Sleep Circuits and Functions PDF

1519_Title 8/27/04 11:05 AM Page 1
SLEEP
CIRCUITS AND FUNCTIONS
EDITED BY
Pierre-Hervé Luppi, Ph.D.
Université Claude Bernard Lyon I
Lyon, France
CRC PR E S S
Boca Raton London New York Washington, D.C.
Copyright © 2005 CRC Press LLC
1519_C00.fm Page vi Tuesday, August 24, 2004 1:40 PM
Library of Congress Cataloging-in-Publication Data
Sleep : circuits & functions / edited by Pierre-Hervé Luppi.

p. cm. — (Methods & new frontiers in neuroscience series)
Includes bibliographical references and index.
ISBN 0-8493-1519-0 (alk. paper)
1. Sleep—Physiological aspects. 2. Neurophysiology. I. Luppi, Pierre-Hervé. II. Series.
QP425.S6735 2004
612.8¢21—dc22 2004054498
This book contains information obtained from authentic and highly regarded sources. Reprinted material
is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable
efforts have been made to publish reliable data and information, but the author and the publisher cannot
assume responsibility for the validity of all materials or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, microfilming, and recording, or by any information storage or
retrieval system, without prior permission in writing from the publisher.
All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or
internal use of specific clients, may be granted by CRC Press LLC, provided that $1.50 per page
photocopied is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923
USA. The fee code for users of the Transactional Reporting Service is ISBN 0-8493-1519-
0/05/$0.00+$1.50. The fee is subject to change without notice. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.
The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for
creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC
for such copying.
Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation, without intent to infringe.
Visit the CRC Press Web site at www.crcpress.com
© 2005 by CRC Press LLC
No claim to original U.S. Government works

International Standard Book Number 0-8493-1519-0
Library of Congress Card Number 2004054498
Printed in the United States of America 1 2 3 4 5 6 7 8 9 0
Printed on acid-free paper

1519_C00.fm Page vii Tuesday, August 24, 2004 1:40 PM
Methods & New Frontiers

in Neuroscience
Our goal in creating the Methods & New Frontiers in Neuroscience series is to
present the insights of experts on emerging experimental techniques and theoretical
concepts that are or will be at the vanguard of the study of neuroscience. Books in
the series cover topics ranging from methods to investigate apoptosis to modern
techniques for neural ensemble recordings in behaving animals. The series also
covers new and exciting multidisciplinary areas of brain research, such as compu-
tational neuroscience and neuroengineering, and describes breakthroughs in classical
fields such as behavioral neuroscience. We want these to be the books every neuro-
scientist will use in order to get acquainted with new methodologies in brain research.
These books can be given to graduate students and postdoctoral fellows when they
are looking for guidance to start a new line of research.
Each book is edited by an expert and consists of chapters written by the leaders
in a particular field. Books are richly illustrated and contain comprehensive bibli-
ographies. Chapters provide substantial background material relevant to the partic-
ular subject; hence, they are not only “methods” books. They contain detailed tricks
of the trade and information as to where these methods can be safely applied. In
addition, they include information about where to buy equipment and about Web
sites that are helpful in solving both practical and theoretical problems.
We hope that as the volumes become available, the effort put in by us, by the
publisher, by the book editors, and by the individual authors will contribute to the
further development of brain research. The extent to which we achieve this goal will
be determined by the utility of these books.
Sidney A. Simon, Ph.D.

Miguel A.L. Nicolelis, M.D., Ph.D.
Series Editors

1519_C00.fm Page ix Friday, August 27, 2004 11:43 AM
Foreword
This book presents up-to-date research on sleep mechanisms and functions. Each
chapter describes the latest findings and provides a synthesis and bibliography. The
chapter by Luppi et al. concerns the anatomical network of sleep, and it summarizes
the newest model of the neural system responsible for paradoxical sleep (PS) or
rapid eye movement (REM) sleep. In the rat the neurons responsible for PS onset
and maintenance seem to be clustered in a sphere of tissue smaller than 1 mm3,
centered on the sublaterodorsal nucleus of the pontine reticular formation. It is well
known that a discrete bilateral lesion of this nucleus is followed by the disappearance
during many months of the tonic aspect of PS (fast cortical activity and decrease of
muscle tone), whereas pontine-geniculate-occipital (PGO) activity may still occur.
Why such an important state, which is responsible for dreaming, is dependent upon
such a small system without recuperation is still unknown, because for waking the
systems responsible are quite diffuse and redundant (locus coeruleus, posterior
hypothalamus, hypocretin neurons, etc.) so that lesion of one or even several of them
is followed by recovery of EEG arousal. The same is true for slow-wave sleep
because either lesion of the ventro-lateral preoptic region (VLPO) or prebulbar
transection (midpontine preparation) is followed by some recovery of slow cortical
activity after 1 or 2 weeks.
Fort et al. have identified in vitro the presumed sleep-promoting neurons local-
ized in the VLPO. They have been able to establish firmly that two subtypes of
sleep-promoting neurons may be segregated according to their modulation by 5-HT.
Their results resuscitate the almost moribund 5-HT theory of sleep in suggesting
“that 5-HT released during waking in the preoptic area may participate concomitantly
to seemingly opposite mechanisms by strengthening arousal through the inhibition
of Type I neurons and prepare sleep via the subthreshold excitation of Type II
neurons.” Fort et al. describe in detail a model in which PGD2 and adenosine would
contribute to the homeostatic regulation of sleep.
The molecular mechanisms of sleep-wake regulation by PGD2 and PGE2 have
been described by Osamu Hayaishi, who dedicated 20 years of his life to the explo-
ration of the role of prostaglandins in the sleep-waking cycle. In his chapter, he
summarizes the experiments which have permitted to localize PGD synthase in the
arachnoid membranes and the choroids plexuses, and the PGD2 receptors at the ventral
surface of the rostral basal forebrain. Then, the binding of PGD2 to its receptor is
followed by a transduction by adenosine through the adenosine A2 receptor. This is
a good example of the cascade of events that start in the leptomeninges (a system
unknown by the majority of electrophysiologists) and end in the ventral hypothalamus.
In addition to PGD2 and adenosine, new players have entered the arena of sleep-
or waking-inducing substances. In his paper De Lecea emphasizes that “small is
beautiful and interesting.” Given the extraordinary cellular complexity of the brain,

1519_C00.fm Page x Friday, August 27, 2004 11:43 AM
it can be estimated that a few hundred mRNAs are expressed in a small population
of cells (less than 106 neurons). The expression of these rare mRNAs would confer
particular physiological properties to the neurons that produce them. By studying
small populations, De Lecea has been able to isolate by differential gene expression
analysis two peptidergic systems that are well-known newcomers in the jet set of
waking- or sleep-responsible substances: Corticostatin at the cortical level is impli-
cated in slow-wave activity and the famous hypocretins (or orexins) in the lateral
hypothalamus. This latter system is responsible for waking and for inhibiting PS,
because of the discoveries that hypocretins knockout (KO) mice present narcoleptic
attacks, and narcoleptic dogs display a mutation in the hypocretin receptor 2 gene.
De Lecea also describes new methods for deciphering the mechanisms of action
of these new peptides:
• The use of cholera toxin A, which is equivalent to placing a stimulating

electrode in situ
• The utilization of pseudorabies viruses expressing green fluorescent pro-
tein (GFP), which may be used for the mapping of afferents
• The use of expression cameleon to monitor changes in intracellular calcium
• The development of MRI for rodents in vivo that will give moving images
in three spatial planes
All of these techniques will provide the “how” but not necessarily the “why” of the
function of sleep.
The approach to the function of sleep states is the aim of numerous recent genetic
studies. This field, which was pioneered by Valatx in 1972, is now fully developed,
as demonstrated by the review written by Tafti et al: “A revolution in the under-
standing of the molecular basis of circadian rhythm has led to the identification of
a number of clock genes and of their interaction to generate a circadian rhythm.”
Moreover the recent progresses in molecular genetics have permitted the identifica-
tion of genetic factors responsible for the pathology of sleep disorders (narcolepsy
and advanced sleep-phase syndrome). In rodents it has also been shown that theta
oscillations during sleep may be modulated by the metabolic fatty-acid beta-oxida-
tion pathway.
Tononi and Cirelli have followed another genetic approach. They screened
20,000 transcripts expressed in the cerebral cortex during sleep, wake, or sleep
deprivation in the rat, and they found that about 100 genes related to protein
synthesis and neural plasticity increase their expression during sleep. During sleep
deprivation Tononi and Cirelli found an increase in the expression of the mRNA
for the arylsulfotransferase enzyme (regulating a major step in the catabolism of
catecholamines), which led them to the hypothesis that “sleep function might be to
interrupt the continuous catecholaminergic activity that occurs during waking.” They
have also thoroughly investigated the rest-activity cycle in the fruit fly. They bring
conclusive evidence that the rest of a fly is sleeplike, because it is modulated by
both the circadian clock and the need for sleep as evidenced by the homeostatic
increase of rest after rest deprivation. These fly hypnologists also have been able
to obtain by mutation a very short sleep line (3 H out of 24 H), which can be

1519_bookTOC.fm Page xix Tuesday, August 24, 2004 1:41 PM
Contents
Chapter 1 Sleep and Neuronal Plasticity: Cellular Mechanisms of
Corticothalamic Oscillations
Mircea Steriade
Chapter 2 Role of Basalo-Cortical System in Modulating Cortical Activity

and Sleep-Wake States
Maan Gee Lee and Barbara E. Jones
Chapter 3 In Vitro Identification of the Presumed Sleep-Promoting Neurons

of the Ventrolateral Preoptic Nucleus (VLPO)
Patrice Fort, Pierre-Hervé Luppi, and Thierry Gallopin
Chapter 4 Molecular Mechanisms of Sleep-Wake Regulation: A Role of

Prostaglandin D2 and Adenosine
Osamu Hayaishi
Chapter 5 The Network Responsible for Paradoxical Sleep Onset and

Maintenance: A New Theory Based on the Head-Restrained
Rat Model
Pierre-Hervé Luppi, Romuald Boissard, Damien Gervasoni, Laure Verret,
Romain Goutagny, Christelle Peyron, Denise Salvert, Lucienne Léger,
Bruno Barbagli, and Patrice Fort
Chapter 6 Reverse Genetics and the Study of Sleep-Wake Cycle:

The Hypocretins and Cortistatin
Luis de Lecea
Chapter 7 Genetic Regulation of Sleep

Yves Dauvilliers, Paul Franken, and Mehdi Tafti
Chapter 8 Searching for Sleep Mutants of Drosophila melanogaster

Chiara Cirelli and Giulio Tononi

1519_bookTOC.fm Page xx Tuesday, August 24, 2004 1:41 PM
Chapter 9 Sleep Phylogeny: Clues to the Evolution and Function

of Sleep
Jerome M. Siegel
Chapter 10 Sleep, Synaptic Plasticity, and the Developing Brain

Marcos Gabriel Frank
Chapter 11 Changes in Brain Gene Expression between Sleep and

Wakefulness
Giulio Tononi and Chiara Cirelli
Chapter 12 Neuronal Reverberation and the Consolidation of New

Memories across the Wake-Sleep Cycle
Sidarta Ribeiro, Damien Gervasoni, and Miguel A.L. Nicolelis
Chapter 13 Cerebral Functional Segregation and Integration during

Human Sleep
Pierre Maquet, Fabien Perrin, Steven Laureys, Tahn Dang-Vu, Martin Desseilles,
Mélanie Boly, and Philippe Peigneux

1519_C01.fm Page 1 Tuesday, August 24, 2004 1:42 PM
1 Sleep and Neuronal

Plasticity: Cellular
Mechanisms of
Corticothalamic
Oscillations
Mircea Steriade
CONTENTS
Introduction
Sleep Rhythms: Their Grouping by Cortical Slow Oscillation
into Unified Activities
Spindles, a Thalamic Rhythm under Cortical Influence
Clock-Like Thalamic Delta, an Intrinsic Cell Oscillation
Synchronized by Cortical Activity
The Slow Cortical Oscillation and Its Actions in Grouping
Other Sleep Rhythms
Sleep Rhythms Leading to Neuronal Plasticity in Cortical Networks
Intrathalamic and Thalamocortical Neuronal Circuits
Underlying Augmenting Responses
Neuronal Plasticity Outlasting Augmenting Responses
and Sleep Spindles
Functional Significance of Sleep Oscillations
Acknowledgments
References
INTRODUCTION
Three pioneers of sleep research, Frédéric Bremer, Giuseppe Moruzzi, and Michel
Jouvet, developed their concepts based on data from experiments conducted on
various brainstem neuronal systems.1–3 All three eventually reached the conclusion
that sleep is an active process, but Bremer and Moruzzi initially considered the stage
of sleep with highly synchronized brain electrical activity as a deafferented, passive
0-8493-1519-0/05/$0.00+$1.50

state, due to a fall in the cerebral tonus produced by disconnection from sensory
systems1 or decreased activity in the brainstem reticular core.2 These two views are
not irreconcilable, because corticipetal activities in discrete sensory systems con-
tribute to widespread forebrain activation by neocortical projections to the brainstem
reticular formation.4
The passive theory of sleep genesis by brain deafferentation is still alive and
well because some data that pointed to actively hypnogenic neurons did not yet
elucidate the multiple targets and chemical codes of the presumably sleep-promoting
elements. Neuronal systems that are hypothesized to induce sleep would exert their
inhibitory actions on neurons located within ascending activating systems, thus
disconnecting the forebrain, as postulated in passive sleep theories.
Slow-wave sleep (SWS) is far from being a resting or inactive state associated
with general inhibition of cortex and subcortical systems,5 which would give rise to
an “abject annihilation of consciousness.”6 Recent studies using intracellular record-
ings in naturally sleeping animals7 demonstrate intense activity of neocortical neu-
rons during SWS (Figure 1.1) and suggest that brain oscillations during SWS are
actively implicated in the consolidation during this sleep state of memory traces
acquired during the wakefulness. This chapter discusses the experimental basis of
this hypothesis.
SLEEP RHYTHMS: THEIR GROUPING BY CORTICAL

SLOW OSCILLATION INTO UNIFIED ACTIVITIES
In the intact brain there are no pure rhythms, as those generated by neuronal networks
in simplified preparations, such as thalamic or cortical slices maintained in vitro.
Instead a coalescence of different oscillatory types is observed during SWS, due to
the impact of cortically generated slow oscillation (~0.5–1 Hz) upon neuronal syn-
aptic interactions in the thalamus that give rise to spindles (7–15 Hz) and upon the
interplay between intrinsic currents of thalamocortical neurons that produce clock-
like delta waves (1–4 Hz). Moreover, fast rhythms in the beta-to-gamma frequency
range (20–60 Hz), conventionally regarded as only characteristic for the behavioral
states of waking and REM sleep, occur during the active (depolarizing) phase of the
slow cortical oscillation in SWS. This combination of low-frequency (<15 Hz) and
fast (>20 Hz) rhythms defies a strict dissociation between different sleep and waking
rhythms, and justifies our concept8 that sleep oscillations are generated in intercon-
nected neuronal loops between the cerebral cortex and thalamus under the control
of generalized modulatory systems arising in the brainstem core, hypothalamus, and
basal forebrain. This condition can only be investigated in vivo9 and, at best, using
intracellular recordings in naturally awake and sleeping preparations.
This section discusses:
• Neuronal circuitry that underlies different sleep oscillations

• Comparison of the results of experiments conducted in vivo and in vitro9
• Morphological and physiological bases of SWS rhythms’ coalescence by
the slow cortical oscillation

FIGURE 1.1 Natural slow-wave sleep (SWS) characterized by prolonged hyperpolarizations

in neocortical neurons but rich spontaneous firing during depolarizing phases of slow oscil-
lation. Chronically implanted cat. Five traces in top panel depict EEG from depth of left
cortical areas 4 (motor) and 21 (visual association), intracellular recording from area 21
regular-spiking neuron (resting membrane potential is indicated), electro-oculogram (EOG),
and electromyogram (EMG). Part marked by horizontal bar is expanded below left (arrow).
Note relation between hyperpolarizations and depth-positive EEG field potentials. Below right
histograms of membrane potential (10-sec epochs) during period of transition from waking
to SWS depicted above. Note membrane potential around –64 mV during the 20 sec of waking
and progressively increased tail of hyperpolarizations, up to –90 mV, during SWS. Data from
experiments by M. Steriade, I. Timofeev and F. Grenier (details in Steriade et al.7).
• Human studies that corroborate experimental work and together empha-

size the role of SWS oscillations in neuronal plasticity and learning
SPINDLES, A THALAMIC RHYTHM UNDER CORTICAL INFLUENCE

Sleep spindles are generated in thalamic networks and are initiated in thalamic
reticular neurons (Figure 1.2). Briefly, GABAergic reticular neurons impose spike-
bursts in the frequency range of spindles onto thalamocortical neurons, which display

FIGURE 1.2 Spindle oscillations in reticular thalamic (RE), thalamocortical (Th-Cx, vent-
rolateral nucleus), and cortical (Cx, motor area) neurons. A, circuit of three neuronal types.
B, two rhythms (7–14 Hz and 0.1–0.2 Hz) of spindle oscillations in cortical EEG. C, one
EEG spindle sequence is depicted below with intracellular recordings in cats under barbiturate
anesthesia. See explanations in text. Modified from Steriade and Deschênes (1988).
rhythmic inhibitory postsynaptic potentials (IPSPs) that de-inactivate the Ca2+-

dependent current (IT) and produce low-threshold spikes (LTSs) crowned by high-
frequency bursts consisting of fast, Na+-mediated action potentials. These spike-
bursts are transferred to cortical neurons, where they elicit excitatory postsynaptic
potentials (EPSPs), occasionally leading to action potentials.

The pacemaking role of thalamic reticular GABAergic neurons was demon-

strated by absence of spindles in thalamocortical systems after lesions of thalamic
reticular perikarya or transections separating this inhibitory nucleus from the remain-
ing thalamus10 and by preservation of spindles in the deafferented thalamic reticular
nucleus.11 The failure to record spindles in isolated thalamic slices from the posterior
part of the reticular nucleus was explained by the absence of an intact collection of
reticular cells in that experimental condition.12 The slicing procedure may cut the
very long dendrites of these neurons, which generate spindles through an avalanche
of dendro-dendritic synaptic interactions.11 The requirement of high-density IT in
intact dendrites of thalamic reticular neurons for the production of spindles, similar
to those seen during natural SWS, was demonstrated in combined in vivo, in vitro,
and in computo work.13 Because the thalamic origin of spindles was discussed at
length in previous reviews and a monograph,14 here emphasis is placed on the cortical
control of thalamically generated spindles, which explains discrepant results between
extremely simplified and intact-brain preparations.
One of the dissimilarities between in vitro and in vivo results is the systematic
propagation of spindles in thalamic slices15 versus the quasi-simultaneity of spindle
sequences over widespread thalamic and cortical territories during natural SWS in
animals and humans.16,17 It is known that the most efficient experimental method to
elicit spindles are corticofugal volleys, applied either ipsilaterally, which directly
activate thalamic reticular neurons,18 or contralaterally (through callosal and corti-
cothalamic pathways), to avoid antidromic invasion of thalamocortical cells’ axons
and axon-reflex activation of pacemaking reticular neurons.19
It was natural to hypothesize that spindle propagation in thalamic slices was due
to the absence of cortex in the isolated thalamus. Indeed, decortication prevented
the simultaneity of spindle sequences and disorganized their spatio-temporal coher-
ence.16,17 The role of cortex in spindles’ simultaneity is also shown by diminished
coherence of spindles during cortical spreading depression, during which cortico-
thalamic neurons display no or negligible spontaneous activity.20 The powerful role
of corticothalamic projections in the high synchronization of spindle oscillations was
demonstrated in humans by showing that cortical-damaged patients display significantly
reduced coherence spectra from derivations ipsilateral to the lesion.21
Corticothalamic activity is not only implicated in the long-range synchronization
of spindles but also in the termination of individual spindle sequences. The termi-
nation of spindle sequences is due, at least partially, to asynchrony in the thalamic
circuit, stemming from the different durations of spindle-related IPSPs in thalamo-
cortical cells, resulting in different times at which postinhibitory rebound spike-
bursts are fired, so that the synchrony in the circuit between thalamocortical and
thalamic reticular neurons is disrupted and spindles are terminated.
During the late phase of spindles, neocortical neurons become tonically depolar-
ized leading to firing, and spike-triggered-averages by cortical neurons do not reveal
a phase relationship between cortical and thalamocortical neurons.22 The sustained
depolarization of cortical neurons during the late part of a spindle sequence may be
effective in desynchronizing thalamic networks and terminate spindles. Another
intrinsic cellular factor may be the depolarizing action of a hyperpolarization-activated
cation current (IH).23 This hypothesis was tested in a computational model, but the

isolated network between thalamocortical and thalamic reticular neurons oscillated

infinitely, and up-regulation of IH alone was not sufficiently strong to terminate
spindling; however, with the addition of the corticothalamic feedback, spindles in the
thalamic network were shorter.22
To sum up,
• The first part of a spindle sequence is generated in the pacemaker thalamic

reticular nucleus.
• During the first two-to-four IPSPs composing the spindles, thalamocorti-
cal neurons do not display rebound spike-bursts, do not return signals to
thalamic reticular neurons, and do not contribute to this phase of spindles.
• The middle part of a spindle sequence is due to the activity in the recip-
rocal loop between thalamic reticular and thalamocortical neurons.
• The termination of spindles is due to the depolarizing action of cortico-
thalamic neurons, possibly assisted by the depolarizing action of IH.
Importantly, although generated within the thalamus, spindles can be triggered

by the synchronous firing of corticothalamic neurons, as naturally occurring during
the depolarizing phase of the slow sleep oscillation, which is associated with depth-
negative field potentials in cortex (Figure 1.3 A). This combination gives rise to
slowly oscillatory cycles that include both the cortically generated slow oscillation
and the thalamically generated spindles.
FIGURE 1.3 (See facing page.) Coalescence of cortical slow oscillation with other slow-
wave sleep (SWS) rhythms generated in the thalamus. In left column two traces represent
field potential from depth of cortical association area 5 and intracellular recording from
thalamic reticular neuron (top and bottom traces, respectively); below traces represent field
potential from depth of cortical area 5 and intracellular recording from thalamocortical neuron
in ventrolateral nucleus. In right column circuits involved in generation of respective SWS
pattern. Synaptic projections are indicated with small letters, corresponding to arrows at left,
which indicate time sequence of events. (A) Combination of slow oscillation with a spindle
sequence. Depolarizing phase of field slow oscillation (depth-negative, downward deflection,
also called K-complex) in cortex (Cx) travels through corticothalamic pathway (a) and triggers
in thalamic reticular nucleus (RE) a spindle sequence that is transferred to thalamocortical
cells (ThCx) of dorsal thalamus (b) and back to cortex (c), where it shapes tail of slow
oscillatory cycle. (B) Modulation of slow oscillation by a sequence of clock-like delta waves
originating in thalamus by interplay between two inward currents (IH and IT) of thalamocortical
neurons. Synchronous activity of cortical neurons during slow oscillation (depth-negative
peak of cortical field potential) travels along corticothalamic pathway (a’) eliciting an EPSP,
curtailed by an IPSP produced along cortico-RE (a) and RE-ThCx (b) projections. Hyperpo-
larization of thalamocortical cell generates a sequence of low-threshold potentials crowned
by high-frequency spike-bursts at delta frequency that may reach cortex through thalamocor-
tical link (c). Diagrams modified from Amzica and Steriade,30 with intracellular staining of
three neuronal types, and intracellular recordings by Steriade et al.29 and Contreras and
Steriade.18

CLOCK-LIKE THALAMIC DELTA, AN INTRINSIC CELL OSCILLATION

SYNCHRONIZED BY CORTICAL ACTIVITY
The thalamic component of delta waves has a clock-like pattern and depends on two
inward currents of thalamocortical neurons: the hyperpolarization-activated current,
IH, carried by Na+ and K+, which is expressed as a depolarizing sag of membrane
potential toward rest, and a transient Ca2+ current, IT, underlying the LTS. The
mechanisms of generation and synchronization of this intrinsic-cell thalamic oscillation
were revealed using intracellular studies in vitro24–26 and in vivo.27,28 The prerequisite
FIGURE 1.3

for the appearance of the clock-like rhythm is the hyperpolarization of thalamocortical

neurons to levels more negative than –65 or –70 mV, and their depolarization leads to
abolition of the rhythm. In contrast to the spindle oscillation that is generated by
synaptic interactions that necessarily include the thalamic reticular nucleus, the delta
oscillation is an intrinsic oscillation of thalamocortical neurons.
However intrinsically generated, the thalamic delta oscillation is subject to the
influence of corticothalamic synaptic volleys, which excite thalamic reticular neurons
that set the membrane potential of thalamocortical neurons at adequate levels of
hyperpolarization, at which clock-like delta is generated, and synchronize pools of
thalamocortical cells.27 In turn the thalamic component of sleep delta waves is
projected to cortex and sculpts the slow oscillation29,30 (Figure 1.3 B). Then a
thalamic oscillation generated by intrinsic neuronal properties becomes expressed
at the cortical level as a result of synchronization among different thalamocortical
neurons due to synaptic activities evoked by corticofugal volleys.
THE SLOW CORTICAL OSCILLATION AND ITS ACTIONS IN GROUPING OTHER

SLEEP RHYTHMS
The slow oscillation was first described using intracellular recordings of cortical
neurons in anesthetized animals as well as EEG recordings in human sleep.31 Its
cortical origin was demonstrated by survival after thalamectomy,32 presence in large
isolated cortical slabs in vivo,33 in cortical slices maintained in vitro,34 and absence
in the thalamus of decorticated animals.35 The slow oscillation was also described
using extracellular36 and intracellular7 recordings of cortical neurons during natural
SWS in animals as well as in EEG37–38 and magnetoencephalographic (MEG)39
recordings during night SWS in humans.
The slow oscillation consists of prolonged depolarizations, associated with brisk
firing (~10–40 Hz), and long-lasting hyperpolarizations during which neurons are
silent (Figure 1.1). Generally, the depolarization lasts for ~0.3–0.6 sec and consists
of non-NMDA- and NMDA-mediated EPSPs, fast prepotentials (FPPs), a voltage-
dependent persistent Na+ current (INa(p)), and fast IPSPs induced in pyramidal neurons
by synaptically coupled GABAergic local-circuit cortical cells.31 The presence of
fast waves within the beta and gamma frequency bands (generally 20–60 Hz) over
the depolarizing phase of the slow sleep oscillation may be surprising for those who
think that these fast rhythms are necessarily associated with consciousness and states
of cognition. In reality, fast rhythms are voltage-dependent and occur as a function
of membrane depolarization in cortical neurons.
The transition from beta to gamma oscillations may take place over short (0.5–1
sec) time periods36 without being related to change in behavioral or cognitive state.
There is no need to distinguish between these two types of fast rhythms unless
intracellular recordings in behaving and performing animals would demonstrate their
distinction and the depolarization phase associated with fast rhythms characterizes
the slow oscillation during natural SWS or deep anesthesia.36 As to the hyperpolar-
izing phase of the slow oscillation, it is not due to the action of inhibitory interneurons
but to disfacilitation (removal of synaptic, mainly excitatory, inputs) in intracortical
and thalamocortical networks, and to some K+ currents.

Several pieces of evidence support this conclusion:
• Neurons identified morphologically as basket (aspiny or sparsely spiny)

cells or electrophysiologically characterized as fast-spiking (presumably
GABAergic) cells, during either natural sleep7 or ketamine-xylazine anes-
thesia,40 behave in phase with regular-spiking (pyramidal) neurons; i.e.,
they fire during the depolarizing phase and are silent during the hyperpo-
larizing phase.
• Intracellular recordings with Cl-filled pipettes during naturally sleeping
animals do not affect the prolonged hyperpolarizations of the slow oscil-
lation in SWS.7,41
• Recordings with Cs+-filled pipettes strikingly reduce or abolish the hyper-
polarizations.41 As Cs+ blocks nonspecifically K+ currents, hyperpolariza-
tions during the slow oscillation are produced, at least partially, by a series
of K+ currents, most probably IK(Ca).
• Disfacilitation in cortical networks is the other factor accounting for the
prolonged hyperpolarizations, under anesthesia as well as during natural
SWS, as the apparent input resistance was almost double during the
hyperpolarizing phase of the slow oscillation in SWS, compared to the
depolarizing phase of this oscillation.7,42
The disfacilitation is explained by a progressive depletion of [Ca2+]o during the

depolarizing phase of the slow oscillation,43 which would produce a decrease in
synaptic efficacy that would eventually lead to the functional disconnection of
cortical networks. Realistic models of the SWS slow oscillation in corticothalamic
systems propose that summation of miniature EPSPs during the hyperpolarizing
(silent) phase of the slow oscillation activates INa(p) and depolarizes the membrane
of pyramidal neurons sufficiently for triggering spikes and generating the next
depolarizing phase.33,44 The transition from the SWS slow oscillation to brain-
activated states is produced by the erasure of prolonged hyperpolarizing phases in
cortical neurons45 and their increased input resistance as tested during the behavioral
state of wakefulness.7
The concept of grouped SWS rhythms, mainly slow and spindle oscillations but
also slow and fast oscillations, derived from animal and human studies7,8,32,36,38,40
(Figure 1.4), is corroborated by recent studies using d.c. EEG signals during stages
2 and 3 of human sleep and showing the grouping of slow oscillation with spindles
and beta rhythms.46
SLEEP RHYTHMS LEADING TO NEURONAL

PLASTICITY IN CORTICAL NETWORKS
Plasticity is defined as a short- or long-term alteration in neuronal responsiveness
that depends on the history of a given neuronal network, a change that may evolve
from the transient strengthening or depression of synapses to permanent formation
of new connections. Besides synaptic activities in neuronal networks, which depend

on the behavioral state of vigilance, the mechanisms underlying plasticity include

modifications in the release of neurotransmitters and postsynaptic sensitivity47 and
changes in intrinsic currents that modify neuronal responsiveness.48,49 The impact
of network synaptic activity on voltage- and transmitter-gated conductances of single
thalamic and neocortical neurons, and the transformation of firing patterns produced
by intrinsic cellular properties during shifts in natural states of vigilance, are dis-
cussed elsewhere.8,9
Moruzzi50 proposed the first hypothesis relating sleep to plasticity by postulating
that SWS does not concern the fast recovery processes in routine synapses underlying
stereotyped activities but the slow recovery of learned synapses. During the past
decade the development from brain oscillations occurring spontaneously during SWS
FIGURE 1.4

or their experimental models to neuronal plasticity was investigated in thalamocor-

tical systems. Here we focus on short- and medium-term increases in cellular respon-
siveness that appear as a consequence of augmenting responses that mimic naturally
occurring sleep spindles. This leads to a discussion of data from humans and animal
experiments showing the role of sleep oscillations in memory and learning.
INTRATHALAMIC AND THALAMOCORTICAL NEURONAL CIRCUITS

UNDERLYING AUGMENTING RESPONSES
Augmenting responses are the experimental classical model of sleep spindles51 and
are defined as thalamically evoked cortical potentials that grow in size during the first
stimuli at a frequency of 5 to 15 Hz, usually ~10 Hz, like the waxing of waves at the
onset of spontaneously occurring spindle sequences. Although augmentation occurs
in the thalamus of decorticated animals52 (like spindles) and in the intact cortex of
athalamic preparations32 or even in isolated cortical slabs in vivo53 and in cortical
slices maintained in vitro,54 the full development of augmenting responses, leading
to self-sustained activities, requires interacting thalamic and cortical networks.
The old idea that incremental thalamocortical responses are of two basically
different types, augmenting and recruiting, was suggested by invoking a different
cortical layer distribution and a longer latency of recruiting responses that was
ascribed to a “diffuse multineuronal system” with intralaminar nuclei serving as an
intrathalamic association system.55 It is now known that augmenting responses may
precede recruiting responses, or vice versa, within the same sequence of rhythmic
potentials (because of the multi-laminar distribution of thalamic projections to cor-
tex), that some cortical recruiting (depth-positive) responses may display latencies
as short as those of augmenting (depth-negative) responses, and that the longer
latency of cortical recruiting responses is not due to the intrathalamic spread of
activity but to slower conduction velocities of axons from some thalamic nuclei.8
FIGURE 1.4 (See facing page.) Cortical slow oscillation groups thalamically generated
spindles. CAT (top), intracellular recording in cat under urethane anesthesia from area 7 (1.5
mm depth). Electrophysiological identification (at right) shows orthodromic response to
stimulation of thalamic centrolateral (CL) intralaminar nucleus and antidromic response to
stimulation of lateroposterior (LP) nucleus. Neuron and related EEG wave oscillation is slow.
One cycle of slow oscillation is framed in dots. Part marked by horizontal bar below intra-
cellular trace (at left) is expanded above (right) to show spindles following depolarizing
envelope of slow oscillation. CAT (bottom left), dual simultaneous intracellular recordings
from right and left cortical area 4. Note spindle during depolarizing envelope of slow oscil-
lation and synchronization of EEG when both neurons synchronously display prolonged
hyperpolarizations. HUMAN, the K-complex (KC) in natural sleep. Scalp monopolar record-
ings with respect to contralateral ear are shown (see figurine). Traces show a short episode
from a stage 3 non-REM sleep. The two arrows point to two K-complexes, consisting of a
surface-positive wave, followed (or not) by a sequence of spindle (sigma) waves. Note
synchrony of K-complexes in all recorded sites. At right, frequency decomposition of electrical
activity from C3 lead (see 1) into three frequency bands: slow oscillation (S, 0 to 1 Hz), delta
waves (D, 1 to 4 Hz) and spindles (s, 12 to 15 Hz). Modified from Steriade et al.,32 Contreras
and Steriade40 (CAT), and Amzica and Steriade38 (HUMAN).

The distinction between augmenting and recruiting responses is no longer necessary

and we simply designate these responses as augmenting or incremental.
In the thalamus of decorticated animals, thalamocortical neurons display two
types of augmenting responses to local thalamic stimulation at 10 Hz52: One type
is associated with progressively decreased IPSPs elicited by successive stimuli in
the train and with progressive depolarization of neurons leading to high-threshold
spike-bursts with increasing number of action potentials and spike inactivation
(Figure 1.5 A). The other type of intrathalamic augmenting responses is based on
progressively increased LTSs, which are deinactivated by the increasing hyperpo-
larization produced by repetitive stimuli in the train (Figure 1.5 B). This type of
augmentation (with progressively increased LTSs and rebound spike-bursts) is due
to the parallel excitation, whereas the high-threshold form of augmenting is due to
decremental responses, in a pool of thalamic reticular GABAergic neurons.56 As
augmenting responses mimic spindles, and spindles have been recorded in the
deafferented RE nucleus,11 augmenting responses as well as spindles were also
obtained in computational models of isolated RE nucleus, with synaptic intercon-
nections including GABAA and GABAB components.57,58
Augmenting responses are also generated within the neocortical circuitry, as
demonstrated by stimulating the callosal pathway in thalamectomized cats.32 Rhyth-
mic pulse-trains with the frequency range of sleep spindles (10 Hz) eventually lead
to intrinsically bursting cell’s depolarization and dramatic increase in the number
of action potentials within each evoked spike-burst (Figure 1.5 C). As in other forms
of augmenting responses, such an enhancement in neuronal responsiveness may lead
to self-sustained activities and, in some instances, to various patterns of electro-
graphic seizures32 (see below).
NEURONAL PLASTICITY OUTLASTING AUGMENTING RESPONSES AND

SLEEP SPINDLES
This section discusses the self-sustained activity that follows evoked responses in
the frequency range of spindles within the thalamus of decorticated animals52 and
FIGURE 1.5 (See facing page.) Intrathalamic and corticocortical augmenting responses
leading to neuronal plasticity. (A, B) Unilaterally decorticated cats under ketamine-xylazine
anesthesia. Intracellular recordings from thalamocortical neurons in ventrolateral (VL)
nucleus. Stimulation in VL nucleus (pulse-trains of 5 stimuli at 10 Hz). (A) Pulse-train at 10
Hz evoked high-threshold spike-bursts containing progressively more action potentials, with
spike inactivation. (B) Low-threshold augmenting responses developing from progressive
increase in IPSP-rebound sequences and followed by a self-sustained spindle. Arrow indicates
expanded spike-burst (action potentials truncated). Part marked by horizontal bar and indi-
cating augmenting responses is expanded at right. (C) Cat with extensive thalamic lesion by
kainate lesion, unilateral to cortical recording in area 7. Repetitive callosal stimulation (10
Hz) of homotopic point in contralateral hemisphere. Responses to pulse-trains (each consisting
of 5 stimuli at 10 Hz), repeated every 3 sec, applied to contralateral area 7. Intracortical
augmenting responses to first and eighth pulse-trains are illustrated. Depolarization is about
7 mV, and action potentials within bursts are increased in number after repetitive stimulation.
Modified from Steriade and Timofeev52 (A, B) and Steriade et al.32 (C).

presents evidence of memory processes in the more complex corticothalamic net-

works after prolonged and rhythmic stimuli that mimic spindles.59 The self-sustained
activity is virtually identical to that of responses during the prior period of stimu-
lation. Similar changes have previously been reported in amygdalo-hippocampal
synaptic networks and were followed by self-sustained seizures in those circuits.60
FIGURE 1.5

Finally, medium-term neuronal plasticity, lasting several minutes, will be discussed

from experiments on neocortical neurons.53
During repetitive (10 Hz) thalamic stimuli in decorticated animals, the IPSPs of
thalamocortical neurons are progressively diminished, and the depolarization area
of augmenting responses increases continuously (Figure 1.6 A).
In intact thalamocortical networks, cortical augmenting responses to thalamic
volleys are characterized by the appearance of a secondary depolarization that
mainly depends on spike-bursts generated by an intrinsic property (the de-inactiva-
tion of IT) of thalamocortical neurons,61 however the cortex has the necessary
equipment to develop some forms of augmentation even after thalamectomy (see
Figure 1.5 C).52 The self-sustained oscillations following internally generated
incoming signals during SWS61 suggest that this deafferented behavioral state may
sustain mental events. Indeed repeated spike-bursts evoked by volleys applied to
corticothalamic pathways as well as occurring during spontaneous oscillations may
lead to self-sustained activity patterns resembling those evoked in the late stages of
stimulation (Figure 1.6 B). Such changes are due to resonant activities in closed
loops, as in memory processes.
During the depolarizing envelope of spindle sequences, associated with firing
in neocortical neurons, cortical stimuli elicit an enhancement of the control response,
which may last from tens of seconds to several minutes (Figure 1.7). Repeated pulse-
trains giving rise to augmenting responses produce progressively reduced in ampli-
tude of the IPSP of the control response and its replacement by depolarization.
Moreover single stimuli applied after the rhythmic pulse-trains elicit exclusively
depolarizing responses whose enhancement remained unchanged for several minutes
(see bottom panel in Figure 1.7). Similar phenomena occur in cortical neurons when
testing cortical stimuli are applied during the depolarizing phases of naturally occur-
FIGURE 1.6 (See facing page.) Short-term plasticity from repetitive intrathalamic augment-
ing responses of high-threshold type, and development from corticothalamic augmenting
responses to self-sustained activity. (A) Intracellular recording of thalamocortical neuron in
VL nucleus of cat with ipsilateral hemidecortication and callosal cut. Ketamine-xylazine
anesthesia. Progressive and persistent increase in area of depolarization by repeating pulse-
trains. Pulse-trains consisting of 5 stimuli at 10 Hz were applied to VL every 2 sec. Responses
to four pulse-trains (1–4) are illustrated (1 and 2 were separated by 2 sec; 3 and 4 were also
separated by 2 sec and followed 14 sec after 2). Responses to 5-shock train consisted of an
early antidromic spike, followed by orthodromic spikes displaying progressive augmentation
and spike inactivation. With repetition of pulse-trains, IPSPs elicited by preceding stimuli in
train were progressively reduced until their complete obliteration and spike-bursts contained
more action potentials with spike inactivation. Increased area of depolarization from first to
fifth responses in each pulse-train as well as from pulse-train 1 to pulse-trains 3 and 4. (B)
Brainstem-transected cat. Cortically evoked spike-bursts in thalamic VL neuron (1). Motor
cortex stimulation was applied with pulse-trains at 10 Hz delivered every 1.3 sec. In 1 the
pattern of cortically evoked responses at onset of rhythmic pulse-trains (faster speed than in
2–4). Responses in 2–4 at later stages of stimulation. Stimuli are marked by dots. In 2–4
stimuli and evoked spike-bursts are aligned. Spontaneous spike-bursts appear progressive,
resembling evoked ones, as a form of memory in corticothalamic circuit. Modified from
Steriade and Timofeev,52 and Steriade.59

ring spindle sequences.53 Among the mechanisms that may explain the long-term
increased responsiveness is the high-frequency firing in response to repeated pulse-
trains that may result in activation of high-threshold Ca2+ currents and enhanced
[Ca2+]i that may activate protein kinase A62 or Ras/mitogen-activated protein kinase,63
which are thought to be involved in memory consolidation.
FIGURE 1.6

FIGURE 1.7 Cortical augmenting responses lead to long-lasting enhancement of depolariz-

ing responses in intact cortex. Cat under barbiturate anesthesia. Intracellular recording from
electrophysiologically (left upper panel) and morphologically (left middle panel) identified
area 7 pyramidal regular-spiking neuron with thin spike (see expanded action potential close
to stained neuron). Right panel shows (from top to bottom): control response to a single
stimulus to cortex, early responses to pulse-trains at 10 Hz, responses to pulse-train with same
parameters applied 12 min later, and response to a single stimulus applied 16 min after onset
of rhythmic stimulation. Below amplitude of stimulus-evoked response at 20 ms after stimulus
onset. Initially hyperpolarizing responses became depolarizing after pulse-trains at 10 Hz.
From Timofeev et al.53

In addition to sleep spindles, the cortically generated slow oscillation is also

implicated in neuronal plasticity. Oscillations within the frequency range of the slow
(0.5–1 Hz) and delta (1–4 Hz) rhythms are implicated in cortical plasticity in the
developing visual cortex, as microelectrode recording and optical imaging show that
the effects of monocular deprivation on cortical responses are increased by a 6-hr
SWS period in the dark, and SWS deprivation blocks this enhancement.64
A puzzling issue is the development of paroxysmal activity, such as epileptiform
seizures with spike-wave (SW) complexes at ~3 Hz or SW and polyspike-wave
(PSW) complexes intermingled with fast (10–20 Hz) runs, after the progressive
enhancement of neocortical cells’ responsiveness induced by rhythmic stimulation
in the frequency range of spindles that evoke augmenting responses.65,66 Such
seizures are generated intracortically because thalamocortical neurons are steadily
hyperpolarized and silent, due to the powerful inhibition exerted by thalamic retic-
ular neurons that faithfully follow each paroxysmal depolarization of corticotha-
lamic neurons.67,68 This transformation from normal phenomena (sleep spindles or
augmenting responses leading to neuronal plasticity that may be implicated in
memory) into pathological episodes (seizures) was unexpected, as epilepsy is a state
during which memory is suspended. There is a subtle threshold beyond which
augmentation and enhanced responsiveness to control stimuli are rapidly trans-
formed into epileptiform patterns. The mechanisms and significance of this devel-
opment are now under investigation.
FUNCTIONAL SIGNIFICANCE OF SLEEP OSCILLATIONS

Sleep oscillations may determine the behavioral quiescence during this behavioral
state, rather than being simple electrical signs of it. Indeed data show that the
neuronal substrates of widely synchronized thalamic and cortical sleep oscillations
are the same as those that produce the disconnection and unresponsiveness to signals
from the external world, which are the defining features of SWS. The brain oscilla-
tions that define the transition from wakefulness to SWS and occur during early stages
of SWS, such as spindles, are associated with long periods of hyperpolarization69 and
increased membrane conductance70 in thalamocortical cells, with the consequence that
the incoming messages are blocked71 and the cerebral cortex is deprived of information
from the outside world.59,68
The thalamus is the first relay station in which afferent signals are obliterated
from the very onset of SWS. The role of spindles in disconnecting the brain from
external stimuli was also demonstrated by investigating event-related-potentials in
humans and showing that this thalamically generated oscillation gates information
processing and protects the sleeper from disturbing stimuli.72 Following the appear-
ance of these initial signs, other oscillatory types mark the late stage of SWS and
they further deepen the unresponsiveness of thalamic and cortical neurons, discon-
necting the brain from the external world.
Spindles and slow oscillations are not only operational in passively deafferenting
thalamocortical systems, but are also implicated in active cerebral functions. During
spindles, rhythmic and synchronized spike-bursts of thalamocortical neurons depo-
larize the dendrites of neocortical neurons, which is associated with massive Ca2+

entry.73 It was hypothesized74 that the massive Ca2+ entry in cortical cells’ dendrites
may provide an effective signal to efficiently activate Ca2+ calmodulin-dependent
protein kinase II (CaMKII), which is implicated in synaptic plasticity of excitatory
synapses in cortex.75 Similar phenomena occur in SWS during the rhythmic spike-
trains associated with oscillations in the frequency band of the slow (0.5–1 Hz)
oscillation, and could provide the mechanisms that have been hypothesized to con-
solidate memory traces acquired during the state of wakefulness.29 This idea is
supported by human studies demonstrating that the improvement of discrimination
tasks and formation of procedural memory depends on SWS76–78 and that training
on a declarative learning task leads to a significant enhancement of spindles’ density
in humans.79
Similar relations between SWS and memory consolidation have been postulated
in work on the hippocampus. The hypothesis that neuronal synchrony associated
with sharp potentials during SWS consolidates and transfers information to neocor-
tical fields80,81 was worked out and dendritic recordings from CA1 pyramidal
neurons82 suggested that sleep patterns are important for preservation of experience-
induced synaptic changes.81 The firing rate of a hippocampal “place cell” and the
correlation between neuronal pairs during wakefulness are increased during subse-
quent SWS epochs.83
All the above data show that, far from being a period of complete inactivity,
SWS oscillations are implicated in mental processes. Dreaming mentation appears
also during SWS, the content of dreams is closer to real life events84 than dreaming
during REM sleep, the recall rate of dreaming mentation in SWS is quite high,85
and the suggestion has been made that cortically consolidated memories, stored
during SWS by rhythmic spike-trains associated with neocortically generated
oscillations29,68 as well as the information outflow from the hippocampus, would be
integrated with other stored memories during REM sleep.86
ACKNOWLEDGMENTS
Personal experiments discussed in this chapter have been supported by grants from
the Canadian Institutes for Health Research (MT-3689 and MOP-36545), Human
Frontier Science Program (RG-0131), and National Institutes of Health-USA (RO1-
NS40522). I thank the following collaborators for their creative work: F. Amzica,
D. Contreras, R. Curró Dossi, F. Grenier, A. Nuñez, D. Paré, and I. Timofeev.
REFERENCES
1. Bremer, F., Cerveau isolé et physiologie du sommeil, C.R. Soc. Biol. (Paris) 118,
1235–1241, 1935.
2. Moruzzi, G. and Magoun, H.W., Brain stem reticular formation and activation of the
EEG, Electroencephalogr. Clin. Neurophysiol., 1, 455–473, 1949.
3. Jouvet, M., The role of monoamines and acetylcholine-containing neurons in the
regulation of the sleep–waking cycle, Ergeb. Physiol., 64, 166–307, 1972.

4. Bremer, F. and Terzuolo, C., Contribution à l’étude des mécanismes physiologiques

du maintien de l’activité vigile du cerveau. Interaction de la formation reticulée et
de l’écorce cérébrale dans le processus de l’éveil, Arch. Int. Pharmacodyn. Physiol.,
62, 157–178, 1954.
5. Pavlov, I.P., “Innere Hemmung” der bedingten Reflexe und der Schlaf — ein und
derselbe Prozess, Skand. Arch. Physiol., 44, 42–58, 1923.
6. Eccles, J.C., Chairman’s opening remarks in The Nature of Sleep, Wolstenholme,
G.E.W. and O’Connor, M., Eds., Churchill, London, 1961, pp. 1–3.
7. Steriade, M., Timofeev, I., and Grenier, F., Natural waking and sleep states: a view
from inside neocortical neurons, J. Neurophysiol., 85, 1969–1985, 2001.
8. Steriade, M., Impact of network activities on neuronal properties in corticothalamic
systems, J. Neurophysiol., 86, 1–39, 2001.
9. Steriade, M., The Intact and Sliced Brain, The MIT Press, Cambridge, MA, 2001.
10. Steriade, M., Deschênes, M., Domich, L., and Mulle, C., Abolition of spindle oscil-
lations in thalamic neurons disconnected from nucleus reticularis thalami, J. Neuro-
physiol., 54, 1473–1497, 1985.
11. Steriade, M., Domich, L., Oakson, G., and Deschênes, M., The deafferented reticularis
thalami nucleus generates spindle rhythmicity, J. Neurophysiol. 57, 260–273, 1987.
12. Steriade, M., McCormick, D.A., and Sejnowski, T.J., Thalamocortical oscillation in
the sleeping and aroused brain, Science, 262, 679–685, 1993.
13. Destexhe, A., Contreras, D., Steriade, M., Sejnowski, T.J., and Huguenard, J.R., In
vivo, in vitro and computational analysis of dendritic calcium currents in thalamic
reticular neurons, J. Neurosci., 16, 169–185, 1996.
14. Steriade, M., Jones, E.G., and Llinás, R.R., Thalamic Oscillations and Signaling,
Wiley-Interscience, New York, 1990.
15. Kim, U., Bal, T., and McCormick, D.A., Spindle waves are propagating synchronized
oscillations in the ferret LGNd in vitro, J. Neurophysiol., 74, 1301–1323, 1995.
16. Contreras, D., Destexhe, A., Sejnowski, T.J., and Steriade, M., Control of spatiotem-
poral coherence of a thalamic oscillation by corticothalamic feedback, Science, 274,
771–774, 1996.
17. Contreras, D., Destexhe, A., Sejnowski, T.J., and Steriade, M., Spatiotemporal patterns
of spindle oscillations in cortex and thalamus, J. Neurosci., 17, 1179–1196, 1997.
18. Contreras, D. and Steriade, M., Spindle oscillation: the role of corticothalamic feed-
back in a thalamically generated rhythm, J. Physiol. (Lond.), 490, 159–179, 1996.
19. Steriade, M., Wyzinski, P., and Apostol, V., Corticofugal projections governing rhyth-
mic thalamic activity, in Corticothalamic Projections and Sensorimotor Activities,
Frigyesi, T.L., Rinvik, E., and Yahr, M.D., Eds., Raven Press, New York, 1972, pp.
221–272.
20. Contreras, D., Destexhe, A., and Steriade, M., Spindle oscillations during cortical
spreading depression in naturally sleeping cats, Neuroscience, 77, 933–996, 1997.
21. Gottselig, J.M., Bassetti, C.L. and Achermann, P., Power and coherence of sleep
spindle frequency activity following hemispheric strokes, Brain, 125, 373–383, 2002.
22. Timofeev, I., Bazhenov, M., Sejnowski, T.J., and Steriade M., Contribution of intrinsic
and synaptic factors in the desynchronization of thalamic oscillatory activity, Thal.
& Rel. Syst., 1, 53–69, 2001.
23. Bal, T. and McCormick, D.A., What stops synchronized thalamocortical oscillations?
Neuron, 17, 297–308, 1996.
24. Leresche, N., Jassik-Gerschenfeld, D., Haby, M., Soltesz, I., and Crunelli, V., Pace-
maker-like and other types of spontaneous membrane potential oscillations of
thalamocortical cells, Neurosci. Lett., 113, 72–77, 1990.

25. McCormick, D.A. and Pape, H.C., Properties of a hyperpolarization-activated cation

current and its role in rhythmic oscillation in thalamic relay neurones, J. Physiol.
(Lond.), 431, 291–318, 1990.
26. Soltesz, I., Lightowler, S., Leresche, N., Jassik-Gerschenfeld, D., and Crunelli, V.,
Two inward currents and the transformation of low-frequency oscillations of rat and
cat thalamocortical cells, J. Physiol. (Lond.), 441, 175–197, 1991.
27. Steriade, M., Curró Dossi, R., and Nuñez, A., Network modulation of a slow intrinsic
oscillation of cat thalamocortical neurons implicated in sleep delta waves: cortical
potentiation and brainstem cholinergic suppression, J. Neurosci., 11, 3200–3217,
1991.
28. Curró Dossi, R., Nuñez, A., and Steriade, M., Electrophysiology of a slow (0.5–4
Hz) intrinsic oscillation of cat thalamocortical neurones in vivo, J. Physiol. (Lond.),
447, 215–234, 1992.
29. Steriade, M., Contreras, D., Curró Dossi, R., and Nuñez, A., The slow (<1 Hz)
oscillation in reticular thalamic and thalamocortical neurons: scenario of sleep rhythm
generation in interacting thalamic and neocortical networks, J. Neurosci., 13,
3284–3299, 1993.
30. Amzica, F. and Steriade, M., The functional significance of K-complexes, Sleep Med.
Rev., 6, 139–149, 2002.
31. Steriade, M., Nuñez, A., and Amzica, F., A novel slow (<1 Hz) oscillation of neo-
cortical neurons in vivo: depolarizing and hyperpolarizing components, J. Neurosci.,
13, 3252–3265, 1993.
32. Steriade, M., Nuñez, A. and Amzica, F., Intracellular analysis of relations between
the slow (<1 Hz) neocortical oscillation and other sleep rhythms, J. Neurosci., 13,
3266–3283, 1993.
33. Timofeev, I., Grenier, F., Bazhenov, M., Sejnowski, T.J., and Steriade, M. Origin of
slow oscillations in deafferented cortical slabs, Cerebr. Cortex, 10, 1185–1199, 2000.
34. Sanchez-Vives, M.V. and McCormick, D.A., Cellular and network mechanisms of
rhythmic recurrent activity in neocortex, Nat. Neurosci., 3, 1027–1034, 2000.
35. Timofeev, I. and Steriade, M., Low-frequency rhythms in the thalamus of intact-
cortex and decorticated cats, J. Neurophysiol., 76, 4152–4168, 1996.
36. Steriade, M., Amzica, F., and Contreras, D., Synchronization of fast (30–40 Hz)
spontaneous cortical rhythms during brain activation, J. Neurosci., 16, 392–417, 1996.
37. Achermann, P. and Borbély, A., Low-frequency (<1 Hz) oscillations in the human
sleep EEG, Neuroscience, 81, 213–222, 1997.
38. Amzica, F. and Steriade, M., The K-complex: its slow (<1 Hz) rhythmicity and
relation to delta waves, Neurology, 49, 952–959, 1997.
39. Simon, N.R., Mandshanden, I., and Lopes da Silva, F.H., A MEG study of sleep,
Brain Res. 860, 64–76, 2000.
40. Contreras, D. and Steriade, M., Cellular basis of EEG slow rhythms: a study of
dynamic corticothalamic relationships, J. Neurosci., 15, 604–622, 1995.
41. Timofeev I., Grenier F, and Steriade M., Disfacilitation and active inhibition in the
neocortex during the natural sleep–wake cycle: an intracellular study, Proc. Natl.
Acad. Sci USA, 98, 1924–1929, 2001.
42. Contreras, D., Timofeev, I., and Steriade, M., Mechanisms of long-lasting hyperpo-
larizations underlying slow sleep oscillations in cat corticothalamic networks, J.
Physiol. (Lond.), 494, 251–264, 1996.
43. Massimini M. and Amzica F., Extracellular calcium fluctuations and intracellular
potentials in the cortex during the slow sleep oscillation, J. Neurophysiol., 85,
1346–1350, 2001.

44. Bazhenov, M., Timofeev, I., Steriade, M., and Sejnowski, T.J., Model of thalamocor-
tical slow-wave sleep oscillations and transitions to activated states, J. Neurosci., 22,
8691–8704, 2002.
45. Steriade, M., Amzica, F., and Nuñez, A., Cholinergic and noradrenergic modulation of
the slow (~0.3 Hz) oscillation in neocortical cells, J. Neurophysiol., 70, 1384–1400, 1993.
46. Mölle, M., Marshall, L., Gais, S., and Born, J., Grouping of spindle activity during
slow oscillations in human non-REM sleep, J. Neurosci., 22, 10941–10947, 2002.
47. Buonomano, D.V. and Merzenich, M.M., Cortical plasticity: from synapses to maps,
Ann. Rev. Neurosci., 21, 149–186, 1998.
48. Marder, E., From biophysics to models of network function, Ann. Rev. Neurosci., 21,
25–45, 1998.
49. Turrigiano, G.G., Leslie, K.R., Desai, N.S., Rutherford, L.C., and Nelson, S.B.,
Activity-dependent scaling of quantal amplitude in neocortical neurons, Nature, 391,
892–896, 1998.
50. Moruzzi, G., The functional significance of sleep with particular regard to the brain
mechanisms underlying consciousness, in Brain and Conscious Experience, Eccles,
J.C., Ed., Springer, New York, 345–379, 1966.
51. Morison, R.S. and Dempsey, E.W., Mechanism of thalamocortical augmentation and
repetition, Am. J. Physiol., 138, 297–308, 1942.
52. Steriade, M. and Timofeev, I., Short-term plasticity during intrathalamic augmenting
responses in decorticated cats, J. Neurosci., 17, 3778–3795, 1997.
53. Timofeev, I., Grenier, F., Bazhenov, M., Houweling, A., Sejnowski, T.J., and Steriade,
M., Short- and medium-term plasticity associated with augmenting responses in
cortical slabs and spindles in intact cortex of cats in vivo, J. Physiol. (Lond.), 542,
583–598, 2002.
54. Castro-Alamancos, M.A. and Connors, B.W., Cellular mechanisms of the augmenting
response: short-term plasticity in a thalamocortical pathway, J. Neurosci., 16,
7742–7756, 1996.
55. Jasper H.H., Diffuse projection systems: the integrative action of the thalamic reticular
system, Electroencephalogr. Clin. Neurophysiol., 1, 405–420, 1949.
56. Timofeev, I. and Steriade, M., Cellular mechanisms underlying intrathalamic augment-
ing responses of reticular and relay neurons, J. Neurophysiol., 79, 2716–2729, 1998.
57. Bazhenov, M., Timofeev, I., Steriade, M., and Sejnowski, T.J., Cellular and network
models for intrathalamic augmenting responses during 10-Hz stimulation, J. Neuro-
physiol., 79, 2730–2748, 1998.
58. Bazhenov, M., Timofeev, I., Steriade, M., Sejnowski, T.J., Self-sustained rhythmic
activity in the thalamic reticular nucleus mediated by depolarizing GABAA receptor
potentials, Nat. Neurosci., 2, 168–174, 1999.
59. Steriade, M., Alertness, quiet sleep, dreaming, in Cerebral Cortex (vol. 9, Normal
and Altered States of Function), Peters, A. and Jones, E.G., Eds., Plenum, New York,
1991, pp. 279–357.
60. Steriade, M., Development of evoked responses into self-sustained activity within
amygdalo-hippocampal circuits, Electroencephalogr. Clin. Neurophysiol., 16,
221–236, 1964.
61. Steriade, M., Timofeev, I., Grenier, F., and Dürmüller, N., Role of thalamic and
cortical neurons in augmenting responses: dual intracellular recordings in vivo, J.
Neurosci., 18, 6425–6443, 1998.
62. Abel, T., Nguyen, P.V., Barad, M., Deuel, T.A., Kandel, E.R., and Bourtchouladze,
R., Genetic demonstration of a role for PKA in the late phase of LTP and in hippo-
campus-based long-term memory, Cell, 88, 615–626, 1997.

63. Dolmetsch, R.E., Pajvani, U., Fife, K., Spotts, J.M., and Greenberg, M.E., Signaling
to the nucleus by an L-type calcium channel-calmodulin complex through the MAP
kinase pathway, Science, 294, 333–339, 2001.
64. Frank, M.G., Issa, N.P., and Stryker, M.P., Sleep enhances plasticity in the developing
visual cortex, Neuron, 30, 275–287, 2001.
65. Steriade, M. and Amzica, F., Dynamic coupling among neocortical neurons during
evoked and spontaneous spike-wave seizure activity, J. Neurophysiol., 72, 2051–2069,
1994.
66. Steriade, M., Amzica, F., Neckelmann, D., and Timofeev, I., Spike-wave complexes
and fast runs of cortically generated seizures, II, Extra- and intracellular patterns. J.
Neurophysiol., 80, 1456–79, 1998.
67. Steriade, M. and Contreras, D., Relations between cortical and thalamic cellular
events during transition from sleep pattern to paroxysmal activity, J. Neurosci., 15,
623–642, 1995.
68. Steriade, M., Neuronal Substrates of Sleep and Epilepsy, Cambridge Univ. Press,
Cambridge (UK), 2003.
69. Hirsch, J.C., Fourment, A., and Marc, M.E., Sleep-related variations of membrane
potential in the lateral geniculate body relay neurons of the cat, Brain Research, 259,
308–312, 1983.
70. Timofeev, I., Contreras, D., and Steriade, M., Synaptic responsiveness of cortical and
thalamic neurons during various phases of slow oscillation in cat, J. Physiol. (Lond.),
494, 265–278, 1996.
71. Steriade, M., Iosif, G., and Apostol, V., Responsiveness of thalamic and cortical motor
relays during arousal and various stages of sleep, J. Neurophysiol., 32, 251–265, 1969.
72. Elton, M., Winter, O., Heslenfeld, D., Loewy, D., Campbell, K., and Kok, A., Event-
related potentials to tones in the absence and presence of sleep spindles, J. Sleep
Res., 6, 78–83, 1997.
73. Yuste, R. and Tank, D.W., Dendritic integration in mammalian neurons, a century
after Cajal, Neuron, 16, 701–716, 1996.
74. Sejnowski, T.J. and Destexhe, A., Why do we sleep?, Brain Res., 886, 208–223, 2000.
75. Soderling, T.R. and Derkach, V.A., Postsynaptic protein phosphorylation and LTP,
Trends Neurosci., 23, 75–80, 2000.
76. Gais, S., Plihal, W., Wagner, U., and Born, J., Early sleep triggers memory for early
visual discrimination skills, Nat. Neurosci., 3, 1335–1339, 2000.
77. Stickgold, R., James, L., and Hobson, J.A., Visual discrimination learning requires
sleep after training, Nat. Neurosci., 3, 1237–1238, 2000.
78. Stickgold, R., Whitbee, D., Schirmer, B., Patel, V., and Hobson, J.A., Visual discrim-
ination improvement. A multi-step process occurring during sleep, J. Cogn. Neurosci.,
12, 246–254, 2000.
79. Gais, S., Mölle, M., Helms, K., and Born, J., Learning-dependent increases in sleep
density, J. Neurosci., 22, 6830–6834, 2002.
80. Buzsáki, G., Two-stage model of memory trace formation: a role for “noisy” brain
states, Neuroscience, 31, 551–570, 1989.
81. Buzsáki, G., Memory consolidation during sleep: a neurophysiological perspective,
J. Sleep Res., 7 (Suppl. 1), 17–23, 1998.
82. Kamondi, A., Acsády, L., and Buzsáki, G., Dendritic spikes are enhanced by coop-
erative network activity in the intact hippocampus, J. Neurosci., 18, 3919–3928, 1998.
83. Wilson, M.A. and McNaughton, B.L., Reactivation of hippocampal ensemble mem-
ories during sleep, Science, 265, 676–679, 1994.

84. Hobson, J.A., Pace-Schott, E., and Stickgold, R., Dreaming and the brain: toward a
cognitive neuroscience of conscious states, Brain Behav. Sci., 23, 793–842, 2000.
85. Nielsen, T., Cognition in REM and NREM sleep, Brain Behav. Sci., 23, 851–866,
2000.
86. Hobson, J.A. and Pace-Schott, E.F., The cognitive neuroscience of sleep: neuronal
systems, consciousness and learning, Nat. Rev. Neurosci., 3, 679–693, 2002.

