(4): 646–654.
Bjorklund, L.M., Sanchez-Pernaute, R., Chung, S., Andersson,
T., Chen, I.Y., McNaught, K.S., Brownell, A.L., Jenkins,
B.G., Wahlestedt, C., Kim, K.S. and Isacson, O. (2002) Em-
bryonic stem cells develop into functional dopaminergic neu-
rons after transplantation in a Parkinson rat model. Proc.
Natl. Acad. Sci. U.S.A., 99(4): 2344–2349.
Brustle, O., Jones, K.N., Learish, R.D., Karram, K., Choudh-
ary, K., Wiestler, O.D., Duncan, I.D. and McKay, R.D.
(1999) Embryonic stem cell-derived glial precursors: a source
of myelinating transplants. Science, 285(5428): 754–756.
Bulte, J.W., Ben-Hur, T., Miller, B.R., Mizrachi-Kol, R.,
Einstein, O., Reinhartz, E., Zywicke, H.A., Douglas, T. and
Frank, J.A. (2003) MR microscopy of magnetically labeled
neurospheres transplanted into the Lewis EAE rat brain.
Magn. Reson. Med., 50(1): 201–205.
Bulte, J.W., Douglas, T., Witwer, B., Zhang, S.C., Strable, E.,
Lewis, B.K., Zywicke, H., Miller, B., van Gelderen, P.,
Moskowitz, B.M., Duncan, I.D. and Frank, J.A. (2001) Mag-
netodendrimers allow endosomal magnetic labeling and in
vivo tracking of stem cells. Nat. Biotechnol., 19: 1141–1147.
Bulte, J.W., Hoekstra, Y., Kamman, R.L., Magin, R.L., Webb,
A.G., Briggs, R.W., Go, K.G., Hulstaert, C.E., Miltenyi, S.,
The, T.H., et al. (1992) Specific MR imaging of human
lymphocytes by monoclonal antibody-guided dextran-
magnetite particles. Magn. Reson. Med., 25(1): 148–157.
Bulte, J.W. and Kraitchman, D.L. (2004) Iron oxide MR con-
trast agents for molecular and cellular imaging. NMR Bio-
med., 17(7): 484–499.
Bulte, J.W., Zhang, S.C., van Gelderen, P., Herynek, V.,
Jordan, E.K., Duncan, I.D. and Frank, J.A. (1999) Neuro-
transplantation of magnetically labeled oligodendrocyte
progenitors: magnetic resonance tracking of cell migration
and myelination. Proc. Natl. Acad. Sci. U.S.A., 96:
15256–15261.
Bulte, J.W., Zhang, S.C., van Gelderen, P., Herynek, V.,
Jordan, E.K., Janssen, C.H., Duncan, I.D. and Frank, J.A.
(2002) Magnetically labeled glial cells as cellular MR contrast
agents. Acad. Radiol., 9(Suppl 1): S148–S150.
Emsley, J.G., Mitchell, B.D., Kempermann, G. and Macklis,
J.D. (2005) Adult neurogenesis and repair of the adult CNS
with neural progenitors, precursors, and stem cells. Prog.
Neurobiol., 75(5): 321–341.
Fairless, R. and Barnett, S.C. (2005) Olfactory ensheathing
cells: their role in central nervous system repair. Int. J. Bio-
chem. Cell. Biol., 37(4): 693–699.
Frank, J.A., Miller, B.R., Arbab, A.S., Zywicke, H.A., Jordan,
E.K., Lewis, B.K., Bryant Jr., L.H. and Bulte, J.W. (2003)
Clinically applicable labeling of mammalian and stem cells by
combining superparamagnetic iron oxides and transfection
agents. Radiology, 228(2): 480–487.
Hawrylak, N., Ghosh, P., Broadus, J., Schlueter, C., Green-
ough, W.T. and Lauterbur, P.C. (1993) Nuclear magnetic
resonance (NMR) imaging of iron oxide-labeled neural
transplants. Exp. Neurol., 121: 81–92.
Horak, D., Sykova, E., Babic, M., Jendelova, P. and Hajek, M.
(2006) Superparamagnetic iron oxide nanoparticles with
modified surface and their use as probes for stem cell labe-
ling. Patent application no: PV 1006–120, Czech Republic.
Jendelova, P., Herynek, V., De Croos, J., Glogarova, K.,
Andersson, B., Hajek, M. and Sykova, E. (2003) Imaging
the fate of implanted bone marrow stromal cells labeled with
superparamagnetic nanoparticles. Magn. Reson. Med., 50:
767–776.
Jendelova, P., Herynek, V., Urdzikova, L., Glogarova, K.,
Kroupova, J., Bryja, V., Andersson, B., Burian, M., Hajek,
M. and Sykova, E. (2004) MR tracking of transplanted bone
marrow and embryonic stem cells labeled by iron oxide
nanoparticles in rat brain and spinal cord. J. Neurosci. Res.,
76(2): 232–243.
Jendelova, P., Herynek, V., Urdzikova, L., Glogarova, K., Rah-
matova, S., Fales, I., Andersson, B., Prochazka, P., Zamecnik,
J., Eckschlager, T., Kobylka, P., Hajek, M. and Sykova, E.
(2005) MR tracking of human CD34+ progenitor cells sep-
arated by means of immunomagnetic selection and trans-
planted into injured rat brain. Cell Transplant., 14: 173–182.
Jiang, Y., Jahagirdar, B.N., Reinhardt, R.L., Schwartz, R.E.,
Keene, C.D., Ortiz-Gonzalez, X.R., Reyes, M., Lenvik, T.,
Lund, T., Blackstad, M., Du, J., Aldrich, S., Lisberg, A.,
Low, W.C., Largaespada, D.A. and Verfaillie, C.M. (2002)
Pluripotency of mesenchymal stem cells derived from adult
marrow. Nature, 418(6893): 41–49.
Kalish, H., Arbab, A.S., Miller, B.R., Lewis, B.K., Zywicke,
H.A., Bulte, J.W., Bryant Jr., L.H. and Frank, J.A. (2003)
Combination of transfection agents and magnetic resonance
contrast agents for cellular imaging: relationship between re-
laxivities, electrostatic forces, and chemical composition.
Magn. Reson. Med., 50(2): 275–282.
Kostura, L., Kraitchman, D.L., Mackay, A.M., Pittenger, M.F.
and Bulte, J.W. (2004) Feridex labeling of mesenchymal stem

cells inhibits chondrogenesis but not adipogenesis or osteo-
genesis. NMR Biomed., 17(7): 513–517.
Lee, I.H., Bulte, J.W., Schweinhardt, P., Douglas, T., Trifu-
novski, A., Hofstetter, C., Olson, L. and Spenger, C. (2004)
In vivo magnetic resonance tracking of olfactory ensheathing
glia grafted into the rat spinal cord. Exp. Neurol., 187(2):
509–516.
Lesny, P., Pradny, M., Jendelova, P., Michalek, J., Vacik, J.
and Sykova, E. (2006) Macroporous hydrogels based on
2-hydroxyethyl methacrylate. Part 4: growth of rat bone
marrow stromal cells in three-dimensional hydrogels with
positive and negative surface charges and in polyelectrolyte
complexes. J. Mater. Sci. Mater. Med., 17: 829–833.
McDonald, J.W., Liu, X.Z., Qu, Y., Liu, S., Mickey, S.K.,
Turetsky, D., Gottlieb, D.I. and Choi, D.W. (1999) Trans-
planted embryonic stem cells survive, differentiate and pro-
mote recovery in injured rat spinal cord. Nat. Med., 5(12):
1410–1412.
McKay, R.D. (2004) Stem cell biology and neurodegenerative
disease. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 359(1445):
851–856.
Mendonca Dias, M.H. and Lauterbur, P.C. (1986) Ferromag-
netic particles as contrast agents for magnetic resonance im-
aging of liver and spleen. Magn. Reson. Med., 3(2): 328–330.
Newman, M.B., Davis, C.D., Borlongan, C.V., Emerich, D.
and Sanberg, P.R. (2004) Transplantation of human umbil-
ical cord blood cells in the repair of CNS diseases. Expert
Opin. Biol. Ther., 4(2): 121–130.
Norman, A.B., Thomas, S.R., Pratt, R.G., Lu, S.Y. and
Norgren, R.B. (1992) Magnetic resonance imaging of neural
transplants in rat brain using a superparamagnetic contrast
agent. Brain Res., 594: 279–283.
Orlic, D., Kajstura, J., Chimenti, S., Bodine, D.M., Leri, A. and
Anversa, P. (2001a) Transplanted adult bone marrow cells
repair myocardial infarcts in mice. Ann. N.Y. Acad. Sci., 938:
221–230.
Orlic, D., Kajstura, J., Chimenti, S., Jakoniuk, I., Anderson,
S.M., Li, B., Pickel, J., McKay, R., Nadal-Ginard, B.,
Bodine, D.M., Leri, A. and Anversa, P. (2001b) Bone mar-
row cells regenerate infarcted myocardium. Nature,
410(6829): 701–705.
Pachernik, J., Bryja, V., Esner, M., Kubala, L., Dvorak, P. and
Hampl, A. (2005) Neural differentiation of pluripotent
mouse embryonal carcinoma cells by retinoic acid: inhibi-
tory effect of serum. Physiol. Res., 54(1): 115–122.
Park, K.I., Lachyankar, M., Nissim, S. and Snyder, E.Y. (2002)
Neural stem cells for CNS repair: state of the art and future
directions. Adv. Exp. Med. Biol., 506(Pt B): 1291–1296.
Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K.,
Douglas, R., Mosca, J.D., Moorman, M.A., Simonetti,
D.W., Craig, S. and Marshak, D.R. (1999) Multilineage po-
tential of adult human mesenchymal stem cells. Science,
284(5411): 143–147.
383
Pluchino, S., Zanotti, L., Deleidi, M. and Martino, G. (2005)
Neural stem cells and their use as therapeutic tool in neu-
rological disorders. Brain Res. Brain Res. Rev., 48(2):
211–219.
Renshaw, P.F., Owen, C.S., McLaughlin, A.C., Frey, T.G. and
Leigh Jr., J.S. (1986) Ferromagnetic contrast agents: a new
approach. Magn. Reson. Med., 3(2): 217–225.
Roitberg, B. (2004) Transplantation for stroke. Neurol. Res.,
26(3): 256–264.
Sasaki, M., Honmou, O., Akiyama, Y., Uede, T., Hashi, K. and
Kocsis, J.D. (2001) Transplantation of an acutely isolated
bone marrow fraction repairs demyelinated adult rat spinal
cord axons. Glia, 35: 26–34.
Sykova, E. and Jendelova, P. (2005) Magnetic resonance
tracking of implanted adult and embryonic stem cells in in-
jured brain and spinal cord. Ann. N.Y. Acad. Sci., 1049:
146–160.
Sykova, E. and Jendelova, P. (2006) Magnetic resonance track-
ing of transplanted stem cells in rat brain and spinal cord.
Neurodegener. Dis., 3: 62–67.
Sykova, E., Jendelova, P., Urdzikova, L., Lesny, P. and Hejcl,
A. (2006) Bone marrow stem cells and polymer hydrogels:
two strategies for spinal cord injury repair. Cell. Mol. Ne-
urobiol., 26(7–8): 1111–1127.
Takahashi, T., Kalka, C., Masuda, H., Chen, D., Silver, M.,
Kearney, M., Magner, M., Isner, J.M. and Asahara, T.
(1999) Ischemia- and cytokine-induced mobilization of bone
marrow-derived endothelial progenitor cells for ne-
ovascularization. Nat. Med., 5(4): 434–438.
Toma, C., Pittenger, M.F., Cahill, K.S., Byrne, B.J. and
Kessler, P.D. (2002) Human mesenchymal stem cells differ-
entiate to a cardiomyocyte phenotype in the adult murine
heart. Circulation, 105(1): 93–98.
Uccelli, A., Zappia, E., Benvenuto, F., Frassoni, F. and
Mancardi, G. (2006) Stem cells in inflammatory demyelina-
ting disorders: a dual role for immunosuppression and
neuroprotection. Expert Opin. Biol. Ther., 6(1): 17–22.
Urdzikova, L., Jendelova, P., Glogarova, K., Burian, M.,
Hajek, M. and Sykova, E. (2006) Transplantation of bone
marrow stem cells as well as mobilization by granulocyte-
colony stimulating factor promote recovery after spinal cord
injury in rat. J. Neurotrauma, 23: 1379–1391.
Zhang, Z.G., Jiang, Q., Zhang, R., Zhang, L., Wang, L., Arni-
ego, P., Ho, K.L. and Chopp, M. (2003) Magnetic resonance
imaging and neurosphere therapy of stroke in rat. Ann. Ne-
urol., 53(2): 259–263.
Zhang, Z.Y. and Smith, B.D. (2000) High-generation polycat-
ionic dendrimers are unusually effective at disrupting anionic
vesicles: membrane bending model. Bioconjug. Chem., 11(6):
805–814.
Zhao, C., Fancy, S.P., Magy, L., Urwin, J.E. and Franklin, R.J.
(2005) Stem cells, progenitors and myelin repair. J. Anat.,
207(3): 251–258.
This page intentionally left blank
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 27

Intrathecal drug delivery strategy is safe and

efficacious for localized delivery to the spinal cord

Molly S. Shoichet1,2,3,, Charles H. Tator4, Peter Poon1,2, Catherine Kang1,2 and

M. Douglas Baumann1,2

1
Department of Chemical Engineering and Applied Chemistry, University of Toronto, Donnelly Center for Cellular and
Biomolecular Research, Toronto, ON, M5S 3E1, Canada
2
Institute of Biomaterials and Biomedical Engineering, 164 College St., Toronto, ON, M5S 3G9, Canada
3
Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, ON, M5S 3H6, Canada
4
Krembil Neuroscience Centre, Toronto Western Research Institute, and Department of Surgery, University of Toronto,
399 Bathurst Street, Toronto, ON, M5 T 2S8, Canada

Abstract: Neuroprotective and neuroregenerative strategies for spinal cord injury repair are limited in part
by poor delivery techniques. A novel drug delivery system is being developed in our laboratory that can
provide localized release of therapeutically relevant molecules from an injectable hydrogel. Design criteria
were established for the hydrogel to be — injectable, fast-gelling, biocompatible, biodegradable and able to
release biologically active therapeutics when injected into the intrathecal space that surrounds the spinal
cord. This novel way of localized drug delivery to the spinal cord was tested first with a collagen gel and
then with a new hydrogel blend of hyaluronan and methylcellulose (HAMC). The underlying principle that
this novel methodology is both safe and able to provide localized delivery was proven with a fast gelling
collagen solution. Using a recombinant human epidermal growth factor, rhEGF, dispersed in collagen, we
demonstrated localized release to the injured spinal cord. We extended this technology to other fast-gelling
systems and found that HAMC was injectable due to the shear thinning property of hyaluronan (HA),
biocompatible and had some therapeutic benefit when injected into the intrathecal space using a com-
pression injury model in rats.

Keywords: spinal cord injury; localized drug delivery; hydrogels; hyaluronan; methyl cellulose; collagen;
epidermal growth factor

Introduction therapeutic strategies that have been investigated

Of the several therapeutic strategies investigated
for spinal cord injury repair (Pearse et al., 2004;
Tsai et al., 2004; Ramer et al., 2005), only systemic
delivery of methylprednisolone (MP) is used clini-
cally; however, results from its clinical trial have
been openly criticized (Hurlbert, 2001). Other
Corresponding author. Tel.: +1-416-978-1460; Fax: +1-416-
978-4317; E-mail: molly@ecf.utoronto.ca
hold great promise (Gorio et al., 2002), yet sys-
temic delivery may cause side effects, and many
promising therapeutic proteins degrade when de-
livered systemically. Moreover, many therapeutic
molecules are unable to cross the blood-spinal
cord barrier (BSCB). These difficulties suggest that
local delivery strategies are required. Two in-
trathecal techniques have been used to test local-
ized delivery: (1) bolus injection, the effects of
which are short-lived because the therapeutic

DOI: 10.1016/S0079-6123(06)61027-3 385

386
agent is washed away by the cerebrospinal fluid
(CSF) flow (Terada et al., 2001; Yaksh et al.,
2004); and (2) mini-pump delivery, which is inva-
sive and can lead to complications due to infection
and/or blockage of the catheter (Jones and
Tuszynski, 2001). A third extradural delivery strat-
egy is being investigated clinically for the delivery
of rho-kinase inhibitors; however, in this methodo-
logy where the therapeutic molecule is applied in a
fibrin glue to the external surface of the dura, the
molecule must diffuse across the dura, the in-
trathecal space and then the spinal cord, resulting
in only a very small fraction of the molecule
reaching the target tissue (Dergham et al., 2002;
McKerracher and Higuchi, 2006).
Injectable delivery strategy
Given the limitations of current clinical strategies,
we sought to develop an improved localized deli-
very strategy. As with any localized strategy, the
therapeutic benefits should include decreased sys-
temic toxicity and decreased dose applied. In the
proposed model, we sought to develop a delivery
strategy that would localize the therapeutic mole-
cules at the site of injection in the intrathecal space
(or subarachnoid space) to localize delivery to the
injured spinal cord. To this end, our goal was to
develop a biocompatible, biodegradable, injecta-
ble, fast gelling polymer within which therapeutic
molecules could be dispersed and released. We
aimed to inject the drug delivery system via a
30 gauge needle into the intrathecal cavity and
have it gel quickly (less than 1 min) in order to
localize release at this site (see Fig. 1 for concep-
tual understanding of new strategy).
In the first set of studies we aimed to demon-
strate safety because this was a completely new
and novel approach toward drug delivery to the
spinal cord. Several questions arose in the deve-
lopment of this system, such as: What would hap-
pen if we injected a polymeric hydrogel into the
intrathecal space? Would this strategy cause any
harm to either non-injured or spinal cord injured
rats? Given that the volume of CSF in the adult rat
is 250 ml, we chose to investigate the injection of
20 ml, which represents less than 10% of the total
CSF volume, a volume which we expected to be
tolerated based on previous studies where liquids
of a similar volume had been injected (Rieselbach
et al., 1962). All animal procedures were per-
formed in accordance with the Guide to the Care
and Use of Experimental Animals (Canadian
Council on Animal Care) and protocols were ap-
proved by the Animal Care Committee of the Re-
search Institute of the University Health Network.
Importantly, injection into the intrathecal space
of both non-injured and spinal cord injured rats
was tolerated and had no deleterious affects rela-
tive to the injection of an artificial CSF (aCSF).
The safety of the delivery strategy methodology
was proven with a fast gelling collagen solution
(Jimenez Hamann et al., 2003) that was injected
into adult rats either non-injured or after acute
spinal cord compression injury. The compression
injury was achieved at thoracic level 2 (T2) using a
modified aneurysm clip with a closing force rang-
ing from 35 to 56 g, corresponding to injuries
ranging from moderate to severe (Rivlin and
Tator, 1977, 1978). Safety was assessed by histo-
logy, immunohistochemistry and open-field motor
function using the Basso, Beattie and Bresnahan
(BBB) scoring scale (Basso et al., 1996). This is a
21-point scale that ranks no locomotion at 0 and
normal gait at 21. Neither healthy nor injured spi-
nal cord tissue was affected by this injection strat-
egy involving collagen gel. Similarly, locomotor
function was unaffected.
Localized release of bioactive factors
Having demonstrated the safety of this new in-
trathecal delivery strategy (Jimenez Hamann et al.,
2003), our next goal was to demonstrate localized
delivery and efficacy. Based on previous research
which demonstrated that mini-pump delivery into
the subarachnoid space of EGF and FGF2 stimu-
lated endogenous stem cell proliferation (Kojima
and Tator, 2000), we chose to study the effects of
their delivery by the injectable collagen system.
Using recombinant human (rh) EGF and rhFGF2,
we first investigated the bioactivity and release
profile of each factor from the collagen gel in vitro.
Released EGF was bioactive for 14 days and

Fig. 1. Injectable delivery strategy to the intrathecal space of the spinal cord. A fine, 30 gauge needle pierces the dura of the spinal cord
and is used to deliver a fast gelling, biodegradable, biocompatible hydrogel for localized delivery of therapeutic molecules to the injured
spinal cord. (Adapted with permission, copyright Michael Corrin (2005).)
FGF2 for 56 days. Confident that our system al-
lowed release of bioactive factors, we investigated
the release of these factors in vivo (in the presence
of heparin as a stabilizer). Using a similar animal
model — 35 g clip compression injury at T2 — we
compared the delivery of rhEGF/rhFGF2 vs.
aCSF. We had determined that there was no
cross-reactivity between: (1) the rhEGF and the
endogenous EGF in the rat; and (2) the rhFGF2
and endogenous FGF2, so that we could follow
the depth of penetration into the tissue by
immunohistochemistry. We determined that
rhEGF penetrated deeply into the injured spinal
cord tissue as evident at 30 min and 6 h after in-
jection of the collagen gel. At 1 and 7 days, it was
difficult to detect the rhEGF in the injured spinal
cord due to either a fast release from the collagen
gel or the limited sensitivity of the immunohisto-
chemical method. Interestingly, in non-injured
controls, rhEGF penetrated the spinal cord tissue,
but the penetration was superficial and not deep.
The greater depth of penetration observed in the
injured tissue likely reflects its greater permeabil-
ity. Figure 2 summarizes the depth of rhEGF pen-
etration observed in the injured and non-injured
387
spinal cord. Unlike rhEGF, rhFGF2 did not pene-
trate the injured (or non-injured) spinal cord tissue
beyond the pia. It is not clear why EGF penetrated
the tissue and FGF2 did not; however, this may
reflect the greater molecular weight of FGF2 re-
lative to EGF or its charge. Importantly, release of
EGF/FGF2 from the collagen did stimulate the
ependymal cells that line the central canal, which is
believed to be the source of endogenous stem cells
(Martens et al., 2002). There was also evidence of
reduced cavitation around the site of injury in
those animals injected with EGF/FGF2 relative to
aCSF; however, there was no improvement in
function. Stimulation of endogenous stem cells
was also observed in studies where EGF/FGF2
was delivered by injection into the ventricles in the
brain (Martens et al., 2002).
The novel injectable drug delivery system was
shown to be safe and to localize release of these
agents (Jimenez Hamann et al., 2005). This pro-
vided a new paradigm for drug delivery to the
spinal cord that we have continued to investigate.
However, cellular build-up in the intrathecal space
was observed when EGF/FGF2 was incorporated
into the collagen gel, likely attracting fibroblasts
388

Fig. 2. Injection of rhEGF via the fast-gelling collagen gel to the (A) uninjured and (B) injured spinal cord demonstrates localized
delivery. Using immunohistochemical labeling against rhEGF, we observed a superficial penetration of rhEGF into the uninjured
spinal cord (A) and deep penetration into the injured spinal cord (B) as demarcated by the dotted lines. Penetration of rhEGF is
observed after 30 min and 6 h of injection, but not after 1 and 7 days. Figure is at 6 h after injection. (Adapted with permission from
Jimenez Hamann et al. (2005).)

and other cell types. To overcome this limitation
of the delivery vehicle, it was redesigned to be non-
adhesive to cells while still meeting the other de-
sign criteria of biocompatibility, biodegradability,
injectability, and fast gelation.
New fast-gelling blend shows therapeutic promise
Of the many available chemical and physical gels
available, none met all of our design criteria, re-
quiring us to develop a new gel system. While
physical gels are not as stable or robust as chemi-
cal gels, we believed that they could be sufficiently
stable in the intrathecal space. Both methylcellu-
lose and hyaluronan met all of our design criteria,
except fast gelation. Methylcellulose (MC) is an
inverse gelling polymer that forms a weak gel at
371C in water (Li et al., 2001), but does not gel fast
enough for our injectable drug delivery system.
Hyaluronan (HA) does not gel on its own. How-
ever, a blend of HA and MC (HAMC) gel quickly
and in fact form a gel at room temperature, likely
due to a ‘‘salting out’’ effect that HA has on MC
(Xu et al., 2004).
In addition to promoting faster gelation of MC,
HA is attractive because it is non-immunogenic,
biocompatible (Vercruysse and Prestwich, 1998)
and known to promote wound healing by reducing
inflammation and minimizing tissue adhesion and
scar formation (Balazs, 1991). HA is also shear
thinning — that is it flows in the direction of stress
(Weiss, 2000) — and it is this unique property that
allows HAMC to meet the design criteria of in-
jectable and fast gelling.
Prior to initiating the in vivo study in rats,
HAMC was evaluated in vitro for degradation and
cell adhesion. HAMC was found to degrade (or
erode) within 14 days and was non-adhesive to
cells. The lack of adhesion of 3T3 fibroblasts to
HAMC gels demonstrated the non-adhesive prop-
erty of HAMC in vitro.
Thus this new fast-gelling, injectable HAMC
was investigated as a drug delivery vehicle to the
injured spinal cord. The first test was one of safety.
Since this was a new blend, we had to first demon-
strate that it would not cause any deleterious
effects in the rat animal model. Thus the biocom-
patibility of HAMC within the intrathecal space
was examined in vivo in both uninjured and spinal
cord injured rat animal models relative to control
animals receiving aCSF injections. In this set of
studies HAMC was injected at T2 after a moderate
clip compression injury using a 35 g clip, similar to
the in vivo design with the injectable collagen sys-
tem. Injection of HAMC in spinal cord injured
and uninjured rats were compared to a control
injection of aCSF. The only difference other than
the choice of material was the injection volume,
where 10 ml was used instead of the 20 ml of col-
lagen injected.