2 Role of Basalo-Cortical
System in Modulating
Cortical Activity and
Sleep-Wake States
Maan Gee Lee and Barbara E. Jones
CONTENTS
Introduction
Discharge Properties of Identified Cholinergic and GABAergic
Basal Forebrain Neurons in Anesthetized Rats
Discharge Properties of Basal Forebrain Neurons in Head-Fixed Rats
Role of Cholinergic and GABAergic Basal Forebrain Neurons
Summary
Acknowledgments
References
INTRODUCTION
Cortical activity varies in association with behavior and sleep-wake states. It is
predominantly fast during waking and paradoxical sleep (PS or rapid eye movement
sleep, REMS) and slow during quiet or slow wave sleep (SWS; see Figure 2.1). The
fast activity reflects cortical activation that is stimulated and maintained by subcor-
tical activating systems. These systems originate in the rostral brainstem where
glutamatergic neurons of the reticular formation, cholinergic pontomesencephalic
tegmental neurons, and noradrenergic locus coeruleus neurons collectively comprise
critical activating systems.1,2 They project forward through a dorsal pathway to the
midline and intralaminar thalamic nuclei that form the nonspecific thalamo-cortical
projection system. Excited by brainstem inputs, this thalamo-cortical system stim-
ulates in turn widespread cortical activation3 (see for review Jones4 and Steriade,
this volume). In parallel with this pathway, ascending projections from the brainstem
pass ventrally through the hypothalamus to reach the basal forebrain. Excited by the
brainstem inputs, the basalo-cortical system also stimulates widespread cortical
0-8493-1519-0/05/$0.00+$1.50

FIGURE 2.1 Cortical activation and slow wave (SW) promoting systems. Schematic sagittal
view of rat brain showing the major relays of the arousal systems and their excitatory pathways
(arrows) involved in promoting EEG fast activity (bottom EEG trace) characteristic of waking
and paradoxical sleep (PS or REMS). The major ascending pathways emerge from the
brainstem reticular formation to ascend along a dorsal trajectory into the thalamus where they
terminate upon nuclei of the nonspecific thalamo-cortical projection system and a ventral
trajectory through the lateral hypothalamus up to the basal forebrain, where they terminate
upon neurons of the basalo-cortical projection system (in the substantia innominata).
From the basal forebrain neurons containing acetylcholine (ACh, circle) give rise to
widespread projections to the cerebral cortex to excite cortical neurons and promote fast
activity. Evidence is presented to show that the cholinergic neurons discharge (On) in asso-
ciation with cortical activation (bottom EEG trace). They are excited by inputs from the
brainstem arousal systems, including glutamatergic neurons of the reticular formation, nora-
drenergic neurons of the locus coeruleus and cholinergic neurons of the laterodorsal and
pedunculopontine tegmental nuclei, as well as histaminergic and orexinergic neurons of the
tuberomammillary nucleus and posterior hypothalamus.
Neurons containing GABA (triangle) are also located in the basal forebrain. They give
rise to inhibitory projections (ending as blocks) that go to the cortex, the posterior hypothal-
amus, or brainstem as well as to local neurons in the basal forebrain. Evidence is presented
to show that particular GABAergic neurons discharge maximally with cortical slow waves
(top EEG trace) and minimally with cortical activation (Off). Particular GABAergic cell
groups can thus promote SWS by directly modulating cortical activity or by indirectly
attenuating cortical activation through inhibition of hypothalamic and brainstem arousal
systems or local cholinergic neurons. Illustration adapted with permission from Jones.53
Abbreviations: 7g, 7th nerve genu; ac, anterior commissure; CPu, caudate-putamen; Cx, cortex;
Cu, cuneate nucleus; GP, globus pallidus; Hi, hippocampus; ic, internal capsule; LC, locus
coeruleus; LDTg, laterodorsal tegmental nucleus; opt, optic tract; PH, posterior hypothalamus;
POA, preoptic area; PPTg, pedunculopontine tegmental nucleus; RF Gi, gigantocellular retic-
ular formation; RF Mes, mesencephalic reticular formation; RF PnC, pontis caudalis reticular
formation; RF PnO, pontis oralis reticular formation; Rt, reticularis nucleus of the thalamus;
s, solitary tract; scp, superior cerebellar peduncle; SI, substantia innominata; Sol, solitary
tract nucleus; SN, substantia nigra; Th, thalamus; TM, tuberomammillary nucleus; VN,
vestibular nuclei; VTA, ventral tegmental area.

activation. This ventral extra-thalamic relay to the cerebral cortex can even be
sufficient since lesion of the thalamus does not eliminate cortical activation.4,5
The basalo-cortical projection is composed importantly of cholinergic neurons6,7
(Figure 2.1 and Figure 2.2). Acetylcholine depolarizes and excites cortical neurons
to stimulate tonic firing and resulting fast cortical activity.8,9 Pharmacological block
FIGURE 2.2 Basal forebrain cells and Neurobiotin (Nb)-labeled cell. (A) Distribution of
cholinergic cells (small open circles) and GABAergic cells (small open triangles) in basal
forebrain. (Adapted with permission from Gritti et al.21) Identified cells are indicated that
were recorded and labeled in urethane-anesthetized rats as Nb+/ChAT+ (filled circle, see
Figure 2.3) and Nb+/GAD+ Off (burst) (filled triangle, see Figure 2.4). Units recorded in
unanesthetized head-fixed rats are indicated that were “like” the identified Nb+/ChAT+ cells
(large open circle, see Figure 2.5) and Nb+/GAD+ Off (burst) cells (large open triangle, see
Figure 2.6). Magnification bar = 1 mm. (B) A cell recorded and juxtacellularly labeled in the
head-fixed rat. The Nb-labeled cell was revealed in fluorescence using Streptavidin-Cy2.
(Adapted with permission from Lee et al.32) Magnification bar = 20 mm.
Abbreviations: CPu, caudate-putamen; GP, globus pallidus; FStr, fundus striatum; LPO, lateral
preoptic area; MCPO, magnocellular preoptic nucleus; MPO, medial preoptic nucleus; OTu,
olfactory tubercle; Pir, piriform cortex; SIa, substantia innominata pars anterior.

of muscarinic receptors with atropine diminishes fast cortical activity which is

replaced by slow wave activity.10,11 Yet the way in which cholinergic neurons spe-
cifically modulate the cerebral cortex is not known, since single-unit recording has
found that basal forebrain neurons comprise a very heterogeneous population of
cells. Indeed, although many units discharge at their highest rate in association with
cortical fast activity, others discharge at their highest rate in association with cortical
slow wave activity.12–14 Although electrical stimulation can evoke cortical activa-
tion,15 it can also induce slow wave EEG activity and SWS.16,17 Although lesions
have been reported to diminish cortical activation,11,18 they have also been shown
to decrease sleep.19,20 Collectively, these results suggest the presence of sleep-
promoting in addition to wake- or cortical activation-promoting neurons in the basal
forebrain that would presumably correspond to noncholinergic and cholinergic
neurons, respectively.
In addition to cholinergic neurons, a large population of GABAergic neurons is
distributed through the basal forebrain21 (Figure 2.1 and Figure 2.2 A). These cells
are heterogeneous in their size and projections, magnocellular neurons projecting
together with the cholinergic neurons to the cerebral cortex, smaller neurons pro-
jecting caudally to the posterior hypothalamus or brainstem, and others presumably
projecting locally onto cholinergic and other basal forebrain cells.22–24 GABAergic
neurons could thus play varied and different roles as compared to cholinergic neurons
in the modulation of cortical activity and sleep–wake states. Clearly, knowing the
transmitter phenotype and projection pathway of recorded units is critical for under-
standing the specific roles of GABAergic and cholinergic basal forebrain neurons
in cortical modulation and sleep–wake states.
DISCHARGE PROPERTIES OF IDENTIFIED

CHOLINERGIC AND GABAERGIC BASAL FOREBRAIN
NEURONS IN ANESTHETIZED RATS
The prominent properties of cholinergic and noncholinergic basal forebrain neurons
were originally studied in vitro by intracellular recording and labeling with Neuro-
biotin (Nb) for subsequent identification using immunohistochemical staining for
choline acetyltransferase (ChAT).25,26 The distinctive intrinsic properties manifested
by cholinergic neurons in vitro provided clues for their subsequent identification in
vivo. For the in vivo studies, extracellular recordings were performed using glass
micropipettes, and recorded units were labeled with Nb by applying the juxtacellular
technique27,28 (Figure 2.2 B). Equivalent to single cell electroporation,29 juxtacellular
labeling is achieved by passing current pulses through an Nb-filled micropipette that
is touching or in close proximity to the membrane of the recorded cell. Sections
were immunostained for ChAT or glutamic acid decarboxylase (GAD) to determine
whether the Nb-labeled cells were cholinergic or GABAergic. Using urethane anes-
thesia, it was possible to characterize units according to their discharge properties
and the relationship of their discharge to slow wave versus stimulation-induced faster
EEG activity.28,30

FIGURE 2.3 Discharge pattern of Nb+/ChAT+ neuron in the MCPO (see Figure 2.2 A)
recorded in a urethane-anesthetized rat. As shown by the EEG (in A) and unit discharge rate
(from the peristimulus histogram, PSH, in B), the average spike rate increases in association
with somatosensory stimulation-evoked cortical activation. In the expanded traces (C and D)
for the prestimulation (left) and stimulation (right) conditions, it is evident that the unit
discharge changes from irregular tonic discharge in association with cortical slow wave
activity to rhythmic bursting discharge in association with cortical activation characterized
by theta-like activity together with enhanced gamma activity. (Copied with permission from
Manns, I.D., Alonso, A., and Jones, B.E., J. Neurosci., 20, 1505–1518, 2000.)
In urethane-anesthetized rats, all identified cholinergic neurons increased their

discharge rate in association with stimulation-induced cortical activation (identified
as Nb+/ChAT+ “On” cells28 (Figure 2.2 A and Figure 2.3). Cortical activation was
marked by the replacement of irregular slow wave activity with rhythmic theta-like
activity upon which was riding high frequency gamma activity (30–60 Hz, Figure
2.3 C). The majority of cholinergic neurons not only increased their rate of discharge
with cortical activation but also changed their pattern of discharge from an irregular
slow pattern to a rhythmic bursting discharge (Figure 2.3 D, with an average
intraburst frequency of ~70 Hz). This bursting discharge likely emerged from the
low threshold calcium spikes that had been shown in vitro to provide the cholinergic
neurons with the capacity to discharge in rhythmically recurring high frequency
spike bursts.26 In vivo the rhythmic bursting was correlated with the rhythmic theta-
like EEG activity. These results suggested that cholinergic basal forebrain neurons

could elicit cortical activation by stimulating high-frequency gamma activity together

with theta through rhythmic modulation of the cerebral cortex (Figure 2.1).
In the same preparation, identified GABAergic neurons were found to be het-
erogeneous in their discharge profiles and properties.30 A substantial group of
GABAergic neurons (40%) increased their discharge rate in association with cortical
activation and discharged tonically at relatively high frequencies in the gamma EEG
range of activity (identified as Nb+/GAD+ “On” [tonic] cells). Some of these could
be antidromically activated from the cerebral cortex.
However, the majority of GABAergic neurons (60%) decreased their rate of
discharge in association with cortical activation (identified as Nb+/GAD+ “Off”
cells). Some of these discharged in a very slow, irregular tonic manner (<10 Hz)
during slow-wave EEG activity and turned off during cortical activation (called
Nb+/GAD+ Off [tonic] cells). None of these could be antidromically activated from
the cerebral cortex and were thus thought to exert their influence either caudally on
the posterior hypothalamus or brainstem or locally in the basal forebrain. Other
GABAergic neurons discharged in bursts of spikes (~200 Hz) in association with
cortical slow wave activity to turn off with stimulation-induced cortical activation
(called Nb+/GAD+ Off [burst] cells, Figure 2.2 A and Figure 2.4). Their discharge
was correlated with the irregular slow waves. They could be antidromically activated
from the cerebral cortex and accordingly exert a direct inhibitory influence upon
cortical neurons during slow wave EEG activity. It was thus hypothesized that
particular GABAergic basal forebrain neurons could be more active with slow wave
EEG during SWS than with cortical activation during waking and could thus inhibit
the discharge of other subcortical or cortical neurons involved in stimulating cortical
activation (Figure 2.1).
Recording of identified cholinergic and GABAergic neurons in relation to EEG
activity in urethane-anesthetized rats thus revealed different cell groups that could
differentially modulate cortical activity and sleep-wake states. However, study of
their discharge profiles in unanesthetized animals is necessary to determine their
different roles in modulating natural EEG activity and sleep-wake states. For this
purpose recording and juxtacellular labeling were undertaken in head-fixed rats.
DISCHARGE PROPERTIES OF BASAL FOREBRAIN

NEURONS IN HEAD-FIXED RATS
In order to record and label neurons during natural sleep–wake states, rats were
implanted with a u-frame, along with chronically indwelling EEG and EMG elec-
trodes, to fix their heads to a special carriage in the stereotaxic frame and permit
recording of single units together with other polygraphic variables.31 Following an
~1-week adaptation to the head-fixation, recording could be performed during the
afternoon when the rat passes naturally through states of wake, SWS, and PS.
Except during very large movements, units could usually be held through at least
one full sleep–wake cycle to be labeled with Nb at the end of at least one complete
cycle. Although many units could be recorded in this manner in one animal, only

FIGURE 2.4 Discharge pattern of Nb+/GAD+ Off (burst) neuron (see Figure 2.2 A) recorded
in a urethane-anesthetized rat. As shown by the EEG (in A) and unit discharge rate (from the
peristimulus histogram, PSH, in B), the average spike rate decreases in association with
somatosensory stimulation-evoked cortical activation. In the expanded traces (C and D) for
the prestimulation (left) and stimulation (right) conditions, it is evident that the unit discharges
in an irregular bursting pattern in association with cortical slow wave activity and stops firing
in association with cortical activation. (Copied with permission from Manns, I.D., Alonso,
A., and Jones, B.E., J. Neurosci., 20, 9252–9263, 2000.)
one unit could be labeled with Nb per side to insure unequivocal identification of
the recorded unit.
Applying this procedure a large sample of units has been recorded in the basal
forebrain cholinergic cell area and a small sample of those labeled with Nb using
the juxtacellular technique (Figure 2.2). In this sampling, a diverse population of
cells was characterized according to discharge profiles in relation to sleep–wake
states.32 The vast majority (90%) manifested significant variation in their average
discharge rate as a function of state and showed maximum and minimum rates in
different states to form (12) multiple distinct cell groups. Some of these shared
properties with the cells previously identified in the urethane-anesthetized rats as
the cholinergic cells that discharged with cortical activation (Nb+/ChAT+ On) and
others with the GABAergic cells which discharged maximally with slow wave
activity (Nb+/GAD+ Off). Awaiting immunohistochemical identification of an ade-

quate number of Nb-labeled cells with these properties, we describe these two cell
types here as “cholinergic-like” and “GABAergic-Off-like.”
The major proportion of cells (~75%) in the basal forebrain (MCPO and SI
area, Figure 2.2 A) discharged maximally during wake or PS in association with
cortical activation. Comprising a small proportion (<10%), one group of cells
discharged in rhythmic bursts of spikes in association with EEG theta activity
during wake and PS in the head-fixed rats (Figure 2.2 A and Figure 2.5 A–C32) in
a manner similar to the identified cholinergic cells recorded in urethane-anesthe-
tized rats (above). According to the frequency of the primary mode of the interspike
interval histogram (ISIH, Figure 2.5 D), the average instantaneous firing frequency
within the bursts was remarkably similar to that of identified cholinergic cells in
the anesthetized rats (above). According to the autocorrelation histogram (ACH,
Figure 2.5 E), their discharge was rhythmic during waking and PS or transition to
PS, when theta occurred. Their average discharge rate was highest in PS (thus
classified as P-max), lower in active wake (aW) and lowest in SWS (Figure 2.5
F). Across epochs it was significantly positively correlated with gamma and theta
and negatively correlated with delta EEG activity, while not being significantly
correlated with EMG activity (see Figure 2.5 G–J). These “cholinergic-like” cells
thus discharge in rhythmic bursts in association with theta activity during active
FIGURE 2.5 (See facing page.) Discharge properties and profile of a unit firing with
cortical activation of PS and aW in the head-fixed rat and having similar discharge
properties as identified Nb+/ChAT+ cells in the anesthetized rat (see Figure 2.3). Classified
as P-max (from group 11: wsP [Fast, Phasic, Rhythmic] [#u210]), it discharges on average
at an intermediate rate during aW (A), a minimal rate during SWS (B), and a maximal
rate during PS (C). It fires in bursts of spikes during PS (with a peak mode of ~68 Hz in
the ISIH, D) that recur rhythmically at a theta frequency (with a frequency of 7.0 Hz in
the ACH, E). The rhythmic bursting also occurs with the appearance of theta during the
transition into PS (tPS). It is correlated with the theta EEG activity of the retrosplenial
cortex (see expanded trace of 1-sec period of unit activity and RS EEG in C (a) during
PS, tPS and active periods of wake showing theta. The average spike rate (F) is moderately
high in aW (7.28 Hz), minimal in SWS (0.74 Hz) and maximal in PS (11.23 Hz). The
spike rate is significantly positively correlated with gamma (G, r = 0.43, n = 84 observations,
p <.001) and theta (H, r = 0.50), significantly negatively correlated with delta (I, r = – 0.80),
while being weakly positively correlated with EMG (J, r = 0.26). In this and Figure 2.6,
the unit discharge is presented with EEG and EMG activity for 10 sec epochs of aW (A),
SWS (B), and PS (C). Spike amplitude is cut at 1 mV. The unit discharge is analyzed for
instantaneous firing frequency (from the peak of the primary mode of the interspike interval
histogram, ISIH, D) and rhythmicity of discharge (from the autocorrelation histogram,
ACH, E). Average spike rate (per second displayed on a linear scale, F), gamma EEG
power (30–58 Hz, G), ratio of theta (4.5–8 Hz)/delta (1–4 Hz) EEG power (H), delta EEG
power (1–4.5 Hz, I) and EMG amplitude (30–100 Hz, J) are displayed per state together
with the s.e.m.s. Abbreviations: active wake (aW), quiet wake (qW), transition to SWS
(tSWS), slow wave sleep (SWS), transition to PS (tPS) and paradoxical sleep (PS). (Copied
with permission from Lee, M.G., Manns, I.D., Alonso, A., and Jones, B.E., J. Neurophysiol.,
in press.)

waking and PS and may accordingly stimulate theta in addition to gamma during
these states.
A small proportion of cells (<10%) recorded in the basal forebrain discharged
at lower average rates with cortical activation than cortical slow wave activity and
thus at lower rates during wake and/or PS than SWS in head-fixed rats32 in a manner
similar to that of identified GABAergic Off cells in urethane-anesthetized rats
(above). Commonly discharging maximally in SWS (called S-max), these cells
nonetheless had varying properties. Half of them discharged in an irregular tonic
fashion like the Nb+/GAD+ Off (tonic) cells, showing slow average rates and
FIGURE 2.5

instantaneous firing frequencies (<10 Hz) during SWS and lower rates (~50% on
average) during wake and PS.
Another half discharged in high frequency bursts of spikes with cortical slow
wave activity (Figure 2.2 A and Figure 2.6 A–C) in a manner similar to that of the
identified Nb+/GAD+ Off (burst) cells in the anesthetized animals. According to the
frequency of the primary mode of the ISIH (Figure 2.6 D), their average instanta-
neous firing frequency within the bursts was within the range of that of the identified
Nb+/GAD+ Off (burst) cells (above). According to the ACH (Figure 2.6 E), their
burst discharge was not rhythmic but irregular. Their average discharge rate was
highest in SWS and equivalently low in active wake (aW) and PS (Figure 2.6 F).
Across epochs, it was significantly positively correlated with delta and negatively
correlated with gamma and theta EEG activity, while not correlated with EMG
activity (see Figure 2.6 G–J). These GABAergic Off (burst)-like cells accordingly
discharge in irregular bursts in association with delta activity during SWS and could
thus contribute to modulating this cortical slow activity. Collectively, particular
putative GABAergic cell groups could promote SWS by inhibiting neurons of the
posterior hypothalamus, brainstem or basal forebrain activating systems and also by
directly inhibiting cortical neurons to attenuate cortical activation and promote slow
wave EEG activity (Figure 2.1).
ROLE OF CHOLINERGIC AND GABAERGIC BASAL

FOREBRAIN NEURONS
Neurotoxic lesions of the basal forebrain have been reported to diminish cortical
activation on the one hand11 and SWS20 on the other. Such different effects may be
explained by the diverse cell population of the basal forebrain comprised by cortical
activation On and promoting neurons on the one hand and cortical activation Off
and slow wave promoting cells on the other. Evidence is presented here that these
FIGURE 2.6 (See facing page.) Discharge properties and profile of a unit firing with slow
wave activity during SWS in the head-fixed rat and in a manner similar to Nb+/GAD+ Off
(burst) cells in the anesthetized rat (see Figure 2.4). Classified as an S-max unit (from group
6: wSp [Fast, Phasic] unit [#c16u02]), it discharges on average at a minimal rate during aW
(A), a maximal rate during SWS (B), and equivalent minimal rate during PS (C). As evident
in the recording (see expanded trace of 500 msec period of unit activity in B [a]), the ISIH
(D) and ACH (E), the unit discharges in a distinctly phasic manner with high frequency bursts
(114 Hz peak frequency of the principal mode of the ISIH). This bursting occurs maximally
during SWS, although it is also evident (by central peak in ACHs) during other states with
a much lower incidence. The average spike rate (F) increased from aW (1.6 Hz) in the tSWS
to be highest during SWS (3.8 Hz) and decreased in tPS to be equivalently low in PS (1.8
Hz) as in aW. The spike rate was not significantly correlated with gamma (G, r = 0.13), was
significantly negatively correlated with theta (H, r = –0.30), and was significantly positively
correlated with delta EEG power (I, r = 0.53, n = 180 observations, p <.001), while being
significantly negatively correlated with EMG amplitude (J, r = –0.38). See Figure 2.5 for
general details. (Copied with permission from Lee, M.G., Manns, I.D., Alonso, A., and Jones,
B.E., J. Neurophysiol., in press.)

two seemingly opposing cell groups are comprised of cholinergic and particular
GABAergic neurons, respectively. Through both muscarinic and nicotinic receptors,
cholinergic neurons would excite cortical neurons,9,33 whereas GABAergic neurons
would inhibit cortical neurons. Such opposing actions from basal forebrain have
been demonstrated by electrical stimulation, whereby cortical neuronal discharge
with acetylcholine (ACh) release was evoked from some sites and inhibition of
discharge with no ACh release was evoked from adjacent sites.34 These different
effects can be explained by differential stimulation of intermingled yet clustered
cholinergic and GABAergic basalo-cortical projecting neurons. The way in which
FIGURE 2.6

these two opponent cell groups might modulate cortical activity and sleep-wake
states is revealed here by their specific activity profiles.
In both the anesthetized and naturally sleeping–waking rat, we have documented
an increased rate and altered pattern of discharge in cholinergic and cholinergic-like
neurons in association with cortical activation of waking and PS. Moreover the
neurons were found to discharge in rhythmic bursts cross-correlated with cortical
theta activity, which these neurons are thus hypothesized to stimulate. This modu-
lation could be effected in the cerebral cortex by fast nicotinic actions of ACh upon
some interneurons,33,35 which in turn would pace the activity of pyramidal neurons
in the cortex as they do in the hippocampus in the production of theta.36 The slower
muscarinic actions of ACh could also excite certain interneurons along with pyra-
midal cells to stimulate high frequency gamma activity that rides upon theta.36,37
The bursting discharge by the cholinergic cells would maximize release of ACh38 to
attain large populations of cortical neurons and thus provide a means for synchro-
nizing activity across distributed cortical networks during the cortical activation of
active waking and PS.
Our pharmacological experiments in freely moving rats have provided support
for the role of cholinergic basalo-cortical neurons in generating theta and gamma
in the cerebral cortex along with the states of wake and PS.39,40 From the pharma-
cological results of in vitro studies upon identified cholinergic neurons, it is known
that cholinergic neurons are excited by the transmitters of the major arousal systems,
including glutamate, noradrenaline (NA), histamine and orexin.41–44 The effect of
these transmitters is through particular receptors upon the cholinergic cells, such as
the a1–adrenergic receptor (a1–AR), which evoke membrane depolarization and
excitation. Microinjections of these transmitters, including importantly NA or their
agonists into the basal forebrain cholinergic cell area increase gamma and theta EEG
activity while suppressing delta EEG activity and concomitantly stimulate waking
while suppressing sleep.40,45,46
The peptide, neurotensin, which was shown in vitro to selectively excite and
elicit rhythmic bursting in the cholinergic cells,47 was found when injected into the
basal forebrain to evoke theta together with gamma activity while suppressing delta
and SWS but enhancing PS in addition to waking.39 These results clearly demonstrate
the robust influence of cholinergic basalo-cortical neurons in stimulating theta and
gamma EEG activity and promoting waking and PS states (Figure 2.1).
GABAergic basal forebrain neurons are heterogeneous in their projections and
properties and thus may fulfill different roles in modulating EEG activity and
sleep–wake states. By fast inhibitory postsynaptic potentials (IPSPs) produced
through GABAA receptors, GABA can serve to pace activity in pyramidal or other
neurons in the cortex. It can also stimulate slow IPSPs through GABAB receptors.
It can thus serve to pace activity of fast or slow waves or to inhibit activity. Also
depending upon their prime target in the cortex,48,49 the GABAergic projection
neurons could facilitate pyramidal cell discharge by inhibiting particular GABAergic
interneurons or could inhibit pyramidal cell discharge by inhibiting other GABAergic
interneurons, given the interconnections of cortical interneuronal networks.
We provide evidence here that different groups of GABAergic neurons may
facilitate cortical activation as On cells, whereas others may attenuate cortical acti-

vation as Off cells. The Nb+/GAD+ Off (burst) cells recorded in the anesthetized
rat and Nb+/GAD+ Off (burst)-like cells recorded in the head-fixed rat would appear
to have the capacity to pace or modulate slow cortical activity or to inhibit cortical
fast activity (Figure 2.1). The Nb+/GAD+ Off (tonic) cells could also via projections
to cholinergic basal forebrain neurons or to other neurons in the posterior hypothal-
amus or brainstem attenuate cortical activation by inhibiting those activating cells
(Figure 2.1).
The SW- or SWS-promoting GABAergic basal forebrain neurons represent
particular subgroups of GABAergic cells that likely would be modulated in partic-
ular ways by transmitters of the arousal systems. Indeed, in vitro pharmacological
studies revealed a small group of non-cholinergic neurons that were inhibited by
NA.50 Previous in vivo studies had also found that SWS-active neurons in the
preoptic area and basal forebrain were inhibited by NA through an a2 -adrenergic
receptor (a2 –AR).51
In our in vivo recording studies in urethane-anesthetized rats, we found that
identified Nb+/GAD+ Off cells were immunostained for the a2 –AR.52 Together with
the results presented here for the Nb+/GAD+ Off and -like cells, these results suggest
that the GABAergic SWS-active neurons would be inhibited by NA during waking
to become disinhibited and active during SWS. Studies examining the effect of NA
microinjections into the basal forebrain (above) could be interpreted in light of these
findings, since NA suppressed delta and SWS during the day when the rat normally
sleeps the majority of the time, while also evoking gamma and theta with waking.45
Our results support the role in the promotion of slow wave activity and SWS of
particular GABAergic cell groups that bear a2 –AR and would thus be inhibited by
NA released from the locus coeruleus neurons of the brainstem arousal systems.
Through differential modulation by the transmitters of the brainstem arousal
systems, cholinergic cortical activation (On) and GABAergic SW-promoting (Off)
basal forebrain neurons, respectively, serve to stimulate cortical activation with wake
and PS and reciprocally promote cortical slow wave activity and SWS (Figure 2.1).
The basal forebrain thus has the capacity to regulate both the cortical activity and
sleep–wake state of the animal across the sleep–wake cycle.
SUMMARY
The basal forebrain is known to serve as the ventral extra-thalamic relay to the cerebral
cortex from the brainstem activating systems and thus to stimulate cortical activation.
Yet it is also known to have the capacity to promote slow wave cortical activity and
SWS. By using juxtacellular labeling with Nb in association with extracellular record-
ing of neurons in urethane-anesthetized and in head-fixed rats, we have identified
particular cell groups which discharge at their highest rate with cortical activation as
cholinergic or cholinergic-like. Others discharge at their highest rate with slow wave
activity and SWS as GABAergic or GABAergic-like. The cholinergic cells discharge
in rhythmic bursts with rhythmic theta activity and stimulate theta and gamma EEG
activity with waking and PS. Particular GABAergic cells discharge in arrhythmic
bursts with slow, irregular EEG activity and may modulate this cortical slow wave
activity while promoting SWS. Other GABAergic cells discharge in a slow, irregular

tonic manner with slow wave activity and may attenuate cortical activation by inhib-
iting other neurons of the activating systems, including the local cholinergic neurons
or the neurons in the posterior hypothalamus or brainstem.
These cortical activation and slow wave promoting cell groups are differentially
modulated by transmitters of the activating systems, including importantly NA,
which excites cholinergic cells and inhibits the slow wave promoting GABAergic
cells. Through the cholinergic cells, the basal forebrain thus serves as a relay for
the activating influences of the brainstem yet also directly modulates activity of the
cerebral cortex and promotes wake or PS states. Through particular GABAergic cell
groups, the basal forebrain may close the relay by inhibiting cholinergic and other
neurons of the activating systems and also directly modulate slow cortical activity
and promote SWS. The basal forebrain thus has the capacity to regulate cortical
activity and sleep–wake states across the sleep–wake cycle.
ACKNOWLEDGMENTS
This research was supported by grants from the Canadian Institute of Health
Research (13458) and National Institute of Mental Health (RO1 MH-60119-01A).
REFERENCES
1. Jones, B.E., Reticular formation. Cytoarchitecture, transmitters, and projections, in
The Rat Nervous System, 2nd ed., Paxinos, G., Academic Press Australia, New South
Wales, 1995, pp. 155–171.
2. Jones, B.E., Arousal systems, Front. Biosci., 8, S438–S451, 2003.
3. Steriade, M., Curro Dossi, R., Pare, D., and Oakson, G., Fast oscillations (20–40Hz)
in thalamocortical systems and their potentiation by mesopontine cholinergic nuclei
in the cat, Proc. Natl. Acad. Sci. USA, 88, 4396–4400, 1991.
4. Jones, B.E., Basic Mechanisms of Sleep-wake States, in Principles and Practice of
Sleep Medicine, 3rd ed., Kryger, M.H., Roth, T., and Dement, W.C. Eds., W.B.
Saunders, Philadelphia, 2000, pp. 134–154.
5. Vanderwolf, C.H. and Stewart, D.J., Thalamic control of neocortical activation: a
critical re-evaluation, Brain Res. Bull., 20, 529–538, 1988.
6. Rye, D.B., Wainer, B.H., Mesulam, M.-M., Mufson, E.J., and Saper, C.B., Cortical
projections arising from the basal forebrain: a study of cholinergic and noncholinergic
components employing combined retrograde tracing and immunohistochemical local-
ization of choline acetyltransferase, Neuroscience, 13, 627–643, 1984.
7. Jones, B.E., Activity, modulation and role of basal forebrain cholinergic neurons
innervating the cerebral cortex, Prog. Brain Res., 145, 157–169, 2004.
8. Metherate, R., Cox, C.L., and Ashe, J.H., Cellular bases of neocortical activation:
modulation of neural oscillations by the nucleus basalis and endogenous acetylcho-
line, J. Neurosci., 12, 4701–4711, 1992.
9. McCormick, D.A., Neurotransmitter actions in the thalamus and cerebral cortex and
their role in neuromodulation of thalamocortical activity, Prog. Neurobiol., 39,
337–388, 1992.
10. Longo, V.G., Behavioral and electroencephalographic effects of atropine and related
compounds, Pharmacol. Rev., 18, 965–996, 1966.

11. Stewart, D.J., MacFabe, D.F., and Vanderwolf, C.H., Cholinergic activation of the
electrocorticogram: Role of the substantia innominata and effects of atropine and
quinuclidinyl benzilate, Brain Res., 322, 219–232, 1984.
12. Detari, L. and Vanderwolf, C.H., Activity of identified cortically projecting and other
basal forebrain neurones during large slow waves and cortical activation, Brain Res.,
437, 1–8, 1987.
13. Szymusiak, R. and McGinty, D., Sleep-related neuronal discharge in the basal fore-
brain of cats, Brain Res., 370, 82–92, 1986.
14. Szymusiak, R. and McGinty, D., Sleep-waking discharge of basal forebrain projection
neurons in cats, Brain Res. Bull., 22, 423–430, 1989.
15. Starzl, T.E., Taylor, C.W., and Magoun, H.W., Ascending conduction in reticular
activating system, with special reference to the diencephalon, J. Neurophysiol., 14,
461–477, 1951.
16. Sterman, M.B. and Clemente, C.D., Forebrain inhibitory mechanisms: Sleep patterns
induced by basal forebrain stimulation in the behaving cat, Exp. Neurol., 6, 103–117,
1962.
17. Sterman, M.B. and Clemente, C.D., Forebrain inhibitory mechanisms: Cortical syn-
chronization induced by basal forebrain stimulation, Exp. Neurol., 6, 91–102, 1962.
18. Buzsaki, G., Bickford, R.G., Ponomareff, G., Thal, L.J., Mandel, R., and Gage, F.H.,
Nucleus basalis and thalamic control of neocortical activity in the freely moving rat,
J. Neurosci., 8, 4007–4026, 1988.
19. McGinty, D.J. and Sterman, M.B., Sleep suppression after basal forebrain lesions in
the cat, Science, 160, 1253–1255, 1968.
20. Szymusiak, R. and McGinty, D., Sleep suppression following kainic acid-induced
lesions of the basal forebrain, Exp. Neurol., 94, 598–614, 1986.
21. Gritti, I., Mainville, L., and Jones, B.E., Codistribution of GABA — with acetylcho-
line-synthesizing neurons in the basal forebrain of the rat, J. Comp. Neurol., 329,
438–457, 1993.
22. Gritti, I., Mainville, L., Mancia, M., and Jones, B.E., GABAergic and other noncho-
linergic basal forebrain neurons project together with cholinergic neurons to meso-
and iso-cortex in the rat, J. Comp. Neurol., 383, 163–177, 1997.
23. Gritti, I., Mainville, L., and Jones, B.E., Projections of GABAergic and cholinergic
basal forebrain and GABAergic preoptic-anterior hypothalamic neurons to the pos-
terior lateral hypothalamus of the rat, J. Comp. Neurol., 339, 251–268, 1994.
24. Zaborszky, L. and Duque, A., Local synaptic connections of basal forebrain neurons,
Behav. Brain Res., 115, 143–158, 2000.
25. Alonso, A., Khateb, A., Fort, P., Jones, B.E., and Muhlethaler, M., Differential
oscillatory properties of cholinergic and noncholinergic nucleus basalis neurons in
guinea pig brain slice, Eur. J. Neurosci., 8, 169–182, 1996.
26. Khateb, A., Muhlethaler, M., Alonso, A., Serafin, M., Mainville, L., and Jones, B.E.,
Cholinergic nucleus basalis neurons display the capacity for rhythmic bursting activity
mediated by low threshold calcium spikes, Neuroscience, 51, 489–494, 1992.
27. Pinault, D., A novel single-cell staining procedure performed in vivo under electro-
physiological control: morpho-functional features of juxtacellularly labeled thalamic
cells and other central neurons with biocytin or Neurobiotin, J. Neurosci. Methods,
65, 113–136, 1996.
28. Manns, I.D., Alonso, A., and Jones, B.E., Discharge properties of juxtacellularly
labeled and immunohistochemically identified cholinergic basal forebrain neurons
recorded in association with the electroencephalogram in anesthetized rats, J. Neu-
rosci., 20, 1505–1518, 2000.

29. Haas, K., Sin, W.C., Javaherian, A., Li, Z., and Cline, H.T., Single-cell electroporation
for gene transfer in vivo, Neuron, 29, 583–591, 2001.
30. Manns, I.D., Alonso, A., and Jones, B.E., Discharge profiles of juxtacellularly labeled
and immunohistochemically identified GABAergic basal forebrain neurons recorded
in association with the electroencephalogram in anesthetized rats, J. Neurosci., 20,
9252–9263, 2000.
31. Souliere, F., Urbain, N., Gervasoni, D., Schmitt, P., Guillemort, C., Fort, P., Renaud,
B., Luppi, P.H., and Chouvet, G., Single-unit and polygraphic recordings associated
with systemic or local pharmacology: a multi-purpose stereotaxic approach for the
awake, anesthetic-free, and head-restrained rat, J. Neurosci. Res., 61, 88–100, 2000.
32. Lee, M.G., Manns, I.D., Alonso, A., and Jones, B.E., Sleep-wake related discharge
of basal forebrain neurons recorded with micropipettes in head-fixed rats, J. Neuro-
physiol., in press.
33. Porter, J.T., Cauli, B., Tsuzuki, K., Lambolez, B., Rossier, J., and Audinat, E.,
Selective excitation of subtypes of neocortical interneurons by nicotinic receptors, J.
Neurosci., 19, 5228–5235, 1999.
34. Jimenez-Capdeville, M.E., Dykes, R.W., and Myasnikov, A.A., Differential control
of cortical activity by the basal forebrain in rats: a role for both cholinergic and
inhibitory influences, J. Comp. Neurol., 381, 53–67, 1997.
35. Xiang, Z., Huguenard, J.R., and Prince, D.A., Cholinergic switching within neocor-
tical inhibitory networks, Science, 5379, 985–988, 1998.
36. Soltesz, I. and Deschenes, M., Low- and high-frequency membrane potential oscil-
lations during theta activity in CA1 and CA3 pyramidal neurons of the rat hippoc-
ampus under ketamine-xylazine anesthesia, J. Neurophysiol., 70, 97–116, 1993.
37. McCormick, D.A. and Prince, D.A., Mechanisms of action of acetylcholine in the
guinea-pig cerebral cortex in vitro, J. Physiol., 375, 169–194, 1986.
38. Lisman, J.E., Bursts as a unit of neural information: making unreliable synapses
reliable, Trends Neurosci., 20, 38–43, 1997.
39. Cape, E.G., Manns, I.D., Alonso, A., Beaudet, A., and Jones, B.E., Neurotensin-
induced bursting of cholinergic basal forebrain neurons promotes gamma and theta
cortical activity together with waking and paradoxical sleep, J. Neurosci., 20,
8452–8461, 2000.
40. Cape, E.G., and Jones, B.E., Effects of glutamate agonist versus procaine microin-
jections into the basal forebrain cholinergic cell area upon gamma and theta EEG
activity and sleep-wake state, Eur. J. Neurosci., 12, 2166–2184, 2000.
41. Eggermann, E., Serafin, M., Bayer, L., Machard, D., Saint-Mleux, B., Jones, B.E.,
and Muhlethaler, M., Orexins/hypocretins excite basal forebrain cholinergic neurones,
Neuroscience, 108, 177–181, 2001.
42. Fort, P., Khateb, A., Pegna, A., Muhlethaler, M., and Jones, B.E., Noradrenergic
modulation of cholinergic nucleus basalis neurons demonstrated by in vitro pharma-
cological and immunohistochemical evidence in the guinea pig brain, Eur. J. Neuro-
sci., 7, 1502–1511, 1995.
43. Khateb, A., Fort, P., Pegna, A., Jones, B.E., and Muhlethaler, M., Cholinergic nucleus
basalis neurons are excited by histamine in vitro, Neuroscience, 69, 495–506, 1995.
44. Khateb, A., Fort, P., Serafin, M., Jones, B.E., and Muhlethaler, M., Rhythmical bursts
induced by NMDA in cholinergic nucleus basalis neurones in vitro, J. Physiol.
(Lond.), 487.3, 623–638, 1995.

45. Cape, E.G., and Jones, B.E., Differential modulation of high frequency gamma
electroencephalogram activity and sleep-wake state by noradrenaline and serotonin
microinjections into the region of cholinergic basalis neurons, J. Neurosci., 18,
2653–2666, 1998.
46. Espana, R.A., Baldo, B.A., Kelley, A.E., and Berridge, C.W., Wake-promoting and
sleep-suppressing actions of hypocretin (orexin): basal forebrain sites of action,
Neuroscience, 106, 699–715, 2001.
47. Alonso, A., Faure, M.-P., and Beaudet, A., Neurotensin promotes oscillatory bursting
behavior and is internalized in basal forebrain cholinergic neurons, J. Neurosci., 14,
5778–5792, 1994.
48. Freund, T.F. and Meskenaite, V., Gamma-aminobutyric acid-containing basal fore-
brain neurons innervate inhibitory interneurons in the neocortex, Proc. Natl. Acad.
Sci. USA, 89, 738–742, 1992.
49. Freund, T.F. and Gulyas, A.I., GABAergic interneurons containing calbindin D28K
or somatostatin are major targets of GABAergic basal forebrain afferents in the rat
neocortex, J. Comp. Neurol., 314, 187–199, 1991.
50. Fort, P., Khateb, A., Serafin, M., Muhlethaler, M., and Jones, B.E., Pharmacological
characterization and differentiation of non-cholinergic nucleus basalis neurons in
vitro, NeuroReport, 9, 1–5, 1998.
51. Osaka, T. and Matsumura, H., Noradrenaline inhibits preoptic sleep-active neurons
through a2-receptors in the rat, Neurosci. Res., 21, 323–330, 1995.
52. Manns, I.D., Lee, M.G., Modirrousta, M., Hou, Y.P., and Jones, B.E., Alpha 2 adren-
ergic receptors on GABAergic, putative sleep-promoting basal forebrain neurons,
Eur. J. Neurosci., 18, 723–727, 2003.
53. Jones, B.E., Neurotransmitter systems regulating sleep-wake states, in Biological
Psychiatry, D’Haenen, D., den Boer, J.A., and Willner, P., Eds., John Wiley & Sons,
New York, 2002, pp. 1215–1228.

1519_book.fm Page 43 Tuesday, August 24, 2004 3:05 PM
3 In Vitro Identification
of the Presumed
Sleep-Promoting
Neurons of the
Ventrolateral Preoptic
Nucleus (VLPO)
Patrice Fort, Pierre-Hervé Luppi,
and Thierry Gallopin
CONTENTS
The Discovery of the VLPO as a Potent Sleep Center of the Brain

Serotonin Modulation Reveals Two Types of Presumed Sleep
Promoting Neurons in the VLPO
Hypnogenic Substances Differentially Modulate Type-1 and
Type-2 Neurons: An Emphasis on the PGD2-Adenosine Axis
A Functional Model of the Neuronal Network Responsible
for Sleep Promotion
Summary
Acknowledgments
References
THE DISCOVERY OF THE VLPO AS A POTENT SLEEP

CENTER OF THE BRAIN
Between World War I and World War II, von Economo proposed that a basal forebrain
area would be the place for a brain center involved in sleep regulation. He reported
that comatose patients struck down with an encephalitis lethargica had prominent
parenchyma injury at the level of the preoptic area (POA) near the optic tract.1 These
seminal clinical studies, indicating that an intact rostral hypothalamus is critical for
0-8493-1519-0/05/$0.00+$1.50

the production of normal sleep, definitely represents a founding step for the research
aimed at discovering the neurobiological mechanisms regulating behavioral states
(namely wakefulness, slow waves sleep, and paradoxical sleep). During the subse-
quent period until early 1990, the experimental process for basic research, using
standard lesion, neuronal unit recording, neuropharmacological, and neurochemical
approaches in animals, led to the statement of a few fundamental concepts, which
can be summarized as follows:
1. In lines with early predictions by von Economo, POA, more especially

its lateral part, is the unique brain structure that fulfilled necessary and
sufficient criteria for an hypnogenic center containing neurons that directly
promote sleep.2–6
2. This simplicity highly contrasts with the redundant network responsible
for arousal, involving numerous brain areas and neurotransmitter systems
such as acetylcholine-, noradrenaline-, serotonin-, histamine-, and
recently discovered orexin-containing neurons, with widespread projec-
tions from rhombencephalon to cerebral mantle. Collectively these com-
ponents, with an activity specific to wakefulness, form the so-called
ascending reticular activating system (ARAS) that regulates cortical acti-
vation during waking.7–10
3. Soon as drowsiness begins, the hypnogenic center would put out of func-
tion the ARAS system through sustained inhibitions.
4. The sleep pressure as well as drowsiness would be owed to the conjunction
of homeostatic and circadian processes that are able to directly modulate
the sleep center.11,12
Despite these crude consensual concepts and the real progress of our knowledge
since von Economo’s proposal, we have to admit that basic neurobiological mech-
anisms involved in sleep promotion and the harmonious succession of behavioral
states remain largely underestimated. Indeed, LPOA is a vast forebrain region that
contains multiple contingents of intermingled and loosely arranged neurons, gov-
erning vital functions. This cytoarchitectonic configuration and the lack of a precise
plotting of the sleep-promoting neurons have hindered the decoding of cellular,
synaptic, or molecular mechanisms used by the sleep center to play its functional
role. However, a decisive stage was exceeded in 1996 by the individuation of the
ventrolateral preoptic nucleus (VLPO), a small neuronal core (radius 300 mm)
located in the most ventral part of the LPOA. This was made possible by means of
a functional neuroimaging paradigm at the cellular level, using the expression of
the early gene c-Fos as a marker of neuronal activity in rats having slept for a long
period before sacrifice.13 This hypersomnia, also coined sleep-rebound, is the typical
behavioral response following sleep deprivation in rats.
While neurons specifically activated during sleep and immunostained for c-Fos
(c-Fos+) were diffusely distributed in LPOA, they were more densely packed within
the VLPO. Furthermore, the density of c-Fos+ neurons correlated closely with the
sleep quantities during the last two hours preceding sacrifice. This labeling pattern
of sleep-active neurons would be related rather to the production of sleep itself than

to a homeostatic regulation induced by its deprivation. Although drowsiness mark-

edly increased in deprived rats, little staining was observed in rats sacrificed before
the sleep-rebound.13,14 By the same functional approach, it has been demonstrated
that VLPO and suprachiasmatic nucleus (SCN) have synchronized activity.15 Fur-
ther, they are interconnected and received inputs from retina ganglionic cells. Sim-
ilarly the dorso-median hypothalamic nucleus, a SCN relay, projects strongly to the
VLPO. Considered together, these anatomical data suggests that circadian- and
photic-linked information may be conveyed to modulate the VLPO activity across
the nyctemeral period.16–23
Electrophysiological experiments in freely-moving rats have shown that neurons
that doubled their firing rate at sleep onset are more frequently recorded in VLPO
than in other LPOA parts.24 Furthermore, their recruitment and firing activation are
positively correlated to both sleep depth and duration. Of particular functional
interest, this sleep-specific activity of VLPO neurons (i.e., sleep-on neuron) is inverse
to that of wake-active neurons.25–28 Functionally the bilateral neurotoxic destruction
of VLPO neurons (more than 70%) is followed by a profound and long-lasting
insomnia with a reduction of 56% of sleep quantities in rats.29 In line with these
data, we further demonstrated that iontophoretic application of carbachol, a cholin-
ergic agonist, targeted to the VLPO suppressed sleep in anesthetic-free head-restraint
rats.30,31 These physiological data support the necessity of VLPO for producing
normal sleep. Retrograde and anterograde tract-tracing studies indicate that VLPO
neurons are reciprocally connected with cerebral areas containing wake-active neu-
rons such as the histaminergic tuberomammillary nucleus (TMN), serotonergic mid-
brain raphe nuclei (RN), noradrenergic locus coeruleus (LC), cholinergic pontine
(LDT/PPT) and magnocellular preoptic (NB) nuclei, as well as orexinergic perifor-
nical area of the lateral hypothalamus.13,14,32–37 More than 90% of c-Fos+ sleep-active
neurons in VLPO express galanin mRNA while 80% of neurons projecting to the
TMN contain both galanin and GAD, the GABA-synthesizing enzyme, suggesting
that projections to the waking systems are inhibitory in nature.13,14,36
Taken together these data indicate that the VLPO plays a key role in coordinating
the inhibition of arousal systems to promote sleep and thus occupies a privileged
place within the complex neuronal network involved in behavioral states. Its indi-
viduation had opened new fields for investigation of the underlying regulatory
mechanisms of sleep. For us the fact that VLPO neurons are:
• Specifically active during sleep

• Endowed with reciprocal inhibitory connections with the wake-promoting
areas
• Densely packed in a small-sized nucleus
offers a unique opportunity and evident methodological advantage to study at cel-

lular, synaptic and molecular levels the neurons responsible for sleep. A special
effort to characterize neurotransmitters and pathways that control VLPO sleep-active
neurons would thus contribute to understand the mechanisms that manage their
excitability across the sleep-waking cycle and should provide key insight into the
regulation of behavioral states.

To strengthen our proposal, we have undertaken during the last years electro-
physiological recordings of VLPO neurons in rat brain slices. This in vitro experi-
mental approach is proved to be suitable for exploring electrophysiological, phar-
macological, and chemo-morphological properties of neurons and thus for drawing
up the so-called functional ID card of the sleep-active neurons. One of our primary
objectives was to determine whether neurons inhibited by neurotransmitters released
from wake-promoting areas could be frequently recorded in VLPO. In close collab-
oration with our colleagues of the Geneva University (Switzerland), we thus iden-
tified successfully a homogeneous neuronal group with a specific set of intrinsic
membrane properties and a clear-cut chemo-morphology that are inhibited by the
major neurotransmitters of waking. Their high proportion (80% of the recorded
neurons), matching that of cells active during sleep in VLPO, and their pharmaco-
logical profile represent convincing arguments about their presumed status as sleep-
promoting neurons (PSP). We showed that PSP neurons are GABAergic, multipolar,
triangular shaped, and endowed with a potent low threshold calcium potential. These
neurons are inhibited by noradrenaline (NA).38 Recent works determined that this
inhibitory effect is mediated by post-synaptic alpha-2 adrenoceptors.39,40
We further revealed that NA-inhibited neurons are also inhibited by acetylcholine
(ACH). In contrast histamine (HA) and orexin (Ox) did not modulate PSP neurons,
although an inhibitory influence was expected.38,41 However, it should be noted that
TMN neurons contain both histamine and GABA42 and are thus in position, as
noradrenergic and cholinergic drives, to inhibit PSP neurons during waking. Con-
sidering their unique profile of neuromodulation (since the remaining recorded cells
are excited by NA, ACH, HA, and Ox), the overall inhibition of the PSP neurons
by neurotransmitters of waking is in agreement with their inactivity during waking.24
We previously suggested that the reciprocal inhibitory interaction of PSP neurons
with the multiple waking systems to which they project is a key factor for promoting
sleep by coordinating their inhibition at sleep onset.38 More recently a consensual
model has been proposed suggesting that this reciprocity of projections is analogous
to a flip-flop switch electrical circuit.6
When VLPO neurons start to fire at sleep onset and fire rapidly during sleep,
they would inhibit the waking-promoting neurons allowing for their own disinhibi-
tion and reinforced firing. During arousal, waking-promoting neurons fire at a high
rate, thus inhibiting VLPO neurons and resulting in the disinhibition of their own
firing. Either sleep or waking is self-reinforcing when its component neurons are
sufficiently active. The reciprocal inhibitory interaction of these systems provides a
mechanism for the maintenance of one of the two stable configurations. Accordingly,
disruption of wake- and sleep-promoting pathways would result in behavioral insta-
bility due to a destabilization of the reciprocal inhibitory interactions. This is likely
the case in murine models of narcolepsy, a human sleep pathology, with functional
failure of the orexin system concomitant to pronounced vigilance disturbances and
sudden transitions in behavioral states.43,44 An increasing number of data agreed that
orexin-containing neurons would play a major role for the maintenance of arousal.
The widespread excitatory projection to waking-promoting neurons provides to this
neuronal system an ideal position to orchestrate their respective activity.45 Turn on
during waking, orexin-containing neurons would strengthen the activity of the wake-

promoting neurons, which in turn, via their inhibitory projections to PSP neurons,
would prevent sleep onset and thus stabilize waking.6,45–47
Functional properties of the flip-flop switch model may easily support the pro-
duction by a simple neuronal network of stable states of wakefulness and sleep and
an important resistance to switching by limiting inappropriate changes when inputs
to VLPO or wake-promoting areas fluctuate. In great contrast, this model does not
take account for the necessary instability or un-balanced relationship between wake-
and sleep-promoting neurons that should occur for rapid transitions between sleep
and waking (drowsiness or awaking), switching events that are frequently encoun-
tered across the sleep-waking cycle (75% of all transition states in rats). In this
context mechanisms responsible for the firing increase of sleep-on neurons just
before or at sleep onset remain unknown. They would be the result of a disinhibition
linked to a decreased activity of wake-promoting neurons, thus releasing PSP neu-
rons from potent inhibitions during waking or an increase of a sleep-dependent
excitatory drive, thus inducing the inhibition of the wake-promoting neurons and
reinforcing sleep.
It is tempting to hypothesize that such excitatory drive would be related to
thermoregulation48 or homeostatic process, involving hypnogenic factors that
directly excite PSP neurons. In an attempt to test this second hypothesis, we thus
pursued the pharmacological and molecular characterization of the PSP neurons in
VLPO slices with a special interest for their interactions with presumed sleep
neuromodulators. Numerous substances contributing to the sleep homeostasis have
been described.49–51 The following parts are focused on two well-recognized sleep
factors, namely serotonin and adenosine.
SEROTONIN MODULATION REVEALS TWO TYPES

OF PRESUMED SLEEP PROMOTING NEURONS IN
THE VLPO
In the subsequent experiments we performed recordings in loose cell-attach config-
uration to study the effect of multiple drug applications on a single PSP cell. In
contrast to intracellular and patch-clamp electrophysiological techniques, this mode
allows stable recordings of healthy neurons for long periods of time necessary to
complete pharmacological experiments. Infrared differential interference videomi-
croscopy was used to locate VLPO neurons according to their typical size and
triangular shape. Cell-attached recordings were made from the soma with patch
micropipettes filled with ACSF and attached to an electric microdrive to place it
under visual control in contact with the soma of the cell chosen (Figure 3.1). In this
mode cells were classified as PSP neurons when application of NA and ACH induced
a decrease of their firing rate.38 We thus showed that 47% of the PSP neurons are
also inhibited by 5-HT (Type-1 cells) while 53% are excited (Type-2 cells).52 Our
data indicate that the modulation of PSP neurons by 5-HT is complex and that two
types of PSP neurons are present in VLPO according to their pharmacological profile
(Figure 3.1).
This unexpected segregation led us to resume the characterization of both types
of neurons to arrest their respective neuronal and functional specificities regarding

A1 B1
A2 B2
NA 100 µM NA 100 µM
18 18
Hz Hz
0 0
60 s 60 s
A3 B3
5-HT 100 µM 5-HT 100 µM
18 18
Hz Hz
0 0
60 s 60 s
FIGURE 3.1 Pharmacological identification of the Type-1 and Type-2 neurons of the VLPO.
(A1–B1) Microphotographies showing the typical morphology (triangular shaped) of the
Type-1 and Type-2 neurons respectively, as observed in slices by means of the infrared
differential interference videomicroscopy. Bars: 20mm. (A2–A3) Firing frequency versus time
diagrams illustrating a Type-1 neuron extracellularly recorded in loose-attached configuration
since its firing rate is strongly reduced after the bath application of NA (A2) and 5-HT (A3).
(B2–B3) Firing frequency versus time diagrams illustrating a Type-2 neuron since it is
inhibited by NA (B2) but reversibly excited following 5-HT application (B3).
sleep production. First assessed as Type-1 or Type-2 cells according to their

responses to NA and 5-HT, PSP neurons were then systematically re-recorded in
the whole-cell patch-clamp configuration. We found that the standard electrophysi-
ological parameters such as the resting potential, membrane resistance, amplitude
and duration of the sodium-dependant action potential, as well as duration of the
post-hyperpolarization (AHP) are not significantly different in both types of PSP
neurons (Figure 3.2). The only difference was the higher AHP amplitude in Type-
2 versus Type-1 neurons. In agreement with our previous data,38 an electrophysio-
logical landmark of both types of cells is the presence of a powerful low threshold
calcium spike, due to an IT current.
We further demonstrated that both types of PSP neurons present abnormal
membrane rectifications underlying the activation of sodium persisting (INaP) and
potassium time-dependant (Ih) currents (Figure 3.2). For the morphological identi-
fication, PSP neurons were filled with biocytin during the recording sessions. These
staining experiments demonstrate that both types of neurons were recorded in the

A B
TTX 10ΠM
* *
2
1
20 mV
* 20 mV
0,5 nA 0,5 nA
200 ms 100 ms
C D
I (nA)
-1 -0,8 -0,6 -0,4 -0,2 0
0
-20
20 mV
-40
-60
-80
-100
0,5 nA -120
-140
500 ms V (mV)
FIGURE 3.2 Intrinsic membrane properties of both types of presumed sleep-promoting neu-
rons recorded in the VLPO. (A) Response of a neuron submitted to a depolarizing pulse at
rest or held hyperpolarized by continuous negative current injection. Notice the presence of
a potent low threshold spike (LTS, *), probably calcium-dependent because it remained
following TTX treatment and sodium spikes disappeared. (B) At the resting membrane
potential, the application of short-lasting depolarizing current pulses revealed the presence
of a plateau due to the activation of a persistent sodium current (trace 1), which in some cases
is able to reach the spike threshold (trace 2). (C) Response of a neuron following injection
of hyperpolarizing current pulses with progressive amplitude increase showing the presence
of a time-dependent rectification of the membrane potential (sag indicated by filled versus
open circle). (D) Current-voltage curves obtained for the cell illustrated in D (I–V). Note the
diminution of the membrane resistance during the current pulses revealing the presence of
an Ih current in PSP neurons. Arrowheads indicated the level of the resting membrane potential.
VLPO proper, avoiding the extended VLPO36 or adjacent basal forebrain structures
(lateral or median preoptic areas, nucleus basalis, supraoptic nucleus). Both types
of cells were morphologically undistinguishable since they were typically medium-
sized and mainly triangular shaped with three primary dendrites (Figure 3.3).
Biocytin-labeled dendrites occasionally extended ventrally over long distances,
where they exit the parenchyma and/or travel along the brain surface.20
A required feature to ensure that both types are PSP neurons was to determine
their respective neurochemical nature, by evaluating first the well-known cellular
markers of neurons that are sleep-active, namely galanin and GABA. By double-
fluorescent staining, we demonstrated that biocytin-filled Type-1 and Type-2 neu-
rons both contain galanin (Figure 3.3). Furthermore, we applied the single-cell RT-
PCR technique53 coupled to patch-clamp recordings of neurons beforehand assessed
by their responses to NA and 5-HT. We thus determined that Type-1 and Type-2
neurons express mRNAs coding for GAD 65 and GAD 67 (Figure 3.4). It is thus
likely that Type-1 and Type-2 neurons match up the VLPO neurons that are GABA
and galanin in nature, selectively activated during sleep and that project directly to
the waking systems.14,36

Type 1
A1 A2
Type 2
B1 B2
FIGURE 3.3 Both types of presumed sleep promoting neurons contain galanine as a neu-
rotransmitter. (A1–B1) Photomicrographies of one Type-1 (C1) and one Type-2 (D1) neuron,
both filled with neurobiotine revealed using a streptavidine-Cy3 fluorochrome (yellow).
(C2–D2) The same slices were submitted simultaneously to galanin immunodetection by
using a secondary antibody labeled with Cy2 fluorochrome (green). The superposition evi-
denced that both types of PSP neurons are double-labeled, indicating that they match the
galanin VLPO neurons that are activated during sleep. Bars: 20mM.
A B
7
a
c
a
G A tr
tr
5
D6
T5
T2
T1
D6
D6
D6
-in
-in
5-H
5-H
5-H
GA
GA
GA
SS
SS
603 bp 603 bp
310 bp 310 bp
FIGURE 3.4 Molecular characterization of the Type-1 (A) and Type-2 (B) neurons by the
multiplex single-cell RT-PCR technique. Agarose gel of PCR products showing expression
of GAD65, GAD67, and 5-HT1a receptors in one Type-1 neuron (A) and GAD65, GAD67,
and serotonergic 5-HT2c and 5-HT5a in one Type-2 neuron (B4). F is a marker for relative
molecular mass. Genomic DNA amplification, which could occur if the nucleus was harvested
during the patch-clamp recordings, was systematically assessed using a somatostatin gene
intron (ss-intr) as a genomic control. These data indicate that both types of neurons are likely
GABAergic in nature but expressed different sets of serotonergic postsynaptic receptors.

Type-2 Inhibition
Type-1 No effect
Excitation
A1 B1 C1 100% 100%
100
% of neurons
80
PGD2 1µM PGD2 1µM
60
10 10 40
Hz Hz 20
6 6 0% 0% 0% 0%
0
60 s 60 s Type-1 Type-2
A2 B2 C2 100 93%
AD 100µM AD 100µM 83%
% of neurons
12 12 80
60
Hz Hz 40
20
17%
0 0 0% 7%
0
0%
A3 B3 C3 100
87% 86%
% of neurons
80
CGS 1µM 12 16
CGS 1µM 60
Hz Hz 40
20 13% 14%
4 4 0
0% 0%
A4 B4 C4 100
78%
% of neurons
80
77%
Galanin 1µM 12 Galanin 1µM 12 60
Hz Hz 40
22% 23%
20
6 6 0% 0%
0
FIGURE 3.5 Pharmacological characterization of VLPO Type-1 and Type-2 neurons

recorded in loose cell-attach configuration. Following systematic applications of both NA and
5-HT to assess the subtype of the recorded cell, the effect on its spontaneous firing rate was
examined following by PGD2 (A1, B1, C1), ADO (B1, C1, D1), CGS, an A2A agonist (C1,
C2, C3), and galanin (D1, D2, D3) applications. Note that although ADO clearly inhibited
the Type-2 neurons, CGS application was followed by a reversible increase of the firing rate.
The experiments clearly demonstrated that Type-1 and Type-2 neurons present a specific
pharmacological profile and thus expressed different sets of postsynaptic receptors for these
different sleep factors.
It has been shown that galanin inhibited noradrenergic LC neurons,54 seroton-

ergic midbrain raphe neurons55 and histaminergic TMN neurons.56 Besides, galanin
is also co-localized with 5-HT and NA.57,58 It seems that galanin could participate
in combination with amines to the inhibition during waking of PSP neurons. Partly
supporting this hypothesis, we observed that galanin inhibits the firing activity of
Type-2 neurons but had no effect on Type-1 neurons (Figure 3.5). Although three
types of post-synaptic receptors have been cloned (GalR1, GalR2 and GalR3), only
GalR2 are specifically expressed in the VLPO, while GalR1 and GalR2 are widely
distributed in the POA.59
Taken together our data indicate that the neuromodulation by 5-HT would be a
relevant criterion for the pharmacological segregation of the PSP neurons in VLPO.
Numerous 5-HT receptors have already been identified and dispatched in seven
major classes according to their functional properties.26 The extreme diversity of the
pharmacological effects observed with 5-HT highly supports its duality of modula-
tion of PSP neurons. We thus attempted to identify the subtypes of 5-HT receptors

responsible for the excitatory and inhibitory effects. For this purpose we used the
single-cell multiplex RT-PCR technique for an initial screening of 5-HT receptor
mRNAs expressed in PSP neurons. Although preliminary, this study evidences that
Type-1 and Type-2 neurons express different sets of 5-HT receptors. Type-2 neurons
express mRNA for 5-HT2c, 5-HT5a and probably 5-HT4 receptors. Previous studies
have shown that 5-HT2c receptors mediate excitatory effects in numerous brain
regions60 and would be involved in sleep regulation. Indeed, intraperitoneal admin-
istration of 5-HT2a/2c or 5-HT2c (m-CPP) agonists decreased sleep quantities, in
contrast to 5-HT2a/2c antagonist (ritanserine).61–63
Furthermore, ritanserine reversed the effect produced by 5-HT2a/2c agonist but
not that induced by 5-HT2c agonist, rather supporting the crucial role of 5-HT2a in
sleep regulation.61,62,64,65 With regard to the Type-1 neurons, they express mRNAs
coding for 5-HT6, 5-HT5b, 5-HT1d and probably 5-HT1a receptors (Figure 3.4). The
presence of 5-HT1a receptors was assessed because 8-OH-DPAT, a 5-HT1a agonist,
mimicked the inhibition induced by 5-HT. It has been already demonstrated that
activation of 5-HT1a receptors hyperpolarized neurons in many brain regions. Their
involvement in the 5-HT-induced inhibition of Type-1 neurons is consistent with
data reporting that systemic administration of 8-OH-DPAT is followed by a decrease
of sleep quantities in rats.66–68
Although our present data shed new light on the functional links between VLPO
neurons and sleep, they however raise new questions regarding the specific role of
the two populations of PSP neurons, differently modulated by 5-HT, and their
integration within the flip-flop switch model.6,38 The firing rate of 5-HT neurons is
maximal during waking and greatly decreases during sleep to become silent during
paradoxical sleep.27,69 In addition the 5-HT release is lower during sleep versus
waking in target areas of the midbrain raphe nuclei.70,71 This suggests that 5-HT
would act preferentially as a waking neurotransmitter. In lines with this statement,
the inhibition of Type-1 neurons highly supports the participation of 5-HT during
waking to the inhibition of the VLPO sleep-active neurons. On the other hand, our
unexpected findings that 5-HT may excite a subset of PSP neurons revive the Jouvet’s
hypothesis, suggesting a major contribution of 5-HT to sleep production.72–74 This
proposal was initially supported by the profound insomnia induced by lesion of
midbrain raphe nuclei.72,75
Similar behavioral effects have been obtained following the down-regulation of
5-HT synthesis with systemic para-chloro-phenylalanine treatment (PCPA) in cats
and rats.76 In PCPA-treated insomniac animals, sleep can be restored by intraven-
tricular or systemic administration of 5-HTP, the precursor of 5-HT.77–79 A potential
target for sleep induction is the ventral LPOA that a posteriori includes VLPO,
because 5-HTP microinjection in this forebrain area restored, with a delayed latency
(1 hour), natural sleep in PCPA-treated insomniac cats, while injections in neigh-
boring zones were unable to reverse such insomnia.80 These data suggest that the 5-
HT release in VLPO during waking could not be responsible for sleep onset, through
direct excitation of PSP neurons, but would rather prepare the local physiological
conditions necessary for sleep to occur. This diachronic action as stated by Jouvet’s
team74,81,82 would facilitate the 5-HT-dependant synthesis of hypnogenic factors (see
in the following) during waking that could mediate the activation of PSP neurons.

Considering this hypothesis, Type-2 neurons would stay inactive during waking due
to combined NA and ACH inhibitions, despite the 5-HT subthreshold excitatory
drive that would participate to intracellular mechanisms downstream post-synaptic
serotonergic receptors and prepare their firing activation for sleep onset.
Both functional neuroanatomy and electrophysiology in awake animals indicate
that the activation of sleep-on neurons in VLPO is related rather to the sleep
occurrence than to the sleep need associated to long-lasting awaking.13,24 However,
a subset of VLPO neurons increased their firing rate and anticipated for a few seconds
the sleep onset (drowsiness).24 At the same moment, 5-HT release increased within
the LPOA due to either a sudden firing acceleration of afferent raphe neurons or
presynaptic mechanisms monitoring local 5-HT release.83 The obvious temporal
correlation between these two events led us to suggest that Type-2 neurons could
match this subset of VLPO sleep-on neurons and would thus be directly involved,
rather than Type-1 neurons, in 5-HT mechanisms of sleep induction.52 In conclusion,
our present data suggest that 5-HT, released during waking in the VLPO, may
participate concomitantly to seemingly opposite mechanisms, by strengthening
arousal through the inhibition of Type-1 neurons and preparing sleep via the sub-
threshold excitation of Type-2 neurons.
HYPNOGENIC SUBSTANCES DIFFERENTIALLY

MODULATE TYPE-1 AND TYPE-2 NEURONS: AN
EMPHASIS ON THE PGD2-ADENOSINE AXIS
Among processes that are also inclined to modulate the activity of PSP neurons,
homeostatic mechanisms had long been thought to play a crucial role in sleep
triggering.11,12,49,50 The homeostasis is supposed to explain the sleep pressure and
sleep-rebound following its deprivation, due to the synthesis and accumulation of
natural factors during prolonged awakening. Their increasing concentration during
waking in wake and sleep-promoting areas up to a critical threshold would contribute
the preparation and promotion of sleep. Prostaglandin D2 (PGD2) and adenosine
(ADO) have long been functionally implicated in sleep regulation, although their
respective targets and mechanisms of action remain largely unknown.84 In this
context we tried to determine whether these sleep factors would directly modulate
PSP neurons in a way that supported their role in sleep.
We observed that the spontaneous firing of Type-1 and Type-2 neurons was not
modified by PGD2 (Figure 3.5), suggesting that its behavioral effects are not due
to direct activation of PSP neurons. However previous in vivo studies showed that
PGD2 when injected LPOA85 promoted sleep and excited one third of sleep-on
neurons in rats.86,87 Other studies rather support an indirect mechanism by blocking
the brainstem noradrenergic inputs that strongly inhibit sleep-on neurons during
waking.88,89 To test this hypothesis, it turns out necessary to determine in vitro
whether PGD2 would modulate noradrenergic inhibitory post-synaptic potentials in
Type-1 and Type-2 cells. Another possibility is that PGD2 would promote sleep by
inducing meningeal targets to release paracrine-signaling molecules.84 PGD2 con-
centration in the CSF increases during waking and sleep deprivation.90,91 Furthermore

PGD2 infusion into the subarachnoid space ventral to the basal forebrain, namely
the PGD2-sensitive zone, induced sleep92 and consequently c-Fos expression in
VLPO neurons.93
Because PSP neurons are unresponsive to PGD2 and likely do not express PGD2
receptors, their activation during sleep would require a secondary messenger to
transmit the PGD2-mediated signal into the brain parenchyma. Among these mes-
sengers ADO is a concrete candidate because:
• Its local concentration increased following PGD2 infusion into the sub-
arachnoid space.93,94
• Its infusion or that of specific A2a receptors (A2aR) agonists into the same
space mimicked the PGD2 effects.95–98
• Sleep quantities are greatly reduced in mice deleted for functional A2aR.99
In the following, we thus tried to determine whether ADO modulates Type-1 and
Type-2 neurons.
We observed that ADO inhibited the firing of both types of PSP neurons (Figure
3.5). These effects involved A1R since a specific agonist, CPA, reproduces the ADO
effect. It is not easy to reconcile the inhibition by ADO of PSP neurons with its
hypnogenic properties. However ADO inhibits ubiquitously the neuronal activity,
via A1R, by combining post-synaptic hyperpolarization and pre-synaptic inhibition
of neurotransmitter release.100–102 ADO acts also as a homeostatic regulator to slow
down cell metabolism and is usually considered as a witness of neuronal energy
use.103,104 Indeed, ADO cerebral concentrations are generally low, but considerably
increase in conditions of ischemia or massive energy depletion.105 It is then possible
that, in our experimental conditions, ADO would be interpreted by recorded PSP
neurons as a signal of energy deficiency, leading to homeostatic protection by their
own firing inhibition without relevance with sleep regulation. In wake-promoting
areas as cholinergic forebrain and pontine nuclei, ADO concentration is higher
during waking versus sleep and increases during prolonged awakening.106–108 Thus
ADO via A1R would facilitate sleep by reducing the cellular metabolism of waking-
promoting neurons according to homeostatic mechanisms.109
Specific A2aR agonists and antagonists were used to determine their potential
role in the modulation of PSP neurons. We demonstrated that Type-1 neurons express
exclusively A1R since the antagonist, DPCPX, reversed totally the ADO-induced
inhibition. In contrasts, DPCPX treatment revealed a reversible firing increase in
Type-2 neurons, due to post-synaptic A2aR. In fact, this excitatory effect is similarly
observed following application of CGS 21680, a specific agonist (Figure 3.5) and
reversed in presence of a specific antagonist, ZM 241385. Excitatory post-synaptic
effects mediated by A2aR had already been described in several brain regions110–113
but never in LPOA or VLPO. Our data clearly indicate that Type-2 neurons:
• Are under a permanent inhibitory control by ADO

• Express A2aR in addition to A1R
• Are able to be turned on by ADO

This the first tangible experimental demonstration supporting the consensual

hypothesis that ADO would promote sleep through direct excitation of VLPO sleep-
active, likely Type-2, neurons expressing A2aR.6,51,84,114 Previous microdialysis exper-
iments failed to evidence an increase of ADO concentration in LPOA, in contrast
to wake-promoting areas, during prolonged awakening.115 However this technique
is not suitable to measure faint variations of ADO concentration in LPOA, where
PSP neurons are loosely arranged. To date the kinetics of ADO release in VLPO
across the sleep-waking cycle remain unknown.
We already reported that PGD2 infusion in the subarachnoid space is followed
simultaneously by sleep increase, c-Fos expression in VLPO neurons and local ADO
release.94 Similar effects are obtained following CGS 21680 infusion in the PGD2-
sensitive zone.98 It is tempting to hypothesize that ADO in the CSF may rapidly
diffuse until nearby A2aR-expressing PSP neurons in VLPO. It is therefore interesting
to recall that PSP neurons have primary dendrites that course and travel along the
brain floor to exit in some cases the cerebral parenchyma into the subarachnoid space.
Although more experiments are needed to work out detailed mechanisms, our
data suggest that Type-2 neurons would function as a neuronal probe to detect local
increases of ADO concentration. During waking ADO would be gradually accumu-
lated in CSF, diffusing into the neighboring parenchyma. When the critical ADO
concentration is attained in VLPO, Type-2 neurons would be turned on by direct
excitation or presynaptic mechanisms of local neurotransmitter release. As 5-HT,
ADO may have seemingly opponent actions in VLPO, both promoting waking, via
inhibition of Type-1 neurons, and preparing sleep via excitation of Type-2 neurons.
A FUNCTIONAL MODEL OF THE NEURONAL

NETWORK RESPONSIBLE FOR SLEEP PROMOTION
Our initial in vitro work led to the first identification of PSP neurons within VLPO
with characteristics matching with their potential role in sleep promotion. It has
been subsequently suggested that alternance of behavioral states is settled to recip-
rocal inhibitory interactions between PSP and waking-promoting neurons. The PSP
neurons would be strongly inhibited during arousal but totally released from this
inhibition and thus active during sleep. The recently acquired data, pursuing this
successful experimental process to complete the functional ID card of PSP neurons,
establish firmly that two subtypes of PSP neurons can be segregated, based on their
differential pharmacological profile regarding two major sleep neuromodulators.
However major unknowns in the matter remain to be elucidated for an accurate
understanding of the events leading to sleep. The most obvious concern is the
respective functional contribution of Type-1 and Type-2 neurons and their anatomical
integration within the flip-flop switching model. It has been previously shown that
VLPO galanin-containing neurons projecting to the TMN or LC are less numerous
than c-Fos+ galanin-containing neurons encountered following sleep rebound.14
These data support our proposal that two types of sleep-active neurons are
present in VLPO, one assigned to the inhibition of waking-promoting areas during
sleep, the other with unknown efferent projection and function. A priority is to

determine whether Type-1 or Type-2 neurons send their axons to the waking-
promoting areas. While this question is currently under investigation in our labora-
tory, it is already possible to propose some preliminary clues. The reciprocal inhib-
itory interactions and flip-flop switching models both supposed that, during waking,
PSP neurons are strongly inhibited by wake-active cells. Therefore it is more likely
that Type-1 neurons would project to the waking-promoting areas because they
should be maintained completely silent during waking, when inhibitory neurotrans-
mitters NA, 5-HT, ACH, and GABA are maximally released in VLPO. The consol-
idated inhibition of Type-1 neurons during waking allows the required stability for
the system.
In contrast, Type-2 neurons are able to procure a subtle instability required for
behavioral switching (drowsiness). Although under NA and ACH inhibitory drives
during waking, Type-2 neurons would be less hyperpolarized and thus more excitable
because they are responsive for homeostatic signals. They could start firing before
Type-1 neurons, when the release of sleep factors increased in VLPO during long-
lasting awaking. Supporting this hypothesis, a contingent of VLPO sleep-active
neurons begins to fire during drowsiness, just before sleep onset. It is finally possible
that Type-2 neurons, secondary to their own excitation, stimulate, by direct excitation
or desinhibition, the neighboring Type-1 neurons. The increased activity of Type-1
neurons would initiate the down-regulation of waking-promoting areas, definitely
releasing all PSP neurons from their powerful inhibition.
Our current functional model of the neuronal network responsible switching of
behavioral states is summarized by the scheme given in Figure 3.6. During waking
PSP neurons (Type-1 and Type-2) are maintained inactive by combined inhibitions
of NA and ACH waking-promoting systems. Simultaneously 5-HT and ADO are
released in VLPO. In concert with homeostatic, metabolic, and circadian drives,
both compounds gradually activate Type-2 neurons and drowsiness occurs. Likely
by presynaptic mechanisms, Type-2 neurons in turn stimulate neighboring Type-1
neurons, which inhibit the waking-promoting systems, thus reinforcing their own
firing to finally promote sleep. We propose that Type-2 neurons would be responsible
for the preparation and initiation of sleep (permissive neurons) and Type-1 neurons
would be responsible for sleep maintenance (executive neurons).
SUMMARY
In agreement with our initial statement, the main contribution of the present work
is the concrete identification of neurons that are responsible for sleep and the
disclosure that VLPO is a suitable and simple model for additional basic research
to approach at cellular and molecular levels the neurobiological processes underlying
sleep production and regulation.
ACKNOWLEDGMENTS
CNRS FRE 2469 and UMR5167, INSERM U480, Université Claude Bernard Lyon
1 and CNRS UMR 7637 ESPCI in Paris supported this work.