Fig. 3. Representative histology sections of each experimental group stained with Luxol Fast Blue and counterstained with Hema-
toxylin and Eosin. Sections showing injection site in dura of: (A) uninjured rat injected with aCSF (arrow points to unsealed dura that
was punctured with needle); and (B) uninjured rat injected with HAMC (arrow points to resealed dura). (Scale bar ¼ 200 mm.)
(Adapted with permission from Gupta et al. (2006).)
It is important to realize that this new technique
requires that the dura be punctured, which clini-
cally can result in serious post-dural puncture
headaches. Interestingly, in the spinal cord injured
animals that received HAMC, the dura was ob-
served to re-seal whereas no re-sealing was ob-
served in animals that received aCSF (Fig. 3). This
key result is critical to the long-term application of
this material clinically. After 28 days there was no
evidence of scar formation, arachnoiditis, syringo-
myelia or of HAMC itself in the intrathecal space.
The benefits of HAMC were observed at the
tissue level and the functional level. Both cavita-
tion volume and inflammatory response were less
in those animals injected with HAMC relative to
aCSF. For example, the cavitation volume in an-
imals injected with aCSF was 52.1720.1 mm3
whereas for HAMC the cavitation volume was
36.676.0 mm3. While not statistically different (at
95% confidence), the trend of reduced cavitation
for animals injected with HAMC is important and
demonstrates that injection of HAMC is safe and
affected neither the normal spinal cord nor the
severity of injury. The inflammatory response was
determined by quantification of ED-1 labeled
macrophages/microglia based on pixel inten-
sity. For animals injected with aCSF, the area oc-
cupied was approximately 3.7971.39
105 mm2
389
vs. 2.4270.35
105 mm2 for animals injected with
HAMC. Thus, the inflammatory response at
28 days after injection, was significantly reduced
in those animals injected with HAMC relative to
aCSF controls (po0.05, n ¼ 6). This was an un-
expected, yet beneficial effect of HAMC delivery,
potentially due to the beneficial anti-inflammatory
effects attributed to HA in other tissues (Dougados,
2000; Sheehan et al., 2003; Cabrera et al., 2004;
Roth et al., 2005). The re-sealing of the dura pro-
vides another advantage over other delivery tech-
niques, such as bolus and intrathecal mini-pump
where the dura may not seal at the punctured site,
resulting in the potential for CSF leakage and/or
infection.
The benefit or detriment of the inflammatory
response is widely debated in the neuroscience
community (Schwartz, 2003). For example,
Michael Schwartz is investigating the therapeutic
benefit of macrophage delivery (Rapalino et al.,
1998; Knoller et al., 2005). In the case of HAMC
delivery, the reduced inflammatory response ob-
served, judged by ED-1 immunohistochemistry,
was thought to be beneficial and there was also
improved locomotor function as described below.
To gain a greater perspective on the safety of
HAMC delivery, the locomotor function of ani-
mals was scored by the BBB and grid walk
390

A
21

18

15
BBB Score

12

0
0 5 10 15 20 25 30
Days after surgery
B
18

16

14

12
# of Foot Falls

10

0
0 5 10 15 20 25 30
Days after Surgery

Fig. 4. (A) BBB scores over 1 month to measure the effect of HAMC on animal function: (&) HAMC uninjured, (W) aCSF
uninjured, (’) HAMC injured, and (m) aCSF injured. The two uninjured groups scored normal (21 points) at all times; (B) % foot
falls of total steps per grid walk run for (&) HAMC uninjured, (W) aCSF uninjured, (’) HAMC injured, and (m) aCSF injured. The
two uninjured groups had virtually zero footfalls at all times. Data is shown as mean7standard deviation (uninjured animals: n ¼ 4;
injured animals: n ¼ 8). The asterisk (*) at day 7 in (A) indicates significant difference (p ¼ 0.035). (Adapted with permission from
Gupta et al. (2006).)

analysis. As shown in Fig. 4A, uninjured animals
injected with HAMC or aCSF showed no differ-
ence in locomotor function, demonstrating
safety of the new injectable HAMC material and
confirming safety of the methodology. Impor-
tantly, injection of HAMC in spinal cord injured
animals showed a general trend of improved
locomotor function relative to injection of
aCSF, and a statistically significant difference
was observed at day 7 (p ¼ 0.035). This sug-
gests a mild neuroprotective effect of HAMC
(Rabchevsky et al., 1999) possibly due to the
decreased inflammatory response observed by
immunohistochemistry.

To confirm the functional benefit observed, grid
walk analysis was completed following the meth-
odology described by Metz et al. (2000), where
previously trained animals were scored on the
number of foot falls that occurred when walking
across a horizontal ladder. As shown in Fig. 4B,
uninjured animals had nearly perfect scores of zero
foot falls throughout a run for both HAMC-
injected and aCSF-injected animals, thereby corrob-
orating the normal BBB scores shown (cf. Fig. 4A).
It is important to note that uninjured animals
often miss 1–2 steps on the horizontal ladder (Metz
et al., 2000; Hains et al., 2004), likely due to their
carelessness when crossing the grid quickly. For
injured animals, the HAMC-injected animals had
fewer foot falls than aCSF-injected animals, but
the difference between HAMC-injected and aCSF-
injected animals was not statistically different.
Conclusions and future outlook
We have demonstrated a new and safe way to
achieve localized delivery to the injured spinal
cord. First with collagen and then with HAMC,
we demonstrated that the underlying methodology
is safe, having neither a negative impact on the
tissue nor locomotor function. Moreover, there
may be a small therapeutic benefit of HAMC
alone, indicating that this material may be useful
as a delivery and for its inherent ‘‘wound healing’’
properties. Lastly, we found that this delivery
strategy resulted in localized release of growth
factors into the injured spinal cord, thereby prov-
ing the hypothesis that this strategy could achieve
local release into the injured tissue.
In on-going studies we continue to examine the
therapeutic benefit of the release of neuroprotec-
tive and neuroregenerative factors to limit degen-
eration and promote regeneration, respectively,
following traumatic injury to the spinal cord. Lo-
calized release from hydrogel systems continues to
be explored for its potential therapeutic benefit in
spinal cord injury.
Acknowledgments
We are grateful to Maria Jimenez-Hamann and
Dimpy Gupta for initial studies. We thank the
391
Canadian Institute of Health Research, the Canada
Foundation for Innovation and the Ontario
Innovation Trust and the Ontario Branch of the
Canadian Paraplegic Association for funding.
References
Balazs, E.A. (1991) Cosmetic and pharmaceutical applications
of polymers. In: Gebelein C.G. (Ed.), Medical Applications
of Hyaluronan and its Derivatives. Plenum Press, New York,
pp. 293–310.
Basso, D.M., Beattie, M.S. and Bresnahan, J.C. (1996) Graded
histological and locomotor outcomes after spinal cord con-
tusion using the NYU weight-drop device versus transection.
Exp. Neurol., 139: 244–256.
Cabrera, P.V., Blanco, G., Alaniz, L., Greczanik, S., Garcia,
M., Alvarez, E. and Hajos, S.E. (2004) CD44 and hyaluronic
acid regulate in vivo iNOS expression and metalloproteinase
activity in murine air-pouch inflammation. Inflamm. Res.,
53: 556–566.
Dergham, P., Ellezam, B., Essagian, C., Avedissian, H., Lubell,
W.D. and McKerracher, L. (2002) Rho signaling pathway
targeted to promote spinal cord repair. J. Neurosci., 22:
6570–6577.
Dougados, M. (2000) Sodium hyaluronate therapy in osteoar-
thritis: arguments for a potential beneficial structural effect.
Semin. Arthritis Rheum., 30: 19–25.
Gorio, A., Gokmen, N., Erbayraktar, S., Yilmaz, O., Mad-
aschi, L., Cichetti, C., Di Giulio, A.M., Vardar, E., Cerami,
A. and Brines, M. (2002) Recombinant human erythropoietin
counteracts secondary injury and markedly enhances neuro-
logical recovery from experimental spinal cord trauma. Proc.
Natl. Acad. Sci. U.S.A., 99: 9450–9455.
Gupta, D., Tator, C.H. and Shoichet, M.S. (2006) Fast-gelling
injectable blend of hyaluronan and methylcellulose for in-
trathecal, localized delivery to the injured spinal cord. Bio-
materials, 27: 2370–2379.
Hains, B.C., Saab, C.Y., Lo, A.C. and Waxman, S.G. (2004)
Sodium channel blockade with phenytoin protects spinal
cord axons, enhances axonal conduction, and improves func-
tional motor recovery after contusion SCI. Exp. Neurol., 188:
365–377.
Hurlbert, R.J. (2001) The role of steroids in acute spinal cord
injury: an evidence-based analysis. Spine, 26: S39–S46.
Jimenez Hamann, M.C., Tator, C.H. and Shoichet, M.S. (2005)
Injectable intrathecal delivery system for localized adminis-
tration of EGF and FGF-2 to the injured rat spinal cord.
Exp. Neurol., 194: 106–119.
Jimenez Hamann, M.C., Tsai, E.C., Tator, C.H. and Shoichet,
M.S. (2003) Novel intrathecal delivery system for treatment
of spinal cord injury. Exp. Neurol., 182: 300–309.
Jones, L.L. and Tuszynski, M.H. (2001) Chronic intrathecal
infusions after spinal cord injury cause scarring and com-
pression. Microsc. Res. Tech., 54: 317–324.
Knoller, N., Auerbach, G., Fulga, V., Zelig, G., Attias, J.,
Bakimer, R., Marder, J.B., Yoles, E., Belkin, M., Schwartz, M.
392
and Hadani, M. (2005) Clinical experience using incubated
autologous macrophages as a treatment for complete spinal
cord injury: phase I study results. J. Neurosurg. Spine, 3:
173–181.
Kojima, A. and Tator, C.H. (2000) Epidermal growth factor
and fibroblast growth factor 2 cause proliferation of
ependymal precursor cells in the adult rat spinal cord in
vivo. J. Neuropathol. Exp. Neurol., 59: 687–697.
Li, L., Thangamathesvaran, P.M., Yue, C.Y., Tam, K.C., Hu,
X. and Lam, Y.C. (2001) Gel network structure of methyl-
cellulose in water. Langmuir, 17: 8062–8068.
Martens, D.J., Seaberg, R.M. and van der Kooy, D. (2002) In
vivo infusions of exogenous growth factors into the fourth
ventricle of the adult mouse brain increase the proliferation of
neural progenitors around the fourth ventricle and the central
canal of the spinal cord. Eur. J. Neurosci., 16: 1045–1057.
McKerracher, L. and Higuchi, H. (2006) Targeting Rho to
stimulate repair after spinal cord injury. J. Neurotrauma, 23:
309–317.
Metz, G.A., Merkler, D., Dietz, V., Schwab, M.E. and Fouad,
K. (2000) Efficient testing of motor function in spinal cord
injured rats. Brain Res., 883: 165–177.
Pearse, D.D., Pereira, F.C., Marcillo, A.E., Bates, M.L.,
Berrocal, Y.A., Filbin, M.T. and Bunge, M.B. (2004) cAMP
and Schwann cells promote axonal growth and functional
recovery after spinal cord injury. Nat. Med., 10: 610–616.
Rabchevsky, A.G., Fugaccia, I., Fletcher-Turner, A., Blades,
D.A., Mattson, M.P. and Scheff, S.W. (1999) Basic fibroblast
growth factor (bFGF) enhances tissue sparing and functional
recovery following moderate spinal cord injury. J. Neuro-
trauma, 16: 817–830.
Ramer, L.M., Ramer, M.S. and Steeves, J.D. (2005) Setting the
stage for functional repair of spinal cord injuries: a cast of
thousands. Spinal Cord, 43: 134–161.
Rapalino, O., Lazarov-Spiegler, O., Agranov, E., Velan, G.J.,
Yoles, E., Fraidakis, M., Solomon, A., Gepstein, R., Katz,
A., Belkin, M., Hadani, M. and Schwartz, M. (1998) Im-
plantation of stimulated homologous macrophages results in
partial recovery of paraplegic rats. Nat. Med., 4: 814–821.
Rieselbach, R.E., Di Chiro, G., Freireich, E.J. and Rall, D.P.
(1962) Subarachnoid distribution of drugs after lumbar in-
jection. N. Engl. J. Med., 267: 1273–1278.
Rivlin, A.S. and Tator, C.H. (1977) Objective clinical assess-
ment of motor function after experimental spinal cord injury
in the rat. J. Neurosurg., 47: 577–581.
Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute
spinal cord compression in a new acute cord injury model in
the rat. Surg. Neurol., 10: 38–43.
Roth, A., Mollenhauer, J., Wagner, A., Fuhrmann, R., Straub,
A., Venbrocks, R.A., Petrow, P., Brauer, R., Schubert, H.,
Ozegowski, J., Peschel, G., Muller, P.J. and Kinne, R.W.
(2005) Intra-articular injections of high-molecular-weight
hyaluronic acid have biphasic effects on joint inflammation
and destruction in rat antigen-induced arthritis. Arthritis
Res. Ther., 7: R677–R686.
Schwartz, M. (2003) Macrophages and microglia in central
nervous system injury: are they helpful or harmful? J. Cereb.
Blood Flow Metab., 23, 385–394.
Sheehan, K.M., DeLott, L.B., Day, S.M. and DeHeer, D.H.
(2003) Hyalgan has a dose-dependent differential effect on
macrophage proliferation and cell death. J. Orthop. Res., 21:
744–751.
Terada, H., Kazui, T., Takinami, M., Yamashita, K., Washi-
yama, N. and Muhammad, B.A. (2001) J. Thorac. Card-
iovasc. Surg., 122: 979–985.
Tsai, E.C., Dalton, P.D., Shoichet, M.S. and Tator, C.H. (2004)
Synthetic hydrogel guidance channels facilitate regeneration
of adult rat brainstem motor axons after complete spinal
cord transection. J. Neurotrauma, 21: 789–804.
Vercruysse, K.P. and Prestwich, G.D. (1998) Hyaluronate de-
rivatives in drug delivery. Crit. Rev. Ther. Drug Carrier Syst.,
15: 513–555.
Weiss, C. (2000) New frontiers in medical sciences: redefining
Hyaluronan. In: Weigel, G. A. a. P. H. (Eds.), Why Visco-
elasticity is Important for the Medical Uses of Hyaluronan
and Hylans. Excerpta Medica, pp. 89–103.
Xu, Y., Wang, C., Tam, K.C. and Li, L. (2004) Salt-assisted
and salt-suppressed sol-gel transitions of methylcellulose in
water. Langmuir, 20: 646–652.
Yaksh, T.L., Horais, K.A., Tozier, N., Rathbun, M., Richter,
P., Rossi, S., Grafe, M., Tong, C., Meschter, C., Cline, J.M.
and Eisenach, J. (2004) Toxicol. Sci., 80: 322–334.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 28

Decompression craniectomy after traumatic brain

injury: recent experimental results

Nikolaus Plesnila

Laboratory of Experimental Neurosurgery, Department of Neurosurgery and Institute for Surgical Research, University of
Munich Medical Center, GroX hadern, Marchioninistr 15, 81377 Munich, Germany

Abstract: Among the secondary events occurring after traumatic brain injury (TBI) pathologically in-
creased intracranial pressure (ICP) correlates most closely with poor outcome. In addition to infusion of
hypertonic solutions, e.g. mannitol, and other medical measures, decompression of the brain by surgical
removal of a portion of the cranium (craniectomy) has been used for many decades as an intuitive strategy
for the treatment of post-traumatic ICP increase. The lack of evidence-based clinical and controversial
experimental data, however, resulted in decompressive craniectomy to be recommended by most national
and international guidelines only as a third tier therapy for the treatment of pathologically elevated ICP.
Ongoing clinical trials on the use of decompressive craniectomy after TBI may clarify many aspects of the
clinical application of this technique, however, some important pathophysiological issues, e.g. the timing of
decompression craniectomy, its effect on brain edema formation, and its role for secondary brain damage,
are still widely discussed and can only be addressed in experimental settings. The aim of the current review
was therefore to summarize and discuss recent experimental data dealing with the use of decompression
craniectomy following TBI. The present results suggest that surgical decompression effectively prevents
secondary brain damage when performed early enough. Although caution should be taken when trans-
ferring conclusions drawn from experimental settings to the clinical situation, the current literature suggests
that the timing of decompression may be of utmost importance in order to exploit the full neuroprotective
potential of craniectomy following TBI.

Keywords: traumatic brain injury; decompression; craniectomy; intracranial pressure; mice; review

Introduction History of decompression craniectomy

Despite significant improvements in the manage-
ment of traumatic brain injury (TBI) within the
past two decades, brain edema formation resulting
in refractory intracranial hypertension remains the
leading cause of unfavorable outcome and death
among affected patients (Murray et al., 1999).
Corresponding author. Tel.: +49-89-2180-76-535; Fax: +49-
89-2180-76-532; E-mail: plesnila@med.uni-muenchen.de
Opening of the cranial cavity is the most intuitive
treatment option for increased intracranial pres-
sure (ICP). Accordingly, as soon as increased ICP
was considered an important component of the
pathophysiology of TBI, i.e. at the beginning of the
last century, surgical decompression was used for
the treatment of patients suffering from the sequels
of severe TBI. The first report on the use of de-
compression craniectomy following TBI was pub-
lished by the legendary Swiss surgeon and Noble
laureate for medicine Theodor Emil Kocher (1901).

DOI: 10.1016/S0079-6123(06)61028-5 393

394
The intense collaboration of Kocher with Harvey
Cushing resulted — among others — in the use of
decompression craniectomy also for the treatment
of other cerebral disorders, e.g., brain tumors and
vascular malformations (Cushing, 1905).
Decompression craniectomy after TBI in man
Kocher’s initial positive experience with decom-
pression craniectomy was followed by a plethora of
controversial clinical findings on the use of surgical
decompression following brain trauma during the
past 100 years (Diemath, 1966; Clark et al., 1968;
Kerr, 1968; Kjellberg and Prieto, 1971; Ransohoff
and Benjamin, 1971; Ogawa et al., 1974; Venes
and Collins, 1975; Pereira et al., 1977; Makino and
Yamaura, 1979; Yamaura et al., 1979; Gerl and
Tavan, 1980). These early reports helped to deter-
mine the most effective surgical technique for de-
compression craniectomy in TBI patients. Another
important finding was that decompression craniec-
tomy, most effectively together with opening of the
dura mater, lowers pathologically elevated ICP, a
fact confirmed also by more recent studies (Dam
et al., 1996; Polin et al., 1997; Yoo et al., 1999;
Berger et al., 2002; Schneider et al., 2002; Ruf et al.,
2003; Aarabi et al., 2006; Josan and Sgouros, 2006;
Sahuquillo and Arikan, 2006). However, despite
the positive effect of decompression craniectomy
on otherwise uncontrollable intracranial hyperten-
sion, the effect of surgical decompression on mor-
tality and overall functional outcome following
TBI remained controversial. For example, a recent
review of 400 patients with severe TBI reported
that craniectomised patients may have benefited
from the procedure (Meier and Grawe, 2003). An-
other study matching TBI patients with craniec-
tomy with non-craniectomised controls from the
Traumatic Coma Data Bank reported good recov-
ery or moderate disability in 37% of decompressed
patients as compared with only 16% in the historic
control group (Polin et al., 1997). Although a num-
ber of other studies support these findings (Kunze
et al., 1998; Guerra et al., 1999; De Luca et al.,
2000; Kontopoulos et al., 2002; Schneider et al.,
2002; Ziai et al., 2003), the general problem of all
these investigations is the lack of a prospective,
randomized, and controlled study design and in
some cases an insufficient number of observations.
In additions to these limitations, there are also a
number of publications reporting no benefit from
decompressive craniectomy (Clark et al., 1968;
Munch et al., 2000; Soukiasian et al., 2002;
Messing-Junger et al., 2003). For example, a
retrospective study comparing craniectomised
patients from one center with a matched control
population form the Traumatic Coma Data Bank
reported no differences between these two groups
regarding ICP, therapy intensity level, and overall
mortality (Munch et al., 2000), a finding also cor-
roborated by others (Soukiasian et al., 2002). Ac-
cordingly, despite a report of a randomized trial
demonstrating beneficial use of decompression
craniectomy in children (Taylor et al., 2001) and
clear benefits of surgical decompression in stroke
patients (Schwab et al., 1998; Fraser and Hartl,
2005), the current literature does not provide
evidence-based proof for the general use of decom-
pression surgery following TBI in adults. As a re-
sult current American and European TBI
guidelines still recommend decompression craniec-
tomy only as a ‘‘second tier’’ therapeutic option for
adult patients with TBI (Maas et al., 1997; The
Brain Trauma Foundation. The American Associ-
ation of Neurological Surgeons. The Joint Section
on Neurotrauma and Critical Care, 2000).
Decompression craniectomy after experimental TBI
Because heterogeneous patient populations and
mechanisms of injury may have hampered the
investigation of decompression craniectomy in the
clinical environment, many researchers tried to
evaluate the therapeutic potential of decompres-
sion craniectomy under more homogenous and
controlled conditions, i.e. in experimental models
of TBI. Most surprisingly also experimental data
on the therapeutic use of decompression craniec-
tomy after brain trauma were not univocal
(Moody et al., 1968; Cooper et al., 1979; Gaab
et al., 1979; Burkert and Plaumann, 1989; Rinaldi
et al., 1990). Although most authors agree with the
clinical finding that craniotomy lowers pathologi-
cally elevated ICP (Moody et al., 1968; Gaab et al.,

1979; Rinaldi et al., 1990), an increase in brain
edema formation and enhanced histopathological
damage were observed after craniectomy in experi-
mental animals (Moody et al., 1968; Gaab et al.,
1979; Rinaldi et al., 1990). Although it is difficult
to retrospectively evaluate why experimental stud-
ies show such heterogeneous results, the consider-
ation of some pathophysiological concepts not
available at the time when these experiments were
performed may deepen our understanding on the
use of decompression craniectomy after TBI.
Pathophysiological considerations
Experimental and clinical findings demonstrate that
the pathophysiology of TBI consists of two com-
ponents: one is determined by the immediate me-
chanical damage to the brain by the traumatic
insult and the other is determined by a plethora of
secondary events on the functional, neurochemical,
and molecular level, which ultimately result in addi-
tional loss of initially viable brain tissue. While the
first, or primary injury component occurs within a
few milliseconds during the mechanical impact and
32
30
Contusion volume (mm3)
28
26
24
22
20
18
16
0
Fig. 1. Secondary expansion of the volume of a traumatic cortical contusion after brain injury. As compared with the primary damage
measured immediately after the traumatic impact (0 h; controlled cortical impact in mice) secondary processes result in delayed and
progressive expansion of the contusion by more than 60% of its initial size during the first 24 h after experimental TBI. Adapted with
permission from Zweckberger et al. (2006).
395
can only be avoided by prophylactic measures, sec-
ondary injury is a dynamic process which develops
gradually over several hours (maybe days) and
offers therefore the possibility for therapeutic inter-
ventions (Murray et al., 1999; Hartl and Ougorets,
2004; Nortje and Menon, 2004). In case of experi-
mental traumatic contusions, as observed after the
controlled cortical impact (CCI) brain injury model,
these two components can easily be distinguished
by standard histopathological techniques, e.g. Nissl
staining. Determination of the contusion volume
15 min after injury reveals a value of
20 mm3,
while 24 h after TBI the contusion volume plateaus
at
32 mm3. This indicates that the primary
contused tissue, which is necrotic already minutes
after TBI, i.e. it contains only pyknotic nuclei and
does not show any perfusion, increases by 60%
(Zweckberger et al., 2006). Determination of the
contusion volume at intermediate time points, i.e.
2, 6, and 12 h after TBI, shows a fast and linear
expansion of contused tissue volume within the first
12 h after trauma with only limited additional tissue
loss from 12 to 24 h (Fig. 1). Since blood-brain
barrier leakage, brain edema formation, and ICP
increase show a very similar early time course
: p<0.05 vs 15min; n=7
+63%
+52%
+32%
0 2 6 12 24
Time after trauma (h)
396
(Zweckberger et al., 2003, 2006) it seem to be clear
that secondary brain damage and its sequel, i.e.
contusion expansion, is a linearly ongoing process
which starts immediately after TBI and levels off
approximately 12 h after injury. The consequence of
this finding is, that the later a therapeutic interven-
tion starts the less effective it will be. In the case of
craniectomy this may indicate that delayed initiation
of decompression, i.e. at a time point when brain
edema and ICP already developed, may not have
any beneficial potential. In the contrary, under cir-
cumstances of massive brain edema formation and
high ICP craniectomy, which among others lowers
vascular–interstitial pressure gradients, may also
have adverse effects, e.g. an increased risk of inter-
nal or external herniation by enhanced vasogenic
edema formation with subsequent cerebral ischemia
or detrimental brain stem compression, respectively.
Evaluating the protocols of the early experimen-
tal studies investigating decompression craniectomy
following TBI (Moody et al., 1968; Cooper et al.,
1979; Gaab et al., 1979; Burkert and Plaumann,
1989; Rinaldi et al., 1990) it becomes clear that,
most logically, a clinically relevant treatment sce-
nario was chosen. In parallel to the situation in
patients, animals were decompressed several hours
after TBI when pathologically elevated ICP already
developed. In view of our current understanding
of the gradual and irreversible development of
secondary brain damage, the point in time of de-
compression in these studies may well have been
chosen too late, i.e. at a time when the vicious
circle of secondary brain damage already started
and sequels became irreversible.
Accordingly, the experimental data available up
to the beginning of the new millennium did not
necessarily allow to draw definitive conclusions
on the therapeutic potential of decompression
craniectomy after TBI.
Experimental results on early decompression after
TBI
Based on the above-mentioned pathophysiological
understanding of the gradual and irreversible pro-
gression of parenchymal brain damage following
TBI, we designed a series of experiments in a
kinetic experimental TBI model, CCI, investigating
the maximal treatment potential of decompression
craniectomy. This was achieved by comparing one
group of animals which was allowed to develop
post-traumatic ICP, while in a second group of
animals pathological ICP increase was completely
prevented by a large parietal craniectomy (Fig. 2).
Under these circumstances the contusion volume of
the not craniectomised animals increased by almost
60% within 24 h as compared with the primary
damage inflicted by the mechanical impact imme-
diately (15 min) after trauma (Fig. 3). In craniecto-
mised animals the secondary contusion expansion
was completely blunted thereby demonstrating that
decompression craniectomy completely and effec-
tively prevents the secondary loss of brain tissue
following TBI.
Since two studies from the late 70 s using cold
lesion-induced brain injury reported up to seven-
fold increase of brain edema volume 8 h following
decompressive craniectomy in dogs (Cooper et al.,
1979; Gaab et al., 1979), we performed wet–dry
ratio measurements for the determination of brain
edema in our kinetic TBI model. Brain water con-
tent in mice increased dramatically by almost 3%
(from 78 to 81%) following CCI. In craniectomised
animals surgical decompression prevented brain
edema formation by more than 50%. Together
with the lack of secondary contusion expansion
these findings indicate that approximately 50% of
post-traumatic brain edema formation following
CCI are primary, i.e. caused by the initial mechan-
ical impact, while only 50% are secondary and are
therefore amenable to therapeutic interventions.
Finally, we were interested to evaluate the thera-
peutic window of decompression craniectomy.
Control animals were traumatized but not craniec-
tomised, while treated animals were craniectomised
immediately after TBI or with a delay of 1, 3, or 8 h.
Contusion volume in not craniectomised mice was
over 30 mm3 24 h after TBI indicating progressive
secondary brain damage, while animals craniecto-
mised immediately after trauma had contusion vol-
umes of
20 mm3, i.e. comparable with primary
contusions. If craniectomy was delayed by 1 or 3 h
a significant proportion of secondary brain damage
was prevented, while decompression 8 h after TBI
showed a marginal but statistically not significant
 397

30 Mean +/- SD; n=3 p<0.001 vs. control

p<0.001 vs. 0 h (baseline)
25

20 Control
ICP (mmHg)

15

10

5
Craniectomy

0
0 1 2 3 6 12 24
Time after trauma (h)

Fig. 2. Intracranial pressure of craniectomised and not craniectomised mice after experimental TBI (CCI). While intracranial pressure
increases significantly in not decompressed animals (control; open circles; n ¼ 6), no significant ICP increase is observed in animals
receiving a large parietal craniectomy immediately after TBI (closed circles; n ¼ 6). Adapted with permission from Zweckberger et al.
(2003).

45
Mean +/- SD; n=7 Closed skull
40 p<0.05 vs. Control 24h Craniectomized

35
Contusion volume (mm3)

30

25

20

15

10

0
15 min 24 h 24 h

Fig. 3. Effect of decompression craniectomy on the secondary expansion of a traumatic cortical contusion. In not craniectomised mice
(gray bars) significant secondary expansion of a traumatic contusion (streaked area) is observed (compare also with Fig. 1). In contrast,
no secondary lesion expansion is present in craniectomised mice indicating that craniectomy completely prevents secondary brain
damage. Adapted with permission from Zweckberger et al. (2006).

effect. These results clearly indicate that the patho-
physiology of secondary lesion expansion follows a
two step paradigm: a slow starting phase where
treatment is effective which is ultimately followed
by a self amplifying final phase when even the most
effective treatment options, e.g. craniectomy, are
not or only marginally effective.
Taken together, the recent literature indicates
that the timing of craniectomy is of major signifi-
cance for its therapeutic benefit: performed in the
398
presence of massive brain swelling decompression
may be detrimental, but when performed early
enough craniectomy may completely prevent seco-
ndary brain damage (including secondary brain
edema formation).
Although we are aware of the caution which
should be taken when extrapolating experimental
data to the human situation one has also to take
into consideration that pure physical phenomena
like pressure changes in a closed cavity may not be
as different between species as compared with other
parameters, e.g. gene expression or the initiation of
intracellular signal transduction cascades. Accord-
ingly, our results suggest that early decompressive
craniectomy may have a clinical potential and
should clearly be favored in comparison with de-
layed surgical decompression, which may even be
detrimental. In view of the present data we conclude
that timing seems to be of utmost importance and
that decompression has to be performed as early as
possible in order to exert its therapeutic potential.
Abbreviations
CCI controlled cortical impact
ICP intracranial pressure
TBI traumatic brain injury
Acknowledgments
The excellent work of Klaus Zweckberger who
performed most experiments investigating the role
of decompression craniectomy on secondary brain
damage after TBI is highly appreciated. Special
thanks also to Dr. Alexander Baethmann, who
was always a source of inspiration and help during
this project.
References
Aarabi, B., Hesdorffer, D.C., Ahn, E.S., Aresco, C., Scalea,
T.M. and Eisenberg, H.M. (2006) Outcome following de-
compressive craniectomy for malignant swelling due to
severe head injury. J. Neurosurg., 104: 469–479.
Berger, S., Schwarz, M. and Huth, R. (2002) Hypertonic saline
solution and decompressive craniectomy for treatment of
intracranial hypertension in pediatric severe traumatic brain
injury. J. Trauma, 53: 558–563.
Burkert, W. and Plaumann, H. (1989) The value of large pres-
sure-relieving trepanation in treatment of refractory brain
edema. Animal experiment studies, initial clinical results.
Zentralbl. Neurochir., 50: 106–108.
Clark, K., Nash, T.M. and Hutchison, G.C. (1968) The failure
of circumferential craniotomy in acute traumatic cerebral
swelling. J. Neurosurg., 29: 367–371.
Cooper, P.R., Hagler, H., Clark, W.K. and Barnett, P. (1979)
Enhancement of experimental cerebral edema after decom-
pressive craniectomy: implications for the management of
severe head injuries. Neurosurgery, 4: 296–300.
Cushing, H. (1905) The establishment of cerebral hernia as a
decompressive measure for inaccessible brain tumor: with the
description of intramuscular methods of making the bone
defect in temporal and occipital regions. Surg. Gynecol.
Obstet., 1: 297–314.
Dam, H.P., Sizun, J., Person, H. and Besson, G. (1996) The
place of decompressive surgery in the treatment of uncon-
trollable post-traumatic intracranial hypertension in children.
Childs Nerv. Syst., 12: 270–275.
De Luca, G.P., Volpin, L., Fornezza, U., Cervellini, P.,
Zanusso, M., Casentini, L., Curri, D., Piacentino, M.,
Bozzato, G. and Colombo, F. (2000) The role of decompres-
sive craniectomy in the treatment of uncontrollable post-
traumatic intracranial hypertension. Acta Neurochir. Suppl.,
76: 401–404.
Diemath, H.E. (1966) The value of bitemporal relief surgery in
severe closed skull brain trauma. Wien. Med. Wochenschr.,
116: 1043–1044.
Fraser, J.F. and Hartl, R. (2005) Decompressive craniectomy as
a therapeutic option in the treatment of hemispheric stroke.
Curr. Atheroscler. Rep., 7: 296–304.
Gaab, M., Knoblich, O.E., Fuhrmeister, U., Pflughaupt,
K.W. and Dietrich, K. (1979) Comparison of the effects of
surgical decompression and resection of local edema in the
therapy of experimental brain trauma. Investigation of ICP,
EEG and cerebral metabolism in cats. Childs Brain, 5:
484–498.
Gerl, A. and Tavan, S. (1980) Bilateral craniectomy in the
treatment of severe traumatic brain edema. Zentralbl.
Neurochir., 41: 125–138.
Guerra, W.K., Gaab, M.R., Dietz, H., Mueller, J.U., Piek, J.
and Fritsch, M.J. (1999) Surgical decompression for trau-
matic brain swelling: indications and results. J. Neurosurg.,
90: 187–196.
Hartl, R. and Ougorets, I. (2004) Critical care of neurotrauma.
Curr. Neurol. Neurosci. Rep., 4: 481–488.
Josan, V.A. and Sgouros, S. (2006) Early decompressive
craniectomy may be effective in the treatment of refractory
intracranial hypertension after traumatic brain injury. Childs
Nerv. Syst., 22: 1268–1274.

Kerr, F.W. (1968) Radical decompression and dural grafting in
severe cerebral edema. Mayo Clin. Proc., 43: 852–864.
Kjellberg, R.N. and Prieto Jr., A. (1971) Bifrontal decompres-
sive craniotomy for massive cerebral edema. J. Neurosurg.,
34: 488–493.
Kocher, T. (1901) Die Therapie des Hirndruckes. In: Hölder
A. (Ed.), Hirnerschütterung, Hirndruck und chirurgische
Eingriffe bei Hirnkrankheiten. Hölder Verlag, Wien, pp.
262–266.
Kontopoulos, V., Foroglou, N., Patsalas, J., Magras, J., Foro-
glou, G., Yiannakou-Pephtoulidou, M., Sofianos, E., Anast-
assiou, H. and Tsaoussi, G. (2002) Decompressive
craniectomy for the management of patients with refractory
hypertension: should it be reconsidered? Acta Neurochir.
(Wien.), 144: 791–796.
Kunze, E., Meixensberger, J., Janka, M., Sorensen, N. and
Roosen, K. (1998) Decompressive craniectomy in patients
with uncontrollable intracranial hypertension. Acta Neuro-
chir. Suppl. (Wien.), 71: 16–18.
Maas, A.I., Dearden, M., Teasdale, G.M., Braakman, R., Co-
hadon, F., Iannotti, F., Karimi, A., Lapierre, F., Murray, G.,
Ohman, J., Persson, L., Servadei, F., Stocchetti, N. and Un-
terberg, A. (1997) EBIC-guidelines for management of severe
head injury in adults. European Brain Injury Consortium.
Acta Neurochir. (Wien.), 139: 286–294.
Makino, H. and Yamaura, A. (1979) Assessment of outcome
following large decompressive craniectomy in management of
serious cerebral contusion. A review of 207 cases. Acta
Neurochir. Suppl. (Wien.), 28: 193–194.
Meier, U. and Grawe, A. (2003) The importance of decom-
pressive craniectomy for the management of severe head in-
juries. Acta Neurochir. Suppl., 86: 367–371.
Messing-Junger, A.M., Marzog, J., Wobker, G., Sabel, M. and
Bock, W.J. (2003) Decompressive craniectomy in severe
brain injury. Zentralbl. Neurochir., 64: 171–177.
Moody, R.A., Ruamsuke, S. and Mullan, S.F. (1968) An eval-
uation of decompression in experimental head injury. J.
Neurosurg., 29: 586–590.
Munch, E., Horn, P., Schurer, L., Piepgras, A., Paul, T. and
Schmiedek, P. (2000) Management of severe traumatic brain
injury by decompressive craniectomy. Neurosurgery, 47:
315–322.
Murray, G.D., Teasdale, G.M., Braakman, R., Cohadon, F.,
Dearden, M., Iannotti, F., Karimi, A., Lapierre, F., Maas,
A., Ohman, J., Persson, L., Servadei, F., Stocchetti, N.,
Trojanowski, T. and Unterberg, A. (1999) The European
Brain Injury Consortium survey of head injuries. Acta Ne-
urochir. (Wien.), 141: 223–236.
Nortje, J. and Menon, D.K. (2004) Traumatic brain injury:
physiology, mechanisms, and outcome. Curr. Opin. Neurol.,
17: 711–718.
Ogawa, M., Minami, T., Katsurada, K. and Sugimoto, T.
(1974) Evaluation of external cranial decompression for
traumatic acute brain swelling. Med. J. Osaka Univ., 25:
73–78.
399
Pereira, W.C., Neves, V.J. and Rodrigues, Y. (1977) Bifrontal
decompressive craniotomy as the treatment for severe cere-
bral edema. Arq. Neuropsiquiatr., 35: 99–111.
Polin, R.S., Shaffrey, M.E., Bogaev, C.A., Tisdale, N.,
Germanson, T., Bocchicchio, B. and Jane, J.A. (1997)
Decompressive bifrontal craniectomy in the treatment of se-
vere refractory posttraumatic cerebral edema. Neurosurgery,
41: 84–92.
Ransohoff, J. and Benjamin, V. (1971) Hemicraniectomy in the
treatment of acute subdural haematoma. J. Neurol. Ne-
urosurg. Psychiatry, 34: 106.
Rinaldi, A., Mangiola, A., Anile, C., Maira, G., Amante, P.
and Ferraresi, A. (1990) Hemodynamic effects of decom-
pressive craniectomy in cold induced brain oedema. Acta
Neurochir. Suppl. (Wien.), 51: 394–396.
Ruf, B., Heckmann, M., Schroth, I., Hugens-Penzel, M., Reiss,
I., Borkhardt, A., Gortner, L. and Jodicke, A. (2003) Early
decompressive craniectomy and duraplasty for refractory in-
tracranial hypertension in children: results of a pilot study.
Crit. Care, 7: R133–R138.
Sahuquillo, J. and Arikan, F. (2006) Decompressive craniec-
tomy for the treatment of refractory high intracranial pres-
sure in traumatic brain injury. Cochrane Database Syst.
Rev., CD003983.
Schneider, G.H., Bardt, T., Lanksch, W.R. and Unterberg, A.
(2002) Decompressive craniectomy following traumatic brain
injury: ICP, CPP and neurological outcome. Acta Neurochir.
Suppl., 81: 77–79.
Schwab, S., Steiner, T., Aschoff, A., Schwarz, S., Steiner, H.H.,
Jansen, O. and Hacke, W. (1998) Early hemicraniectomy in
patients with complete middle cerebral artery infarction.
Stroke, 29: 1888–1893.
Soukiasian, H.J., Hui, T., Avital, I., Eby, J., Thompson, R.,
Kleisli, T., Margulies, D.R. and Cunneen, S. (2002) Decom-
pressive craniectomy in trauma patients with severe brain
injury. Am. Surg., 68: 1066–1071.
Taylor, A., Butt, W., Rosenfeld, J., Shann, F., Ditchfield, M.,
Lewis, E., Klug, G., Wallace, D., Henning, R. and Tibballs, J.
(2001) A randomized trial of very early decompressive craniec-
tomy in children with traumatic brain injury and sustained
intracranial hypertension. Childs Nerv. Syst., 17: 154–162.
The Brain Trauma Foundation. The American Association of
Neurological Surgeons. The Joint Section on Neurotrauma
and Critical Care. (2000) Critical pathway for the treatment
of established intracranial hypertension. J. Neurotrauma, 17:
537–538.
Venes, J.L. and Collins, W.F. (1975) Bifrontal decompressive
craniectomy in the management of head trauma. J.
Neurosurg., 42: 429–433.
Yamaura, A., Uemura, K. and Makino, H. (1979) Large
decompressive craniectomy in management of severe cere-
bral contusion. A review of 207 cases. Neurol. Med. Chir.
(Tokyo), 19: 717–728.
Yoo, D.S., Kim, D.S., Cho, K.S., Huh, P.W., Park, C.K. and
Kang, J.K. (1999) Ventricular pressure monitoring during
400
bilateral decompression with dural expansion. J. Neurosurg.,
91: 953–959.
Ziai, W.C., Port, J.D., Cowan, J.A., Garonzik, I.M., Bhardwaj,
A. and Rigamonti, D. (2003) Decompressive craniectomy for
intractable cerebral edema: experience of a single center. J.
Neurosurg. Anesthesiol., 15: 25–32.
Zweckberger, K., Eros, C., Zimmermann, R., Kim, S.W., En-
gel, D. and Plesnila, N. (2006) Effect of early and delayed
decompressive craniectomy on secondary brain damage after
controlled cortical impact in mice. J. Neurotrauma, 23:
1083–1093.
Zweckberger, K., Stoffel, M., Baethmann, A. and Plesnila,
N. (2003) Effect of decompression craniotomy on increase
of contusion volume and functional outcome after con-
trolled cortical impact in mice. J. Neurotrauma, 20:
1307–1314.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 29

Novel neuroproteomic approaches to studying

traumatic brain injury

Andrew K. Ottens1,2,, Firas H. Kobeissy1,2, Brian F. Fuller1, Ming Chen Liu2,

Monika W. Oli3, Ronald L. Hayes2,3 and Kevin K.W. Wang1,2,3

1
Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the
University of Florida, PO Box 100256, Gainesville, FL 32610, USA
2
Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight Brain Institute of the University
of Florida, PO Box 100244, Gainesville, FL 32610, USA
3
Banyan Biomarkers, Inc., 12085 Research Dr., Alachua, FL 32615, USA

Abstract: Neuroproteomics entails wide-scope study of the nervous system proteome in both its content
and dynamics. The field employs high-end analytical mass spectrometry and novel high-throughput
antibody approaches to characterize as many proteins as possible. The most common application has been
differential analysis to identify a limited set of highly dynamic proteins associated with injury, disease, or
other altered states of the nervous system. Traumatic brain injury (TBI) is an important neurological
condition where neuroproteomics has revolutionized the characterization of protein dynamics, leading to
a greater understanding of post-injury biochemistry. Further, proteins of altered abundance or post-
translational modifications identified by neuroproteomic studies are candidate biochemical markers of TBI.
This chapter explores the use of neuroproteomics in the study of TBI and the validation of identified
putative biomarkers for subsequent clinical translation into novel injury diagnostics.

Keywords: proteomics; neuroproteomics; brain injury; neurotrauma; TBI

Introduction to TBI is estimated that between 1.6 and 3.8 million

Traumatic brain injury (TBI) is neurotrauma
caused by a mechanical force applied to the head
(Wang et al., 2004). There are approximately 1.4–2
million TBI incidents annually in the United
States, resulting in 1.1 million emergency depart-
ment visits, 235,000 hospitalizations, 90,000 left
with long-term disabilities, and 50,000 deaths
(Langlois et al., 2006b; Ragnarsson, 2006). Yet
the true burden of TBI may not be reflected in
these numbers as many TBI incidents go unre-
ported or are not classified as TBI. For example it
Corresponding author. Tel.: +1-352-392-8060; Fax: +1-352-
392-2579; E-mail: aottens@mbi.ufl.edu
sports-related TBI incidents actually occur in the
US as compared with the 300,000 reported. TBI
costs the American economy an estimated $60 bil-
lion per year in medical expenses and lost produc-
tivity (Finkelstein et al., 2006). All told, TBI
represents a major public health problem with
significant societal and economic consequences
(Sosin et al., 1995; Wang et al., 2004; Yi and
Hazell, 2006). Despite this, there remains no ap-
proved therapy for TBI (Narayan et al., 2002;
Wieloch and Nikolich, 2006).
TBI results from a number of etiologies (Fig. 1)
primarily affecting male (78.8% of incidents) ado-
lescents and young adults (31.7% of incidents are
among 15- to 24-year olds). In fact, TBI is

DOI: 10.1016/S0079-6123(06)61029-7 401

402
Fig. 1. Breakdown of reported TBI etiologies in the United States between 1995 and 2001 (Langlois et al., 2006a).
considered the leading cause of death and disabil-
ity in children and young adults. A sharp rise in
TBI incidence among, typically young, American
combat casualties recently added to this equation
(Warden, 2006). Of Iraq and Afghanistan conflict
casualties evacuated to the Walter Reed Army
Medical Center as of January 2006, 28% (1700
cases) had TBI. An unfortunate benefit to in-
creased incidents of military head trauma has been
a recent boost in funded research for development
of new neurotrauma diagnostics and therapies.
This is an important opportunity as military
medicine has historically driven medical advance-
ments in trauma, while progress in head injury
treatment has been slow to proceed (Narayan
et al., 2002).
A major limitation is that TBI continues to be
difficult to assess clinically, even with modern im-
aging techniques such as magnetic resonance im-
aging (MRI) and computer tomography (CT),
which are limited by sensitivity, specificity and
availability (Ottens et al., 2006). These diagnostics
often do not provide reliable prognoses, particu-
larly for mild TBI patients that account for 80% of
those who suffer lifelong impairments (Alexander,
1995). There is clearly a need for more defined TBI
diagnostics for patient management and progno-
sis. Likewise, these tools are needed as quantifiable
outcome measures for development of new thera-
pies, a recognized deficiency of past TBI thera-
peutics studies (Denslow et al., 2003).
To this end, the present strategy is to develop
TBI diagnostics ahead of new therapy trials. For
this purpose, it is important to have accurate
models of TBI that mimic human sequelae and
outcomes (Narayan et al., 2002; Cernak, 2005).
Several experimental TBI models have been intro-
duced to investigate the pathobiology and mole-
cular dynamics of brain injury. Generally, a
mechanical force is used to impact or torque the
brain. The primary injury is comprised of contu-
sion and hemorrhage, characterized by rapid cell
death followed by diffuse axonal injury. Conse-
quently, brain function is rapidly disrupted at the
site of injury and in distal regions interconnected
through white matter tracts, leading to likely def-
icits in higher cognitive and vital sensory-motor
functions (Wieloch and Nikolich, 2006).
Secondary injury occurs in proximal brain re-
gions marked by massive edema, intracranial
hemorrhages through blood-brain barrier (BBB)
disruption, and further cell death and axonal in-
jury. Those biochemical events are mediated by
neurotoxic increases in excitatory amino acids
(EAAs), ischemic factors (free radicals/oxidative
damage), proteolysis, and an inflammatory re-
sponse mediated by microglia and astroglia acti-
vation and proliferation (McIntosh et al., 1998;
Wieloch and Nikolich, 2006). A major causal fac-
tor for secondary degeneration is the disruption of
intracellular calcium homeostasis following EAA
glutamate overflow into extra-synaptic regions

(glutamate excitotoxicity), and the additional
glutamate release induced by a calcium influx.
These biochemical events can lead to prolonged
cell membrane depolarization, ATP depletion and
subsequent ionic imbalance, manifested clinical as
an increase in intracranial pressure (ICP) (Kupina
et al., 2003; Yi and Hazell, 2006). Furthermore,
secondary insults continue for days, potentially
worsening a patient’s condition with unpredicted
results (coma or death).
Several studies have demonstrated that acti-
vated cysteine proteases are major intracellular ef-
fectors in neurotrauma pathology. These proteases
act on a number of substrates including cytoske-
letal, cytosolic, synaptic and membrane proteins,
which upon proteolysis yield signature fragments
indicative of neuronal cell death dynamics (Farkas
et al., 2005). Necrotic oncosis is mediated by cal-
pains activated by the inrush of calcium (Kupina
et al., 2003). Apoptosis is activated via more in-
volved caspase pathways during secondary injury
(Wennersten et al., 2003). Intrinsic and extrinsic
apoptotic effectors invoke caspase-8 or -9 activa-
tion that activates the executioner caspases (pri-
marily caspase-3) (Rink et al., 1995; Ottens et al.,
2006). Other molecular events occur tangential to
cell death. Shifts in pro-survival and pro-death
signals (Rink et al., 1995; Shimamura et al., 2005)
can result in cells that survive but are susceptible
to degeneration or later death. Another feature
of TBI is the marked long-term accumulation of
proteins in damaged cells. Some of these proteins,
such as synuclein (Syn), amyloid-b (Ab), and
neurofilament (NF), are major pathological com-
ponents of neurodegenerative diseases such as
Alzheimer’s and Parkinson’s disease (Smith
et al., 2003).
A detailed understanding of molecular dynamics
following TBI can be of high value to elucidate
injury mechanism for diagnostics and treatments
(Shimamura et al., 2005). Recently, novel ne-
uroproteomic approaches have been utilized to
expand our understanding of TBI molecular mech-
anisms. Neuroproteomics is a new field that pro-
vides a wide-scope, more global analysis of protein
dynamics to investigate differential alteration in
protein levels, modifications and functions post
neurotrauma (Jenkins et al., 2002; Denslow et al.,
403
2003; Kobeissy et al., 2006). A number of papers
have been published recently discussing the appli-
cation of neuroproteomics in the field of neuro-
trauma (Buonocore et al., 1999; Jenkins et al.,
2002; Satchell et al., 2003; Conti et al., 2004;
Haskins et al., 2005; Kobeissy et al., 2006;
Kochanek et al., 2006). One major application of
neuroproteomics lies in its ability to identify bio-
markers that are of great value in providing in-
sights into injury severity, patient management,
and outcome. In this chapter we will explore the
application of neuroproteomics to TBI, discussing
the relevant preclinical animal models, the mass
spectrometry and antibody based neuroproteomic
methods employed, and reviewing validation and
translation of neuroproteomic data into viable
clinical biomarkers of TBI.
Modeling TBI in animals
Traumatic brain injury (TBI) has been modeled in
animals to replicate clinical trauma in a controlled
experimental setting. A variety of animal models
are available, and careful selection is required to
determine which will work best with the scientific
question at hand. A mechanical force is utilized to
inflict injury, which must be controlled, reproduc-
ible and quantifiable. Invariably, the intensity of
the mechanical force should dictate the severity of
injury. The injury should also be comparable to
the human injury, modeled with a direct correla-
tion between injury and outcome.
The following section will cover the basic mecha-
nisms, advantages and limitations of current TBI
animal models. A summary of the models can be
found in Fig. 2, which is based on a classification
scheme from Ommaya (1995; see also Ommaya et
al., 2002) with mechanical paradigms added by
Cernak (2005) in his recent review.
Static and dynamic brain trauma differs from
each other based on amplitude, duration, velocity
and acceleration. Static brain trauma is studied
utilizing models that have a defined amplitude and
duration. A simple example of a static injury
model would be to crush a region of the brain
using a mechanical force (e.g., a vice) for a defined
period of time. If velocity and acceleration were
404
Fig. 2. Schematic representation of in vivo experimental models of traumatic brain injury.
factored into the static model then it would be-
come dynamic.
Dynamic brain injury is the result of a mechan-
ical force with a well-defined amplitude, duration,
velocity, and acceleration. Dynamic brain injury
can be sub-classified into indirect and direct inju-
ries. Indirect dynamic brain injury occurs from
blast, explosive phenomena in which the entire
body is affected by the kinetic energy released.
Recent studies have noted that exposure to blast
waves can cause peripheral blast trauma to the
brain without inflicting external cranial trauma
(Saljo et al., 2000; Cernak et al., 2001). Animal
models consist of exposing a test subject, often a
rodent, to a blast wave with an over-pressurization
range of 154–340 kPa. Importantly, models of ro-
dent blast exposure have produced similar out-
comes to studies of human blast injury patients
(Cernak et al., 1999, 2000).
Direct dynamic brain injury can be further clas-
sified into acceleration and impact models. Accel-
eration brain injury models study the effects of
brain movement within the skull without direct
contact with the brain tissue. Rapid head rotation
is most commonly used to simulate the effects of
forcing the brain into the side of the skull. In this
case special attention has been paid to the rela-
tionship between brain mass and acceleration in
order to regulate the inertial effects that determine
the extent of damage. These models have utilized
miniature swine, rabbits, rats and primates as test
subjects. Unconstrained head motion is standard
in acceleration models; however, it has been noted
that unconstrained head motion can lead to sig-
nificant variability in the outcome. Non-human
primates and miniature swine acceleration models
have been observed to produce injuries that are
strikingly similar to those found in human patients
(Povlishock et al., 1994). Often the result is profuse
axial injury that causes the test subject to go into a
prolonged coma. Rotational acceleration models
are disadvantaged by the excessive cost involved in
the equipment and use of large animals, as well as
the lack of adequate outcome measures of torque
induced brain injury.
Impact dynamic head injury models reproduce
high energy concussive or penetrating injuries that
are either open or closed headed. Concussive
closed head injury models are utilized in order to
replicate human concussive and diffuse brain in-
jury. There have been two controlled concussion
(CC) models put forth that utilize cats and rats as