FIGURE 3.6 (See color insert following page 108.) Model of the neuronal network respon-
sible for reciprocal interactions between sleep- and waking-promoting areas, regulating the
sleep-waking switching. Inhibitory pathways are shown in red and the excitatory pathways
in green. The black circle indicates the brain areas involved in sleep regulation. As traffic
lights, circles are colored in green when areas are strongly active, in red when they are inactive,
and yellow when they are in a transitory state (either increasing or decreasing their firing rate).
Abbreviations: 5-HT, serotonin; Ach, acetylcholine; ADO, adenosine; RN, mesencephalic
raphe nuclei; Gal, galanine; Hist, histamine; LC, locus coeruleus; LDT, laterodorsal tegmental
nuclei; NA, noradrenalin; NB, nucleus basalis; TMN, tuberomammillary nucleus; VLPO,
ventrolateral preoptic nucleus.
REFERENCES
1. Von Economo, C., Schlaftheorie, Ergeb. Physiol 55, 121-135, 1929.
2. Szymusiak, R., Steininger, T., Alam, N., and McGinty, D., Preoptic area sleep-
regulating mechanisms, Arch Ital Biol 139 (1-2), 77-92, 2001.
3. McGinty, D. and Szymusiak, R., Hypothalamic regulation of sleep and arousal, Front
Biosci 8, 1074-1083, 2003.
4. Shiromani, P., Scammell, T., Sherin, J. E., and Saper, C. B., Hypothalamic Regulation
of Sleep, in Handbook of Behavioral State Control: Cellular and Molecular Mecha-
nisms, Lydic, R. and Baghdoyan, H. A., CRC Press, Boca Raton, FL, 1999, pp. 311-325.

5. Schmidt, M. H., Valatx, J. L., Sakai, K., Fort, P., and Jouvet, M., Role of the lateral
preoptic area in sleep-related erectile mechanisms and sleep generation in the rat, J
Neurosci 20 (17), 6640-6647, 2000.
6. Saper, C. B., Chou, T. C., and Scammell, T. E., The sleep switch: hypothalamic
control of sleep and wakefulness, Trends Neurosci 24 (12), 726-731, 2001.
7. Moruzzi, G., The sleep-waking cycle, Ergeb Physiol 64, 1-165, 1972.
8. Moruzzi, G. and Magoun, H. W., Brain stem reticular formation and activation of the
EEG., J Neuropsychiatry Clin Neurosci 7 (2), 251-267, 1949.
9. Jones, B., The organization of central cholinergic systems and their functional impor-
tance in sleep-waking states, in Cholinergic Function and Dysfunction. Progress in
Brain Research., Cuello, A.C., Ed., Elsevier, Amsterdam, 1993, p. 61.
10. Jones, B. E., Basic mechanisms of sleep-wakes states, in Principles and Practice of
Sleep Medecine, Kryger, M. H., Roth, T., and Dement, W. C., Eds., W.B. Saunders,
Philadelphia, 1994, pp. 145-162.
11. Borbely, A. A., From slow waves to sleep homeostasis: new perspectives, Arch Ital
Biol 139 (1-2), 53-61, 2001.
12. Borbely, A. A., Sleep regulation. Introduction, Hum Neurobiol 1 (3), 161-162, 1982.
13. Sherin, J. E., Shiromani, P. J., McCarley, R. W., and Saper, C. B., Activation of
ventrolateral preoptic neurons during sleep, Science 271 (5246), 216-219, 1996.
14. Sherin, J. E., Elmquist, J. K., Torrealba, F., and Saper, C. B., Innervation of hista-
minergic tuberomammillary neurons by GABAergic and galaninergic neurons in the
ventrolateral preoptic nucleus of the rat, J Neurosci 18 (12), 4705-4721, 1998.
15. Novak, C. M. and Nunez, A. A., Daily rhythms in Fos activity in the rat ventrolateral
preoptic area and midline thalamic nuclei, Am J Physiol 275, R1620-R1626, 1998.
16. Thompson, R. H., Canteras, N. S., and Swanson, L. W., Organization of projections
from the dorsomedial nucleus of the hypothalamus: a PHA-L study in the rat, J Comp
Neurol 376 (1), 143-173, 1996.
17. Chou, T. C., Bjorkum, A. A., Gaus, S. E., Lu, J., Scammell, T. E., and Saper, C. B.,
Afferents to the ventrolateral preoptic nucleus, J. Neurosci 22 (3), 977-990, 2002.
18. Chou, T. C., Scammell, T. E., Gooley, J. J., Gaus, S. E., Saper, C. B., and Lu, J.,
Critical role of dorsomedial hypothalamic nucleus in a wide range of behavioral
circadian rhythms, J Neurosci 23 (33), 691-702, 2003.
19. Watts, A. G., Swanson, L. W., and Sanchez-Watts, G., Efferent projections of the
suprachiasmatic nucleus: I. Studies using anterograde transport of Phaseolus vulgaris
leucoagglutinin in the rat, J Comp Neurol 258 (2), 204-229, 1987.
20. Sun, X., Whitefield, S., Rusak, B., and Semba, K., Electrophysiological analysis of
suprachiasmatic nucleus projections to the ventrolateral preoptic area in the rat, Eur
J Neurosci 14 (8), 1257-1274, 2001.
21. Novak, C. M. and Nunez, A. A., A sparse projection from the suprachiasmatic nucleus
to the sleep active ventrolateral preoptic area in the rat, Neuroreport 11 (1), 93-96,
2000.
22. Lu, J., Shiromani, P., and Saper, C. B., Retinal input to the sleep-active ventrolateral
preoptic nucleus in the rat, Neuroscience 93 (1), 209-214, 1999.
23. Deurveilher, S., Burns, J., and Semba, K., Indirect projections from the suprachias-
matic nucleus to the ventrolateral preoptic nucleus: a dual tract-tracing study in rat,
Eur J Neurosci 16 (7), 1195-213, 2002.
24. Szymusiak, R., Alam, N., Steininger, T. L., and McGinty, D., Sleep-waking discharge
patterns of ventrolateral preoptic/anterior hypothalamic neurons in rats, Brain Res
803 (1-2), 178-188, 1998.

25. Aston-Jones, G. and Bloom, F. E., Activity of norepinephrine-containing locus coer-

uleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle, J
Neurosci 1 (8), 876-886, 1981.
26. Barnes, N. M. and Sharp, T., A review of central 5-HT receptors and their function,
Neuropharmacology 38 (8), 1083-1152, 1999.
27. Jacobs, B., Overview of the activity of brain monoaminergic neurons across the sleep-
wake cycle, in Sleep: Neurotransmitters and Neuromodulators, Wauquier A. M. J.,
Gaillard J.M, et al. Eds., Raven Press, New York, 1985, p. 1.
28. Sakai, K., Central mechanisms of paradoxical sleep, Brain Dev 8 (4), 402-407, 1986.
29. Lu, J., Greco, M. A., Shiromani, P., and Saper, C. B., Effect of lesions of the
ventrolateral preoptic nucleus on NREM and REM sleep, J Neurosci 20 (10), 3830-
3842, 2000.
30. Schmidt, M. H., Gervasoni, D., Luppi, P. H., and Fort, P., Carbachol administration
into the lateral preoptic area induces penile erections and wakefulness, Neurosci Abstr
522.19, 2001.
31. Schmidt, M. H., Gervasoni, D., Luppi, P. H., and Fort, P., The ventrolateral preoptic
area: role and origin of cholinergic input in the control of wakefulness and penile
erections, Sleep 25 (supp), 2002.
32. Steininger, T. L., Gong, H., McGinty, D., and Szymusiak, R., Subregional organiza-
tion of preoptic area/anterior hypothalamic projections to arousal-related monoam-
inergic cell groups, J Comp Neurol 429, 638-653, 2001.
33. Fort, P., Gervasoni, D., Peyron, C., Rampon, C., Boissard, R., and Luppi, P. H.,
GABAergic projections to the magnocellular preoptic area and substantia innominata
in the rat, Neurosci Abstr 1998.
34. Schmidt, M. H., Gervasoni, D., Luppi, P. H., and Fort, P., Quantitative analysis of
cholinergic afferents to the ventrolateral preoptic area: role in waking mechanisms,
Sleep 26 (supp), 0089, 2003.
35. Luppi, P. H., Aston-Jones, G., Akaoka, H., Chouvet, G., and Jouvet, M., Afferent
projections to the rat locus coeruleus demonstrated by retrograde and anterograde
tracing with cholera-toxin B subunit and Phaseolus vulgaris leucoagglutinin, Neuro-
science 65 (1), 119-160, 1995.
36. Lu, J., Bjorkum, A. A., Xu, M., Gaus, S. E., Shiromani, P. J., and Saper, C. B.,
Selective activation of the extended ventrolateral preoptic nucleus during rapid eye
movement sleep, J Neurosci 22 (11), 4568-4576, 2002.
37. Gervasoni, D., Peyron, C., Rampon, C., Barbagli, B., Chouvet, G., Urbain, N., Fort,
P., and Luppi, P. H., Role and origin of the GABAergic innervation of dorsal raphe
serotonergic neurons, J Neurosci 20 (11), 4217-4225, 2000.
38. Gallopin, T., Fort, P., Eggermann, E., Cauli, B., Luppi, P. H., Rossier, J., Audinat,
E., Mühlethaler, M., and Serafin, M., Identification of sleep-promoting neurons in
vitro, Nature 404 (6781), 992-995, 2000.
39. Gallopin, T., Luppi, P. H., Rambert, F., Frydman, A., and Fort, P., Effect of the wake
promoting agent modafinil on sleep promoting neurons from the ventrolateral preoptic
nucleus: an in-vitro pharmacologic study, Sleep 27 (1), 19-25, 2004.
40. Matsuo, S., Jang, I. S., Nabekura, J., and Akaike, N., alpha 2-Adrenoceptor-mediated
presynaptic modulation of GABAergic transmission in mechanically dissociated rat
ventrolateral preoptic neurons, J Neurophysiol 89 (3), 1640-1648, 2003.
41. Eggermann, E., Serafin, M., Bayer, L., Machard, D., Saint-Mleux, B., Jones, B. E.,
and Mühlethaler, M., Orexins/hypocretins excite basal forebrain cholinergic neurones,
Neuroscience 108 (2), 177-181, 2001.

42. Airaksinen, M. S., Alanen, S., Szabat, E., Visser, T. J., and Panula, P., Multiple
neurotransmitters in the tuberomammillary nucleus: comparison of rat, mouse, and
guinea pig, J Comp Neurol 323 (1), 103-116, 1992.
43. Nishino, S., Ripley, B., Overeem, S., Lammers, G. J., and Mignot, E., Hypocretin
(orexin) deficiency in human narcolepsy, Lancet 355 (9197), 39-40, 2000.
44. Lin, L., Faraco, J., Li, R., Kadotani, H., Rogers, W., Lin, X., Qiu, X., de Jong, P. J.,
Nishino, S., and Mignot, E., The sleep disorder canine narcolepsy is caused by a
mutation in the hypocretin (orexin) receptor 2 gene, Cell 98 (3), 365-376, 1999.
45. Peyron, C., Tighe, D. K., van den Pol, A. N., de Lecea, L., Heller, H. C., Sutcliffe,
J. G., and Kilduff, T. S., Neurons containing hypocretin (orexin) project to multiple
neuronal systems, J Neurosci 18 (23), 9996-10015, 1998.
46. de Lecea, L., Kilduff, T. S., Peyron, C., Gao, X., Foye, P. E., Danielson, P. E.,
Fukuhara, C., Battenberg, E. L., Gautvik, V. T., Bartlett, F. S., Frankel, W. N., van
den Pol, A. N., Bloom, F. E., Gautvik, K. M., and Sutcliffe, J. G., The hypocretins:
hypothalamus-specific peptides with neuroexcitatory activity, Proc Natl Acad Sci U
S A 95 (1), 322-327, 1998.
47. Kilduff, T. S. and Peyron, C., The hypocretin/orexin ligand-receptor system: impli-
cations for sleep and sleep disorders, Trends Neurosci 23 (8), 359-65, 2000.
48. McGinty, D., Alam, M. N., Szymusiak, R., Nakao, M., and Yamamoto, M., Hypo-
thalamic sleep-promoting mechanisms: coupling to thermoregulation, Arch Ital Biol
139 (1-2), 63-75, 2001.
49. Krueger, J. M., Cytokines and Sleep Regulation, in Handbook of Behavioral State
Control: Cellular and Molecular Mechanisms, Lydic, R. and Baghdoyan, H. A., Eds.,
CRC Press, Boca Raton, FL, 1999, pp. 609-622.
50. Krueger, J. M. and Obal, F., Jr., Sleep function, Front Biosci 8, d511-9, 2003.
51. Obal, F., Jr. and Krueger, J. M., Biochemical regulation of non-rapid-eye-movement
sleep, Front Biosci 8, d520-50, 2003.
52. Gallopin, T., Cauli, B., Luppi, P. H., Rossier, J., Lambolez, B., and Fort, P., Serotonin
modulation reveals two types of sleep-promoting neurons in the ventrolateral preoptic
nucleus, FENS Abstr 1, 024-13, 2002.
53. Lambolez, B., Audinat, E., Bochet, P., Crepel, F., and Rossier, J., AMPA receptor
subunits expressed by single Purkinje cells, Neuron 9 (2), 247-258, 1992.
54. Seutin, V., Verbanck, P., Massotte, L., and Dresse, A., Galanin decreases the activity
of locus coeruleus neurons in vitro, Eur J Pharmacol 164 (2), 373-376, 1989.
55. Xu, Z. Q., Zhang, X., Pieribone, V. A., Grillner, S., and Hokfelt, T., Galanin-5-
hydroxytryptamine interactions: electrophysiological, immunohistochemical and in
situ hybridization studies on rat dorsal raphe neurons with a note on galanin R1 and
R2 receptors, Neuroscience 87 (1), 79-94, 1998.
56. Schonrock, B., Busselberg, D., and Haas, H. L., Properties of tuberomammillary his-
tamine neurones and their response to galanin, Agents Actions 33 (1-2), 135-137, 1991.
57. Melander, T., Hokfelt, T., Rokaeus, A., Cuello, A. C., Oertel, W. H., Verhofstad, A.,
and Goldstein, M., Coexistence of galanin-like immunoreactivity with catechola-
mines, 5-hydroxytryptamine, GABA and neuropeptides in the rat CNS, J Neurosci
6 (12), 3640-3654, 1986.
58. Hokfelt, T., Xu, Z. Q., Shi, T. J., Holmberg, K., and Zhang, X., Galanin in ascending
systems. Focus on coexistence with 5-hydroxytryptamine and noradrenaline, Ann N
Y Acad Sci 863, 252-263, 1998.
59. Gundlach, A. L., Burazin, T. C., and Larm, J. A., Distribution, regulation and role of
hypothalamic galanin systems: renewed interest in a pleiotropic peptide family, Clin
Exp Pharmacol Physiol 28 (1-2), 100-105, 2001.

60. Aghajanian, G. K., Electrophysiology of serotonin receptor subtypes and signal

transduction pathways, in Psychopharmacology: The Fourth Generation of Progress,
Bloom, F. R. and Kupfer, D. J., Eds., Raven Press, New York, 1995, pp. 1451-1459.
61. Dugovic, C., Functional activity of 5-HT2 receptors in the modulation of the
sleep/wakefulness states, J Sleep Res 1 (3), 163-168, 1992.
62. Dugovic, C., Wauquier, A., Leysen, J. E., Marrannes, R., and Janssen, P. A., Functional
role of 5-HT2 receptors in the regulation of sleep and wakefulness in the rat, Psy-
chopharmacology (Berl) 97 (4), 436-442, 1989.
63. Martin, J. R., Bos, M., Jenck, F., Moreau, J., Mutel, V., Sleight, A. J., Wichmann, J.,
Andrews, J. S., Berendsen, H. H., Broekkamp, C. L., Ruigt, G. S., Kohler, C., and
Delft, A. M., 5-HT2C receptor agonists: pharmacological characteristics and thera-
peutic potential, J Pharmacol Exp Ther 286 (2), 913-924, 1998.
64. Dugovic, C., Role of serotonin in sleep mechanisms, Rev Neurol (Paris) 157 (11 Pt
2), S16-S19, 2001.
65. Landolt, H. P., Meier, V., Burgess, H. J., Finelli, L. A., Cattelin, F., Achermann, P.,
and Borbely, A. A., Serotonin-2 receptors and human sleep: effect of a selective
antagonist on EEG power spectra, Neuropsychopharmacology 21 (3), 455-466, 1999.
66. Bjorvatn, B., Fagerland, S., Eid, T., and Ursin, R., Sleep/waking effects of a selective
5-HT1A receptor agonist given systemically as well as perfused in the dorsal raphe
nucleus in rats, Brain Res 770 (1-2), 81-88, 1997.
67. Monti, J. M. and Jantos, H., Stereoselective antagonism by the pindolol enantiomers
of 8-OH-DPAT-induced changes of sleep and wakefulness, Neuropharmacology 33
(5), 705-708, 1994.
68. Monti, J. M. and Jantos, H., Dose-dependent effects of the 5-HT1A receptor agonist
8-OH-DPAT on sleep and wakefulness in the rat, J Sleep Res 1 (3), 169-175, 1992.
69. McGinty, D. J. and Harper, R. M., Dorsal raphe neurons: depression of firing during
sleep in cats, Brain Res 101 (3), 569-575, 1976.
70. Houdouin, F., Cespuglio, R., and Jouvet, M., Effects induced by the electrical stim-
ulation of the nucleus raphe dorsalis upon hypothalamic release of 5-hydroxyindole
compounds and sleep parameters in the rat, Brain Res 565 (1), 48-56, 1991.
71. Puizillout, J. J., Gaudin-Chazal, G., Daszuta, A., Seyfritz, N., and Ternaux, J. P.,
Release of endogenous serotonin from “encephale isole” cats. II - Correlations with
raphe neuronal activity and sleep and wakefulness, J Physiol (Paris) 75 (5), 531-537,
1979.
72. Jouvet, M., Biogenic amines and the states of sleep, Science 163 (862), 32-41, 1969.
regulation of the sleep-waking cycle, Ergeb Physiol 64, 166-307, 1972.
74. Jouvet, M., Sleep and serotonin: an unfinished story, Neuropsychopharmacology 21
(Suppl), 24S-27S, 1999.
75. Arpa, J. and De Andres, I., Re-examination of the effects of raphe lesions on the
sleep/wakefulness cycle states in cats, J Sleep Res 2 (2), 96-102, 1993.
76. Delorme, F., Froment, J. L., and Jouvet, M., Suppression of sleep with p-chlo-
romethamphetamine and p-chlorophenylalanine, C R Seances Soc Biol Fil 160 (12),
2347-2351, 1966.
77. Borbely, A. A., Neuhaus, H. U., and Tobler, I., Effect of p-chlorophenylalanine and
tryptophan on sleep, EEG and motor activity in the rat, Behav Brain Res 2 (1), 1-22,
1981.
78. Petitjean, F., Buda, C., Janin, M., Sallanon, M., and Jouvet, M., Insomnia caused by
administration of para-chlorophenylalanine: reversibility by peripheral or central
injection of 5-hydroxytryptophan and serotonin, Sleep 8 (1), 56-67, 1985.

79. Touret, M., Sarda, N., Gharib, A., Geffard, M., and Jouvet, M., The role of 5-
hydroxytryptophan (5-HTP) in the regulation of the sleep/wake cycle in parachlo-
rophenylalanine (p-CPA) pretreated rat: a multiple approach study, Exp Brain Res 86
(1), 117-124, 1991.
80. Denoyer, M., Sallanon, M., Kitahama, K., Aubert, C., and Jouvet, M., Reversibility
of para-chlorophenylalanine-induced insomnia by intrahypothalamic microinjection
of L-5-hydroxytryptophan, Neuroscience 28 (1), 83-94, 1989.
81. Jouvet, M., Neuromediators and hypnogenic factors, Rev Neurol (Paris) 140 (6-7),
389-400, 1984.
82. Sallanon, M., Buda, C., and Janin, M., Implication of serotonin in sleep mechanisms:
induction, facilitation?, in Sleep: Neurotransmitters and Neuromodulators, A. Wau-
quier, J.M. Monti, and J.M. Gaillard, Eds., Raven Press, New York, 1985, p. 136.
83. Python, A., Steimer, T., De Saint, H., Mikolajewski, R., and Nicolaidis, S., Extracel-
lular serotonin variations during vigilance states in the preoptic area of rats: a microdi-
alysis study, Brain Res 910 (1-2), 49-54, 2001.
84. Hayaishi, O., Molecular genetic studies on sleep-wake regulation, with special empha-
sis on the prostaglandin D(2) system, J Appl Physiol 92 (2), 863-868, 2002.
85. Ueno, R., Ishikawa, Y., Nakayama, T., and Hayaishi, O., Prostaglandin D2 induces
sleep when microinjected into the preoptic area of conscious rats, Biochem Biophys
Res Commun 109 (2), 576-582, 1982.
86. Koyama, Y. and Hayaishi, O., Modulation by prostaglandins of activity of sleep-
related neurons in the preoptic/anterior hypothalamic areas in rats, Brain Res Bull 33
(4), 367-372, 1994.
87. Koyama, Y. and Hayaishi, O., Firing of neurons in the preoptic/anterior hypothalamic
areas in rat: its possible involvement in slow wave sleep and paradoxical sleep,
Neurosci Res 19 (1), 31-38, 1994.
88. Osaka, T. and Matsumura, H., Noradrenergic inputs to sleep-related neurons in the
preoptic area from the locus coeruleus and the ventrolateral medulla in the rat,
Neurosci Res 19 (1), 39-50, 1994.
89. Osaka, T. and Matsumura, H., Noradrenaline inhibits preoptic sleep-active neurons
through alpha 2-receptors in the rat, Neurosci Res 21, 323-330, 1995.
90. Pandey, H. P., Ram, A., Matsumura, H., and Hayaishi, O., Concentration of prosta-
glandin D2 in cerebrospinal fluid exhibits a circadian alteration in conscious rats,
Biochem Mol Biol Int 37 (3), 431-437, 1995.
91. Ram, A., Pandey, H. P., Matsumura, H., Kasahara-Orita, K., Nakajima, T., Takahata,
R., Satoh, S., Terao, A., and Hayaishi, O., CSF levels of prostaglandins, especially
the level of prostaglandin D2, are correlated with increasing propensity towards sleep
in rats, Brain Res 751 (1), 81-89, 1997.
92. Matsumura, H., Nakajima, T., Osaka, T., Satoh, S., Kawase, K., Kubo, E., Kantha,
S. S., Kasahara, K., and Hayaishi, O., Prostaglandin D2-sensitive, sleep-promoting
zone defined in the ventral surface of the rostral basal forebrain, Proc Natl Acad Sci
U S A 91 (25), 11998-12002, 1994.
93. Scammell, T., Gerashchenko, D., Urade, Y., Onoe, H., Saper, C., and Hayaishi, O.,
Activation of ventrolateral preoptic neurons by the somnogen prostaglandin D2, Proc
Natl Acad Sci U S A 95 (13), 7754-7759, 1998.
94. Mizoguchi, A., Eguchi, N., Kimura, K., Kiyohara, Y., Qu, W. M., Huang, Z. L.,
Mochizuki, T., Lazarus, M., Kobayashi, T., Kaneko, T., Narumiya, S., Urade, Y., and
Hayaishi, O., Dominant localization of prostaglandin D receptors on arachnoid tra-
becular cells in mouse basal forebrain and their involvement in the regulation of non-
rapid eye movement sleep, Proc Natl Acad Sci U S A 98 (20), 11674-11679, 2001.

95. Satoh, S., Matsumura, H., and Hayaishi, O., Involvement of adenosine A2A receptor
in sleep promotion, Eur J Pharmacol 351 (2), 155-162, 1998.
96. Satoh, S., Matsumura, H., Koike, N., Tokunaga, Y., Maeda, T., and Hayaishi, O.,
Region-dependent difference in the sleep-promoting potency of an adenosine A2A
receptor agonist, Eur J Neurosci 11 (5), 1587-1597, 1999.
97. Satoh, S., Matsumura, H., Suzuki, F., and Hayaishi, O., Promotion of sleep mediated
by the A2a-adenosine receptor and possible involvement of this receptor in the sleep
induced by prostaglandin D2 in rats, Proc Natl Acad Sci U S A 93 (12), 5980-5984, 1996.
98. Scammell, T. E., Gerashchenko, D. Y., Mochizuki, T., McCarthy, M. T., Estabrooke,
I. V., Sears, C. A., Saper, C. B., Urade, Y., and Hayaishi, O., An adenosine A2a agonist
increases sleep and induces Fos in ventrolateral preoptic neurons, Neuroscience 107
(4), 653-663, 2001.
99. Urade, Y., Eguchi, N., Qu, W. M., Sakata, M., Huang, Z. L., Chen, J. F., Schwarzschild,
M. A., Fink, J. S., and Hayaishi, O., Minireview: Sleep regulation in adenosine A(2A)
receptor-deficient mice, Neurology 61 (Suppl 6), S94-6, 2003.
100. Barrie, A. P. and Nicholls, D. G., Adenosine A1 receptor inhibition of glutamate
exocytosis and protein kinase C-mediated decoupling, J Neurochem 60 (3), 1081-
1086, 1993.
101. Dunwiddie, T. V., The physiological role of adenosine in the central nervous system,
Int Rev Neurobiol 27, 63-139, 1985.
102. Proctor, W. R. and Dunwiddie, T. V., Pre- and postsynaptic actions of adenosine in
the in vitro rat hippocampus, Brain Res 426 (1), 187-190, 1987.
103. Mcllwain, H., Adenosine and its mononucleotides as regulatory and adaptative signals
in the brain, in Physiological and Regulatory Functions of Adenosine and Adenine
Nucleotides, Baer, H. P. and Drummond, G. I., Eds., Raven Press, New York, 1979,
pp. 361-376.
104. Daval, J. L. and Nicolas, F., Non-selective effects of adenosine A1 receptor ligands
on energy metabolism and macromolecular biosynthesis in cultured central neurons,
Biochem Pharmacol 55 (2), 141-149, 1998.
105. Benington, J. H. and Heller, H. C., Restoration of brain energy metabolism as the
function of sleep, Prog Neurobiol 45 (4), 347-360, 1995.
106. Alam, M. N., Szymusiak, R., Gong, H., King, J., and McGinty, D., Adenosinergic
modulation of rat basal forebrain neurons during sleep and waking: neuronal record-
ing with microdialysis, J Physiol (London) 521, 679-690, 1999.
107. Rainnie, D. G., Grunze, H. C., McCarley, R. W., and Greene, R. W., Adenosine
inhibition of mesopontine cholinergic neurons: implications for EEG arousal, Science
263 (5147), 689-692, 1994.
108. Strecker, R. E., Morairty, S., Thakkar, M. M., Porkka-Heiskanen, T., Basheer, R.,
Dauphin, L. J., Rainnie, D. G., Portas, C. M., Greene, R. W., and McCarley, R. W.,
Adenosinergic modulation of basal forebrain and preoptic/anterior hypothalamic neu-
ronal activity in the control of behavioral state, Behav Brain Res 115 (2), 2000.
109. Cunha, R. A., Adenosine as a neuromodulator and as a homeostatic regulator in the
nervous system: different roles, different sources and different receptors, Neurochem
Int 38 (2), 107-125, 2001.
110. Barajas-Lopez, C., Surprenant, A., and North, R. A., Adenosine A1 and A2 receptors
mediate presynaptic inhibition and postsynaptic excitation in guinea pig submucosal
neurons, J Pharmacol Exp Ther 258 (2), 490-495, 1991.
111. Umemiya, M. and Berger, A. J., Activation of adenosine A1 and A2 receptors differ-
entially modulates calcium channels and glycinergic synaptic transmission in rat
brainstem, Neuron 13 (6), 1439-1446, 1994.

112. Li, H. and Henry, J. L., Adenosine A2 receptor mediation of pre- and postsynaptic
excitatory effects of adenosine in rat hippocampus in vitro, Eur J Pharmacol 347 (2-
3), 173-182, 1998.
113. Sebastiao, A. M. and Ribeiro, J. A., Adenosine A2 receptor-mediated excitatory
actions on the nervous system, Prog Neurobiol 48 (3), 167-189, 1996.
114. Adrien, J., Adenosine in sleep regulation, Rev Neurol (Paris) 157 (11 Pt 2), S7-11,
2001.
115. Porkka-Heiskanen, T., Strecker, R. E., and McCarley, R. W., Brain site-specificity of
extracellular adenosine concentration changes during sleep deprivation and sponta-
neous sleep: an in vivo microdialysis study, Neuroscience 99 (3), 507-517, 2000.

1519_C04.fm Page 65 Friday, August 27, 2004 9:50 AM
4 Molecular Mechanisms
of Sleep-Wake
Regulation: A Role of
Prostaglandin D2 and
Adenosine
Osamu Hayaishi
CONTENTS
Prostaglandins and Sleep

Prostaglandins
Prostaglandin D2 Induces Physiological Sleep in Rats and Monkeys
Experiments with Rodents
Experiments with Primates
PGE2 Promotes Wakefulness
Mechanisms Underlying Sleep-Wake Regulation by PGD2
Prostaglandin D Synthase (PGDS)
Structure and Function
Localization of PGDS
PGD Receptor (DPR, D-Type PG Receptor)
Adenosine and A2A Receptor
VLPO and TMN
Molecular Genetic Studies
TG Mice for PGDS
KO Mice for PGDS and DPR
A2AR KO Mice
H1R KO Mice and EP4 Receptor
Summary
Acknowledgments
References
0-8493-1519-0/05/$0.00+$1.50

PROSTAGLANDINS AND SLEEP

PROSTAGLANDINS
The prostaglandins (PGs) are a family of naturally occurring lipids with a unique
five-membered carbon ring and are derived from various polyunsaturated fatty acids
such as arachidonic acid. These compounds are widely distributed in virtually all
types of cells in mammalian tissues and organs, and exhibit numerous and diverse
biological effects on a wide variety of physiological and pathological activities such
as contraction and relaxation of muscles, platelet aggregation, inflammation, fever,
and so forth; hence they are generally referred to as tissue hormones or local
hormones. However relatively little attention was paid to the PGs in the central
nervous system (CNS) of mammals until in early 1980s, when we found PGD2 to
be the major prostanoid in the brain of rats1,2 and other mammals, including humans.3
PGD2 had long been considered as a minor and biologically relatively inactive
prostanoid, and therefore our findings suggested that PGD2 may be a unique con-
stituent of the brain of mammals and that it may play some important and possibly
specific function in this organ.
PROSTAGLANDIN D2 INDUCES PHYSIOLOGICAL SLEEP IN RATS AND MONKEYS

Experiments with Rodents
In an attempt to elucidate the neural function of PGD2, we microinjected nmolar

quantities of PGD2 in the preoptic area (POA) of rats and discovered that the amount
of slow-wave sleep (SWS) was increased dose-dependently up to approximately
sixfold with a concomitant drop in the colonic temperature.4 Subsequently we
examined the somnogenic activity of PGD2 in more detail by using the continuous
infusion circadian sleep bioassay system originally developed by Honda and Inoué.5
The effect of this PG was dose-dependent, and as little as 60 fmol/min of PGD2 was
effective in inducing both SWS and paradoxical sleep (PS). The sleep-inducing effect
of PGD2 was specific, and other PGs were much less effective or totally inactive.
Sleep induced by PGD2 was indistinguishable from physiological sleep as judged
from EEG and EMG recordings, heart rate, locomotor activities, and the general
behavior of the rat.6 PGD2 was not pyrogenic and actually caused a slight decrease
in the body temperature, as is observed to occur in physiological sleep. The PGD2
concentration in rat CSF showed a circadian change coupled to the sleep-wake cycle7
and elevates with an increase in sleep propensity during sleep deprivation.8 Further-
more as PGD2 was found to be actively synthesized and metabolized in the brain,
it was tentatively concluded to be an endogenous regulator of physiological sleep.9
Experiments with Primates
To extend our studies to primates, we then developed a continuous sleep bioassay

system for use in monkeys and investigated the sleep-inducing activity of PGD2 in the
rhesus monkey, Macaca mulatta.10 When PGD2 was infused into the lateral or the third
ventricle of the cerebrum during the light period, the amount of total sleep increased

up to three- to fourfold over the control level. As in the case of rats, sleep induced by
PGD2 in this primate appears to be normal on the basis of the following criteria:
• Polygraphic characteristics such as EEG, EMG, EOG, and heart rate, as

well as general behavior, were indistinguishable from those observed
during natural sleep.
• Power spectral analyses were carried out with an analyzer equipped with
a Fast Fourier Transform System. The data obtained were also similar to
those collected during natural sleep. On the other hand, those data obtained
after oral administration of benzodiazepine derivatives such as nitrazepam
were clearly different from the latter in that there was a sharp peak in the
d wave region about 0.5–3 Hz, and the q band about 4–7 Hz seen during
natural sleep was almost nonexistent. Instead the so-called benzodiazepine
fast wave with its peak about 20 Hz was clearly visible.10
• Sleep occurred episodically, and both REM and NREM sleep were
observed.
• Body temperature and heart rate fluctuated according to the change in
sleep stages.
• The monkeys were easily aroused by external stimuli, indicating their
sleep to be clearly different from that induced by common hypnotic agents
such as barbiturates, benzodiazepines, and so forth.
PGD2 was reported to be involved in the pathogenesis of mastocytosis, a disorder

characterized by massive, episodic, and endogenous production of PGD2 accompa-
nied by lethargy and deep sleep episodes.11 Subsequently by using radioimmunoas-
say, Pentreath and coworkers12 determined the levels of PGD2, PGE2, and related
compounds in samples of CSF from patients with African sleeping sickness, which
is caused by Trypanosoma. PGD2 concentrations were selectively and progressively
elevated in the advanced-stage patients. This correlation may indicate that sleep in
the late stage of sleeping sickness may be caused, at least in part, by increased
production of endogenous PGD2. The possibility that PGD2 is produced by enzymic
action in the cell body of the parasites, such as Trypanosoma brucei gambiense, can
not be completely ruled out, because recent available evidence indicates that PGs
are not only produced and widely distributed in higher animals but also in parasites,
such as cestodes, trematodes, nematodes, and protozoa. PGF2a synthetase, the
enzymes catalyzing the synthesis of one of these PGs was highly purified from
Trypanosoma brucei and characterized and PGD2 was also shown to be produced
in and secreted from trypanosomes both in vitro and in vivo.13 These results show
that PGD2 is a somnogenic agent not only in rats and monkeys but in humans also.
PGE2 PROMOTES WAKEFULNESS

PGE2 and D2 are positional isomers (Figure 4.1) and are known to exhibit opposite
biological effects. For example, PGD2 lowers body temperature, suppresses secretion
of luteinizing hormone-releasing hormone, and decreases the transmucosal potential

Cell membrane
Phospholipase A 2
COOH
Arachidonic acid
COOH
Cyclooxygenase
O
O
COOH COOH
O
O O
OH OH OH OH
PGH2 TXA 2
PGI2 PGD
synthase
OH OH O
COOH COOH COOH
OH OH O OH OH OH
PGF 2 α PGD2 PGE 2
FIGURE 4.1 Arachidonate cascade system.
difference in rat colon mucosa, whereas PGE2 has the opposite effect; i.e., it causes
an increase in body temperature, stimulates the hormone secretion, and increases
the potential difference. As for the effect of the E-series PGs on sleep, reports by
previous investigators have been inconsistent, probably due to differences in animal
species, site of application, and other conditions.
In 1988 Matumura et al.14 demonstrated that microinjection of PGE2 into the
POA reduced the amount of diurnal sleep of rats, indicating that PGE2 may induce
wakefulness. The awaking effect of this PG was further examined by use of a long
term sleep bioassay system.15 Under more physiological conditions, both NREM
and REM sleep were dose-dependently reduced. The rebound of both NREM and
REM sleep was observed during the night after PGE2 infusion. NREM sleep reduc-
tion was due to the shortened duration of episodes while REM sleep reduction
resulted from both the shortened duration and the decreased number of episodes.
Under the experimental conditions, PGE2 also induced hyperthermia. However there
seems to be no evidence to support the cause-effect relationships between changes
in sleep-wake activities and temperature alterations.
In order to investigate whether endogenous PGE2 is indeed involved in physio-
logical regulation of sleep-wake cycle in rats, Matumura et al.16 tested the effect of
AH6809 (6-isopropoxy-9-oxoxantheme-2-carboxylic acid, Glaxo), an antagonist of
the PGE2 receptor. When AH 6809 was infused at a rate of 20 pmol/min into the
third ventricle of a rat during the night, when the animal is normally awake, the
amount of NREM sleep was increased by 22% over the control, and that of REM
sleep, by 89%. If PGE2 induces wakefulness or inhibits sleep under physiological

conditions, such a PGE2 antagonist should counteract the effect of endogenous PGE2
and increase the amount of sleep or decrease the amount of wakefulness, provided
that it has no effect on the PGD2 system. Such results were interpreted to mean that
PGE2 is involved in the maintenance of the arousal state under physiological con-
ditions. It is interesting to note that the infusion of AH6809 increased the amount
of sleep but that the brain temperature was hardly affected. When both PGE2 and
AH6809 were infused simultaneously, the waking effect of PGE2 was completely
neutralized, whereas the pyrogenic effect of PGE2 was not inhibited. These results
indicate that the waking effect of PGE2 is probably independent of the pyrogenic
effect and that the PGE2 receptors for waking activity and temperature regulation
are probably different, one being sensitive to AH6809 and the other insensitive to
this antagonist.
These studies were further extended to rhesus monkeys. PGE2 was administered
through a microdialysis probe into 11 brain loci and the promotion of wakefulness
and elevation of brain temperature were monitored.17 The hyperthermic effect of
PGE2 was dose-dependent and most potent in the POA. Its waking effect was also
dose-dependent and was most pronounced in the tuberomammilary nucleus (TMN)
region in the posterior hypothalamus (PH). The waking response of PGE2 was not
correlated with the change in brain temperature. For example when a low dose of
PGE2 (< 100 pmol/m) was administered into the TMN, the time spent awake during
the infusion period increased up to 3.5-fold, and the amount of SWS decreased to
50% of that of the control level, with negligible changes in brain temperature.17
These results clearly indicate that PGE2 is a wakefulness inducing substance in
monkeys also and that its arousal-promoting activity is independent of its hyperther-
mia effect, and is mediated in a specific site in the TMN/PH region.
The earlier studies mentioned in this section up to April 1991 were reviewed
mainly in two previous review articles.9,18
MECHANISMS UNDERLYING SLEEP-WAKE

REGULATION BY PGD2
PROSTAGLANDIN D SYNTHASE (PGDS)
Structure and Function
PGs of the two series and thromboxane A2 are produced from a common substrate,
arachidonic acid, through the arachidonate cascade system, as shown in Figure 4.1.
In this pathway the oxygenation of arachidonate to PGH2 is catalyzed by cycloox-
ygenases (COX-I and -II), and this step has been believed to be the rate-limiting
step in the production of the final metabolites such as PGD2, E2 etc.
PGDS (EC 5. 3. 99. 2) catalyzes the isomerization of PGH2 to PGD2 (Figure
4.1). Because it is the key enzyme in sleep-wake regulation, we studied its structure,
properties, and function in detail. The detailed account has been published in several
previous reviews19–21 and so only a brief summary is presented here. Two distinct
types of PGDS have been reported, one the lipocalin type PGDS (L-PGDS), also
known as the brain-type or glutathione (GSH)-independent enzyme and the other

hematopoietic PGDS, the spleen-type or GSH-requiring enzyme. Hereafter PGDS

refers to L-PGDS unless specified otherwise.
L-PGDS was purified from the brains of rats, humans and frogs. From cDNA
cloning studies and sequence analyses of crystalline enzyme preparations, it was
found to be an N-glycosylated monomeric protein with a molecular weight of about
25,000. The primary structure indicated it to be a member of the lipocalin super-
family and an ectoprotein that is easily secreted from the cells. The enzyme requires
free sulfhydryl compounds for catalysis, but this requirement is not specific to GSH.
All these catalytic and structural properties are intimately related to its function in
sleep regulation, as will be described later.
Although PGDS is a very stable enzyme and highly resistant to heat treatment,
it is inhibited reversibly and specifically by relatively low concentrations of inorganic
quadrivalent selenium (Se4+) compounds.22 This inhibition is probably due to their
interaction with the free SH group in the active center, because this inhibition can
be easily reversed by the addition of excess amounts of sulfhydryl compounds, such
as GSH or diethyl-dithiothreitol (DTT). When picomolar quantities of SeCl4 were
infused into the third ventricle of a rat during the day, both SWS and REMS were
inhibited time- and dose-dependently. After about 2 h from the start of infusion,
both types of sleep were almost completely inhibited. The effect was reversible, and
when the infusion was interrupted or SH compounds, such as GSH or DTT, were
infused simultaneously, either sleep was restored in agreement with results or the
in vitro enzyme activity.23 Further studies showed that intravenous administration of
tetravalent selenium compounds inhibited the sleep of freely moving rats.24 Recently
intracerebroventricular infusion of a PGD receptor (DPR) antagonist (ONO4127)
was also shown to inhibit sleep reversibly and dose-dependently.25 These results,
taken together, clearly show that PGD2 is an endogenously produced, natural sleep-
promoting substance, or a sleep hormone, and that PGDS and DPR play a crucial
role in sleep regulation under physiological conditions.
Localization of PGDS
To determine the localization of PGDS in the rat brain, we employed three indepen-
dent approaches, namely: (1) in situ hybridization to detect messenger RNA (mRNA)
of PGDS, (2) immunohistochemical staining of the enzyme protein, and (3) the
direct determination of enzyme activity. Our results yielded much important and
some unexpected information, which gave us a new insight into the mechanism of
the somnogenic activity of PGD2. The results of the in situ hybridization studies
revealed that the mRNA was expressed intensely in the membrane system surround-
ing the brain rather than in the brain parenchyma, namely, in the leptomeninges;
i.e., the arachnoid membrane of the brain and spinal cord, and also in the choroid
plexus in the ventricles. The mRNA was only faintly and diffusely expressed in the
brain parenchyma, mainly in the white matter rather than in the gray matter, espe-
cially in the corpus callosum.26 Immunohistochemical detection of the PGDS enzyme
protein also revealed essentially the same results. The oligodendrocytes were positive
for both mRNA and protein staining, but little, if any, of either was observed in
other types of cells including neurons. Further studies on the mouse brain27 were in

essential agreement with the results obtained with rats and clearly showed that
mRNA for PGDS and the enzyme protein were mainly localized in the trabecular
cells of the entire leptomeninges and also in the epithelial cells of the choroid plexus.
We then determined the specific activity of PGDS in different parts of the rat
brain parenchyma as well as in these membranous tissues.26 The specific activities
of the PGDS of the choroid plexus and arachnoid membrane were several-fold higher
than that activity in the whole brain. Both rat and human CSF contained a remarkably
large amount of PGDS activity. In the meantime Zahn et al.28 and Hoffmann et al.29
independently and concurrently reported the amino acid sequence of human brain
PGDS to be highly homologous to that of b-trace, a major protein of unknown
function in the human CSF. Watanabe et al.30 quickly confirmed these findings and
further established the fact that b-trace and PGDS are not only structurally but also
enzymatically and immunologically identical. These results taken together were
interpreted to mean that PGDS is mainly, if not exclusively, present in the membrane
system surrounding the brain, namely the arachnoid membrane and choroid plexus,
where PGD2 is dominantly produced. PGDS, being a secretory protein, is then
secreted into the CSF to become b-trace. This b-trace and the PGD2 thus produced
circulate in the CSF in the ventricular and subarachnoid space between the arachnoid
membrane and pia mater.
PGD RECEPTOR (DPR, D-TYPE PG RECEPTOR)

For determination of the site at which PGD2 acts to induce sleep, three different
approaches were employed: (1) autoradiography to detect the PGD2 binding protein,
(2) microinfusion of PGD2 via a microdialysis probe into different areas of the brain,
and (3) immunohistochemical detection of the DPR.
Autoradiographic image analyses by computerized densitometry and color cod-
ing of the binding protein with 3H-labeled PGD2 showed that the PGD2 binding
occurred in the POA of the rat brain.31 In good agreement with this result, when a
pmolar amount of PGD2 was infused through a microdialysis probe into more than
200 different areas in the rat brain, PGD2 failed to induce sleep in all parts of the
brain parenchyma except in the POA, where a weak somnogenic activity was con-
sistently observed. The most pronounced sleep-inducing activity was observed,
however, when PGD2 was applied to the subarachnoid space in the medial ventral
region of the rostral basal forebrain.32 Finally the location of DPR in the rat and
mouse brains was visualized with antibody highly specific for DPR.27 The DPR
immunoreactivity was localized almost exclusively in the leptomeninges on the
ventral surface of the basal forebrain with weak immunoreactivity in the pia/arach-
noid membrane in the choroid plexus of the lateral and third ventricles. In contrast,
the PGDS immunoreactivity was localized in the leptomeninges surrounding the
entire brain and in the choroid plexus, in good agreement with our previous prelim-
inary studies with rat brain. The DPR immunoreactivity was not found in the dura
mater, pia mater, or brain parenchyma. Electron microscopic studies on the mouse
brain clearly showed that DPR-expressing cells were arachnoid trabecular cells and
that the immunogold particles were mainly located on the plasma membranes and
with less frequency on the intracellular membrane structures such as the vesicles

and endoplasmic reticulum.27 Little, if any, immunoreactivity was seen in arachnoid

barrier cells and pia mater cells. Most DPR-expressing cells were also positive for
PGDS, indicating that PGD2 acts as an autocrine as well as paracrine agent, although
PGD2 produced in other parts of the brain, such as CSF, may also contribute to
promoting sleep.
More than 700 serial coronal sections were used to find the exact location of
DPR-expressing cells in the mouse brain.27 The DPR-positive cells were highly
concentrated in the ventral surface of the rostral basal forebrain whereas other areas
were almost completely negative. The region with concentrated DPRs was clearly
defined as bilateral wings in the rostral basal forebrain lateral to the optic chiasm
in the proximity of the ventrolateral preoptic (VLPO) area, a known sleep center,
and the TMN, a known wake center. The rostral and main portions of this region
were associated with the visual pathway composed of the optic nerves, optic chiasm,
and optic tracts.
PGD2 infusion into the lateral ventricle of mice increased preferentially NREM
sleep rather than REM sleep27 in good agreement with our previous observation that
NREM sleep was selectively induced by PGD2 infusion into the DPR-rich area in
the basal forebrain of rats.32 In the DPR knockout (KO) mice, the amount of NREM
sleep did not increase after PGD2 infusion into the brain.27 These results taken
together clearly show that PGD2 produced in the leptomeninges and choroid plexus,
and possibly in the CSF, circulates in the ventricular and subarachnoid space, and
binds to DPRs in the basal forebrain to initiate the NREM sleep.
ADENOSINE AND A2A RECEPTOR

To find out how the sleep signal initiated by the binding of PGD2 to the DPR in the
surface of the basal forebrain is transduced into the brain parenchyma, we applied
numerous neurotransmitters, peptides, hormones, and other bioactive substances to
the DPR-enriched sleep-promoting zone to see if any of these compounds could
replace or mimic the somnogenic activity of PGD2. Among several hundred test
compounds, only adenosine and adenosine A2A receptor (A2AR) agonists such as 2-
(4-(2-carboxyethyl) phenylethylamino) adenosine-5’-N-ethykcarboxamideadenos-
ine (CGS21680) and 2-(4-(2-(2-aminoethylamino-carbonyl)ethyl) phenylethy-
lamino) -5’-N-ethylcarboxamidoadenosine (APEC) were effective and induced
NREM, but not REM sleep when infused into rats during the night.33 On the other
hand, A1-receptor agonists, such as N6-cyclohexyladenosine (CHA) and N6-cyclo-
pentyladenosine (CPA), were ineffective. In rats pretreated by intraperitoneal infu-
sion of a selective antagonist of A2AR, KF17837, the sleep-inducing effects of both
A2AR agonists and PGD2 were attenuated, indicating that the somnogenic effect of
PGD2 may be mediated by adenosine via A2AR. The extracellular level of adenosine
in the subarachnoid space of the basal forebrain was increased dose-dependently by
the infusion of PGD2 or the DPR agonist BW245C in rats, and this effect was
attenuated by the simultaneous treatment of the rats with a DPR antagonist,
BWA868C. The PGD2-induced increase in extracellular adenosine was also found
in wild type (WT) mice but was not observed in the DPR KO mice.27 More recent
experimental results obtained from A2AR KO mice showed that PGD2 exerted its

Posterior
Basal forebrain Preoptic area
hypothalamus
PGDS
PGDS
Histamine
GABA
Galanin TMN
VLPO
A 2A R
DPR
PGD2 PGE 2, Orexin
FIGURE 4.2 Schematic representation of the molecular mechanisms of sleep-wake regula-
tion by PGD2, E2, adenosine, histamine, and orexin.
sleep-promoting effects in a manner at least partially dependent on the A2AR.34 We

conclude that A2AR is involved in the sleep-promoting action of PGD2 and that
adenosine plays an important role in the sleep promotion initiated by PGD2.
VLPO AND TMN

The immediate early gene product cFos (cellular feline osteosarcoma) has widely
been used as a useful marker of neuronal activation. Fos, the protein encoded by
the gene, is a transcription factor that triggers transcription in a cascade of cellular
responses. To determine which neuron groups are involved in response to PGD2 or
adenosine, especially A2A agonists, we examined c-Fos immunoreactivity.35–37
When PGD2 or the A2AR agonist CGS 21680 was infused for 2 h into the
subarachnoid space in the PGD2-sensitive zone, a marked increase in the number of
Fos-positive cells was observed in the leptomeningeal membrane on the ventral
surface of the basal forebrain as well as in the VLPO area concomitant with the
induction of NREM sleep. In contrast the number of Fos-positive neurons decreased
markedly in the TMN of the PH. Using Fos immunoreactivity, Sherin et al.38 showed
a discrete cluster of neurons in the VLPO to play a critical role in the generation of
sleep. The VLPO is known to send specific inhibitory GABAergic and galaninergic
efferents to the TMN, which neurons contain the ascending histaminergic arousal
system (Figure 4.2).
PGD2 does not induce sleep when infused into the TMN32 and therefore it is
unlikely that putative wake neurons in the TMN are directly inhibited by PGD2.
PGD2 induces sleep most effectively when infused into the subarachnoid space in
the PGD2-sensitive zone. PGD2 increased the firing rates of sleep-active neurons in
the POA,39 where these neurons are most abundant in the VLPO. The VLPO may
induce sleep by inhibiting wake-promoting neurons in the TMN, for both GABA
and galanin were shown to inhibit the firing rate of wake-active TMN neurons.38

PGD2 also induced Fos-IR in the leptomeninges,35 which suggests that PGD2
activates VLPO via leptomeningeal DP receptors. These results taken together
strongly indicate that PGD2 may bind to the DPRs in the PGD2-sensitive zone, where
meningeal cells release paracrine signaling molecules such as adenosine, which
subsequently excite nearby sleep-active VLPO neurons. These VLPO neurons may
directly induce NREM sleep or send inhibitory signals to the TMN to down-regulate
the wake neurons; thus sleep-wake cycle is regulated by a flip-flop mechanism
involving the interaction between these two centers.
The results of biochemical and pharmacological studies up to the end of 2002
described in this section have been reviewed in previous articles.40–42
MOLECULAR GENETIC STUDIES

For elucidation of the exact role of the PGD2 system and possibly of the E2 one in
sleep-wake regulation in vivo under physiological conditions and also for delineation
of the genetic mechanisms involved in this process, the sleep behavior of KO mice
for L-PGDS, DPR, adenosine A2AR, and histamine H1 receptor (H1R) generated in
the laboratories of the author and others, was examined. Transgenic (TG) mice over-
expressing the human PGDS gene generated in my laboratory were also tested. In
all these experiments, the circadian profiles of sleep-wake patterns of WT and the
genetically engineered mice were essentially identical under macroscopic examina-
tion. However, gross phenotypic changes were observed under certain specific con-
ditions, such as extraneous physical stimuli, sleep deprivation, and so forth. These
results may be interpreted to mean that sleep is essential for life and that the sleep-
regulatory system is composed of a complicated compensatory network in which
the deficiency of one system may be effectively compensated by other systems during
development.
TG MICE FOR PGDS

In the year 2000 we generated TG mice by incorporating the human PGDS gene
into mice. Northern blot analysis clearly showed that human PGDS mRNA was
expressed in almost all tissues and organs of these mice.43 We expected these mice
to sleep all the time, but the mice appeared to be very healthy and to grow and sleep
normally. As shown in Figure 4.3, there was no significant difference in the circadian
sleep pattern between the WT and TG mice, but when the tails of these mice were
clipped for DNA sampling at 8:00 PM, as indicated by the arrows, we found that the
amount of SWS of the TG mice increased sharply and significantly. This effect lasted
for several hours, and the amount of SWS returned to the control level after 5–6 h.
The maximum increment was almost as high as the maximum amount of sleep
during the daytime. The sleep pattern of the WT mice was essentially unaffected by
the tail clipping, although the amount of SWS slightly decreased, probably due to
the pain. In both cases the amount of REM sleep did not differ significantly.
These somewhat unexpected and puzzling experimental observations may be
explained on the basis of the sequence of enzyme reactions depicted in Figure 4.1.

Wild type TG
Amount of SWS 50
*
(min/hr)
**
tail clip
30 * * *
* **
baseline baseline
*
after tc after tc
10
20:00 8:00 20:00 20:00 8:00 20:00
Clock time
FIGURE 4.3 Increase in SWS in TG mice after tail clipping n = 7 for WT and n = 10 for
TG mice. * P<0.05 ** P<0.01 Compared with the baseline day by the paired t test.
In the arachidonate cascade system, the enzyme cyclooxygenase, existing as con-

stitutive COX I and inducible COX II, is generally believed to be the rate-limiting
enzyme rather than the individual synthases under physiological conditions, which
is why the sleep patterns of WT and TG mice were essentially identical. However,
it is possible that the pain stimulus caused by the tail clipping elicited induction of
the inducible COX II, which then produced an excessive amount of PGH2, the
substrate for PGDS, so the PGDS step now became the rate-limiting step under these
conditions, leading to the larger quantities of PGD2 in the TG mice than in the WT
mice and ultimately to an increased amount of SWS in the TG mice.
In order to prove this interpretation, we then measured the amount of PGD2
in the brains of WT and TG mice before and after the tail clipping.43 The amount
of PGD2 in the brains of TG mice increased sharply and significantly for about
3 h after the tail clipping, and then started to decrease thereafter and returned to
normal after about 6 h. The time course of the changes in the amount of PGD2 in
the TG mice brain was essentially the same as that in the amount of SWS. In
contrast, the PGD2 content in the brains of WT controls after clipping remained
essentially the same as before it. Thus, it seems reasonable to conclude that the
increase in SWS in the TG mice after tail clipping was probably due to the induction
of COX II, or possibly to some other rate-limiting enzyme in the up-stream in
response to the pain stimulus, resulting in the increased level of PGH2 in the brain,
which prostanoid was then converted to PGD2 by the excessive amount of PGDS
in these TG mice.
These results are consistent with the notion that PGD2 is an endogenous sleep
substance in the brain of mice and is primarily involved in the induction and
maintenance of NREM sleep. Furthermore, they indicate that the PGDS gene is the
first gene to be implicated in the homeostatic control of NREM sleep.

Wild type KO
Amount of NREM sleep 100
after SD
after SD
(min/2hr)
** ** *
50 *
SD SD
baseline baseline
0
8:00 20:00 8:00 8:00 20:00 8:00
Clock time
FIGURE 4.4 Effect of SD on NREM sleep in WT and PGDS-KO mice. * P<0.05 ** P<0.01
Compared with the baseline day by the paired t test.
KO MICE FOR PGDS AND DPR

PGDS KO mice were also generated.44 The circadian profiles of NREM and REM
sleep in these KO mice were similar if not identical to those of the corresponding
WT control mice. Furthermore, the duration and the episode number of each vigi-
lance state in WT and KO mice were almost the same, indicating that the lack of
PGDS does not seem to affect the phenotype.
Because sleep is controlled as a function of prior wakefulness and sleep pressure
increases during waking or sleep deprivation (SD), the effect of SD for 6 h in the
late phase of the light period between 2 PM and 8:00 PM on NREM and REM sleep
in WT and KO mice was examined. As shown in Figure 4.4, a strong rebound of
NREM sleep (approximately 43% increase) was observed after SD in the WT mice.
The rebound in KO mice after SD was only approximately 10%, indicating that
endogenously produced PGD2 by PGDS is involved in the homeostasis of NREM
sleep after SD. This assumption is further supported by the observation that the
PGD2 content in the brains of the WT mice after SD was approximately twofold
higher than that before SD, whereas the amount of PGD2 in the brains of KO mice
was essentially unchanged after SD.45
The cDNA for the DPR has been isolated and identified as a 7-transmembrane
GS protein-coupled rhodopsin-type receptor.46 Again the basal circadian sleep pro-
files of WT and DPR KO mice were essentially identical. Infusion of PGD2 into the
lateral ventricle of the WT mice induced a dose-dependent increase in the amount
of NREM sleep as in the case of rats, but essentially no increase was observed in
the DPR KO mice.27 The extracellular adenosine level in the subarachnoid space of
the basal forebrain, the DPR-enriched zone, was then determined after PGD2 per-
fusion through a microdialysis probe. In the WT mice, the adenosine level increased
more than 100% after PGD2 perfusion of 400 pmol/min but did not increase at all

in DPR KO mice after PGD2 perfusion or in either WT or KO mice after the vehicle
perfusion. These results are in good agreement with our previous observation that
the activation of DPRs in the arachnoid trabecular cells of the basal forebrain in rats
triggers a local increase in the extracellular adenosine level.
A2AR KO MICE
A2AR KO mice were generated in a study on the relationship between A2AR and the
D2 dopamine receptor.47 The A2AR agonist CGS21680, at doses of 0.04, 0.2, 1 and
5 pmol/min dose-dependently increased NREM sleep in WT mice after infusion into
the lateral ventricle, but not at all in A2AR KO mice, as compared with the value for
the baseline day. In contrast, an A1 agonist, cyclopentyladenosine, did not affect sleep
profiles in WT mice at a dose of 1 pmol/min and increased NREM sleep only slightly
at a dose of 5 pmol/min.33 These results suggested that activation of mainly the A2AR
was involved in sleep promotion effect of adenosine. When PGD2 was infused into
the lateral ventricle of A2AR KO and WT mice, it dose-dependently induced sleep
during 6-h PGD2 infusion and 4-h post-dosing. PGD2 at doses of 10 and 50 pmol/min
increased NREM sleep in WT mice by 35% and 90.6%, and in A2AR KO mice by
only 5.6% and 38.1%, respectively, as compared with the baseline-day value.34
These results indicate that PGD2 exerted its somnogenic effect in a manner at
least partially dependent on the A2AR system, somewhat analogous to the interaction
between A2AR and D2 dopamine receptors.47 Alternatively, PGD2 may directly acti-
vate sleep-active neurons as previously shown by in vivo experiments, in which
PGD2 and E2 were applied ionophoretically on to various neurons in the POA/anterior
hypothalamus of unanesthetized rats.39 PGD2 had an excitatory effect on about one-
third of the sleep neurons and approximately the same percentage of wake-neurons
were excited by PGE2. As mentioned earlier, the VLPO containing a dense population
of sleep-active neurons is located in close proximity to the inner surface of the
subarachnoid space and, therefore, it is possible that PGD2 may activate some of
these neurons directly.
H1R KO MICE AND EP4 RECEPTOR

For exploration of the neural mechanisms involved in the PGE2-induced wakefulness
in rats, the effect of PGE2 on the activity of the histaminergic system and the
involvement of PGE2 receptor subtypes in the response were examined. The TMN
of the posterior hypothalamus is the sole source of histaminergic innervation of the
mammalian CNS, and this histaminergic system is considered to play a central role
in mediating wakefulness.
PGE2 perfusion of the TMN significantly increased both synthesis and release
of histamine. Among the agonists of the four distinct subtypes of PGE2 receptors
(EP1-4) tested, only the EP4 receptor-agonist (ONO-AE1-329) mimicked the excita-
tory effect of PGE2. In situ hybridization revealed that EP4 receptor mRNA was
expressed in the TMN region. Furthermore, EP4 agonist perfusion of the TMN
induced wakefulness. These findings thus indicate that PGE2 induced wakefulness
through activation of the histaminergic system via EP4 receptors.48

Other activators of the histaminergic system may also be involved in wakeful-

ness. The neuropeptides orexin (hipocretin) A and B were isolated from rat hypo-
thalamic extracts and identified, and a mutation in the orexin-2 receptor gene was
found to be associated with canine narcolepsy. Mice lacking the orexin peptide
display increases in REM and NREM sleep and a decrease in awake time during
the active period of normal rodents. The exact role of orexin in physiological sleep
and the mechanism involved have not yet been clarified.
Orexin neurons are exclusively localized in the lateral hypothalamic area and
project their fibers to the entire central nervous system, including the TMN, which
is enriched in orexin-2 receptors. Perfusion of orexin A (5–25 pmol/min) for 1 h
into the TMN of rats through a microdialysis probe promptly increased wakefulness,
concomitant with a reduction in REM and NREM sleep.49 Microdialysis studies
showed that orexin A increased histamine release from both the medial preoptic area
and the frontal cortex by approximately twofold over the baseline in a dose-
dependent manner. Infusion of orexin A (1.5 pmol/min) for 6 h into the lateral
ventricle of mice produced a significant increase in wakefulness during the 8 h after
the start of infusion to the same level seen during the active period in WT mice;
however in H1R KO mice no effect of orexin infusion was observed under the same
conditions. These results indicate that orexin is a potent waking substance acting
upon its receptor in the TMN and that the arousal effect of orexin A depends on the
histaminergic neurotransmission mediated by H1R.49
Molecular genetic aspects of our publications up to the end of 2002, mentioned
in this section were reviewed previously.50
SUMMARY
The concept of humoral, rather than neural, regulation of sleep dates as far back as
to almost 100 years ago. Kuniomi Ishimori and Henri Piéron independently and
concurrently took samples of the CSF of sleep-deprived dogs and infused them into
the brain of normal dogs. The recipient dogs soon fell asleep. Thus these two authors
became the first to demonstrate the presence of endogenous sleep-promoting sub-
stances, but the chemical nature of their sleep substances was not identified. Although
it is no longer possible to determine their chemical structure, available evidence
indicates PGD2 as a most plausible candidate. During the next 90 years or so, nearly
50 endogenous sleep substances were reported by numerous investigators to be
present in the brain, CSF, and other organs and tissues of mammals, although their
physiological relevance has remained uncertain in most instances.51
This review has briefly summarized the highlights of the prostaglandin and sleep
paradigm, the study of which has been carried out in my own and other laboratories
over the past 20 years.
Based upon results obtained by biochemical, physiological, and molecular bio-
logical studies, the following tentative conclusions have been drawn as a working
hypothesis for future studies:
• PGD2 and E2 are endogenous sleep and wake substances, respectively,

involved in the regulation of sleep and wakefulness under physiological

conditions not only in rodents but also in monkeys and humans and
possibly other mammals as well.
• PGD2 is produced by L-PGDS mainly present in the membrane system
surrounding the brain, secreted into the CSF, and is bound to its receptor,
DPR, also present in the outer surface of the rostral basal forebrain.
• The above experimental results strongly indicate the presence of a hitherto
unknown signal transduction system in the CNS of mammals. During the
past several decades, the mechanism of signal transduction has been
extensively studied by a number of investigators at the cellular level. These
studies indicated that most, if not all hormones, cytokines, and neurotrans-
mitters do not penetrate the cell membrane. Instead they are bound to
specific receptors on the cell surface, and the signals are then transmitted
through these receptors via so-called second messengers such as cyclic
AMP, Ca2+, and so forth. The mechanisms underlying the sleep regulation
by PGD2 are somewhat reminiscent of the signal transduction mechanisms
at the cellular level; namely, PGD2 is bound to its receptor on the surface
of the meninges, which binding is followed by the transduction via ade-
nosine through the adenosine A2A receptor. This signal is transmitted
across the leptomeninges into the brain parenchyma into the VLPO, a
putative sleep center, and further to TMN, a putative wake center.
• More recent studies with PGDS- and DPR-KO mice and others reveal
that the PGD2 system plays a crucial role in the homeostatic regulation
of NREM sleep.
We have thus witnessed a significant progress in sleep research on the humoral

mechanisms of sleep regulation and opened up a new frontier by elucidating the
interplay between the humoral regulation and the neural network. Obviously many
more important questions remain to be answered. Hopefully the fruits of our studies
described herein will provide a basis for further studies to solve the remaining
formidable problems pertaining to the mystery of sleep.
ACKNOWLEDGMENTS
The author is indebted to Y. Urade, N. Eguchi, Z.-L. Huang, and L. Frye for their
help during the preparation of this manuscript and illustrations, and to N. Ueda for
secretarial assistance. He also wishes to express his deep gratitude to all collabora-
tors, past and present, on this project during the past 20 years.
The work from this laboratory has been supported mainly by grants-in-aid from
the Ministry of Health, Labor and Welfare of Japan; the Ministry of Education, Culture,
Sports, Science, and Technology of Japan; and the Osaka Bioscience Institute.
REFERENCES
1. Narumiya, S. et al., Prostaglandin D2 in rat brain, spinal cord and pituitary: basal
level and regional distribution, Life Sciences, 31, 2093, 1982.

2. Hiroshima, O. et al., Basal level of prostaglandin D2 in rat brain by a solid-phase

enzyme immunoassay, Prostaglandins, 32, 63, 1986.
3. Ogorochi, T. et al., Regional distribution of prostaglandins D2, E2, and F2a and related
enzymes in postmortem human brain, J. Neurochem., 43, 71, 1984.
4. Ueno, R. et al., Prostaglandin D2 induced sleep when microinjected into the preoptic
area of conscious rats, Biochem. Biophys. Res. Commun., 109, 576, 1982.
5. Honda, K. and Inoué, S., Establishment of a bioassay method for the sleep-promoting
substance, Rep. Inst. Med. Dent. Eng., 12, 81, 1978.
6. Ueno, R. et al., Prostaglandin D2, a cerebral sleep-inducing substance in rats, Proc.
Natl. Acad. Sci. USA, 80, 1735, 1983.
7. Pandy, H.P. et al. Concentration of prostaglandin D2 in cerebrospinal fluid exhibits a
circadian alteration in conscious rats, Biochem. Mol. Biol. Int., 37, 431, 1995.
8. Ram, A. et al., CSF levels of prostaglandins, especially the level of prostaglandin D2,
are correlated with increasing propensity towards sleep in rats, Brain Res., 751, 81, 1997.
9. Hayaishi, O., Sleep–wake regulation by prostaglandins D2 and E2, J. Biol. Chem.,
263, 14593, 1988.
10. Onoe, H. et al., Prostaglandin D2, a cerebral sleep-inducing substance in monkeys,
Proc. Natl. Acad. Sci. USA, 85, 4082, 1988.
11. Roberts, II, J.L. et al., Increased production of prostaglandin D2 in patients with
systemic mastocytosis, New Engl. J. Med., 303, 1400, 1980.
12. Pentreath, V.W. et al., The somnogenic T lymphocyte suppressor prostaglandin D2 is
selectively elevated in cerebrospinal fluid of advanced sleeping sickness patients,
Trans. R. Soc. Trop. Med. Hyg., 84, 795, 1990.
13. Kubata, B. K., Identification of a novel prostaglandin F2a synthase in Trypanosoma
brucei, J. Exp. Med., 192, 1327, 2000.
14. Matsumura, H. et al., Awaking effect of PGE2 microinjected into the preoptic area
of rats, Brain Res., 444, 265, 1988.
15. Matumura, H. et al., Awaking effect of prostaglandin E2 in freely moving rats, Brain
Res., 481, 242, 1989.
16. Matumura, H. et al., Evidence that brain prostaglandin E2 is involved in physiological
sleep–wake regulation in rats, Proc. Natl. Acad. Sci. USA, 86, 5666, 1989.
17. Onoe, H. et al., Prostaglandin E2 exerts an awaking effect in the posterior hypothal-
amus at a site distinct from that mediating its febrile action in the anterior hypothal-
amus, J. Neurosi., 12, 2715, 1992.
18. Hayaishi, O., Molecular mechanisms of sleep–wake regulation: roles of prostaglan-
dins D2 and E2, FASEB J., 5, 2575, 1991.
19. Urade, Y. and Hayaishi, O., Prostaglandin D synthase: structure and function, Vitamins
and Hormones, 58, 89, 2000.
20. Urade, Y. and Hayaishi, O., Biochemical, structural, genetic physiological, and patho-
physiological features of lipocalin prostaglandin D synthase, Biochim. Biophys. Acta,
1482, 259, 2000.
21. Urade, Y. and Eguchi, N., Lipocalin-type and hematopoietic prostaglandin D syn-
thases as a novel example of functional convergence, prostaglandins, and other lipid
mediators, 68–69, 375, 2002.
22. Islam, F. et al., Inhibition of rat brain prostaglandin D synthase by inorganic seleno-
compounds, Arch. Biochem. Biophys., 289, 161, 1991.
23. Matsumura, H., Takahata, R., and Hayaishi, O., Inhibition of sleep in rats by inorganic
selenium compounds, inhibitors of prostaglandin D synthase, Proc. Natl. Acad. Sci.
USA, 88, 9046, 1991.

24. Takahata, R. et al., Intravenous administration of inorganic selenium compounds,

inhibitors of prostaglandin D synthase, inhibits sleep in freely moving rats, Brain
Res., 623, 65, 1993.
25. Huang, Z.-L., Urade, Y., and Hayaishi, O., unpublished observation.
26. Urade, Y. et al., Dominant expression of mRNA for prostaglandin D synthase in
leptomeninges, choroids plexus, and oligodendrocytes of the adult rat brain, Proc.
Natl. Acad. Sci. USA, 90, 9070, 1993.
27. Mizoguchi, A. et al., Dominant localization of prostaglandin D receptors on arachnoid
trabecular cells in mouse basal forebrain and their involvement in the regulation of
nonrapid eye movement sleep, Proc. Natl. Acad. Sci. USA, 98, 11674, 2001.
28. Zahn, M. et al., Purification and N-terminal sequence of b-trace, a protein abundant
in human cerebrospinal fluid, Neurosci. Lett., 154, 93, 1993.
29. Hoffmann, A. et al., Purification and chemical characterization of b-trace protein from
human cerebrospinal fluid: its identification as prostaglandin D synthase, J. Neuro-
chem., 61, 451, 1993.
30. Watanabe, K. et al., Identification of b-trace as prostaglandin D synthase, Biochem.
Biophys. Res. Commun., 203, 1110, 1994.
31. Yamashita, A., Watanabe, Y., and Hayaishi, O., Autoradiographic localization of a
binding protein(s) specific for prostaglandin D2 in rat brain, Proc. Natl. Acad. Sci.
USA, 80, 6113, 1983.
32. Matsumura, H. et al., Prostaglandin D2-sensitive, sleep-promoting zone defined in
the ventral surface of the rostral basal forebrain, Proc. Natl. Acad. Sci. USA, 91,
11998, 1994.
33. Satoh, S., Matumura, H., and Hayaishi, O., Involvement of adenosine A2A receptor
in sleep promotion, Eur. J. Pharmacol., 351, 155, 1998.
34. Qu, W.-M. et al., Adenosine A2A receptor deficiency attenuates the somnogenic effect
of prostaglandin D2, Sleep Biol. Rhythms, 2, S55, 2004.
35. Scammell, T. et al., Activation of ventrolateral preoptic neurons by the somnogen
prostaglandin D2, Proc. Natl. Acad. Sci. USA, 95, 7754, 1998.
36. Scammell, T.E. et al., An adenosine A2a agonist increases sleep and induces Fos in
ventrolateral preoptic neurons, Neuroscience, 107, 653, 2001.
37. Satoh, S. et al., Region-dependent difference in the sleep-promoting potency of an
adenosine A2A receptor agonist, Eur. J. Neurosci., 11, 1587, 1999.
38. Sherin, J.E. et al., Activation of ventrolateral preoptic neurons during sleep, Science,
271, 216, 1996.
39. Koyama, Y. and Hayaishi, O., Modulation by prostaglandins of activity of sleep-
related neurons in the preoptic/anterior hypothalamic areas in rats, Brain Res. Bull.,
33, 367, 1994.
40. Hayaishi, O., Molecular mechanisms of sleep-wake regulation: a role of prostaglandin
D2, Philos. Trans. R. Soc. Lond. B. Biol. Sci., 355, 275, 2000.
41. Hayaishi, O. and Urade, Y., Prostalglandin D2 in sleep-wake regulation: recent
progress and perspectives, Neuroscientist, 8, 12, 2002.
42. Hayaishi, O., Unraveling the enigma of sleep — molecular mechanisms of
sleep–wake regulation, in Oxygen and Life — Oxygenases, Oxidases, and Lipid
Mediators, Ishimura, Y. et al., Eds., Elsevier, Amsterdam, 2002, 503.
43. Pinzar, E. et al., Prostaglandin D synthase gene is involved in the regulation of
nonrapid eye movement sleep, Proc. Natl. Acad. Sci. USA, 97, 4903, 2000.
44. Eguchi, N. et al., Lack of tactile pain (allodynia) in lipocalin-type prostaglandin D
synthase-deficient mice, Proc. Natl. Acad. Sci. USA, 96, 726, 1999.

45. Eguchi, N. et al., Sleep in transgenic and gene-knockout mice for lipocalin-type
prostaglandin D synthase, in Oxygen and Life — Oxygenases, Oxidases, and Lipid
Mediators, Ishimura, Y., Eds., Elsevier, Amsterdam, 2002, 429.
46. Hirata et al., Molecular characterization of a mouse prostaglandin D receptor and
functional expression of the cloned gene, Proc. Natl. Acad. Sci. USA, 91, 11192, 1994.
47. Chen, J. F. et al., The role of the D2 dopamine receptor (D2R) in A2A adenosine
receptor (A2AR)-mediated behavioral and cellular responses as revealed by A2A and
D2 receptor knockout mice, Proc. Natl. Acad. Sci. USA, 98, 1970, 2001.
48. Huang, Z.-L. et al., Prostaglandin E2 activates the histaminergic system via EP4
receptor to induce wakefulness in rats, J. Neurosci., 23, 5975, 2003.
49. Huang, Z.-L. et al., Arousal effect of orexin A depends on activation of the histamin-
ergic system, Proc. Natl. Acad. Sci., USA, 98, 9965, 2001.
50. Hayaihsi, O., Molecular genetic studies on sleep-wake regulation, with special empha-
sis on the prostaglandin D2 system, J. Appl. Physiol., 92, 863, 2002.
51. Inoué, S., Biology of Sleep Substances, CRC Press, Boca Raton, FL, 1989.

5 The Network
Responsible for
Paradoxical Sleep Onset
and Maintenance: A New
Theory Based on the
Head-Restrained Rat
Model
Pierre-Hervé Luppi, Romuald Boissard,
Damien Gervasoni, Laure Verret,
Romain Goutagny, Christelle Peyron,
Denise Salvert, Lucienne Léger, Bruno Barbagli,
and Patrice Fort
CONTENTS
Introduction: The Neuronal Network Responsible for Paradoxical Sleep

before the Head-Restrained Rat Model
The PS Neuronal Network Identified with the Head-Restrained Rat Model
Identification and Pharmacology of a Pontine PS-Inducing
Structure in Rats
Efferents of the PS-On Neurons from the SLD
GABAergic and Non-GABAergic Inputs to the SLD PS-On Neurons
Evidence That GABA Is Responsible for the Inactivation of
Monoaminergic Neurons during PS
Localization of the GABAergic Neurons Responsible for the Tonic
Inhibition of Monoaminergic Neurons during PS
Conclusion: A New Network Model for PS Onset and Maintenance
Acknowledgments
References
0-8493-1519-0/05/$0.00+$1.50

INTRODUCTION: THE NEURONAL NETWORK

RESPONSIBLE FOR PARADOXICAL SLEEP BEFORE THE
HEAD-RESTRAINED RAT MODEL
In the middle of the last century, a series of historical observations led to the
discovery of a sleep phase in humans and other mammals characterized by a cortical
activation and conspicuous rapid eye movements (REM) paradoxically associated
with a complete disappearance of the muscle tone.1–3 This phase of sleep, coined
paradoxical sleep (PS) or REM sleep was then shown to correlate with dream
activity.1,4 A large number of studies in cats later demonstrated that the dorsal part
of the pontine reticular formation plays a crucial role in PS onset and maintenance.
It was first shown that PS persists following decortication, cerebellar ablation, or
brain stem transections rostral to the pons. In contrast transection at the posterior
limit of the pons suppresses PS.5 It was later shown that electrolytic and chemical
lesions of the dorsal part of the pontis oralis (PnO) and caudalis (PnC) nuclei
suppress PS.6–10 Since then a large number of results showed that the neurons
responsible for PS are cholinergic or at least cholinoceptif and restricted to a small
area of the dorsal part of these nuclei.11
Jouvet and Michel12 were the first to demonstrate that cholinergic mechanisms
play a major role in PS generation because peripheral atropine administration
suppressed PS, whereas anticholinesterase compounds increase PS. Then George
et al.13 discovered that bilateral injections of carbachol, a cholinergic agonist, into
the PnO and PnC promote PS. It was later shown that PS is induced with the shortest
latency when carbachol is ejected in a small area of the dorsal PnO and PnC,14–18
named peri-locus coeruleus a (peri-LCa) by Sakai et al.19,20 Sakai and coworkers20–23
found that the great majority of the pontine neurons with a tonic activity specific
during PS were localized in the peri-LCa. They recently divided these neurons into
two populations23: The first population of neurons are located in the dorsal and
rostral peri-LCa. They are inhibited by carbachol, a cholinergic agonist and project
rostrally to the intralaminar thalamic nuclei of the thalamus, the posterior hypothal-
amus, and the basal forebrain. The second population of PS-on neurons is excited
by carbachol, distributed in all parts of the peri-LCa, and caudally project to the
nucleus reticularis magnocellularis (Mc) localized in the ventromedial bulbar retic-
ular formation19,20
Based on these and other results, it has been proposed that the first type of
neurons are cholinergic and responsible for the cortical activation during PS, whereas
the second type of neurons are noncholinergic, possibly glutamatergic, and generate
the muscle atonia observed during this sleep state via descending excitatory projec-
tion to glycinergic premotoneurons within the Mc.11,22–27 Supporting this hypothesis,
intracellular recordings of motoneurons combined with strychnine applications dem-
onstrated that glycine is responsible for the tonic hyperpolarization of the spinal,
hypoglossal and trigeminal motoneurons.24,28–30 Further, our anatomical data showed
that the great majority of the neurons in the peri-LCa projecting to the Mc are not
cholinergic.25 In addition glutamate release in the Mc increases specifically during
PS31 and injection of non-NMDA glutamate agonists in the Mc suppresses muscle
tone.32 Spinal-projecting PS-on neurons have been recorded in the Mc33,34 and cyto-

toxic lesions of this structure induced a decrease in PS quantities and an increase

in muscle tone during PS.35 Further, we have shown that the Mc contains a large
contingent of glycinergic neurons.26,27 Glycinergic neurons from the ventral gigan-
tocellular reticular nucleus (the rat equivalent of the Mc) directly project to spinal
motoneurons36 while those of the parvocellular and parvocellular alpha nuclei
directly project to the trigeminal motor nucleus.37,38 It has been shown in cats
following induction of PS by carbachol injections in the peri-LCa that Fos-labeled
cells in the Mc project to the trigeminal motor nucleus.39
Several hypotheses have been proposed concerning the mechanisms responsi-
ble for the activation of the PS-on neurons from the peri-LCa at the onset and
during PS. Hobson and McCarley40,41 and later on Sakai20 proposed that the acti-
vation of these neurons is due to an excitatory interaction between the PS-on and
a reciprocal inhibition with the monoaminergic neurons. This well-accepted
hypothesis was formulated following the findings that serotonergic neurons from
the raphe nuclei and noradrenergic neurons from the locus coeruleus cease firing
during PS.40,42–44 Supporting this theory, drugs enhancing serotonin and noradren-
ergic transmission in particular serotonin and norepinephrine reuptake blockers
suppress PS (review in References 11 and 45). However the sites where the
monoamines in particular serotonin exert their PS-suppressing effect have not been
unambiguously identified.
Applications of norepinephrine, epinephrine, or benoxathian (an a2 agonist)
into the peri-LCa inhibit PS, but that of serotonin has no effect.46–48 Norepinephrine
via a2-adrenoceptor inhibits the noncholinergic PS-on neurons but has no effect on
the cholinergic PS-on neurons from the peri-LCa, and serotonin has no effect on
both types of neurons.23 Monoamines could also act on PS-on neurons localized in
structures other than the peri-LCa such as the Mc25 or the pedunculopontine teg-
mental (PPT) and laterodorsal tegmental cholinergic nuclei (LDT).49 The PPT and
LDT have been indeed reported to contain PS-on neurons, although the great
majority of the neurons from these nuclei are tonically active both during waking
(W) and PS.50–53 In cats, a subset of these neurons is tonically active during W and
exhibit phasic burst discharge just prior to and during ponto-geniculo-occipital
(PGO) waves.54–57 Bilateral lesion of the LDT and PPT in cats leads to the disap-
pearance of the PGO waves.58 These neurons have been antidromically activated
following the stimulation of the intralaminar and lateral geniculate nuclei from the
thalamus.56,57
It has been shown in combining retrograde tracing with cholineacetyltransferase
immunostaining that LDT and PPT neurons projecting to these nuclei are cholin-
ergic.56,59,60 These results indicate that cholinergic neurons from the LDT and PPT
play an important role in the activation of the cortex during W and PS and are
responsible in cats for the genesis of the PGO waves. In this species serotonin or
norepinephrine depletion induce continuous PGO waves, suggesting that the
monoamines are inhibitory on the PGO-on cholinergic neurons, but in vivo and in
vitro pharmacological studies reported excitatory, inhibitory, or no effect of monoam-
ines on cholinergic neurons from the LDT and PPT (review in Reference 61).
Monoamines have also been proposed to contribute to the muscle atonia of PS
by a disfacilitation of motoneurons, and it has been shown that serotonin and

norepinephrine are excitatory on motoneurons.62,63 The role of monoamines might

be particularly important for the hypoglossal motoneurons because the application
of serotonin on hypoglossal motoneurons during PS reverses the atonia of the upper
airway musculature.64
According to the classical reciprocal interaction model,20,41 the cessation of firing
of the noradrenergic and serotonergic neurons at the onset of PS is the result of
active PS-specific inhibitory processes originating from PS-on cells. These neurons
were first hypothesized to be cholinergic and localized in the peri-LCa , LDT and
PPT, but acetylcholine excites LC noradrenergic neurons and is only weakly inhib-
itory on serotonergic DRN neurons.65,66 It has therefore been suggested that they
might use GABA or glycine, rather than acetylcholine, as an inhibitory neurotrans-
mitter.67,68 It has recently been shown that application of non-NMDA agonists such
as kainic acid or that of bicuculline, a GABAA antagonist into the peri-LCa region
induces a strong increase in PS quantities.16,69–72 Altogether these recent results
strongly pointed out that in addition to cholinergic and monoaminergic neurons,
populations of glutamatergic and GABAergic neurons might play a crucial role in
the onset and maintenance of PS.
To test this hypothesis, we developed a new model combining single-unit
recordings, precise and limited local pharmacology by micro-iontophoresis in unan-
aesthetized head-restrained rats, and anterograde and retrograde tracing combined
with Fos and neurochemical identification of labeled cells.73–77 We decided to work
with rats instead of cats because of the cost, the availability of the majority of
recent data, and more limited ethical concerns. With this choice we were neverthe-
less facing the challenge to identify in rats the structures responsible for PS onset
and maintenance already identified in cats, in particular the peri-LCa and the Mc.
Data on the neuronal network responsible for PS onset and maintenance was lacking
in this species. We succeeded in identifying this network in rats and present our
results below in a first part. In a second part we present our results with the same
rat model showing that monoaminergic neurons are silenced during PS by GABAer-
gic neurons from the dorsal paragigantocellular reticular nucleus. Finally, based on
all these results, we propose a new theory on the network responsible for PS onset
and maintenance.
THE PS NEURONAL NETWORK IDENTIFIED WITH

THE HEAD-RESTRAINED RAT MODEL
IDENTIFICATION AND PHARMACOLOGY OF A PONTINE PS-INDUCING
STRUCTURE IN RATS
We found that a long-lasting PS-like hypersomnia can be pharmacologically induced
with a short latency in head-restrained rats by iontophoretic applications of bicu-
culline or gabazine, two GABAA antagonists specifically into a very small area of
the dorso-lateral pontine tegmentum.76 We also recorded neurons in this region
specifically active during PS and excited by bicuculline or gabazine iontophoresis78
(Figure 5.1). This region has been denominated the sublaterodorsal nucleus (SLD)

A gabazine
Spikes/sec Kainic acid Kainic acid
300
PS-like
200
100
seconds
B
Spikes/s
carbachol PS Kainic acid

200
100
seconds
FIGURE 5.1 Pharmacology of a PS-on neuron from the SLD. (A) Effect of the iontophoretic
application of gabazine (100 nA, 265 sec) or kainic acid (50 nA, 3 or 5 sec) on the activity
of a PS-on SLD neuron. Application of kainic acid during SWS induced an excitation of the
neuron. Long gabazine application induced a very strong excitation of the neuron followed
by the onset of a PS-like phase characterized by theta activity on the EEG and a complete
atonia on the EMG. (B) On the same neuron as in A, the application of carbachol (50 nA,
30 sec) during SWS induced no change in the firing rate of the neuron, and that of kainic
acid induced a strong excitation.

by Swanson.79 It approximately corresponds to the dorsal subcoeruleus nucleus in

Paxinos and Watson atlas80 and seems to be the equivalent in rats of the cat peri-
LCa. Our results have been recently reproduced in freely moving rats81 and are in
agreement with a recent study in cats showing that pressure injection of bicuculline
and to a lesser extent phaclofen (a GABAB antagonist) in the dorsal portion of the
nucleus pontis oralis (which roughly corresponds to the peri-LCa) induces a strong
increase in PS quantities with short latencies, whereas the application of muscimol
(a GABAA agonist) or baclofen (a GABAB agonist) induced W.70,71 These and our
data imply that the onset of PS-on neurons of the SLD is mainly due to the removal
of a tonic GABAergic input present during W and SWS.
We also showed that kainic acid (a glutamate agonist) iontophoretic application
into the SLD induces an activation of PS-on neurons (Figure 5.1) and is consistently
associated with a transient PS-like state followed by W with an increase in muscle
activity.76 The PS-like state induced by bicuculline was reversed by the application
of kynurenate.76 In agreement with our results, it has been shown in cats that the
administration of kainic acid in the peri-LCa using microdialysis induces a PS-like
state.69 Altogether these results suggest that PS-on neurons in the SLD receive a
tonic glutamatergic input during all sleep-waking states. Upon removal of the tonic
GABAergic input at the onset of PS, the unmasked glutamatergic input would be
responsible of the tonic activity on the PS-on SLD neurons during PS.
Neurons responsible for PS onset and maintenance seem to be clustered in a
sphere of tissue smaller than 1 mm3 centered on the SLD. Indeed Fos labeled neurons
in the site of injection following 90-min bicuculline applications occupied less than
1mm3. Further, bicuculline, gabazine and kainic acid ejections just 500 mm away
from the positive sites gave rise to W with increased muscle activity.76 The neuro-
chemical nature of the PS-on neurons from the SLD remains to be determined. It
is unlikely that they are cholinergic since in our Fos staining experiments, only
occasional Fos positive neurons in the injection site were immunoreactive to cho-
lineacetyltransferase. Rather, it is likely that at least part of these neurons are
glutamatergic, because our Fos experiments highly suggest that they provide an
excitatory pathway to glycinergic neurons from the medullary reticular formation
and glutamate release is increased at this level during PS.31
Bicuculline or gabazine administration into the SLD induced a PS-like state
characterized by the presence of muscle atonia, EEG activation with or without theta
activity, and nonreactivity to surrounding stimuli but the absence of ocular move-
ments or penile erections.76 The absence of these events may be due to the fact that
SLD neurons responsible for phasic activities during PS do not receive tonic
glutamatergic and GABAergic inputs during W and SWS. It is also possible that
the SLD neurons are not involved in all PS-related phenomena. Cholinergic and
non-cholinergic neurons from the LDT or neurons from other structures could be
responsible for the missing phasic events. Supporting this hypothesis, neurons spe-
cifically active during PS-related erections have been recently recorded in the LDT,82
and an increased number of Fos positive neurons has been observed in the LDT
following PS rebound.83 In our experiments, bicuculline injections into the LDT
induce W in apparent contradiction with this hypothesis. This finding may be

explained by the fact that the majority of the LDT neurons exhibit a high firing rate
both during W and PS, whereas only a minority are specifically active during PS.84,85
We found that carbachol iontophoresis into the rat SLD induced a W state with
increased muscle and that SLD PS-on neurons do not respond to carbachol
iontophoresis76 (Figure 5.1). These results indicate important species differences
between rats and cats in the pharmacological sensitivity of the pontine PS-on
neurons. In agreement with our results, following carbachol administration into
the rat pontine reticular formation, the enhancement of PS was of small
magnitude86–89 or not reliably obtained.90 The increase in PS was less than 100%,
compared to the 300% increase obtained in cats, the first PS episode appeared at
least 50 min after the carbachol injection, and the duration of the episodes was
similar to natural, spontaneous PS. In cats, however, PS is induced almost imme-
diately after the injection and the episodes last longer than in control PS. The
effective sites in rats were widely distributed in the pontine reticular formation. In
contrast, the most effective site in cats is the peri-LCa that corresponds to the rat
SLD.15 The absence of effect of carbachol ejection in the SLD does not rule out
a role of cholinergic processes in PS onset and maintenance in the rat. It is indeed
possible that PS-on neurons in the SLD have muscarinic or nicotinic receptors but
that the activation of these receptors by carbachol is unable to modify their activity
due to the strong GABAergic tonic inhibition revealed in our study. Supporting
this hypothesis it has been shown that carbachol applications in the region of the
SLD are able to induce with a short latency a long period of atonia in anesthetized
or decerebrate rats models91,92 in which the GABAergic inhibitory tone on SLD
neurons could be decreased or even absent. Another possibility is that the cholin-
ergic system plays an important role in PS in rats via an action on populations of
neurons controlling PS localized in other pontine regions than the SLD. An increase
in the number of cholinergic neurons containing Fos has been indeed observed
following PS recovery,83 and a strong enhancement in PS quantities was found
following carbachol pressure ejection in the most ventral part of the oral pontine
reticular formation.18,93
EFFERENTS OF THE PS-ON NEURONS FROM THE SLD

We recently attempted to determine the efferents of the SLD responsible for the
cortical activation and the muscle atonia seen following bicuculline, gabazine, or
kainic acid applications in the SLD combining Fos and glycine immunostainings
with PHA-L anterograde tracing.76 Following PHA-L injection in the SLD, a large
number of anterogradely labeled fibers was observed in the medullary reticular
nuclei. In combining Fos, glycine, and PHA-L stainings in rats that received long
bicuculline or gabazine applications in the SLD before perfusion, we observed in
the caudal nucleus raphe magnus, ventral gigantocellular (rat equivalent of the cat
Mc), parvocellular, and parvocellular alpha reticular nuclei a large number of anter-
ogradely labeled fibers in the vicinity of Fos labeled cells and numerous Fos and
glycine-immunoreactive double-labeled neurons.76 These results confirm and extend
previous results demonstrating that PS-on neurons from the peri-LCa provide a

strong excitatory putatively glutamatergic projection to glycinergic neurons from

the Mc.22
These results and ours indicate that glycinergic neurons from the ventral gigan-
tocellular nucleus and the parvocellular and parvocellular alpha reticular nuclei are
responsible for the hyperpolarization of spinal and cranial motoneurons during PS,
respectively. The role of the glycinergic neurons from the caudal nucleus raphe
magnus, active during PS, is less clear. They could be responsible for the known
attenuation of ascending sensory neurotransmission during PS. It has been shown
that glycine mediates the inhibition during PS of ascending sensory transmission
via Clarke’s column dorsal spinocerebellar tract neurons, and raphe magnus neurons
are known to control sensory input via direct projections to sensory neurons from
the dorsal horn of the spinal cord.94
In addition to descending projections to the medullary reticular formation, we
also found that the SLD projects to rostral structures like the intralaminar thalamic
nuclei, intrafascicular thalamic nucleus and zona incerta.76 In agreement with Jones
and Yang,95 these ascending projections could be responsible for the cortical acti-
vation obtained during the administration of bicuculline, gabazine, or kainic acid
into the SLD and more generally than seen during PS. Fos labeled neurons were
observed in these structures following long lasting bicuculline or gabazine ejections
into the SLD.76
GABAERGIC AND NON-GABAERGIC INPUTS TO THE SLD

PS-ON NEURONS
Combining retrograde tracing with cholera toxin B subunit (CTb) and GAD immu-
nostaining, we recently tried to identify the GABAergic neurons potentially respon-
sible for the inhibition during W and SWS of the PS-on neurons localized in the
SLD and the glutamatergic neurons responsible for their constant excitation across
all vigilance states.77 Our results suggest that the GABAergic innervation of SLD
neurons arises both from interneurons and distant neurons located in the pontine
and deep mesencephalic reticular nuclei and to a minor extent hypothalamic and
medullary structures.77 These results are in agreement with previous studies indicat-
ing that the GABAergic neurons responsible for the tonic inhibition during W and
SWS of the PS-on neurons from the SLD could be within the SLD itself or in the
pontine and deep mesencephalic reticular nuclei. A recent study by Xi et al.70
suggested that GABAergic interneurons might be the best candidates for the inhi-
bition of PS-on SLD neurons. They found in cats that administration of antisense
oligonucleotides against glutamic acid decarboxylase (GAD) mRNA in the nucleus
pontis oralis (NPO), a region corresponding to the SLD, produces a significant
decrease in W and an increase in PS. Maloney et al.96 found in rats that the number
of Fos expressing GABAergic neurons in the rostral pontine reticular nucleus
decreased following PS rebound, suggesting that GABAergic neurons from this
structure are active during W and SWS and inactive during PS.
It has been shown in cats97,98 and rats78 that muscimol injections in the most
ventrolateral part of the periaqueductal gray and in the region of the deep mesen-
cephalic reticular nucleus just ventral to it induce a strong increase in PS quantities

Delta
(1.5-4Hz)
Theta
(4.5-8.5Hz)
Sigma Aq
(9-14Hz)
Gamma DRN
(30-50Hz)
4 1000µm
Muscle
Hypnogram Mus 90min

PS
SWS
W
30 60 90 120 150 180 210 min
FIGURE 5.2 Effect of a muscimol application in the region of the deep mesencephalic
reticular nucleus just ventral to the periaqueductal gray. (A) Power spectrum analysis and
hypnogram obtained in a rat before and during the iontophoretic application of muscimol
(100 nA, 90 min). The application of muscimol induced a 250% increase in PS quantities as
compared to Nacl. (B) The CTb injection site obtained by iontophoresis in the positive site
of muscimol ejection with another barrel of the electrode.
(Figure 5.2). More recently Sakai et al.22 reported that muscimol applications limited
to the region of the deep mesencephalic reticular nucleus just ventral to the periaq-
ueductal gray induced an increase in PS quantities but those in the ventrolateral
periaqueductal gray had no effect. We reported a strong non-GABAergic projection
to the SLD from the ventrolateral periaqueductal gray and a mixed GABAergic and
non-GABAergic projection from the region of the deep mesencephalic reticular
nucleus just ventral to the periaqueductal gray.77 Altogether from these results, we
propose that GABAergic neurons located in the dorsal part of the deep mesencephalic
reticular nucleus, the pontine reticular nucleus, and in the SLD itself project to and
directly inhibit the PS-on neurons from the SLD specifically during W and SWS.
Because acetylcholine was thought to be the main neurotransmitter responsible
for the activation of the pontine PS-on neurons, the cholinergic input to the cat peri-
LCa and the pontine reticular nucleus has attracted a lot of attention and several
studies reported that it arises from the LDT and the PPT.99–103 Sakai104 found that
the peri-LCa receives additional cholinergic inputs from the magnocellular, parvo-
cellular and lateral paragigantocellular reticular nuclei. In our study, we only found
a small non-GABAergic projection from the LDT and PPT. In agreement with our
pharmacological results, these results suggest that the LDT and PPT cholinergic
input to the SLD is rather a minor one.77

Following a section between the pons and the medulla in the cat, a state of PS
cannot be recorded on either side of the section.105,106 From these results it has been
hypothesized that, in addition to the descending projections from the SLD to the
medullary reticular nuclei responsible for muscle atonia, reciprocal ascending pro-
jections are crucial for generating PS. We found strong ascending projections to the
SLD from the parvocellular and lateral paragigantocellular reticular nuclei and only
small projections from the dorsal paragigantocellular and magnocellular reticular
nuclei.77 These results suggest that these two later structures thought to contain
respectively the GABA and the glycinergic neurons responsible for the inhibition
of LC noradrenergic neurons107 and the motoneurons76 during PS, play minor roles
in the control of the PS-on neurons from the SLD. In contrast, the parvocellular,
and lateral paragigantocellular reticular nuclei could contain neurons controlling PS-
on neurons from the SLD.
In the classical reciprocal interaction model, serotonin and norepinephrine are
responsible for the inhibition of the PS-on neurons during W and SWS. Supporting
this hypothesis, it has been previously shown in rats that the pontine reticular nucleus
receives noradrenergic inputs from the LC and A5 and A7 noradrenergic groups and
serotoninergic inputs from all pontine and medullary raphe nuclei and the B9
group.102 In general agreement with these results, we observed a small number of
retrogradely labelled neurons in the dorsal raphe and locus coeruleus nuclei follow-
ing CTb injections into the SLD.77 Additional studies in rats are now necessary to
determine the exact role of the monoamines in the regulation of the activity of the
PS-on neurons from the SLD.
From our pharmacological results76 we also hypothesized that the PS-on neurons
from the SLD are constantly excited across all vigilance states by a glutamatergic
input. The majority of the glutamatergic neurons providing a constant excitatory
input to SLD PS-on neurons should be located in the brainstem although forebrain
glutamatergic neurons could also participate. The structures responsible for the onset
and maintenance of PS are indeed restricted to the brainstem.5 Such glutamatergic
inputs can arise from the numerous non-GABAergic neurons projecting to the SLD
localized in the ventrolateral periaqueductal gray, the mesencephalic, pontine, and
parvocellular reticular nuclei. Additional studies are necessary to determine which
one of these structures provides a glutamatergic input to the SLD PS-on neurons.
The histochemical nature and role of the strong non-GABAergic afferents to
the SLD from the primary motor area of the frontal cortex, the bed nucleus of the
stria terminalis, and central nucleus of the amygdala remains to be identified.
Maquet et al.108 found that regional cerebral blood flow is positively correlated with
PS in the amygdaloid complex, and electrical stimulation of the central nucleus of
the amygdala increases the frequency of pontine waves recorded in or just dorsal
to the SLD during PS.109 From these and our results, it might be hypothesized that
the central nucleus of the amygdala and the functionally related bed nucleus of the
stria terminalis provide excitatory glutamatergic projections to PS-on neurons from
the SLD.
The substantial predominantly non-GABAergic projection to the SLD from the
lateral, perifornical and posterior hypothalamic areas could also play an important
role in PS homeostasis. Reversible inactivation of the lateral hypothalamic area by

muscimol (a GABAA agonist) applications induced a decrease in W and a disap-

pearance of PS,110 and neurons specifically active during PS or W have been recorded
in the same area.111,112 A number of recent studies indicate that two intermingled
populations of neurons localized in this region containing the hypocretins (orexins)
or melanin concentrating hormone (MCH) could be involved. It has first been shown
that narcolepsy, a sleep disorder characterized by excessive daytime sleepiness and
cataplexy, is due to the lack of hypocretin mRNA and peptides in humans113 or a
disruption of the hypocretin receptor 2 or its ligand in dogs and mice.114,115 It has
then been shown that intracerebroventricular infusion of hypocretin induce W while
that of MCH increases PS and, to a minor extent, SWS amounts.116,117
Combining Fos and MCH or hypocretin immunostainings, we recently dem-
onstrated that MCH but not hypocretin neurons are active during PS.116 Finally a
substantial number of MCH and hypocretin immunoreactive fibers have been
observed in the SLD.118,119 Altogether these data suggest that MCH and hypocretin
neurons might directly project to PS-on neurons from the SLD. However, it is
unlikely that their effects on PS is mediated by such projection. Indeed, hypocretins
and MCH are excitatory and inhibitory peptides, respectively, and it is therefore
more likely that hypocretins inhibit and MCH promotes PS respectively via an
excitation and an inhibition of PS-off neurons. They could act on the GABAergic
presumably PS-off SLD neurons but also on the GABAergic neurons located in
the deep mesencephalic and pontine reticular nucleus. They could also influence
PS via projections to the monoaminergic PS-off neurons from the locus coeruleus
and raphe nuclei that contain a large number of MCH and hypocretin immunore-
active fibers118,119 (see below).
EVIDENCE THAT GABA IS RESPONSIBLE FOR THE INACTIVATION OF

MONOAMINERGIC NEURONS DURING PS
In anesthetized rats iontophoretic applications of GABA or glycine strongly inhibit
LC and DRN neurons, and co-iontophoresis of bicuculline or strychnine (GABAA
and glycine antagonists, respectively), antagonize these effects.67,120,121 In vitro stud-
ies on slices using focal stimulation and bath-applied bicuculline and strychnine
revealed GABA- and glycine-mediated IPSPs in LC neurons, and GABA-mediated
IPSPs in DRN cells.122–125 In agreement with these results, GABA-and glycine-
immunoreactive varicose fibers as well as GABAA and glycine receptors have been
found in the rat LC and DRN.67,68,126–128 Based on these results, Jones67,68 and we
proposed that GABA or glycine might be responsible for the inhibition of monoam-
inergic neurons during both slow wave sleep (SWS) and PS. To test this hypothesis
we determine the effect of iontophoretic applications of bicuculline and gabazine
(two GABAA antagonists) and strychnine (a glycine antagonist) during W, SWS,
and PS on the activity of LC noradrenergic and DRN serotonergic cells in the head-
restrained unanesthetized rat.73–75
Iontophoretic application of bicuculline, gabazine, or strychnine during SWS or
PS induced a tonic firing in LC noradrenergic and DRN serotonergic neurons73–75
(Figure 5.3). Application of these antagonists during W induced a sustained increase
in discharge rate. These results indicate the existence of tonic GABA and glycinergic

FIGURE 5.3 Effect of gabazine iontophoresis during PS on a serotonergic neurons of the

DRN. The iontophoretic application of gabazine induced a reversible increase of the firing
rate of the neuron recorded. Iontophoretic application of 8OH-DPAT, a classical 5HT1A
agonist on this neuron induced a complete cessation of its activity highly suggesting its
serotonergic nature.
inputs to the LC and DRN that are active during all vigilance states. Importantly
we found that when the strychnine effect occurred during transitions between PS
and W, the discharge rate of the LC or DRN neurons increased at the onset of W.
In the same situation but after bicuculline administration, the discharge rate of a
given neuron was not increased at the transition between PS and W. These results
strongly suggest that the release of GABA but not that of glycine is responsible for
the inactivation of LC noradrenergic neurons and DRN serotonergic during PS.
At variance with our results, Levine and Jacobs129 found in cats that the ionto-
phoretic application of bicuculline reversed the typical suppression of neuronal
activity of DRN serotonergic neurons during SWS but not during PS. Sakai and
Crochet130 did not find in cats an effect of bicuculline microdialysis infusion on
DRN serotonergic neurons during PS and hypothesized that our results were due to
a nonspecific excitatory action of bicuculline. This is unlikely because we reproduced
the effect of bicuculline with gabazine, another specific GABAA antagonist (Figure
5.3). Our results are supported by those of Nitz and Siegel,131,132 who found in cats
with the microdialysis technique a significant increase in GABA release in the DRN
and LC during PS as compared to W and SWS and, in contrast, no detectable changes
in glycine concentrations. Based on these and our results, we suggest that during W
the LC and DRN cells are under a tonic GABAergic inhibition that increases during
SWS, and even further during PS, and that the increase in GABAergic inhibition is
responsible for the inactivation of these neurons during the sleep states. In contrast
the glycinergic tonic inhibition would be constant across the sleep-waking cycle and
thus control the general excitability of LC and DRN neurons.
Instead of GABA, Sakai and Crochet130 proposed that the cessation of activity
of the serotonergic neurons of the DRN is caused by a disfacilitation resulting from
the cessation of discharge of norepinephrine or histamine containing neurons. They
found that the cessation of discharge of presumed serotonergic DRN neurons during
PS is reversed by either histamine or phenylephrine, an a1-adrenergic agonist, while

FIGURE 5.4 Effect of prazosin iontophoresis during W on a serotonergic DRN neuron. The
application of prazosin (150 nA, 82 sec) during W induced a significant decrease in the
discharge rate of a serotonergic neuron from the DRN.
the application of a specific H1 histamine receptor antagonist or prazosin, a specific

a1-adrenoceptor antagonist suppressed the spontaneous discharge of DRN seroton-
ergic neurons during W and SWS. However, it was shown previously that systemic
administration of prazosin or WB4101 in freely moving cats did not block the activity
of DRN neurons.133 We also found in rats that the iontophoretic application of
norepinephrine increases the activity of DRN serotonergic neurons during W, SWS,
and PS, but that of prazosin did not completely suppress their tonic activity during
W (Figure 5.4).
We therefore propose that the inactivation of DRN serotonergic neurons and LC
noradrenergic neurons during PS is due to a tonic GABAergic inhibition. The tonic
activity of DRN serotonergic neurons during W would be due at least in part to
excitatory noradrenergic and histaminergic inputs, and that of LC noradrenergic
neurons would be mainly due to their intrinsic electrophysiological properties.123 It
must be noted that other neuroactive substances such as the hypocretins, MCH, and
serotonin might also play a role in the inactivation of monoaminergic neurons during
PS. We recently showed that hypocretin neurons are inactive during PS while MCH
neurons are strongly active.116 MCH- and hypocretin-containing fibers have been
observed in the monoaminergic nuclei,118,119 and hypocretins strongly excite hista-
minergic neurons from the tuberomammillary nucleus,117,134 serotonergic neurons
from the DRN,135,136 and noradrenergic neurons from the LC.137–139 MCH is known
to be an inhibitory peptide,140 but its effects on monoaminergic neurons remain to
be studied. It seems likely that an increase in MCH and a decrease in hypocretin
releases, respectively, could also contribute to the cessation of activity of monoam-
inergic neurons during PS.
Based on the measurement by voltammetry of 5HIAA, the metabolite of sero-
tonin, it has also been proposed that a dendritic release of serotonin could be
responsible for the cessation of activity of DRN serotonergic neurons during PS.141
However, other authors have shown by microdialysis that serotonin release in the
DRN decreases during PS (review in Reference 142).

LOCALIZATION OF THE GABAERGIC NEURONS RESPONSIBLE FOR THE TONIC

INHIBITION OF MONOAMINERGIC NEURONS DURING PS
Our results obtained with double-staining experiments indicate that the LC and
DRN receive GABAergic inputs from neurons located in a large number of distant
regions from the forebrain to the medulla.75,107 We observed a substantial number
of GAD-immunoreactive neurons in the preoptic area, the lateral hypothalamic area,
the mesencephalic and pontine periaqueductal gray, and the dorsal paragigantocel-
lular reticular nucleus that project to the LC and DRN.75,107 These results indicate
that the GABAergic innervation of these two monoaminergic nuclei arises from
multiple, distant GABAergic groups in addition to interneurons. They suggest that
the serotonergic neurons of the DRN and noradrenergic neurons of the LC could
be inhibited by multiple populations of GABAergic neurons located in different
structures, and raise the question of the functional significance of such redundancy.
One possibility is that only some of these GABAergic afferents are destined to the
serotonergic neurons of the DRN and the noradrenergic neurons of the LC. This
seems likely for the DRN, which is an heterogeneous structure, but not for the LC
which in rats contains nearly exclusively noradrenergic cells. Another possibility is
that some of these afferents are postsynaptic and the others presynaptic, but the
more likely explanation is that each of these afferents is active only under specific
physiological conditions.
Based on physiological and electrophysiological data (see above), we expect
that one or several of these GABAergic afferents are turned on specifically at the
onset of and during PS episodes and are responsible for the inhibition of brainstem
monoaminergic neurons during PS. Although it has recently been proposed that
GABAergic neurons located in the extended ventrolateral preoptic nucleus might
also be involved,143 previous results highly suggest that brainstem GABAergic
neurons are mostly involved. Indeed, it is well known that PS-like episodes occur
in pontine or decerebrate cats.144 It has recently been shown in decerebrate animals
that PS episodes induced by carbachol injections in the pons are still associated
with a cessation of activity of serotonergic neurons of the raphe obscurus and
pallidus nuclei.145
Among the brainstem GABAergic afferents revealed in our study, several are
common to the DRN and the LC and are therefore good candidates for this role.
We observed substantial GABAergic projections to the LC and DRN from the ventro-
lateral periaqueductal gray and the dorsal paragigantocellular nucleus.75,107 In agree-
ment with these results, local application of bicuculline blocked the dorsal paragi-
gantocellular-evoked inhibition of LC neurons,120 and focal iontophoretic application
of NMDA in the ventral periaqueductal gray induced bicuculline sensitive IPSPs in
DRN serotonergic neurons.146 The hypothesis that the GABAergic inhibition is
coming from neurons located in the periaqueductal gray is further supported by two
recent studies. Yamuy et al.147 showed that after a long period of PS induced by
pontine injection of carbachol, a large number of Fos positive cells are visible in
the DRN and a region lateral to it. Maloney et al.83 observed, after a PS rebound
induced by deprivation, an increase in Fos-positive GAD immunoreactive neurons
in the periaqueductal gray.

To directly determine among the GABAergic afferents to the LC, those active
during PS, we recently combined iontophoretic application of CTb in the LC with
Fos staining in rats deprived of PS, rats with enhanced PS during rebound after PS
deprivation, and control rats. Using this method we observed a large number of CTb
and Fos double-immunostained neurons in the dorsal paragigantocellular reticular
nucleus and a substantial number in the ventro-lateral periaqueductal gray and the
lateral paragigantocellular reticular nucleus.148 We propose that the GABAergic
neurons responsible for the inhibition of the LC noradrenergic neurons during PS
are mainly but not exclusively localized in the dorsal paragigantocellular reticular
nucleus. To further test this hypothesis, we recorded the spontaneous activity of
neurons from the dorsal paragigantocellular reticular nucleus across the sleep-waking
cycle in head-restrained rats. Neurons with an activity specific to PS (PS-on neurons)
were found within this nucleus,149 further supporting that it contains the GABAergic
neurons responsible for the cessation of activity of the noradrenergic neurons of the
LC during PS. This hypothesis is also supported by a recent study showing that
electrical stimulation of the area of the dorsal paragigantocellular reticular nucleus
induces an increase in PS quantities.150
CONCLUSION: A NEW NETWORK MODEL FOR PS

ONSET AND MAINTENANCE (FIGURE 5.5)
We propose that the onset and maintenance of PS is due to the activation of PS-on
glutamatergic neurons from the SLD. During W and SWS they would be hyperpo-
larized by tonic GABAergic inputs arising from GABAergic PS-off neurons local-
ized in the SLD itself and the deep mesencephalic and pontine reticular nuclei.
Noradrenergic and serotonergic PS-off neurons would also participate in the hyper-
polarization of SLD neurons particularly during W. The cessation of activity of the
monoaminergic neurons at the onset of and during PS would be due to an active
inhibition by PS-on GABAergic neurons localized in the dorsal paragigantocellular
reticular nucleus and the ventrolateral periaqueductal gray. Although the exact mech-
anism of the cessation of activity of the GABAergic PS-off neurons remains to be
identified, we propose that the GABAergic PS-on neurons inhibiting the monoam-
inergic neurons could, at the same time, inhibit the GABAergic PS-off neurons.
The activation of the SLD PS-on neurons at the onset of PS would be due to
the strong glutamatergic excitatory input present during all vigilance states blocked
during W and SWS by the inhibitory inputs from the GABAergic and monoaminergic
PS-off neurons. It would arise from one or several of the non-GABAergic afferents
to the SLD (e.g., the periaqueductal gray, the deep mesencephalic and pontine
reticular nuclei, and the parvocellular reticular nucleus).
Ascending SLD PS-on glutamatergic neurons would induce cortical activation
via their projections to intralaminar thalamic relay neurons in collaboration with
W/PS-on cholinergic and glutamatergic neurons from the LDT and PPT, mesen-
cephalic and pontine reticular nuclei, and the basal forebrain. Descending PS-on
glutamatergic SLD neurons would induce muscle atonia via their excitatory projec-
tions to glycinergic premotoneurons localized in the magnocellular and parvocellular
reticular nuclei.

W/SWS PeF, HLA PS PeF, HLA

Hcrt Hcrt
Pef/HLA Pef/HLA
MCH MCH
Thalamus Thalamus
EEG activation EEG activation
Glu Glu
SLD LC/DRN SLD LC/DRN

PAG, DPMe, PRN PAG, MRN, PRN
NA/5-HT NA/5-HT
Glu Glu
Glu Glu
DPMe, PRN, SLD
D DPMe, PRN, SLD
LDT/PPT GABA LDT/PPT GABA
Ach Ach
Mc Mc
Muscle atonia Muscle atonia
Gly Gly
vlPAG/DPGi vlPAG/DPGi
GABA GABA
PS-on PS-off PS-on PS-off
FIGURE 5.5 (See color insert following page 108.) Model of the network responsible for PS
onset and maintenance. The onset and maintenance of PS would result from the activation of PS-
on glutamatergic neurons from the SLD. The activation of these neurons would be due to the
removal of tonic inhibitions arising from the monoaminergic PS-off neurons and GABAergic PS-
off neurons localized in the SLD itself and the deep mesencephalic and pontine reticular nuclei.
The cessation of activity of the PS-off neurons would be due to a tonic inhibition issued from
GABAergic PS-on neurons localized in the dorsal paragigantocellular reticular nucleus and the
ventrolateral periaqueductal gray.
Abbreviations: DRN, dorsal raphe nucleus; 5-HT, serotonin; LC, locus coeruleus; NA,
norepinephrine; LDT, laterodorsal tegmental nucleus; Ach, acetylcholine; Mc, magnocellular
reticular nucleus; Gly, glycine; DPMe, deep mesencephalic reticular nucleus; PAG, periaq-
ueductal gray; DPGi, dorsal paragigantocellular reticular nucleus; PPT, pedunculopontine
nucleus; PRN, pontine reticular nucleus; SLD, sublaterodorsal nucleus; Glu, glutamate;
Pef/HLA perifornical/lateral hypothalamic area; Hcrt, hypocretin (orexin).
In addition, two populations of hypothalamic neurons containing the hypocretins

or MCH would participate to PS homeostasis through their direct excitatory and
inhibitory actions respectively on the monoaminergic and GABAergic PS-off neu-
rons and reciprocal inhibitory interactions.
ACKNOWLEDGMENTS
This work was supported by CNRS (ERS 5645, FRE 2469, and UMR5167),
INSERM (U480), Université Claude Bernard Lyon 1. The authors wish to thank C.
Guillemort (GFG Co., Pierre-Bénite, France) for his help in designing the head-
restraining system.
REFERENCES
1. Aserinsky, E. and Kleitman, N., Regularly occurring periods of eye motility and
concomitant phenomena during sleep, Science, 118, 273-274, 1953.
2. Dement, W.C., The occurrence of low-voltage, fast, electroencephalogram patterns
during behavioral sleep in the cat, Electroencephalogr. Clin. Neurophysiol., 10, 291-
296, 1958.

3. Jouvet, M. and Michel, F., Corrélations électromyographiques du sommeil chez

le chat décortiqué et mésencéphalique chronique, C.R. Soc. Biol., 153, 422-425,
1959.
4. Dement, W. and Kleitman, N., The relation of eye movements during sleep to dream
activity: an objective method for the study of dreaming, Journal of Experimental
Psychology. Learning, Memory, and Cognition, 53, 339, 1957.
5. Jouvet, M., Recherches sur les structures nerveuses et les mécanismes responsables
des différentes phases du sommeil physiologique, Arch. Ital. Biol., 100, 125-206,
1962.
6. Carli, G. and Zanchetti, A., A study of pontine lesions suppressing deep sleep in the
cat, Arch. Ital. Biol., 103 (4), 751-88, 1965.
7. Jouvet, M. and Delorme, F., Locus coeruleus et sommeil paradoxal, C.R. Seances
Soc. Biol., 159, 895-899, 1965.
8. Webster, H.H. and Jones, B.E., Neurotoxic lesions of the dorsolateral pontomesen-
cephalic tegmentum-cholinergic cell area in the cat. II. Effects upon sleep-waking
states, Brain Research, 458 (2), 285-302, 1988.
9. Sastre, J.P., Sakai, K., and Jouvet, M., Are the gigantocellular tegmental field neurons
responsible for paradoxical sleep?, Brain Research, 229 (1), 147-161, 1981.
10. Jouvet, M. and Mounier, D., Effets des lésions de la formation réticulée pontique sur
le sommeil du chat, C.R. Seances Soc. Biol. Fil, 154, 2301-2305, 1960.
11. Jones, B.E., Paradoxical sleep and its chemical/structural substrates in the brain,
Neuroscience, 40 (3), 637-656, 1991.
12. Jouvet, M. and Michel, F., Mise en evidence d`un “centre hypnique” au niveau du
rhombencéphale chez le chat., C.R. Acad. Sci., 154, 2301-2305, 1960.
13. George, R., Haslett, W.L., and Jenden, D.J., A cholinergic mechanism in the brainstem
reticular formation: induction of paradoxical sleep, Int. J. Neuropharmacol., 3, 541-
552, 1964.
14. Baghdoyan, H.A., Cholinergic mechanisms regulating REM sleep, in Sleep Science:
Integrating Basic Research and Clinical Practice, Schwartz, W.J.S., Ed., Karger
Publishing, Basel, 1997, pp. 88-116.
15. Vanni-Mercier, G., Sakai, K., Lin, J.S., and Jouvet, M., Mapping of cholinoceptive
brainstem structures responsible for the generation of paradoxical sleep in the cat,
Arch. Ital. Biol., 127 (3), 133-164, 1989.
16. Lai, Y.Y. and Siegel, J.M., Cardiovascular and muscle tone changes produced by
microinjection of cholinergic and glutamatergic agonists in dorsolateral pons and
medial medulla, Brain Research, 514 (1), 27-36, 1990.
17. Yamamoto, K., Mamelak, A.N., Quattrochi, J.J., and Hobson, J.A., A cholinoceptive
desynchronized sleep induction zone in the anterodorsal pontine tegmentum: locus
of the sensitive region, Neuroscience, 39 (2), 279-293, 1990.
18. Garzon, M., De Andres, I., and Reinoso-Suarez, F., Sleep patterns after carbachol
delivery in the ventral oral pontine tegmentum of the cat, Neuroscience, 83 (4), 1137-
1144, 1998.
19. Sakai, K., Sastre, J.P., Salvert, D., Touret, M., Tohyama, M., and Jouvet, M., Teg-
mentoreticular projections with special reference to the muscular atonia during par-
adoxical sleep in the cat: an HRP study, Brain Res., 176 (2), 233-254, 1979.
20. Sakai, K., Sastre, J.P., Kanamori, N., and Jouvet, M., State-specific neurones in the
ponto-medullary reticular formation with special reference to the postural atonia
during paradoxical sleep in the cat, in Brain Mechanisms of Perceptual Awareness
and Purposeful Behavior, Pompeiano, O. and Aimone Marsan, C., Eds., Raven Press,
New York, 1981, pp. 405-429.

21. Sakai, K., Neurons responsible for paradoxical sleep, in Sleep: Neurotransmitters
and Neuromodulators, Wauquier, A. and Janssen Research Foundation, Raven Press,
New York, 1985, pp. 29-42.
22. Sakai, K., Crochet, S., and Onoe, H., Pontine structures and mechanisms involved in
the generation of paradoxical (REM) sleep, Arch. Ital. Biol., 139 (1-2), 93-107, 2001.
23. Sakai, K. and Koyama, Y., Are there cholinergic and non-cholinergic paradoxical
sleep-on neurones in the pons?, Neuroreport, 7 (15-17), 2449-2453, 1996.
24. Chase, M.H., Soja, P.J., and Morales, F.R., Evidence that glycine mediates the
postsynaptic potentials that inhibit lumbar motoneurons during the atonia of active
sleep, J. Neurosci., 9 (3), 743-51, 1989.
25. Luppi, P.H., Sakai, K., Fort, P., Salvert, D., and Jouvet, M., The nuclei of origin of
monoaminergic, peptidergic, and cholinergic afferents to the cat nucleus reticularis
magnocellularis: a double-labeling study with cholera toxin as a retrograde tracer, J.
Comp. Neurol., 277 (1), 1-20, 1988.
26. Fort, P., Luppi, P.H., Wenthold, R., and Jouvet, M., Glycine immunoreactive neurons
in the medulla oblongata in cats, C. R. Acad. Sci. III, 311 (5), 205-212, 1990.
27. Fort, P., Luppi, P.H., and Jouvet, M., Glycine-immunoreactive neurones in the cat
brain stem reticular formation, Neuroreport, 4 (9), 1123-1126, 1993.
28. Soja, P.J., Lopez-Rodriguez, F., Morales, F.R., and Chase, M.H., The postsynaptic
inhibitory control of lumbar motoneurons during the atonia of active sleep: effect of
strychnine on motoneuron properties, J. Neurosci., 11 (9), 2804-2811, 1991.
29. Kohlmeier, K. A., Lopez-Rodriguez, F., Liu, R.H., Morales, F.R., and Chase, M.H.,
State-dependent phenomena in cat masseter motoneurons, Brain Res., 722 (1-2), 30-
38, 1996.
30. Yamuy, J., Fung, S.J., Xi, M., Morales, F.R., and Chase, M.H., Hypoglossal moto-
neurons are postsynaptically inhibited during carbachol-induced rapid eye movement
sleep, Neuroscience, 94 (1), 11-15, 1999.
31. Kodama, T., Lai, Y.Y., and Siegel, J.M., Enhanced glutamate release during REM
sleep in the rostromedial medulla as measured by in vivo microdialysis, Brain Res.,
780 (1), 178-181, 1998.
32. Lai, Y.Y. and Siegel, J.M., Pontomedullary glutamate receptors mediating locomotion
and muscle tone suppression, J. Neurosci., 11 (9), 2931-2937, 1991.
33. Siegel, J.M., Wheeler, R.L., and McGinty, D.J., Activity of medullary reticular for-
mation neurons in the unrestrained cat during waking and sleep, Brain Res., 179, 49-
60, 1979.
34. Sakai, K., Kanamori, N., and Jouvet, M., Neuronal activity specific to paradoxical
sleep in the bulbar reticular formation in the unrestrained cat, C.R. Seances Acad.
Sci. D., 289 (6), 557-561, 1979.
35. Holmes, C.J. and Jones, B.E., Importance of cholinergic, GABAergic, serotonergic
and other neurons in the medial medullary reticular formation for sleep-wake states
studied by cytotoxic lesions in the cat, Neuroscience, 62 (4), 1179-1200, 1994.
36. Holstege, J.C. and Bongers, C.M., A glycinergic projection from the ventromedial
lower brainstem to spinal motoneurons. An ultrastructural double labeling study in
rat, Brain Res., 566 (1-2), 308-315, 1991.
37. Li, Y.Q., Takada, M., Kaneko, T., and Mizuno, N., GABAergic and glycinergic
neurons projecting to the trigeminal motor nucleus: a double labeling study in the
rat, J. Comp. Neurol., 373 (4), 498-510, 1996.
38. Rampon, C., Peyron, C., Petit, J.M., Fort, P., Gervasoni, D., and Luppi, P.H., Origin
of the glycinergic innervation of the rat trigeminal motor nucleus, Neuroreport, 7
(18), 3081-3085, 1996.

39. Morales, F.R., Sampogna, S., Yamuy, J., and Chase, M.H., c-fos expression in brain-
stem premotor interneurons during cholinergically induced active sleep in the cat, J.
Neurosci., 19 (21), 9508-9518, 1999.
40. Hobson, J.A., McCarley, R.W., and Wyzinski, P.W., Sleep cycle oscillation: reciprocal
discharge by two brainstem neuronal groups, Science, 189 (4196), 55-58, 1975.
41. McCarley, R.W. and Hobson, J.A., Neuronal excitability modulation over the sleep
cycle: a structural and mathematical model, Science, 189 (4196), 58-60, 1975.
42. Aston-Jones, G. and Bloom, F.E., Activity of norepinephrine-containing locus coer-
uleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle, J.
Neurosci., 1 (8), 876-886, 1981.
43. McGinty, D.J. and Harper, R.M., Dorsal raphe neurons: depression of firing during
sleep in cats, Brain Res., 101 (3), 569-575, 1976.
44. Aghajanian, G.K. and Vandermaelen, C.P., Intracellular identification of central nora-
drenergic and serotonergic neurons by a new double labeling procedure, J. Neurosci.,
2 (12), 1786-1792, 1982.
45. Gervasoni, D., Panconi, E., Henninot, V., Boissard, R., Barbagli, B., Fort, P., and
Luppi, P.H., Effect of chronic treatment with milnacipran on sleep architecture in rats
compared with paroxetine and imipramine, Pharmacol. Biochem. Behav., 73 (3), 557-
563, 2002.
46. Tononi, G., Pompeiano, M., and Cirelli, C., Suppression of desynchronized sleep
through microinjection of the alpha 2-adrenergic agonist clonidine in the dorsal
pontine tegmentum of the cat, Pflugers Arch., 418 (5), 512-518, 1991.
47. Crochet, S. and Sakai, K., Effects of microdialysis application of monoamines on the
EEG and behavioural states in the cat mesopontine tegmentum, Eur. J. Neurosci., 11
(10), 3738-3752, 1999.
48. Crochet, S. and Sakai, K., Alpha-2 adrenoceptor mediated paradoxical (REM) sleep
inhibition in the cat, Neuroreport, 10 (10), 2199-2204, 1999.
49. Horner, R.L., Sanford, L.D., Annis, D., Pack, A.I., and Morrison, A.R., Serotonin at
the laterodorsal tegmental nucleus suppresses rapid-eye-movement sleep in freely
behaving rats, J. Neurosci., 17 (19), 7541-7552, 1997.
50. Kayama, Y., Ohta, M., and Jodo, E., Firing of “possibly” cholinergic neurons in the
rat laterodorsal tegmental nucleus during sleep and wakefulness, Brain Res., 569 (2),
210-220, 1992.
51. Datta, S., Spoley, E.E., and Patterson, E.H., Microinjection of glutamate into the
pedunculopontine tegmentum induces REM sleep and wakefulness in the rat, Am. J.
Physiol. Regulatory Integrative Comp. Physiol., 280, 752-759, 2001.
52. Datta, S. and Siwek, D.F., Single cell activity patterns of pedunculopontine tegmen-
tum neurons across the sleep-wake cycle in the freely moving rats, J. Neurosci. Res.,
70 (4), 611-621, 2002.
53. Datta, S., Mavanji, V., Patterson, E.H., and Ulloor, J., Regulation of rapid eye move-
ment sleep in the freely moving rat: local microinjection of serotonin, norepinephrine,
and adenosine into the brainstem, Sleep, 26 (5), 513-520, 2003.
54. McCarley, R.W., Nelson, J.P., and Hobson, J.A., Ponto-geniculo-occipital (PGO) burst
neurons: correlative evidence for neuronal generators of PGO waves, Science, 201
(4352), 269-272, 1978.
55. Saito, H., Sakai, K., and Jouvet, M., Discharge patterns of the nucleus parabrachialis
lateralis neurons of the cat during sleep and waking, Brain Res., 134 (1), 59-72,
1977.
56. Sakai, K. and Jouvet, M., Brain stem PGO-on cells projecting directly to the cat
dorsal lateral geniculate nucleus, Brain Res., 194 (2), 500-505, 1980.

57. Steriade, M., Pare, D., Datta, S., Oakson, G., and Curro Dossi, R., Different cellular
types in mesopontine cholinergic nuclei related to ponto-geniculo-occipital waves, J.
Neurosci., 10 (8), 2560-2579, 1990.
58. Sakai, K., Petitjean, F., and Jouvet, M., Effects of ponto-mesencephalic lesions and
electrical stimulation upon PGO waves and EMPs in unanesthetized cats, Electroen-
cephalogr. Clin. Neurophysiol., 41 (1), 49-63, 1976.
59. Pare, D., Smith, Y., Parent, A., and Steriade, M., Projections of brainstem core
cholinergic and non-cholinergic neurons of cat to intralaminar and reticular thalamic
nuclei, Neuroscience, 25 (1), 69-86, 1988.
60. Sofroniew, M.V., Priestley, J.V., Consolazione, A., Eckenstein, F., and Cuello, A.C.,
Cholinergic projections from the midbrain and pons to the thalamus in the rat,
identified by combined retrograde tracing and choline acetyltransferase immunohis-
tochemistry, Brain Res., 329 (1-2), 213-223, 1985.
61. Koyama, Y. and Sakai, K., Modulation of presumed cholinergic mesopontine tegmen-
tal neurons by acetylcholine and monoamines applied iontophoretically in unanes-
thetized cats, Neuroscience, 96 (4), 723–733, 2000.
62. McCall, R.B. and Aghajanian, G.K., Serotonergic facilitation of facial motoneuron
excitation, Brain Res., 169 (1), 11-27, 1979.
63. Morilak, D.A. and Jacobs, B.L., Noradrenergic modulation of sensorimotor processes
in intact rats: the masseteric reflex as a model system, J. Neurosci., 5 (5), 1300-1306,
1985.
64. Kubin, L., Davies, R.O., and Pack, A.I., Control of Upper Airway Motoneurons
During REM Sleep, News Physiol. Sci., 13, 91-97, 1998.
65. Guyenet, P.G. and Aghajanian, G.K., ACh, substance P and met-enkephalin in the
locus coeruleus: pharmacological evidence for independent sites of action, Eur. J.
Pharmacol., 53 (4), 319-328, 1979.
66. Koyama, Y. and Kayama, Y., Mutual interactions among cholinergic, noradrenergic
and serotonergic neurons studied by ionophoresis of these transmitters in rat brainstem
nuclei, Neuroscience, 55 (4), 1117-1126, 1993.
67. Luppi, P.H., Charlety, P.J., Fort, P., Akaoka, H., Chouvet, G., and Jouvet, M., Ana-
tomical and electrophysiological evidence for a glycinergic inhibitory innervation of
the rat locus coeruleus, Neurosci. Lett., 128 (1), 33-36, 1991.
68. Jones, B.E., Noradrenergic locus coeruleus neurons: their distant connections and
their relationship to neighboring (including cholinergic and GABAergic) neurons of
the central gray and reticular formation, Prog. Brain Res., 88, 15-30, 1991.
69. Onoe, H. and Sakai, K., Kainate receptors: a novel mechanism in paradoxical (REM)
sleep generation, Neuroreport, 6 (2), 353-356, 1995.
70. Xi, M.C., Morales, F.R., and Chase, M.H., Evidence that wakefulness and REM sleep
are controlled by a GABAergic pontine mechanism, J. Neurophysiol., 82 (4), 2015-
2019, 1999.
71. Xi, M.C., Morales, F.R., and Chase, M.H., The motor inhibitory system operating
during active sleep is tonically suppressed by GABAergic mechanisms during other
states, J. Neurophysiol., 86 (4), 1908-1915, 2001.
72. Xi, M.C., Morales, F.R., and Chase, M.H., Induction of wakefulness and inhibition
of active (REM) sleep by GABAergic processes in the nucleus pontis oralis, Arch.
Ital. Biol., 139 (1-2), 125-145, 2001.
73. Darracq, L., Gervasoni, D., Souliere, F., Lin, J.S., Fort, P., Chouvet, G., and Luppi,
P.H., Effect of strychnine on rat locus coeruleus neurones during sleep and wakeful-
ness, Neuroreport, 8 (1), 351-355, 1996.

74. Gervasoni, D., Darracq, L., Fort, P., Souliere, F., Chouvet, G., and Luppi, P.H., Elec-
trophysiological evidence that noradrenergic neurons of the rat locus coeruleus are
tonically inhibited by GABA during sleep, Eur. J. Neurosci., 10 (3), 964-970, 1998.
75. Gervasoni, D., Peyron, C., Rampon, C., Barbagli, B., Chouvet, G., Urbain, N., Fort,
P., and Luppi, P.H., Role and origin of the GABAergic innervation of dorsal raphe
serotonergic neurons, J. Neurosci., 20 (11), 4217-4225, 2000.
76. Boissard, R., Gervasoni, D., Schmidt, M.H., Barbagli, B., Fort, P., and Luppi, P.H.,
The rat ponto-medullary network responsible for paradoxical sleep onset and main-
tenance: a combined microinjection and functional neuroanatomical study, Eur. J.
Neurosci., 16 (10), 1959-1973, 2002.
77. Boissard, R., Fort, P., Gervasoni, D., Barbagli, B., and Luppi, P.H., Localization of
the GABAergic and non-GABAergic neurones projecting to the sublaterodorsal
nucleus and potentially gating paradoxical sleep onset, Eur. J. Neurosci., 18, 1627-
1639, 2003.
78. Boissard, R., Gervasoni, D., Fort, P., Henninot, V., Barbagli, B., and Luppi, P.H.,
Neuronal networks responsible for paradoxical sleep onset and maintenance in rats:
a new hypothesis, Sleep, 23 Suppl., 107, 2000.
79. Swanson, L.W., Brain Maps: Structure of the Rat Brain: A Laboratory Guide with
Printed and Electronic Templates for Data, Models, and Schematics, 2nd rev. ed.
Elsevier, New York, 1998.
80. Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Compact 3rd
ed., Academic Press, Sydney, Orlando, 1997.
81. Pollock, M.S. and Mistlberger, R.E., Rapid eye movement sleep induction by micro-
injection of the GABA-A antagonist bicuculline into the dorsal subcoeruleus area of
the rat, Brain Res., 962 (1-2), 68-77, 2003.
82. Koyama, Y., Schmidt, M.S., Takahashi, K., and Kayama, Y., Cholinergic neurones in
the laterodorsal tegmental nucleus regulate penile erection during paradoxical sleep,
Soc. Neurosci., (Abstr.) 522, 20, 2001.
83. Maloney, K.J., Mainville, L., and Jones, B.E., Differential c-Fos expression in cho-
linergic, monoaminergic, and GABAergic cell groups of the pontomesencephalic
tegmentum after paradoxical sleep deprivation and recovery, J. Neurosci., 19 (8),
3057-3072, 1999.
84. El Mansari, M., Sakai, K., and Jouvet, M., Unitary characteristics of presumptive
cholinergic tegmental neurons during the sleep–waking cycle in freely moving cats,
Exp. Brain Res., 76 (3), 519-529, 1989.
85. Steriade, M., Datta, S., Pare, D., Oakson, G., and Curro Dossi, R.C., Neuronal
activities in brain-stem cholinergic nuclei related to tonic activation processes in
thalamocortical systems, J. Neurosci., 10 (8), 2541-2559, 1990.
86. Gnadt, J.W. and Pegram, G.V., Cholinergic brainstem mechanisms of REM sleep in
the rat, Brain Res., 384 (1), 29-41, 1986.
87. Shiromani, P.J. and Fishbein, W., Continuous pontine cholinergic microinfusion via
mini-pump induces sustained alterations in rapid eye movement (REM) sleep, Phar-
macol. Biochem. Behav., 25 (6), 1253-1261, 1986.
88. Velazquez-Moctezuma, J., Gillin, J.C., and Shiromani, P.J., Effect of specific M1,
M2 muscarinic receptor agonists on REM sleep generation, Brain Res., 503 (1), 128-
131, 1989.
89. Bourgin, P., Escourrou, P., Gaultier, C., and Adrien, J., Induction of rapid eye move-
ment sleep by carbachol infusion into the pontine reticular formation in the rat,
Neuroreport, 6 (3), 532-536, 1995.