test subjects (Tornheim and McLaurin, 1981;
Tornheim et al., 1990; Goldman et al., 1991). In
both, a force is applied to either the coronal struc-
ture (cats) or the skull midline (rat). The cat model
has been found to be suitable for the study of cere-
bral contusion and resulting edema. The downside
to the cat model is that the inflicted injury usually
results in hemorrhage, which limits the ability
of the researcher to use this model for the study
of molecular or cellular mechanisms related exclu-
sively to the impact. The rat model has been a solid
model for the study of mild to moderate injury
depending on the amount of force used. It also
does not cause skull fracture or significant hem-
orrhage, and is comparable to similar injuries
found in human TBI. These factors, as well as the
low cost in using rats, make this model a prime
method for the study of both TBI mechanisms and
therapy.
Open head contusion injuries apply a mechanical
force directly to the brain through a hole in the
skull (removed by craniotomy). An example is the
fluid percussion injury (FPI) model where a fluid
pressure pulse is delivered to the revealed dura to
cause brain deformation (Erb and Povlishock,
1988; Faden et al., 1989, 2003; McIntosh et al.,
1994; Conti et al., 1998; Albensi et al., 2000). FPI is
a popular model for use with many different spe-
cies (rats, mice, cats, pigs, rabbits, dogs, and
sheep). The model has been adapted to apply ei-
ther a central force to the vertex at the midline
between bregma and lambda, or a lateral force over
the left parietal bone between bregma and lambda
to produce different desired outcomes. Craniotomy
position is critical for impairment determination
and reproducibility. While both central and lateral
FPI have comparable pathobiology, it has been
observed that central FPI causes death due to au-
tonomic dysfunctions at hypothalamic levels. Thus,
lateral FPI can reproduce an injury with clinical
characteristics of contusion without skull fracture,
but suffers from increased morbidity due to dis-
proportionate brainstem injury and limited control
of injury magnitude.
Controlled cortical impact (CCI) is a more re-
cent open-head concussive model that applies a
mechanical force to the intact dura by means of a
pneumatic piston (Dixon et al., 1991; Gennarelli,
405
1994; Povlishock et al., 1994), which allows for
finer control of the mechanical force than with
FPI. The use of pneumatics however practically
limits this model for use with rodents such as fer-
rets, rats and mice. CCI is also considered to de-
liver a more focused injury than FPI. The model
successfully replicates the clinical pathobiology of
brain injury with skull deformation and cortical
compression. CCI is widely used for the study of
molecular and cellular mechanisms related to TBI
and is a useful tool in the development of novel
therapies.
The weight drop model is another widely uti-
lized direct impact model, simulating an impact
acceleration injury by dropping a brass weight
onto a fixed target up against the surgically ex-
posed brain of a rat or mouse. In this model there
is a direct correlation between the severity of in-
jury and the weight size and height from which it is
dropped (Marmarou et al., 1994; Piper et al., 1996;
De Mulder et al., 2000). Historically, this model
has suffered from a few technical limitations. Al-
though the weigh is dropped though a Plexiglas
tube, there is still a small amount of lateral mo-
tion. There is also the likelihood of a second im-
pact as the weight rebounds, since it is dropped
freely. More recent improvements to the weight
drop model have utilized a laser guided, air-driven,
high-velocity impactor, which is more controlled
and reproducible then earlier setups (Cernak et al.,
2004). The model is able to replicate the biochem-
ical and neurological aspects of TBI in rats and
mice as observed in clinical cases.
High-velocity penetrating injury models have
been designed in order to study the effects of pro-
jectile injury to the brain. Initially models were
designed for use in cats and sheep (Carey et al.,
1989, 1990; Finnie, 1993; Carey, 1995). In each
case, a small projectile (relative to the animal’s
size) was used to inflict a head wound on the
anesthetized subject. Each of the models produced
sever tissue damage as well as brain displacement.
While these models are both useful in the histo-
pathological features of projectile wounds, they
are not practical for molecular studies due to the
use of large animal test subjects. Williams et al.
(2005) recently developed a rat penetrating ballis-
tic brain injury (PBBI) model that mimics aspects
406
of a high energy bullet wound as demonstrated by
histopathological and physiological studies.
Through the insertion of a rapidly inflatable
probe, they are able to reproduce the permanent
injury track left by a bullet, as well as the temporal
cavity generated by the dissipating energy from the
penetrating projectile.
In all, there are a number of preclinical animal
models that adequately reproduce facets of clinical
TBI, though the pathobiology of each model will
differ somewhat in mirroring the reality of human
TBI incidents. In the end, these models are useful
for generating preclinical tissue samples in a repro-
ducible, controlled manner for use in neuro-
proteomic studies reviewed in the following two
sections.
Neuroproteomic analysis with mass spectrometry in
TBI
The introduction of peptide analysis by mass
spectrometry in combination with bioinformatics
for data processing has revolutionized the field
of proteomics. Figure 3 illustrates how the field
of proteomics (and genomics) has been driven
by technology. By definition proteomics entails
wide-scale protein analysis made possible by
high-throughput methods and instrumentation to
approach characterization of the complete pro-
teome, or more feasibly a distinct subset such as
the neuroproteome, the protein content of the
central nervous system (CNS). What is also ap-
parent in Fig. 3 is that further, significant techno-
logical advancements are poised to occur in
proteomics, which will continue to revolutionize
the field, bringing us closer to complete ne-
uroproteome characterization.
Though mass spectrometers have been used to
characterize macro biomolecules (e.g., polypep-
tides) since the early 1990s (Kaufmann, 1995;
Nguyen et al., 1995), it has only been in the last
few years that instrument manufacturers have be-
gun to cater to proteomic applications. Since, the
proteomics market has swelled with a discernable
rise in products and software and correlated pub-
lications. Despite this, investigators remain limited
in their ability to characterize a given proteome.
Available technology has become quite good for
analyzing relative abundance change (differential
analysis) among moderately or highly expressed
proteins; however, extremes in dynamic range (1010
orders) confound complete proteome coverage.
Further, proteins encoded only by some 30,000
genes transcribe to multiple isoforms and can be
post-translationally modified by some 400 different
processes (Morrison et al., 2002; Patton, 2002)
producing a large array of functionally distinct
biomolecules. Though progress has been made in
characterizing the more common of these proc-
esses, there remains no adequate means to capture
the fully functionalized state of proteins, let alone
in a wide-scale, high-throughput paradigm. To
complicate matters further, any given proteome is
highly dynamic, able to shift in its protein compli-
ment in a matter of minutes to hours. As yet pro-
teomic experiments have been limited to temporal
snap shots due to resource constraints.
In light of the challenges, progress has been
made in characterizing neuroproteome dynamics
initiated by TBI — the primary motivators being
to identify biochemical markers for development
of diagnostics and potential therapeutic avenues.
Prominent protein dynamics are more easily trans-
lated into clinical diagnostics; therefore, the capa-
bilities of current proteomic technologies have
coalesced well with the monitoring and prognostic
medical needs in neurotrauma (Denslow et al.,
2003; Wang et al., 2005).
A general platform has emerged for performing
mass spectrometry based differential proteomics
for TBI research (Fig. 4), though the details for
each step do vary. The first important considera-
tion is, what adjustments must be made to the
proteomics experiment to suit the sample type
(e.g., neuronal culture, brain tissue, cerebrospinal
fluid (CSF)). Confounding factors introduced by
the mechanism of insult and sample processing
must be controlled, generally by including
matched naı̈ve or sham controls. Biological vari-
ability also will increase with sample heterogeneity
(Molloy et al., 2003). Ideally, experiments should
be processed multiple times with separate samples,
though this is often too costly; thus, subsequent
validation of individual protein dynamics is often
a more manageable scenario.
 407

Fig. 3. Timeline and publication history of proteomics vs. genomics. PubMed entries (searched on Oct. 4, 2006) for genomic studies
(search phrase ‘‘genome or genomics’’) and proteomic studies (search phrase ‘‘proteome or proteomics or 2D-PAGE’’) are normalized
against the number of all entries per publication year. The graph is overlaid by key technological advancements that have shaped each
field.

After extraction, proteins must be resolved from
one another. The most common approach to pro-
teome separation in general and in TBI research
(Buonocore et al., 1999; Jenkins et al., 2002;
Satchell et al., 2003; Conti et al., 2004; Kochanek
et al., 2006) has been two-dimensional polyacryl-
amide gel electrophoresis (2D-PAGE). Newer
multidimensional separation approaches have
been introduced recently, overcoming some limi-
tation of 2D-PAGE (Ottens et al., 2006). For
example, we have opted to use ion-exchange liq-
uid chromatography in tandem with gel elect-
rophoresis, in an approached called CAX-PAGE,
to enhance differential comparison without hav-
ing to mix samples and to extend the mass range
(Ottens et al., 2005; Kobeissy et al., 2006). How-
ever, 2D-PAGE remains the most common ap-
proach, having been established over 30 years
(Beranova-Giorgianni, 2003). 2D-PAGE involves
resolving proteins by their isoelectric point (pI)
first and then by relative molecular mass (Mr). The
peak capacity for each dimension (between 50 and
100) is then multiplied to provide separation of
roughly 5000 proteins (gel spots in Fig. 4). The
neuroproteome contains more than this number of
proteins though, and does not resolve evenly
across the separation space; thus, it is important
to have additional dimensions of separation, as
afforded by peptide chromatography and mass
spectrometry.
Mass spectrometry provides accurate and pre-
cise mass-to-charge (m/z) information that can be
aligned against an established, species specific,
protein database(s) that has been processed in
silico into sequence specific peptides. A general
caveat to mass spectrometry is that proteins must
be fragmented into polypeptides (generally from 7
to 20 amino acids) in order to permit accurate
m/z assignment, though recent progress has been
made to alleviate this limitation (Kelleher, 2004;
Bogdanov and Smith, 2005). Nevertheless, mass
spectrometry provides a degree of selectivity as yet
unrivalled for the identification of proteins on a
large scale (technologies such as Edmund degrada-
tion or immunological assays are also selective,
but on a more limited scale). Time-of-flight mass
408

Fig. 4. Neuroproteomic analysis scheme using mass spectrometry. The general platform for neuroproteomic research is outlined in six
steps: (1) extract proteins from the biological sample and matched control; (2) resolve proteins via multidimensional separations; (3)
isolate target (e.g., differential) proteins and digest into tryptic fragments; (4) resolve peptides by chromatography (used with tandem
mass spectrometry); (5) analyze with mass spectrometry; and (6) search mass spectra against protein database to identify target
proteins.

spectrometers have often been used to identify peptide mass mapping (Aebersold and Mann,
proteins by precise and accurate measurement of 2003), multiple peptide m/z values (at least 4 or
peptide m/z values alone (no tandem mass spectra more) combined with the Mr and pI of the orig-
sequence information). In this approach, called inating protein (from 2D-PAGE) is selective

enough to provide identification. However, it is of-
ten difficult to attain accurate measurement of 4 or
more peptides per protein, particularly for those of
lower abundance. Conti et al. (2004) encountered
this limitation with lower abundant differential
proteins displayed on 2D-PAGE separation of
control and TBI CSF. Further, peptide mass map-
ping will not work well for proteins that have been
modified and therefore shifted in Mr or pI value.
The added selectivity afforded by tandem mass
spectrometry alleviates these concerns where only
the sequence of the originating protein is needed to
provide an accurate match to data of a single pep-
tide (condoned when the data are of excellent spec-
tral quality, though data for two or more peptides
per protein are generally advised). Tandem mass
spectrometry coupled with peptide chromatogra-
phy is more resource intensive, and therefore is of-
ten reserved for identification when peptide mass
mapping fails, as used by Kochanek et al. (2006)
for differential analysis of control and TBI tissue.
Tandem mass spectrometry provides multidi-
mensional data via the use of gas-phase molecular
fragmentation — traditionally through collision
induced dissociation (Abersold and Mann, 2003),
though more recent techniques such as electron
capture dissociation (Cooper et al., 2005) and
electron transfer dissociation (Good and Coon,
2006) have shown some improved functionality.
The intact peptide (parent ion) m/z value is analy-
zed in the first stage of mass spectral analysis. In
the second stage, the peptide ion is broken up at
integral points between amino acids into sequence
specific fragment ions (daughter ions), which are
mass analyzed to produce a tandem mass spectrum
(Fig. 5). Fragmentation efficiency is not uniform
between all amino acids of a peptide; however,
enough information is generally attained to corre-
late the data with a database derived peptide se-
quence having a matched parent ion m/z value.
Bioinformatic software is used to automatically
process tandem mass spectra against a protein da-
tabase(s). A score and/or a statistical measure (XC
and P (pep), respectively in Fig. 5) of how well the
data matches with the peptide sequence is pro-
vided. The originating protein(s) is then listed with
all associated peptides. As illustrated in Fig. 5, ex-
cised gel bands or spots likely contain multiple
409
proteins, even following multidimensional protein
separation. It therefore must be noted that it can
still be difficult to associate a single protein with
a given differential gel spot or band even after
tandem mass spectrometry due to protein co-
migration. To help ascertain the differential nature
of a protein, peptide tandem mass spectrometry
data can also be quantitative. The number of
peptides per protein is a useful relative quantita-
tive measure of differential abundance (Sanders
et al., 2002; Peng et al., 2004; Ottens et al., 2005;
Kobeissy et al., 2006), e.g., six MAP2 peptides for
naı̈ve indicates greater abundance than one pep-
tide for TBI (Fig. 5). This information can be used
for data reduction in eliminating unlikely candi-
date proteins, those with no apparent change or
opposite change in quantity, and to confirm the
differential nature of the final protein list.
In all, mass spectrometry combined with protein
and peptide separations is an effective platform
for identifying protein dynamics following TBI
(Satchell et al., 2003; Conti et al., 2004; Haskins
et al., 2005; Kobeissy et al., 2006; Kochanak et al.,
2006). To date, a number of candidate biochemical
markers of TBI have been identified in a hand-
ful of neuroproteomic studies, which are in the
processes of validation and clinical translation.
Characterization of global protein dynamics and
pathways following neurotrauma is more elusive,
and will require more refined and exhaustive
proteomic experiments (further technological ad-
vancement will ultimately be required). Greater
protein and/or peptide separation, improved sen-
sitivity, and more refined m/z assignment of larger
polypeptides will drive the neuroproteomics field
forward, which will further the understanding of
the protein chemistry of brain injury and avenues
for new treatments.
TBI neuroproteomics by antibody arrays and high-
throughput blotting
Alternative to the mass spectrometry-based pro-
teomic approach, several antibody-based appro-
aches are also emerging through the availability
of various antibody array or protein array plat-
forms (Zyomyx protein Biochips, BD PowerBlot
410

Fig. 5. Bioinformatics of mass spectrometry data. Illustrated is an example of the bioinformatic output of processed tandem mass
spectrometry data of one differential target from our recent TBI neuroproteomics study (Kobeissy et al., 2006). Corresponding naı̈ve
and TBI gel bands at position 23A (first differential band in the 23 protein fraction) were excised, digested and analyzed by reversed
phase liquid chromatography–tandem mass spectrometry. Peptide tandem mass spectra (peptide probability, po0.001) indicated the
presence of the protein MAP2. Protein coverage was lower in the TBI sample than in naı̈ve control (1 and 6 peptides, respectively),
correlating this protein with the less intense TBI gel band.

and BD Clontech antibody microarrays 500)
(Moody et al., 2001; Graslund et al., 2002; James,
2002; Kusnezow and Hoheisel, 2002). The follow-
ing is a brief description of several of these tech-
nologies as applicable to TBI research, primarily
biomarker development.
Antibody microarray
Antibody microarrays are modeled after DNA
microarrays. A collection of highly specific pri-
mary antibodies (to different protein antigens) is
printed on a standard-size glass slide (e.g., BD-
Clontech microarray 500) (Gu et al., 2006). Bio-
logical samples (brain tissue lysate, CSF or serum/
plasma) from control and treatment groups are
labeled with green and red fluorescent Cyanine dye
(Cy-3 and Cy-5) pairs. An equal amount of each is
mixed and applied to the slide. After washing, the
fluorescence intensity for Cy-3 and Cy-5 are then
respectively measured. The bound Cy-3 to Cy-5
ratio will then represent a relative differential
quantitative measure of each antigen. The method
was pioneered with a cytokine panel due to the
multitude of cytokine antibodies available. Using
this kit, pro-inflammatory cytokines such as IL-1b
and TNF-alpha were found to be elevated in CSF
following TBI (Shiozaki et al., 2005), a general
indication for the involvement of CNS inflamma-
tion following neurotrauma. A drawback of this
method is the technical challenge of even sample
labeling. As an alternative, it is also possible to
construct a ‘‘sandwich format’’ antibody array in a

partitioned (e.g., 96-well) plate. In this case, two
highly specific primary antibodies are required to
detect each protein antigen. While the coating step
of the first set of antibodies (capture antibody) is
the same as above, the proteins in the biological
samples are not labeled or mixed. The second set
of detecting antibodies is applied to the plate,
forming an antibody–antigen–antibody ‘‘sand-
wich’’. Finally, the abundance of the antigen is
quantified by measuring the amplifiable detection
antibody in each well (Huang, 2001).
High-throughput immunoblotting
Another novel technique called high-throughput
immunoblotting (HTPI) has found some recent
utility for neuroproteomics research (Malakhov
et al., 2002; Yoo et al., 2002; Lorenz et al., 2003).
This proteomic approach involves multiplexing
traditional Western blot analysis. Using our own
recent TBI study as an example (Liu et al., 2006),
two pooled hippocampal lysate samples, naı̈ve
and CCI (200–1000 mg), were prepared and sub-
jected to two sets of five SDS-polyacrylamide gels
(13
10 cm). Proteins in the gel were transferred to
a blotting membrane, which was then clamped to a
blotting manifold that isolates 40 longitudinal
channels. A cocktail of between 4 and 6 antibodies
was added to each channel and allowed to hybrid-
ize for 1 h at 371C. After washing, a secondary
antibody conjugated to Alexa680 fluorescent dye
was added, then quantified by an Odyssey Infrared
Imaging system (LI-COR). Each of the antibodies
per channel is well characterized and resolved from
one another by relative molecular mass, thus
achieving coverage of 1000 proteins over five blots.
As HTPI is still a Western blot in principle, it
is useful for detecting protein modifications that
alter gel mobility (e.g., proteolytic breakdown
products). Using HTPI, we identified more than
40 decreased proteins following CCI relative to
naı̈ve, many of which were seemingly subjected
to calpain and/or caspase proteolysis. Another
positive feature is the ease for hit validation, as
the same antibodies used for HTPI can also be
used for subsequent conventional Western blot
confirmation. In this manner, we successfully
411
confirmed several novel TBI-linked proteolytic
substrates (Fig. 6), including betaII-spectrin, stria-
tin, synaptotagmin-1, synaptojanin-1 and NSF
(N-ethylmaleimide sensitive fusion protein) (Liu
et al., 2006).
In summary, antibody-based neuroproteomic
techniques offer several attractive features. First,
instrumentation requirements for antibody-based
platforms are generally far less formidable in cost
and operation than those for mass spectrometry
based approaches. Second, since the target antigen
is known for each antibody probe, initial protein
identification is rapid and straightforward, and as
a result the bioinformatic component of this ap-
proach is far less labor-intensive. Third, validation
is again relatively simple, since an antibody to the
target is already available. One does need to be
cautious about antibody cross reactivity to unre-
lated protein(s) that could give rise to false positive
identification. One major challenge of this ap-
proach is the lack of available antibodies for all
30,000 (approx.) proteins in the average mamma-
lian proteome, let alone the multitude of post-
translational forms, making this approach less
than exhaustive (Kusnezow and Hoheisal, 2002).
Across-species reactivity is another issue, i.e., can
an antibody raised against a human protein anti-
gen cross detect the rat protein homologue? Fur-
thermore, the antibodies used are not standardized
in their affinity to target antigens; however, the
Human Proteome Organization (HUPO) is taking
steps to standardize specific antibody sets for the
whole human proteome (Haab et al., 2005).
Validation of proteomic data for TBI biomarker
development
In recent years, the term proteomics is often men-
tioned together with biomarker discovery, as pro-
teomic studies have the capability of identifying
unique and unobvious protein biomarkers from
tissue or biofluids obtained from human diseased
patients or from animal models of disease and
disorders. Biomarkers of acute brain injury have
also attracted a lot of recent research inter-
est (Wang et al., 2006). First, it is commonly
recognized that better patient monitoring and
412

Fig. 6. Example of TBI differential TBI proteome as analyzed by HTPI. HTPI Template A is shown for naı̈ve rat hippocampus (upper
panel) and the controlled cortical impact injury counterpart (lower panel). Molecular mass markers (lane 40) are indicated on the right.
Thirteen proteins in template A were reduced in average intensity (abundance) after TBI (solid boxes in upper panel). In addition,
several breakdown products were observed (dotted boxes in lower panel). Two proteins found to be up-regulated in TBI were CASK
(A3) and Psme3 (A29) (in dotted box). Adapted with permission from Liu et al. (2006).

management often leads to better clinical outcome.
Thus, a simple point-of-care (POC) CSF- or
blood-based diagnostic test would be extremely
useful, and significantly less expensive than any
brain image scanner (such as computed tomo-
graphy). Second, numerous failures in clinical drug
trials for the treatment of ischemic strokes and
TBI over the last 10–15 years point to the difficulty
of translating preclinical successes to clinical suc-
cesses. One remedy of the situation is the develop-
ment of robust quantitative biomarkers that are
predictive of disease progression and treatment
response (Lesko and Atkinson, 2001; Pineda et al.,
2004; Wang et al., 2006). In practice, proteins
identified from differential neuroproteomic studies
of brain injury (human or animal models) can
only be considered potential biomarkers for
brain injury, particularly if the goal is to develop
a CSF- or blood-based brain injury biomarker
test. This is even truer of candidate proteins iden-
tified from differential tissue lysate analysis, not in
biofluids.

Here are several criteria for an ideal biomarker
of brain injury:
i. Levels of the biomarker in a relevant biofluid
(CSF and/or blood) correlate with disease
severity, e.g., Glasgow Coma Score (GCS)
for TBI.
ii. Levels of the biomarker correlate with clin-
ical outcome, e.g., as measured by the Ex-
tended Glasgow Outcome Scale (GOS-E).
iii. Biomarker signals are not confounded by
other non-brain injury neurological condi-
tions.
iv. Biomarker levels are responsive to therapeu-
tic interventions.
To date, the most common validation technol-
ogy for neuroproteomics data to biomarker assay
relies heavily on antibody assays, since antibodies
(both polyclonal and monoclonal) have the ability
to selectively identify and quantify a single protein
(antigen) of interest in a rapid, high-throughput
manner. Due to the maturity of immunological
methods, antibodies-based diagnostic assays have
been refined almost to an art form, namely, en-
zyme-linked immunosorbant assays (ELISA).
There are two forms of ELISAs: direct ELISA
utilizes a single detection antibody; or sandwich
ELISA (swELISA) that requires the formation of
an immobilized capture antibody to protein anti-
gen to detection antibody complex. For biomark-
er-based diagnostic, swELISA is indeed the
method of choice, since it provides antigen enrich-
ment that significantly improves signal fidelity.
The work flow for developing biomarker sw-
ELISA assays is as follows. Once a putative bio-
marker is identified and confirmed, one must
obtain a pair of compatible and high affinity anti-
bodies. Antibodies from commercial sources are
explored first; however, custom-made antibodies
are often needed. Custom antibodies are generally
raised against a peptide epitope-carrier protein
complex, a recombinant protein or a fusion pro-
tein (e.g., GST fusion protein) as the antigen,
which is injected into host animals for an immune
response (rabbit, chicken, goat, and mouse). Care
must be taken that the antibodies thus raised
will only recognize the protein isoform or post-
translationally modified form of interest (e.g.,
413
phosphorylated form or breakdown products,
such as c-tau or alphaII-spectrin breakdown prod-
ucts (SBDPs) (Pineda et al., 2004). Afterwards,
antigen affinity purification is highly recom-
mended. Finally, some form of the antigen (puri-
fied native protein or recombinant protein) must
be available for the development and testing of the
antibody, as well as for use to generate a standard
curve for absolute quantification.
What follows are iterations of trial and error to
find the best antibody pair (capture and detection),
and the optimal assaying conditions to give the
best signal to noise ratio as well as the lowest de-
tection limit. The term ELISA implies that the de-
tection antibody is coupled to an enzyme-linked
substrate to product system for signal amplifica-
tion (e.g., alkaline phosphatase or horse radish
peroxidase). An ideal biomarker ELISA assay
should strive to achieve high (490%) selectivity
(i.e., the ability to detect the presence of injury)
and high (490%) specificity (i.e., the ability to
detect the absence of injury).
A major challenge of brain injury biomarker
assay development is the availability of high qual-
ity antibodies. On this front, there have been sig-
nificant developments in capturing and detecting
agents beyond monoclonal antibodies, such as ap-
tamers (Bunka and Stockley, 2006) and phage-
displayed single chain antibody fragments (Tomizaki
et al., 2005). Another major challenge is the de-
tection limits, since it is conceivable that the best
brain injury marker might be very low in abun-
dance. ELISA sensitivity can be improved with an
enzymatic signal-enhancement method known as
tyramide signal amplification (TSA). It is also
apparent that technological advances in ELISA sen-
sitivity improvement will greatly benefit this field.
Animal models are commonly used for preclin-
ical biomarker validation, as they confer a number
of advantages. One important advantage of pre-
clinical animal models is the ability to prioritize
the potential utility of biomarkers by use of rela-
tively simple Western blot studies prior to com-
mitment of substantial time and resources to the
development of more sensitive quantitative ap-
proaches such as ELISA. Employing a CCI model
of TBI, we capitalized on the ability to simultane-
ously characterize the appearance of biomarkers in
414
injured tissues as well as cerebrospinal fluid (Pike
et al., 2001, 2003). This often overlooked advan-
tage allows the rigorous characterization of rela-
tionships between biomarker generation in injured
brain tissue and appearance in other compart-
ments such as CSF and ultimately in blood.
Validations with preclinical animal models
make possible comparisons of distribution bioki-
netics and the temporal profile of biomarker
appearance in different, clinically relevant com-
partments such as CSF and blood, and their re-
lationship to the time of injury. These values can
include the half-life and time to maximal concen-
trations of biomarkers. A clear understanding of
the biokinetics of biomarkers is an important
component and accurate interpretation of the
clinical relevance of changes seen in biomarker
levels. Apart from straightforward comparisons of
appearance and temporal profile of biomarkers,
preclinical animal models provide important in-
formation on the potential relationship of bio-
marker levels to injury magnitudes, lesion volume
and outcome. For example, our laboratory (Ringger
et al., 2004) has confirmed that a biochemical
marker of calpain proteolysis, the cytoskeletal pro-
tein alphaII-spectrin, when measured in CSF relia-
bly correlated with the magnitude of CCI injury,
lesion volume and neurological outcome.
Translation to clinical diagnostics
The primary challenge facing translation of bio-
markers to practical clinical diagnostics resides in
the approaches to confirm the clinical utility of
putative biomarkers of TBI identified and vali-
dated with preclinical animal models. Ideally, the
confirmation platform should integrate preclinical
validation with clinical assessments of the poten-
tial applicability of the candidate biomarkers. A
direct comparison of biomarker generation bet-
ween preclinical models and biomarker data from
human clinical studies allows investigators to gain
considerable insight into the validity (or challenges
to the validity) of the employed preclinical animal
models. For example, we have shown (unpublished
data) similar profiles of alphaII-spectrin degrada-
tion by calpain in both CCI and severe human
TBI. Thus, CCI reproduces important features of
calpain mediated pathology seen in human TBI.
The development of a clinical research platform
for TBI poses considerable challenges. An impor-
tant limitation of clinical models, as typically im-
plemented, is the failure to incorporate systematic
studies of the role of secondary insults and age,
significant determinants of injury response in hu-
man TBI. Mild TBI and the frequent absence of
objective evidence of injury by computerized tomo-
graphic studies or MRI are especially challenging.
Neurological diagnosis of varying magnitudes of
mild or moderate TBI is unreliable, since the GCS
scale was developed primarily to assess more se-
verely injured comatose patients. Outcome meas-
ures are equally controversial and ambiguous.
However, greater progress has been made in estab-
lishing a clinical research platform for severe TBI.
Our program has recently completed the first
extensive examination of the potential clinical util-
ity of biochemical markers of calpain and caspase-
3 proteolysis in severe (GCS ¼ 4 8) TBI patients
(Pineda, 2006, in press). This research incorporated
important elements of a clinical research platform
for biomarkers of TBI. Such a platform should be
able to test hypotheses about the potential clinical
utility and validity of biomarkers assessed, alone or
in combination. Generally validation effects should
focus on whether biomarkers are reliably associ-
ated with the magnitude of brain injury, occurrence
of secondary insults and outcome. The platform
should allow rigorous determination of whether
appropriate markers can provide information criti-
cal for diagnosis, triage and management, as well
as data on possible injury mechanisms and cellular
localization of injury. In the research of Pineda
(2006, in press), investigators were able to confirm
that a protein biomarker of calpain proteolysis
(degradation of alphaII-spectrin) measured in CSF
was reliably associated with the magnitude of di-
chotomized GCS on admission (GCS 3–6 vs. 6–8),
while a biomarker of caspase-3 alphaII-spectrin
proteolysis showed no such association. The bio-
marker of calpain proteolysis was also reliably as-
sociated with secondary insult, (preadmission
hypotension) while a marker of caspase-3 pro-
teolysis showed no significant relationship. The
biomarker of calpain proteolysis, but not caspase-3