90. Deurveilher, S., Hars, B., and Hennevin, E., Pontine microinjection of carbachol does
not reliably enhance paradoxical sleep in rats, Sleep, 20 (8), 593-607, 1997.
91. Taguchi, O., Kubin, L., and Pack, A.I., Evocation of postural atonia and respiratory
depression by pontine carbachol in the decerebrate rat, Brain Res., 595 (1), 107-115,
1992.
92. Fenik, V., Ogawa, H., Davies, R.O., and Kubin, L., Pontine carbachol produces a
spectrum of REM sleep-like and arousal-like electrocortical responses in urethane-
anesthetized rats, Sleep Res. Online, 2 Suppl., 30, 1999.
93. De Andres, I., Gomez-Montoya, J., Gutierrez-Rivas, E., and Reinoso-Suarez, F.,
Differential action upon sleep states of ventrolateral and central areas of pontine
tegmental field, Arch. Ital. Biol., 123 (1), 1-11, 1985.
94. Taepavarapruk, N., McErlane, S.A., and Soja, P.J., State-related inhibition by GABA
and glycine of transmission in Clarke’s column, J. Neurosci., 22 (13), 5777-5788,
2002.
95. Jones, B.E. and Yang, T.Z., The efferent projections from the reticular formation and
the locus coeruleus studied by anterograde and retrograde axonal transport in the rat,
J. Comp. Neurol., 242 (1), 56-92, 1985.
96. Maloney, K.J., Mainville, L., and Jones, B.E., c-Fos expression in GABAergic, sero-
tonergic, and other neurons of the pontomedullary reticular formation and raphe after
paradoxical sleep deprivation and recovery, J. Neurosci., 20 (12), 4669-4679, 2000.
97. Sastre, J.P., Buda, C., Kitahama, K., and Jouvet, M., Importance of the ventrolateral
region of the periaqueductal gray and adjacent tegmentum in the control of paradox-
ical sleep as studied by muscimol microinjections in the cat, Neuroscience, 74 (2),
415-426, 1996.
98. Sastre, J.P., Buda, C., Lin, J.S., and Jouvet, M., Differential c-fos expression in the
rhinencephalon and striatum after enhanced sleep-wake states in the cat, Eur. J.
Neurosci., 12 (4), 1397-1410, 2000.
99. Mitani, A., Ito, K., Hallanger, A.E., Wainer, B.H., Kataoka, K., and McCarley, R.W.,
Cholinergic projections from the laterodorsal and pedunculopontine tegmental nuclei
to the pontine gigantocellular tegmental field in the cat, Brain Res., 451 (1-2), 397-
402, 1988.
100. Shiromani, P.J., Armstrong, D.M., and Gillin, J.C., Cholinergic neurons from the
dorsolateral pons project to the medial pons: a WGA-HRP and choline acetyltrans-
ferase immunohistochemical study, Neurosci. Lett., 95 (1-3), 19-23, 1988.
101. Quattrochi, J.J., Mamelak, A.N., Madison, R.D., Macklis, J.D., and Hobson, J.A.,
Mapping neuronal inputs to REM sleep induction sites with carbachol-fluorescent
microspheres, Science, 245 (4921), 984-986, 1989.
102. Semba, K., Aminergic and cholinergic afferents to REM sleep induction regions of
the pontine reticular formation in the rat, J. Comp. Neurol., 330 (4), 543-556, 1993.
103. Semba, K., Reiner, P.B., and Fibiger, H.C., Single cholinergic mesopontine tegmental
neurons project to both the pontine reticular formation and the thalamus in the rat,
Neuroscience, 38 (3), 643-654, 1990.
104. Sakai, K., Executive mechanisms of paradoxical sleep, Arch. Ital. Biol., 126 (4), 239-
257, 1988.
105. Webster, H.H., Friedman, L., and Jones, B.E., Modification of paradoxical sleep
following transections of the reticular formation at the pontomedullary junction,
Sleep, 9 (1), 1-23, 1986.
106. Vanni-Mercier, G., Sakai, K., Lin, J.S., and Jouvet, M., Carbachol microinjections in
the mediodorsal pontine tegmentum are unable to induce paradoxical sleep after caudal
pontine and prebulbar transections in the cat, Neurosci. Lett., 130 (1), 41-45, 1991.

107. Luppi, P.H., Gervasoni, D., Peyron, C., Barbagli, B., Boissard, R., and Fort, P.,
Norepinephrine and REM Sleep, in Rapid Eye Movement Sleep, Mallick, B. N. and
Inoue, S., Eds., Norosa Publishing Hoouse, New Delhi, 1999, pp. 107-122.
108. Maquet, P., Peters, J., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A., and Franck,
G., Functional neuroanatomy of human rapid-eye-movement sleep and dreaming,
Nature, 383 (6596), 163-166, 1996.
109. Deboer, T., Sanford, L.D., Ross, R.J., and Morrison, A.R., Effects of electrical stim-
ulation in the amygdala on ponto-geniculo-occipital waves in rats, Brain Res., 793
(1-2), 305-310, 1998.
110. Lin, J.S., Sakai, K., Vanni-Mercier, G., and Jouvet, M., A critical role of the posterior
hypothalamus in the mechanisms of wakefulness determined by microinjection of
muscimol in freely moving cats, Brain Res., 479 (2), 225-240, 1989.
111. Steininger, T.L., Alam, M.N., Gong, H., Szymusiak, R., and McGinty, D., Sleep-
waking discharge of neurons in the posterior lateral hypothalamus of the albino rat,
Brain Res., 840 (1-2), 138-147, 1999.
112. Alam, M.N., Gong, H., Alam, T., Jaganath, R., McGinty, D., and Szymusiak, R.,
Sleep–waking discharge patterns of neurons recorded in the rat perifornical lateral
hypothalamic area, J. Physiol., 538 (Pt 2), 619-631, 2002.
113. Peyron, C., Faraco, J., Rogers, W., Ripley, B., Overeem, S., Charnay, Y., Nevsimalova,
S., Aldrich, M., Reynolds, D., Albin, R., Li, R., Hungs, M., Pedrazzoli, M., Padigaru,
M., Kucherlapati, M., Fan, J., Maki, R., Lammers, G.J., Bouras, C., Kucherlapati,
R., Nishino, S., and Mignot, E., A mutation in a case of early onset narcolepsy and
a generalized absence of hypocretin peptides in human narcoleptic brains, Nat. Med.,
6 (9), 991-997, 2000.
114. Lin, L., Faraco, J., Li, R., Kadotani, H., Rogers, W., Lin, X., Qiu, X., de Jong, P.J.,
mutation in the hypocretin (orexin) receptor 2 gene, Cell, 98 (3), 365-376, 1999.
115. Chemelli, R.M., Willie, J.T., Sinton, C.M., Elmquist, J.K., Scammell, T., Lee, C.,
Richardson, J.A., Williams, S.C., Xiong, Y., Kisanuki, Y., Fitch, T.E., Nakazato, M.,
Hammer, R.E., Saper, C.B., and Yanagisawa, M., Narcolepsy in orexin knockout
mice: molecular genetics of sleep regulation, Cell, 98 (4), 437-451, 1999.
116. Verret, L., Goutagny, R., Fort, P., Cagnon, L., Salvert, D., Leger, L., Boissard, R.,
Salin, P., Peyron, C., and Luppi, P.H., A role of melanin-concentrating hormone
producing neurons in the central regulation of paradoxical sleep, B.M.C. Neurosci.,
4 (1), 19, 2003.
117. Yamanaka, A., Tsujino, N., Funahashi, H., Honda, K., Guan, J.L., Wang, Q.P., Tom-
inaga, M., Goto, K., Shioda, S., and Sakurai, T., Orexins activate histaminergic
neurons via the orexin 2 receptor, Biochem. Biophys. Res. Commun., 290 (4), 1237-
1245, 2002.
118. Bittencourt, J.C., Presse, F., Arias, C., Peto, C., Vaughan, J., Nahon, J.L., Vale, W.,
and Sawchenko, P.E., The melanin-concentrating hormone system of the rat brain:
an immuno- and hybridization histochemical characterization, J. Comp. Neurol., 319
(2), 218-245, 1992.
119. Peyron, C., Tighe, D.K., van den Pol, A.N., de Lecea, L., Heller, H.C., Sutcliffe, J.
G., and Kilduff, T.S., Neurons containing hypocretin (orexin) project to multiple
neuronal systems, J. Neurosci., 18 (23), 9996-10015, 1998.
120. Ennis, M. and Aston-Jones, G., GABA-mediated inhibition of locus coeruleus from
the dorsomedial rostral medulla, J. Neurosci., 9 (8), 2973-2981, 1989.
121. Gallager, D.W. and Aghajanian, G.K., Effect of antipsychotic drugs on the firing of
dorsal raphe cells. II. Reversal by picrotoxin, Eur. J. Pharmacol., 39 (2), 357-364, 1976.

122. Cherubini, E., North, R.A., and Williams, J.T., Synaptic potentials in rat locus coer-
uleus neurones, J. Physiol., 406, 431-442, 1988.
123. Williams, J.T., Bobker, D.H., and Harris, G.C., Synaptic potentials in locus coeruleus
neurons in brain slices, Prog. Brain Res., 88, 167-172, 1991.
124. Osmanovic, S.S. and Shefner, S.A., gamma-Aminobutyric acid responses in rat locus
coeruleus neurones in vitro: a current-clamp and voltage-clamp study, J. Physiol.,
421, 151-170, 1990.
125. Pan, Z.Z., Colmers, W.F., and Williams, J.T., 5-HT-mediated synaptic potentials in
the dorsal raphe nucleus: interactions with excitatory amino acid and GABA neu-
rotransmission, J. Neurophysiol., 62 (2), 481-486, 1989.
126. Luque, J.M., Malherbe, P., and Richards, J.G., Localization of GABAA receptor
subunit mRNAs in the rat locus coeruleus, Brain Res. Mol. Brain Res., 24 (1-4), 219-
226, 1994.
127. Wang, Q.P., Ochiai, H., and Nakai, Y., GABAergic innervation of serotonergic neurons
in the dorsal raphe nucleus of the rat studied by electron microscopy double immu-
nostaining, Brain Res. Bull., 29 (6), 943-948, 1992.
128. Zarbin, M.A., Wamsley, J.K., and Kuhar, M.J., Glycine receptor: light microscopic
autoradiographic localization with [3H] strychnine, J. Neurosci., 1 (5), 532-547, 1981.
129. Levine, E.S. and Jacobs, B.L., Neurochemical afferents controlling the activity of
serotonergic neurons in the dorsal raphe nucleus: microiontophoretic studies in the
awake cat, J. Neurosci., 12 (10), 4037-4044, 1992.
130. Sakai, K. and Crochet, S., Serotonergic dorsal raphe neurons cease firing by disfa-
cilitation during paradoxical sleep, Neuroreport, 11 (14), 3237-3241, 2000.
131. Nitz, D. and Siegel, J., GABA release in the dorsal raphe nucleus: role in the control
of REM sleep, Am. J. Physiol., 273 (1 Pt 2), R451-455, 1997.
132. Nitz, D. and Siegel, J.M., GABA release in the locus coeruleus as a function of
sleep/wake state, Neuroscience, 78 (3), 795-801, 1997.
133. Heym, J., Trulson, M.E., and Jacobs, B.L., Effects of adrenergic drugs on raphe unit
activity in freely moving cats, Eur. J. Pharmacol., 74 (2-3), 117-125, 1981.
134. Bayer, L., Eggermann, E., Serafin, M., Saint-Mleux, B., Machard, D., Jones, B., and
Muhlethaler, M., Orexins (hypocretins) directly excite tuberomammillary neurons,
Eur. J. Neurosci., 14 (9), 1571-1575, 2001.
135. Brown, R.E., Sergeeva, O.A., Eriksson, K.S., and Haas, H.L., Convergent excitation
of dorsal raphe serotonin neurons by multiple arousal systems (orexin/hypocretin,
histamine and noradrenaline), J. Neurosci., 22 (20), 8850-8859, 2002.
136. Liu, R.J., van den Pol, A. N., and Aghajanian, G.K., Hypocretins (orexins) regulate
serotonin neurons in the dorsal raphe nucleus by excitatory direct and inhibitory
indirect actions, J. Neurosci., 22 (21), 9453-9464, 2002.
137. Bourgin, P., Huitron-Resendiz, S., Spier, A.D., Fabre, V., Morte, B., Criado, J.R.,
Sutcliffe, J.G., Henriksen, S.J., and de Lecea, L., Hypocretin-1 modulates rapid eye
movement sleep through activation of locus coeruleus neurons, J. Neurosci., 20 (20),
7760-7765, 2000.
138. Hagan, J.J., Leslie, R.A., Patel, S., Evans, M.L., Wattam, T.A., Holmes, S., Benham,
C.D., Taylor, S.G., Routledge, C., Hemmati, P., Munton, R.P., Ashmeade, T.E., Shah,
A.S., Hatcher, J.P., Hatcher, P.D., Jones, D.N., Smith, M.I., Piper, D.C., Hunter, A.J.,
Porter, R.A., and Upton, N., Orexin A activates locus coeruleus cell firing and
increases arousal in the rat, Proc. Natl. Acad. Sci. USA, 96 (19), 10911-10916, 1999.
139. Horvath, T.L., Peyron, C., Diano, S., Ivanov, A., Aston-Jones, G., Kilduff, T.S., and
van Den Pol, A.N., Hypocretin (orexin) activation and synaptic innervation of the
locus coeruleus noradrenergic system, J. Comp. Neurol., 415 (2), 145–159, 1999.

140. Gao, X.B. and van den Pol, A.N., Melanin concentrating hormone depresses synaptic
activity of glutamate and GABA neurons from rat lateral hypothalamus, J. Physiol.,
533 (Pt 1), 237-252, 2001.
141. Cespuglio, R., Sarda, N., Gharib, A., Chastrette, N., Houdouin, F., Rampin, C., and
Jouvet, M., Voltammetric detection of the release of 5-hydroxyindole compounds
throughout the sleep–waking cycle of the rat, Exp. Brain Res., 80 (1), 121-128, 1990.
142. Portas, C.M., Bjorvatn, B., and Ursin, R., Serotonin and the sleep-wake cycle: special
emphasis on microdialysis studies, Prog. Neurobiol., 60 (1), 13-35, 2000.
143. Lu, J., Bjorkum, A.A., Xu, M., Gaus, S.E., Shiromani, P.J., and Saper, C.B., Selective
activation of the extended ventrolateral preoptic nucleus during rapid eye movement
sleep, J. Neurosci., 22 (11), 4568-4576, 2002.
regulation of the sleep–waking cycle, Ergeb. Physiol., 64, 166-307, 1972.
145. Woch, G., Davies, R.O., Pack, A.I., and Kubin, L., Behaviour of raphe cells projecting
to the dorsomedial medulla during carbachol-induced atonia in the cat, J. Physiol.,
490 (Pt 3), 745-758, 1996.
146. Jolas, T., Nestler, E.J., and Aghajanian, G.K., Chronic morphine increases GABA
tone on serotonergic neurons of the dorsal raphe nucleus: association with an up-
regulation of the cyclic AMP pathway, Neuroscience, 95 (2), 433-443, 2000.
147. Yamuy, J., Sampogna, S., Lopez-Rodriguez, F., Luppi, P.H., Morales, F.R., and Chase,
M.H., Fos and serotonin immunoreactivity in the raphe nuclei of the cat during
carbachol-induced active sleep: a double-labeling study, Neuroscience, 67 (1), 211-
223, 1995.
148. Verret, L., Localization of the neurons responsible for the inhibition of locus voeruleus
noradrenergic neurons during paradoxical sleep in the rat., Sleep, 26, 2003.
149. Goutagny, R., Fort, P., and Luppi, P.H., Role of dorsal paragigantocellular nucleus
in paradoxical sleep regulation: a study combining electrophysiology and pharma-
cology across vigilance states in the rat., The paradox of sleep: an unfinished study,
Lyon (France), 2003.
150. Kaur, S., Saxena, R.N., and Mallick, B.N., GABAergic neurons in prepositus hypo-
glossi regulate REM sleep by its action on locus coeruleus in freely moving rats,
Synapse, 42 (3), 141-150, 2001.

6 Reverse Genetics and the

Study of Sleep-Wake
Cycle: The Hypocretins
and Cortistatin
Luis de Lecea
CONTENTS
Differential Gene Expression as a Tool to Study Sleep

The Hypocretins (Orexins): Two Hypothalamus-Specific Peptides
The Hypocretins in Normal Sleep-Wake Cycle
Cortical Gene Expression: Cortistatin
Transgenic Approaches to Study the Role of Neuropeptides on Sleep
Neuropeptides in Control of Sleep
Acknowledgments
References
DIFFERENTIAL GENE EXPRESSION AS A TOOL TO

STUDY SLEEP
From the vast amount of information provided by the sequencing of vertebrate
genomes,1,2 there are several numbers that are striking, regarding the 20–30% of
nonannotated sequences that are expressed in the brain. An estimated 10% of the
mouse or human brain transcriptome belongs to the low abundance class; that is, it
is expressed in fewer than five copies per cell.3–5 Given the extraordinary cellular
complexity of the central nervous system, it can be estimated that a few hundred
mRNAs are expressed in small populations (less than106 neurons) of cells.4 Expres-
sion of these rare mRNAs in small populations of neurons would confer particular
physiological properties to the neurons that produce them, thus there are a few
hundred populations of neurons that are functionally distinct and whose character-
ization may provide essential information about the fine mechanisms of brain func-
tion and, in particular, sleep-wake regulation.
In spite of major advances in our understanding of the neuronal circuits that
govern the sleep-wakefulness cycle,6 the cell groups involved in the different stages
0-8493-1519-0/05/$0.00+$1.50

FIGURE 6.1 Distribution of the peptide precursors described in this chapter. The hypocretins
are two peptides derived from the same precursor and share a 7/7 match with the hormone
secretin. The hypocretin precursor is restricted to the lateral hypothalamus. Cortistatin is very
similar to the neuropeptide somatostatin and is expressed in GABAergic neurons in the cortex
hippocampus and amygdala. Both of these peptides precursors were discovered by differential
gene expression methods in the brain. (Modified from Sutcliffe, J.G. and De Lecea, L., The
hypocretins: setting the arousal threshold, Nat. Rev. Neurosci., 3, 339–349, 2002. With
permission.)
of sleep and in the control of the boundaries between sleep states are poorly under-
stood. The development of molecular markers that define neuronal cell groups with
distinct physiological properties will enhance our understanding of the regulation
of the states of vigilance.
The search of molecular markers that define populations of neurons in areas
important for arousal is broadly warranted. This chapter describes the isolation, by
differential gene expression analysis, of two peptidergic systems that modulate
different aspects of the sleep-wakefulness cycle. The success of this strategy dem-
onstrates the need for new markers of neuronal cell types, which may define
populations of neurons critical for our understanding of cortical activity and sleep
(Figure 6.1).
THE HYPOCRETINS (OREXINS): TWO

HYPOTHALAMUS-SPECIFIC PEPTIDES
The hypothalamus can be considered a federation of nuclei with distinct functions
that include energy homeostasis, circadian rhythms, sex behavior, and arousal. It is
thus expected that mRNAs specifically expressed in restricted areas of the hypothal-

amus will carry selective functions. Analysis of a collection of the most prevalent
cDNAs expressed in the hypothalamus revealed that as many as 40% of these
sequences encode secreted proteins.7 Further characterization of a cDNA encoding
a novel putative secreted protein revealed that it was restricted to the perifornical
area of the lateral hypothalamus. The deduced protein sequence contained a putative
signal secretory sequence and several pairs of dibasic residues that were possible
substrates of prohormone convertases. Cleavage at these sites would generate two
putative products of proteolysis had 13 amino acid identities across 19 residues.
This region of one of the peptides contained a 7/7 match with secretin, suggesting
that the prepropeptide gave rise to two peptide products that were structurally related
both to each other and to secretin. These putative peptides were named hypocretin
(hcrt)-1 and -2 to reflect their hypothalamic origin and the similarity to the incretin
neuropeptide family.8 The peptides showed neuroexcitatory activity in mature, cul-
tured hypothalamic neurons and were localized in large, dense, core vesicles by
immuno electron microscopy. Shortly after the peptides were discovered, Sakurai et
al. reported the isolation of the orexins, which are identical to the hypocretins, as
the endogenous ligands of two orphan G-protein coupled receptors. These authors
named the peptides orexins because they showed feed-inducing activity when
injected into the brain ventricles.
A great deal of interest was sparked by three reports linking the hypocretinergic
system with narcolepsy. The discovery that canine narcolepsy is caused by mutations
in hypocretin receptor 2, together with the narcolepsy-like phenotype of hypocretin
deficient mice, and the practical absence of hypocretin neurons in the hypothalamus
of narcoleptic patients has demonstrated that this system is involved in the state
boundary control. Comprehensive reviews of the hypocretinergic system are avail-
able elsewhere in the literature.9–11
THE HYPOCRETINS IN NORMAL SLEEP-WAKE CYCLE

Even though it appears that the main function of the hypocretins involves stability
of the boundaries between wakefulness and sleep, the precise role of hcrts in normal
sleep is still a matter of debate.
The actions of hcrt on sleep may be integrated into the reciprocal interaction
model of REM sleep generation by McCarley and Hobson.12 This model considers
two populations of neurons: REM-off cells, which are silent during REM sleep, and
REM-on neurons, which generate REM sleep bouts. REM-off cells, which include
noradrenergic neurons of the LC, serotoninergic neurons of the raphe nucleus, and
the histaminergic neurons of the tuberomammilary nucleus (TMN), are highly active
during wake and silent during REM sleep. During wakefulness REM-off neurons
inhibit REM-on cells, which include cholinergic neurons of the laterodorsal tegmen-
tum and pedunculo pontine nucleus (LDT/PPT). During REM sleep, REM-on cells
show a higher activity after the inhibitory action of REM-off cells is removed.
Considering the wake-promoting properties of hcrt-1, it has been suggested that
hcrt increases arousal and inhibits REM sleep by activating REM-off cells, in
particular the noradrenergic ones in the LC which receive the densest hcrt innerva-

tion.13,14 This hypothesis is in line with in vitro and in vivo experiments that have
shown that hcrt-1 excites this cell population.15–17 Further, local administration of
hcrt-1 promotes wakefulness and suppresses REM sleep (Bourgin et al., 2000).
Several studies have revealed the ability of hcrt to excite other REM-off neurons18–21
as well as REM-on cells in the LDT/PPT22 and cholinergic neurons in the basal
forebrain.23,24 Part of the wake-promoting effects of hcrt seem to be mediated by
histaminergic neurons in the tuberomammilary nucleus, as histamine H1 receptor
knockout mice are impervious to hcrt administration.21 Similar analyses in mutant
animals with alterations in specific neurotransmitter systems will lead to a better
understanding of the interaction of the hcrts with the sleep-wake circuitry.
Based on the framework of the reciprocal interaction model of McCarley and
Hobson, two alternative models have been proposed to integrate the activity of hcrt
neurons in the reciprocal inhibitory model for REM sleep regulation. Mignot and
collaborators11 considered hcrt neurons as wake neurons. During arousal, hcrt neu-
rons are activated by metabolic, circadian, and stress circuits, and stimulate both
REM-off and REM-on neurons leading to the awake state. In contrast during REM
sleep hcrt neurons exhibit minimal activity, reducing the firing of REM-off neurons
and subsequently activating REM-on neurons. Kilduff and Peyron25 proposed that
hcrt cells are wake-on and REM-on neurons. According to this model, hcrt neurons
drive the tonus of both REM-on and REM-off neurons during wakefulness. This
same model postulates that during REM sleep, REM-off cells will be inhibited by
GABAergic neurons from the periaqueductal gray and disinhibit REM-on cells
(Figure 6.2).
Other authors have suggested that hcrt neurons are activated during wakefulness
but only when somato-motor activity is present, independently of the state of vigi-
lance.26 The experimental data currently available is compatible with these three
previous models that, while they are verified, provided testable working hypotheses.
Investigating the activity of hcrt neurons during states of vigilance is essential
to understanding the physiological role of hcrt in the regulation of sleep-wakefulness.
Several studies correlate hcrt release with the sleep-wakefulness cycle. Prolonged
waking produced by pharmacological and instrumental sleep deprivations produces
an increase in extracellular hcrt levels or c-Fos/hcrt mRNA positive cells,27–29 which
initially may suggest that hcrt is a factor that accumulates during wakefulness.
However there is no correlation between hcrt levels and wake or sleep amounts,29
strongly suggesting that hcrt may be primarily related to the regulation of the
transitions between states of vigilance, rather than a particular sleep-wake stage.
Electrophysiological studies have investigated the activity of neurons in the lateral
hypothalamus in parallel with sleep recording in freely moving rats. These studies
have identified two cell types in the lateral hypothalamus (LH): wake-on/REM-on
neurons and REM-off cells, indicating that LH neurons, which include hcrt neurons,
exhibit a discharge pattern that correlates with arousal and sleep.30
CORTICAL GENE EXPRESSION: CORTISTATIN

Cortistatin was discovered as a result of the effort to characterize cortex-specific
gene expression modulated by synaptic activity and named after its cortical expres-

FIGURE 6.2 (See color insert following page 108.) Hypocretinergic activity dependent on
the states of vigilance. During wakefulness, metabolic, circadian, and behavioral inputs
converge on hypocretin neurons, which activate noradrenergic neurons in the locus coeruleus
and promote arousal. During non REM sleep, the activity of hypocretin neurons decreases,
but the inhibition of REM-off neurons over REM-on cells is still effective. During REM sleep,
hypocretin and REM-off cells are silent disinhibiting REM-on cells. (From Sutcliffe, J.G. and
De Lecea, L., The hypocretins: setting the arousal threshold, Nat. Rev. Neurosci., 3, 339–349,
2002. With permission.)
sion and sequence homology to somatostatin.31 The characterization of this peptide

is yet another example of the use of reverse genetics to study the molecular com-
ponents of the sleep machinery.
Cortistatin is synthesized as a precursor of 116 amino acids, which gives rise to
a C-terminal mature peptide, cortistatin-14 (CST-14), that shares 11 of its 14 residues
with the neuropeptide somatostatin. However the similarity between cortistatin and

somatostatin is restricted to the mature peptide and are the products of different genes.
CST-14 binds to all five somatostatin receptors in vitro, although several authors
suggest that CST-14 exerts its actions in vivo by binding to its own specific receptor.32
Cortistatin expression is restricted to scattered cells in the cerebral cortex and
hippocampus. These neurons use GABA as their neurotransmitter and are different
from the population of cortical neurons that express somatostatin.33 Networks
of GABAergic inhibitory neurons are known to be critical for synchronization of
cortical activity and have been proposed to have a major role in the maintenance
of slow wave sleep.34–37 Experiments and models have shown how the network
frequency depends on excitation of the interneurons and on the parameters of
GABAA-mediated IPSCs between the interneurons (conductance and time course).38
The electrophysiological firing properties of interneurons are substantially different
from those of pyramidal cells, and they are thought to be based on the expression
of particular ionic conductances (e.g., HCN2, KCNQ2-4, Kv3.1, etc). Further proof
that cortical GABAergic neurons and these conductances are important for cortical
activity and slow-wave sleep are the significant differences in delta power of mice
deficient in Kv3.1 channels.39
Intracerebroventricular infusion of CST14 dramatically increases the amount of
slow-wave activity in rats, at the expense of wakefulness. The mechanism by which
CST-14 enhances cortical synchronization has been established through the interac-
tion of CST-14 with acetylcholine, a neurotransmitter known to be involved in the
maintenance of cortical desynchronization. Application of Ach in the anesthetized
animal increases fast activity, and this effect is blocked with the simultaneous
addition of CST-14. These data suggest that CST-14 increases slow-wave sleep by
antagonizing the effects of acetylcholine on cortical excitability. In addition to this
mechanism, cortistatin may enhance cortical synchronization by enhancing Ih, a
cation conductance shown to be important in thalamocortical synchronization.32
A set of experiments suggests that cortistatin expression correlates with the sleep
homeostat. The concentration of cortistatin mRNA oscillates along the light/dark
cycle in rats, with maximal levels at the end of the dark (active) period. Further, the
steady-state concentration of cortistatin mRNA increases fourfold upon sleep dep-
rivation and returns to normal levels after sleep rebound, indicating that the expres-
sion of the peptide is associated with sleep demand.32
TRANSGENIC APPROACHES TO STUDY THE ROLE OF

NEUROPEPTIDES ON SLEEP
The most challenging aspect of reverse genetics is the characterization of the function
of the newly discovered genes. Genetically modified mice, transgenic and knockout
animals, give a wealth of information about the phenotypic consequences of the
missing gene. Here we are presenting several transgenic and knockout mice affecting
the hypocretinergic and cortistatin systems as examples of the possibilities of this
technology in the study of sleep circuits.
In addition to the study of canine model of narcolepsy,40 which resulted in the
discovery of the hypocretinergic system as a main regulator of transition of states

of vigilance, the development of new mouse models with altered hypocretinergic

transmission is increasing our knowledge on this neurotransmitter system. In par-
ticular the generation and phenotypic characterization of hypocretin and orexin
knockout mice41 has extended the observations and the importance of the peptide
transmitters, rather than the receptors, as critical components of the circuits that
control the stability of the states of vigilance.
To further support the role for hcrt in narcolepsy, hcrt receptor 2 ko mice have
been recently generated. These animals present a milder cataplexyphenotype than
the ligand knockouts, suggesting hcrtr2-dependent and hcrtr-2-independent mecha-
nisms in the narcolepsy/cataplexy syndrome.42
Likewise the analysis of mice deficient in cortistatin is uncovering additional
functions for this peptidergic system. Preliminary data indicate that these mice are
hyperexcitable and display enhanced long-term potentiation in the hippocampus. In
contrast, transgenic mice overexpressing cortistatin show decreased long-term poten-
tiation and impaired spatial learning.
In addition to knockout animals, expression of different transgenes under the
control of selective promoters, such as hypocretin or cortistatin, can modify the
neuronal system in such a way that we obtain additional information about its
physiology and connectivity. For instance, elegant studies in transgenic mice have
demonstrated that hypocretin-containing neurons are important components in the
control of homeostasis. Selective degeneration of hypocretin expressing cells by the
use of the toxic gene ataxin 3 under the control of the hypocretin promoter, caused
narcolepsy, hypophagia, and obesity,43 thus demonstrating a pivotal role for the cells
expressing hcrt in the coordination and integration of the homeostatic circuitry.
Animal models, in which hcrt-expressing cells are constitutively activated (e.g.,
by the use of cholera toxin under the control of the hcrt promoter) may shed important
light to the effect of stimulus-independent hcrt-ergic activation. Cholera toxin A
inhibits the inhibition of adenylyl cyclase and, when expressed intracellularly, causes
constitutive high levels of cAMP.44 This constitutive activation is equivalent to
placing a stimulating electrode in hypocretin neurons.
The recent development of genetically encoded fluorescent reporters of cell
activity45 opens a unique opportunity to allow monitoring of ensembles of neurons
defined by transgenic promoters. The hypocretinergic system is an outstanding
candidate for such endeavors because of its restricted localization to a few thousand
neurons and relative electrophysiological homogeneity.46
Other transgenic approaches can be used to map the anatomical afferents to sets
of neurons defined by a transgenic promoter. Following the strategy described by
DeFalco et al.,47 transgenic mice expressing the cre recombinase can be infected
with a recombinant pseudorabies virus that depends on cre-mediated recombination
for replication. In addition, this modified virus expresses green fluorescent protein
following cre recombination. Thus one could map polysynaptic connections to
hypocretin neurons expressing cre in transgenic mice. This would allow a cell-type
specific mapping of affererents, to better define the neural system.
The signals that regulate hypocretin can be functionally investigated using in
vitro cell-specific visualization of neuronal activity through the use of cellular tag-
ging by EGFP. Van den Pol et al. have determined that hcrt neurons can be depo-

larized by hcrt-1 via a glutamatergic interneuron.46 In contrast, hcrt neurons can be

hyperpolarized by norepinephrin or serotonin, suggesting an inhibitory feedback
loop between these transmitters and hypocretin cells.46
An alternative method to intracellular recordings in fluorescently labeled hypo-
cretin neurons is the cameleon technology.48 The cameleons are recombinant con-
structs that consist of two fluorescent molecules, a yellow-emitting molecule (YFP)
and a cyan-emitting molecule, linked by a calmodulin domain and a peptide (called
M13) that binds calmodulin in the presence of calcium. When the concentration of
intracellular calcium, which corresponds to increased neuronal activity, reaches a
maximum, the calmodulin domain of the cameleon binds to the M13 peptide and
brings the blue and yellow emitting proteins together, resulting in an increase in the
ratio of blue fluorescence known as FRET (fluorescence resonance emission trans-
fer).48,49 Thus changes in intracellular calcium concentrations in hypocretin neurons
can be detected by expressing cameleons in transgenic mice under the control of
the hypocretin promoter. This method has the advantage of measuring the activity
of the entire hypocretinergic system in response to defined stimuli.
Additional important data of neuronal activity related with sleep can be achieved
in vitro by the application of voltage sensitive fluorescent dies to cortical slices. This
methodology has demonstrated the importance of specific conductances (Ih) on
spontaneous neuronal activity in the cortical micro-circuitry.50 When applied to
transgenic and knockout mice, this technology can reveal new mechanisms of cor-
tical physiology relevant to sleep processes.
Development of magnetic resonance (MR) imaging in rodents will allow in the
future the visualization of activity of hypocretin cells in vivo. MR uses radio frequency
energy pulses and a strong magnetic field to provide pictures of internal organs and
tissues. The images have remarkable detail and can be visualized in each of the three
spatial planes (axial, sagittal, and coronal) using data collected from a single imaging
procedure. MR imaging has proven to be an invaluable tool for the diagnosis of a
wide range of pathologies including cancer, cardiovascular disease, joint and muscu-
loskeletal disorders, and neurological disease. Although the spatial resolution of MR
in most magnets available for in vivo applications does not reach the micron level, it
is expected that the development of equipment, parallel to that of protein NMR will
allow cellular and subcellular resolution in freely behaving animals. Also, recent
development of contrasting agents that depend on ion flux, such as DOPTA-Gd for
detection of intracellular calcium,51 may allow the monitoring of ensembles of neurons
(e.g., defined by the hypocretin promoter) in freely moving mice.
NEUROPEPTIDES IN CONTROL OF SLEEP

Here we have described two examples of new neuropeptides that have different
activities and regulate different aspects of the sleep-wakefulness cycle. The use of
reverse genetics has uncovered basic properties of the peptides (structure, distribu-
tion, electrophysiological activity, connectivity). The development of mouse models
has allowed a better understanding of their function and their role on sleep regulation.
It is not surprising that neuropeptides have a major role in the modulation of sleep
circuits. First the time frame and kinetics of peptidergic action (seconds to minutes)

is consistent with the reversible modification of synaptic transmission that would be

required during sleep. Second peptide release is associated with high frequency firing
of presynaptic neurons, which provides another code for the stability of neuronal
systems. New peptidergic systems with restricted localizations in areas critical for
sleep-wakefulness regulation are likely to come to light in the near future.
ACKNOWLEDGMENTS
I thank our collaborators for keeping these stories exciting and especially Raphaelle
Winsky, Pilar Ruiz-Lozano, and Steven Henriksen for critical reading. This work
was supported by grants from NIH.
REFERENCES
1. Venter, J.C., Adams, M.D., Myers, E.W., Li, P.W., Mural, R.J., Sutton, G.G., Smith,
H.O., Yandell, M., Evans, C.A., Holt, R.A., Gocayne, J.D., Amanatides, P., Ballew,
R.M., Huson, D.H., Wortman, J. R., Zhang, Q., Kodira, C.D., Zheng, X.H., Chen,
L., Skupski, M., Subramanian, G., Thomas, P.D., Zhang, J., Gabor Miklos, G.L.,
Nelson, C., Broder, S., Clark, A.G., Nadeau, J., McKusick, V.A., Zinder, N., Levine,
A.J., Roberts, R.J., Simon, M., Slayman, C., Hunkapiller, M., Bolanos, R., Delcher,
A., Dew, I., Fasulo, D., Flanigan, M., Florea, L., Halpern, A., Hannenhalli, S., Kravitz,
S., Levy, S., Mobarry, C., Reinert, K., Remington, K., Abu-Threideh, J., Beasley, E.,
Biddick, K., Bonazzi, V., Brandon, R., Cargill, M., Chandramouliswaran, I., Charlab,
R., Chaturvedi, K., Deng, Z., Di Francesco, V., Dunn, P., Eilbeck, K., Evangelista,
C., Gabrielian, A.E., Gan, W., Ge, W., Gong, F., Gu, Z., Guan, P., Heiman, T.J.,
Higgins, M.E., Ji, R.R., Ke, Z., Ketchum, K.A., Lai, Z., Lei, Y., Li, Z., Li, J., Liang,
Y., Lin, X., Lu, F., Merkulov, G.V., Milshina, N., Moore, H.M., Naik, A.K., Narayan,
V.A., Neelam, B., Nusskern, D., Rusch, D.B., Salzberg, S., Shao, W., Shue, B., Sun,
J., Wang, Z., Wang, A., Wang, X., Wang, J., Wei, M., Wides, R., Xiao, C., Yan, C.
et al., The sequence of the human genome, Science, 291, 1304–1351, 2001.
2. Lander, E.S., Linton, L.M., Birren, B., Nusbaum, C., Zody, M.C., Baldwin, J., Devon,
K., Dewar, K., Doyle, M., FitzHugh, W., Funke, R., Gage, D., Harris, K., Heaford,
A., Howland, J., Kann, L., Lehoczky, J., LeVine, R., McEwan, P., McKernan, K.,
Meldrim, J., Mesirov, J.P., Miranda, C., Morris, W., Naylor, J., Raymond, C., Rosetti,
M., Santos, R., Sheridan, A., Sougnez, C., Stange-Thomann, N., Stojanovic, N.,
Subramanian, A., Wyman, D., Rogers, J., Sulston, J., Ainscough, R., Beck, S., Bentley,
D., Burton, J., Clee, C., Carter, N., Coulson, A., Deadman, R., Deloukas, P., Dunham,
A., Dunham, I., Durbin, R., French, L., Grafham, D., Gregory, S., Hubbard, T.,
Humphray, S., Hunt, A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer,
S., Milne, S., Mullikin, J.C., Mungall, A., Plumb, R., Ross, M., Shownkeen, R., Sims,
S., Waterston, R.H., Wilson, R.K., Hillier, L.W., McPherson, J.D., Marra, M.A.,
Mardis, E.R., Fulton, L.A., Chinwalla, A.T., Pepin, K.H., Gish, W. R., Chissoe, S.L.,
Wendl, M.C., Delehaunty, K.D., Miner, T.L., Delehaunty, A., Kramer, J.B., Cook,
L.L., Fulton, R.S., Johnson, D.L., Minx, P.J., Clifton, S.W., Hawkins, T., Branscomb,
E., Predki, P., Richardson, P., Wenning, S., Slezak, T., Doggett, N., Cheng, J.F., Olsen,
A., Lucas, S., Elkin, C., Uberbacher, E., Frazier, M. et al., Initial sequencing and
analysis of the human genome, Nature, 409, 860–921, 2001.

3. Pietu, G., Mariage-Samson, R., Fayein, N.A., Matingou, C., Eveno, E., Houlgatte,
R., Decraene, C., Vandenbrouck, Y., Tahi, F., Devignes, M.D., Wirkner, U., Ansorge,
W., Cox, D., Nagase, T., Nomura, N., and Auffray, C., The Genexpress IMAGE
knowledge base of the human brain transcriptome: a prototype integrated resource
for functional and computational genomics, Genome Research, 9, 195–209, 1999.
4. de Chaldee, M., Gaillard, M.C., Bizat, N., Buhler, J.M., Manzoni, O., Bockaert, J.,
Hantraye, P., Brouillet, E., and Elalouf, J.M., Quantitative assessment of transcriptome
differences between brain territories, Genome Research, 13, 1646–1653, 2003.
5. Gustincich, S., Batalov, S., Beisel, K.W., Bono, H., Carninci, P., Fletcher, C.F.,
Grimmond, S., Hirokawa, N., Jarvis, E.D., Jegla, T., Kawasawa, Y., LeMieux, J.,
Miki, H., Raviola, E., Teasdale, R. D., Tominaga, N., Yagi, K., Zimmer, A., Hayash-
izaki, Y., and Okazaki, Y., Analysis of the mouse transcriptome for genes involved in
the function of the nervous system, Genome Research, 13, 1395–1401, 2003.
6. Pace-Schott, E.F. and Hobson, J.A., The neurobiology of sleep: genetics, cellular
physiology and subcortical networks, Nat. Rev. Neurosci., 3, 591–605, 2002.
7. Gautvik, K.M., de Lecea, L., Gautvik, V.T., Danielson, P.E., Tranque, P., Dopazo, A.,
Bloom, F.E., and Sutcliffe, J.G., Overview of the most prevalent hypothalamus-
specific mRNAs, as identified by directional tag PCR subtraction, Proc. Natl. Acad.
Sci. USA, 93, 8733–8738, 1996.
8. de Lecea, L., Kilduff, T.S., Peyron, C., Gao, X., Foye, P.E., Danielson, P.E., Fukuhara,
C., Battenberg, E.L., Gautvik, V.T., Bartlett, F.S., 2nd, Frankel, W.N., van den Pol,
A.N., Bloom, F.E., Gautvik, K.M., and Sutcliffe, J.G., The hypocretins:
hypothalamus-specific peptides with neuroexcitatory activity, Proc. Natl. Acad. Sci.
USA, 95, 322–327, 1998.
9. Willie, J.T., Chemelli, R.M., Sinton, C.M., and Yanagisawa, M., To eat or to sleep?
orexin in the regulation of feeding and wakefulness, Annu. Rev. Neurosci., 24,
429–458, 2001.
10. Sutcliffe, J.G. and De Lecea, L., The hypocretins: setting the arousal threshold, Nat.
Rev. Neurosci., 3, 339–349, 2002.
11. Hungs, M. and Mignot, E., Hypocretin/orexin, sleep and narcolepsy, Bioessays, 23,
397–408, 2001.
12. McCarley, R.W. and Hobson, J.A., Neuronal excitability modulation over the sleep
cycle: a structural and mathematical model, Science, 189 (4196), 58–60, 1975.
13. Peyron, C., Tighe, D.K., van den Pol, A.N., de Lecea, L., Heller, H.C., Sutcliffe, J.G.,
and Kilduff, T.S., Neurons Containing Hypocretin (Orexin) Project to Multiple Neu-
ronal Systems, J. Neurosci., 18, 9996–10015, 1998.
14. Date, Y., Ueta, Y., Yamashita, H., Yamaguchi, H., Matsukura, S., Kangawa, K.,
Sakurai, T., Yanagisawa, M., and Nakazato, M., Orexins, orexigenic hypothalamic
peptides, interact with autonomic, neuroendocrine and neuroregulatory systems [In
Process Citation], Proc. Natl. Acad. Sci. USA, 96, 748–753, 1999.
15. Hagan, J.J., Leslie, R.A., Patel, S., Evans, M.L., Wattam, T.A., Holmes, S., Benham,
C.D., Taylor, S.G., Routledge, C., Hemmati, P., Munton, R.P., Ashmeade, T.E., Shah,
A.S., Hatcher, J.P., Hatcher, P.D., Jones, D.N., Smith, M.I., Piper, D.C., Hunter, A.J.,
Porter, R.A., and Upton, N., Orexin A activates locus coeruleus cell firing and
increases arousal in the rat, Proc. Natl. Acad. Sci. USA, 96, 10911–10916, 1999.
16. Horvath, T.L., Peyron, C., Diano, S., Ivanov, A., Aston-Jones, G., Kilduff, T.S., and
van Den Pol, A.N., Hypocretin (orexin) activation and synaptic innervation of the
locus coeruleus noradrenergic system, J. Comp. Neurol., 415, 145–159, 1999.

17. Bourgin, P., Huitrón-Reséndiz, S., Spier, A., Fabre, V., Morte, B., Criado, J., Sutcliffe,
J., Henriksen, S., and de Lecea, L., Hypocretin-1 modulates REM sleep through
activation of locus coeruleus neurons, J. Neurosci., 20, 7760–7765, 2000.
18. Bayer, L., Eggermann, E., Serafin, M., Saint-Mleux, B., Machard, D., Jones, B., and
Muhlethaler, M., Orexins (hypocretins) directly excite tuberomammillary neurons,
Eur. J. Neurosci., 14, 1571–1575, 2001.
19. Brown, R.E., Sergeeva, O.A., Eriksson, K.S., and Haas, S.L., Orexin A excites
serotonergic neurons in the dorsal raphe nucleus of the rat, Neuropharmacology, in
press, 2001.
20. Eriksson, K.S., Sergeeva, O., Brown, R.E., and Haas, H.L., Orexin/hypocretin excites
the histaminergic neurons of the tuberomammillary nucleus, J. Neurosci., 21,
9273–9279, 2001.
21. Huang, Z.L., Qu, W.M., Li, W.D., Mochizuki, T., Eguchi, N., Watanabe, T., Urade,
Y., and Hayaishi, O., Arousal effect of orexin A depends on activation of the hista-
minergic system, Proc. Natl. Acad. Sci. USA, 98, 9965–9970, 2001.
22. Xi, M., Morales, F.R., and Chase, M.H., Effects on sleep and wakefulness of the
injection of hypocretin-1 (orexin-A) into the laterodorsal tegmental nucleus of the
cat, Brain Res., 901, 259–264, 2001.
23. Espana, R.A., Baldo, B.A., Kelley, A.E., and Berridge, C.W., Wake-promoting and
sleep-suppressing actions of hypocretin (orexin): basal forebrain sites of action,
Neuroscience, 106, 699–715, 2001.
24. Thakkar, M.M., Ramesh, V., Strecker, R.E., and McCarley, R.W., Microdialysis
perfusion of orexin-A in the basal forebrain increases wakefulness in freely behaving
rats, Arch. Ital. Biol., 139, 313–328, 2001.
25. Kilduff, T.S. and Peyron, C., The hypocretin/orexin ligand-receptor system: implica-
tions for sleep and sleep disorders, Trends Neurosci., 23, 359–365, 2000.
26. Torterolo, P., Yamuy, J., Sampogna, S., Morales, F.R., and Chase, M.H., Hypocret-
inergic neurons are primarily involved in activation of the somatomotor system, Sleep,
26, 25–28, 2003.
27. Scammell, T.E., Estabrooke, I.V., McCarthy, M.T., Chemelli, R.M., Yanagisawa, M.,
Miller, M.S., and Saper, C.B., Hypothalamic arousal regions are activated during
modafinil-induced wakefulness, J. Neurosci., 20, 8620–8628, 2000.
28. Estabrooke, I.V., McCarthy, M.T., Ko, E., Chou, T.C., Chemelli, R.M., Yanagisawa,
M., Saper, C. B., and Scammell, T., Fos expression in orexin neurons varies with
behavioral state, J. Neurosci., 21, 1656–1662, 2001.
29. Yoshida, Y., Fujiki, N., Nakajima, T., Ripley, B., Matsumura, H., Yoneda, H., Mignot,
E., and Nishino, S., Fluctuation of extracellular hypocretin-1 (orexin A) levels in the
rat in relation to the light-dark cycle and sleep-wake activities, Eur. J. Neurosci., 14,
1075–1081, 2001.
30. Alam, M.N., Gong, H., Alam, T., Jaganath, R., McGinty, D., and Szymusiak, R.,
Sleep-waking discharge patterns of neurons recorded in the rat perifornical lateral
hypothalamic area, J. Physiol., 538, 619–631, 2002.
31. de Lecea, L., Criado, J.R., Prospero-Garcia, O., Gautvik, K.M., Schweitzer, P., Daniel-
son, P.E., Dunlop, C.L., Siggins, G.R., Henriksen, S.J., and Sutcliffe, J.G., A cortical
neuropeptide with neuronal depressant and sleep-modulating properties, Nature, 381,
242–245, 1996.
32. Spier, A.D. and de Lecea, L., Cortistatin: a member of the somatostatin neuropeptide
family with distinct physiological functions, Brain Res. Rev., 33, 228–241, 2000.

33. de Lecea, L., del Rio, J.A., Criado, J.R., Alcantara, S., Morales, M., Henriksen, S.J.,
Soriano, E., and Sutcliffe, J. G., Cortistatin is expressed in a distinct subset of cortical
interneurons, J. Neurosci., 17, 5868–5880, 1997.
34. Whittington, M.A., Traub, R.D., and Jefferys, J.G., Synchronized oscillations in
interneuron networks driven by metabotropic glutamate receptor activation, Nature,
373, 612–615, 1995.
35. Traub, R.D., Whittington, M.A., Stanford, I.M., and Jefferys, J.G., A mechanism for
generation of long-range synchronous fast oscillations in the cortex, Nature, 383,
621–624, 1996.
36. Traub, R.D., Jefferys, J.G., and Whittington, M.A., Simulation of gamma rhythms in
networks of interneurons and pyramidal cells, J. Comput. Neurosci., 4, 141–150,
1997.
37. Jefferys, J.G. and Whittington, M.A., Review of the role of inhibitory neurons in
chronic epileptic foci induced by intracerebral tetanus toxin, Epilepsy Res., 26, 59–66,
1996.
38. Amzica, F. and Steriade, M., Electrophysiological correlates of sleep delta waves,
Electroencephalogr. Clin. Neurophysiol., 107, 69–83, 1998.
39. Joho, R.H., Ho, C.S., and Marks, G.A., Increased gamma- and Decreased delta-
Oscillations in a Mouse Deficient for a Potassium Channel Expressed in Fast-Spiking
Interneurons, J. Neurophysiol., 82, 1855–1864, 1999.
40. Lin, L., Faraco, J., Li, R., Kadotani, H., Rogers, W., Lin, X., Qiu, X., de Jong, P.J.,
mutation in the hypocretin (orexin) receptor 2 gene, Cell, 98, 365–376, 1999.
41. Chemelli, R.M., Willie, J.T., Sinton, C.M., Elmquist, J.K., Scammell, T., Lee, C.,
Richardson, J.A., Williams, S.C., Xiong, Y., Kisanuki, Y., Fitch, T.E., Nakazato, M.,
Hammer, R.E., Saper, C.B., and Yanagisawa, M., Narcolepsy in orexin knockout
mice: molecular genetics of sleep regulation, Cell, 98, 437–451, 1999.
42. Willie, J.T., Chemelli, R.M., Sinton, C.M., Tokita, S., Williams, S.C., Kisanuki, Y.Y.,
Marcus, J.N., Lee, C., Elmquist, J.K., Kohlmeier, K.A., Leonard, C.S., Richardson,
J.A., Hammer, R.E., and Yanagisawa, M., Distinct narcolepsy syndromes in Orexin
receptor-2 and Orexin null mice: molecular genetic dissection of Non-REM and REM
sleep regulatory processes, Neuron, 38, 715–730, 2003.
43. Hara, J., Beuckmann, C.T., Nambu, T., Willie, J.T., Chemelli, R.M., Sinton, C.M.,
Sugiyama, F., Yagami, K., Goto, K., Yanagisawa, M., and Sakurai, T., Genetic ablation
of orexin neurons in mice results in narcolepsy, hypophagia, and obesity, Neuron,
30, 345–354, 2001.
44. Burton, F.H., Hasel, K.W., Bloom, F.E., and Sutcliffe, J. G., Pituitary hyperplasia and
gigantism in mice caused by a cholera toxin transgene, Nature, 350, 74–77, 1991.
45. Zhang, J., Campbell, R. E., Ting, A. Y., and Tsien, R. Y., Creating new fluorescent
probes for cell biology, Nat. Rev. Mol. Cell Biol., 3, 906–918, 2002.
46. Li, Y., Gao, X.B., Sakurai, T., and van den Pol, A.N., Hypocretin/Orexin excites
hypocretin neurons via a local glutamate neuron-A potential mechanism for orches-
trating the hypothalamic arousal system, Neuron, 36, 1169–1181, 2002.
47. DeFalco, J., Tomishima, M., Liu, H., Zhao, C., Cai, X., Marth, J.D., Enquist, L., and
Friedman, J.M., Virus-assisted mapping of neural inputs to a feeding center in the
hypothalamus, Science, 291, 2608–2013., 2001.
48. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J.M., Adams, J.A., Ikura, M., and
Tsien, R.Y., Fluorescent indicators for Ca2+ based on green fluorescent proteins and
calmodulin, Nature, 388, 882–887, 1997.

49. Miyawaki, A., Griesbeck, O., Heim, R., and Tsien, R.Y., Dynamic and quantitative
Ca2+ measurements using improved cameleons, Proc. Natl. Acad. Sci. USA, 96,
2135–2140, 1999.
50. Mao, B.Q., Hamzei-Sichani, F., Aronov, D., Froemke, R.C., and Yuste, R., Dynamics
of spontaneous activity in neocortical slices, Neuron, 32, 883–898, 2001.
51. Li, W.H., Parigi, G., Fragai, M., Luchinat, C., and Meade, T.J., Mechanistic studies
of a calcium-dependent MRI contrast agent, Inorg. Chem., 41, 4018–4024, 2002.

7 Genetic Regulation
of Sleep
Yves Dauvilliers, Paul Franken, and Mehdi Tafti
CONTENTS
Introduction
Evidence for a Genetic Contribution to Sleep
Candidate Gene Studies
Mutagenesis Studies
Quantitative Trait Loci (QTL) Studies
QTL Analysis of Sleep Duration, Distribution, and Architecture
QTL Analysis of the Sleep EEG
QTL Analysis of Homeostatic Regulation of Sleep
Gene Expression Studies
Genetics of Circadian Rhythms
Conclusions
Acknowledgments
References
INTRODUCTION
The behavior sleep is conserved across many species including birds and mammals.
Sleep-like behavior has also been characterized in invertebrates,1 notably the fruit
fly.2,3 Sleep is of vital importance, although the neurobiological substrates of the
functions of sleep remain elusive. Substantial progress has been achieved during the
last two decades in our understanding of neurobiology underlying the expression
and regulation of sleep. Sleep is regulated by two major processes, one circadian
that determines its timing and one homeostatic that determines its need.4 The neu-
ronanatomical and neurochemical pathways involved in sleep initiation and main-
tenance are now well described5 but, in contrast to impressive advances in the
molecular genetics of circadian rhythms, little is known about the molecular bases
of sleep. We do know that the expression and the regulation of most of sleep
components are under genetic control.
Mignot and colleagues identified a mutation in the hypocretin receptor 2 gene
as the cause of canine form of narcolepsy,6 a sleep disorder found in several species,
including humans. Based on this major discovery, the role of the hypocretin system
0-8493-1519-0/05/$0.00+$1.50

in human narcolepsy is now well established.7,8 This example demonstrated that with
a genetic approach to sleep, unexpected molecular pathways remain to be discovered.
Another more recent example of a successful forward genetics approach applied to
sleep is the discovery that the metabolic fatty acid beta-oxidation pathway is impli-
cated in the regulation of theta oscillations during sleep.9 This pathway has not been
previously implicated in sleep and may also play a role in cognitive functions. The
genetic dissection of sleep therefore constitutes a promising approach in understand-
ing the molecular basis of sleep physiology and sleep disorders. Hence, the contri-
bution of genetic components to the pathology of sleep disorders is increasingly
recognized as of major importance. In this review we will focus on advances toward
genetic approaches to sleep in mice, although it is recognized that the fruit fly offers
a powerful alternative.
EVIDENCE FOR A GENETIC CONTRIBUTION TO SLEEP

For a wide range of phenotypes, considerable variation has been observed among
species, strains, and individuals within a species. These phenotypes include many
aspects of sleep such as the daily amount of rapid-eye-movement sleep (REMS) and
non-REMS (NREMS) and their distribution across the day. These variations may
be attributed to environmental factors such as light, temperature, diet, and behavioral
conditioning. The role of these factors in sleep is well established, in terms of both
physical and behavioral effects. These factors, however, do not fully account for the
variability observed, and the differences in sleep-related phenotypes in animals and
humans suggest the presence of an important contribution of genetic factors. The
implication of genetic factors is strongly indicated by the many interindividual
variations in the different aspects of sleep in laboratory animals kept under identical
environmental conditions from birth. Significant differences in NREMS and REMS
have been observed in inbred strains of rodents, with far greater interstrain than
intrastrain variability, suggesting that environmental factors play a less important
role.10–17 Furthermore, these differences are highly resistant to prolonged manipula-
tions such as immobilization, forced activity, or sleep deprivation.18,19
Sleep is a complex behavior both in its manifestation and its regulation. The
various aspects of sleep differ in their regulation, and each of these aspects is likely
to be under genetic control. Each component of sleep needs to be considered as a
complex phenotype. A systematic genetic approach is therefore needed for their
identification.20,21 Early work on waking EEG recordings by Vogel22,23 had strongly
suggested the effect of single genes. Pioneering work by Valatx in inbred mice had
also indicated that several aspects of sleep are controlled by genetic factors.17,24
These studies, together with a diallelique sleep experiment in mice performed by
Friedmann,12 clearly indicated that although some aspects of sleep may follow a
simple segregation, the classical genetic laws cannot predict most others.
As a first step to identify some of the underlying genes, most studies have focused
on sleep of pure inbred strains, mainly of mice and rats. Mouse models constitute
the best animal tool for discovering genes related to complex behavior such as
sleep.25 Mice are easy to breed and study, and high-density genetic marker maps
and the sequenced mouse genome are now available. Moreover there are numerous

other mapping panels, such as the recombinant inbred (RI) lines for which genome-
wide markers have been genotyped.26
A pure strain is established by successive inbred crossing, and the genome is
considered as being at homozygous state after 60 generations of brother-sister
mating. All the individuals of a specific strain thus have the same genetic combina-
tion, unlike outbred lines. To follow the segregation of a phenotype, two inbred
mouse strains differing in the trait of interest are crossed, and their F1 offspring are
either intercrossed to generate F2 or backcrossed to one of the progenitor strains to
generate backcross populations. Further random intercross or backcross generations
can be performed to generate advanced inter- and backcross populations. To generate
RI sets, F2 mice are brother-sister mated until full homozygosity, thereby fixing a
unique set of recombinations in several inbred lines.
Heterogenous stocks are generated by intercrossing several inbred mouse strains
over many generations and therefore represent higher rates of recombination and
polymorphism useful for fine mapping.27 The first step is thus to compare several
inbred strains to identify the differences for a given phenotype. It is usually necessary
to combine several approaches to localize a gene and determine its function.
Genome-wide search for genes affecting a phenotype of interest, through quantitative
trait loci-QTL analysis, mutagenesis, molecular genetic, and candidate gene studies
are the most commonly used strategies.
CANDIDATE GENE STUDIES

Candidate gene studies address whether a gene, which is already known, is also
implicated in sleep regulation. The known physiological role of a gene may lead to
its suspected implication to account for phenotype variations between different
genotypes. A study of the polymorphism of this gene will look for a relationship
(association) between an allele and the phenotype. The use of genetically modified
mice has provided efficient tools to study how genes are implicated in sleep and
appears to have a particularly promising future. This approach consists of construct-
ing lines of transgenic animals whose genetic material has been altered by either
adding supplementary copies of a given gene (transgenic), or by changing the gene
of interest (knockout, knockin).28,29 These techniques can be used to address the
consequences of overexpression, ectopic expression, time- and tissue-specific
expression, and gain or loss of function of a candidate gene. Potential advantages
and problems have been discussed in several reviews.20,21,30 If the mutant animal is
viable, analysis of its phenotype provides information about the normal function of
the modified gene. It is then easy to study the sleep of these transgenic mice and
draw conclusions about their implication in sleep regulation.
Sleep physiology and pharmacology have identified most of the essential systems
from which candidate genes can be chosen. The sleep of several transgenic mice
has been studied, each of which shows abnormalities in terms of sleep architecture.
Studies in mice transgenic for prion protein (PrP) suggest that PrP plays a role in
promoting sleep continuity.31 Also, after sleep deprivation slow-wave activity
increased, but the changes in EEG power density were more prominent and lasted
longer in the PrP knockout mice.31

Sleep of transgenic mice producing an excess of growth hormone shows an

increase in NREMS; in contrast NREMS was significantly suppressed during both
the light and the dark period in the transgenic mice with deficiency in the somato-
tropic axis without any changes in REMS.32,33 Sleep of TNF receptor knockout mice
shows an increase in NREM and a decrease in REMS after interleukin-1beta treat-
ment, results providing new evidence that TNF alpha is involved in physiological
sleep regulation.34 Respectively, 5-HT1B receptor, GH-IGF-1, and dopamine trans-
porter (DAT) knockout mice demonstrate a reduction in NREMS, thus testifying to
their likely implication in the regulation of NREMS.32,35,36 REMS for its part was
significantly increased in 5-HT1B-R knockout mice in contrast to the absence of
change in GHRH knockout mice. Transgenic mice overexpressing prostaglandin D2
(PGD2) showed a significant increase in NREMS after nociceptive stimulation
(pinching the tail) without alterations in REMS.37 These findings appear to be related
to a local inflammation, linked to an increase in the level of cerebral PGD2 release.
Sleep of albumin-D binding protein (DBP) knockout mice showed a reduction in
circadian amplitude of NREMS and alterations in REMS.38
All these studies suggest a potential role in sleep for each of the investigated
genes; however, the invariable presence of sleep abnormalities, irrespective of the
transgenic model studied, probably indicates a nonspecific effect of the different
genetic manipulations or the high sensitivity of sleep to diverse alterations in general
physiology. The first sleep-related gene discovery concerns the orexin (hypocretin)
system. Orexin-A (or 1) and -B (or 2) are hypothalamic neuropeptides acting on
orexin-A and -B receptors and first thought to be involved in feeding behavior.39–41
Mignot’s group identified, through linkage analysis and positional cloning, muta-
tions in the orexin-B receptor as the cause of canine narcolepsy,6 an animal model
of the human sleep disorder narcolepsy. Almost simultaneously Yanagisawa’s group,
interested in the role of orexins in feeding behavior, discovered in the mouse a
phenotype similar to canine and human narcolepsy after a targeted deletion of the
prepro-orexin gene.42
The lack of discovery of new genes could be a limitation inherent to the strategy
of the candidate gene approach, but studies in clock-gene knockout and mutant flies
and mice revealed a potential new pathway of transcriptional regulators in sleep
homeostasis. The genome-wide search or mapping experiments make no a priori
assumptions on gene systems involved and, although the outcome may favor a
known physiological mechanism, unknown and unexpected systems may be dis-
covered. The systematic search for genes affecting a particular phenotype needs to
cover the whole genome of an organism. The hypocretin success story in canine
narcolepsy is the unique and the best example of this approach in the field of sleep
research; therefore genome-wide search constitutes the method of choice if we are
to discover new sleep genes. Another approach, the genome-wide mutagenesis, is
also of special interest.
MUTAGENESIS STUDIES
Mutagenesis is an important strategy to search for new genes implicated in sleep.
Its aim is primarily to produce mutants, which present abnormalities in sleep or

circadian rhythms, then to localize the defective genes and identify them by means
of candidate gene analysis or positional cloning.43,44 Ethylnitrosurea (ENU) is the
mutagen most commonly used to randomly induce point mutations. Both dominant
and recessive mutations can be screened in the same way as a single gene mutation
in a pathological condition. This technique has already been used successfully in
the field of circadian rhythms and led to the discovery of the gene Clock, one of the
key mammalian genes in circadian rhythmicity.45 More recently, with the same
technique, Rab3a gene (coding for the most abundant ras-associated binding brain
protein) was found to alter both circadian period and homeostatic response to sleep
loss in the mouse.46
Nevertheless, the relatively high number of animals, which have to be studied
to isolate abnormal sleep phenotypes of interest, limits the feasibility of this
approach. Moreover genetic screens by mutagenesis are for fully penetrant dominant
and recessive mutations and therefore cannot identify small-effect sequence varia-
tions that may turn out to be essential for some aspects of the phenotype.
QUANTITATIVE TRAIT LOCI (QTL) STUDIES

A number of limitations inherent to each of the techniques described above com-
plicate the process of identifying new genes implicated in sleep physiology. The
Quantitative Trait Loci (QTL) analysis was developed to overcome some of these
obstacles, and it consists of identifying all the loci controlling a given quantitative
trait and responsible for interindividual differences (even small).29,47 Sleep physiol-
ogy and regulation are particularly complex and probably involve numerous genes
and many interactions among these genes and environmental factors. QTL analysis
has been proposed as a powerful approach in the genetic dissection of complex
traits.16,20,47–53 The QTL technique is particularly appropriate when the trait is com-
plex and the number of genes is high. The QTL analysis effectively allows numerous
loci to be identified, some of which may have a major effect and others a minor
effect on the different phenotypes related to sleep. This analysis can be performed
in several segregating mouse populations including inter- and backcross, advanced
inter- and backcross, RI and heterogeneous stocks. Although RIs are usually not
suitable for QTL mapping due to their limited progenitor strains and number, for
QTLs of large effects they may provide significant mapping accuracy because of
the fourfold increase in recombination as compared to an F2 population.49 Also, for
complex phenotypes with high variability, the use of RIs is advantageous because
several individuals per strain are tested instead of, for instance, a single F2 or
backcross animal.
QTL analysis does not map a gene but a genetic effect in a large chromosomal
region (usually 20–30 cM). These large regions may contain a single gene or several
genes with variable effects. QTLs may interact with each other (epistasis), an effect
which is difficult to detect in QTL mapping experiments. The first step is to make
sure that the region contains QTLs of large enough effect. When a QTL has been
localized in RI strains, segregating F2 mice is the strategy of choice. Another
possibility is to transfer the QTL region from one inbred strain background to another
inbred background through repeated backcrossing and selection to produce a con-

genic strain. Sleep amounts available in eight congenic strains generated by trans-
ferring histocompatibility genes from the inbred strain BALB/c (C) to the inbred
background of C57BL/6 (B) were analyzed.54 In almost all cases, the results indicated
that even if the transferred pieces of chromosomes contained a large effect QTL
detected in CXB-RIs, a clear effect could not be observed in the resulting congenic
strain. The different loci (QTLs) may interact (epistasis), producing a variable QTL
effect between genetically different strains.55 Several subsequent crossing experi-
ments are necessary to obtain recombinant animals and thereby reduce the size of
the region of interest, and the responsible genes are finally determined by means of
either candidate genes or positional cloning techniques.49
Although natural allelic variation of genes with small effect can be mapped
through QTL analysis, the final identification of sequence variants (quantitative trait
nucleotides or QTNs) in the QTL region with biologically significant effects on the
phenotype may represent prohibitive efforts in terms of both phenotyping and geno-
typing.49,53 The final identification of functional QTNs is the most difficult part since
QTNs may be found every few kilo bases. Therefore a combination of several
approaches is necessary for mapping and candidate gene analysis. High-resolution
QTL mapping in conjunction with the availability of whole genome sequences of
several major mouse strains should identify candidate genes to be investigated.
Because most QTNs will probably be involved in gene regulation rather than being
mutations affecting the protein function, further gene expression profiling with high
throughput genomics technologies (e.g., microarray or TaqMan), gene translation,
and post-translational protein analyses should be used to uncover the molecular
mechanisms involved.
An excellent example of how QTL analysis can further our understanding of
complex traits was provided by Takahashi’s group for the circadian behavior in
mice.55 Although most of the molecular machinery of the circadian timekeeping
system has been discovered mainly by direct molecular techniques and mutagenesis,
the identified genes do not explain the complexity of the observed circadian behavior.
For instance none of the clock genes has been found to be involved in the approx-
imately 1-hour difference in free-running circadian period between BALB/c and
C57BL/6 inbred mouse strains.56 Instead QTL analysis in a BALB/c x C57BL/6
intercross revealed several loci with epistatic interaction.55
QTL ANALYSIS OF SLEEP DURATION,

DISTRIBUTION, AND ARCHITECTURE
The complex nature of sleep is reflected in high intra- and interspecies phenotypic
variability. The 24-hour amount of sleep is highly dependent on the environment.57
Hence many animals of the same strain and of different strains need to be recorded
in strictly similar experimental conditions (light, temperature, diet) to verify whether
interline variability is clearly higher than intraline variability and ultimately the
presence of an underlying genetic factor.
Pharmacogenetics studies have been first applied in this field; mouse strains LS
and SS (long and short sleepers) were initially the most studied models.58 These

mouse strains were created in the 1970s by selecting mice that were the most or
least sensitive to the hypnotic effect of ethanol by testing their righting reflex. After
18 generations of selection, the resulting strains had a mean sleep time of 10 minutes
(SS) or 2 hours (LS), respectively, after similar doses of ethanol. Analyses of the
sensitivity of these strains to different pharmacological compounds determined to
what extent genetic control of the soporific or anesthetic effect of these hypnotics
was similar or different.58–61 Also the different pharmacological effects of alcohol
(sedation, hypothermia, toxicity) appeared to be controlled by different genes.58,60
QTL analysis in SS X LS hybrids traced intrastrain variations in the differences in
alcohol sensitivity to at least seven or eight loci.62,64 More recent studies have
confirmed this result with a total of seven QTLs, accounting for 60% of the variation
between SS and LS strains.63,64 These pharmacogenetics studies are nevertheless
considerably restricted: The relationship between these mouse models and the
genetic control of alcohol-induced sleep is uncertain because sleep and circadian
rhythms in LS and SS animals have not been studied. Moreover the effects of
benzodiazepines and alcohol on sleep appear to be indirect and highly dependent
on prior sleep and waking history.
The 24-hour amount of sleep shows highly significant differences between
inbred mouse strains; for example AKR mice sleep approximately 3 hours more
than DBA mice over the 24-hour day.11 Obviously, as for the difference in the
period of circadian rhythms, not a single gene may be found to account for this
difference but many genes with complex interactions. A series of experiments have
been initiated to dissect different phenotypic aspects of sleep in mice through QTL
analysis. The distribution of NREM and REM sleep time over 24 hours also varies
according to the genetic background. AKR, C57BL/6 and C57BR strains are
characterized by long episodes of REMS and C57BL/6, BALB/c and 129/Ola
strains had the longest episodes of NREMS.11 At the other end of the spectrum,
the DBA/2 strain is characterized by short episodes of NREM and REM sleep
resulting in a very fragmented sleep.11 Genetic studies in F1 and F2 mice also
indicate the complex nature of the genetic control of these parameters, implicating
the presence of several genes. Regulation of NREMS clearly differs from that of
REMS; the genetic control of these two types of sleep probably reflects this
difference.11,16,20 A first QTL analysis of REMS identified several loci involved in
the variability between two inbred mouse strains (BALB/c and C57BL/6).16 The
loci were different for the duration of diurnal REMS (chromosome 7), nocturnal
REMS (chromosome 5, near the clock gene), and for total REMS time during 24
hours, suggesting that several genes are involved in the expression and regulation
of REMS. Another group using the same methodology reported other QTLs (on
chromosomes 4, 16, and 17) for differences in the diurnal amount of REMS
between the same mouse strains.65
A QTL analysis between two other mouse strains, C57BL/6 and DBA/2, found
another QTL on chromosome 1 with a highly significant effect on the amount of
REMs in the 12-hour light period.54 Numerous genes are therefore implicated in the
regulation of REMS; about 50% of the variance in REMS time is explained by the
presence of at least six different loci. It is worth noting that no significant QTL is
so far found for the amount of NREMS.16,20,54 One of the genomic regions relevant

to REMS regulation contains the candidate gene albumin-D binding protein (DBP).
This gene is a transcription factor expressed with strong circadian rhythmicity.66
DBP knockout mice are characterized by a reduction of their circadian period and
by an overall drop in locomotor activity. The study of their sleep, apart from revealing
a total sleep time identical to that of wild-type mice, showed a reduction in circadian
amplitude of NREMS, as well as alterations in REMS regulation.38
QTL ANALYSIS OF THE SLEEP EEG

The amplitude and frequency of rhythmic EEG activity (delta, theta, and alpha
oscillations) are well quantified using spectral analysis algorithms such as the Fast-
Fourier Transform. Spectral analysis of EEG activity during sleep has demonstrated
significant variations between different mouse strains for both NREM and REM
sleep.9,11,67 EEG activity during NREMS is usually characterized by slow waves and
sleep spindles associated with high spectral power in the 0.5–4.5 Hz and the 12–15
Hz, respectively. The main EEG activity in REMS, especially in rodents, is in the
theta rhythm frequency range (5–9 Hz). The dominant frequency of theta oscillations
depends strongly on genetic background.
The theta peak frequency (TPF) is particularly slow (5.75–6.25 Hz) in A/J,
C3H/HeJ, AKR/J, and BALB/cByJ (C) strains, and fast (6.75–7.75 Hz) in DBA/2J,
SPRET/Ei, PL/J, LP/J, C57BL/6J (B), 129/Ola, and C57BR/cdJ strains.9,11,67 The
TPF during REMS shows the highest phenotypic difference between inbred strains
and almost all of the inter-strain variability can be attributed to genetic effects (F10,66
= 56, p < 2.10–8; heritability = 0.97).9 To follow the segregation of TPF, a slow (C)
and a fast (B) TPF strain were reciprocally crossed. TPF in CXB- and BXC-F1 mice
was similar to that of B6 and significantly different from C mice, indicating that the
C allele was recessive without maternal effect. An F2 population was generated by
crossing CXB-F1 mice to map the underlying genes. The distribution of TPF was
approximately normal, which prompted us to use this phenotype in a QTL analysis.
QTL interval mapping with 89 polymorphic markers identified a single, highly
significant location on midchromosome 5 (LOD score = 11.5, p < 4.10–2), explaining
over 65% of total variance.9 Accordingly a backcross population was generated by
crossing CXB-F1 mice to the progenitor strain C (CXB-BC). Based on the F2 data,
CXB-BC mice were classified as slow or fast TPF, if 6.5 Hz or slower, or 6.75 Hz
or faster. Polymorphic markers of chromosome 5 were genotyped in these CXB-
BCs and a strong linkage to D5Mit240 was found (LOD score = 8.8, p <2.10–10).
Genotyping of 200 additional CXB-BC mice and selective sleep recording in 31
recombinant mice narrowed the region of interest to a 2.4 cM interval. Candidate
genes within the region included the short-chain acyl-coenzyme A dehydrogenase
(Acads).68 A spontaneous mutation in Acads appeared in the BALB/cByJ subline
with a highly significantly difference in TPF when compared to the BALB/cBy wild-
type subline (7.10 ± 0.22 Hz).9
There was no correlation between waking and REMS TPF, and there was no
significant linkage between the waking TPF and any markers of chromosome 5,
suggesting that the Acads mutation has a highly specific effect on TPF during sleep

only. Waking TPF showed no significant difference between the mutant and the wild
type BALB/cBy (mutant = 7.6 ± 0.27 Hz, wild type = 7.7 ± 0.36 Hz), clearly
indicating that the Acads mutation affects theta oscillations only during sleep.9 This
unexpected finding indicates a major role for the mitochondrial b-oxidation during
sleep, which is fatty acid chain-length specific because long-chain acyl-coenzyme
A dehydrogenase (Acadl) deficiency does not affect theta frequency.
High-density cDNA microarrays were then used to evaluate changes in the brain
gene expression caused by the Acads mutation.9 Glyoxalase I (Gl01) gene, involved
in metabolic detoxification, was identified as the only gene up-regulated in the brain
of Acads deficient mice. Increased Gl01 expression was evidenced in all inbred
strains that display slow theta oscillations during REMS. Both the slow theta and
the increased glyoxalase I expression could be partially reversed by acetyl-L-car-
nitine treatement, probably through detoxification of excess short-chain fatty acids.9
Brain short-chain fatty acid b-oxidation during sleep might represent a previously
unrecognized metabolic pathway in the adult brain with potential role in REMS and
brain maturation.
QTL ANALYSIS OF HOMEOSTATIC REGULATION

OF SLEEP
Two main processes regulate sleep: a circadian and a homeostatic process.69 Delta
power (activity in the 0.5–4 Hz range) is an EEG measure of NREMS intensity and
a reliable quantitative marker of NREMS homeostasis, thus, sleep loss evokes a
proportional increase in delta power, and excess sleep a decrease.70–72 The dynamics
of this homeostatically regulated process, referred to as Process S,4,69 have been
studied extensively, and mathematical simulations that quantify the relationship
between the sleep-wake distribution and delta power predicted the time course of
delta power remarkably well.73–75 The neurophysiological substrate of the homeostatic
process is unknown, however, and genetic studies might help to elucidate the nature
of what is depleted during wakefulness and recovered during NREMS.
A recent study involving six inbred mouse strains deprived of 6 hours of sleep,
in strictly similar circumstances, revealed significant differences in the intensity of
the homeostatic rebound of NREMS.74 The time constant for the accumulation of a
need for NREMS varied between genotypes but not for its decline during NREMS.
To confirm these predictions, a dose-response experiment was performed in which
sleep was deprived of for varying durations (i.e., dose) to verify that the delta power
increase in subsequent NREMS (i.e., response) depends on both the duration of prior
wakefulness and genotype. The segregation of the rebound of delta power after sleep
deprivation in 25 BXD recombinant inbred strains by quantitative trait loci (QTL)
analysis was performed and established that additive genetic factors accounted for
67% of the total variance (i.e., heritability in inbred strains).74,76
After performing a genome-wide scan (788 MIT-markers), two genomic regions
were identified, a significant QTL was found on chromosome 13 (LOD = 3.57,
nominal correlation p <0.00005, genome-wide <0.01) and a suggestive QTL on
chromosome 2.74 The QTL on chromosome 13 has a large effect, explaining 49%

of genetic variance in delta power rebound and 33% of total phenotypic variance.
These findings were not related to or influenced by differences in REM or NREM
sleep time expressed during the period over which delta power was calculated.
These results demonstrate that the increase of NREMS need is under a strong
genetic control and provide a basis for identifying genes (significant QTLs) under-
lying NREMS homeostasis.
Homeostatic regulation of sleep has also been addressed with reversed genetic
studies. Parallel studies of the homeostatic regulation of both REM and NREM
sleep have been conducted in transgenic mice. Mice overexpressing growth hormone
show more REMS under baseline conditions but show normal recovery pattern after
sleep deprivation.33
Clock mutant mice, in addition to demonstrating important changes in circadian
sleep architecture, also present an alteration in homeostatic sleep regulation with a
decreased rebound in REMS after sleep deprivation.77 DBP knockout mice also
showed a decreased REMS rebound after sleep deprivation without any difference
in the rebound of NREMS.38 Mice lacking functional genes for the serotonin-2C
receptor,78 Cry 1 and 2,79 and Rab3a46 show an altered NREMS rebound after sleep
deprivation. In addition, double cry 1 and 2 knockout mice show a higher amount
of NREMS compared to wild-type controls79; therefore the loss of circadian genes
(at least Cry 1 and 2) does not only affect the circadian rhythms (see section titled
Genetics of Circadian Rhythms) but also the sleep homeostatic process.
Recently loss-of-function mutations in other circadian genes (Per, Tim, Clock,
and Cycle) in fruit flies were also found to be responsible for a higher sleep rebound
after sleep deprivation compared to wild-type flies.80 Cycle-mutant fruit flies show
a disproportionally larger sleep rebound and die after 10 hours of sleep deprivation,
although they are more resistant than other clock mutants to various stressors.80
GENE EXPRESSION STUDIES

The first attempt to identify the molecular basis of sleep has been based on the
assumption that there must be genes that change their expression as a function of
behavioral states (sleep versus wake) and time spent in a particular state. Because
sleep need and intensity are homeostatically regulated, sleep deprivation should lead
to a change in the expression of those genes, of which the products are necessary
for recovery or at least involved in sleep regulation. The genes studied in most detail
are the immediate early genes (IEG), generally the proto-oncogene c-Fos. The
changes occurring in the IEG expression are interesting in at least two respects: the
expression of c-Fos and the other IEG increases with neuronal activity,81 and the
level of expression of their proteins or RNA messengers (mRNA) in discreet brain
regions can be determined, thus providing a map of their variations as a function
of sleep or wakefulness. Moreover as IEG products are themselves transcription
factors, they also alter the expression of numerous other genes. During the day c-
Fos expression in the brain correlates directly with the rest/activity cycle. In noc-
turnal rodents, expression is high at night and low during the day in most regions
of the brain.82,83 This effect is reversed when the animals are deprived of sleep during
the day, provoking a sleep rebound the following night.83–85 Diurnal rodents present

the opposite profile of c-Fos expression, with high levels during the day and low
levels at night.
IEG genes have also been used to dissect the neuroanatomy of sleep and wake-
fulness more finely. The locus coeruleus, in particular, appears to play an important
role, not only because c-Fos expression changes according to states of sleep and
wakefulness in this nucleus, but also because the locus coeruleus appears to control
a large part of the c-Fos expression in the entire forebrain in the waking state.
Unilateral lesions of the locus coeruleus reduce c-Fos level during wakefulness
ipsilaterally rather than controlaterally.86,87 The reduced levels of c-Fos as well as
Ngfi-A in wakefulness on the lesioned side is comparable to the levels observed
during periods of prolonged sleep. Studies have also been conducted during phar-
macological REMS, induced by injecting cholinergic agonists into the pontine retic-
ular formation.
This manipulation activates the transcription of the c-Fos gene in several nuclei
implicated in the regulation of REMS.88,89 Recently an exception to this wakefulness
and high c-Fos levels correlation was objectified in certain cells in the ventrolateral
preoptic region, the neurons expressing high c-Fos levels during sleep.90 These
neurons probably play a key role in initiating NREMS. Despite such correlations,
the functional role of IEGs has yet to be established. In fact only two studies suggest
a direct c-Fos role in the regulation of sleep.86,89 The former concerns the observation
of a reduction in spontaneous sleep and in sleep rebound after deprivation in c-Fos
knockout mice. In the second study, c-Fos antisense oligonucleotide injections in
the medial preoptic region reduced c-Fos protein levels and increased wakefulness
the following day.
Other approaches are needed in gene expression during sleep, such as substrac-
tive hybridization, PCR differential display (cDNA display), cDNA microarrays
(DNA chips), or real-time RT-PCR (TaqMan) methods.29,87 The substractive hybrid-
ization method has been used on rats deprived of sleep for 24 hours.91 Four mRNA
clones were isolated with lower levels after sleep deprivation and six with higher
mRNA levels. An analysis of the structure of two of these clones identified neuro-
granin and dendrin proteins.92,93 Several laboratories have used molecular biology
techniques, sometimes with sleep deprivation, to determine alterations in the tran-
scriptional activity of several genes including growth factors. Thus variations in
mRNA levels over 24 hours of GHRH in the hypothalamus94 of BDNF and its
receptor in the hippocampus95 were demonstrated.
Adding sleep deprivation sometimes increased these variations. For example,
mRNA and GHRH levels and those of the adenosine A1 receptor, respectively,
increase at paraventricular and basal telencephalic level after sleep deprivation.96,97
Furthermore, mRNA and interleukin 1 beta protein only increased significantly in
the hypothalamus and the brainstem after sleep deprivation.98 Other experiments
have been carried out with selective REMS deprivation.99–101 Finally, corticostatin
and hypocretin proteins, strongly implicated in the control of different states of sleep
and wakefulness have been identified using mRNA screening approaches, demon-
strating the importance of this field of investigation.39,102 It should be noted that
paradoxically, prepro-hypocretin mRNA levels are not modified in the hypothalamus
after 6-hour sleep deprivation.103

The mammalian brain expresses roughly half of the estimated 30,000 genes, so
it is likely that many genes change their level of expression during the states of sleep
and wakefulness.104 These alterations in gene transcriptional activities might reflect
a change in neuronal activity, although the role of these genes in sleep regulation is
still difficult to ascertain. The sensitization of these variations by sleep deprivation
merely reflects the effects of prolonged wakefulness on the brain. In addition to
using sleep deprivation techniques, future studies should systematically take recov-
ery after sleep deprivation into account, considering the marked NREMS rebound,
which may also coincide with alterations in the level of gene expression.79,105 In any
case, these studies may reveal negative results because alterations in the system may
occur at a post-transcriptional level.
GENETICS OF CIRCADIAN RHYTHMS

Circadian rhythmicity is a virtually universal property in all unicellular organisms,
plants, and animals.106 The alterations in circadian rhythms related to variations in
our environment are now well established, with temperature and light able to affect
our internal biological clock. However a number of genetic factors have also been
implicated in regulating these rhythms. Behavioral genetic research on circadian
rhythms is one of the most advanced in biology. This is greatly facilitated by the
fact that circadian rhythms in mammals are generated or synchronized by a discrete
region of the hypothalamus, the suprachiasmatic nuclei (SCN).107 Lesioning of the
SCN abolishes all rhythmicity, and circadian fluctuations are restored by transplant-
ing fetal hypothalamic tissue to lesioned animals.108 This furthers the idea that the
SCN is necessary and sufficient for the generation of behavioral rhythms.
It can thus be demonstrated that the neuronal, metabolic, and neurochemical
activity of the SCN themselves vary in a circadian fashion even when the tissue is
isolated in vitro.107 When the activity of several of these neurons is recorded simul-
taneously, each cell presents a different circadian rhythmicity (period or phase) with
no apparent synchronized activity.109,110 More recently the cloning and characteriza-
tion of mammalian clock genes have revealed that they are expressed in a circadian
manner throughout the body and even in cultured cells and organs.111–113
It is now generally accepted that peripheral cells contain a circadian clock, which
is similar to the one present in SCN neurons, except that the latter seems to be self-
sustained. It is still unclear how the central SCN clock synchronizes these peripheral
clocks, albeit humoral signals appear to be crucial. Unlike the rhythms entrained by
the pacemaker SCN, some rhythms of peripheral origin may be entrained by our
eating habits (dominant Zeitgeber for peripheral circadian oscillators).113 The phase
of peripheral clocks can be completely uncoupled from the SCN by restricted
feeding. The glucocorticoid hormones seem to inhibit the uncoupling of peripheral
and central circadian oscillators by altered feeding time.113
Genetic research in this field has already led to the isolation of several genes,
whose mutations in the fruit fly Drosophila melanogaster can considerably alter the
period of circadian rhythms or completely abolish rhythmicity as a function of the
implicated allele.114–118 In mammals, notably in mice, considerable variations exist
in circadian period length among inbred lines: Some are particularly long, others

particularly short117,119; i.e., C57BL and BALB/c mice have a 1-hour period differ-
ence.55,56 Numerous genetic factors are likely to be involved to account for these
different phenotypes. Mutations responsible for altering circadian rhythmicity have
been reported in mammals.106,117,119,121
Two genes that are essential for the production of behavioral rhythmicity, Clock
(Circadian Locomotor Output Cycles Kaput) and Wheel, have indeed been isolated,
both resulting from a mutagenesis screens in mice using the ENU mutagen.117,121
The mutant Wheel gene (chromosome 4 of the mouse) exerts a dominant effect and
causes complex neurological disruption associating hyperactivity, rotating behavior,
and circadian rhythmicity.121 The mutant Clock gene (chromosome 5 of the mouse)
exerts a semidominant effect, responsible only for altering the circadian period,
which becomes abnormally long.117 These mutant mice are capable of following a
rhythm entrained by alternating light/dark but loose endogenous circadian rhythm
in constant environment (dark/dark). The Clock gene was functionally identified
using a combination of techniques including positional cloning45 and a transgenic
rescue approach.122 The CLOCK protein has sequence motifs with direct DNA
binding properties (the basic Helix Loop Helix or bHLH domain), enhancing its
implication in regulating the transcription of several genes. In another rodent species,
the golden hamster, a spontaneous semidominant mutation of the Tau gene has been
responsible for the isolated alteration of the circadian period.119 This mutation in the
homozygous state exclusively induces a shorter circadian period (approximately 20
hours) resistant to variations in alternating light and dark. This gene codes for a
protein belonging to the casein kinase Ie family; the mutation may be responsible
for deactivating the protein via its inability to fix and phosphorylize PER protein.123
After cloning the gene Clock, three homologues of the Drosophila Per gene,
coding for PERIOD protein, were isolated in the mouse and in humans, mPer1,
mPer2, and mPer3.124–127 These Per genes are expressed in several cerebral regions,
but significant rhythmic daily fluctuations are only found in the SCN, indicating
their implication in generating circadian rhythms.124,127,128 Only mPer1 and mPer2
appear to be strongly implicated because their knockout mice develop abnormal
circadian rhythmicity. mPer3 has only added effects.
Two homologues of the Drosophila Cry gene, mCry1 and mCry2, coding for
photoreceptive flavoproteins Cryptochromes, have been isolated in mammals with
demonstrated oscillatory activity,129,130 moreover these genes appear to have stronger
rhythmic activity in mice than in Drosophila, with highly altered circadian period-
icity in knockout mice. Proteins mCRY1 and -2 are transcription inhibitors of their
own genes, but also of mPER1, 2, and 3 via the protein complex CLOCK-BMAL1.125
More precisely mCRY1 and mCRY2 are nuclear proteins that interact with each of
the mPer proteins, translocate each mPER protein from cytoplasm to nucleus, and
are rhythmically expressed in the SCN. The mPER and mCRY proteins appear to
inhibit the transcriptional complex differentially. Analysis of Cryptochrome inhibi-
tion of CLOCK-BMAL1 mediated transcription shows that the inhibition is through
direct protein-to-protein interactions, independent of the PERIOD and TIMELESS
proteins. PER2 is a positive regulator of the BMAL1 loop; and Cryptochromes are
the negative regulators of the PER and CRY cycles involved in the negative limb of
the feedback loop. Hence two other genes affecting circadian rhythms have been

localized in mammals, Bmal1 (homologue of cyc gene in Drosophila) coding for

Bmal1 protein, which also belongs to the bHLH family131,132 and mTim gene. A
single mTim gene has been isolated in humans and mice, but its function, unlike its
homologue tim in Drosophila, is still unknown.133 Because knockout mice for mTim
die at the embryonic stage, it has not been possible to distinguish any transcriptional
rhythmic activity, as light does not appear to exert any effect on protein mTim, and
the nuclear translocation of mPer would depend on heterodimerization between
proteins mPer and mCry and not mTim.130
The role of Drosophila protein TIM seems to be equivalent to that of CRY in
mammals. Finally the two proteins CLOCK and BMAL1, which are activators of mPer
transcription and coded respectively by mClock and Bmal1, also have strong homology
with Drosophila genes dClock and Cyc; however the activity of these genes differs
between the two species. In Drosophila the expression of dClock follows a circadian
rhythmicity with oscillation regulated by the protein complex PER-TIM-CLOCK
inhibiting its own mRNA synthesis. In mice the transcriptional activity of Clock is not
rhythmical, contrary to the Bmal1 gene whose regulation is thought to be carried out
by mPer2. Recently the orphan nuclear receptor REV-ERB alpha (transcription factor)
was found to be a major regulator of cyclic Bmal1 transcription.134
The circadian REV-ERB alpha expression is controlled by components of the
general feedback loop in which BMAL1 and CLOCK, players of the positive limb,
activate transcription of the cryptochrome and period genes, components of the
negative limb. Thus REV-ERB alpha constitutes a molecular link through which
components of the negative limb drive antiphasic expression of components of the
positive limb. REV-ERB alpha influences the period length and affects the phase-
shifting properties of the clock, but it is not required for circadian rhythm generation.
On the whole, and irrespective of the species studied, Drosophila or mouse, Per and
Clock genes are central to the cellular machinery for circadian regulation. It is thus
likely that PER and CLOCK proteins work together in producing 24-hour rhythmic-
ity in the SCN.
The effect of the environment and, more precisely, the alternation between light
and darkness in regulating circadian rhythmicity is well known, and its mechanism
of action is beginning to be understood. In the mammalian retina, besides the
conventional rod-cone system, a melanopsin-associated photoreceptive system exists
that conveys photic information for accessory visual functions such as circadian
photoentrainment.135 Melanopsin is expressed in retinal ganglion cells (RGCs) which
are intrinsically photosensitive. Melanopsin knockout mice entrained to a light and
dark cycle, phase-shifted after a light pulse and increased circadian period when
light intensity increased; however the magnitude of these behavioral responses in
knockout mice was 40% lower than in wild-type mice.136 Although melanopsin is
not essential for the circadian clock to receive photic input, it contributes significantly
to the magnitude of photic responses.
The effect of the alternation between light and darkness also involves the tran-
scription regulation of several genes. It seems that light has a direct influence on
the TIM protein in Drosophila and CRY in mammals via phosphorylation, ubiquit-
ination, and finally degradation.129 The transcriptional activity of Per and Tim genes
is not altered by light, however the formation of the protein complex PER-TIM and

its nuclear translocation are clearly affected by light-dark alternation. But the inter-
mittent presence of light is not essential for the generation of rhythmic activity. The
duration of formation of the PER-TIM complex once in the nucleus appears to be
determined by two different endogenous processes: a self-regulation phenomenon
and the presence of the Double-Time (DBT) protein. DBT is capable of binding to
PER in vitro and in Drosophila cells, suggesting that a physical association of PER
and DBT regulates PER phosphorylation and accumulation in vivo.137 DBT belongs
to the same family as TAU protein, the casein kinase Ie family.123,138 The function
of these two proteins, in Drosophila and in mammals, appears to be relatively well
conserved; this kinase protein allows PER phosphorylation, the inhibition of its
translocation to the nucleus, and the reduction of its stability. The expression of
genes implicated in circadian regulation also occurs outside the central nervous
system, in Drosophila as in mammals.113,139
However, although these genes function independently of each other, they remain
photosensitive in Drosophila, contrary to the case for mammals. An accumulation
of the neuropeptide vasopressin relies on the presence of other proteins CLOCK
and CYC, and establishes the circadian phase via coordination of the rhythmic
activity of different neurons.112,140 In mammals numerous studies have pointed to
the presence of soluble factors diffusing from the SCN to other cerebral regions,
thus entraining sleep and wake rhythms and locomotor activity. These factors are in
the process of being identified. One of the factors, TGF alpha, was recently isolated
in relation to a yeast secretion tap system.141 This peptide appears to play an
inhibiting role in locomotion by acting on the subparaventricular zone. Lastly it
appears that certain peripheral circadian oscillators depend on food intake through
a hormonal glucocorticoid signal.113
CONCLUSIONS
Sleep is a complex behavior both in its manifestation and regulation, which can be
studied at many different levels. The complexity of sleep-wake regulation, in addi-
tion to the many environmental influences, implies a predisposing genetic deter-
minism that is beginning to be understood. Most of the current progress in the study
of sleep genetics comes from animal (mainly mice and Drosophila) studies. Multiple
approaches using both animal models and genetic techniques are needed to deter-
mine new sleep genes and molecular bases of sleep. Over the past few years, a
revolution in the understanding of the molecular basis of circadian rhythm genera-
tion has led to the identification of a number of core clock genes and the development
of feedback models that explain how these core components interact to generate a
circadian rhythm.
Recent progress in molecular genetics and the development of detailed human
genomic map have already led to the identification of genetic factors in the contri-
bution of the pathology of sleep disorders. At least eight human orthologs of mouse
core clock genes have been identified, and a mutation in hPer2 is responsible for
the autosomal-dominant familial advanced-sleep-phase syndrome in humans. In
addition, the successful identification of a mutation in the hypocretin-2 receptor
underlying canine narcolepsy, leading to the discovery of the hypothalamic hypo-

cretin (orexin) neurotransmitter system as a key target for human narcolepsy, is one
of the best examples of how a genetic approach can not only further our under-
standing of the pathophysiology of sleep disorders but also bring new insights into
sleep physiology.
ACKNOWLEDGMENTS
Y.D. is supported by the “Association pour l’Etude du Sommeil,” Montpellier-
France, M.T. is supported by The Swiss National Science Foundation and the Geneva
University Hospitals, P.F. is supported by The NIH Heart, Lung, and Blood Institute.
REFERENCES
1. Campbell, S.S., and Tobler, I., Animal sleep, a review of sleep duration across
phylogeny, Neurosci. Biobehav. Rev., 8, 269, 1984.
2. Hendricks, J.C. et al., Rest in Drosophila is a sleeplike state, Neuron, 25,129, 2000.
3. Shaw, P.J. et al., Correlates of sleep and waking in Drosophila melanogaster, Science,
287, 1834, 2000.
4. Daan, S., Beersma, D.G., and Borbely, A.A., Timing of human sleep, recovery process
gated by a circadian pacemaker, Am. J. Physiol., 246, 161, 1984.
5. PaceSchott, E.F., and Hobson, J.A., The neurobiology of sleep, genetics, cellular
physiology and subcortical networks, Nat. Rev. Neurosci., 3, 591, 2002.
6. Lin, L. et al., The sleep disorder canine narcolepsy is caused by a mutation in the
hypocretin (orexin) receptor 2 gene, Cell, 98, 365, 1999.
7. Willie, J.T. et al., To eat or to sleep? orexin in the regulation of feeding and wake-
fulness, Annu. Rev. Neurosci., 24, 429, 2001.
8. Taheri, S., and Mignot, E., The genetics of sleep disorders, Lancet. Neurol., 1, 242,
2002.
9. Tafti, M. et al., Deficiency in shortchain fatty acid betaoxidation affects theta oscil-
lations during sleep, Nat., Genet., 34, 320, 2003
10. Benca, R.M. et al., Rat strain differences in response to dark pulse triggering of REM
sleep, Physiol. Behav., 49, 83, 1991.
11. Franken, P., Malafosse, A., and Tafti, T., Genetic determinants of sleep regulation in
inbred mice, Sleep, 22, 155,1999.
12. Friedmann, J.K., A diallel analysis of the genetic underpinnings of mouse sleep,
Physiol. Behav., 12, 169, 1974.
13. Leung, C. et al., Heritalility of dark pulse triggering of REM sleep in rats, Physiol.
Behav., 52, 127, 1994.
14. Kitahama, K., and Valatx J.L., Instrumental and pharmacological REM sleep depri-
vation in mice, strain differences, Neuropharmacology, 19, 529, 1980.
15. Rosenberg, R.S. et al., Strain differences in the sleep of rats, Sleep, 10, 537, 1987.
16. Tafti, M. et al., Localization of candidate genomic regions influencing paradoxical
sleep in mice, Neuroreport, 8, 3755, 1997.
17. Valatx, J.L., Bugat, R., and Jouvet, M., Genetic studies of sleep in mice, Nature, 238,
226, 1972.
18. Van Twyver, H. et al., Effects of environment and strain differences on EEG and
behavioral measurement of sleep, Behav. Biol., 9, 105, 1973.

19. Webb, W.B., and Friedmann, J.K., Attempts to modify the sleep patterns of the rats,
Physiol. Behav., 6, 459, 1971.
20. Tafti, M., and Franken, P., Invited review, genetic dissection of sleep, J. Appl. Physiol.,
92, 1339, 2002
21. Franken, P, and Tafti, M., Genetics of sleep and sleep disorders. Front. Biosci., 8,
381, 2003.
22. Vogel, F., Über die Erblichkeit des Normalen Electroencephalogramms, Stuttgart,
Thieme, 1958.
23. Vogel, F., The genetic basis of the normal human electroencephalogram, Humange-
netik, 10, 91, 1970.
24. Valatx, J.L., and Bugat, R., Facteurs génétiques dans le déterminisme du cycle veille
sommeil chez la souris, Brain. Res., 69, 315, 1974.
25. Takahashi, J.S., Pinto, L.H., Vitaterna, M.H., Forward and reverse genetic approaches
to behavior in the mouse, Science, 264, 1724, 1994.
26. Bailey, D.W., Definition of inbred strains, in P.L. Altman and D.D. Katz, eds., Inbred
and Genetically Defined Strains of Laboratory Animals. Part 1. Mouse and Rat,
FASEB, Bethesda, 1979, p. 47.
27. Talbot, C.J. et al., High resolution mapping of quantitative trait loci in outbred mice.
Nat. Genet., 21, 305, 1999.
28. Roemer, K., Johnson, P.A., and Friedmann, T., Knockin and knockout. Transgenes,
development and disease, New Biologist, 3–4, 331, 1991.
29. Schibler, U., and Tafti, M., Molecular approaches towards the isolation of sleep related
genes, J. Sleep. Res., 8, 1, 1999.
30. Jaenisch, R., Transgenic animals, Science, 240, 1468, 1988.
31. Tobler, I. et al., Altered circadian activity rhythms and sleep in mice devoid of prion
protein, Nature, 380, 639, 1996.
32. Zhang, J. et al., Nonrapid eye movement sleep is suppressed in transgenic mice with
a deficiency in the somatotropic system, Neurosci. Lett., 220, 97, 1996.
33. Hajdu, I. et al., Sleep of transgenic mice producing excess rat growth hormone, Am.
J. Physiol. Regul. Integr. Comp. Physiol., 282, 70, 2002.
34. Fang, J., Wang W., and Krueger J.M., Mice lacking the TNF 55 kDa receptor fail to
sleep more after TNFalpha treatment, J. Neurosci., 17, 5949, 1997.
35. Boutrel, B. et al., Key role of 5HT1B receptors in the regulation of paradoxical sleep
as evidenced in 5HT1B knockout mice, J. Neurosci., 19, 3204, 1999.
36. Wisor, J.P. et al., Dopaminergic role in stimulant induced wakefulness, J. Neurosci.,
21, 1787, 2001.
37. Pinzar, E. et al., Prostaglandin D synthase gene is involved in the regulation of
nonrapid eye movement sleep, Proc. Natl. Acad. Sci. USA., 97, 4903, 2000.
38. Franken, P. et al., The transcription factor DBP affects circadian sleep consolidation
and rhythmic EEG activity, J. Neurosci., 20, 617, 2000.
39. de Lecea, L. et al., The hypocretins, hypothalamus specific peptides with neuroexci-
tatory activity, Proc. Natl. Acad. Sci., 95, 322, 1998.
40. Sakurai, T. et al., Orexins and orexin receptors, a family of hypothalamic neuropep-
tides and G proteincoupled receptors that regulate feeding behavior, Cell, 92, 573,
1998.
41. Kilduff, T.S., and Peyron, C., The hypocretin/orexin ligandreceptor system, implica-
tions for sleep and sleep disorders, Trends. Neurosci., 23, 359, 2000.
42. Chemelli, R.M. et al., Narcolepsy in orexin knockout mice, molecular genetics of
sleep regulation, Cell, 98, 437, 1999.

43. Nadeau, J.H. and Frankel, W.N., The roads from phenotypic variation to gene dis-
covery, mutagenesis versus QTLs, Nat. Genet., 25, 381, 2000.
44. Nadeau, J.H., Modifier genes in mice and humans, Nat. Rev. Genet., 2, 165, 2001.
45. King, D.P. et al., Positional cloning of the mouse circadian clock gene, Cell, 89, 641,
1997.
46. Kapfhamer, D. et al., Mutations in Rab3a alter circadian period and homeostatic
response to sleep loss in the mouse, Nat. Genet., 32, 290, 2002.
47. Lander, E.S., and Kruglyak, L., Genetic dissection of complex traits, guidelines for
interpreting and reporting linkage results, Nat. Genet., 11, 241, 1995.
48. Mott, R. et al., A new method for fine mapping quantitative trait loci in outbred
animal stocks, Proc. Natl. Acad. Sci. USA, 97, 12649, 2000.
48. Darvasi, A., Experimental strategies for the genetic dissection of complex traits in
animal models, Nat. Genet., 18, 19, 1998.
50. Flint, J., and Mott, R., Finding the molecular basis of quantitative traits, successes
and pitfalls, Nat. Rev. Genet., 2, 437, 2001.
51. Lander, E.S., and Botstein, D., Mapping mendelian factors underlying quantitative
traits using RFLP linkage maps, Genetics, 121, 185, 1989.
52. Churchill, G.A., and Doerge, R.W., Empirical threshold values for quantitative trait
mapping, Genetics, 138, 963, 1994.
53. Lyman, R.F., Lai, C., and MacKay, T.F., Linkage disequilibrium mapping of molecular
polymorphisms at the scabrous locus associated with naturally occurring variation in
bristle number in Drosophila melanogaster, Genet. Res., 74, 303, 1999.
54. Tafti, M. et al., Quantitative trait loci approach to the genetics of sleep in recombinant
inbred mice, J. Sleep. Res., 8, 37, 1999.
55. Shimomura, K., et al. Genomewide epistatic interaction analysis reveals complex
genetic determinants of circadian behavior in mice, Genome. Res., 11, 959, 2001.
56. Schwartz, W.J., and Zimmerman, P., Circadian time keeping in BALB/c and C57BL/6
inbred mouse strains, J. Neurosci., 10, 3685, 1990.
57. Roussel, B., Turrillot, P., Kitahama, K., Effect of ambient temperature on the sleep-
waking cycle in two strains of mice, Brain Research, 294, 67, 1984.
58. Phillips, T.J., Feller, D.J., and Crabbe, J.C., Selected mouse lines, alcohol and behav-
ior, Experientia, 45, 805, 1989.
59. De fiebre, C.M., and Collins, A., Classical genetic analyses of responses to nicotine
and ethanol in crosses derived from longand shortsleep mice, J. Pharmacol. Exp.
Ther., 261, 173, 1992.
60. Erwin, V.G., Jones B.C., and Radcliffe, R., Further characterisation of LS x SS
recombinant inbred strains of mice activating and hypothermic effects of ethanol,
Alcoholism. Clin. Exp. Res., 2, 200, 1990.
61. Marley, R.J., Freund, R.K., and Whener J.M., Differential response to flurazepam in
long sleep and short sleep mice, Pharmacol. Biochem. Behav., 31, 453, 1988.
62. Dudek, B.C., and Abbott M.E., A biometrical genetic analysis of ethanol response
in selectively bred long sleep and short sleep mice, Behav. Genet., 14, 1, 1984.
63. Crabbe, J.C., Belknap, J., and Buck, K.J., Genetic animal models of alcohol and drug
abuse, Science, 264, 1715, 1994.
64. Markel, P.D. et al., Quantitive trait loci for ethanol sensitivity in the LS X SS
recombinant inbred stains, interval mapping, Behav. Genet., 26, 447, 1996.
65. Toth, L. A., and Williams, R.W., A quantitative genetic analysis of slow wave sleep
and rapid eye movement sleep in CXB recombinant inbred mice, Behav. Genet., 29,
329, 1999.

66. Lopez Molina, L. et al., The DBP gene is expressed according to a circadian rhythm
in the suprachiasmatic nucleus and influences circadian behavior. EMBO. J., 16, 6762,
1997.
67. Franken, P., Malafosse, A., and Tafti, T., Genetic variation in EEG activity during
sleep in inbred mice, Am. J. Physiol., 275, 1127, 1998.
68. Guerra, C. et al., Abnormal nonshivering thermogenesis in mice with inherited defects
of fatty acid oxidation, J. Clin. Invest., 102, 1724, 1998.
69. Borbély, A.A., A two process model of sleep regulation, Hum. Neurobiol., 1, 195,
1982.
70. Tobler, I., and Borbély, AA., Sleep EEG in the rat as a function of prior waking,
Electroencephalogr. Clin. Neurophysiol., 64, 74, 1986.
71. Dijk, D.J., Beersma, D.G., and Daan, S., EEG power density during nap sleep,
reflection of an hourglass measuring the duration of prior wakefulness, J. Biol.
Rhythms, 2, 207, 1987.
72. Werth, E. et al., Dynamics of the sleep EEG after an early evening nap, experimental
data and simulations, Am. J. Physiol., 271, 501, 1996.
73. Franken, P., Tobler, I., and Borbély, A.A., Sleep homeostasis in the rat, simulation
of the time course of EEG slowwave activity, Neurosci. Lett., 130, 141, 1991.
74. Franken, P., Chollet, D., and Tafti, T., The homeostatic regulation of sleep need is under
genetic control, J. Neurosci., 21, 2610, 2001.
75. Achermann, P. et al., A model of human sleep homeostasis based on EEG slowwave
activity, quantitative comparison of data and simulations, Brain. Res. Bull., 31, 97,
1993.
76. Hegmann, J.P., and Possidente, B., Estimating genetic correlations from inbred
strains, Behav. Genetics, 11, 103, 1981.
77. Naylor, E. et al., The circadian clock mutation alters sleep homeostasis in the mouse,
J, Neurosci., 20, 8138, 2000.
78. Frank, M.G., Stryker, M.P., and Tecott, L.H., Sleep and sleep homeostasis in mice
lacking the 5HT2c receptor, Neuropsychopharmacology, 27, 869, 2002.
79. Wisor, J.P. et al., A role for cryptochromes in sleep regulation, BMC Neurosci., 3,
20, 2002.
80. Shaw, P.J. et al., Stress response genes protect against lethal effects of sleep depri-
vation in Drosophila, Nature, 417, 287, 2002.
81. Morgan, J.I., and Curran, T., Stimulus transcription coupling in the nervous system,
involvement of the inducible protooncogenes Fos and jun, Ann. Rev. Neurosci., 14,
421, 1991.
82. Baasheer, R. et al., Effects of sleep on wake-induced cFos expression. J. Neurosci.,
17, 9746, 1997.
83. Grassi-zucconi, G. et al., cFos spontaneous expression during wakefulness is reversed
during sleep in neuronal subsets of the rat cortex. J. Physiol., 88, 91, 1994.
84. O’Hara, B.F. et al., Immediate early gene expression in brain during sleep deprivation:
preliminary observations, Sleep, 16, 1, 1993.
85. Pompeiano, M. et al., NGFIA expression in the rat brain after sleep deprivation, Brain
Research, 46, 143, 1997.
86. Cirelli, C., Pompeiano, M., and Tononi, G., Neuronal gene expression in the waking
state, a role for the locus coeruleus, Science, 274, 1211, 1996.
87. Cirelli, C., and Tononi, G., Differences in brain gene expression between sleep and
waking as revealed by mRNA differential display and cDNA microarray technology,
J. Sleep. Res., 8, 44, 1999.

88. Shiromani, P.J. et al., Time course of Foslike immunoreactivity associated with
cholinergically induced REM sleep, J. Neurosci., 15, 3500, 1995.
89. Shiromani, P.J. et al., cFos knock out mice have reduced non-REM sleep, Sleep
Research, 26, 42, 1997.
90. Sherin, J.E. et al., Activation of ventrolateral preoptic neurons during sleep, Science,
271, 216, 1996.
91. Rhyner, T.A., Borbely, A.A., and Mallet, J., Molecular cloning of forebrain mRNAs
which are modulated by sleep deprivation, Eur. J. Neurosci., 2, 1063, 1990.
92. NeunerJehle, M., Rhyner, T.A., and Borbely, A.A., Sleep deprivation differentially
alters the mRNA and protein levels of neurogranin in rat brain, Brain Research, 685,
143, 1995.
93. NeunerJehle, M. et al., Characterization and sleep deprivation induced expression
modulation of dendrin, a novel dendritic protein in rat brain neurons, J. Neurosci.
Res., 46, 138, 1996.
94. Bredow, S. et al., Hypothalamic growth hormonereleasing hormone mRNA varies
across the day in rats, Neuroreport, 7, 2501, 1996.
95. Bova, R.M. et al., BDNF and trkB mRNAs oscillate in rat brain during the light-dark
cycle, Brain Res. Mol. Brain. Res., 57, 321, 1998.
96. Baasheer, R. et al., Opposite changes in adenosine A1 and A2A receptor mRNA in
the rat following sleep deprivation, Neuroreport, 12, 1577, 2001.
97. Toppila, J. et al., Sleep deprivation increases somatostatin and growth hormone releas-
ing hormone messenger RNA in the rat hypothalamus, J. Sleep. Res., 6, 171, 1997.
98. Mackiewicz, M. et al., Modulation of IL1 beta gene expression in the rat CNS during
sleep deprivation, Neuroreport, 7, 522, 1996.
99. Kushida, C.A., Zoltoski, R.K., and Gillin, J.C., The expression of m1m3 muscarinic
receptor mRNAs in rat brain following REM sleep deprivation, Neuroreport, 6, 1705,
1995.
100. Porkka-Heiskanen, T. et al., Noradrenergic activity in rat brain during rapid eye
movement sleep deprivation and rebound sleep, Am. J. Physiol., 268, 1456, 1995.
101. Toppila, J. et al., The effect of REM sleep deprivation on somatostatin and growth
hormone-releasing hormone gene expression in the rat hypothalamus, J. Sleep. Res.,
5, 115, 1996.
102. de Lecea, L. et al., A cortical neuropeptide with neuronal depressant and sleep
modulating properties, Nature, 38, 242, 1996.
103. Terao, A., et al., Preprohypocretin (preproorexin) expression is unaffected by short-
term sleep deprivation in rats and mice, Sleep, 23, 867, 2000.
104. Cirelli, C., and Tononi, G., Gene expression in the brain across the sleep-waking
cycle, Brain Research, 885, 303, 2000.
105. Terao, A. et al., Region specific changes in immediate early gene expression in
response to sleep deprivation and recovery sleep in the mouse brain, Neuroscience,
120, 1115, 2003.
106. Takahashi, J.S., Molecular neurobiology and genetics of circadian rhythms in mam-
mals, Ann. Rev. Neurosci., 18, 531, 1995.
107. Klein, D., Moore, R.Y. and Reppert S.M., Suprachiasmatic nucleus, The Mind’s Clock,
Oxford University Press, New York, 1991, p. 467.
108. Ralph, M.R. et al., Transplanted suprachiasmatic nucleus determines circadian period,
Science, 247, 9758, 1990.
109. Liu, C. et al., Cellular construction of a circadian clock, period determination in the
suprachiasmatic nuclei, Cell, 91, 855, 1997.

110. Welsh, D.K. et al., Individual neurons dissociated from rat suprachiasmatic nucleus
express independently phased circadian firing rhythms, Neuron, 14, 697, 1995.
111. Balsalobre, A., Damiola, F., and Schibler, U., A serum shock induces circadian gene
expression in mammalian tissue culture cells, Cell, 93, 929, 1998.
112. Jin, X. et al., A molecular mechanism regulating rhythmic output from the suprach-
iasmatic circadian clock, Cell, 96, 57, 1999.
113. Le Minh, N. et al., Glucocorticoid hormones inhibit food induced phase shifting of
peripheral circadian oscillators, EMBO J., 20, 7128, 2001.
114. Hardin, P.E., Hall, J.C., and Rosbash, M., Feedback of the Drosophila period gene
product on circadian cycling of its messenger RNA levels, Nature, 343, 536, 1990.
115. Kronopka, R.J. and Benzer, S., Clock mutants of drosophila melanomaster, Proc.
Natl. Acad. Sci., 68, 2112, 1971.
116. Sehgal, A. et al., Loss of behavioral rythms and per RNA oscillations in the Drosophila
mutant timeless, Science, 263, 1603, 1994.
117. Vitaterna, M.H. et al., Mutagenesis and mapping of a mouse gene, Clock, essential
for circadian behavior, Science, 264, 719, 1994.
118. Young, M.W., The molecular control of circadian behavioral rhythms and their
entrainment in Drosophila, Annu. Rev. Biochem., 67, 135, 1998.
119. Ralph, M.R., and Menaker, M., A mutation of the circadian system in golden hamsters,
Science, 241, 1225, 1988.
120. Valatx, J.L., Genetics as a model for studying the sleepwaking cycle, Exp. Brain.
Res., 8, 135, 1984.
121. Nolan, P. et al., Heterozygosity mapping of partially congenic lines, mapping of a
semi dominant neurological mutation, Wheels, on mouse chromosome 4, Genetics,
140, 245, 1995.
122. Antoch, M.P. et al., Functional identification of the mouse circadian Clock gene by
transgenic BAC rescue, Cell, 89, 655, 1997.
123. Lowrey, P.L. et al., Positional syntenic cloning and functional characterization of the
mammalian circadian mutation tau, Science, 288, 483, 2000.
124. Shearman, L.P. et al., Two period homologs, circadian expression and photic regula-
tion in the suprachiasmatic nuclei, Neuron, 19, 1261, 1997.
125. Shearman, L.P. et al., Interacting molecular loops in the mammalian circadian clock,
Science, 288, 1013, 2000.
126. Sun, Z.S. et al., RIGUI, a putative mammalian ortholog of the Drosophila period
gene, Cell, 90, 1003, 1997.
127. Tei, H. et al., Circadian oscillation of a mammalian homologue of the Drosophila
period gene, Nature, 389, 512, 1997.
128. Albrecht, U. et al., A differential response of two putative mamalian circadian regu-
lators, mper1 and mper2, to light, Cell, 91, 1055, 1997.
129. Sancar, A., Cryptochrome, the second photoactive pigment in the eye and its role in
circadian photoreception, Annu. Rev. Biochem., 69, 31, 2000.
130. Kume, K. et al., mCRY1 and mCRY2 are essential components of the negative limb
of the circadian clock feedback loop, Cell, 98, 193, 1999.
131. Dunlap, J.C., Molecular bases for circadian clocks, Cell, 96, 271, 1999.
132. Lee, C., Bae, K., and Edery, I., PER and TIM inhibit the DNA binding activity of a
Drosophila CLOCK-CYC/dBMAL1 heterodimer without disrupting formation of the
heterodimer: a basis for circadian transcription, Mol. Cell. Biol., 19, 5316, 1999.
133. Barnes, J.W. et al., Requirement of mammalian Timeless for circadian rhythmicity,
Science, 302, 439, 2003.

134. Preitner, N. et al., The orphan nuclear receptor REVERBalpha controls circadian
transcription within the positive limb of the mammalian circadian oscillator, Cell,
110, 251, 2002.
135. Berson, D.M. Strange vision, ganglion cells as circadian photoreceptors, Trends.
Neurosci., 26, 314, 2003
136. Ruby, N.F. et al., Role of melanopsin in circadian responses to light, Science, 298,
2211, 2002.
137. Kloss, B. et al., The Drosophila clock gene doubletime encodes a protein closely
related to human casein kinase, I epsilon, Cell, 94, 97, 1998.
138. Keesler, G.A. et al., Phosphorylation and destabilization of human period I clock
protein by human casein kinase I epsilon, Neuroreport, 11, 951, 2000.
139. Silver, R. et al., A diffusible coupling signal from the transplanted suprachiasmatic
nucleus controlling circadian locomotor rhythms, Nature, 382, 810, 1996.
140. Park, J.H. et al., Differential regulation of circadian pacemaker output by separate
clock genes in Drosophila, Proc. Natl. Acad. Sci. USA, 97, 3608, 2000.
141. Kramer, A. et al., Regulation of daily locomotor activity and sleep by hypothalamic
EGF receptor signaling, Science, 294, 2511, 2001.

8 Searching for Sleep

Mutants of Drosophila
Melanogaster
Chiara Cirelli and Giulio Tononi
CONTENTS
Introduction
Forward Genetics to Understand Sleep Regulation and Functions
Fly Sleep Shares Many Features with Mammalian Sleep
Complex Behaviors and Single-Gene Mutations
The Genetics of Sleep
Short Sleepers, Sleep Deprivation, and Sleep Restriction
The Sleep Phenotype in Wild-Type Drosophila Lines
The Sleep Phenotype in Drosophila Mutant Lines
Conclusions
References
INTRODUCTION
Sleep is present in all species where it has been studied, but its functions remain
unknown. A sufficient amount of sleep constitutes a fundamental biological need.
Curtailing the amount of sleep in normal sleepers affects performance, vigilance,
memory, and health. Like all complex behaviors, sleep is both environmentally
modulated and genetically determined; however the responsible genes have not been
discovered. To identify them we have initiated a genetic screening for short sleepers
in the fruit fly Drosophila melanogaster. Mutagenesis screening in drosophila has
helped unravel cellular mechanisms that are highly conserved across species; e.g.,
those controlling development, aging, stress, memory, and circadian rhythms. For
the past few years, our laboratory and others have shown that fly sleep shares many
key features with mammalian sleep. As in mammals, sleep in drosophila is charac-
terized by increased arousal thresholds and by changes in brain electrical activity.
Fly sleep is regulated independent of the circadian clock, modulated by stimulants
and hypnotics, affected by age, and is associated with changes in brain gene expres-
sion similar to those observed in mammals.
0-8493-1519-0/05/$0.00+$1.50

In the past 2 years, our laboratory has screened ~7000 mutant drosophila lines,
most of which were carrying single-gene mutations. We found that the amount and
regulation of sleep are highly conserved: Almost all flies sleep 400–800 min in 24
hours and show increased sleep duration and continuity after sleep deprivation. We
have also identified several short sleeper lines that sleep <280 min in 24 hours. The
short sleep mutation is often due to the genomic insertion of a P element whose
mobilization reverts the flies to normal sleep, suggesting a single gene effect. The
current work is aimed at characterizing these mutant lines genetically, molecularly,
and behaviorally to identify the genes responsible for the short sleep phenotype and
investigate the molecular pathways controlled by these genes. This research will
help to identify the molecular mechanisms regulating the need for sleep and provide
novel clues to its functions.
FORWARD GENETICS TO UNDERSTAND SLEEP

REGULATION AND FUNCTIONS
The importance of sleep is strongly suggested by the time spent in this state and by
its ubiquitous occurrence in all animal species studied so far (Rattenborg et al., 2000;
Tobler, 2000). In both humans and animals, sleep need is tightly regulated and sleep
pressure becomes overwhelming after just a few hours without sleep (Borbély and
Achermann, 1999). Total sleep deprivation and chronic sleep restriction cause sig-
nificant cognitive deficits, including decrease in attention and short-term memory,
speech impairment, and inflexible thinking (Horne, 1988; Forest and Godbout, 2000;
Belenky et al., 2003; Van Dongen et al., 2003). Sleep loss also affects host defense
systems and can lead to a decrease in glucose tolerance and an increase in peripheral
metabolic rate and in cortisol level (Spiegel et al., 1999; Rogers et al., 2001). If
sustained for several days, sleep deprivation is fatal in rats and other mammals
(Rechtschaffen et al., 1989; Rechtschaffen and Bergmann, 2002). Like hunger or
thirst, the drive for sleep appears to satisfy an elementary need, but unlike eating
and drinking, the purpose of sleep remains obscure: Sleep is the one major biological
process whose functions have not yet been identified (Rechtschaffen, 1998).
It is generally thought that sleep is “by the brain and for the brain” (Horne 1988;
Hobson 1989). Some evidence indicates that sleep may represent a favorable time
for brain protein synthesis (Ramm and Smith, 1990; Nakanishi et al., 1997). Another
possibility, suggested by behavioral studies, is that sleep may promote memory
consolidation (Stickgold et al., 2001; Walker et al., 2002). It is also widely thought
that the functions of sleep may ultimately relate to cellular and molecular aspects
of neural function (Moruzzi, 1972; Rechtschaffen, 1998; Cirelli and Tononi, 2000;
Steriade and Timofeev, 2003). In line with this assumption, we have used mRNA
differential display and, more recently, high-density microarrays to perform a
genome-wide expression profiling to identify brain transcripts whose expression
changes as a function of sleep and wakefulness (Cirelli and Tononi, 2000; Cirelli et
al., 2004). We found that the expression of hundreds of genes is modulated in the
brain as a function of behavioral state and independently of circadian time, support-
ing the notion that wakefulness and sleep differ not only at the behavioral, electro-
physiological, and metabolic level, but also at the molecular level.

Gene expression studies can only provide correlative evidence, and any link
between a putative sleep function and a specific gene (or gene category) needs to
be supported by causal experimental approaches. Forward genetic approaches consist
in mutating (ideally) all the genes expressed in the brain and in studying how each
single mutation affects sleep, its regulatory mechanisms, and its functional conse-
quences. A systematic mutagenesis screening of sleep mutants in mammals remains
a daunting task (Tafti and Franken, 2002; Dugovic et al., 2003), but it has recently
been shown that drosophila sleep shares many features with mammalian sleep
(Hendricks et al., 2000; Shaw et al., 2000). This finding has advanced our knowledge
of the phylogeny of sleep, supporting the notion that sleep fulfills at least one
fundamental function in many divergent animal species. Moreover D. melanogaster
may now be used as a powerful tool for the genetic dissection of sleep with forward
genetics, an approach that has greatly benefited research on circadian rhythms. In
forward genetics either a P element insertion or a chemical such as N-ethyl-N-
nitrosurea (ENU, in mice) or ethyl methanesulfonate (EMS, in flies) is used to mutate
at random the whole genome; this is followed by a high throughput screening of all
mutant offspring to detect major effects on the phenotype of interest. The power of
forward genetics is that mutant screens make no assumption concerning the mech-
anisms underlying a behavior and require only a clear phenotype to be expressed.
For the past 2 years, our laboratory has embarked on a large-scale mutagenesis
screening in search for flies that need little sleep or show abnormal homeostatic
response after sleep deprivation (Cirelli et al., 2003). The final goal is to screen as
many mutant fly lines as there are fly genes. So far we have screened ~7000 mutant
lines and shown that sleep amount and response to sleep deprivation are highly
conserved phenotypes in wild-type flies as well as in mutant lines. Most importantly
this work has demonstrated that sleep mutants can be isolated and that the identifi-
cation of the corresponding genes is feasible.
FLY SLEEP SHARES MANY FEATURES WITH

MAMMALIAN SLEEP
Sleep is a complex integrative phenomenon that needs to be defined using multiple
criteria. As mentioned above, drosophila had been extensively used in circadian
research long before it became of interest to sleep researchers (Konopka and Benzer,
1971). Circadian studies had shown that fruit flies are active and move around during
the day and much less so during the night, but until 3 years ago it was not known
whether the sustained periods of immobility during the night represented a sleep-
like state or just quiet wakefulness; i.e., a state of behavioral inactivity in which the
ability to respond to the environment is preserved. The demonstration that flies sleep,
much as other animals and humans do, was achieved using behavioral, pharmaco-
logical, molecular, and genetic techniques (Hendricks et al., 2000; Shaw et al., 2000).
More recently classical electrophysiological methods have proven that EEG corre-
lates on sleep and wakefulness are also present in the fruit fly (Nitz et al., 2002).
Fly behavior was monitored using visual observation, an ultrasound activity
monitoring system, and an automatic infrared system (Drosophila Activity Monitor-
ing System, DAMS; Trikinetics, Waltham, MA). The ultrasound method (Shaw et

al., 2000) allows a continuous, high-resolution measurement of the behavior of a

single fly housed inside an ultrasound standing wave chamber. Whenever the fly
moves its head, wings, or limbs, a perturbation of the standing wave is produced
and is counted as a movement. Although very precise, this method is impractical
for evaluating sleep-waking parameters in a large-scale project. The DAMS is instead
designed to monitor hundreds or thousands of flies simultaneously. One DAMS
monitor contains 32 glass tubes, each housing a single fly and enough food for 1-
week recording (Figure 8.1 A). As each fly moves back and forth in its tube, it
interrupts an infrared light beam that bisects the tube. Each crossing is counted as
a movement and the number of movements every minute are summed up and
expressed as an activity index. Both the ultrasound and the infrared system had been
validated by visual observation and give similar results: Flies are mostly active and
moving around during the day, and during the night they show long periods of
immobility that can last several hours.
Immobility qualifies as sleep only if it is accompanied by a reversible increase
in arousal threshold. Arousal threshold in flies has been measured using vibratory,
visual, auditory (Shaw et al., 2000; Nitz et al., 2002), and more recently, thermal
stimuli (Huber et al., 2004). In all cases it was found that flies that had been
behaviorally awake immediately before the stimulus readily responded to low and
medium stimulus intensities. By contrast flies that had been behaviorally quiescent
for 5 min or more rarely showed a motor response, although they quickly responded
when the stimulus intensity was increased. Thus sleep can be operatively defined in
FIGURE 8.1 (See facing page.) Analysis of locomotor activity and sleep in fruit flies. (A)
A Drosophila Activity Monitoring System (DAMS) monitor containing thirty-two 6.5-mm
(5 mm I.D.) glass tubes, each housing a single fly. (B) Typical pattern of sleep in a population
of 96 female wild-type Canton-S flies as measured in a DAMS monitor. DAMS measures
activity as counts (number of crossings) per minute. Wakefulness is defined as any period
of at least one minute characterized by activity (one or more counts per minute). Based on
arousal threshold data, sleep is defined as any period of uninterrupted behavioral quiescence
(no counts/min) lasting for at least 5 min. Mean values of the amount of sleep are calculated
on consecutive 30-min time intervals, and the time course is graphically shown over the
entire day. In female flies most of the sleep occurs at night. (C) Increase in sleep duration
following 6,12, and 24 hours of sleep deprivation (SD) in female Canton-S flies (n = 20–40
for each experiment). Each diagram shows the daily amount of sleep for baseline day (blue
line), SD day (red line), and the first recovery day after SD (green line). Time and duration
of SD are indicated by the red bars below the x axis. An increase in sleep duration is present
after all 3 periods of SD, and occurs mainly during the first 6 hours following the end of
SD. Flies were maintained in a 12:12 light dark cycle (light on at 8:00 A.M.). (D) To measure
sleep fragmentation, a sleep continuity score is calculated, which increases during contin-
uous epochs with no locomotor activity and decreases during epochs with one or more
counts of activity. The sleep continuity score is high if sleep is continuous and undisturbed,
and low if sleep is fragmented. Blue lines in the upper diagram represents sleep scores for
16 female Canton-S flies during baseline. Green lines in the lower diagram show the sleep
score for the same flies the day following 24 hours SD. Note the significant increase in the
sleep score immediately after the end of SD. In several flies this increase persists during
the following night.

flies as any period of behavioral quiescence (no counts detected by the DAMS)
lasting longer than 5 min (Figure 8.1 B).
All animals studied so far show a homeostatic regulation of sleep (Tobler, 1995,
2000). Flies, as well as other invertebrates such as cockroaches (Tobler, 1983; Tobler
and Neuner-Jehle, 1992), scorpions (Tobler and Stalder, 1988), and honey bees
(Kaiser and Steiner-Kaiser, 1983; Sauer et al., 1999) are no exception. Sleep depri-
vation can be performed by gentle tapping on the glass tube whenever the fly stops
moving for more than 5 min, or automatically. Currently in our laboratory wakeful-
ness is enforced by placing the DAMS monitors vertically within a framed box able
to rotate along its major axis under the control of a motor. The box can rotate 180∞
clockwise or counter-clockwise (2–3 revolutions per min). At the nadir of each
rotation, the monitors are dropped 1 cm. This causes the flies to fall from their
current position to the bottom of the tube. This method can effectively sleep-deprive
thousands of flies simultaneously for one or more days.
FIGURE 8.1 (See color insert following page 108.)

Wild-type flies sleep longer after being sleep-deprived (Figure 8.1 C). As in
mammals, this sleep rebound occurs mainly immediately after the end of the sleep
deprivation period, is more pronounced after longer (12–24 hours) than after shorter
(6 hours) periods of sleep loss, and the recovered sleep only represents a fraction of
what was lost (Figure 8.1 C). There is no increase in sleep duration when female flies
are subjected to 12 hours of the same stimulation during the day (when they are
normally awake), ruling out aspecific effects (Shaw et al., 2000). In mammals sleep
after sleep deprivation is also qualitatively different; i.e., is richer in slow-wave activity,
a well-characterized EEG marker of sleep intensity and sleep pressure, and less
fragmented (there are fewer periods of brief awakenings during sleep) (Borbély and
Achermann, 1999; Huber et al., 2000). New evidence from our laboratory shows that
in flies sleep continuity is increased and the number of brief awakenings is reduced
after sleep deprivation (Biesiadecki et al., 2003; Huber et al., 2004; Figure 8.1 D).
The homeostatic regulation of sleep in mammals can be dissociated in part from
circadian factors. A similar dissociation between circadian and homeostatic regula-
tion of sleep can be seen in flies in which the central circadian clock has been
genetically destroyed by a mutation in one canonical circadian gene; e.g., cycle,
period, or Clock. These mutant flies sleep across the entire 24-hour period rather
than just at night; however, after 24 hours of sleep deprivation, they still show a
sleep rebound (Shaw et al., 2000, 2002).
Fly sleep seems to be sensitive to at least some of the same stimulants and
hypnotics that modulate behavioral states in mammals. For example, when given
caffeine (Hendricks et al., 2000; Shaw et al., 2000) or modafinil (Hendricks et al.,
2003), flies stay awake longer. By contrast, when fed with antihistamines, they go
to sleep earlier (Shaw et al., 2000).
As mentioned above, hundreds of genes change their expression in the rat brain
between sleep and wakefulness, suggesting that in mammals sleep and wakefulness
differ significantly at the molecular level (Cirelli et al., 2004). A first systematic screen-
ing of state-dependent gene expression in drosophila using mRNA differential display
suggested that this might also be the case in fruit flies. We identified (Shaw et al., 2000;
Cirelli and Tononi, 2001) several wakefulness-related genes in the fly that corresponded
to wakefulness-related genes in the rat, including, for instance, those coding for the
mitochondrial enzyme cytochrome oxidase C (subunit I), the endoplasmic reticulum
chaperone BiP, and the transcription factor Stripe A (homologue to the rat immediate
early gene NGFI-A). Whether molecular similarities between flies and rats extend to
sleep-related genes is now been tested using high-density cDNA microarrays.
Recently Nitz et al. (2002) were able to obtain prolonged recordings of local
field potentials (LFPs) from the medial part of the fly brain between the mushroom
bodies. They found that LFPs from awake, moving fruit flies are dominated by spike-
like potentials and that these spikes largely disappear during the quiescent state when
arousal thresholds are increased. Targeted genetic manipulations demonstrated that
LFPs had their origin in brain activity and were not merely an artifact of movement
or electromyographic activity. Thus as in mammals, wakefulness and sleep in fruit
flies are accompanied by different patterns of brain electrical activity.
Sleep in mammals is prominent in the very young, stabilizes during adolescence
and adulthood, and declines during old age. Sleep in drosophila follows a similar

pattern (Shaw et al., 2000). On the first full day after eclosion, the amount of sleep
is high but declines steadily until day 3, when it reaches an adult pattern. As the
flies ages, the amount of sleep during the night declines and by 33 days of age is
significantly below that found in young adults (Shaw et al., 2000). Thus, as for
mammalian sleep, sleep in drosophila is characterized by increased arousal threshold,
changes in brain electrical activity, and is homeostatically regulated independent of
the circadian clock. As in mammals, sleep is abundant in young flies and it is reduced
in older flies, and it is modulated by stimulants and hypnotics. Several molecular
markers modulated by sleep and wakefulness in mammals are also modulated by
behavioral state in drosophila.
COMPLEX BEHAVIORS AND SINGLE-GENE

MUTATIONS
Complex biological processes such as development, learning and memory, and aging
are both environmentally modulated and genetically determined. It is often assumed
that complex behaviors are under polygenic control. It is also widely assumed that
the mechanisms underlying complex behaviors may differ in simple organisms, such
as invertebrates, compared to mammals in general and humans in particular; however
several recent examples demonstrate that complex behaviors can be strongly regu-
lated by single genes and that the cellular and molecular mechanisms underlying
complex biological processes are often shared between invertebrates and humans
(Sokolowski, 2001). The completion of the sequencing and annotation of the fly,
mouse, and human genomes has made these findings less surprising.
All the major gene families are present in both mammals and nonmammalian
species and the total number of genes is much more similar between species than
previously thought: The human genome contains ~40000 genes, compared to ~30000
and 14000 genes in the mouse and fly genome, respectively. The majority of fly
genes are shared with humans. It is becoming increasingly apparent that the verte-
brate genome arose from the amplification of a core set of genes not much larger
than that of the fly; for instance, the sequence diversity between the various potas-
sium channels is greater within either drosophila or mouse than the divergence of
a particular channel between drosophila and mouse (Miklos and Maleszka, 2000).
The majority (77%) of the genes involved in human diseases have fly counterparts
(Reiter et al., 2001), and the expression of human genes into flies very often results
in phenotypes that mimic human diseases (e.g., human a-synuclein in drosophila
causes a phenotype that resembles human Parkinson’s disease; Auluck et al., 2002).
Moreover single-gene mutations identified in drosophila have been instrumental in
understanding complex biological processes in mammals, including humans (Ben-
zer, 1971). Compelling examples include aging, memory, and the circadian clock.
In drosophila, a mutation in the gene methuselah or in the gene Indy (for I’m
not dead yet) results in a significant (up to twofold) increase in the average adult
life span (Lin et al., 1998; Rogina et al., 2000). Indy is involved in intermediary
metabolism (Rogina et al., 2000), and caloric restriction is the only intervention
that extends life span in mammals, suggesting that flies and mammals share at least
one molecular mechanism to extend life span. Forward genetics in drosophila has

identified several single-gene mutations of the cAMP and CREB pathways that
affect learning and memory (e.g., Waddell and Quinn, 2001; Sanyal et al., 2002).
Related studies in mice have shown that cAMP and CREB are also crucial for
memory formation in mammals (e.g., Barco et al., 2002; Kida et al., 2002). Finally,
Konopka and Benzer in 1971 showed that a single-gene mutation of the period
locus can abolish a complex behavior such as locomotor and eclosion rhythms.
Since then mutagenesis screening in both flies and mice have identified all the
currently known canonical circadian genes and have demonstrated that all the major
components of the molecular clock are shared between drosophila and mammals
(e.g., Blau, 2003).
THE GENETICS OF SLEEP

Several sleep disorders, from narcolepsy and somnambulism to REM sleep behavior
disorder and fatal familial insomnia, are under strong genetic control (O’Hara and
Mignot, 2000; Toth, 2001; Franken and Tafti, 2003). The circadian regulation of
sleep is also partly genetically determined; e.g., morningness-eveningness tendencies
in humans show substantial heritability (Katzenberg et al., 1998). Familial advanced
sleep phase syndrome (FASPS), in which the sleep cycle occurs 4 hours earlier than
normally, is a highly penetrant autosomal dominant circadian rhythm variant due a
point mutation in a casein kinase-epsilon phosphorylation site of the circadian gene
Per2 (Toh et al., 2001). By contrast the delayed sleep phase syndrome has been
associated with a structural (Ebisawa et al., 2001) and a length (Archer et al., 2003)
polymorphism in the human Per3 gene. Several aspects of normal human sleep,
from EEG pattern to the duration of total sleep and of REM sleep, are also strongly
genetically determined, suggesting that allelic variants or gene mutations must be
responsible for these variations. There is long-standing and compelling evidence
(reviewed in Franken and Tafti, 2003) that sleep duration and EEG patterns of
monozygotic twins are more similar than those of dizygotic twins or unrelated
subjects, confirming that these complex traits are controlled by genes more than by
environmental factors.
The genetic control of sleep has been confirmed in mice by QTL analysis that
has shown that the amount of total sleep, NREM sleep, and REM sleep are under
genetic control (Friedmann, 1974; Valatx et al., 1972, 1974; Tafti et al., 1997, 1999;
Franken et al., 1998; Toth and Williams, 1999). The increase in sleep pressure during
sleep deprivation, as measured by an increase in slow wave activity in the 0.75–4Hz
range, is also controlled by genetic factors in mice (Franken et al., 2001).
Finally, it has been shown that a mutation in the gene coding for short-chain
acylcoenzyme A dehydrogenase (Acads), an enzyme involved in brain fatty acid
beta oxidation, affects theta oscillations during sleep, but not during wakefulness
(Tafti et al., 2003). Thus classical genetic sleep in humans and QTL analysis in mice
have shown that there is a strong genetic control for sleep phenotypes such as amount
of sleep, EEG patterns, and the response to sleep deprivation. However, with the
exception of Acads, these approaches have yet to identify the underlying mecha-
nisms; i.e., the genes involved and their downstream targets.