proteolysis was also differentially associated with
outcome. Significantly higher level of a biomarker
of calpain proteolysis was detected in patients within
the first 24 h following injury who subsequently had
a worse outcome (assessed by the dichotomized
GCS 6 months after injury). Importantly, calpain
proteolysis is primarily associated with oncotic
necrosis while caspase-3 proteolysis is more promi-
nently associated with apoptosis. Thus, these data
collectively suggest that oncotic necrosis may more
importantly contribute to pathology in severe TBI
than apoptosis. This hypothesis is consistent with
the temporal profile of accumulation of these
biomarkers in CSF of severe TBI patients for the
first 5 days following injury. Markers of calpain
proteolysis were increased more than caspse-3 pro-
teolysis, although caspase-3 proteolysis tended to be
more sustained.
Our laboratory has recently begun an important
series of studies that will significantly enhance our
ability to validate relationships between changes in
biomarkers and secondary insults following TBI.
An important limitation in such studies has been
the absence of the ability to continuously record
physiological data that would indicate the presence
of a secondary insult (e.g., reduced ICP, hypoten-
sion, compromised tissue oxygenation). In fact,
Hemphill et al. (2005) has recently documented
that studies conducted by collecting data of medi-
cal records, frequently missed secondary insults
documented by continuous recording of physio-
logical parameters. Thus, our program has recently
initiated the systematic comparison of biomarker
changes assessed from both CSF and blood with
physiological variables recorded continuously from
severe TBI patient. The ultimate goal of these
studies is to determine if changes in biomarker
levels can, in fact, predict onset of secondary in-
sults as well as document their occurrence. If so,
rapid assessment of biomarkers could provide criti-
cal information guiding management of severely
injured TBI patients in the intensive care unit.
Closing remarks
In conclusion, neuroproteomics has revolutionized
the exploration of protein dynamics following TBI.
415
In the near-term, whether by mass spectrometry or
antibody-based arrays, neuroproteomics serves as
a high-throughput means to discover potentially
novel biochemical markers of TBI. Significant ad-
vances have also been made in developing strate-
gies for preclinical and clinical validation of
biomarkers of TBI. While often limited by the fail-
ure to incorporate the role of secondary insults, age
and gender, preclinical animal models in combina-
tion with neuroproteomic techniques have proven
invaluable for rapidly assessing the potential utility
of the relatively large numbers of biomarkers
generated. While significant challenges remain in
implementing preclinical platforms from mild
and moderate TBI, the establishment of effective
platforms to validate biomarkers in severe TBI
promises to define clinical useful subsets of bio-
markers of TBI within the next few years. Looking
forward, neuroproteomic technology will rapidly
evolve to more effectively and more efficiently
characterize the neuroproteome as altered by TBI,
much as has been seen in the genomic revolution.
This impending opportunity will confer unparal-
leled detail on the biochemical processes occurring
following TBI, and will inspire novel therapies tri-
als, to be evaluated with the biomarkers presently
in development.
Disclaimer
KKW and RLH own stock, receive royalties from
and are executive officers of Banyan Biomarkers
Inc. and as such may benefit financially as a result
of the outcomes of this research or work reported
in this publication.
References
Aebersold, R. and Mann, M. (2003) Mass spectrometry-based
proteomics. Nature, 422: 198–207.
Albensi, B.C., Knoblach, S.M., Chew, B.G., O’Reilly, M.P.,
Faden, A.I. and Pekar, J.J. (2000) Diffusion and high res-
olution MRI of traumatic brain injury in rats: time course
and correlation with histology. Exp. Neurol., 162: 61–72.
Alexander, M.P. (1995) Mild traumatic brain injury: patho-
physiology, natural history, and clinical management.
Neurology, 45: 1253–1260.
416
Beranova-Giorgianni, S. (2003) Proteome analysis by two--
dimensional gel electrophoresis and mass spectrometry:
strengths and limitations. Trac-Trend Anal. Chem., 5:
273–281.
Bogdanov, B. and Smith, R.D. (2005) Proteomics by FTICR
mass spectrometry: top down and bottom up. Mass Spect-
rom. Rev., 24: 168–200.
Bunka, D.H. and Stockley, P.G. (2006) Aptamers come of age
— at last. Nat. Rev. Microbiol., 4: 588–596.
Buonocore, G., Liberatori, S., Bini, L., Mishra, O.P., Delivoria-
Papadopoulos, M., Pallini, V. and Bracci, R. (1999) Hypoxic
response of synaptosomal proteins in term guinea pig fetuses.
J. Neurochem., 73: 2139–2148.
Carey, M.E. (1995) Experimental missile wounding of the
brain. Neurosurg. Clin. N. Am., 6: 629–642.
Carey, M.E., Sarna, G.S. and Farrell, J.B. (1990) Brain edema
after an experimental missile wound. Adv. Neurol., 52:
301–305.
Carey, M.E., Sarna, G.S., Farrell, J.B. and Happel, L.T. (1989)
Experimental missile wound to the brain. J. Neurosurg., 71:
754–764.
Cernak, I. (2005) Animal models of head trauma. NeuroRx, 2:
410–422.
Cernak, I., Savic, J., Ignjatovic, D. and Jevtic, M. (1999) Blast
injury from explosive munitions. J. Trauma, 47: 96–104.
Cernak, I., Savic, V.J., Kotur, J., Prokic, V., Veljovic, M. and
Grbovic, D. (2000) Characterization of plasma magnesium
concentration and oxidative stress following graded trau-
matic brain injury in humans. J. Neurotrauma, 17: 53–68.
Cernak, I., Vink, R., Zapple, D.N., Cruz, M.I., Ahmed, F.,
Chang, T., Fricke, S.T. and Faden, A.I. (2004) The patho-
biology of moderate diffuse traumatic brain injury as iden-
tified using a new experimental model of injury in rats.
Neurobiol. Dis., 17: 29–43.
Cernak, I., Wang, Z., Jiang, J., Bian, X. and Savic, J. (2001)
Ultrastructural and functional characteristics of blast injury-
induced neurotrauma. J. Trauma, 50: 695–706.
Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh,
T.K. (1998) Experimental brain injury induces regionally
distinct apoptosis during the acute and delayed post-
traumatic period. J. Neurosci., 18: 5663–5672.
Conti, A., Sanchez-Ruiz, Y., Bachi, A., Beretta, L., Grandi, E.,
Beltramo, M. and Alessio, M. (2004) Proteome study of
human cerebrospinal fluid following traumatic brain in-
jury indicates fibrinogen degradation products as trauma-
associated markers. J. Neurotrauma, 21: 854–863.
Cooper, H.J., Hakansson, K. and Marshall, A.G. (2005) The
role of electron capture dissociation in biomolecular analysis.
Mass Spectrom. Rev., 24: 201–222.
De Mulder, G., Van Rossem, K., Van Reempts, J., Borgers, M.
and Verlooy, J. (2000) Validation of a closed head injury
model for use in long-term studies. Acta Neurochir. Suppl.,
76: 409–413.
Denslow, N.D., Michel, M.E., Temple, M.D., Hsu, C.Y., Sa-
atman, K. and Hayes, R.L. (2003) Application of proteomics
technology to the field of neurotrauma. J. Neurotrauma, 20:
401–407.
Dixon, C.E., Clifton, G.L., Lighthall, J.W., Yaghmai, A.A. and
Hayes, R.L. (1991) A controlled cortical impact model of
traumatic brain injury in the rat. J. Neurosci. Methods, 39:
253–262.
Erb, D.E. and Povlishock, J.T. (1988) Axonal damage in severe
traumatic brain injury: an experimental study in cat. Acta
Neuropathol., 76: 347–358.
Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989)
The role of excitatory amino acids and NMDA receptors in
traumatic brain injury. Science, 244: 798–800.
Faden, A.I., Knoblach, S.M., Cernak, I., Fan, L., Vink, R.,
Araldi, G.L., Fricke, S.T., Roth, B.L. and Kozikowski, A.P.
(2003) Novel diketopiperazine enhances motor and cognitive
recovery after traumatic brain injury in rats and shows
neuroprotection in vitro and in vivo. J. Cereb. Blood Flow
Metab., 23: 342–354.
Farkas, O., Polgar, B., Szekeres-Bartho, J., Doczi, T., Povli-
shock, J.T. and Buki, A. (2005) Spectrin breakdown products
in the cerebrospinal fluid in severe head injury — preliminary
observations. Acta Neurochir., 147: 855–861.
Finnie, J.W. (1993) Pathology of experimental traumatic
craniocerebral missile injury. J. Comp. Pathol., 108: 93–101.
Finkelstein, E., Corso, P. and Miller, T. (2006) The Incidence
and Burden of Injuries in the United States. Oxford Univer-
sity Press, New York.
Gennarelli, T.A. (1994) Animate models of human head injury.
J. Neurotrauma, 11: 357–368.
Goldman, H., Hodgson, V., Morehead, M., Hazlett, J. and
Murphy, S. (1991) Cerebrovascular changes in a rat model of
moderate closed-head injury. J. Neurotrauma, 8: 129–144.
Good, D.M. and Coon, J.J. (2006) Advancing proteomics with
ion/ion chemistry. Biotechniques, 40: 783–789.
Graslund, S., Falk, R., Brundell, E., Hooq, C. and Stahl, S.
(2002) A high-stringency proteomics concept aimed for gen-
eration of antibodies specific for cDNA encoded proteins.
Biotechnol. Appl. Biochem., 35: 75–82.
Gu, Q., Sivanandam, T.M. and Kim, C.A. (2006) Signal sta-
bility of Cy3 and Cy5 on antibody microarrays. Proteome
Sci., 4: 21.
Haab, B.B., Geierstanger, B.H., Michailidis, G., Vitzthum, F.,
Forrester, S., Okon, R., Saviranta, P., Brinker, A., Sorette,
M., Perlee, L., Suresh, S., Drwal, G., Adkins, J.N. and
Omenn, G.S. (2005) Immunoassay and antibody microarray
analysis of the HUPO Plasma Proteome Project reference
specimens: systematic variation between sample types and cali-
bration of mass spectrometry data. Proteomics, 5: 3278–3291.
Haskins, W.E., Kobeissy, F.H., Wolper, R.A., Ottens, A.K.,
Kitlen, J.W., McClung, S.H., O’Steen, B.E., Chow, M.M.,
Pineda, J.A., Denslow, N.D., Hayes, R.L. and Wang,
K.K.W. (2005) Rapid discovery of putative protein biomark-
ers of traumatic brain injury by SDS-PAGE-capillary liquid
chromatography-tandem mass spectrometry. J. Neurotrau-
ma, 22: 629–644.
Hemphill III, J.C., Barton, C.W., Marabito, D. and Manley,
G.T. (2005) Influence of data resolution and interpolation
method on assessment of secondary brain insults in neuro-
critical care. Physiol. Meas., 26: 373–386.

Huang, R.P. (2001) Simultaneous detection of multiple proteins
with an array-based enzyme-linked immunosorbent assay
(ELISA) and enhanced chemiluminescence (ECL). Clin.
Chem. Lab. Med., 39: 209–214.
James, P. (2002) Chips for proteomics: a new tool or just hype?
Biotechniques, Dec(Suppl): 4–13.
Jenkins, L.W., Peters, G.W., Dixon, C.E., Zhang, X., Clark,
R.S.B., Skinner, J.C., Marion, D.W., Adelson, P.D.
and Kochanek, P.M. (2002) Conventional and functional
proteomics using large format two-dimensional gel electro-
phoresis 24 hours after controlled cortical impact in postnatal
day 17 rats. J. Neurotrauma, 19: 715–740.
Kaufmann, R. (1995) Matrix-assisted laser desorption ioniza-
tion (MALDI) mass spectrometry: a novel analytical tool in
molecular biology and biotechnology. J. Biotechnol., 41:
155–175.
Kelleher, N.L. (2004) Top-down proteomics. Anal. Chem., 76:
197A–203A.
Kobeissy, F.H., Ottens, A.K., Zhang, Z., Liu, M.C., Denslow,
N.D., Dave, J.R., Tortella, F.C., Hayes, R.L. and Wang,
K.K. (2006) Novel differential neuroproteomics analysis of
traumatic brain injury in rats. Mol. Cell. Proteomics, 5:
1887–1898.
Kochanek, A.R., Kline, A.E., Gao, W.M., Chadha, M., Lai,
Y., Clark, R.S.B., Dixon, C.E. and Jenkins, L.W. (2006) Gel-
based hippocampal proteomic analysis 2 weeks following
traumatic brain injury to immature rats using controlled
cortical impact. Dev. Neurosci., 28: 410–419.
Kupina, N.C., Detloff, M.R., Bobrowski, W.F., Snyder, B.J.
and Hall, E.D. (2003) Cytoskeletal protein degradation and
neurodegeneration evolves differently in males and females
following experimental head injury. Exp. Neurol., 180: 55–73.
Kusnezow, W. and Hoheisel, J.D. (2002) Antibody microar-
rays: promises and problems, 33: S14–S23
Langlois, J.A., Rutland-Brown, W. and Thomas, K.E. (2006a)
Traumatic Brain Injury in the United States: Emergency De-
partment Visits, Hospitalizations, and Deaths. Center for
Disease Control and Prevention, National Center for Injury
Prevention and Control, Atlanta, GA.
Langlois, J.A., Rutland-Brown, W. and Wald, M.M. (2006b)
The epidemiology and impact of traumatic brain injury: a
brief overview. J. Head Trauma Rehabil., 21: 375–378.
Lesko, L.J. and Atkinson Jr., A.J. (2001) Use of biomarkers
and surrogate endpoints in drug development and regulatory
decision making: criteria, validation, strategies. Annu. Rev.
Pharmacol. Toxicol., 41: 347–366.
Liu, M.C., Akle, V., Zheng, W.R., Dave, J.R., Tortella, F.C.,
Hayes, R.L. and Wang, K.K.W. (2006) Comparing calpain-
and caspase-3-degradation patterns in traumatic brain injury
by differential proteome analysis. Biochem. J., 394: 715–725.
Lorenz, P., Ruschpler, P., Koczan, D., Stiehl, P. and Thiesen,
H.J. (2003) From transcriptome to proteome, differentially ex-
pressed proteins identified in synovial tissue of patients suffer-
ing from rheumatoid arthritis and osteoarthritis by an initial
screen with a panel of 791 antibodies. Proteomics, 3: 991–1002.
Malakhov, M.P., Kim, K.I., Malakhova, O.A., Jacobs, B.S.,
Borden, E.C. and Zhang, D.E. (2002) High-throughput
417
immunoblotting. Ubiquitiin-like protein ISG15 modifies
key regulators of signal transduction. J. Biol. Chem., 278:
16608–16613.
Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J.,
Kita, H. and Demetriadou, K. (1994) A new model of diffuse
brain injury in rats. Part I: Pathophysiology and biomechan-
ics. J. Neurosurg., 80: 291–300.
McIntosh, T.K., Juhler, M. and Wieloch, T. (1998) Novel
pharmacologic strategies in the treatment of experimental
traumatic brain injury. J. Neurotrauma, 15: 731–769.
McIntosh, T.K., Yu, T. and Gennarelli, T.A. (1994) Alterations
in regional brain catecholamine concentrations after experi-
mental brain injury in the rat. J. Neurochem., 63: 1426–1433.
Molloy, M.P., Brzenzinski, E.E., Hang, J., McDowell, M.T.
and VanBogelen, R.A. (2003) Overcoming technical varia-
tion and biological variation in quantitative proteomics. Pro-
teomics, 3: 1912–1919.
Moody, M.D., Van Arsdell, S.W., Murphy, K.P., Orencole,
S.F. and Burns, C. (2001) Array-based ELISAs for high-
throughput analysis of human cytokines. Biotechniques, 31:
186–194.
Morrison, R.S., Kinoshita, Y., Johnson, M.D., Uo, T., Ho,
J.T., McBee, J.K., Conrads, T.P. and Veenstra, T.D. (2002)
Proteomic analysis in the neurosciences. Mol. Cell. Pro-
teomics, 1: 553–560.
Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A.,
Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C.,
Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D.,
Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D.,
Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller,
N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W.,
Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E.,
Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P.,
Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G.,
Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich,
M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L.
(2002) Clinical trials in head injury. J. Neurotrauma, 19:
503–557.
Nguyen, D.N., Becker, G.W. and Riggin, R.M. (1995) Protein
mass spectrometry: applications to analytical biotechnology.
J. Chromatogr. A, 705: 21–45.
Ommaya, A.K. (1995) Head injury mechanisms and the con-
cept of preventive management: a review and critical synthe-
sis. J. Neurotrauma, 12: 527–546.
Ommaya, A.K., Goldsmith, W. and Thibault, L. (2002) Bio-
mechanics and neuropathology of adult and paediatric head
injury. Br. J. Neurosurg., 16: 220–242.
Ottens, A.K., Kobeissy, F.H., Haskins, W.E., Golden, E.C.,
Zhang, Z.Q., Chen, S., Hayes, R.L., Wang, K.K.W. and
Denslow, N.D. (2006) Neuroproteomics in neurotrauma.
Mass Spectrom. Rev., 25: 380–408.
Ottens, A.K., Kobeissy, F.H., Wolper, R.A., Haskins, W.E.,
Hayes, R.L., Denslow, N.D. and Wang, K.K.W. (2005) A
multidimensional differential proteomic platform using dual
phase ion-exchange chromatography — polyacrylamide gel
electrophoresis/reverse phase liquid chromatography tandem
mass spectrometry. Anal. Chem., 77: 4836–4845.
418
Patton, W.F. (2002) Detection technologies in proteome anal-
ysis. J. Chromatogr. B, 771: 3–31.
Peng, J., Kim, M.J., Cheng, D., Duong, D.M., Gygi, S.P. and
Sheng, M. (2004) Semiquantitative proteomics analysis of rat
forebrain postsynaptic density fractions by mass spectrome-
try. J. Biol. Chem., 279: 21003–21011.
Pike, B.R., Flint, J., Dave, J.R., Lu, X.C., Wang, K.K.,
Tortella, F.C. and Hayes, R.L. (2003) Accumulation of cal-
pain and caspase-3 proteolytic fragments of brain-derived
alphaII-spectrin in CSF after middle cerebral artery occlu-
sion in rats. J. Cereb. Blood Flow Metab., 24: 98–106.
Pike, B.R., Flint, J., Dutta, S., Johnson, E., Wang, K.K.W. and
Hayes, R.L. (2001) Accumulation of non-erythroid alphaII-
spectrin and calpain-cleaved alphaII-spectrin breakdown
products in cerebrospinal fluid after TBI in rats. J. Neuroc-
hem., 78: 1297–1306.
Pineda, J.A. (2006) Accumulation of alphaII-spectrin break-
down products in CSF of severely injured traumatic brain
injury patients. J. Neurotrauma, in press.
Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers
of proteolytic damage following traumatic brain injury. Brain
Pathol., 14: 202–209.
Piper, I.R., Thomson, D. and Miller, J.D. (1996) Monitoring
weight drop velocity and foam stiffness as an aid to quality
control of a rodent model of impact acceleration neurotrau-
ma. J. Neurosci. Methods, 69: 171–174.
Povlishock, J.T., Hayes, R.L., Michel, M.E. and McIntosh,
T.K. (1994) Workshop on animal models of traumatic brain
injury. J. Neurotrauma, 11: 723–732.
Ragnarsson, K.T. (2006) Traumatic brain injury research since
the 1998 NIH consensus conference. J. Head Trauma Re-
habil., 21: 379–387.
Ringger, N.C., O’Steen, B.E., Brabham, J.G., Siler, X., Pineda,
J., Wang, K.K.W. and Hayes, R.L. (2004) A novel marker
for traumatic brain injury: CSF alphaII-spectrin breakdown
product levels. J. Neurotrauma, 21: 1443–1456.
Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neuge-
bauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic
cell death after experimental traumatic brain injury in the rat.
Am. J. Pathol., 147: 1575–1583.
Saljo, A., Bao, F., Haglid, K.G. and Hansson, H.A. (2000)
Blast exposure causes redistribution of phosphorylated ne-
urofilament subunits in neurons of the adult rat brain. J.
Neurotrauma, 17: 719–726.
Sanders, S.L., Jeffings, J., Canutescu, A., Link, A.J. and Well,
P.A. (2002) Proteomics of the eukaryotic transcription ma-
chinery: identification of proteins associated with compo-
nents of yeast TFIID by multidimensional mass
spectrometry. Mol. Cell. Biol., 22: 4723–4738.
Satchell, M.A., Zhang, X., Kochanek, P.M., Dixon, C.E.,
Jenkins, L.W., Melick, J., Szabo, C. and Clark, R.S.B. (2003)
A dual role for poly-ADP-ribosylation in spatial memory
acquisition after traumatic brain injury in mice involving
NAD+ depletion and ribosylation of 14-3-3g. J. Neuroc-
hem., 85: 697–708.
Shimamura, M., Garcia, J.M., Prough, D.S., Dewitt, D.S.,
Uchida, T., Shah, S.A., Avila, M.A. and Hellmich, H.L.
(2005) Analysis of long-term gene expression in neurons of
the hippocampal subfields following traumatic brain injury in
rats. J. Neurosci., 131: 87–97.
Shiozaki, T., Hayakata, T., Tasaki, O., Hosotubo, H., Fuijita,
K., Mouri, T., Tajima, G., Kajino, K., Nakae, H., Tanaka,
H., Shimazu, T. and Sugimoto, H. (2005) Cerebrospinal fluid
concentrations of anti-inflammatory mediators in early-phase
severe traumatic brain injury. Shock, 23: 406–410.
Smith, D.H., Uryu, K., Saatman, K.E., Trojanowski, J.Q. and
McIntosh, T.K. (2003) Protein accumulation in traumatic
brain injury. Neuromolecular Med., 4: 59–72.
Sosin, D.M., Sniezek, J.E. and Waxweiler, R.J. (1995) Trends in
death associated with traumatic brain injury, 1979 through
1992. Success and failure. JAMA, 273: 1778–1780.
Tomizaki, K.Y., Usui, K. and Mihara, H. (2005) Protein-
detecting microarrays: current accomplishments and require-
ments. Chembiochem, 6: 782–799.
Tornheim, P.A., McDermott, F. and Shiguma, M. (1990) Effect
of experimental blunt head injury on acute regional cerebral
blood flow and edema. Adv. Neurol., 52: 377–384.
Tornheim, P.A. and McLaurin, R.L. (1981) Acute changes in
regional brain water content following experimental closed
head injury. J. Neurosurg., 55: 407–413.
Wang, K.K.W., Larner, S.F., Robinson, G. and Hayes, R.L.
(2006) Neuroprotection targets after traumatic brain injury.
Curr. Opin. Neurol., 19: 514–519.
Wang, K.K.W., Ottens, A., Haskins, W.E., Liu, M.C., Den-
slow, N., Chen, S.S. and Hayes, R.L. (2004) Neuroproteomic
studies of traumatic brain injury. In: Neuhold L.A. (Ed.),
Human Brain Proteome — International Review of Neuro-
biololgy, Vol. 61. Elsevier, San Diego, pp. 215–240.
Wang, K.K.W., Ottens, A.K., Liu, M.C., Lewis, S.B., Meegan,
C., Oli, M.W., Tortella, F.C. and Hayes, R.L. (2005) Proteo-
mic identification of biomarkers of traumatic brain injury.
Expert Rev. Proteomics, 2: 603–614.
Warden, D. (2006) Military TBI during the Iraq and Afghani-
stan wars. J. Head Trauma Rehabil., 21: 398–402.
Wennersten, A., Holmin, S. and Mathiesen, T. (2003) Charac-
terization of Bax and Bcl-2 in apoptosis after experimental
traumatic brain injury in the rat. Acta Neuropathol., 105:
281–288.
Wieloch, T. and Nikolich, K. (2006) Mechanisms of neural
plasticity following brain injury. Curr. Opin. Neurobiol., 16:
258–264.
Williams, A.J., Hartings, J.A., Lu, X.C., Rolli, M.L., Dave,
J.R. and Tortella, F.C. (2005) Characterization of a new rat
model of penetrating ballistic brain injury. J. Neurotrauma,
22: 313–331.
Yi, J.H. and Hazell, A.S. (2006) Excitotoxic mechanisms and
the role of astrocytic glutamate transporters in traumatic
brain injury. Neurochem. Int., 48: 394–403.
Yoo, G.H., Piechocki, M.P., Ensley, J.F., Nguyen, T., Oliver,
J., Meng, H., Kewson, D., Shibuya, T.Y., Lonardo, F. and
Tainsky, M.A. (2002) Docetaxel induced gene expression
patterns in head and neck squamous cell carcinoma using
cDNA microarray and PowerBlot. Clin. Cancer Res., 8:
3910–3921.
Weber & Maas (Eds.)
Progress in Brain Research, Vol. 161
ISSN 0079-6123
Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 30

Remyelination of the injured spinal cord

Masanori Sasaki1,2, Bingcang Li1,2,3, Karen L. Lankford1,2, Christine Radtke1,2,4 and

Jeffery D. Kocsis1,2,

1
Department of Neurology and Center for Neuroscience and Regeneration Research, Yale University School of Medicine,
New Haven, CT 06510, USA
2
Rehabilitation Research Center, Veterans Affairs Connecticut Healthcare System, West Haven, CT 06516, USA
3
Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing, P.R. China
4
Department of Plastic, Hand and Reconstructive Surgery, Medical School Hannover, Hannover, Germany

Abstract: Contusive spinal cord injury (SCI) can result in necrosis of the spinal cord, but often long white
matter tracts outside of the central necrotic core are demyelinated. One experimental strategy to improve
functional outcome following SCI is to transplant myelin-forming cells to remyelinate these axons and
improve conduction. This review focuses on transplantation studies using olfactory ensheathing cell (OEC)
to improve functional outcome in experimental models of SCI and demyelination. The biology of the OEC,
and recent experimental research and clinical studies using OECs as a potential cell therapy candidate are
discussed.

Keywords: spinal cord injury; remyelination; olfactory ensheathing cells

Introduction Cellular transplantation of appropriate cells into

The clinical pathophysiology of contusive spinal
cord injury (SCI) is complex. A combination of
hemorrhage, ischemia and/or edema develops re-
sulting in necrosis with tissue loss (Schwab et al.,
2006, review). Additionally long tracts of the
injured spinal cord may survive, but become
demyelinated. Although regeneration of spinal cord
axons is an ultimate objective, the presence in many
patients with non-penetrating SCI of a population
of surviving axons that do not conduct impulse
due to demyelination (Keirstead, 2005; Kocsis
and Sasaki, 2005, reviews), suggests a cell-based
approach for remyelination as a potential strategy
for inducing recovery of function in SCI. One
approach to this goal capitalizes on recent progress
in the transplantation of cells into the injured CNS.
Corresponding author. Tel.: +1-(203)-937-3802;
Fax: +1-(203)-937-3801; E-mail: Jeffery.Kocsis@yale.edu
experimental models of SCI can promote axonal
regeneration, provide neuroprotective effects by
secretion of neurotrophins and remyelinate axons.
One cell of particular interest as a cell therapy
candidate to both encourage axonal remyelination
and regeneration is a specialized glial cell, the
olfactory ensheathing cell (OEC). Adult olfactory
receptor neurons continually undergo turnover
from an endogenous progenitor pool, and their
nascent axons grow through the olfactory nerves
and cross the PNS–CNS interface, where they form
new synaptic connections in the olfactory bulb
(Graziadei et al., 1978). OECs associate with olfac-
tory receptor neurons from their peripheral origin
to their central projection in the outer nerve layer of
the olfactory bulb (Doucette, 1991). This putative
support role of OECs in axonal growth within the
adult CNS has spawned extensive research to study
the potential of OEC transplants encouraging ax-
onal regeneration and functional recovery in SCI

DOI: 10.1016/S0079-6123(06)61030-3 419

420
models (Li et al., 1997, 1998; Ramon-Cueto et al.,
1998, 2000; Imaizumi et al., 2000a, b).
Transplantation of OECs after SCI is associated
with functional improvement even when trans-
plantation is delayed by weeks (Lu et al., 2002;
Keyvan-Fouladi et al., 2003). While the precise
mechanisms of the functional recovery after OEC
transplantation are not fully understood, several
mechanisms including elongative axonal regen-
eration, axonal sparing, sprouting and plasticity
associated with novel polysynaptic pathways, re-
cruitment of endogenous SCs and remyelination
have been proposed (Raisman, 2001; Bareyre
et al., 2004; Sasaki et al., 2004; Keyvan-Fouladi
et al., 2002). In animals with SCI and OEC trans-
plants, myelinated axons spanning the lesion site
display a characteristic peripheral pattern of my-
elination similar to that of Schwann cell (SC) my-
elination (Franklin et al., 1996; Li et al., 1997,
1998; Imaizumi et al., 1998, 2000b; Sasaki et al.,
2004).
Remyelination by transplantation may be of
particular importance because contusive SCI often
results in loss of function from demyelination
induced by spinal cord trauma. OECs can form
myelin when transplanted into demyelinated spinal
cord (Franklin et al., 1996; Imaizumi et al., 1998;
Sasaki et al., 2004). Moreover, human OECs can
form myelin in the immunosuprressed rat (Barnett
et al., 2000; Kato et al., 2000) and OECs trans-
planted into the non-human primate can remyelin-
ate spinal cord axons (Radtke et al., 2004).
Remyelination in demyelinated spinal cord
Three types of SCI models are commonly used
in experimental studies in rodents: transection,
compression and contusion (Rosenzweig and
McDonald, 2004, review). However, because of
the complex pathophysiological conditions of SCI
in these models, the study of remyelination is more
complex. Aside from the heterogeneity of the le-
sion, endogenous remyelination by SCs occurs
in these model systems thus making distinction of
endogenous remyelination from that of trans-
planted cells more complex. A model used by sev-
eral groups including ours to study remyelination
by transplanted cells is the X-irradiation/ethidium
bromide model (X-EB). Ethidium bromide (EB), a
nucleic acid chelator, kills cells in the target zone
(dorsal white matter) and focal X-irradiation to
blocks mitosis and kills oligodendrocyte progeni-
tors (Blakemore and Crang, 1985). In this model
system, endogenous repair in the center of the le-
sion is delayed by 6–8 weeks thus allowing a stable
time window to assess transplanted cells (Honmou
et al., 1996). Generally our analyses are carried out
within 3 weeks after X-EB lesion induction, which
is well within the time at which persistent demyeli-
nation is confirmed in a large number of control
studies (Blakemore and Crang, 1985; Franklin
et al., 1996; Honmou et al., 1996; Kato et al.,
2000). The lesion induced by this procedure is
characterized by virtually complete loss of endog-
enous glial elements (astrocytes and oligodendro-
cytes) with preservation of axons (Fig. 1A). The
higher power light micrograph of the lesion shows
compact fields of demyelinated axons in close ap-
position separated by areas of myelin debris and
macrophages (Figs. 1A2, A3). The lesion occupies
nearly the entire dorso-ventral extent of the dorsal
columns for 5–7 mm longitudinally.
A large body of work supports the proposal
that transplantation of OECs into various SCI
and demyelination models can promote axonal re-
generation, remyelination and functional recovery
(Franklin et al., 1996; Li et al., 1998; Ramon-Cueto
et al., 1998; Imaizumi et al., 2000a, b; Ramon-
Cueto et al., 2000; Lu et al., 2002; Keyvan-Fouladi
et al., 2003; Plant et al., 2003). Yet, there is an
important controversy as to whether the trans-
planted OECs associate with axons and form
peripheral myelin, as opposed to recruiting endog-
enous SCs that form myelin (Takami et al., 2002;
Boyd et al., 2004). OECs can express a number of
trophic factors, transcription factors and extracel-
lular matrix molecules (Ramon-Cueto et al., 1998;
Chuah and West, 2002; Au and Roskams, 2003;
Ramer et al., 2004), which could facilitate endog-
enous SC cell invasion, angiogenesis and activation
of progenitor cells to facilitate repair.
To address this issue, we used OECs from
GFP-expressing rats. After 3 weeks of transplan-
tation into X-EB model, methylene blue/Azure II
semithin plastic sections demonstrated extensive
Fig. 1. Light micrographs of transverse sections of the dorsal spinal cord stained with methylene blue/azure II showing the dorsal
funiculus in demyelinated (A) and OEC transplanted (B) rats. Examination at higher magnification show demyelinated (A2, A3) and
remyelinated (B2, B3) axons in the dorsal columns. Note the extensive myelination after transplantation of OEC (B). Many myelin-
forming cells were similar to peripheral myelin-forming cells, characterized by large nuclear and cytoplasmic regions (B3). A2, B2 and
A3, B3 were prepared from the boxed area in A1, B1 and A2, B2 respectively. Immuno-EM for GFP shows numerous GFP+ cells
remyelinating the demyelinated axons (C, D). Counterstaining for these sections was minimal, and dark staining shows electron dense
immunoperoxidase reaction product. Note the electron dense reaction product in the cytoplasm and nuclei of most cells forming
myelin (C, D). One cell associated with myelin in this field displays distinct reaction product, whereas an adjacent cell does not have
electron dense reaction product in its cytoplasm (D1). D2 is an enlargement of the box in D1. Note the dense reaction product in the
cytoplasm of the myelin-forming cell on the right and the basement membrane surrounding the cell. Scale bars ¼ 700 mm (A1, B1);
70 mm (A2, B2); 1 mm (A3, B3); 5 mm (C); 2 mm (D1); 0.4 (D2). C and D are modified with permission from Sasaki et al. (2006a).
422
remyelination in the demyelinated axons (Fig. 1B).
To fully establish that the OECs derived from
GFP rats were indeed responsible for the remyeli-
nation, immuno-electron microscopy (EM) for
GFP was performed. Intense reaction product
was observed within the cytoplasm and nuclei
of cell profiles surrounding myelinated axons
(Fig. 1C). To establish a first approximation of
the relative contribution of donor cell to host cell
remyelination, quantitative analysis was performed
on GFP immuno-EM experiments of two trans-
planted animals. A total of 136 axons were counted
in the lesions, and of these 120 were GFP+ cells.
Of the GFP+ cells, four (3.3%) showed no direct
contact with axons. The myelin-forming status of
five GFP+ cells (4.2%) could not be determined
because of poor membrane preservation. Of the
remaining 111 GFP+ cells, 90 cells (81%; 90 of 111)
showed a distinct myelin structure. The remaining
21 GFP+ cells (19%; 21 of 111) appeared to be in
varying stages of loose wrapping or ensheathing the
axons but did not show distinct compact myelin.
Within this same region, 16 myelinated axon pro-
files were detected, which were associated with my-
elin-forming cells that did not contain DAB
reaction product. An example of a non-GFP my-
elinated axon adjacent to a GFP+ myelinating cell
is shown in Figs. 1, D1 and D2. Note the presence
of a basement membrane surrounding both myelin-
ated axons. In frozen sections, OECs from donor
GFP rats (GFP-OEC) exhibited a robust distribu-
tion within the X-EB lesioned spinal cord dorsal
columns. The GFP-OECs were easily distinguished
at this time point by their green fluorescence
and were distributed longitudinally along the dor-
sal columns for
8 mm (Fig. 2A). GFP-OECs were
mostly confined to the lesion zone of the dorsal
funiculus. GFAP immunostaining indicated a near
absence of astrocytes within the lesion site, but in-
tense GFAP staining was observed at the outer
boundary of the lesion zone (Figs. 2B, C). In con-
trast to GFAP immunolabeling, immunostaining
for P0, a specific marker of peripheral myelin
(Greenfield et al., 1973), was primarily localized
within the lesion and transplantation site (Figs. 2D,
E). Ringlet-like P0 immunostaining was associ-
ated with GFP-OECs and was surrounded by
GFP+-OEC cytoplasm, indicating the transplanted
cells within the dorsal column lesion formed that
peripheral-type myelin. Immunostaining for P0 and
NF in coronal section reveals a central NF+ axonal
core surrounded by P0-identified myelin, which is
wrapped by cytoplasm of a GFP-OEC (Fig. 2E),
indicating that transplanted OECs remyelinate the
demyelinated axons.
Moreover, we also transplanted highly purified
OECs isolated from transgenic pigs expressing the
alpha1, 2 fucosyltransferase gene (H-transferase or
HT) into a demyelinated lesion of the African
green monkey spinal cord (Radtke et al., 2004).
Four weeks post-transplantation, robust remyeli-
nation was found in 62.5% of the lesion sites,
whereas there was virtually no remyelination in the
non-transplanted controls. This was the first dem-
onstration that xenotransplantation of character-
ized OECs into the primate spinal cord results in
remyelination. This together with the immuno-
histochemical demonstration of the grafted cells
within the lesioned area confirmed that remyeli-
nation was indeed achieved by OECs.
Nodal reconstruction of remyelinated spinal cord
axons
The restoration of rapid and secure impulse con-
duction after demyelination is dependent on the
acquisition of myelin sheaths and the clustering of
specific molecules within discrete domains of the
myelinated axon membrane. In myelinated axons,
voltage-gated sodium (Nav) channels are aggre-
gated in high density at nodes of Ranvier, whereas
Shaker-type potassium (Kv1) channels are sepa-
rated from nodal Nav channels by septate-like
paranodal junctions (Peles and Salzer, 2000;
Rasband and Trimmer, 2001; Girault and Peles,
2002). Of the seven Nav channel isoforms expressed
in nervous tissue (Goldin et al., 2000), Nav1.6 is the
predominant one at mature nodes in both the PNS
and CNS (Caldwell et al., 2000; Boiko et al., 2001)
following a transition from Nav1.2 (Boiko et al.,
2001; Kaplan et al., 2001; Jenkins and Bennett,
2002; Rios et al., 2003). The channel clustering
(Vabnick et al., 1997; Rasband et al., 1999) and the
transition from Nav1.2 to Nav1.6 (Boiko et al.,
2001; Rios et al., 2003) are dependent on interac-
tion of the axon with myelinating cells (Kaplan
 423

Fig. 2. Sagittal frozen sections through the lesion site demonstrate the distribution of transplanted GFP-OEC. Transplanted cells are
primarily confined to the lesion site. Some cells migrated into the deep white matter. The dashed line demarcates lesion edge (A).
Coronal frozen sections in the lesion show the presence of GFP-OEC within a lesion site. Transplanted cells survived primarily in the
dorsal funiculus. There was little GFAP staining within the lesion zone (B). GFAP-positive cells were present at the peripheral margin
of the lesion. These results indicate that few astrocytes are present in the transplant region, and that there is a preponderance of GFP-
OEC in the lesion zone (B, C). P0-immunostaining (red) of the frozen coronal section reveals that most axons remyelinated by the
transplanted OECs are surrounded by peripheral type of myelin. Red-P0 rings are associated with green cellular elements, indicating
that transplanted OECs remyelinate the demyelinated axons (D). Expansion of a cell indicated by an arrow (D inset). P0 and
neurofilament (NF) staining at lesion boundary (dashed line) and transplant zone (E). Higher magnification showing neurofilament-
defined axon cores surrounded by P0 myelin rings enwrapped by GFP-OECs (E2). B–E are coronal. Scale bars: 1 mm (A); 400 mm (B);
30 mm (C); 10 mm (D); 20 mm (A, inset); 10 mm (B, inset); 30 mm (E1); 10 mm (E2); 2.5 mm (E2, inset). Modified with permission from
Sasaki et al. (2006a).
424
Fig. 3. Nodal structure of remyelinated axons (A, B) at 3 weeks after OEC transplantation. Sagittal section showing a field of
myelinated axons interspersed among donor OECs. Note the node structure in the center of the field (A). High-power electron
micrograph showing node and paranodal loops (B1). B2 is an enlargement of the boxed area in B1. Scale bars ¼ 5 mm (A); 1 mm (B1);
0.5 mm (B2). Modified with permission from Sasaki et al. (2006a).
et al., 1997; Eshed et al., 2005). Remyelinated
axons display inappropriately short internodal
lengths (Gledhill and McDonald, 1977; Weiner
et al., 1980; Blakemore and Murray, 1981;
Hildebrand et al., 1985), indicating that new nodes
are formed. Despite their location at formerly
internodal sites, remyelinated PNS axons have
been shown to display high densities of Nav chan-
nels at nodes (Novakovic et al., 1996, 1998) and
Kv1 aggregations within juxtaparanodal domains
(Rasband et al., 1998). The expression and organi-
zation of specific isoforms of Nav and Kv1 channels
in remyelinated CNS axons have not been exam-
ined. Recently we reported that EM analysis of
spinal cords performed at 3 weeks after GFP-OEC
transplantation demonstrated distinct nodes of
Ranvier (Fig. 3). Large cytoplasmic and nuclear
compartments were present in cells associated
with the myelin profiles (Fig. 3A). Longitudinal
sections of the dorsal columns revealed well
formed nodal and paranodal regions; paranodal
loops from adjacent myelin-forming cells were
readily recognized flanking nodes (Figs. 3B1, B2).
Also, in immunohistochemical analysis we obser-
ved Nav1.6 staining at most nodes, whereas detect-
able Nav1.2 immunostaining was not apparent at
nodes (Fig. 4). In dorsal columns 3 weeks after
transplantation, virtually all nodes bounded by
GFP-OEC myelin sheaths exhibited Nav1.6 stain-
ing (Figs. 4A–D); similar to control spinal cord
axons, Nav1.2 immunolabeling was not observed at
any nodes (Figs. 4E–H). The Nav1.6 labeling was
localized to the nodal domain and was not obser-
ved in paranodal or juxtaparanodal regions or
beneath the myelin sheath in remyelinated axons,
suggesting that the transplanted GFP-OECs are
competent to contribute to the specific clustering
of Nav channels at nodes. As an additional deter-
minant of the ability of axons myelinated by
GFP-OECs to support the asymmetric organiza-
tion of ion channels within remyelinated nodal
regions, we examined the distribution of Kv1.2
in the juxtaparanodal region. Kv1.2, as well as
Kv1.1, form heteromultimers with Kv1.4 and Kvb2
(Wang et al., 1993; Rasband et al., 1998) and
have been shown previously to be aggregated in
juxtaparanodal regions of most spinal cord axons
(Rasband et al., 1999; Rasband and Trimmer,
2001). At 3 weeks (Figs. 4I–L), Kv1.2 is aggregated
within juxtaparanodal regions of the remyelinated
axons, with some nodes exhibiting incursion of
Kv1.2 channels into adjacent paranodal regions.
Kv1.2 was not observed within nodal areas in
these axons.
Transected spinal cord
Distribution and myelin formation by transplanted
GFP-OECs within the spinal cord
Spinal cord injuries without OEC transplants can
show limited SC-like myelination, presumably from
invasion of the injury site from endogenous SCs
(Brook et al., 1998; Imaizumi et al., 2000a, b;
Namiki et al., 2000; Takami et al., 2002) or possibly
from precursor cells. The degree to which OECs
can integrate into injury sites and survive and
 425

Fig. 4. Nav1.2 and Nav1.6 at GFP-OEC nodes in remyelinated dorsal columns. At 3 weeks after transplantation, Nav1.6 clustering is
displayed at most Caspr-delimited (A–D) nodes formed by GFP-OECs at 3 weeks. In contrast, Caspr-delimited nodes formed by GFP-
OECs do not exhibit Nav1.2 immunostaining (E–H). Merged images of A–C and E–G are shown in D and H, respectively. Juxta-
paranodal Kv1.2 immunolabeling at 3 weeks after GFP-OEC transplantation dorsal columns. Paranodes display Caspr staining (I)
that is flanked by Kv1.2 aggregations within juxtaparanodal regions. Merged images of I–K are shown in L, respectively. Scale
bars ¼ 10 mm. Modified with permission from Sasaki et al. (2006a).

whether they form myelin or facilitate endogenous
myelin repair mechanisms was controversial. OECs
in culture are diverse and exhibit characteristics of
astrocytes, SCs and oligodendrocytes. Moreover,
they express a number of trophic factors, transcrip-
tion factors and extracellular matrix molecules
(Ramon-Cueto and Avila, 1998; Chuah and West,
2002; Au and Roskams, 2003) that could facilitate
endogenous cell invasion, as well as angiogenesis
and activation of progenitor cells. A recent study
was unable to find evidence of myelination in the
compressed spinal cord by the OECs isolated from
426

Fig. 5. Montage image of sagittal frozen section showing distribution of GFP-OECs within and beyond the transection site (A).
Semithin plastic sections stained with methylene blue/Azure II through the transection site 5 weeks after transplantation of OECs.
Low-power micrograph showing completeness of the transection through the entire dorsal funiculus and beyond (B). EM from the
same lesion showing myelinated axons surrounded by a cellular element forming a tunnel (C). EM of anti-GFP immunoperoxidase
staining of OEC transplant (D). Reaction product can be seen in cytoplasmic regions of the myelin-forming cell but not in the myelin.
A crosscut section of axon showing cytoplastic reaction product. Enlargement of the boxed area is shown in D2. Note the presence of
extracellular fibrous elements in D1. Scale bars ¼ 250 mm (A); 0.75 mm (B); 5 mm (C); 1 mm (D1); 0.25 mm (D2). Modified with
permission from Sasaki et al. (2004).

embryonic day 18 rat, infected with a LacZ-
expressing retrovirus, and suggests that OEC trans-
plantation may facilitate endogenous SC invasion
into the lesion site (Boyd et al., 2004).
We transplanted genetically labeled GFP-OECs
into a dorsal funiculus transection model in the
rat. Five weeks after OEC transplantation, the
cells survived and distributed extensively within
the transection site and more limitedly beyond
the lesion site (Fig. 5A1). High-power micro-
graphs (Fig. 5A2) of AnkyrinG and Caspr stain-
ing demonstrate nodal and paranodal regions,

respectively, associated with some transplanted
GFP-OECs, suggesting forming a new node of
Ranvier. Methylene blue/Azure II semithin plastic
sections through this region of the lesion 5 weeks
after transplantation revealed greater structural
detail (Fig. 5B). As reported initially by Li et al.
(1997), small groups of myelinated axons were
often surrounded by a non-myelinating cell form-
ing tube-like structures around the myelinated
axons. This organization is not observed after en-
dogenous repair or after transplantation of SCs
(Imaizumi et al., 2000b). These profiles were ob-
served in both dorsal and ventral regions of the
lesion zone (Fig. 5B). EM further indicates that
the myelinated axons in the lesion zone are sur-
rounded by cytoplasmic extensions of cells form-
ing tunnels (Fig. 5C). These clusters of myelinated
axons were variable from animal to animal and
confined mostly to and near the lesion zone. Rings
of P0 labeling were observed in some regions of the
lesion and were often associated with GFP cellular
elements. EM examination of anti-GFP immuno-
peroxidase-reacted sections revealed that many
detectable GFP+ cells were in direct contact with
host axons. Reaction product was clearly evident
in the cytoplasm of many cells that formed well-
defined multi-laminate structures characteristic of
Fig. 6. Open-field locomotor scores for OEC transplant (n ¼ 20) and sham injection (n ¼ 6) groups tested 1 week before and for 5
weeks after transplantation. Modified with permission from Sasaki et al. (2004).
427
myelin (Fig. 5D). In a longitudinal section of a
myelinated axon, intense reaction product can be
seen in the cytoplasm of the myelin-forming cell.
Large cytoplasmic and nuclear regions as well as
the presence of basal lamina and extracellular fi-
brils (Fig. 5D) indicate a peripheral pattern of
myelination.
Taken together, we conclude that transplanted
GFP-OECs integrate into the injury site of a dor-
sal funiculus transection, distribute and associate
longitudinally with axons spanning the lesion site
and do form myelin.
Improved hindlimb locomotor function
All OEC-transplanted animals in this model system
exhibited a gradual improvement in hindlimb
locomotor function during the 5-week recovery
period (Fig. 6). The sham injection group recov-
ered to a Basso, Beattie and Bresnahan (BBB) lo-
comotor rating scale (Basso et al., 1995) of
9 and
did not display weight-bearing plantar stepping.
However, the OEC transplant group recovered to

18 on the BBB score and displayed consistent
weight-bearing plantar stepping. Statistical analysis
indicated that the open-field locomotor scores at
428

3, 4 and 5 weeks after injury and OEC transplan-

tation was significantly higher than sham injection.