SHORT SLEEPERS, SLEEP DEPRIVATION, AND

SLEEP RESTRICTION
A sufficient amount of sleep constitutes a fundamental biological need, but such a
need may be fulfilled by different amounts of sleep in different people. Most people
require more than 6 hours of sleep to feel well rested, but there is a group of people
who only need between 3 and 6 hours of sleep. There are a few reported cases of
extreme short sleepers belonging to the same family (Jones and Oswald, 1968).
Natural short and long sleepers have similar amounts of stage 3 and 4 sleep, but
long sleepers have more stage 2 and REM sleep (Webb and Friel, 1971; Hartmann
and Brewer, 1976; see also refs in Horne, 1988).
Several studies have focused on differences in personality and life styles between
long and short sleepers (e.g., Hartmann and Brewer, 1976), but very little is known
relative to cognitive and physiological functions in short sleepers relative to normal
and long sleepers. It is well established now that sleep deprivation and sleep restric-
tion can impair cognitive performance when sustained for more than a few hours or
a few days, respectively (see Van Dongen et al., 2003 and refs therein), but these
studies were performed in normal sleepers, thus it is not known whether sleep loss
will equally affect daytime performance in short and normal sleepers.
A few studies used slow wave activity as a measure of sleep pressure and showed
that, although the homeostatic sleep regulatory mechanisms do not differ between
short and normal sleepers, short sleepers may tolerate a higher sleep pressure than
long sleepers (Aeschbach et al., 1996, 2001). Whether living under a higher sleep
pressure has behavioral consequence on daytime performance, vigilance, and mem-
ory, remains unknown.
The elucidation of the genes and pathways that regulate sleep need in drosophila
is significant for several reasons. Judging from the remarkable similarities between
drosophila and mammalian circadian control, an understanding of how sleep need
is controlled in drosophila should open the way to dissecting similar mechanisms
in mammals. Knowledge of the genes and molecular pathways that control the
amount of sleep will also shed light on how sleep is reduced or increased in
pathological conditions. Such knowledge will permit the delineation of crucial
molecular targets and the development of appropriate pharmacological interven-
tions. Understanding the differences between normal and short sleeper lines at the
molecular level will indicate through which mechanisms the latter can afford to
sleep much less while preserving normal levels of performance. We expect that
these molecular differences will be closely related to the biological functions
fulfilled by sleep and thereby provide a mechanistic approach toward unraveling
such functions.
THE SLEEP PHENOTYPE IN WILD-TYPE

DROSOPHILA LINES
The full characterization of the fly sleep phenotype discussed above was mainly
performed in one wild-type strain, the Canton-S strain, which has been maintained

in the laboratory for several decades. To establish whether the sleep phenotype is
stable among other wild-type strains, we examined sleep patterns and the response
to sleep deprivation in 123 lines derived from single female flies (isofemale lines)
collected in the wild between 1994 and 2002. We found that the amount of sleep
over the 24-hour period and the homeostatic response to sleep deprivation are well-
conserved phenotypes across wild-type strains: All flies tested so far are diurnal,
most flies sleep between 400 and 800 min/day, and sleep deprivation for 24 hours
is in all cases followed by an increase in sleep duration and in sleep continuity as
measured by the sleep continuity score (Holladay et al., 2003; Huber et al., 2004).
The analysis of wild-type strains has also confirmed a significant difference between
male and female flies: Female flies sleep almost exclusively during the night, while
males show also a long period of siesta in the middle of the day (Figure 8.2 A). The
daily amount of sleep in 123 isofemale lines is shown in Figure 8.2 B. For both
female and male flies, mean values are similar to those of the originally described
Canton-S flies (Shaw et al., 2000).
THE SLEEP PHENOTYPE IN DROSOPHILA

MUTANT LINES
The ~7000 mutant lines tested so far include deficiency lines, lines obtained through
insertional mutagenesis with transposable elements (P and EP elements), and lines
chemically mutagenized with EMS. The collection of deficiencies included ~150
lines, each carrying a deletion of a relatively large portion of the fly genome. The
advantage of this collection is that as a whole it covers ~80% of the fly genome,
thus allowing a quick and comprehensive screening of most of the fly genome. The
main disadvantage is that the identification of the gene of interest may be difficult
because each deletion includes many different genes, sometimes more than 100.
The insertional lines tested so far include the ~1000 lines of the Berkeley
Drosophila Genome Project primary collection (Spradling et al., 1999) and the ~2300
EP lines of the Rörth collection (Rörth et al., 1998). They include both loss-of-
function mutations and gain-of-function mutations. The former are often due to the
insertion of a transposon inside a transcription unit, the latter to gene overexpression
following the transposon insertion upstream of the transcription start site. Insertional
mutagenesis usually permits the rapid identification of the mutated gene by sequenc-
ing the flanking sequences from one or both ends of the transposon insertion.
Moreover the mobilization of the inserted element can generate new alleles, and
expression patterns can be characterized by lacZ staining of tissues.
A limitation of this approach, however, is that transposons do not insert at random
into the genome but have preferred hot spots (Liao et al., 2000). For this reason we
are also using chemical mutagenesis with EMS, which is known to randomly induce
small (point) mutation over the entire genome at a reasonable rate. EMS has been
the most frequently used chemical mutagen in drosophila over the last 25 years,
such as in the search for learning mutants, circadian mutants, and paralytic mutants
(Roberts, 1998). The disadvantage of chemical mutagenesis is that the molecular
characterization of the gene of interest may be not as straightforward as with
insertional mutagenesis.

FIGURE 8.2 (See color insert.) Sleep pattern and sleep amount in 123 wild-type lines. (A)
Daily amount of sleep in wild-type Canton-S female (blue line, n = 14) and male (red line, n
= 15) flies. (B) Daily amount of sleep in 123 isofemale lines derived from single wild-type
female flies. Most female flies (blue line) sleep between 500 and 800 min/day, with a mean of
650 ± 126 (mean ± SD; median = 670, min = 289, max = 935; female Canton-S flies = 664 ±
137). Male flies of the same isofemale lines (red line) sleep between 600 and 1000 min/day
with a mean of 786 ± 170 (mean ± SD; median = 799, min = 109, max = 1106; Canton-S male
flies = 864 ± 137).
We are currently screening 50–100 mutant lines every week. Flies are continu-
ously recorded in a DAMS monitor for one week, including 2–3 baseline days, 24
hours of sleep deprivation, and 1–3 days of recovery after sleep deprivation. Ten to
sixteen flies (4–7 days old at the beginning of the experiment) are tested for each
line. In agreement with the results obtained with isofemale lines, we have found

FIGURE 8.3 (See color insert.) Daily sleep amount in 1547 mutant fly lines. Mean ± SD
is 616 ± 169 (min 131, max 1155). Shaded areas show one (dark red) and two (light red)
standard deviations from the mean.
that daily sleep amount and response to sleep deprivation are highly conserved
phenotypes in most mutant lines. Figure 8.3 shows the daily sleep amount in female
flies of 1547 insertional lines. The amount of sleep in 24 hours is normally distrib-
uted, with a mean of 616 ± 169 (mean ± SD; min 131, max 1155). As shown in
Figure 8.3, few fly lines qualify as short sleepers, defined here as those in which
sleep amount is less than 280 min in 24 hours for female flies, and less than 450
min in 24 hours in male flies (i.e. less than two standard deviations from the mean
of all fly mutant lines screened so far). Of the 7000 lines screened so far, only 15
lines qualify as short sleeper lines.
As observed with Canton-S flies, the great majority of the isofemale lines and
of the mutant lines tested so far showed a sleep rebound after 24 hours of sleep
deprivation. As in wild-type flies (Figure 8.1 C), a 24-hour sleep deprivation was
followed by an increase in sleep duration that was most pronounced during the first
4–6 hours immediately after the end of the sleep deprivation period and was over
in most cases by the end of the second day of recovery. Moreover, similar to wild-
type flies, most mutant lines also showed an increase in sleep continuity after sleep
deprivation. Finally, like wild-type lines, mutant fly lines only recovered a fraction
(10–40%) of the sleep lost during the sleep deprivation period.
Three short sleeper lines are shown in Figure 8.4. These are some of the lines
that we are currently characterizing. Each of these lines is homozygous for a single
P element insertion, and our initial revertant analysis indicates that the mutation

FIGURE 8.4 (See color insert.) Daily amount of sleep in 3 short sleeper lines. Red lines in each panel represents daily amount of sleep during baseline
(min/24 H; mean ± SD) of 16 female (upper panel) and male (lower panel) flies of a short sleeper line. For comparison the blue line in each panel
represents the daily amount of sleep in wild-type flies (n = 16/panel). Daily amount of sleep (mean ± SD, n = 40–80 flies, at least 3 independent
experiments) is as follows: ss1 flies = 190 ± 40, 200 ± 50 (females and males, respectively); ss2 flies = 210 ± 30 and 280 ± 90; ss3 flies = 230 ± 40, 270 ± 70.

caused by the transposon insertion is solely responsible for the short sleeper phe-
notype, pointing to a single-gene effect. It should be mentioned that the overall
baseline performance of these mutant flies, assessed by measuring levels of loco-
motor activity, sensitivity to anesthetics, geotaxic response, sensitivity to heat, and
vigilance tests, is normal, ruling out major aspecific abnormalities as responsible
for the short sleeper phenotype.
CONCLUSIONS
Work in other laboratories and ours for the past 5 years has demonstrated that fly
sleep shares many fundamental features of mammalian sleep. We have shown that
sleep phenotypes in flies are well defined and stable, and can thus be employed to
screen for mutations affecting sleep amount, sleep quality and the homeostatic
regulation of sleep. Our ongoing, high-throughput screening of >7000 drosophila
lines (~1/3 of the fly genome) has led to the identification of several mutations that
significantly shorten daily sleep amount, indicating that key features of sleep are
under genetic control. The three lines shown in Figure 8.4 are all insertional lines
for which the location of the transposon insertion is known, thus it is possible to
identify and characterize the genes responsible for the short sleep phenotype. The
demonstration that the short sleeper phenotype can be reverted in these lines by
transposon jumping indicates that a single gene is likely to be involved. By charac-
terizing the molecular pathways involved in producing the short sleeper phenotype,
we should be able to determine whether the three genes act through the same pathway
or whether each of them has specific downstream targets.
It should be mentioned here that our screening also identified a few long sleeper
lines. The reason why we do not focus our efforts currently on some of these lines
is largely practical. Behavioral characterization of long sleepers may be challenging
and time consuming because any mutation affecting the general health of the fly
(e.g., paralytic mutations) is likely to result in a decrease in locomotor activity and
therefore can affect our calculation of sleep amount. It should also be mentioned
that so far our screening has not identified any mutation that can produce a no-sleep
fly. This finding per se is yet another proof that sleep must serve a very important
function and that wakefulness cannot substitute for sleep.
In line with this conclusion, the fly mutagenesis screening discussed here shows
that while the daily sleep quota differs among mutant lines, very few mutations can
shorten sleep time to less than 300 min per day. Whatever the function of sleep
might be, this seems to be the minimum time required in flies to carry out that
function (interestingly, 2–3 hours is also the limit for human short sleepers). It could
be argued that the identification of a no-sleep mutant is only a question of time;
however compelling evidence from sleep deprivation experiments suggests that this
may not be the case. Sleep deprivation is fatal in rats if prolonged for several days
(Rechtschaffen et al., 1989; Rechtschaffen and Bergmann, 2002), and a recent study
has shown that flies also die when kept awake for more than 60 hours (Shaw et al.,
2002), thus the no-sleep phenotype might be missed altogether by mutagenesis
screenings performed at a late developmental stage or in adulthood. It cannot be
excluded, however, that such phenotype could be identified in the future by screens

performed at an earlier developmental stage should such studies become possible

in flies and other invertebrates.
REFERENCES
Aeschbach D., Cajochen C., Landolt H., and Borbely A.A., Homeostatic sleep regulation in
habitual short sleepers and long sleepers, Am. J. Physiol., 270, R41, 1996.
Aeschbach D., Postolache T.T., Sher L., Matthews J.R., Jackson M.A., and Wehr T.A.,
Evidence from the waking electroencephalogram that short sleepers live under higher
homeostatic sleep pressure than long sleepers, Neuroscience, 102, 493, 2001.
Archer S.N., Robilliard D.L., Skene D.J., Smits M., Williams A., Arendt J., and von Schantz
M., A length polymorphism in the circadian clock gene Per3 is linked to delayed
sleep phase syndrome and extreme diurnal preference, Sleep, 26, 413, 2003.
Auluck P.K., Chan H.Y., Trojanowski J.Q., Lee V.M., and Bonini N.M., Chaperone suppres-
sion of alpha-synuclein toxicity in a drosophila model for Parkinson’s disease, Sci-
ence, 295, 865, 2002.
Barco A., Alarcon J.M., and Kandel E.R., Expression of constitutively active CREB protein
facilitates the late phase of long-term potentiation by enhancing synaptic capture,
Cell, 108, 689, 2002.
Belenky G., Wesensten N.J., Thorne D.R., Thomas M.L., Sing H.C., Redmond D.P., Russo
M.B., and Balkin T.J., Patterns of performance degradation and restoration during
sleep restriction and subsequent recovery: a sleep dose-response study, J. Sleep. Res.,
12, 1, 2003.
Benzer, S., From the gene to behavior, JAMA, 218, 1015, 1971.
Biesiadecki M., Huber R., Holladay C., Hill S., Tononi G., and Cirelli C., Sleep homeostasis
in the fruit fly, Sleep, 26(S), A22, 2003.
Blau J., A new role for an old kinase: CK2 and the circadian clock, Nature Neuroscience, 6,
208, 2003.
Borbély A.A. and Achermann P., Sleep homeostasis and models of sleep regulation, J. Biol.
Rhythms, 14, 557, 1999.
Cirelli C. and Tononi G., Gene expression in the brain across the sleep-wakefulness cycle,
Brain Res., 885, 303, 2000.
Cirelli C. and Tononi G., Molecular correlates of sleep and waking in Drosophila, Actas de
Fisiologia, 7, 131, 2001.
Cirelli C., Hill S., Holladay C., Biesiadecki M., Martinez-Gonzalez D., Kreber R., Ganetzky
B., and Tononi G., Sleep in Drosophila melanogaster: a mutagenesis screening, Sleep,
26S, A416, 2003.
Cirelli C., Gutierrez C.M., and Tononi G., Extensive and divergent effects of sleep and
wakefulness on brain gene expression, Neuron, 41, 35–43, 2004.
Dugovic C., Laposky A.D., Losee-Olson S., and Turek F.W., Novel mutagenesis model to
screen and discover genes that affect the sleep-wake cycle, Sleep, 26S, A421, 2003.
Ebisawa T. et al., Association of structural polymorphisms in the human period3 gene with
delayed sleep phase syndrome, EMBO Rep., 2, 342, 2001.
Forest G., and Godbout R., Effects of sleep deprivation on performance and EEG spectral
analysis in young adults, Brain Cognition, 43, 195, 2000.
Franken P., Malafosse A., and Tafti M., Genetic variation in EEG activity during sleep in
inbred mice, Am. J. Physiol., 275, R1127, 1998.
Franken P., Chollet D., and Tafti M., The homeostatic regulation of sleep need is under genetic
control, J. Neurosci., 21, 2610, 2001.

Franken P. and Tafti M., Genetics of sleep and sleep disorders, Front. Biosci., 8, E381, 2003.
Friedmann J.K., A diallel analysis of the genetic underpinnings of mouse sleep, Physiol.
Behav., 12, 169, 1974.
Hartmann E. and Brewer V., When is more or less sleep required? A study of variable sleepers,
Comp. Psychiatry, 17, 275, 1976.
Hendricks J.C., Finn S.M., Panckeri K.A., Chavkin J., Williams J.A., Sehgal A., and Pack
A.I., Rest in Drosophila is a sleep-like state, Neuron, 25, 129, 2000.
Hendricks J.C., Kirk D., Pancheri K., Miller M.S., and Pack A.I. Modafinil maintains waking
in the fruit fly Drosophila melanogaster, Sleep, 26, 139, 2003.
Hobson J.A., Sleep, Scientific American Library HPHLP, New York, 1989.
Holladay C., Huber R., Biesiadecki M., Martinez-Gonzalez D., Hill S., Kreber R., Ganetzky
B., Tononi G., and Cirelli C., Natural variation in the sleep phenotype in Drosophila
melanogaster, Sleep, 26S, A23, 2003.
Horne J.A., Why we sleep: The functions of sleep in humans and other mammals, Oxford
University Press, Oxford, 1988.
Huber R., Deboer T., and Tobler I., Effects of sleep deprivation on sleep and sleep EEG in
three mouse strains: empirical data and simulations, Brain Res., 857, 8, 2000.
Huber R., Hill S., Holladay C., Biesiadecki M., Tononi G., and Cirelli C., Sleep homeostasis
in drosophila melanogaster, Sleep, 27, 628–639, 2004.
Jones H.S. and Oswald I., Two cases of healthy insomnia, Electroencephalogr. Clin. Neuro-
physiol., 24, 378, 1968.
Kaiser W., Steiner-Kaiser J., Neural correlates of sleep, wakefulness, and arousal in a diurnal
insect, Nature, 301, 707, 1983.
Katzenberg D., Young T., Finn L., Lin L., King D.P., Takahashi J.S., Mignot E., A CLOCK
polymorphism associated with human diurnal preference, Sleep, 21, 569, 1998.
Kida S., Josselyn S.A., de Ortiz S.P., Kogan J.H., Chevere I., Masushige S., Silva A.J., CREB
required for the stability of new and reactivated fear memories, Nat. Neurosci., 5,
348, 2002.
Konopka R.J., Benzer S., Clock mutants of Drosophila melanogaster, Proc. Natl. Acad. Sci.
USA, 68, 2112, 1971.
Liao G.C., Rehm E.J., Rubin G.M., Insertion site preferences of the P transposable element
in Drosophila melanogaster, Proc. Natl. Acad. Sci. USA, 97, 3347, 2000.
Lin Y.J., Seroude L., and Benzer S., Extended life-span and stress resistance in the Drosophila
mutant methuselah, Science, 282, 943, 1998.
Miklos G.L. and Maleszka R., Deus ex genomix, Nature Neuroscience, 3, 424, 2000.
Moruzzi G., The sleep-waking cycle, Ergeb. Physiol., 64, 1, 1972.
Nakanishi H. et al., Positive correlations between cerebral protein synthesis rates and deep
sleep in Macaca mulatta, Eur. J. Neurosci., 9, 271, 1979.
Nitz D.A., van Swinderen B., Tononi G., and Greenspan R.J., Electrophysiological Correlates
of Rest and Activity in Drosophila melanogaster, Curr. Biol., 12, 1934, 2002.
O’Hara B.F. and Mignot E., Genetics of sleep and its disorders, in Genetic Influences on
Neural and Behavioral Functions, Pfaff D.W., Berrettini W.H., Joh T.H., and Maxon
S.C., Eds., CRC Press, Boca Raton, FL, 2000.
Ramm P. and Smith C.T., Rates of cerebral protein synthesis are linked to slow wave sleep
in the rat, Physiol. Behav., 48, 749, 1990.
Rattenborg N.C., Amlaner C.J., and Lima S.L., Behavioral, neurophysiological and volution-
ary perspectives on unihemispheric sleep, Neurosci. Biobehav. Rev., 24, 817, 2000.
Rechtschaffen, A., Current perspectives on the function of sleep, Perspect. Biol. Med., 41,
359, 1998.

Rechtschaffen A., Bergmann B.M., Everson C.A., Kushida C.A., Gilliland M.A., Sleep dep-
rivation in the rat: X. Integration and discussion of the findings, Sleep, 12, 68, 1989.
Rechtschaffen A. and Bergmann B.M., Sleep deprivation in the rat: an update of the 1989
paper, Sleep, 25, 18–24, 2002.
Reiter L.T., Potocki L., Chien S., Gribskov M., and Bier E., A systematic analysis of human
disease-associated gene sequences in Drosophila melanogaster, Genome Res., 11,
1114, 2001.
Roberts D.B. (Ed)., Drosophila, A Practical Approach, 2nd ed., Oxford University Press, 1998.
Rogers N.L., Szuba M.P., Staab J.P., Evans D.L., Dinges D.F., Rogers N.L., Szuba M.P., Staab
J.P., Evans D.L., and Dinges D.F., Neuroimmunologic aspects of sleep and sleep loss,
Semin. Clin. Neuropsychiatry, 6, 295, 2001.
Rogina B., Reenan R.A., and Nilsen S.P., Helfand S.L., Extended life-span conferred by
cotransporter gene mutations in Drosophila, Science, 290, 2137, 2000.
Rörth P., Szabo K., Bailey A., Laverty T., Rehm J., Rubin G.M., Weigmann K., Milan M.,
Benes V., Ansorge W., and Cohen S.M., Systematic gain-of-function genetics in
Drosophila. Development, 125, 1049, 1998.
Sanyal S., Sandstrom D.J., Hoeffer C.A., and Ramaswami M., AP functions upstream of
CREB to control synaptic plasticity in Drosophila, Nature, 416, 870, 2002.
Sauer S., Herrmann E. and Kaiser W., The effect of forced activity on a behavioral sleep sign
in honey bees, Sleep Research Online, 2 (S1), 217, 1999.
Shaw P.J., Cirelli C., Greenspan R.J., and Tononi G., Correlates of sleep and waking in
Drosophila melanogaster, Science, 287, 1834, 2000.
Shaw P.J., Tononi G., Greenspan R.J., and Robinson D.F., Stress response genes protect
against lethal effects of sleep deprivation in Drosophila, Nature, 417, 287, 2002.
Sokolowski M.B., Drosophila: genetics meets behavior, Nature Reviews, 2, 879, 2001.
Spiegel K., Leproult R., and Van Cauter E., Impact of sleep debt on metabolic and endocrine
functions, Lancet, 354, 1435, 1999.
Spradling A.C., Stern D., Beaton A., Rhem E.J., Laverty T., Mozden N., Misra S., and Rubin
G.M., The Berkeley Drosophila Genome Project gene disruption project: Single P-
element insertions mutating 25% of vital Drosophila genes, Genetics, 153, 135, 1999.
Steriade M. and Timofeev I., Neuronal plasticity in thalamocortical networks during sleep
and waking oscillations, Neuron, 37, 563, 2003.
Stickgold R., Hobson J.A., Fosse R., and Fosse M., Sleep, learning, and dreams: off-line
memory reprocessing, Science, 294, 1052, 2001.
Tafti M., Franken P., Kitahama K., Malafosse A., Jouvet M., and Valatx J.L., Localization of
candidate genomic regions influencing paradoxical sleep in mice, Neuroreport, 8,
3755, 1997.
Tafti M., Chollet D., Valatx J.L., Franken P., Quantitative trait loci approach to the genetics
of sleep in recombinant inbred mice, J. Sleep Res., 8 (S1), 37, 1999.
Tafti M. and Franken P., Genetic dissection of sleep, J. Appl. Physiol., 92, 1339, 2002.
Tafti M., Petit B., Chollet D., Neidhart E., de Bilbao F., Kiss J.Z., Wood P.A., and Franken
P., Deficiency in short-chain fatty acid beta-oxidation affects theta oscillations during
sleep, Nat. Genet., 34, 320, 2003.
Tobler I., Effect of forced locomotion on the rest-acticity cycle of the cockroach, Behavior
Brain Res., 8, 351, 1983.
Tobler I., Is sleep fundamentally different between mammalian species?, Behavior Brain Res.,
69, 35, 1995.
Tobler I., Phylogeny of sleep regulation, in Principles and Practice of Sleep Medicine, M.H.
Kryger, T. Roth, and W.C. Dement, Eds., W.B. Saunders, Philadelphia, 2000, p. 72.

Tobler I. and Neuner-Jehle M., 24-H variation of vigilance in the cockroach Blaberus gigan-
teus, J. Sleep Res., 1, 231–239, 1992.
Tobler I. and Stalder J., Rest in the scorpion — a sleep-like state?, J. Comp. Physiol., A163,
227, 1988.
Toh K.L., Jones C.R., He Y., Eide E.J., Hinz W.A., Virshup D.M., Ptacek L.J., and Fu Y.H.,
Per2 phosphorylation site mutation in familial advanced sleep phase syndrome, Sci-
ence, 291, 1040, 2001.
Toth L.A. and Williams R.W., A quantitative genetic analysis of slow-wave sleep and rapid-
eye movement sleep in CXB recombinant inbred mice, Behav. Genet., 29, 329, 1999.
Toth L.A., Identifying genetic influences on sleep: an approach to discovering the mechanisms
of sleep regulation, Behav. Genet., 31, 39, 2001.
Valatx J.L., Bugat R., and Jouvet M. Genetic studies of sleep in mice, Nature, 238, 226, 1972.
Valatx J.L., Bugat R., Genetic factors as determinants of the waking-sleep cycle in the mouse,
Brain Res., 69, 315, 1974.
Van Dongen H.P.A., Maislin G., Mullington J.M., Dinges D.F., The cumulative cost of
additional wakefulness: dose-response effects on neurobehavioral functions and sleep
physiology from chronic sleep restriction and total sleep deprivation, Sleep, 26, 117,
2003.
Waddell S. and Quinn W.G., Flies, genes, and learning, Ann. Rev. Neurosci., 24, 1283, 2001.
Walker M.P., Brakefield T., Morgan A., Hobson J.A., and Stickgold R., Practice with sleep
makes perfect: sleep-dependent motor skill learning, Neuron, 35, 205, 2002.
Webb W.B. and Friel J., Sleep stage and personality characteristics of “natural” long and
short sleepers, Science, 171, 587, 1971.

9 Sleep Phylogeny:
Clues to the Evolution
and Function of Sleep
Jerome M. Siegel
CONTENTS
Introduction
Terrestrial Mammals
Aquatic Mammals
Reptiles
Conclusions
References
INTRODUCTION
A persuasive argument for the importance of sleep rests on its ubiquity among
animals. All mammals sleep.1 Reptiles appear to sleep, although by some measures
they may not.2–9 It has not been conclusively demonstrated that fish sleep, although
some species show marked circadian rhythms of activity.10,11 To meet the accepted
definition of sleep, animals must show periods of inactivity with raised arousal
thresholds and must show sleep debt when deprived, leading to rebound sleep when
deprivation is ended. Fruit flies (Drosophila melanogaster) show periods of inactivity
with raised arousal thresholds and sleep rebound after deprivation.12–14 If such periods
are homologous to sleep in vertebrates, one must consider any reported absence of
sleep in higher vertebrates as an error due to inadequate assessment or to be an
evolved adaptation to particular ecological niches that has done away with a sleep
state present in ancestral animals. This chapter discusses the special situation of
marine mammals, which appear to have evolved adaptations that at the very least
mask some aspects of sleep and certainly dispense with the need for immobility
during what otherwise appears to be sleep. A further issue is the nature of sleep. Most
mammals1,15 and birds16 show evidence of REM sleep — also known as paradoxical
sleep (PS) — although this state may not exist in certain marine mammals.17
Amounts of sleep differ substantially between species, with some sleeping as
little as 2 h per day and others as much as 20 h.1,15 Surely these enormous variations
0-8493-1519-0/05/$0.00+$1.50

offer some insight into the physiological needs responsible for sleep. Although
animals in different ecological niches might most adaptively have evolved differing
durations of activity and inactivity, it is unlikely that no animals would have evolved
a complete or nearly complete absence of sleep unless it served some vital function.
The cost of sleep in terms of vulnerability, loss of time to eat, procreate, and gain
an edge in competition with other animals is considerable. Certainly contemporary
humans make major efforts to reduce sleep time to achieve their goals. Differences
in sleep amounts seem to be systematically related to certain constitutional variables,
suggesting that underlying physiological factors, rather than ecological niche, deter-
mine sleep need. The study of sleep phylogeny can help explain the essence of sleep
debt; i.e., which physiological, neurochemical, and genetic events are conserved
across sleep in differing animals.
TERRESTRIAL MAMMALS
Although there are approximately 4,000 mammalian species, fewer than 100 have
been studied under laboratory conditions. Most of these have been observed in only
a single study. Perhaps an additional 100 have been observed in zoos. Certainly there
is no need to study sleep in all mammalian species; however it is likely that a thorough
examination of sleep physiology, exploring the genetic variations and adaptations
that have occurred over more than 100,000,000 years of mammalian evolution, may
reveal aspects of sleep not seen in the four or five laboratory species that have been
most thoroughly studied. For example, humans and rats have been shown to have a
clear link between REM sleep and penile erections.16,19 A recent study of sleep in
the armadillo revealed that penile erections occur in non-REM sleep, but not in REM
sleep in this species.20 Such observations are not merely a curiosity but speak to the
issue of which aspects of sleep are core phenomena and which are perhaps epiphe-
nomena not linked to particular sleep-waking states. In this case, the findings suggest
that certain aspects of sympathetic and parasympathetic control during sleep differ
across species. One may speculate that other aspects of standard sleep signs in rats,
cats, and humans, such as high voltage electroencephalogram (EEG) during non-
REM sleep, low voltage EEG during REM sleep or high voltage EEG occurring
simultaneously in both hemispheres, may not be essential for sleep. Many of the
largest mammals such as elephants, and giraffes have only been studied by visual
observation. Understanding sleep in these animals is crucial, because the extreme
points in any cross species comparison can be most informative as to the underlying
variables that determine sleep amounts and physiology.
Perhaps the most surprising conclusion from studies of mammalian sleep is that
knowing the order to which an animal belongs tells you very little about the amount
of total sleep or REM sleep they have.1,15 In other words, as a group rodents do not
have characteristic sleep patterns that differentiate them from carnivores, primates,
artiodactyls, insectivores, and so on. Each of these groups shows a wide and over-
lapping range of total and REM sleep amounts. Each order is characterized by a
common genetic inheritance that produces characteristic behaviors, brain and body
anatomy, intelligence, diet, and reproductive physiology that tends to differentiate
it from other orders. Yet their sleep is not characteristic of the group, suggesting that

TABLE 9.1
Correlations between Sleep Parameters and Constitutional Variables
Total Daily Quiet Sleep REM Sleep REM Sleep Cycle
Sleep Time Time Time Sleep% Length
Body weight –0.53a –0.53a –0.45a –0.12 0.83a

Brain weight –0.55a –0.48a –0.52a –0.25 0.89a
Metabolic rate 0.33b 0.30b 0.13 –0.09 0.82
Encephalization quotient –0.17 –0.10 –0.20b –0.30b 0.52b
a P<0.001.
b P£0.05.
these variables do not determine sleep amount. A comprehensive analysis of the

determinants of sleep time looking at body weight, metabolic rate, brain weight,
encephalization quotient (brain-to-body weight ratio), body temperature, neonatal
brain weight as a percentage of adult weight, gestation period, and litter size con-
cluded that total sleep time was most closely correlated with body weight15 (Table
9.1). This was a negative correlation; i.e., big animals sleep less. Body weight is
inversely related to metabolic rate, so we can say that animals with higher metabolic
rate have more total sleep time. The implications of this is discussed in the Conclu-
sions section.
Among terrestrial mammals REM sleep amounts are positively correlated with
total sleep amounts; however this explains only a small amount of the total variance
in REM sleep time. It has been noted that predator animals and animals with safe
sleeping sites have relatively larger amounts of REM sleep.21 This makes some sense,
because arousal thresholds are elevated in some animals in REM sleep, so it might
be dangerous for prey animals to have large amounts of REM sleep., It is not true,
however, that REM sleep is deep sleep in all animals.22 In humans, for example,
arousal from REM sleep is more rapid than from non-REM sleep, and there is
evidence that in general animals aroused from REM sleep function better than those
aroused from non-REM sleep.23 It is difficult to quantify safety of sleep site by
measures such as frequency of death during sleep. Sites that seem exposed may in
fact be safe, and there is little evidence that animals are disproportionately hunted
during sleep. Therefore, while there is little doubt that certain predator animals have
large amounts of REM sleep, this relation does not appear to adequately explain
REM sleep time.
An alternate correlate of REM sleep time is how immature animals are at birth.
At birth all mammals so far examined have their maximal amounts of REM sleep.
Amounts diminish with age to adult levels.24 Animals such as rats or cats that are
born relatively immature have a greater elevation in REM sleep at birth. Animals
such as horses or guinea pigs that are born relatively mature have little elevation
of REM sleep at birth. This suggests that REM sleep may have some role in brain
or body development or in the protection of small animals without substantial
thermoregulatory capacity from hypo- or hyperthermia. This role has not been
identified. Furthermore, even though animals that are immature at birth have

TABLE 9.2
Correlations of REM Sleep Parameters with Measures of
Neonatal Maturity and Reproductive Variables
REM Sleep REM Sleep% of Total
Time Sleep Time
Altricial-precocial rating –0.66a –0.45b

Neonatal brain weight (% adult) –0.61a –0.55b
Litter size 0.51a 0.41b
a P<0.001.
b P£0.05.
decreasing REM sleep amounts as they age, they continue to have higher REM
sleep amounts when they reach adulthood (Table 9.224). No theory has been offered
as to why this is; however, from a statistical standpoint, the correlation between
immaturity at birth and REM sleep time in adulthood accounts for a large amount
of the interspecies variability in REM sleep time between mammals. Figure 9.125
shows some mammals with relatively high and low amounts of REM sleep. It is
important to note that humans do not have unusual amounts of REM sleep either
in terms of the number of hours per day or the percent of sleep time devoted to
REM sleep (Figure 9.1); rather the amount of REM sleep time shown by humans
is in line with our intermediate state of maturity at birth. This is obviously a problem
for any theory hypothesizing that REM sleep amount is linked to intellectual capac-
ity or any other characteristic in which humans are believed to be at an extreme
within the animal kingdom.25
The monotremes are one of the three branches of the mammalian line, the other
two being the placentals and the marsupials.1 The extant monotremes are the short-
and long-nosed echidna and the platypus. The monotremes are egg-laying mammals
that have relatively low but regulated body temperature (approximately 32∞C). They
nurse their young from milk-secreting patches, rather than nipples and have thick
fur. Their bone structure contains some reptilian characteristics, and genetic analysis
indicates that they are more similar to reptiles and birds than other mammals. The
platypus has a bill that responds to electric fields and a poison spur, characteristics
typically seen in reptiles or fish but not in mammals. Despite the origin of
monotremes early in the mammalian line, relatively little speciation has occurred,
with only five monotreme species known to have evolved, presumably because their
geographic isolation from other species reduced evolutionary pressure.1 Thus the
physiology of monotremes is likely to more closely resemble that of the first mam-
mals than any other mammals, and an early report that echidnas did not have REM
sleep generated considerable interest.26,27 It suggested that REM sleep was a more
recently evolved state with some higher cognitive function.
Because of the possibility that a REM sleep-like state might be missed in the
echidna, we reexamined this issue. In addition to recording electroencephalograms
and electromyograms, we monitored brainstem neuronal activity.28 We know that

FIGURE 9.1 Total sleep amounts and REM sleep amounts. Humans are not unusual either in terms of their total sleep or REM sleep amounts. (From
Siegel J.M., The REM sleep-memory consolidation hypothesis, Science, 294, 1058–1063, 2001. With permission.)

brainstem neuronal activity generates REM sleep29; therefore it might be possible

to detect a REM sleep-like state in the echidna even if the forebrain EEG did not
resemble that of REM sleep in placental mammals. We found that brainstem activity
during sleep in the echidna did not resemble the activity seen in other mammals
during non-REM sleep. Rather it resembled that seen in REM sleep (Figure 9.2).
We concluded that echidnas did have a REM sleep-like state, but one that was not
accompanied by low voltage cortical EEG as is seen in adult mammals. In this
respect the REM sleep-like state resembled that of many neonatal animals, which
have high voltage activity during periods of REM sleep. It has also been reported
that a state looking like REM sleep, with low voltage EEG, may occur in echidnas.30
We saw no such state in our studies, and the Nicol et al. study did not demonstrate
that the state they were observing was a sleep state, rather than a quiet waking state,
so further investigation of this issue may be warranted. Both studies agree, however,
that the echidna has a REM sleep-like state, in contrast to the earlier work.
Because we saw a state that resembled REM sleep in the echidna, we next
studied sleep in the platypus, which is considered the most primitive of the mam-
mals.31 This raised special problems, because these animals are very delicate, are
dangerous to handle because of their poison spurs, and are partially aquatic and
cannot be housed in conventional cages. The platypus requires special animal hus-
bandry procedures. Water pumps, needed to circulate water in their pool, generate
large electrical fields, which stress and thereby can cause the death of the animals.
Shielding procedures have to be used to minimize this stimulus. Telemetry, both on
land and under water, is necessary to allow continuous recording. When we suc-
ceeded in recording from the platypus, we found that it has a particularly vigorous
motor activation during sleep, equal to or greater than that of other mammals in
REM sleep (Figure 9.3). A video of this can be viewed at our web site http://www.
npi.ucla.edu/sleepresearch. When we calculated the amount of REM sleep, we were
surprised to discover that the platypus had more REM sleep than any other animal,
up to 8 hours per day. Both the echidna and platypus are extremely immature at
birth, resembling a worm. They crawl out of the birth canal and into a pouch where
they are protected, warmed, and nourished for several months. The high levels of
REM sleep in monotremes strengthen the relation between immaturity at birth and
REM sleep amounts in adulthood.
AQUATIC MAMMALS
Aquatic mammals have sleep patterns that are quite different from those in terrestrial
mammals, so investigation of sleep in marine mammals may be instructive in under-
standing sleep as a whole, as well as the role of REM versus non-REM sleep.
Under some conditions dolphins swim 24 hours a day for long periods. During
their swimming they breathe regularly and are able to avoid the sides of the pool.
Lilly first noticed that dolphins often close one of their eyes but rarely close both.
The significance of this was discovered by Lev Mukhametov and colleagues, who
developed reversible, relatively noninvasive techniques for recording EEG during
swimming. They found that dolphins generated the high voltage EEG typical of non-
REM sleep in either the right or left side of their cortex, but never in both sides.32

FIGURE 9.2 Instantaneous rate plots of medial reticular neurons during sleep and waking states. In the echidna, rate varies in a way resembling that
seen in placental mammals during REM sleep, rather than that in the regular manner of non-REM sleep. (From Siegel J.M., Manger P., Nienhuis R.,
Fahringer H.M., Pettigrew J., The echidna Tachyglossus aculeatus combines REM and non-REM aspects in a single sleep state: implications for the
evolution of sleep, J. Neuroscience, 16, 3500–3506, 1996. With permission.)

FIGURE 9.3 Sleep states in the platypus. The platypus has periods of rapid eye movements
during a state characterized by high voltage EEG. (From Siegel J.M., Manger P.R., Nienhuis
R., Fahringer H.M., Shalita T., Pettigrew J.D., Sleep in the platypus, Neuroscience 91,
391–400, 1999. With permission.)
The same unihemispheric sleep has now been seen in several cetacean species.33
Figure 9.4 shows an EEG recording we carried out in a beluga whale. USWS appears
to be at least partially an adaptation to the complex brain activity required for
breathing in the dolphin. In contrast to other animals that breathe automatically during
sleep, dolphins and other cetaceans need to be at the surface to breathe, need to sense
wave action, and minimize water ingestion during breathing movements. Adminis-
tration of light doses of barbiturates to dolphins will stop breathing (long before it
produces effective analgesia). This is quite different from terrestrial mammals that
breathe and regulate blood gasses effectively even when deeply anesthetized.
In the dolphin the optic chiasm is completely crossed so that all visual input to
each hemisphere comes from the opposite eye. Because visual input can block certain

EEG rhythms,34 it is necessary to determine if any EEG change observed contralat-

eral to a closed eye is secondary to the reduction of visual input. Generally the eye
contralateral to the hemisphere that was sleeping was closed, as one would expect,
because it is unlikely that visual input could be effectively processed in the hemi-
sphere with EEG synchronization; however we also demonstrated that eye closure
was not the cause of the high voltage EEG Closure or covering of one eye did not
necessarily produce high-voltage activity in the opposite hemisphere. Conversely
we noted that the eye opposite the sleeping hemisphere could open without blocking
the high voltage EEG33; therefore the EEG-defined sleep state is not simply a
consequence of reduced visual input to the contralateral eye. A similar control for
vision-related EEG changes has not been done to support claims of USWS in birds,
who in any case show relatively subtle differences between the EEG in the two
hemispheres, consistently detectable only with power spectral analysis,35 in contrast
to the USWS visible in cetaceans. If one hemisphere is prevented from showing
high-voltage EEG by gently stimulating dolphins, a rebound of sleep in the deprived
hemisphere is seen when deprivation stops,36 important evidence that sleep debt can
be localized to one hemisphere.
Despite many studies, no convincing evidence of REM sleep in dolphins or any
other EEG instrumented cetacean has been produced. The absence of such evidence
may result from the conditions under which these observations have been made.
Often the animals are restrained during recordings. In those cases where they have
been unrestrained, the stimulation caused by the recording cable may have prevented
appearance of REM sleep but, considering the strong REM sleep pressure shown by
all terrestrial mammals that have been deprived and that allows high levels of REM
sleep even under uncomfortable sleeping conditions, one would expect that some
unequivocal REM sleep would have been seen even under less than natural condi-
tions. One can conclude that if REM sleep exists in the dolphin, REM sleep amounts
are among the smallest of any mammal or are uniquely sensitive to disturbance.
A more subtle issue is whether REM sleep may take some novel form in dolphins
that has escaped detection. An alternate hypothesis is that unihemispheric slow-wave
sleep may eliminate the need for REM sleep; for example, if REM sleep has evolved
to stimulate brainstem areas after non-REM sleep to allow optimal functioning in
subsequent waking, the presence of virtually continuous brainstem activity required
for the continuous movement and breathing shown by dolphins may make REM
sleep unnecessary.
REPTILES
The presence of REM sleep in large amounts in the most primitive mammals and
in birds suggests that it may have been present in a common ancestor of these two
classes of animals. That would indicate that at least some reptiles have REM sleep.
The alternate theory, that REM sleep evolved twice, once in mammals and once in
birds, suggests that a REM sleep precursor state must have existed in pre-avian,
premammalian reptiles. According to both hypotheses, examination of state organi-
zation in reptiles would provide an insight into the primitive aspects of REM sleep.
The key challenge is devising a method that would be effective in detecting such a

state. To do this we decided to follow the approach we used in the echidna. Because
we knew what the pattern of brainstem neuronal activity is in the mammalian REM
sleep state, and because midbrain and pontine brainstem regions are both necessary
and sufficient for generating the major neurological changes seen in REM sleep,29
we decided to conduct the first investigations searching for aspects of REM sleep
at the neuronal level in reptiles.37 We chose the turtle as a representative reptile
because excellent prior behavioral studies had been conducted on these animals7 and
because they adapted well to the laboratory.
FIGURE 9.4

At the neuronal level, there are two consistent brainstem activity changes under-
lying REM sleep. One is the burst-pause discharge pattern that gives rise to the rapid
eye movements and twitching characteristic of REM sleep. This pattern is present
in most medial reticular neurons and therefore should be relatively easy to detect.
The second is the cessation of release of norepinephrine, serotonin and histamine
during REM sleep. This would be much more difficult to detect, because these
monoaminergic cell groups are intermingled with other cell types and there is no
easy way to determine the transmitter phenotype of any recorded cell. Our immu-
nohistochemical analyses showed, however, that these cell groups were present in
the turtle, so we focused our effort on recording neuronal activity from medial
reticular cells during quiescent states, expecting to record from non-monoaminergic
and possibly monoaminergic cells.
Our results were clear. We saw no acceleration of discharge and no burst-pause
pattern of discharge analogous to that seen in mammalian reticular cells during sleep.
It appears that this aspect of REM sleep is not present in any form in turtles. We
do not know the pattern of discharge of monoamine cells in the sleep of the turtle,
and it is possible that a cessation of discharge occurs during behavioral immobility.
However, we did not see cells that had the tonic waking discharge with cessation
of discharge within the sleep period even though some of the neurons we recorded
were within the serotonergic raphe region. Further investigations are necessary to
test for this possibility. What we can conclude is that the periodic occurrence of
brainstem activation that is so characteristic of REM sleep in terrestrial mammals
is absent in the turtle. REM sleep precursor states may be present in the reptilian
species that gave rise to mammals and birds but not in modern day turtles. Alterna-
tively the phasic motor activation seen in REM sleep may have evolved rapidly at
the onset of the avian and mammalian lines, perhaps in relation to homeothermy.
FIGURE 9.4 (See facing page.) Relationship between EEG and the state of eyes in a beluga
whale. (A) The state of eyelids and EEG spectral power (1–3 Hz; 5-sec epochs) from the two
hemispheres (R, right; L, left) in a white whale recorded over a 3-h period. EEG power was
normalized as a percentage of the maximal power in each hemisphere during this period. The
state of each eye (R, right; L, left) was scored in real time (O, open; I, intermediate; or C,
closed) and then categorized for 5-sec epochs as described. Compressed figure does not show
short-lasting changes in eye state. (B) Expansion of the two 2.5-min recordings of the EEG
and the state of both eyes. The examples show the EEG asynchrony and parallel changes in
eye state recorded in this whale at the times marked as 1 and 2 in Figure 4 A. Note that the
EEG does not change immediately with changes in eye position. The right eye did not close
during episode 1, and the left eye did not close during episode 2. (C) The average EEG
spectral power in the two hemispheres during episodes with unilateral eye opening (LO/RC,
left open and right closed; LC/RO, left closed and right open). EEG power was normalized
as a percentage of the average 1–3 Hz power recorded in each hemisphere during SWS with
the contralateral eye closure. Reported values are the means S.E. (LO/RC, n = 238 epochs;
LC/RO, n = 441 epochs). (From Lyamin, O.I., Mukhametov, L.M., Siegel, J.M., Nazarenko,
E.A., Polyakova, I.G., and Shpak, O.V., Unihemispheric slow wave sleep and the state of the
eyes in a white whale, Behavioral Brain Res., 129, 125, 2002. With permission.)

CONCLUSIONS
The ultimate question for sleep researchers is the function of REM and non-REM
sleep. Phylogenetic evidence constrains any theory attempting to answer these ques-
tions. We know that sleep amounts vary by more than an order of magnitude across
mammalian species. Either the amount of time spent sleeping has no relation to
underlying function, which would distinguish sleep from many other homeostatically
regulated processes, or sleep need varies considerably across species. The correlates
of this variation should provide some insight into sleep functions.
A survey of the available data indicates that phylogenetic order does not explain
much of the variation of sleep time across species. There is extensive overlap of
both REM and non-REM sleep time between orders, despite the genetic, anatomical,
physiological, and behavioral commonalities within order. Prior data and new data
on primitive mammals and cetaceans indicate a strong negative correlation between
total sleep time and weight. Because metabolic rate is strongly and negatively
correlated with body mass, this is also a positive correlation between metabolic rate
and sleep time. Some evidence suggests that brain regions with high metabolic rate
have higher levels of sleep deprivation-induced damage. We hypothesized that sleep
serves to repair damage caused by oxidative stress.38,39
REM sleep amounts are positively correlated with non-REM sleep amounts,
suggesting that REM sleep may work in concert with non-REM sleep. One persistent
hypothesis that has been raised in several forms is that REM sleep serves to stimulate
the brain to prepare for waking after a period of non-REM sleep.23,40,41
Most of the variation in REM sleep amounts is independent of non-REM sleep
duration. The phylogenetic data indicate that animals born in a relatively immature
state have more REM sleep early in development. One may hypothesize that in these
immature animals REM sleep’s activation of the brain facilitates development. In
animals that are more mature at birth, this process may have occurred in utero and
continued postnatally in their direct interactions with the environment in waking.
Immature animals are obviously not able to interact with the environment in the
same way. A major mystery that remains is why immaturity at birth should be
correlated with REM sleep time in adulthood.
Marine mammals have sleep patterns that differ greatly from those seen in other
animals. They show unihemispheric sleep, with both hemispheres never being in
deep sleep at the same time. They can sleep while swimming, apparently controlling
muscles bilaterally. Finally, they appear to have little or no REM sleep. Understand-
ing the mechanisms and functional relations underlying these unusual sleep adap-
tations of marine mammals can offer a major insight into the function and mecha-
nisms of sleep.
REFERENCES
1. Siegel, J.M., The evolution of REM sleep, in Lydic, R., Baghdoyan, H.A., Eds.,
Handbook of Behavioral State Control, CRC Press, Boca Raton, FL, 1999, p. 87.
2. Flanigan, W.F., Sleep and wakefulness in iguanid lizards, Ctenosaura pectinata, and
Iguana iguana, Brain Behav. Evol., 8, 401, 1973.

3. Huntley, A.C., Electrophysiological and behavioral correlates of sleep in the desert

iguana, Dipsosaurus dorsalis hallowell, Comp. Biochem. Physiol., 86A, 325, 1987.
4. Ayala-Guerrero, F., Huitron-Resendiz, S., and Mancilla, R., Characterization of the
raphe nuclei of the reptile Ctenosaura pectinata, Physiol. Behav., 50, 717, 1991.
5. DeVera, L., Gonzalez, J., and Rial, R.V., Reptilean waking EEG: slow waves, spindles,
and evoked potentials, Electroenceph. Clin. Neurophysiol., 90, 298, 1994.
6. Flanigan, W.F., Sleep and Wakefulness in Chelonian Reptiles II. The Red-Footed
Tortoise, Gechelone Carbonaria, Arch. Ital. Biol., 112, 253, 1974.
7. Flanigan, W.F., Knight, C., Hartse, and K., Rechtschaffen, A., Sleep and wakefulness
in chelonian reptiles. 1. The box turtle, Terrapene carolina, Arch. Ital. Biol., 112,
227, 1974.
8. Flanigan, W.F., Knight, C.P., Hartse, and K.M., Rechtschaffen, A., Sleep and wake-
fulness in chelonian reptiles I. — the box turtle, Terrapene carolina, Arch. Ital. Biol.,
112, 227, 1974.
9. Peyrethon, J. and Dusan-Peyrethon, D., Etude polygraphique de cycle veille-sommeil
chez trois genres de reptiles, Seanc. Soc. Biol., 162, 181, 1968.
10. Shapiro, C.M. and Hepburn, H.R., Sleep in a Schooling Fish, Tilapia, mossambica,
Physiology and Behavior, 16, 613, 1976.
11. Tobler, I. and Borbely, A.A., Effect of rest deprivation on motor activity of fish, J.
Comp. Physiol. A., 157, 817, 1985.
12. Shaw, P.J., Tononi, G., Greenspan, R.J., and Robinson, D.F., Stress response genes
protect against lethal effects of sleep deprivation in Drosophila, Nature, 417, 287,
2002.
13. Shaw, P.J., Cirelli, C., Greenspan, R.J., and Tononi, G., Correlates of sleep and waking
in Drosophila melanogaster, Science, 287, 1834, 2000.
14. Hendricks, J.C., Finn, S.M., Panckeri, K.A., Chavkin, J., Williams, J.A., Sehgal, A.,
Pack, A.I., Rest in Drosophila is a sleep-like state, Neuron, 25, 129, 2000.
15. Zepelin, H., Mammalian sleep, in Kryger, M.H., Roth, T., and Dement, W.C., Eds.,
Principles and Practice of Sleep Medicine, W.B. Saunders, Philadelphia, 2000, p. 82.
16. Amlaner, C.J. and Ball, N.J., Avian sleep, in Kryger, M.H., Roth, T., Dement, W.C.,
Eds., Principles and Practice of Sleep Medicine, W.B. Saunders Company, Philadel-
phia, 1994, p. 81.
17. Mukhametov, L.M., Supin, A.Y., and Polyakova, I.G., Interhemispheric asymmetry
of the electroencephalographic sleep patterns in dolphins, Brain Res., 134, 581, 1977.
18. Schmidt, M.H., Valatx, J.L., Sakai, K., Fort, P., and Jouvet, M., Role of the lateral
preoptic area in sleep-related erectile mechanisms and sleep generation in the rat, J.
Neurosci., 20, 6640, 2000.
19. Moore, C.A., Fishman, I.J., and Hirshkowitz, M., Evaluation of erectile dysfunction
and sleep-related erections, J. Psychosom. Res., 42, 531, 1997.
20. Affanni, J.M., Cervino, C.O., and Marcos, H.J., Absence of penile erections during
paradoxical sleep, Peculiar penile events during wakefulness and slow wave sleep in
the armadillo, J. Sleep Res., 10, 219, 2001.
21. Allison, T. and Cicchetti, D.V., Sleep in mammals: Ecological and constitutional
correlates, Science, 194, 732, 1976.
22. Rechtschaffen, A., Siegel, J.M., Sleep and Dreaming, in Kandel, E.R., Schwartz, J.H.,
and Jessel, T.M., Eds., Principles of Neuroscience, McGraw Hill, New York, 2000, p. 936.
23. Snyder, F., Toward an evolutionary theory of dreaming, Amer. J. Psychiat., 123, 121,
1966.
24. Jouvet-Mounier, D., Astic, L., and Lacote, D., Ontogenesis of the states of sleep in
rat, cat, and guinea pig during the first postnatal month, Dev. Psychobiol., 2, 216, 1970.

25. Siegel, J.M., The REM sleep-memory consolidation hypothesis, Science, 294, 1058,
2001.
26. Allison, T., Van Twyver, H., and Goff, W.R., Electrophysiological studies of the
echidna, Tachyglossus aculeatus, I. Waking and sleep, Arch. Ital. Biol., 110, 145,
1972.
27. Allison, T. and Van Twyver, H., Electrophysiological Studies of the Echidna, Tachy-
glossus aculeatus, II. Dormancy and Hibernation, Arch. Ital. Biol., 110, 185, 1972.
28. Siegel, J.M., Manger, P., Nienhuis, R., Fahringer, H.M., and Pettigrew, J., The echidna
Tachyglossus aculeatus combines REM and non-REM aspects in a single sleep state:
implications for the evolution of sleep, J. Neuroscience, 16, 3500, 1996.
29. Siegel, J.M., Brainstem mechanisms generating REM sleep. In, Kryger, M.H., Roth,
T., Dement, and W.C., Eds., Principles and Practice of Sleep Medicine, W.B. Saunders
Company, 2000, p. 112.
30. Nicol, S.C., Andersen, N.A., Phillips, N.H., and Berger, R.J., The echidna manifests
typical characteristics of rapid eye movement sleep, Neurosci. Lett. 2000, Mar. 31;
28, 3 (1): 49–52, 283, 49, 2000.
31. Siegel, J.M., Manger, P.R., Nienhuis, R., Fahringer, H.M., Shalita, T., and Pettigrew,
J.D., Sleep in the platypus, Neuroscience, 91, 391, 1999.
32. Mukhametov, L.A., Supin, A.Y., and Polyakova, I.G., Interhemispheric asymmetry
of the electroencephalographic sleep patterns in dolphins, Brain Res., 134, 581, 1977.
33. Lyamin, O.I., Mukhametov, L.M., Siegel, J.M., Nazarenko, E.A., Polyakova, I.G.,
and Shpak, O.V., Unihemispheric slow wave sleep and the state of the eyes in a white
whale, Behavioral Brain Res., 129, 125, 2002.
34. Pfurtscheller, G., Neuper, C., and Mohl, W., Event-related desynchronization (ERD)
during visual processing, Int. J. Psychophysiol., 16, 147, 1994.
35. Rattenborg, N.C., Amlaner, C.J., and Lima, S.L., Unilateral eye closure and inter-
hemispheric EEG asymmetry during sleep in the pigeon (Columba livia), Brain Behav.
Evol., 58, 323, 2002.
36. Oleksenko, A.I., Mukhametov, L.M., Polykova, I.G., Supin, A.Y., and Kovalzon, V.M.,
Unihemispheric sleep deprivation in bottlenose dolphins, J. Sleep Res., 1, 40, 1992.
37. Eiland, M.M., Lyamin, O.I., and Siegel, J.M., State-related discharge of neurons in
the brainstem of freely moving box turtles, Terrapene carolina major, Arch. Ital.
Biol., 139, 23, 2001.
38. Eiland, M.M., Ramanathan, L., Gulyani, S., Gilliland, M., Bergmann, B.M.,
Rechtschaffen, A., and Siegel, J.M., Increases in amino-cupric-silver staining of the
supraoptic nucleus after sleep deprivation, Brain Res., 945, 1, 2002.
39. Ramanathan, L., Gulyani, S., Nienhuis, R., and Siegel, J.M., Sleep deprivation
decreases superoxide dismutase activity in rat hippocampus and brainstem, Neurore-
port, 13, 1387, 2002.
40. Ephron, H.S. and Carrington, P., Rapid eye movement sleep and cortical homeostasis,
Psychol. Rev., 73, 500, 1966.
41. Vertes, R.P., A life-sustaining function for REM sleep: a theory, Neurosci. Biobehav.
Rev., 10, 371, 1986.

10 Sleep, Synaptic Plasticity,

and the Developing Brain
Marcos Gabriel Frank
CONTENTS
Introduction
Ontogenesis of Mammalian Sleep and Sleep Regulation
Dissociation
Concordance
Maturation
Experimental Approaches to Neonatal Sleep Function
Correlation- and Association-Based Studies and Pharmacological
Sleep Suppression
Sleep and Visual System Development
Sleep and the Developing Lateral Geniculate Nucleus (LGN)
Sleep and Developmentally Regulated Cortical Plasticity
Further Considerations
Theories of Sleep Function in Developing Animals
The Ontogenetic Hypothesis
Sleep and the Consolidation of Experience
Summary
References
INTRODUCTION
In a variety of mammalian species, sleep amounts are highest during neonatal periods
of rapid brain development and synaptic plasticity than at any other time of life;24,39,64
therefore if sleep contributes to synaptic plasticity, one would expect this to be
especially true in developing animals. This chapter reviews evidence in support of
this hypothesis. It begins with an overview of several landmark events in the onto-
genesis of sleep and sleep regulation to provide context to the more function-based
discussions that follow. It then discusses the results of several studies that provide
indirect or suggestive evidence of a role for sleep in general brain maturation. This
is followed by a review of findings in the developing visual system that more
specifically address a possible role for sleep in developmental synaptic plasticity.
0-8493-1519-0/05/$0.00+$1.50

The chapter concludes with a discussion of several theories regarding sleep function
in developing animals.
ONTOGENESIS OF MAMMALIAN SLEEP AND

SLEEP REGULATION
The ontogenesis of mammalian sleep can be broadly divided into three general stages,
here referred to as the dissociation, concordance, and maturation stages.14 The dis-
sociation stage of sleep ontogeny is characterized by the absence of clear polysom-
nographic features of REM and non-REM sleep. The concordance stage represents
the period of time when distinct, immature forms of REM and non-REM sleep are
first detected. The further maturation of these immature sleep states and the emer-
gence of regulatory sleep mechanisms occur in the third developmental stage.14
DISSOCIATION
Recordings of electrographic and autonomic activities in very young, developing
mammals do not reveal clear signs of REM and non-REM sleep, reflecting the
extreme immaturity of the nervous system at this time.14,26 Although distinct cou-
plings of autonomic and brain activities typical of adult sleep are not observed,
independent oscillations in these systems can occur. In precocial species, such as
the guinea pig, dissociation is present in the fetal period. In altricial species, which
complete a larger portion of their neural development ex utero, this stage appears
to extend into the postnatal period.14,26
CONCORDANCE
During the concordance stage of sleep ontogeny, independent oscillations in auto-
nomic function and brain activity begin to coalesce into discrete episodes that appear
to be immature forms of REM and non-REM sleep. In precocial species and humans,
this concordance begins in utero, whereas in altricial species this begins ex utero,
generally in the second postnatal week. The precise timing of this event in altricial
species is not known, with some investigators reporting the presence of pre-EEG
‘precursor’ states several days before the appearance of EEG defined vigilance
states.14,23 The nature of these putative precursor states is controversial. According
to some investigators, the precursor states known as active sleep and quiet sleep are
homologous to REM and non-REM sleep; however it is also possible that they are
more closely related to the spontaneous cyclic activity typical of the immature
nervous system.14,23
MATURATION
In the third stage of sleep ontogeny, the now polysomnographically identified states
of REM and non-REM sleep rapidly develop and begin to more closely resemble
adult forms of sleep. There are rapid increases in the amplitude of the EEG in both
non-REM and REM sleep, and stereotyped patterns of neuronal activity, such as

pontine-geniculate-occipital (PGO) waves in REM sleep are first observed.14 In

addition, distinct ultradian, homeostatic, and circadian regulatory mechanisms begin
to organize sleep into patterns similar to those found in adult mammals. In the rat
homeostatic regulation of non-REM sleep is present soon after the appearance of
EEG-defined states (postnatal day [P]12). Sleep deprivation in preweanling neonatal
rats results in robust compensatory changes in non-REM sleep amounts and con-
solidation, followed in the fourth postnatal week by increases in non-REM slow-
wave activity.28 Conversely homeostatic regulation of REM sleep has a much more
protracted development. Total and selective REM sleep deprivation fails to induce
compensatory REM rebounds until the third postnatal week in the rat.21,28 Much less
is known regarding ultradian and circadian organization of neonatal sleep states;
e.g., in the developing cat, ultradian periodicities in REM and non-REM sleep are
generally not observed until the third or fourth postnatal week.36 Circadian regulation
of sleep and wake may begin as early as the second postnatal week in the rat but is
not consistently reported until the fourth-fifth postnatal week.23
In addition to the initial absence of regulatory mechanisms and the presence of
rudimentary forms of neurophysiologic activity, neonatal sleep during this stage
differs from adult sleep in several important ways:
• The amount of REM sleep is greatly elevated, declining to adult levels

during the course of postnatal maturation.24,39,64
• Latencies to REM sleep are shortened, and sleep-onset REM periods
(SOREMs) are quite frequent.30,39,48
• Sleep is generally more fragmented, possibly reflecting the absence of
strong circadian and ultradian mechanisms at these ages.14
EXPERIMENTAL APPROACHES TO NEONATAL

SLEEP FUNCTION
CORRELATION- AND ASSOCIATION-BASED STUDIES AND PHARMACOLOGICAL
SLEEP SUPPRESSION
This section reviews findings from two classes of experiments that provide evidence
that sleep has a generally permissive effect on brain development. The first class of
experiments has shown associations or correlations between the amount of sleep, or
sleep phasic activity, and certain indices of brain development. Mirmiran and col-
leagues reported that placing juvenile rats in enriched environments resulted in
increased brain weight and increased amounts of REM sleep.53 Similar enhancements
of REM sleep are also reported in adult animals during learning tasks and may be
necessary for the consolidation of the learned material.5 Associations were also
reported between the frequency of REMs and subsequent eye-opening in the rat,
suggesting that the former events represent preparatory activation of visual motor
circuits.78 Further suggestive evidence for a role for sleep in brain development comes
from a study in the ovine fetus. In agreement with findings in adult animals,54,61
Czikk and colleagues found that cerebral protein synthesis (as measured by 14[C]L

eucine uptake) was elevated during fetal non-REM sleep, suggesting that this sleep
state may promote morphological or structural changes in the developing brain.10
The second class of experiments employ REM sleep deprivation (RSD) in the
postnatal period followed by behavioral, neurological, and biochemical assessments
in adulthood. Because sleep pressure is very high in developing animals, the majority
of these experiments have used pharmacological means of RSD (anti-depressant
medications, or related REM sleep-inhibiting compounds). Pharmacological RSD
in the neonatal period is reported to induce a number of neurochemical and behav-
ioral changes in adult rats, including changes in REM sleep architecture,8,51,52,80
circadian rhythms,17,41,83 anxiety, and sexual behavior,32–34,79 and alterations in neu-
rotransmission in cholinergic and monoaminergic sytems.31,34,59,60 However many of
the behavioral effects are not uniform across studies (even within the same labora-
tory) and vary depending on the drug used in a given experiment.22,25 The interpre-
tation of these results is further complicated by the fact that it is unknown if the
observed deficits are caused by REM sleep loss or non-specific teratogenetic effects
induced by these compounds.
There are additional reasons to doubt the claim that neonatal REM sleep sup-
pression is an important factor in the reported results following pharmacological
RSD. Gentle forms of mechanical RSD do not produce the suite of deficits reported
after neonatal clomipramine exposure.50 More vigorous mechanical RSD is reported
to produce some effects similar to drug-induced behavioral changes,20 but regrettably
the technique employed (periodic shaking of the rat pup) replaces one confounding
variable (nonspecific teratogenicity) with another (neonatal stress).
Many deficits reported after pharmacological REM sleep deprivation are more
easily explained by persistent alterations in monoaminergic function. For example,
the changes in adult sleep architecture ascribed to pharmacological RSD are only
observed following neonatal treatments with agents that alter serotonergic neurotrans-
mission (e.g., serotonin uptake inhibitors). Other REM-sleep inhibiting compounds
delivered neonatally have no effect on subsequent adult sleep patterns.25 Likewise
the changes in anxiety and sexual behavior reported after neonatal REM sleep
deprivation are more likely due to changes in serotonergic neurotransmission than
pharmacological RSD; for example, compounds that reduce serotonin and REM sleep
in neonatal rats decrease anxiety and increase sexual behavior in adulthood, effects
that are precisely opposite to those reported after neonatal clomipramine adminis-
tration.2,19,82 Because both compounds decrease REM sleep but have opposite effects
on serotonin levels, it is unlikely that RSD contributes in a significant way to the
behavioral changes noted in adult rats following neonatal antidepressant treatment.
SLEEP AND VISUAL SYSTEM DEVELOPMENT

Sleep and the Developing Lateral Geniculate Nucleus (LGN)
More persuasive evidence for a role for sleep in developmental plasticity comes
from experiments that combine sleep manipulation with assays of visual system
development. In the developing visual system, endogenous activity in retinal and
thalamocortical circuits helps establish initial patterns of synaptic circuitry that are

elaborated and sculpted by experience during subsequent critical periods of postnatal

development.71,76 The initial development of the central visual pathways and their
subsequent sculpting by experience occur at ages when sleep amounts are very high
or during dramatic changes in sleep expression.14 The following sections consider
the evidence that sleep contributes to both of these processes in the development of
two important components of the central visual pathway, the lateral geniculate
nucleus (LGN) and primary visual cortex (V1).
A role for REM sleep in brain maturation has been examined by studying the
effects of RSD (using mechanical techniques), or the elimination of REM sleep PGO
waves (RSPD), on subsequent visual system development. Davenne and Adrien
examined changes in neuronal morphology in the LGN in kittens after lesioning
PGO generating centers in the brainstem.12 Bilateral electrolytic lesions in the rostral
pontine tegmentum abolished PGO waves in the neonatal cat, resulting in smaller
LGN volumes and reduced LGN soma sizes. These findings were extended in a
second study, which showed that PGO wave elimination in kittens produced much
slower LGN responses to stimulation of the optic chiasm (compared to sham or
unilaterally lesioned control cats) and more LGN cells with mixed On-Off responses
(as opposed to pure On or Off responses to an annulus of light centered in the
receptive field) and fewer X cell responses (relative to On-Off responses).13 These
morphological and functional changes in LGN cells are consistent with delayed
development in the LGN11,81 and suggest that REM sleep neuronal activity may be
necessary for normal LGN development.
Sleep may also be important for the later occurring, critical periods of visual
development. The critical period has been traditionally investigated by surgically
closing an eye (monocular deprivation [MD]), which rapidly alters responsiveness
and morphology in both subcortical and cortical neurons.67,73 Pompeiano et al (1995)
reported that total sleep deprivation combined with MD increased the effects of MD
on LGN cell morphology.57 Unfortunately, this study is difficult to interpret because
the amount of visual experience was not controlled and very little quantitative data
on sleep architecture were presented. Stronger evidence for a role for sleep in
subcortical development is reported when different forms of selective RSD or RSDP
are combined with MD. Oksenberg et al. found that 1 week of RSD in kittens (using
the pedestal technique) augmented the effects of MD on cell morphology in the
binocular segment of the LGN.56 LGN cells innervated by the occluded eye were
smaller in kittens deprived of REM sleep and vision in one eye, resulting in a greater
difference in the size of LGN cells activated by the open and deprived eyes.
A comparable increase in LGN cell size disparity was found when MD was
combined with brainstem lesions that remove PGO waves. In this case LGN cells
receiving input from the open eye appeared to increase in size.69 An additional finding
is that RSD combined with MD also reduces cell sizes in the monocular segment
of the LGN, which does not depend upon competitive interactions between the two
eyes.68 Work from this laboratory has also shown that RSD for one week reduces
immunoreactivity for the calcium binding protein parvalbumin in GABAergic inter-
neurons in the developing LGN.35 These latter findings are particularly interesting
because parvalbumin may play a role in certain forms of synaptic plasticity.7 In sum,

these results suggest that RSD or RSDP may influence LGN maturation during
critical periods of visual system development.
Sleep and Developmentally Regulated Cortical Plasticity
Sleep may also play an important role in developmentally regulated forms of cortical
plasticity. REM sleep, for example, appears to influence a form of long-term poten-
tiation (LTP) elicited during the critical period for visual system development.70 In
this type of LTP, high-frequency white-matter stimulation in cortical slices prepared
P28–P30 rats produces synaptic potentiation in cortical layers II and III. This form
of LTP decreases with age (P35+) and is not observed in cortical slices from adult
rats.40 Using a less stressful version of the pedestal technique of RSD (multiple
small-platforms), Shaffery et al. measured the effects of 1 week of RSD on this form
of LTP in rat visual cortex.70
The authors reported that 1 week of RSD prolonged the critical period for the
developmentally regulated form of LTP; LTP was evoked from slices of visual cortex
from RSD rats at ages when this type of LTP is not normally observed (P34–P40).
A similar extension of the critical period was not seen in cortical slices from control
rats that were left in their nests, or from rats placed on larger platforms (large-
platform control) that presumably permitted REM sleep. Conversely, RSD had no
effect on a nondevelopmentally regulated form of LTP evoked by layer IV stimula-
tion. The extension of the critical period by RSD was similar to effects produced
by dark rearing, which also prolongs the period of induction of this form of LTP.70
These findings suggest a maturational delay in visual cortex, and are in general
agreement with previous findings from the same group suggesting that RSD impairs
normal brain maturation.
Evidence that sleep contributes to developmental cortical plasticity has also been
demonstrated in vivo. We investigated the role of sleep in cortical plasticity by
combining MD with periods of sleep or sleep deprivation.27 Kittens at the height of
the critical period were divided into four experimental groups, all of which had one
eye closed and were kept awake in a lighted environment for 6 hours. This MD
period provided a common stimulus for the synaptic remodeling in all groups. The
four groups differed in their experience thereafter. Cats in the baseline group (MD6)
were quickly anesthetized for physiological measurement of ocular dominance in
primary visual cortex using optical imaging of intrinsic cortical signals and extra-
cellular unit recording. Cats in a second group (MDS) were allowed to sleep for an
additional 6 hours in complete darkness before making optical and unit recordings.
The third group of kittens (MDSD) were treated identically to those in the MDS
group except that they were kept awake during the 6 hours in complete darkness
before the recordings. The fourth group (MD12) was also kept awake for 6 additional
hours but remained in a lighted environment, effectively giving them 6 additional
hours of monocular deprivation before the recordings.
These experiments determined whether:
• The effects of MD were enhanced by subsequent sleep (MD6 compared

to MDS).