Neuroprotection of dorsal corticospinal tract

A recent study demonstrates that primary motor

cortex (M1) pyramidal neurons undergo apoptotic
cell death after axotomizing SCI (Hains et al.,
2003). Moreover, earlier work has shown that cor-
ticospinal neurons become atrophic after spinal
cord transection (McBride et al., 1989). To explore
the possibility that OECs are neuroprotective for
injured corticospinal tract (CST) neurons, we trans-
planted OECs into the dorsal transected spinal cord
(T9) and examined M1 to assess apoptosis and
neuronal loss at 1 and 4 weeks post-transplantation.
Triple labeling with Hoechst, Fluorogold (FG), and
terminal deoxynucleotidyltransferase-mediated de-
oxyuridine triphosphate–rhodamine nick end-labe-
ling (TUNEL) identified putatively axotomized
pyramidal cells in the SCI group, with and with-
out spinal OEC transplantation. The neuronal sub-
set undergoing apoptosis was readily identified in 1-
week tissue. Within these cells, nuclear morph-
ologies of both apoptotic and surrounding normal
neurons and glia could be identified (Figs. 7A, B).
Hoechst staining was observed in all cells and uni-
formly filled the spherical nuclear compartment. In
a proportion of cells undergoing apoptosis, Hoe- Fig. 7. Hoechst 33342, Fluorogold (FG), and TUNEL triple
chst staining revealed abnormal morphology, in- labeling of corticospinal neurons 1 week after injury. Hoechst
cluding the formation of chromatin condensates staining of non-TUNEL-positive (arrows with tails) and
within a less-distinct nucleus (Hains et al., 2003). TUNEL-positive (arrowheads) neurons with corresponding
FG-backfilling are shown. In SCI+FG+OEC animals (B),
Merging of Hoechst, FG and TUNEL staining fewer TUNEL-positive FG-backfilled neurons are observed
(Figs. 7A, B) showed that irregular nuclear frag- compared with SCI+FG+DMEM (A). Insets in A show two
ments identified by Hoechst staining overlap se- TUNEL-positive neurons exhibiting nuclear compartment-
curely with FG- and TUNEL-positive cells, alization and formation of nucleosomes, hallmarks of apoptosis.
permitting the confident designation of an apopto- Quantification of neurons that are both TUNEL- and FG-
positive (C) reveals that OEC transplantation significantly
tic status. The nuclear morphology of TUNEL- (*Po0.05) reduces apoptotic cell death at 1 week. No evidence
positive neurons was identical to that seen in our of death was observed at any other time-point. Scale
earlier paper (Hains et al., 2003) (Fig. 7A, insets). bars ¼ 125 mm in A, B; 20 mm in inset in A. Modified with per-
Apoptotic (FG- and TUNEL-positive cells) pyram- mission from Sasaki et al. (2006b).
idal neurons were observed both after SCI+
FG+DMEM (Fig. 7A, arrowheads) and in the neurons was significantly (Po0.05) reduced by
SCI+FG+OEC group (Fig. 7B, arrowheads), but 45% (16.476.40 vs. 36.175.69) in the OEC trans-
the number of apoptotic neurons was significantly plant group compared with the SCI+FG+
reduced in the SCI+FG+OEC group (Fig. 7C). DMEM group (Fig. 7C). At 4 weeks, both the
At 1 week after injury, the number of apoptotic SCI+FG+DMEM and SCI+FG+OEC groups
 429

showed apoptotic activity comparable with sham
controls (0.9870.81 and 0.7570.56) (Fig. 7C).
We described the supraspinal effects of OEC trans-
plantation on apoptosis and cell survival of CST
neurons within the M1 cortex after transection
of their axons in the spinal cord. Our results indi-
cate that apoptosis of primary motor cortical neu-
rons is reduced and that cortical neuronal density is
increased after OEC transplantation. Enhanced
levels of brain-derived neurotrophic factor (BDNF)
were observed in the OEC transplanted lesion.
Thus, transplantation of OECs into injured spinal
cord has a neuroprotective effect on corticospinal
neurons. The relative contribution of this effect to
the observed functional improvement after OEC
transplantation is uncertain, but this data indicates
that OEC transplantation results in a larger pool of
surviving corticospinal neurons. Thus, OEC trans-
plantation into the injured spinal cord has distant
neuroprotective effects on descending cortical
projection neurons as well (Sasaki et al., 2006b).
OEC into spinal cord contusion injury
We also evaluated the in vivo fate of OECs trans-
planted into the contused rat spinal cord with the
weight-drop device developed at New York Uni-
versity (NYU impactor) (Li et al., 2004). Trans-
planted GFP-OECs were distributed widely within
transplanted spinal cords. In sagittal sections,
GFP-OECs were widely distributed along the
length of the lesion (Fig. 8A). Immuno-EM using
an anti-GFP antibody revealed that the identified
OECs made direct and extensive contact with host
axons (Figs. 8B, C). Transplanted GFP-OECs pro-
duced multi-laminate structures surrounding axons
in a one-to-one relationship with large amounts of
cytoplasm and basal lamina characteristic of pe-
ripheral myelin. The typical relationships of trans-
planted SCs and OECs with host axons were quite
different. While the overwhelming majority of GFP
positive cells in both transplant conditions were in
direct contact with host axons, GFP positive SCs

Fig. 8. Distribution of transplanted OECs into a contused rat spinal cord (A). Sagittal section of a segment of the spinal cord
8 mm
rostral and caudal to the injured area showed transplanted OECs concentrated near the center and distributed along the longitudinal
axes of the spinal cord at 3 weeks after transplantation (A), arrows indicated transplantation points. A2 and A3 are enlargements of
boxed areas in A1. Immuno EM for GFP revealed that many detectable GFP+ cells were in direct contact with host axons (Arrows).
Reaction product was clearly evident in the cytoplasm of many cells that formed well-defined multi-laminate structures characteristic
of myelin at 4 weeks after transplantation (8B, C). Scale bars ¼ 1 mm (A1); 100 mm (A2, A3); 5 mm (B); 1 mm (C).
430

cells were observed almost exclusively in one-to- strategy as an approach that may elicit some de-
one associations with host axons, with a high pro- gree of functional recovery, raising the possibility
portion of those contacts associated with myelina- that this approach might be useful in the treatment
tion, while GFP positive OECs typically wrapped of SCI. Experimental work indicates that cell
or engulfed several axons with a small percentage transplantation approaches can facilitate axonal
of contacts associated with myelination. regeneration, remyelination, neuroprotection and
possible neovascularization. Importantly, axons
remyelinated by transplanted OECs form appro-
Clinical perspectives
priate nodal sodium channel and conduction is
improved. Several clinical studies are ongoing us-
Clinical investigations in SCI using OEC trans-
ing cell therapy approaches for SCI (Senior, 2002;
plantation are in progress. A recent report tested
Dobkin et al., 2006). Future translational studies
the feasibility and safety of transplantation of
are highly expected to bridge the gap between
autologous OECs into the injured spinal cord in
basic and clinical research in OEC transplantation
human paraplegia in a single blind, Phase I clinical
for SCI.
trial (Feron et al., 2005). In this study, 1 year after
cell implantation, there were no medical, surgical
or other complications to indicate that the proce- Abbreviations
dure is unsafe. They conclude that transplantation
of autologous OECs into the injured spinal cord is BBB Basso, Bresnahan and Beattie (1995;
feasible and is safe up to 1 year post-implantation. i.e., and their rating scale for loco-
Long-term safety studies to compare neurological, motion)
functional and psychosocial outcomes are under- BDNF brain-derived neurotrophic factor
way. Furthermore, fetal brain tissue has been CNS central nervous system
transplanted into the lesions of more than 400 CST corticospinal tract
patients with SCI in China (Senior, 2002; Huang et DMEM Dulbecco’s modified Eagle’s medium
al., 2003). Dobkin et al. (2006) compared available EB ethidium bromide
reports from China with the experiences and EM electron microscopy
objective findings of patients who received preop- FG Fluorogold
erative and post-operative assessments before and K v1 Shaker-type potassium
up to 1 year after receiving cellular implants. They M1 primary motor cortex
concluded that the phenotype and the fate of the Nav voltage-gated sodium
transplanted cells, described as OECs in the China NYU New York University
study, are unknown. Perioperative morbidity and OEC olfactory ensheathing cells
lack of functional benefit were identified as the PNS peripheral nervous system
most serious clinical shortcomings. The proce- SC Schwann cells
dures observed did not attempt to meet interna- SCI spinal cord injury
tional standards for either a safety or efficacy trial. TUNEL terminal deoxynucleotidyltrans-
In the absence of a valid clinical trial protocol, ferase (TdT)-mediated deoxyuridine
Dobkin et al. (2006) suggest that physicians should triphosphate (dUTP)-rhodamine
not recommend this procedure to patients. nick end-labeling

Concluding remarks
Remyelination of the injured spinal cord is one of
the key elements for the functional recovery in SCI.
Extensive experimental rodent models of SCI de-
scribe the feasibility of OEC transplantation
Acknowledgments
This work was supported in part by the Depart-
ment of Veterans Affairs, the NIH, and the
National Multiple Sclerosis Society. The Center

for Neuroscience and Regeneration Research is a
collaboration of the Paralyzed Veterans of America
and the United Spinal Association with Yale Uni-
versity. We thank Heather Mallozzi and Margaret
Borelli for excellent technical assistance.
References
Au, E. and Roskams, A.J. (2003) Olfactory ensheathing cells of
the lamina propria in vivo and in vitro. Glia, 41: 224–236.
Bareyre, F.M., Kerschensteiner, M., Raineteau, O., Mettenlei-
ter, T.C., Weinmann, O. and Schwab, M.E. (2004) The
injured spinal cord spontaneously forms a new intraspinal
circuit in adult rats. Nat. Neurosci., 10: 269–277.
Barnett, S.C., Alexander, C.L., Iwashita, Y., Gilson, J.M.,
Crowther, J., Clark, L., Dunn, L.T., Papanastassiou, V.,
Kennedy, P.G. and Franklin, R.J. (2000) Identification of a
human olfactory ensheathing cell that can effect transplant-
mediated remyelination of demyelinated CNS axons. Brain,
123: 1581–1588.
Basso, D.M., Beattie, M.S. and Bresnahan, J.C. (1995) A sen-
sitive and reliable locomotor rating scale for open field testing
in rats. J. Neurotrauma, 12: 1–21.
Blakemore, W.F. and Crang, A.J. (1985) The use of cultured
autologous Schwann cells to remyelinate areas of persistent
demyelination in the central nervous system. J. Neurol. Sci.,
70: 207–223.
Blakemore, W.F. and Murray, J.A. (1981) Quantitative exam-
ination of internodal length of remyelinated nerve fibres in
the central nervous system. J. Neurol. Sci., 49: 273–284.
Boiko, T., Rasband, M.N., Levinson, S.R., Caldwell, J.H.,
Mandel, G., Timmer, J.S. and Matthews, G. (2001) Compact
myelin dictates the differential targeting of two sodium chan-
nel isoforms in the same axon. Neuron, 30: 91–104.
Boyd, J.G., Lee, J., Skihar, V., Doucette, R. and Kawaja, M.D.
(2004) LacZ-expressing olfactory ensheathing cells do not
associate with myelinated axons after implantation into the
compressed spinal cord. Proc. Natl. Acad. Sci. U.S.A., 17:
2162–2166.
Brook, G.A., Plate, D., Franzen, R., Martin, D., Moonen, G.,
Schoenen, J., Schmitt, A.B., Noth, J. and Nacimiento, W.
(1998) Spontaneous longitudinally orientated axonal regen-
eration is associated with the Schwann cell framework within
the lesion site following spinal cord compression injury of the
rat. J. Neurosci. Res., 53: 51–65.
Caldwell, J.H., Schaller, K.L., Lasher, R.S., Peles, E. and
Levinson, S.R. (2000) Sodium channel Na (v) 1.6 is localized
at nodes of Ranvier, dendrites, and synapses. Proc. Natl.
Acad. Sci. U.S.A., 97: 5616–5620.
Chuah, M.I. and West, A.K. (2002) Cellular and molecular bi-
ology of ensheathing cells. Microsc. Res. Tech., 58: 216–227.
Dobkin, B.H., Curt, A. and Guest, J. (2006) Cellular trans-
plants in China: observational study from the largest human
experiment in chronic spinal cord injury. Neurorehabil. Neu-
ral Repair, 20: 5–13.
431
Doucette, R. (1991) PNS-CNS transitional zone of the first
cranial nerve. J. Comp. Neurol., 312: 451–466.
Eshed, Y., Feinberg, K., Poliak, S., Sabanay, H., Sarig-Nadir, O.,
Spiegel, I., Bermingham, J.R. and Peles, E. (2005) Gliomedin
mediates Schwann cell–axon interaction and the molecular as-
sembly of the nodes of Ranvier. Neuron, 47: 215–229.
Feron, F., Perry, C., Cochrane, J., Licina, P., Nowitzke, A.,
Urquhart, S., Geraghty, T. and Mackay-Sim, A. (2005) Au-
tologous olfactory ensheathing cell transplantation in
human spinal cord injury. Brain, 128: 2951–2960.
Franklin, R.J., Gilson, J.M., Franceschini, I.A. and Barnett,
S.C. (1996) Schwann cell-like myelination following trans-
plantation of an olfactory bulb-ensheathing cell line into
areas of demyelination in the adult CNS. Glia, 17: 217–224.
Girault, J.A. and Peles, E. (2002) Development of nodes of
Ranvier. Curr. Opin. Neurobiol., 12: 476–485.
Gledhill, R.F. and McDonald, W.I. (1977) Morphological
characteristics of central demyelination and remyelination: a
single-fiber study. Ann. Neurol., 1: 552–560.
Goldin, A.L., Barchi, R.L., Caldwell, J.H., Hofmann, F., Howe,
J.R., Hunter, J.C., Kallen, R.G., Mandel, G., Meisler, M.H.,
Netter, Y.B., Noda, M., Tamkun, M.M., Waxman, S.G.,
Wood, J.N. and Catterall, W.A. (2000) Nomenclature of
voltage-dated sodium channels. Neuron, 28: 365–368.
Graziadei, P.P., Levine, R.R. and Graziadei, G.A. (1978)
Regeneration of olfactory axons and synapse formation in
the forebrain after bulbectomy in neonatal mice. Proc. Natl.
Acad. Sci. U.S.A., 75: 5230–5234.
Greenfield, S., Brostoff, S., Eylar, E.H. and Morell, P. (1973)
Protein composition of myelin of the peripheral nervous sys-
tem. J. Neurochem., 20: 1207–1216.
Hains, B.C., Black, J.A. and Waxman, S.G. (2003) Primary
cortical motor neurons undergo apoptosis following axoto-
mizing spinal cord injury. J. Comp. Neurol., 462: 328–341.
Hildebrand, C., Kocsis, J.D., Berglund, S. and Waxman, S.G.
(1985) Myelin sheath remodeling in regenerated rat sciatic
nerve. Brain Res., 358: 163–170.
Honmou, O., Felts, P.A., Waxman, S.G. and Kocsis, J.D. (1996)
Restoration of normal conduction properties in demyelinated
spinal cord axons in the adult rat by transplantation of ex-
ogenous Schwann cells. J. Neurosci., 16: 3199–3208.
Huang, H., Chen, L., Wang, H., Xiu, B., Wang, R., Zhang, J.,
Zhang, F., Gu, Z., Li, Y., Song, Y., Hao, W., Pang, S. and
Sun, J. (2003) Influence of patients’ age on functional recov-
ery after transplantation of olfactory ensheathing cells into
injured spinal cord injury. Chin. Med. J. (Engl.), 116:
1488–1491.
Imaizumi, T., Lankford, K.L., Waxman, S.G., Greer, C.A. and
Kocsis, J.D. (1998) Transplanted olfactory ensheathing cells
remyelinate and enhance axonal conduction in the demyelin-
ated dorsal columns of the rat spinal cord. J. Neurosci., 18:
6176–6185.
Imaizumi, T., Lankford, K.L., Burton, W.V., Fodor, W.L. and
Kocsis, J.D. (2000a) Xenotransplantation of transgenic pig
olfactory ensheathing cells promotes axonal regeneration in
rat spinal cord. Nat. Biotechnol., 18: 949–953.
432
Imaizumi, T., Lankford, K.L. and Kocsis, J.D. (2000b) Trans-
plantation of olfactory ensheathing cells or Schwann cells
restores rapid and secure conduction across the transected
spinal cord. Brain Res., 854: 70–78.
Jenkins, S.M. and Bennett, V. (2002) Developing nodes of
Ranvier are defined by ankyrin-G clustering and are inde-
pendent of paranodal axoglial adhesion. Proc. Natl. Acad.
Sci. U.S.A., 99: 2303–2308.
Kaplan, M.R., Cho, M.H., Ullian, E.M., Isom, L.L., Levinson,
S.R. and Barres, B.A. (2001) Differential control of clustering
of the sodium channels Na(v)1.2 and Na(v)1.6 at developing
CNS nodes of Ranvier. Neuron, 30: 105–109.
Kaplan, M.R., Meyer-Franke, A., Lambert, S., Bennett, V.,
Duncan, I.D., Levinson, S.R. and Barres, B.A. (1997)
Induction of sodium channel clustering by oligodendrocytes.
Nature, 386: 724–728.
Kato, T., Honmou, O., Uede, T., Hashi, K. and Kocsis, J.D.
(2000) Transplantation of human olfactory ensheathing cells
elicits remyelination of demyelinated rat spinal cord. Glia, 30:
209–218.
Keirstead, H.S. (2005) Stem cells for the treatment of myelin
loss. Trends Neurosci., 28: 677–683.
Keyvan-Fouladi, N., Li, Y. and Raisman, G. (2002) How do
transplanted olfactory ensheathing cells restore function?
Brain Res. Brain Res. Rev., 40: 325–327.
Keyvan-Fouladi, N., Raisman, G. and Li, Y. (2003) Functional
repair of the corticospinal tract by delayed transplantation of
olfactory ensheathing cells in adult rats. J. Neurosci., 23:
9428–9434.
Kocsis, J.D. and Sasaki, M. (2005) Transplantation of periph-
eral myelin-forming cells to repair demyelinated axons.
In: Waxman S.G. (Ed.), Multiple Sclerosis as a Neuronal
Disease. Elsevier, San Diego, pp. 421–433.
Li, B., Lankford, K.L. and Kocsis, J.D. (2004) Distribution
and migration of GFP-expressing mouse Schwann cells and
olfactory ensheathing cells transplanted into contused spinal
cord of the rat. Abstract of 34th Annual meeting, Society for
Neuroscience.
Li, Y., Field, P.M. and Raisman, G. (1997) Repair of adult rat
corticospinal tract by transplants of olfactory ensheathing
cells. Science, 277: 2000–2002.
Li, Y., Field, P.M. and Raisman, G. (1998) Regeneration
of adult rat corticospinal axons induced by transplanted
olfactory ensheathing cells. J. Neurosci., 18: 10514–10524.
Lu, J., Feron, F., Mackay-Sim, A. and Waite, P.M. (2002)
Olfactory ensheathing cells promote locomotor recovery
after delayed transplantation into transected spinal cord.
Brain, 125: 14–21.
McBride, R.L., Feringa, E.R., Garver, M.K. and Williams Jr.,
J.K. (1989) Prelabeled red nucleus and sensorimotor cortex
neurons of the rat survive 10 and 20 weeks after spinal cord
transection. J. Neuropathol. Exp. Neurol., 48: 568–576.
Namiki, J., Kojima, A. and Tator, C.H. (2000) Effect of
brain-derived neurotrophic factor, nerve growth factor, and
neurotrophin-3 on functional recovery and regeneration after
spinal cord injury in adult rats. J. Neurotrauma, 17:
1219–1231.
Novakovic, S.D., Deerinck, T.J., Levinson, S.R., Shrager, P.
and Ellisman, M.N. (1996) Cluster of axonal Na+ channels
adjacent to remyelinating Schwann cells. J. Neurocytol., 25:
403–412.
Novakovic, S.D., Levinson, S.R., Schachner, M. and Shrager, P.
(1998) Disruption and reorganization of sodium channels in
experimental allergic neuritis. Muscle Nerve, 21: 1019–1032.
Peles, E. and Salzer, J.L. (2000) Molecular domains of myelin-
ated axons. Curr. Opin. Neurobiol., 10: 558–565.
Plant, G.W., Christensen, C.L., Oudega, M. and Bunge, M.B.
(2003) Delayed transplantation of olfactory ensheathing glia
promotes sparing/regeneration of supraspinal axons in the
contused adult rat spinal cord. J. Neurotrauma, 20: 1–16.
Radtke, C., Akiyama, Y., Brokaw, J., Lankford, K.L.,
Wewetzer, K., Fodor, W.L. and Kocsis, J.D. (2004)
Remyelination of the nonhuman primate spinal cord by
transplantation of H-transferase transgenic adult pig
olfactory ensheathing cells. FASEB J., 18: 335–337.
Raisman, G. (2001) Olfactory ensheathing cells – another
miracle cure for spinal cord injury? Nat. Rev. Neurosci., 2:
369–375.
Ramer, L.M., Au, E., Richter, M.W., Liu, J., Tetzlaff, W. and
Roskams, A.J. (2004) Peripheral olfactory ensheathing cells
reduce scar and cavity formation and promote regeneration
after spinal cord injury. J. Comp. Neurol., 473: 1–15.
Ramon-Cueto, A. and Avila, J. (1998) Olfactory ensheathing
glia: properties and function. Brain Res. Bull., 46: 175–187.
Ramon-Cueto, A., Cordero, M.I., Santos-Benito, F.F. and
Avila, J. (2000) Functional recovery of paraplegic rats and
motor axon regeneration in their spinal cords by olfactory
ensheathing glia. Neuron, 25: 425–435.
Ramon-Cueto, A., Plant, G.W., Avila, J. and Bunge, M.B.
(1998) Long-distance axonal regeneration in the transected
adult rat spinal cord is promoted by olfactory ensheathing
glia transplants. J. Neurosci., 18: 3803–3815.
Rasband, M.N., Peles, E., Trimmer, J.S., Levinson, S.R., Lux,
S.E. and Shrager, P. (1999) Dependence of nodal sodium
channel clustering on paranodal axoglial contact in the
developing CNS. J. Neurosci., 19: 7516–7528.
Rasband, M.N. and Trimmer, J.S. (2001) Developmental
clustering of ion channels at and near the node of Ranvier.
Dev. Biol., 236: 5–16.
Rasband, M.N., Trimmer, J.S., Schwarz, T.L., Levinson, S.R.,
Ellisman, M.N., Schachner, M. and Schrager, P. (1998)
Potassium channel distribution, clustering, and function in
remyelinating rat axons. J. Neurosci., 18: 36–47.
Rios, J.C., Rubin, M., St Martin, M., Downey, R.T., Einheber,
S., Rosenbluth, J., Levinson, S.R., Bhat, M. and Salzer, J.L.
(2003) Paranodal interactions regulate expression of sodium
channel subtypes and provide a diffusion barrier for the node
of Ranvier. J. Neurosci., 23: 7001–7011.
Rosenzweig, E.S. and McDonald, J.W. (2004) Rodent models
for treatment of spinal cord injury: research trends and
progress toward useful repair. Curr. Opin. Neurol., 17(2):
121–131.
Sasaki, M., Black, J.A., Lankford, K.L., Tokuno, H.A.,
Waxman, S.G. and Kocsis, J.D. (2006a) Molecular

reconstruction of nodes of Ranvier after remyelination by
transplanted olfactory ensheathing cells in the demyelinated
spinal cord. J. Neurosci., 26: 1803–1812.
Sasaki, M., Hains, B.C., Lankford, K.L., Waxman, S.G. and
Kocsis, J.D. (2006b) Protection of corticospinal tract neurons
after dorsal spinal cord transection and engraftment of
olfactory ensheathing cells. Glia, 53: 352–359.
Sasaki, M., Lankford, K.L., Zemedkun, M. and Kocsis, J.D.
(2004) Identified olfactory ensheathing cells transplanted into
the transected dorsal funiculus bridge the lesion and form
myelin. J. Neurosci., 24: 8485–8493.
Schwab, J.M., Brechtel, K., Mueller, C.A., Failli, V., Kaps, H.P.,
Tuli, S.K. and Schluesener, H.J. (2006) Experimental strate-
gies to promote spinal cord regeneration: an integrative per-
spective. Prog. Neurobiol., 78: 91–116.
Senior, K. (2002) Olfactory ensheathing cells to be used in
spinal-cord repair trial. Lancet Neurol., 1: 269.
433
Takami, T., Oudega, M., Bates, M.L., Wood, P.M., Kleitman,
N. and Bunge, M.B. (2002) Schwann cell but not olfactory
ensheathing glia transplants improve hindlimb locomotor
performance in the moderately contused adult rat thoracic
spinal cord. J. Neurosci., 22: 6670–6681.
Vabnick, I., Messing, A., Chiu, S.Y., Levinson, S.R., Schachner,
M., Roder, J., Li, C., Novakovic, S. and Shrager, P. (1997)
Sodium channel distribution in axons of hypomyelinated and
MAG null mutant mice. J. Neurosci. Res., 50: 321–336.
Wang, H., Kunkel, D.D., Martin, T.M., Schwartzkroin, P.A.
and Tempel, B.L. (1993) Heteromultimeric K+ channels in
terminal and juxtaparanodal regions of neurons. Nature, 365:
75–79.
Weiner, L.P., Waxman, S.G., Stohlman, S.A. and Kwan, A.
(1980) Remyelination following viral-inducted demyelina-
tion: ferric ion-ferrocyanide staining of nodes of Ranvier
within the CNS. Ann. Neurol., 8: 580–583.
This page intentionally left blank
 Subject Index

4-aminopyridine (4-AP) 220, 221 astrocyte swelling 66, 189

a-CaMKII 153 astrocytes 21, 31, 34, 61–70, 72, 73, 83, 88, 89, 98,
aII-spectrin 333, 413, 414 103, 117, 121, 129, 158, 176, 187–190, 196,
acceleration input 29 227, 228, 317, 318, 320, 331–333, 342, 343,
acceleration/deceleration 68, 158, 264, 269–271, 371, 420, 422, 425
284 astrocytic cascades 34
acetylcholine 62, 274 reactive astrocytes 304
acidosis 61, 65, 66, 113, 120, 121 astroglia 61, 175, 282, 318, 402 (see also glial cells)
acute spinal cord injury study 217, 219, 222 athletes 253–258, 263, 280, 282
acute subdural hematoma (ASDH) 118, 239, 266, ATP 33, 34, 62–65, 69, 70, 72, 85, 87, 111,
293, 294, 296, 298, 300 (see also subdural 113–115, 117, 121, 122, 164, 279, 280, 318,
hematoma) 355, 356, 358, 403
Akt 69, 343, 359, 360 autophagy 82, 84, 85 (see also cell death)
Alzheimer’s disease 81, 253, 255, 303–311, 346 autoregulation 112, 113, 131, 210–212, 281, 282
(see also neurodegenerative disease) metabolic 112, 119
AMPA receptor 68, 70, 155, 159–162, 164, 279, pressure 112, 119
332 viscosity 112, 119
amyloid precursor protein (APP) 48, 86–89, 130, axolemma 52, 175, 271, 285
255, 272–274, 303–308, 310, 311 axonal damage 49, 90, 144, 267, 269, 271, 275, 278
amyloid-b peptides 304 axotomy 43, 48, 263
antibody microarray 410, 411 delayed 273, 274
aphasia 346–348 primary 285
apnea 281, 295, 298, 299, 300 secondary 263, 271, 273, 274, 277–279, 285
APOE genotype 303, 310, 311
apolipoprotein E 255, 303 BACE inhibitors 305
apolipoprotein E4 310 Bad 47, 359
apoptosis inducing factor (AIF) 84, 87 basal ganglia 99, 295–297, 306
apoptosis 35, 43–45, 47, 67, 81–88, 90, 121, 127, Bax 47, 84, 86
128, 218, 223, 224, 283, 319, 343, 355, 403, Bcl-2 47, 69, 70, 72, 84, 86, 343
415, 428, 429 (see apoptotic) Bergmann glia 331–333
apoptotic 34, 35, 43–48, 67, 81–87, 126, 128, 131, big black brain 293–295, 297–300
132, 135, 223, 300, 304, 342, 358, 359, 403, biochemical markers 130, 282, 317, 401, 406, 409,
428 414, 415
apparent diffusion co-efficient (ADC) mapping biomarkers 30, 35, 317, 401, 403, 411–415
236 biomechanical mechanisms 28
aquaporins 185 biomechanics 13, 18, 23–25, 27, 329, 335
arachidonic acid 175, 357 acceleration 17
arterio venous difference of oxygen (AVDO2) 112, deceleration 17
119, 120, 208, 210 deformations 14

435
436

force 14 gunshot wounds 4

shear strain 15 traffic injuries 4
strains 14 cell culture 63, 85, 175, 256, 305, 318, 319, 343,
stress 14 356, 369, 376, 379, 381
blood brain barrier (BBB) 88, 97–99, 102, 104, 115, cell death (CD) 13, 20, 21, 23, 24, 33–35, 43–48, 66,
117, 126, 136, 143, 144, 158, 186, 188, 190, 67, 72, 81–87, 114, 122, 125, 126, 128, 131,
237, 246, 282, 317, 319, 321, 356, 359, 395, 132, 135, 144, 147, 153, 164, 218, 255, 257,
402 263, 279, 304, 306, 319, 327, 329–331, 335,
blunt trauma 158, 268, 270 356, 359, 402, 403, 428 (see also apoptosis,
bone marrow stem cells (BMSCs) 225 necrosis, autophagy)
bone marrow stromal cells 217, 226, 227 cell membrane damage 19, 20
brain derived neurotrophic factor (BDNF) 62, cellular 43
229, 343, 359, 429 mechanisms 62, 147, 156, 157, 259, 405
brain edema 28, 86, 187, 235, 356, 393, 395, 396, morphology 61
398 response 13, 16, 18, 19, 21, 22, 43, 132, 177, 284
brain tissue oxygen tension 113, 207–209 cerebellar injury 327
brain water 185–189, 191, 192, 356, 396 cerebellum 44, 62, 65, 175, 256, 257, 327–329, 331,
bromdeoxyuridine (BrdU) 321, 322, 369, 370 333, 335
burst firing 195, 196, 200, 201 cerebral blood flow (CBF) 111, 112, 186, 208, 236,
263, 278, 281, 296
calcineurin 53, 88, 280 cerebral blood volume (CBV) 112, 113
calcium 31–34, 46, 47, 52, 53, 55, 62–64, 66, 68–70, cerebral contusion 27, 66, 235, 266, 283, 405
82, 87, 98, 114, 121, 135, 151, 153, 158, 218, cerebral cortex 48, 83, 130, 135, 145, 146, 165, 257,
274, 275, 279, 280, 304, 318, 319, 331, 402, 307, 327, 335
403 cerebral edema 102, 120, 185–188, 190, 246, 279
calcium ATPase 273, 279 (see also edema)
calcium dysregulation 46, 47, 52, 53 cytotoxic edema 185, 188–190, 192
calcium homeostasis 122, 279, 304, 318, 402 vasogenic edema 185, 188, 190–192
calcium oscillations 62 cerebral energy metabolism 114, 115
calcium signaling 62 cerebral ischemia 65–67, 87, 120, 127, 130, 136,
calcium waves 62 396 (see also ischemia)
intracellular Ca2+ 46, 68 cerebral metabolic rate of oxygen (CMRO2) 112,
intracellular stores 62 119, 120, 208, 210, 281
calpain 45–47, 49, 52, 53, 81, 82, 85–88, 91, cerebral oxygen tension 111
98, 131, 135, 218, 279, 280, 333, 403, 411, cerebral oxygen transport 111, 118
414 cerebral oxygenation 121, 207, 209
calpain inhibitors 87, 88, 91 cerebral perfusion pressure 112, 120, 207
calpain proteolysis 414, 415 cerebral tissue PO2 111, 113, 121
caspases 45–47, 81, 84, 85, 122, 218, 304, 306, 403 cerebrospinal fluid (CSF) 98, 119, 185, 266, 299,
caspase-3 33, 47, 67, 83–85, 87, 89, 132, 306, 307, 320, 386, 406, 414
358, 403, 414, 415 cerebrovascular permeability 237, 238
caspase-12 83, 84 cerebrovascular reactivity 112
caspase-3 proteolysis 414, 415 carbon dioxide (CO2) reactivity 112
caspase-inhibitor 85 Cethrins 217, 223, 224, 229 (see also Rho
causes of neurotrauma 3–6 antagonist)
accidents 4 c-fos 343
falls 4 CGRP 99, 101, 102, 196
firearm injuries 5 child abuse 253, 254, 293

chloride 98
chondroitin sulfate proteoglycan (CSPG) 132
citric acid cycle 111, 115, 117, 118
c-jun 343
cognitive dysfunction 253–255, 257
cognitive performance 320, 321, 345
collagen gel 385–388
community costs 6
computer tomography (CT) 119, 402
concussions 254, 264
controlled cortical impact (CCI) 67, 86, 88, 90,
127, 271, 319, 395, 405, 412
contusion volume 136, 283, 284, 395, 396
contusions 17, 27, 44, 45, 98, 120, 207, 235,
237–239, 241, 266, 270, 283, 298, 395, 396
corpus callosum 55, 126, 266, 267, 334, 342, 371,
378
cortical circuits 200
cortical contusion 144, 270
craniectomy 212, 299, 393–398
critical time frames 244
crossed cerebellar diaschisis (CCD) 329
cross-tolerance 256, 257, 353, 356, 358, 359
CSF production 185–187, 192
CT scan 4, 6, 119, 120, 239, 243, 244, 246–248, 270,
283, 294
0 0
cyclic guanosine 3 ,5 -monophosphate (cGMP)
171, 173
cGMP synthesis 174
cyclosporine A (CsA) 81, 88
cysteine proteases 43, 45, 46, 52, 53,
403 (see also calpains and caspases)
cytochrome C 46, 52, 83, 122, 223
cytokines 87, 89, 127, 176, 225, 279, 358, 359, 368,
410
IL-1b 127
IL-6 127
TNFa 127
cytoskeleton 31, 46, 52, 53, 87, 272, 274, 275, 277,
279, 285, 318
decompression craniectomy 299, 393–398
dementia pugilistica 254, 255, 258, 306
demyelination 125, 128, 131, 132, 220, 221, 227,
228, 419, 420, 422
depolarization 33, 55, 62, 67, 68, 155–157, 164,
274, 276, 278–280, 331, 403
dexanabinol 244, 246, 247
437
diffuse axonal injury (DAI) 17, 27, 31, 49, 53, 55,
126, 264, 266, 298, 306, 402
diffuse head injuries 270
diffuse injuries 17, 27, 29, 43, 44, 55, 56, 90, 335
(see also DTBI)
diffuse traumatic brain injury (DTBI) 43–49, 53,
55, 56 (see also diffuse injuries)
drug delivery 369, 385–388
delivery strategies 385
injectable delivery 386
intrathecal drug delivery 385
dynamic brain injury 404
dynamic loading 15, 16, 265
edema 28, 65, 66, 86, 88, 97–99, 101–103, 119, 120,
126, 131, 136, 185–192, 235, 237–240, 246,
296, 356, 359, 393, 395, 396, 402, 405, 419
(see also cerebral edema)
cytotoxic edema 65, 98, 119, 120, 126, 185, 188,
189, 190, 192
vasogenic edema 97–99, 102, 104, 185, 190–192,
237, 238, 396
electrophysiology 62, 143, 144, 156, 158, 159, 165,
327
energy 45, 65, 82, 84, 85, 87, 90, 111, 114, 115, 117,
118, 121, 268–270, 280, 281, 304, 319, 358,
404, 406
energy metabolism 111, 114, 118, 121, 319
enzyme-linked immunosorbant assays (ELISA) 413
epidemiology 3, 264, 303, 328
EURODERM study 309
MIRAGE study 309, 310
epidermal growth factor 385
erythropoietin (Epo) 353, 357, 358
European Union Directive 244, 247
excitatory amino acid (EAA) 46, 90, 97, 175, 279,
299, 342, 402
excitatory neurotransmitters 279
excitotoxicity 33, 65, 67, 218, 223, 330, 346, 403
experimental models 13, 27–29, 90, 115, 187, 188,
222, 248, 253–256, 285, 296–298, 300, 356,
357, 381, 394, 419
cellular injury models 14
compression injury 18, 385–388
contusion 18
fluid percussion injury 18, 72, 144, 271, 277, 319,
320, 327, 405
of stroke 369
438

piglet subdural injection models 298 GLUT transporters 115

rodent subdural hematoma model 296 GLUT-1 115
weight drop 18, 30, 155, 156, 254, 255, 257, 405, GLUT-3 115
429 glutamate 17, 32, 46, 64–65, 67, 117, 148, 159–162,
extracellular signal regulated kinase (ERK) 218, 223, 254, 279, 296, 318, 319, 332, 342,
34, 69 402, 403
ERK1 343 glutamate receptors 31, 32, 63, 65, 160, 162, 279,
ERK2 343 318
N-methyl D-aspartate receptors (NMDAR)
Fampridines 220, 221, 229 (see also 31–35, 159, 160, 161
4-aminopyridine or 4-AP) a-Amino-3-hydroxy-5-methyl-4-isoxazole-
fas ligand 84 propionic acid receptors (AMPAR) 32, 33,
field extracellular postsynaptic potentials (fEPSP) 160–162, 164
147 metabotropic glutamate receptors (mGluRs) 32,
finite element analysis (FEA) 13, 21, 23, 24 63, 160, 318
fluid percussion injury 18, 144, 271, 277, 279, 319, glutamate transporters 65, 332
320, 327, 405 (see also lateral fluid GLT-1 65
percussion injury) GLAST 65
focal head injuries (FHI) 270 glycolysis 65, 84, 111, 115–117, 121, 279,
focal injuries 16, 27, 44, 270, 331 280
free radicals 35, 46, 47, 85, 136, 159, 177, 279, 342, GM-1 218, 222, 228
357, 402 granulocyte colony stimulating factor (G-CSF)
funding 7, 8, 24, 259, 391 367, 371, 375
fura-2 158 growth factor 135, 226, 320, 358, 368, 379, 385,
391
g-aminobutyric acid (GABA) 33, 62, 146, 148, 150,
156, 162, 163, 171 Hagen–Poiseuille equation 112, 119
GABAA receptor 147, 149–151, 158, 162, 342 head injury 4, 6, 17, 18, 30, 47, 87, 88, 111,
gap junctions 63, 66, 72, 73 118–122, 257, 264–268, 271, 274, 280–284,
gene expression 30, 34, 47, 98, 100, 329, 331, 293, 294, 298, 299, 306, 307, 309, 310, 329,
398 356–357, 402, 404
GFP 371, 420, 422, 427, 429, 430 minor 6
Glasgow Coma Score (GCS) 281, 413 moderate 6
Glasgow outcome scale (GOS) 240, 328 severe 6
Glasgow coma scale (GCS) 240 heat acclimation 257, 353–360
glia 30, 33, 34, 62–64, 67, 81–83, 86, 98, 103, 117, heat shock proteins 331, 355
118, 144, 158, 160, 186, 225, 256, 284, helmets 7, 14, 97, 263, 264
331–333, 342, 428 bicycle helmets 7
glial cells 30, 61, 64, 98, 114, 144, 160, 228, hockey helmets 268
236, 279, 283, 304, 317, 333, 340, motorcycle helmets 7
342, 358 hematoma 16, 27, 44, 66, 67, 118–120, 209, 210,
glial fibrillary acidic protein (GFAP) 61, 64, 67, 228, 239, 266, 268, 270, 293–296, 298, 300
134, 196, 282, 422 hemicraniectomy 299, 300
glial scar 64, 226, 331, 379 high-throughput immunoblotting 411
glial swelling 44, 66, 98, 126 hippocampal 20, 63, 66, 85, 86, 144–148, 156,
glucose 65–67, 111, 115, 117, 120, 122, 278, 280, 160, 162, 164, 253, 256, 304, 306, 308,
281, 319 319–321, 341, 343, 355, 411
glucose metabolism 120, 278, 281 (see also hippocampus)

hippocampus 44, 45, 48, 62, 64, 67, 72, 83, 86, 89,
99, 101, 126, 136, 145–147, 151, 153, 161,
162, 225, 254–257, 307, 308, 310, 320, 321,
327, 329, 341–343, 346
CA1 145
CA3 145
dentate gyrus (DG) 145
perforant pathway 153
histopathology of cerebral contusion 235
hormonal cycling 339, 346
hormonal status 339–341, 344–348
hyaluronan and methylcellulose (HAMC) 385,
388–391
hyaluronan 385, 388
hydrogel 18, 367, 379, 385, 386, 391
hydrolytic enzymes 43, 47, 56
hyperglycolysis 120, 280
hyperventilation 98, 112–114, 119, 120,
210–212
hypotension and hypoxia 118
hypothermia 81, 90, 91, 98, 135, 136, 211, 212, 247
hypoxia 65–67, 90, 102, 115, 118, 119, 121,
122, 210, 211, 296, 298, 300, 301, 343,
357–359
hypoxia inducible factor-1 353, 357, 358
ice hockey 263, 265, 267–271, 275, 286
ICP and cerebral circulation 119, 120
immunoreactivity 64, 67, 87, 89, 99, 100, 103, 177,
255
impact loading 16, 266
impulsive loading 16, 266, 267
in vitro models 18, 21, 24, 27–30, 34, 35, 68, 143,
144, 158, 253, 257
cell/substrate stretch model 30
scratch model/laceration 30
weight drop/compression 30
in vitro TBI models 14, 18, 159
(see also experimental models)
in vivo animal models 17, 278
(see also experimental models)
in vivo imaging 35, 377, 378
in vivo models 17, 18, 21, 27–29, 31, 33, 143, 144,
253, 354, 358
cortical impact model 29, 86
fluid percussion model 29
incapacity 244, 246
acute incapacity 247
439
incidence of spinal cord injury 4
incidence of traumatic brain injury 4
infancy 293, 294, 298, 344
inflammation 98, 127, 128, 135, 160, 176, 178, 188,
218, 224, 300, 340, 346, 353, 357, 359, 388,
410
neurogenic inflammation 97, 100–104
inflammatory processes 127, 130, 134, 348
inflammatory response 45, 46, 101, 128, 130, 136,
176, 177, 224, 284, 357, 389, 390, 402
informed consent 243–248
inhibitory postsynaptic current (sIPSC) 150,
156
inositol-1,4,5 trisphosphate (IP3) 32, 34, 68,
69, 72
IP3 receptor 63
intensive care 6, 118, 243, 249, 281, 299, 415
interleukin-1 127
interleukin-1b (IL-1b) 223, 359
international neurotrauma society (INTS) 3, 8
intracranial pressure (ICP) 6, 28, 49, 66, 98, 112,
118, 120, 186, 207, 235, 281, 296, 299, 300,
393, 403
monitoring 111
ionic fluxes 278, 281
ionic homeostasis 31, 33, 45, 66, 82, 111, 144, 164,
218, 263
ischemia 64–68, 72, 87, 88, 102, 120, 121, 127, 130,
136, 158, 207, 209, 210, 222, 236, 254, 256,
282, 284, 296, 298, 300, 322, 329, 343, 346,
354, 356, 357, 359, 396, 419
ischemic preconditioning (IPC) 256, 343
jugular bulb oximetry 208
jugular bulb oxygen saturation (SjvO2) 111, 120,
121, 207–212
kainate receptors 148
laceration 30, 83, 97, 131, 266
lactate 65, 84, 111, 113, 115, 117, 120, 121, 210,
211, 280, 319
lactate-to-pyruvate ratio 121, 211
laser doppler flowmetry 111
lateral fluid percussion injury (LFPI) 72, 144, 145,
147, 148, 150–153, 155, 156, 164, 279 (see
also fluid precursion injury)
lesion progression 131, 235
440

loads 14–19, 266 mild head injury 6, 274, 309

dynamic loading 15, 16, 265 mild injury 73, 144, 158, 160, 253, 254, 263
impact loading 16, 266 miniature inhibitory postsynaptic currents
impulsive loading 16, 266, 267 (mIPSCs) 150, 151, 156
static loading 15 minocycline 217, 223
long-term depression (LTD) 151, 153 minor traumatic brain injury (mTBI) 263 (see also
long-term potentiation (LTP) 149, 151–153, 161, mild injury)
257, 320, 341 minor head injury 6 (see also mild head injury)
low molecular weight antioxidants (LMWA) mitochondria 33, 34, 44, 52, 68, 72, 84–86, 90,
357 113–115, 117, 118, 121, 271, 275, 279, 280
lucifer yellow 19, 153 electron transport chain 85, 117
lysosomal enzymes 47, 48 (see also hydrolytic mitochondrial function 111, 121, 122, 159
enzymes) mitochondrial permeability transition pore
cathepsin 47 (MPTP) 84, 122
mitogen activated protein kinase (MAPK) 33–35,
macrophage 98, 128, 130, 134, 176–178, 224, 225, 69
227, 370, 373, 381, 389, 420 MK-801 159
magnesium 31, 98, 102, 103 Monro–Kellie doctrine 119
magnetic resonance imaging (MRI) 127, 134, 227, morphological injury 97
228, 235, 270, 283, 296, 341, 348, 367, 368, morris water maze (MWM) 145, 254, 255, 320
377, 381, 402, 414 multi modal monitoring 207
diffusion-weighted MR 236 myelin 129, 136, 137, 220, 223–226, 228, 274,
marker (see biomarker) 282, 342, 371, 420–427, 429
mass spectrometry 401, 403, 406, 407, 409, 411, (see also remyelination, demyelination)
415
tandem mass spectrometry 409, 410 Na+-K+ ATPase 65, 164
matrix metalloproteases 85 Na+/K+ ATPase activity 164
mechanical permeability 32 nanoparticles 369, 373, 375, 377, 381
mechanical response 13–16, 21, 23, 24 iron oxide 367, 368, 371, 379
mechanoporation 32, 48, 49, 84, 88, 98 near infra red spectroscopy (NIRS) 111, 207,
dextrans 48, 50, 51, 53 209–211
horseradish peroxidase 48 necrosis (see necrotic)
membrane resealing 43, 49 aponecrosis 45
mediators 61, 87, 98, 102, 130, 224, 226, 279, 359, necrotic 34, 43–49, 67, 82, 85–88, 90, 126, 218, 235,
360 238–241, 395, 403, 419
metabolic cascade 278, 357 nerve growth factor (NGF) 226, 343
metabotropic glutamate receptors (mGluRs) 32, NeuN 322, 370
34, 63, 69, 70, 160, 318 neural progenitor cells (NPCs) 225, 226, 369
mGluR1 160 neural stem cells (NSCs) 136, 225, 226, 320
mGluR5 160 neurodegenerative disease 45, 125, 253, 255,
methylcellulose 385, 388 256, 403 (see also Alzheimer’s disease,
methylprednisolone 135, 178, 217, 218, 221, 222, Parkinson’s disease)
228, 229, 385 neuroexcitation 46
MicroBeads 367, 377, 378, 381 neurofibrillary tangles 304, 307, 308, 310
microdialysis 46, 111, 120, 121, 175, 211, 278, 296 neurofilaments 46, 88, 271–274, 277, 278, 280, 285,
microglia 31, 35, 89, 98, 130, 176, 196, 304, 329, 379
359, 370, 373, 389, 402 neurogenesis 225, 226, 320–322, 340
microtubules 271–274, 277, 279, 280, 285 neurokinin A (NKA) 99–102

neurokinin receptors 100 (see also tachykinin
receptors)
neuron specific enolase (NSE) 68, 256, 282, 283
neuronal cascades 33
neuronal transmitter 62, 171, 173
neuropeptides 97–102, 104
neuroprotection 3, 61, 64, 217, 223, 342–344, 353,
354, 356–360, 428, 430
neuroproteomics 401, 403, 409, 411, 413, 415
neurotransmitter receptors 62, 158
NF200 333
NF-kappaB 130
nicotinamide adenine dinucleotide
phosphate diaphorase (NADPHd)
171–173, 177
nitric oxide 46, 88, 89, 98, 101, 113, 136, 171–179,
223
nitric oxide synthase (NOS) 46, 88, 171, 318
brain nitric oxide synthase (bNOS) 172
calcium-dependent nitric oxide synthase (cNOS)
176
inducible NOS (iNOS) 136, 176
neuronal nitric oxide synthase (nNOS) 175
NK1 receptor antagonists 103, 104
NMDA antagonists 34
NMDA receptor 46, 64, 68, 84, 148, 159, 160, 162,
175, 279, 341
N-methyl-D-aspartate (NMDA) 148, 175, 279
nodal reconstruction 422
nogo 223, 224
anti-NOGO monoclonal antibodies 217
olfactory ensheathing cells (OECs) 217, 225,
228, 369, 419, 420, 422, 424, 425,
428–430
oligodendrocytes 35, 125, 128, 132, 176, 224, 226,
227, 342, 371, 420, 425
oligodendroglial precursor cells (OPCs) 225, 227
osmotic potential 239
oxidative phosphorylation 65, 111, 115, 117, 118,
280, 281
oxidative stress 97, 122, 308, 356, 357
oxygen and hemoglobin 113, 114
pain 99, 101, 104, 134, 135, 195–201, 229
Parkinson’s disease 81, 130, 371, 403 (see also
neurodegenerative disease)
PARP-1 86
441
patch clamping 158, 333
pathogenesis 45, 51, 53, 56, 67, 83, 87, 89, 130, 187,
190, 303, 304, 357
pathomechanisms 125, 126, 128, 131, 136
pathophysiology 13, 23, 61, 65, 70, 73, 97, 99, 111,
118, 125–127, 131, 185, 195, 235, 275, 283,
293, 296, 299, 300, 327, 335, 359, 393, 395,
397, 419
patient protection 243, 248
peptide chromatography 407, 409
peroxynitrite (ONOO
) 175
pH 66, 72, 117, 120, 356, 368
brain pH 65, 121, 209
pH paradox 121
pharmacological therapies 157, 218, 229, 248
phospholipases 68, 279
phospholipase A2 280
phospholipase C 32, 63, 280, 305
polypharmacia 91
poly-hydroxypropylmethacrylamide 367, 379
population spike amplitudes (PSA) 343, 344
positron emission tomography (PET) 120, 207,
208, 210–212, 278
post-tetanic potentiation (PTP) 151
potassium channel blockers 217, 221
potassium channels 220, 221, 227, 422, 424
antagonism of 220
potassium 33, 98, 117, 217, 218, 220, 279
preconditioning 256, 257, 344, 353, 355
ischemic preconditioning 256, 343
presenilin 304, 305, 306
pressure changes 28, 398
pressure volume index 119
presynaptic hyperexcitability 329–331
primary damage 13, 19, 23, 395, 396
primary phase 14, 19
progenitor cells 136, 225, 226, 320, 367–369, 372,
377, 420, 425
programmed cell death 82, 126 (see also apoptosis)
progressive injury 125, 127, 128, 130, 132, 135,
136
propidium iodide 70, 72, 158, 256
prostaglandins 113, 136, 279
proteases 43, 45, 46, 52, 53, 55, 82, 84, 86, 100, 122,
280, 305, 306, 403
protein kinase C (PKC) 46, 160, 162, 305
proteomics 401, 406, 411
(see also neuroproteomics)
442

proxy consent 243, 244, 247 sexual dimorphism 340, 341, 343
Prussian blue staining 367, 372, 373, 376, 379, 380 shaken baby syndrome 254
purinergic receptors 69, 72 signal transduction 171, 173, 174, 318, 398
P2X receptors 69 small interfering RNA (siRNA) 189
P2Y receptors 69 sodium 32–34, 55, 61, 65, 68–70, 98, 135, 195–200,
Purkinje cell 327–333 218, 221, 227, 422
sodium channel 32, 135, 196, 221, 430
randomized controlled trial (RCT) 219, 243,244 Nav1.3 sodium channel 195, 197–199, 201
enrollment criteria 244 sodium/calcium exchangers 55, 67
reactive oxygen species (ROS) 34, 84, 86, 87, 90, spectrin 46, 47, 52, 87, 88, 273, 274, 280
218, 357, 358 sports 3, 24, 253, 254, 263–265, 269, 275,
rehabilitation 97, 202, 217, 219, 220, 310, 328, 340, 401
346 American football 264
remyelination 131, 132, 134, 226, 342, 419, 420, boxing 268
422, 430 ice hockey 263
repeated injuries in vitro 256 return-to-play guidelines 263, 285
repeated TBI 253, 254, 257, 258 (see also repetitive soccer 264
traumatic brain injury) spreading depression 30, 279
repetitive traumatic brain injury, 253 static injury 403
Rho 223, 224, 386 stem cells 136, 217, 225–227, 320, 367, 369, 371,
Rho antagonist 217, 224 372, 379, 381, 387
risk–benefit ratio 244, 246, 248 bone marrow mesenchymal stem cells 367
road traffic accident mortality 5 embryonic stem cells 367, 369, 371
rotational head injury 307 strain-injury 68 (see also stretch)
strains 14, 16, 17, 19, 21, 23, 28, 55
S-100B protein 256, 317, 320 stretch 30–32, 35, 68, 143, 144, 151, 158–164, 253,
S100B 317–322 256, 257, 271, 275, 277, 279, 318
neurotrophic properties of S100B 318 stretch injury 35, 143, 158–162, 164, 253, 256, 271,
neuroprotective effect of S100B 319 277, 279, 318
S-100a protein 282 striatum 191, 343
Schwann cells 132, 225, 226, 228, 282, 284 stroke 31, 89, 97, 103, 136, 185, 188, 217, 248,
second impact syndrome 285, 286 339, 346, 348, 354, 358, 367, 369, 371,
second messengers 72 394, 412
secondary insults 118, 119, 126, 207, 254, 281, 296, subcellular 43–46, 51, 52, 55, 56, 153
321, 403, 414, 415 subdural hematoma (SDH) 27, 66, 67, 118, 120,
secondary phase 14, 21, 23 210, 239, 266, 280, 293–296, 298, 300 (see
secretases 304, 305 also acute subdural hematoma)
a-secretase 304, 305 substance P 97–104, 196
b-secretase 304, 305 superoxide 46, 175, 177
g-secretase 304, 305 surgery 178, 218, 219, 224, 238–241, 344,
seizure threshold 143, 156 354, 394
seizures 64, 145, 299, 300, 341, 347 surgical decompression 217–219, 393, 394,
sex differences 339 396, 398
sex steroids 86, 345 surgical treatment 217, 219, 235, 239
estradiol 342, 346 synaptic circuits 143, 144
estrogen 86, 309, 340–344, 346, 348 synaptic plasticity 32, 69, 148, 149, 151–153, 257,
progesterone 309, 342, 344–346, 348 333, 335 (see also long-term potentiation,
testosterone 86, 342, 344, 348 long-term depression)

tachykinin receptors 100, 103 (see also neurokinin
receptors)
tensile strain 68, 156, 158, 266–268, 273, 274, 276
(see also stretch)
thalamus 44, 48, 130, 144, 195, 199–201
ventroposterior lateral (VPL) nucleus 195,
199
VPL neurons 195, 199–201 (see also
ventroposterior lateral nucleus)
therapeutic window 33, 88–90, 218, 227, 246, 381,
396
therapies 27, 34, 52, 97, 127, 137, 178, 217, 218,
220, 229, 300, 305, 369, 402, 405, 415
time frames 243–245
time window 243, 244, 246–248, 335, 368, 420
(see also critical time frames)
tissue deformation 15, 16, 28, 31
tissue oxygen response 211
tissue strain 18, 28
tissue tolerances 14, 19, 21
tolerance criteria 13, 14, 18, 23, 24
transcranial Doppler flowmetry 111
443
transplantation 136, 226–228, 367–369, 377, 379,
419–422, 424, 426, 427–430
cellular transplantation 136, 419
OEC transplantation 228, 420, 426, 428–430
traumatic axonal injury (TAI) 48, 85, 271, 327
treatment strategies 14, 17, 90, 111, 125, 300, 353
TrkA receptor 343
tumor necrosis factor (TNF) 84, 89, 177, 223, 359
tumor necrosis factor alpha (TNF-a) 127, 223,
225, 229, 359, 360, 410
TUNEL 82, 87, 89, 306, 428
two-dimensional polyacrylamide gel
electrophoresis (2D-PAGE) 407, 408, 409
voltage gated calcium channels 32, 55
voltage-gated sodium channels 32, 195, 196, 422
Nav1.3 195, 197–199, 201
weight drop model 155, 156, 254, 255, 257, 405,
406
whole-cell patch-clamp recordings 198
whole-cell patch voltage-clamp recordings 148
This page intentionally left blank