• The enhancement of plasticity observed in group MDS required sleep or

simply a period of time following the inducing stimulus (MDS compared
to MDSD).
• The mechanical sleep deprivation indirectly impeded ocular dominance
plasticity (MD12 compared to MDSD).
Optical imaging of intrinsic cortical signals and extracellular unit recording

showed that sleep nearly doubled the effects of MD on visual cortical responses,
and wakefulness in complete darkness tended to erase the effects of the preceding
monocular visual experience. No brain state other than sleep is known to have such
augmenting effects on ocular dominance plasticity because anesthetic states and
cortical inactivation suppress ocular dominance plasticity.29,37,62,63 The enhancement
of plasticity by sleep was similar to that produced by an equal amount of additional
MD. Although the precise contribution of REM and non-REM sleep to this process
is still unknown, we did find that the enhancement of cortical plasticity was highly
correlated with non-REM sleep time, suggesting an important role for non-REM
sleep in the rapid cortical synaptic remodeling elicited by MD.27
Further Considerations
The findings discussed above support a role for sleep in visual system development,
but a number of caveats should be kept in mind. One must consider potential side
effects of the experimental manipulation used in each study; for example, sleep
deprivation indirectly influences behavior and neurochemistry in ways that may
influence the results of an experiment irrespective of sleep changes.5,72 Moreover
manipulations performed in one state may influence neural processes in other vig-
ilance states as well, making it difficult to determine which vigilance state is respon-
sible for the observed effects.
In all of the experiments reviewed above, experimental changes in sleep struc-
ture, or lesions that damage parts of the brain active in sleep, were used to test the
role of sleep in visual development. Many of these manipulations are likely to have
complicated effects on neural development and behavior in addition to their effects
on sleep. In studies using brainstem lesions, it is not clear if the reported deficits
are due to PGO reduction, or the elimination of ascending innervation to target LGN
neurons from cholinergic and monoaminergic brainstem projections.16,75 These affer-
ents not only provide tonic excitatory input to the LGN, they may also promote
neural growth and maturation.43,44 It is therefore possible that bilateral removal of
this input, rather than the elimination of REM sleep PGO waves, may partially
account for the results reported in the Davenne studies (though this an unlikely factor
in the Shaffery study since cell sizes increased following PGO elimination).
The role of stress should also be considered in studies using sleep deprivation.
RSD using the pedestal technique is stressful because the animal periodically con-
tacts water and in some cases is unable to properly groom itself. Repeated stress
can increase neuronal degeneration1,66,84 and modifies synaptic plasticity in complex
ways.15 Thus although the enhancement of the anatomical effects of MD in the LGN
by RSD is consistent with a maturational delay induced by RSD, it may also reflect

the influence of stress hormones on degenerative processes triggered by sensory

deprivation. This seems an especially important consideration, given that the ana-
tomical effects of RSD were reported in the monocular segment of the LGN, an
area not especially influenced by developmental critical periods.
Stress, however, likely plays a minor role in studies using sleep deprivation
combined with in situ LTP, immunohistochemical assays of calcium binding proteins
(parvalbumin), or cortical plasticity in vivo. In the Shaffery et al. investigation, the
chronic stress hormone release induced by long-term RSD should impair, not extend,
susceptibility to LTP.15 The RSD induced down-regulation of parvalbumin in the
LGN is also unlikely due to stress. Increases in stress hormones have no effect on
parvalbumin concentrations in the hippocampus-a brain region sensitive to circulat-
ing glucocorticoid levels.42,66 Moreover, the acute release of stress hormones elicited
by very short periods of TSD tends to enhance, not impair, synaptic plasticity15 and
thus is not a factor in the results reported in Frank et al. (2001).27
A second issue that complicates the interpretation of some studies occurs when
the experimental manipulation alters vigilance states in ways likely to influence
neuronal activity during wake as well as subsequent processing during REM and
non-REM sleep; for example, RSD increases noradrenergic activity in the CNS,38,58
which enhances signal detection in sensory neurons during wake.3,74 Even when total
non-REM sleep amounts are preserved, RSD frequently alters non-REM sleep archi-
tecture (fragmentation, loss of deeper stages of non-REM sleep).4,18 Thus in exper-
imental designs that employ prolonged RSD combined with periods of sensory input,
the observed changes in neuronal morphology and plasticity may result from many
factors including RSD, changes in wake neural processing, and subtle changes in
non-REM sleep. Determining the cause of developmental effects is also difficult in
studies using brainstem lesions, because they interrupt ascending pathways that
profoundly influence neuronal processing in sleep and wake.
THEORIES OF SLEEP FUNCTION IN

DEVELOPING ANIMALS
THE ONTOGENETIC HYPOTHESIS

In their classic study in human infants, Roffwarg and colleagues proposed that the
large amounts of REM sleep in early infancy provide an important source of endog-
enous neural activity necessary for brain maturation.64 In the more recent formula-
tions of the Ontogenetic Hypothesis, it is suggested that REM sleep not only pro-
motes normal brain development but also insulates the brain from excessive
experience-dependent plasticity.56,65 Both functions are thought to be mediated by
the PGO waves, or heightened release of acetylcholine during REM sleep.
The Ontogenetic Hypothesis and its variants,47,49 is intuitively appealing since
REM sleep amounts are unusually high in infants, and decrease as the brain develops.
It is supported by findings that indicate that RSD or RSDP can modify morphological
and electrophysiological development of the LGN, and developmentally regulated
cortical synaptic plasticity in situ (see previous discussion). The theory that REM

sleep offsets waking experience in infants has less direct support, but is suggested
by three findings:
• REM sleep PGO waves in adult cats activate all LGN lamina simulta-
neously, indicating that this activity, in contrast to visual experience, is not
eye-specific.46 Theoretically such nonspecific activation of neural circuits,
if present in developing animals, could counter-balance the more specific,
experience-dependent activation of neural circuitry present in wake.
• In contrast to normal adult mammals, latencies to REM sleep in infants
are very short, and sleep-onset REMs (SOREMs) frequently occur.39,48
• In our study cortical plasticity was negatively correlated with REM sleep
amounts, suggesting that REM sleep inhibits experience-dependent
plasticity.27
Thus it is possible that the short latencies to REM sleep and SOREMs in neonates
interfere with the consolidation of experience-dependent changes in neural circuitry.
Despite the appeal of the Ontogenetic Hypothesis, several issues remain to be
resolved. As discussed previously, potential side effects of the procedures used in
RSD and RSDP stress complicate the experimental support for the Ontogenetic
Hypothesis. A second issue is that sleep in newborns may not be identical or
homologous to adult REM sleep,26 and even when unambiguous periods of REM
sleep are observed, the neurophysiological phenomena typical of adult REM sleep
(e.g., PGO waves) are not always present; for example, REM sleep PGO waves in
the kitten are not reported at ages when REM sleep is maximally expressed.6 It is
also unknown if other aspects of REM sleep, such as heightened cholinergic activity,
are present in newborn animals. Considering the slow maturation of cholinergic
systems,9,45,55 and the late appearance of other REM sleep phenomena,85 this seems
rather unlikely. Indeed, the majority of studies suggesting a developmental role for
REM sleep have been performed at ages when REM sleep has already declined to
near adult levels.35,56,68–70 A fourth and final point is that the Ontogenetic Hypothesis
does not consider the potential role of non-REM sleep in brain development. This
may be a historical oversight, considering the emphasis REM sleep has received in
the past, but there are now several findings linking non-REM sleep to synaptic
plasticity and neuronal development.5
In summary, although there are data to support predictions of the Ontogenetic
Hypothesis, they are limited to a narrow, developmental period and are restricted to
REM sleep. In addition their interpretation must await further experiments that
determine the homology between infant and adult sleep states, and more carefully
control for indirect effects induced by RSD and RSDP.
SLEEP AND THE CONSOLIDATION OF EXPERIENCE

The correlation between non-REM sleep amounts and the effects of MD on cortical
plasticity suggests that non-REM sleep may also play an important role in brain
development.27 The augmentation of experience-dependent synaptic plasticity is
consistent with findings in adult animals, where non-REM sleep has been linked to

learning and memory consolidation, and neuronal events that contribute to synaptic
remodeling (reviewed in Reference 27). A role for non-REM sleep in developmental
cortical plasticity is further suggested by ontogenetic changes in non-REM sleep
that coincide with periods of heightened cortical plasticity. In the cat there is a rapid
decline in REM sleep and a corresponding increase in non-REM sleep amounts
near the beginning of the critical period for visual development.39 In rats the begin-
ning of the critical period for visual development coincides with the development
of non-REM sleep homeostasis. Sleep deprivation does not increase non-REM sleep
EEG activity until the fourth postnatal week, indicating that the regulatory relation-
ship between wake and non-REM sleep develops in parallel with periods of height-
ened cortical plasticity.28 These findings suggest that non-REM sleep may consol-
idate waking experience; a process that begins during critical periods of brain
development when the animal is most sensitive to waking experience, but is retained
throughout life.
While it appears that one function of non-REM sleep may be to consolidate
waking experience, it is likely that this sleep stage has other functions in the
developing brain as well. In altricial species such as the rat and cat, the most rapid
increase in non-REM sleep occurs several weeks before the critical period for visual
development.24,39 In precocial species such as the sheep, non-REM sleep amounts
are near adult levels in utero, a time when exogenous visual experience is minimal.77
Given that the appearance of non-REM sleep coincides with periods of extensive
neocortical development in all mammalian species, and that it is homeostatically
regulated within 24 to 48 hours of its electrographic appearance,14 it is possible that
non-REM sleep promotes the formation of rudimentary circuitry that is subsequently
shaped by experience.
SUMMARY
The abundance of sleep during periods of rapid brain maturation and synaptic
plasticity suggest a role for sleep in brain development. The strongest experimental
support for this view has come from studies in the visual system, where it has been
shown that sleep and sleep loss modify developmental processes in the LGN and in
primary visual cortex. In particular RSD and RSDP trigger several morphological
and electrophysiological changes in the LGN, and modify cortical plasticity in situ.
Non-REM sleep appears to be necessary for the consolidation of visual experience
during critical periods of experience-dependent cortical plasticity in vivo. These
findings indicate that both sleep states could be important for neuronal development
and plasticity, although the contribution of each state might be quite different.
Although the precise role of each sleep state in brain development is still
unknown, current findings suggest that the relative amounts of REM and non-REM
sleep during infancy may critically influence brain maturation. REM sleep is max-
imally expressed when endogenous neuronal activity is critical for the establishment
of rudimentary neural circuitry in the visual system. Although non-REM sleep might
also be important for this latter process, it seems to be more strongly associated
with synaptic changes elicited by experience, because it rapidly matures after eye

opening24,30,39 and becomes homeostatically regulated by wake in a manner similar

to adult non-REM sleep during critical periods of experience-dependent synaptic
plasticity.28 Thus it seems plausible that although both REM and non-REM sleep
promote the formation of rudimentary circuits, non-REM sleep additionally consol-
idates changes in neural circuitry elicited by waking experience.
In conclusion, many phenomena strongly suggest a role for sleep in brain
development and plasticity, but definitive evidence is still lacking. Basic questions
about sleep in developing animals are still unanswered, and we currently know little
about the cellular and molecular mechanisms by which sleep might modify synaptic
plasticity. The opportunities afforded by advances in genetic and pharmacological
manipulation of specific molecules and signaling pathways, and by chronic recording
and manipulation of neural activity, raise the hope that these questions may soon be
answered.
REFERENCES
1. Abraham I., Harkany T., Horvath K.M., Veenema A.H., Penke B., Nyakas C., and
Luiten P.G.M., Chronic corticosterone administration dose-dependently modulates
A(142) and NMDA-Induced Neurodegeneration in Rat Magnocellular Nucleus Basa-
lis, J. Endocrinology, 12, 486, 2000.
2. Adlard B.P.F. and Smart J.L., Some aspects of the behavior of young and adult rats
treated with p-chlorophenylalanine in infancy, Dev. Psychobiol., 7, 135, 1974.
3. Aston-Jones G., Rajkowski J., and Cohen J., Role of locus coeruleus in attention and
behavioral flexibility, Biological Psychiatry, 46, 1309, 1999.
4. Beersma D.G.M., Dijk D.J., Blok C.G., and Everhardus I., REMS deprivation during
5 hours leads to an immediate REMS rebound and to suppression of non-REM
intensity, Elec. Clin. Neurophysiol., 76, 114, 1990.
5. Benington J.H. and Frank M.G., Cellular and molecular connections between sleep
and synaptic plasticity, Progress in Neurobiology, 69, 77, 2003.
6. Bowe-Anders C., Adrien J., and Roffwarg H.P., Ontogenesis of ponto-geniculo-
occipital activity in the lateral geniculate nucleus of the kitten, Exp. Neurol., 43, 242,
1974.
7. Caillard O., Moreno H., Schwaller B., LLano I., Celio M.R., and Marty A., Role of
the calcium-binding protein parvalbumin in short-term synaptic plasticity, PNAS, 97,
13372, 2000.
8. Corner M.A., Mirmiran M., Bour H.L., Boer G.J., van de Poll N.E., van Oyen H.G.,
and Uylings H.B., Does rapid-eye-movement sleep play a role in brain development?,
Prog. Brain Res., 53, 347, 1980.
9. Coyle J.T. and Yamamura H.I., Neurochemical aspects of the ontogenesis of cholin-
ergic neurons in the rat brain, Brain Res., 118, 429, 1976.
10. Czikk M.J., Sweeley J.C., Homan J.H., Milley J.R., and Richardson B.S., Cerebral
leucine uptake and protein synthesis in the near-term ovine fetus: relation to fetal
behavioral state, Am. J. Physiol. Regul. Integr. Comp. Physiol., 284, R200, 2003.
11. Daniels J.D., Pettigrew J.D., and Norman J.L., Development of single-neuron
responses in kitten’s lateral geniculate nucleus, J. Neurophysiol., 41, 1373, 1978.
12. Davenne D. and Adrien J., Suppression of PGO waves in the kitten: anatomical effects
on the lateral geniculate nucleus, Neuroscience Lett., 45, 33, 1984.

13. Davenne D., Fregnac Y., Imbert M., and Adrien J., Lesion of the PGO pathways in
the kitten, II. Impairment of physiological and morphological maturation of the lateral
geniculate nucleus, Brain Res., 485, 267, 1989.
14. Davis F.C., Frank M.G., and Heller H.C., Ontogeny of sleep and circadian rhythms,
in Regulation of Sleep and Circadian Rhythms, Zee P.C. and Turek F.W., Eds., Marcel
Dekker, New York, 1999, p. 19.
15. De Kloet E.R., Oitzl M.S., and Joels M., Stress and cognition: are corticosteroids
good or bad guys?, TINS, 22, 422, 1999.
16. Derrington A., The lateral geniculate nucleus, Current Biol., 11, R635, 2001.
17. Dwyer S.M. and Rosenwasser A.M., Neonatal clomipramine treatment, alcohol intake
and circadian rhythms in rats, Psychopharmacology, 138, 176, 1998.
18. Endo T., Schwierin B., Borbely A.A., and Tobler I., Selective and total sleep depri-
vation: effect on the sleep EEG in the rat, Psychiatry Res., 66, 97, 1997.
19. Farabollini F., Hole D.R., and Wilson C.A., Behavioral effects in adulthood of sero-
tonin depletion by p-chorophenylalanine given neonatally to male rats, Int. J. Neu-
rosci., 41, 187, 1988.
20. Feng P., Postnatal REM sleep deprivation and depression: New findings and hypoth-
esis, Actas de Fisiologica, 7, 141, 2001.
21. Feng P., Ma Y., and Vogel G.W., Ontogeny of REM rebound in postnatal rats, Sleep,
24, 645, 2001.
22. File S.E. and Tucker J.C., Neonatal clomipramine treatment in the rat does not affect
social, sexual and exploratory behaviors in adulthood, Neurobehav. Toxicol. Teratol.,
5, 3, 1983.
23. Frank M.G. and Heller H.C., Development of diurnal organization of EEG slow-wave
activity and slow-wave sleep in the rat, Am. J. Physiol., 273, R472, 1997.
24. Frank M.G. and Heller H.C., Development of REM and slow-wave sleep in the rat,
Am. J. Physiol., 272, R1792, 1997.
25. Frank M.G. and Heller H.C., Neonatal treatments with the serotonin uptake inhibitors
clomipramine and zimelidine, but not the noradrenaline uptake inhibitor desipramine,
disrupt sleep patterns in adult rats, Brain Res., 768, 287, 1997.
26. Frank M.G. and Heller H.C., The ontogeny of mammalian sleep: a reappraisal of
alternative hypotheses, J. Sleep Res., 12, 25, 2003.
27. Frank M.G., Issa N.P., and Stryker M.P., Sleep enhances plasticity in the developing
visual cortex, Neuron, 30, 275, 2001.
28. Frank M.G., Morrissette R., and Heller H.C., Effects of sleep deprivation in neonatal
rats, Am. J. Physiol., 275, R148, 1998.
29. Freeman R.D. Effects of brief uniocular patching on kitten visual cortex, Trans.
Opthal. Soc. U.K., 99, 382, 1979.
30. Gramsbergen A., The development of the EEG in the rat, Developmental Psychobi-
ology, 9, 501, 1976.
31. Henderson M.G., McConnaughey M.M., and McMillen B.A., Long-term conse-
quences of prenatal exposure to cocaine or related drugs: Effects on rat brain monoam-
inergic receptors, Brain Res. Bull., 26, 941, 1991.
32. Hilakivi L. and Sinclair J.D., Effect of neonatal clomipramine treatment on adult
alcohol drinking in the AA and ANA rat lines, Pharmacol. Biochem. Behav., 24,
1451, 1986.
33. Hilakivi L.A. and Hilakivi I., Increased adult behavioral “despair” in rats neonatally
exposed to desipramine or zimeldine: an animal model of depression?, Pharmacol.
Biochem. Behav., 28, 367, 1987.

34. Hilakivi L.A., Hilakivi I., Ahtee L., Haikala H., and Attila M., Effect of neonatal
nomifensine exposure on adult behavior and brain monoamines in rats, J. Neural.
Transm., 70, 99, 1987.
35. Hogan D., Roffwarg H.P., and Shaffery J.P., The effects of 1 week of REM sleep
deprivation on parvalbumin and calbindin immunoreactive neurons in central visual
pathways of kittens, J. Sleep Res., 10(4), 285–296, 2001.
36. Hoppenbrouwers T. and Sterman M.B., Development of sleep-state patterns in the
kitten, Experimental Neurology, 49, 822, 1975.
37. Imamura K. and Kasamatsu T., Ocular dominance plasticity restored by NA infusion
to aplastic visual cortex of anesthetized and paralyzed kittens, Exp. Brain Res., 87,
309, 1991.
38. Irwin M., Thompson J., Miller C., Gillin J.C., and Ziegler M., Effects of sleep and
sleep deprivation on catecholamine and interleukin-2 levels in humans: clinical impli-
cations, J. Clinical Endocrinology and Metabolism, 84, 1979, 1999.
39. Jouvet-Mounier D., Astic L., and Lacote D., Ontogenesis of the states of sleep in rat,
cat and guinea pig during the first postnatal month, Developmental Psychobiology,
2, 216, 1970.
40. Kirkwood A., Lee H.K., and Bear M.F., Co-regulation of long-term potentiation and
experience-dependent synaptic plasticity in visual cortex, Nature, 375, 328, 1995.
41. Klemfuss H. and Gillin C.J., Neonatal scopolamine or antidepressant treatment: effect
on development of hamster circadian rhythms, Pharmacol. Biochem. Behav., 59, 369,
1997.
42. Krugers H.J., Koolhaas J.M., Medema R.M., and Korf J., Prolonged subordination
stress increases Calbindin-D28k immunoreactivity in the rat hippocampal CA1 area,
Brain Res., 729, 289, 1996.
43. Lauder J.M., Hormonal and humoral influence on brain development, Psychoneuroen-
docrinology, 8, 121, 1983.
44. Lauder J.M. and Schambra U.B., Morphogenetic roles of acetylcholine, Environ.
Health Pespect., 107, 65, 1999.
45. Lee W., Nicklaus K.J., Manning D.R., and Wolfe B.B., Ontogeny of cortical musca-
rinic receptor subtypes and muscarinic receptor-mediated responses in rat, J. Phar-
macol. Exp. Ther., 252, 284, 1990.
46. Marks G.A., Roffwarg H.P., and Shaffery J.P., Neuronal activity in the lateral genic-
ulate nucleus associated with ponto-geniculate-occipital waves lacks lamina specific-
ity, Brain Res., 815, 21, 1999.
47. Marks G.A., Shaffery J.P., Oksenberg A., Speciale S.G., and Roffwarg H.P., A func-
tional role for REM sleep in brain maturation, Behavioral Brain Res., 69, 1, 1995.
48. McGinty R.J., Stevenson M., Hoppenbrouwers T., Harper R.M., Sterman M.B., and
Hodgman J., Polygraphic studies of kitten development: sleep state patterns, Devel-
opmental Psychobiology, 10, 455, 1977.
49. Mirmiran M. and Maas Y.G.H., The function of fetal/neonatal REM sleep, in Rapid
Eye Movement Sleep, Mallick B.N. and Inoue S., Eds., Narosa Publishing House,
New Delhi, 1999, p. 326.
50. Mirmiran M., Scholtens J., van de Poll N.E., Uylings H.B., van der Gugten J., and
Boer G.J., Effects of experimental suppression of active (REM) sleep during early
development upon adult brain and behavior in the rat, Brain Res., 283, 277, 1983.
51. Mirmiran M., Scholtens J., van de Poll N.E., Uylings H.B., van der Gugten J., and
Boer G.J., Effects of experimental suppression of active (REM) sleep during early
development upon adult brain and behavior in the rat, Brain Res., 283, 277, 1983.

52. Mirmiran M., van de Poll N.E., Corner M.A., van Oyen H.G., and Bour H.L.,
Suppression of active sleep by chronic treatment with chlorimipramine during early
postnatal development: effects upon adult sleep and behavior in the rat, Brain Res.,
204, 129, 1981.
53. Mirmiran M., van den Dungen, H., and Uylings H.B., Sleep patterns during rearing
under different environmental conditions in juvenile rats, Brain Res., 233, 287, 1982.
54. Nakanishi H., Sun Y., Nakamura R.K., Mori K., Ito M., Suda S., Namba H., Storch
F.I., Dang T.P., and Mendelson W., Positive correlations between cerebral protein
synthesis rates and deep sleep in Macaca mulatta, Eur. J. Neurosci., 9, 271, 1997.
55. Ninomiya Y., Koyama Y., and Kayama Y., Postnatal development of choline acetyl-
transferase activity in the rat laterodorsal tegmental nucleus, Neuroscience Lett., 308,
138, 2001.
56. Oksenberg A., Shaffery J.P., Marks G.A., Speciale S.G., Mihailoff G., and Roffwarg
H.P., Rapid eye movement sleep deprivation in kittens amplifies LGN cell-size dis-
parity induced by monocular deprivation, Developmental Brain Res., 97, 51, 1996.
57. Pompeiano O., Pompeiano M., and Corvaja N., Effects of sleep deprivation on the
postnatal development of visual-deprived cells in the cat’s lateral geniculate nucleus,
Archives Italiennes de Biologie, 134, 121, 1995.
58. Porrka-Heiskanen T., Smith S.E., Taira T., Urban J.H., Levine J.E., Turek F.W., and
Stenberg D., Noradrenergic activity in rat brain during rapid eye movement sleep
deprivation and rebound sleep, Am. J. Physiol. (Regulatory Integrative Comp. Phys-
iol.), 268, R1456, 1995.
59. Prathiba J., Kumar K.B., and Karanth K.S., Effects of REM sleep deprivation on
cholinergic receptor sensitivity and passive avoidance behavior in clomipramine
model of depression, Brain Research, 867, 243, 2000.
60. Prathiba J., Kumar K.B., and Karanth K.S., Hyperactivity of hypothalamic pituitary
axis in neonatal clomipramine model of depression, J. Neural Trans., 105, 1335, 1998.
61. Ramm P. and Smith C.T., Rates of cerebral protein synthesis are linked to slow-wave
sleep in the rat, Physiol. Behav., 48, 749, 1990.
62. Rauschecker J.P. and Hahn S., Ketamine-Xylazine anesthesia blocks consolidation
of ocular dominance changes in kitten visual cortex, Nature, 326, 183, 1987.
63. Reiter H.O., Waitzman D.M., and Stryker M.P., Cortical activity blockade prevents
ocular dominance plasticity in the kitten visual cortex, Exp. Brain Res., 65, 182, 1986.
64. Roffwarg H.P., Muzio J.N., and Dement W.C., Ontogenetic development of the human
sleep-dream cycle, Science, 604, 1966.
65. Roffwarg H.P. and Shaffery J.P., The ontogenetic hypothesis of REM sleep function:
Its history, current status and prospects for confirmation, Sleep Research Online, 2,
714, 1999.
66. Sapolsky R.M. Stress, glucocorticoids, and damage to the nervous system: The current
state of confusion, Stress, 1, 1, 1996.
67. Sengpiel F., Godecke I., Stawinski P., Hubener M., Lowel S., and Bonhoffer T.,
Intrinsic and environmental factors in the development of functional maps in cat
visual cortex, Neuropharmacology, 37, 607, 1998.
68. Shaffery J.P., Oksenberg A., Marks G.A., Speciale S.G., Mihailoff G., and Roffwarg
H.P., REM sleep deprivation in monocularly occluded kittens reduces the size of cells
in LGN monocular segment, Sleep, 21, 837, 1998.
69. Shaffery J.P., Roffwarg H.P., Speciale S.G., and Marks G.A., Ponto-geniculo-occipital
wave suppression amplifies lateral geniculate nuclues cell-size changes in monocu-
larly deprived kittens, Developmental Brain Res., 114, 109, 1999.

70. Shaffery J.P., Sinton C.M., Bissette G., Roffwarg H.P., and Marks G.A., Rapid eye
movement sleep deprivation modifies expression of long-term potentiation in visual
cortex of immature rats, Neuroscience, 110, 431, 2002.
71. Shatz C.J., Emergence of order in visual system development, PNAS, 93, 602, 1996.
72. Siegel J.M., The REM Sleep-Memory Consolidation Hypothesis, Science, 294, 1058,
2001.
73. Singer W., Neuronal mechanisms in experience dependent modification of visual
cortex function, in Development and Chemical Sensitivity of Neurons, Cuenod M.,
Kreutzberg G.W., and Bloom F.E., Eds., Elsevier/North-Holland Biomedical Press,
Amsterdam, 1979, p. 457.
74. Smiley J.F., Monoamines and acetylcholine in primate cerebral cortex: what anatomy
tells us about function, Rev. Bras. Biol., 56, 153, 1996.
75. Steriade M., Arousal: revisiting the reticular activating system, Science, 272, 225,
1996.
76. Sur M. and Leamey C.A., Development and plasticity of cortical areas and networks,
Nat. Rev. Neurosci., 2, 251, 2001.
77. Szeto H. and Hinman D.J., Prenatal development of sleep-wake patterns in sheep,
Sleep, 347, 1985.
78. Van Someren E.J., Mirmiran M., Bos N.P., Lamur A., Kumar A., and Molenaar P.C.,
Quantitative analysis of eye movements during REM-sleep in developing rats, Dev.
Psychobiol., 23, 55, 1990.
79. Vogel G., Neill D., Hagler M., and Kors D., A new animal model of endogenous
depression: a summary of present findings, Neurosci. Biobehav. Rev., 14, 85, 1990.
80. Vogel G., Neill D., Kors D., and Hagler M., REM sleep abnormalities in a new animal
model of endogenous depression, Neurosci. Biobehav. Rev., 14, 77, 1990.
81. Williams A.L. and Jeffery G., Growth dynamics of the developing lateral geniculate
nucleus, J. Comparative Neurology, 430, 332, 2001.
82. Wilson C.A., Pearson J.R., Hunter A.J., Tuohy P.A., and Payne A.P., The effect of
neonatal manipulation of hypothalamic serotonin levels on sexual activity in the adult
rat, Pharmacol. Biochem. Behav., 24, 1175, 1986.
83. Yannielli P.C., Cutrera R.A., Cardinali D.P., and Golombek D.A., Neonatal clomi-
pramine treatment of Syrian hamsters: effect on the circadian system, Eur. J. Phar-
macol., 349, 143, 1998.
84. Yusim A., Ajilore O., Bliss T., and Sapolsky R.U., Glucocorticoids exacerbate insult-
induced declines in metabolism in selectively vulnerable hippocampal cell fields,
Brain Res., 870, 109, 2000.
85. Chase M.H., Brain stem somatic reflex activity in neonatal kittens during sleep and
wakefulness, Physiol. Behav., 7, 165–172, 1971.

11 Changes in Brain Gene

Expression between
Sleep and Wakefulness
Giulio Tononi and Chiara Cirelli
CONTENTS
Introduction
References
INTRODUCTION
In order to determine the molecular changes that occur in the brain during the sleep-
waking cycle and after sleep deprivation, we have performed a systematic screening
of brain gene expression in sleeping, spontaneously awake, and sleep-deprived rats.
The data summarized here refer to the completed analysis of >20,000 transcripts
expressed in the cerebral cortex. The expression of the majority (~95%) of these
genes does not change between sleep and wakefulness or after sleep deprivation,
even when forced wakefulness is prolonged for several days. A few hours of wake-
fulness, either spontaneous or due to sleep deprivation, increase the expression of
several transcripts involved in energy metabolism, excitatory neurotransmission,
transcriptional activation, memory acquisition, and cellular stress. The ~100 genes
whose expression increases during sleep, on the other hand, provide molecular
support for the proposed involvement of sleep in protein synthesis and neural plas-
ticity, and point to a novel role for sleep in membrane trafficking and maintenance.
The pattern of changes in gene expression after long periods of sleep deprivation
is unique and does not resemble that of short-term sleep deprivation or spontaneous
wakefulness. A notable exception is represented, however, by the enzyme arylsul-
fotransferase, whose induction appears to be related to the duration of previous
wakefulness. In rodents this enzyme plays a major role in the catabolism of cate-
cholamines, suggesting that an important role for sleep may be that of interrupting
the continuous activity during wakefulness of brain catecholaminergic systems.
The issue whether changes in gene expression occur in relation to sleep, wake-
fulness, and sleep deprivation is an old one. Early experiments did not focus on
specific genes but examined overall changes in RNA content1 or synthesis2,3 as well
0-8493-1519-0/05/$0.00+$1.50

as global changes in protein synthesis.4–7 Giuditta et al.3 injected [3H]-orotate intra-

ventricularly and measured its incorporation into newly synthesized RNA during
the following hour. In a fraction of neuronal perikarya in the cerebral cortex, the
relative content of radioactive RNA was increased in sleep with respect to waking
in the nuclear but not in the cytoplasmic compartment.
Panov1 found variations in protein and RNA content in individual neurons and
glial cells of some brainstem nuclei after 1–4 days of total or selective REM sleep
deprivation. Bobillier et al.4 reported a generalized decrease of [3H]-amino acid
incorporation into the proteins of telencephalon and brainstem after 3 h of total sleep
deprivation in rats. Conversely a striking increase of labeled proteins was found in
rats that were allowed to sleep for 1.5 h after 1.5 h of total sleep deprivation. Ramm
and Smith6 found that in the rat higher rates of cerebral protein synthesis were
associated with a higher slow-wave sleep score. In rhesus monkeys Nakanishi et al.7
found that protein synthesis, as measured by L-[1-14C]-leucine incorporation, was
positively correlated with deep sleep in most brain regions.
Later studies focused instead on specific genes, the so-called immediate early
genes (IEGs), such as c-fos, NGFI-A, c-jun and junB. IEGs are among the first genes
to be turned on or off in the cascade of molecular events that leads to changes in
the expression of other genes. Their protein products have specific DNA binding
domains by which they act as nuclear transcription factors.8 It was found9 that the
expression of c-fos, NGFI-A, and other IEGs is powerfully modulated by behavioral
state, their expression being low or absent in most brain regions if the animals had
spent most of the previous 3–8 h asleep, and high if the animals had been either
spontaneously awake or sleep deprived for a few hours before sacrifice.10–12 Most
IEGs code for transcription factors, and therefore can regulate the expression of
many other genes. Thus an important implication of these experiments was that the
strong state-dependent modulation of IEGs expression could signal widespread
transcriptional changes across behavioral states.
Over the last several years our laboratory has employed mRNA differential
display (mRNA DD), nylon membrane arrays, and, more recently, GeneChip tech-
nology to systematically establish the differences in gene expression that occur
across behavioral states.13–16 The main goal was to determine whether there are genes
whose expression is specifically increased during sleep and the identity of these
genes. Brain gene expression was compared between rats that had been asleep for
the first 3 or 8 h of the light period, in rats that had been spontaneously awake for
the first 3 or 8 h of the dark period, and in rats that had been sleep deprived during
the light period for 3 or 8 h. This experimental paradigm allowed us to distinguish
between changes in gene expression related to sleep and waking per se as opposed
to circadian time or to the sleep deprivation procedure.
This is an important issue, because several laboratories have recently performed
genome-wide expression analyses to isolate genes regulated by the circadian clock.
These studies have identified hundreds of transcripts cycling in the brain and in
peripheral tissues of mice17–20 and flies21–25 as a function of circadian time. Cycling
genes have been involved in extremely diverse biological functions, from protein
synthesis and immune response to metabolism, and may thus play a role in biological
processes that change between day and night, including wakefulness and sleep.

FIGURE 11.1 Biological functions associated with transcripts with higher expression in
wakefulness (left box) and sleep (right box). The tree on the left (dots and connecting paths)
represents biological processes annotations according to the gene ontology hierarchy.
However these studies did not control for behavioral state and therefore could not
determine to what extent changes in gene expression between day and night depend
on circadian time or on sleep and wakefulness.
We have also examined gene expression in the brain of rats chronically deprived
of sleep for long periods of time, ranging from 4 to 14 days.26 Prolonged sleep loss
was enforced using the disk-over-water method,27 the best controlled method of
long-term sleep deprivation. The main target of our analysis has been the cerebral
cortex because it generates the characteristic electrical rhythms of sleep;28 responds
to prolonged wakefulness with increasing sleep pressure;29 is responsible for the
cognitive defects observed after sleep deprivation;30–31 and is at the center of most
hypotheses concerning the functions of sleep.28,32–35
The studies performed so far have allowed the screening of more than 20,000
transcripts, including an estimated ~15,000 transcripts using Affymetrix GeneChip
technology (GeneChips RGU34A, B, C; [16]) and ~10,000 transcripts using mRNA
DD and nylon membrane arrays.13–15 Since the number of genes expressed in the rat
cerebral cortex is likely to range between 15,000 and 30,000,36–37 this screening
represents the most extensive (yet probably still not exhaustive) analysis of state-
dependent changes in gene expression performed so far. The following conclusions
were derived from this study.

First, up to ~5% of the transcribed sequences tested in the cerebral cortex (~800
out of 15,000) were found to be up- or down-regulated in rats that had slept for 8
h relative to rats that had been spontaneously awake or sleep deprived for a similar
period of time. These sequences included both known (annotated) transcripts as well
as expressed sequence tags (ESTs). In the cerebral cortex of the same animals, a
similar number of transcribed sequences were found to change their expression
because of time of day, rather than because of behavioral state. Day- or nighttime
and sleep or wakefulness appear to influence gene expression in the cerebral cortex
to a similar extent. A direct implication of these results is that changes in behavioral
state should be taken into account in all gene expression studies.
Second, the number of known transcripts upregulated during wakefulness (wake-
related genes) was similar (~100) to the number of transcripts upregulated during
sleep (sleep-related genes). Thus, although sleep is a state of behavioral inactivity,
it is associated with the increased expression of many genes in the brain. Moreover,
the increased expression in the brain during sleep was found to be specific, because
transcripts that were sleep-related in the brain were not sleep-related in other tissues
such as liver and skeletal muscle.16
Third, ~40% of the genes wake-related in the cerebral cortex were also wake-
related in the cerebellum. Similarly, 50% of the cortical sleep-related genes were
also sleep-related in the cerebellum. The finding that molecular correlates of sleep
and wakefulness are found in the cerebellum indicates that cellular processes asso-
ciated with sleep may occur in brain structures that are not known for generating
sleep rhythms. This suggests that, at the cellular level, functions associated with
sleep may take place whether or not electrographic signs of sleep can be recorded.
Finally and most importantly, a functional analysis of transcripts modulated by
behavioral state suggests that sleep and wakefulness may favor different cellular
processes. Several transcripts involved in energy metabolism (mitochondrial genes,
GLUT1), excitatory neurotransmission (Narp, Vesl/Homer), transcriptional activation
(Per2, NGFI-A, NGFI-B, CHOP), memory acquisition (Arc, NGFI-A, BDNF), and
cellular stress (HSPs, Bip) were wakefulness-related. Among sleep-related tran-
scripts was Dbp, which in other tissues is regulated by the circadian clock. Sleep-
related transcripts also included a two-pore domain potassium channel controlling
resting membrane potential (TREK-1); key components of the translational machin-
ery (translation elongation factor 2, initiation factor 4AII); and genes involved in
depotentiation, depression, as well as in the consolidation of long-term memory
(e.g., calcineurin, calmodulin-dependent protein kinase IV). A large number of sleep-
related transcripts are involved in membrane trafficking and maintenance, including
synaptic vesicle turnover (Rab genes, NSF; ARF1, ARF3) glia/myelin function
(MOBP, MAG, plasmolipin carbonic anhydrase II), and synthesis and transport of
glia-derived cholesterol (e.g., HMG-CoA synthase, squalene synthase), the limiting
factor for synapse formation and maintenance.
Wakefulness-related transcripts may help the brain to face high energy demand,
high synaptic excitatory transmission, high transcriptional activity, the need for
synaptic potentiation in the acquisition of new information, and the cellular stress
that may derive from one or more of these processes. An analysis of brain sleep-
related transcripts supports an involvement of sleep in protein synthesis and in

complementary aspects of neural plasticity such as synaptic depression35 and sug-

gests for the first time that sleep may play a significant role in membrane trafficking
and maintenance. Our findings, in line with intracellular recording studies,28 suggest
that sleep, far from being a quiescent state of global inactivity, may actively favor
specific cellular functions.
Many transcripts upregulated during wakefulness are induced diffusely in the
cerebral cortex and in many other brain regions. We hypothesized that a key factor
responsible for their induction might be the level of activity of neuromodulatory
systems, such as the noradrenergic and the serotonergic systems. These systems
project diffusely to most of the brain, and their activity is strictly state-dependent.
Noradrenergic neurons of the locus coeruleus fire regularly at very low rates during
sleep, whereas during wakefulness they fire at higher rates and emit phasic, short
bursts of action potentials in response to salient events.38 Norepinephrine enhances
brain information transmission, promotes attentive processes, and can enable various
forms of activity-dependent synaptic plasticity by stimulating gene transcription.15
To assess the role of the noradrenergic system in the induction of gene expression
during wakefulness, we studied behavior, brain electrical activity, and mRNA levels
of several genes in normal rats and in rats in which the central noradrenergic system
had been lesioned either bilaterally or unilaterally.15,39 We found that after the lesion
of the locus coeruleus waking behavior associated with a normal low-voltage fast
activity EEG was not accompanied by the induction of molecular markers of plas-
ticity such as c-fos, NGFI-A, P-CREB, Arc, and BDNF.
These results indicate that the activation of the EEG can be dissociated from
the activation of gene expression and that the noradrenergic system plays a major
role in the induction of plasticity-related genes during wakefulness. Whether such
a role extends to other functional classes of state-dependent genes is the subject of
ongoing research in our laboratory. Nevertheless the available findings suggest that
the reduced expression of plasticity-related genes due to the reduced firing of locus
coeruleus neurons may be a key factor that determines why the ability to learn new
material is impaired during sleep. Like locus coeruleus cells, serotonergic neurons
of the dorsal raphe also fire at higher levels during wakefulness and decrease their
firing during sleep.40 However in sharp contrast to noradrenergic neurons, dorsal
raphe cells are activated during repetitive motor activity such as locomoting, groom-
ing, or feeding and are inactivated during orientation to salient stimuli.41 Lesions of
the dorsal raphe nucleus were unable to affect the expression of c-fos, NGFI-A, P-
CREB, Arc, and BDNF, either during wakefulness or during sleep.42 Thus the nora-
drenergic, but not the serotonergic, system plays a crucial role in state-dependent
brain gene expression.
Most of the waking-related and sleep-related genes discussed above did not
change their expression if sleep loss was prolonged. One important exception is
represented by the enzyme arylsulfotransferase (AST), which is induced more mark-
edly after several days than after several hours of sleep deprivation.26 The progres-
sively stronger induction of AST is the first demonstration of a molecular response
in the brain that is related to the duration of sleep loss. AST is responsible for the
sulfonation of norepinephrine, dopamine, and to a lesser extent, serotonin. AST
induction during sleep deprivation may therefore constitute a homeostatic response

to the uninterrupted activity of the central noradrenergic system during wakefulness.

This could indicate that at least some of the detrimental effects of sleep loss may
be dependent on the continuous activation of the noradrenergic system and that an
important function of sleep is that of counteracting the effects of continued monoam-
inergic discharge.
REFERENCES
1. Panov A., RNA and protein content of brain stem cells after sleep deprivation, Riv.
Biol., 75, 95–99, 1982.
2. Vitale-Neugebauer A., Giuditta A., Vitale B., and Giaquinto S., Pattern of RNA
synthesis in rabbit cortex during sleep, J. Neurochem., 17, 1263–1273, 1970.
3. Giuditta A., Rutigliano B., and Vitale-Neugebauer A., Influence of synchronized sleep
on the biosynthesis of RNA in two nuclear classes isolated from rabbit cerebral cortex,
J. Neurochem., 35, 1259–1266, 1980.
4. Bobillier P., Sakai F., Seguin S., and Jouvet M., Deprivation of paradoxical sleep and
in vitro cerebral protein synthesis in the rat, Life Sci., 10, 1349–1357, 1971.
5. Brodskii VIa, Gusatinskii V.N., Kogan A.B., and Nechaeva N.V., Variations in the
intensity of H3-leucine incorporation into proteins during slow-wave and paradoxical
phases of natural sleep in the cat associative cortex, Dokl. Akad. Nauk. SSSR, 215,
748–750, 1974.
6. Ramm P. and Smith C.T., Rates of cerebral protein synthesis are linked to slow wave
sleep in the rat, Physiol. Behav., 48, 749–753, 1990.
7. Nakanishi H., Sun Y., Nakamura R.K., Mori K., Ito M., Suda S., Namba H., Storch
F.I., Dang T.P., Mendelson W., Mishkin M., Kennedy C., Gillin J.C., Smith C.B., and
Sokoloff L., Positive correlations between cerebral protein synthesis rates and deep
sleep in Macaca mulatta, Eur. J. Neurosci., 9, 271–279, 1997.
8. Sheng M. and Greenberg M.E., The regulation and function of c-fos and other
immediate early genes in the nervous system, Neuron, 4, 477–485, 1990.
9. Cirelli C. and Tononi G., On the functional significance of c-fos induction during the
sleep/waking cycle, Sleep, 23, 453–469, 2000.
10. Pompeiano M., Cirelli C., and Tononi G., Immediate-early genes in spontaneous
wakefulness and sleep: Expression of c-fos and NGFI-A mRNA and protein, J. Sleep
Res., 3, 80–96, 1994.
11. Cirelli C., Pompeiano M., and Tononi G., Sleep deprivation and c-fos expression in
the rat brain, J. Sleep Res., 4, 92–106, 1995.
12. Pompeiano M., Cirelli C., Ronca-Testoni S., and Tononi G., NGFI-A expression in
the rat brain after sleep deprivation, Mol. Brain Res., 46, 143–153, 1997.
13. Cirelli C. and Tononi G. Differences in gene expression between sleep and waking
as revealed by mRNA differential display, Mol. Brain Res., 56, 293–305, 1998.
14. Cirelli C. and Tononi G., Gene expression in the brain across the sleep-waking cycle,
Brain Research, 885, 303–321, 2000.
15. Cirelli C. and Tononi G., Differential expression of plasticity-related genes in waking
and sleep and their regulation by the noradrenergic system, J. Neurosci., 20,
9187–9194, 2000.
16. Cirelli C., Gutierrez C.M., and Tononi G., Extensive and divergent effects of sleep
and wakefulness on brain gene expression, Neuron, 41, 35–43, 2004.

17. Panda S., Antoch M.P., Miller B.H., Su A.I., Schook A.B., Straume M., Schultz P.G.,
Kay S.K., Takahashi J.S., and Hogenesch J.B., Coordinated transcription of key
pathways in the mouse by the circadian clock, Cell, 109, 307–320, 2002.
18. Ueda H.R., Chen W., Adachi A., Wakamatsu H., Hayashi S., Takasugi T., Nagano
M., Nakahama K., Suzuki Y., Sugano S., Iino M., Shigeyoshi Y., and Hashimoto S.,
A transcription factor response element for gene expression during circadian night,
Nature, 418, 534–539, 2002.
19. Storch K.F., Lipan O., Leykin I., Viswanathan N., Davis F.C., Wong W.H., and Weitz
C.J., Extensive and divergent circadian gene expression in liver and heart, Nature,
417, 78–83, 2002.
20. Akhtar R.A., Reddy A.B., Maywood E.S., Clayton J.D., King V.M., Smith A.G., Gant
T.W., Hastings M.H., and Kyriacou C.P., Circadian cycling of the mouse liver tran-
scriptome, as revealed by cDNA microarray, is driven by the suprachiasmatic nucleus,
Curr. Biol., 12, 540–550, 2002.
21. McDonald M.J. and Rosbash M., Microarray analysis and organization of circadian
gene expression in Drosophila, Cell, 107, 567–578, 2001.
22. Claridge-Chang A., Wijnen H., Naef F., Boothroyd C., Rajewsky N., and Young M.W.,
Circadian regulation of gene expression systems in the Drosophila head, Neuron, 32,
657–671, 2001.
23. Ueda H.R., Matsumoto A., Kawamura M., Iino M., Tanimura T., and Hashimoto S.,
Genome-wide transcriptional orchestration of circadian rhythms in Drosophila, J.
Biol. Chem., 277, 14048–14052, 2002.
24. Lin Y., Han M., Shimada B., Wang L., Gibler T.M., Amarakone A., Awad T.A., Stormo
G.D., Van Gelder R.N., and Taghert P.H., Influence of the period-dependent circadian
clock on diurnal, circadian, and aperiodic gene expression in Drosophila melano-
gaster. Proc. Natl. Acad. Sci. USA 99, 9562–9567, 2002.
25. Ceriani M.F., Hogenesch J.B., Yanovsky M., Panda S., Straume M., and Kay S.A.,
Genome-wide expression analysis in Drosophila reveals genes controlling circadian
behavior, J. Neurosci., 22, 9305–9319, 2002.
26. Cirelli C. and Tononi G., Changes in gene expression in the cerebral cortex of rats
after short-term and long-term sleep deprivation, Sleep, 22 (S1), 113, 1999.
27. Rechtschaffen A. and Bergmann B.M., Sleep deprivation in the rat, an update of the
1989 paper, Sleep, 25, 18–24, 2002.
28. Steriade M. and Timofeev I., Neuronal plasticity in thalamocortical networks during
sleep and waking oscillations, Neuron, 37, 563–576, 2003.
29. Borbély A.A. and Achermann P., Sleep homeostasis and models of sleep regulation,
J. Biol. Rhythms, 14, 557–568, 1999.
30. Horne J.A., Why We Sleep: The Functions of Sleep in Humans and Other Mammals,
Oxford University Press, Oxford, 1988.
31. Van Dongen H.P.A., Maislin G., Mullington J.M., and Dinges D.F., The cumulative
cost of additional wakefulness: dose-response effects on neurobehavioral functions
and sleep physiology from chronic sleep restriction and total sleep deprivation, Sleep,
26, 117–126, 2003.
32. Moruzzi G., The sleep-waking cycle. Ergeb. Physiol., 64, 1–165, 1972.
33. Krueger J.M., Obal F. Jr, Kapas L., and Fang J., Brain organization and sleep function,
Behav. Brain Res., 69, 177–186, 1995.
34. Maquet P., Sleep function(s) and cerebral metabolism, Behav. Brain Res., 69, 75–83,
1995.
35. Tononi G. and Cirelli C., Sleep and synaptic homeostasis: A hypothesis, Brain Res.
Bull., 62, 143–150, 2003.

36. Milner F.D. and Sutcliffe J.G., Gene expression in rat brain, Nucleic Acid Res., 11,
5497–5520, 1983.
37. Velculescu V.E., Madden S.L., Zhang L, Lash A.E., Yu J., Rago C., Lal A., Wang
C.J., Beaudry G.A., Ciriello K.M., Cook B.P., Dufault M.R., Ferguson A.T., Gao Y.,
He T.C., Hermeking H., Hiraldo S.K., Hwang P. M., Lopez M.A., Luderer H.F.,
Mathews B., Petroziello J.M., Polyak K., Zawel L., Kinzler K.W. et al., Analysis of
human transcriptomes, Nat. Genet., 23, 387–388, 1999.
38. Aston-Jones G. and Bloom F.E., Activity of norepinephrine-containing locus coer-
uleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle, J.
Neurosci., 1, 876–886, 1981.
39. Cirelli C., Pompeiano M., and Tononi G., Neuronal gene expression in the waking
state: a role for the locus coeruleus, Science, 274, 1211–1215, 1996.
40. McGinty D.J. and Harper R.M., Dorsal raphe neurons: depression of firing during
sleep in cats, Brain Res., 101, 569–575, 1976.
41. Jacobs B.L. and Fornal C.A., Activity of serotonergic neurons in behaving animals,
Neuropsychopharmacology, 21, 9S–15S, 1999.
42. Tononi G., Cirelli C., and Shaw P.J., The Molecular correlates of sleep, waking, and
sleep deprivation, in Human Frontier Workshop VIII, The Regulation of Sleep, Borbély
A., Hayaishi O., Sejnowski T.J., and Altman J.S., Eds., Human Frontier Scientific
Press, Strasbourg, PA, 2000, pp.155–167.

12 Neuronal Reverberation
and the Consolidation of
New Memories across
the Wake-Sleep Cycle
Sidarta Ribeiro, Damien Gervasoni,
and Miguel A.L. Nicolelis
CONTENTS
Introduction
Searching for Neuronal Reverberation after Novel Sensory Stimulation
Novelty-Induced Neuronal Reverberation Is Sustained and Long-Lasting
Novelty-Induced Neuronal Reverberation Occurs in
Multiple Forebrain Areas
Novelty-Induced Neuronal Reverberation Is Context-Dependent
Neuronal Reverberation Is State-Dependent and Peaks during
Slow-Wave Sleep
Forebrain Reverberation Consists of Low-Fidelity Replay at
Physiological Speeds
A Model for the Complementary Roles of SW and REM Sleep in
Memory Consolidation
Acknowledgments
References
INTRODUCTION
In mammals and birds, long episodes of nondreaming sleep (slow-wave sleep, SW)
are followed by short episodes of dreaming sleep (rapid-eye-movement sleep,
REM).1–9 Despite early insight10 it was not until the 1970s that science began to
recognize the key role of sleep in memory consolidation. The main findings sup-
porting this view are the detrimental effects of sleep deprivation on learning,11–26 the
improved memory retention in rats when REM sleep is enhanced,27 the increase in
sleep amounts following memory acquisition,28–35 and the fact that theta rhythm, a
learning-related36–42 hippocampal oscillation typical of high arousal,43–49 also char-
0-8493-1519-0/05/$0.00+$1.50

acterizes REM sleep.50–53 Given the involvement of the hippocampus in memory

acquisition,54–66 these results indicated that sleep is a privileged offline window for
the processing of novel and ecologically relevant information.67–69
What are the mechanisms underlying the beneficial role of sleep for learning?
Donald Hebb was perhaps the first to point out that the neuronal activity present
during memory encoding must linger in the brain until structural cellular changes
have time to occur, thus transforming a short-lived reverberatory trace in a long-
lasting memory.70 The chase for the neural mechanisms underlying the mnemonic
role of sleep had a major breakthrough with a pioneering investigation of the post-
stimulus activity of hippocampal neurons in rats.71 This elegant study took advantage
of the fact that pyramidal hippocampal neurons fire very selectively according to
the spatial position of the animal in a given environment.72,73
By way of chronic electrode implants in the hippocampus, Pavlides and Winson
first identified and recorded neuronal pairs with nonoverlapping place fields. Then
rats were restricted from entering the place field of either cell overnight. Finally
animals were confined for 10–15 minutes within the place field of one of the cells
to strongly stimulate one neuron while suppressing the activity of the yoked one.
By quantifying the neuronal firing rates during and after spatial confinement across
the wake-sleep cycle, Pavlides and Winson found that the firing rates observed during
waking (WK) experience recur in the hippocampus during ensuing SW and REM
sleep71 (Figure 12.1 A). These results indicated that sleep harbors the first mechanism
postulated by Donald Hebb to be necessary for learning, namely the post-acquisition
neuronal reverberation of memory traces.70
Subsequent exploration was prolific: Post-acquisition neuronal reverberation
during sleep or quiet WK was found to preserve the temporal firing relationships of
alert, exploratory WK in the hippocampus74–80 and the cerebral cortex,81,82 causing
a correlated replay of activity patterns across two-74 or many-neuron ensembles.79
To date, experience-dependent brain reactivation during sleep has been observed in
rodents,71,74–76,78–81 nonhuman primates,82 humans,83 and even songbirds (Figure 12.1
B),84 pointing to a very general biological phenomenon. Furthermore experience-
dependent brain expression of the plasticity-related gene zif-268 was observed during
REM sleep (Figure 12.1 C),85,86 providing a compelling tie between neuronal reac-
tivation during sleep and cellular plasticity able to consolidate memories.87–101 Finally
and most importantly, post-acquisition brain reactivation during sleep has been
shown to be proportional to memory acquisition in rats102 and humans (Figure 12.1
D),103 and to quantitatively predict learning.25,104
In spite of the positive evidence, the neural reverberation hypothesis for memory
consolidation during sleep faces several objections:
• The neocortical reverberation detected to this date is extremely subtle and

decays rapidly within less than 1 H of memory trace formation.81,82 Such
short-lived reverberation falls short of explaining the disruption of mem-
ory traces by sleep deprivation several hours and even days after initial
acquisition.11–23,25,26
• Strictu sensu neuronal reverberation during sleep in mammals has only
been investigated in the hippocampo-cortical loop,71,74–76,78–82 making it

0.60
A Post-Nonexposure
Post-Exposure
B
0.50
*
SPIKES/SEC
0.40
0.30 SONG
*
0.20 SLEEP
0.10
100ms
0.00
WK SWS REM
C WK SWS REM D
4
Brain activation during REM sleep

Control
(adjusted CBF)
1
Enriched Environment
-1
-2
-3
Low High -20 0 20 40 60 80 100
Zif-268 Expression Learning level attained prior to sleep
FIGURE 12.1 (See color insert following page 108.) (A) Hippocampal place cells are
reactivated during SW and REM sleep after WK exposure to their place fields. Asterisk for
p<0.05. (Modified from Pavlides, C. and Winson, J., J. Neurosci., 9, 2907, 1989. With
permission.) (B) Premotor neurons in nucleus RA of a zebra finch accurately replay singing-
specific activity during sleep. Shown are raw traces of neuronal activity (900 ms) recorded
during singing (premotor activity) and sleep (spontaneous activity). A color spectrograph of
the song that the bird sang is shown on top, time-aligned to the premotor activity, with
horizontal bars indicating different song syllables. (Modified from Dave, A.S. and Margoliash,
D., Science, 290, 812, 2000. With permission.) (C) The plasticity-related gene zif-268 is
reinduced during REM sleep in an experience-dependent manner. Autoradiograms of brain
sections hybridized with zif-268 radioactive riboprobes. In controls kept in a familiar envi-
ronment, zif-268 expression decreased from WK to SW and REM sleep. In animals exposed
to a novel enriched environment for 3 H before the experiment, zif-268 levels decreased from
WK to SW sleep but increased from the latter to REM. This effect was particularly noticeable
in the cerebral cortex and the hippocampus. (Modified from Ribeiro, S. et al., Learn. Mem.,
6, 500, 1999.) (D) Learning levels attained prior to sleep modulate regional cerebral blood
flow (CBF) during REM sleep in humans, as measured by positron emission tomography.
(Modified from Peigneux, P. et al., Neuroimage, 20, 125, 2003.)103

difficult to determine whether the phenomenon is particular to this neural

circuit or whether it represents global experience-dependent changes in
the brain.
• Neuronal reverberation has mostly been observed in highly-trained animal
subjects,74–76,79–82,84 raising skepticism about its relevance for the acquisi-
tion and consolidation of novel information.105
• Experience-dependent neuronal reverberation has been reported to occur
in all behavioral states,71,74,75,79–81 including WK.76,78,82
Although the first finding in this regard has hinted at a possible predominance of
reverberation during SW sleep (Figure 12.1 A),71 a comprehensive comparison of
the relative contributions of WK, SW, and REM sleep for neuronal reverberation is
still missing. To further complicate the issue, recent studies have raised the possibility
that neuronal processing may occur at either slower or faster speed than normal
physiological rates during REM79 and SW,76,80 respectively, thus it is uncertain at
the moment how neuronal reverberation relates to different behavioral states.
SEARCHING FOR NEURONAL REVERBERATION

AFTER NOVEL SENSORY STIMULATION
In order to address these objections, we set out to investigate the effects of a transient
novel tactile experience on the long-term evolution of ongoing brain activity across
the major behavioral states of the rat.106 In each of the 5 animals studied, the
extracellular activity of up to 159 neurons per animal and local field potentials (LFP)
representing larger-scale neural rhythms were simultaneously recorded from four
different brain regions: hippocampus (HP), primary somatosensory barrel-field cor-
tex (CX), ventral posteromedial thalamic nucleus (TH), and putamen (PU) (Figure
12.2 A). These brain regions were chosen because they comprise three major fore-
brain circuit loops essential for rodent species-specific behaviors. Rats are nocturnal
gatherers that exhibit a variety of exploratory behaviors during the night, sleeping
intermittently and mostly during the day51 (top panel, Figure 12.2 B). In the wild,
rats rely on spatial navigation and exquisite whisker-based tactile discrimination to
explore new territories in search of food.107 The cortico-thalamic, cortico-hippo-
campal, and cortico-striatal loops probed in this study have been implicated in tactile
information processing,108,109 spatial navigation, and memory formation,59,72 and the
execution of complex motor sequences.110,111
In our study106 neural signals were continuously recorded across the natural
sleep-wake cycle for 48–96 H, with a single 1-H exposure to four complex objects
placed in the four corners of the recording box (Figure 12.2 B). All objects were
strictly novel to the subjects and were designed to maximize shape, texture, and
behavioral value differences (Figure 12.2 C). Objects were presented halfway
through the recording time around midnight, when lights were off and WK reached
a peak (Figure 12.2 B), to maximize the drive for whisker-based tactile exploration
of the environment. As expected, this paradigm strongly increased WK relative to
sleep during the exposure time (Figure 12.2 D), leading to robust novel sensory
stimulation. Other than novel stimulation and the periodic removal of waste and

introduction of food pellets and water, animals were kept undisturbed in the same
environment throughout the recordings. Our paradigm produced marked and acute
exploratory behavior without disrupting the large-scale sleep-wake structure across
the many hours of recording (Figure 12.2 B, top panel). The experiment, therefore,
consisted of a naturalistic behavioral paradigm involving multiple novel sensory and
spatial cues, and was designed to maximize novelty-induced neuronal changes as
opposed to changes caused by behavioral over-training.
In order to investigate the long-term effects of novel stimulation on the spa-
tiotemporal evolution of ongoing neuronal activity, we took advantage of a neuronal
ensemble correlation method previously shown to detect experience-dependent
reactivation of rodent hippocampal ensembles during SW and REM sleep.79 This
method generalizes the concept of pairwise neuronal correlations74,81,82 to an arbi-
trarily large number of neurons, quantifying the degree of similarity between
spatiotemporal patterns of neuronal activity by way of a firing-rate-normalized
template-matching algorithm (Figure 12.2 E). Templates of alert WK neuronal
ensemble activity were selected from moments when animals made whisker contact
with the novel objects (n = 5 templates per animal). Control templates were selected
from epochs of alert WK 24 H (three rats) or 48 H (two rats) before novel stimulation
(n = 5 templates per animal), during which familiar tactile stimulation was produced
by the contact of whiskers with the smooth walls of the recording box to which
animals were habituated. Templates were matched against the entire record of
neuronal activity using the neuronal ensemble correlation method (Figure 12.2 F).
The resulting correlation temporal profiles were averaged for each template set,
aligned with reference to the light-darkness cycle to control for possible circadian
effects, and compared.
NOVELTY-INDUCED NEURONAL REVERBERATION IS

SUSTAINED AND LONG-LASTING
First we tested whether the neuronal ensemble correlation method could detect in
our dataset any trace of increased neuronal reverberation after exposure to the novel
stimuli. For this we examined correlation profiles obtained for all recorded neurons
(3–4 brain areas pooled together) in each animal. As shown for 2 different animals
in Figure 12.3 A, post-novelty average correlation distributions were significantly
right-shifted relative to pre-novelty distributions (ANOVA of mean pre- and post-
novelty correlations over 24 or 48 H, n = 5 animals, F = 9.5, d.f. 1, P = 0.016). This
indicated that the neuronal firing patterns concomitant with novel stimulation per-
sisted significantly more during the ensuing time than patterns sampled 24 or 48 H
before novel stimulation, when animals were in the same behavioral state (alert
WK), but without novel objects to explore. The effect was independently observed,
to a variable degree, in all the five animals studied (Bonferroni, P <0.01), corrobo-
rating the efficacy of the neuronal ensemble correlation method for the detection of
experience-dependent neuronal reverberation.79
Next we assessed whether the neuronal ensemble correlation method could
detect neuronal reverberation lasting at least more than 1 H after exposure to novel
stimulation. Figure 12.3 B shows the temporal evolution of neuronal ensemble

correlations for 2 different animals. Despite the marked inter-animal variability in

the shapes and magnitudes of these profiles, a significant and sustained increase of
neuronal ensemble correlations after exposure to novel stimulation was observed in
all animals. Importantly, these increases lasted well above 1 H. Indeed, post-novelty
increased neuronal ensemble correlations decayed slowly over time, and persisted
A B 100
Time spent awake (%)

CX
HP 80
60
TH
PU
40
20
C
0
Time (Hours)
0 48 96
Pre-Novelty Post-Novelty
200
D Pre
Exploration
E
150 Post
Time Spent in State (%)
100
50
WK SWS REM
F C 1 , C 2 , C 3 ,Ö C 1 , C 2 , C 3 ,Ö
Pre-Novelty Post-Novelty
FIGURE 12.2

above the pre-novelty baseline after 48 H as revealed by the temporal evolution of

associated Bonferroni P values.106
These results indicate that post-stimulus neuronal reverberation occurs in rats
that are completely naïve with respect to the reference stimuli. Such novelty-induced
neuronal reverberation is, at least in principle, capable of implementing the mne-
monic function anticipated by Hebb70 and directly contradicts the notion that only
the performance of highly trained behaviors is followed by neuronal reverberation.105
The new data also demonstrate that sustained experience-dependent neuronal rever-
beration can be detected in the forebrain up to 48 H after exposure to novel
stimulation. As mentioned above, previous measurements of persistent changes in
neuronal firing rates71 or pairwise neuronal correlations74,75,81,82,105 had not been able
to track the reverberation of neuronal activity produced by WK experience beyond
a mere hour after the reference stimulus. Our results indicate that the neuronal
ensemble correlation method79 is more robust in that respect, detecting experience-
dependent changes that persist for several hours after the initial encoding. Thus the
use of the neuronal ensemble correlation method for the analysis of very long neural
records provides evidence for the first time of reverberatory processes that are
compatible with memory impairment effects of sleep deprivation applied hours or
days after training.13,15,18,19,21,23,112
FIGURE 12.2 (See facing page.) (A) Neuroanatomical location of multielectrode implants,
shown on a schematic parasaggital section based on.141 Indicated are the cerebral cortex (CX),
the hippocampus (HP), the thalamus (TH), and the putamen (PU). (B) Experimental design.
Top panel shows a representative example of the strong circadian dynamics of the rat sleep-
wake cycle (rat #5). Grey bands indicate lights-off, white bands indicate lights-on; notice the
fixed 12-h periods of darkness and light. Bottom panels: Animals continuously recorded for
up to 96 hours were kept undisturbed except for a 1-h period of novel sensory stimulation
(white segment) produced by the tactile exploration of 4 distinct novel objects placed at the
corners of the recording box. Neural data from pre- and post-novelty periods (middle panel,
black and red segments, respectively) were compared. (C) Four different objects were used
to produce novel complex stimulation of different shapes and textures: a food cache filled
with Fruit Loops, a shoe brush, a golf ball mounted in a spring, and a spiky object made of
metal pins attached to a wooden axis. (D) Data for rat #3. All animals were highly habituated
to the recording box, so exposure to novel complex objects caused a general increase in time
spent in WK with respect to SW and REM sleep during the exploration of the objects, as
compared to adjacent pre- and post-periods of equal length (60 min). (E) Neuronal ensemble
correlation method. Neuronal activity templates (red boxes) were compared with extensive
recordings of neuronal action potentials (top panel, green ticks) by way of an off-line template-
matching algorithm79 that generalizes the notion of pair-wise correlations to neuronal ensem-
bles of any size. Templates and targets (white boxes) were binned, firing-rate normalized,
and correlated (middle panel). This procedure yields a time series of neuronal ensemble
correlations for each template-target sliding match (bottom panel). (F) Templates of interest
(9 second-long red boxes) were sampled around the origin of pre- and post-novelty periods
during alert WK and slid against their corresponding neuronal targets so as to sample neuronal
correlations every 30 sec for up to 48 H.

A Rat 4 Pre B Rat 4

Number of Observations Post
Neuronal Correlations
Rat 5 Rat 5
0 12 24 36 48
Neuronal Correlations Time (Hours)
C
Cerebral Cortex Hippocampus Putamen Thalamus
.
Neur Correl
.
0 12 24 0 12 24 0 12 24 0 12 24
Time (Hours) Time (Hours) Time (Hours) Time (Hours)
Familiar Environment Novel Environment Novel Environment

(whisker contact) (whisker contact) (no whisker contact)
FIGURE 12.3 (A) Post-novelty neuronal correlations were significantly larger (right-shifted)
than pre-novelty correlations in all animals studied. (B) Temporal profiles of multiple-area
neuronal ensemble correlations reveal long-lasting reverberation. Grey bands indicate lights-
off; white bands indicate lights-on. (C) Long-lasting neuronal reverberation occurs in the
cerebral cortex, hippocampus, putamen, and thalamus. Shown are temporal profiles of neu-
ronal ensemble correlations calculated for single areas (all panels correspond to rat #1 except
the putamen, which corresponds to rat #3). (D) Neural activity sampled when animals were
aroused but not touching the objects yielded enhanced neuronal reverberation that was nearly
identical to that obtained when animals made sensory contact with the objects. This indicates
that neuronal reverberation reflects the novelty context rather than stimulus complexity.

NOVELTY-INDUCED NEURONAL REVERBERATION

OCCURS IN MULTIPLE FOREBRAIN AREAS
In order to assess the anatomical distribution of experience-dependent neuronal
reverberation, we performed the neuronal ensemble correlation analysis for each
area separately. Despite considerable interanimal and interarea variation in the mag-
nitude and shape of correlation profiles, significant changes between pre- and post-
novelty correlations were observed in all areas studied (CX 5/5, HP 3/4, PU 4/4,
and TH 4/5 rats; Bonferroni, P <0.05) (Figure 12.3 C). The temporal evolution of
P values (Bonferroni, P <0.05) associated with single-area correlation profiles con-
firmed that significant reverberation was present in 16 of 18 recording sites for up
to 48 H after exposure to novel stimulation.106
At first glance this indicates that all forebrain areas studied are equally capable
of reverberating neuronal patterns of activity after novel stimulation. Indeed,
experience-dependent changes were not statistically different across different fore-
brain areas (ANOVA, F = 0.24, d.f. 3, P = 0.86). Within the five animals studied,
no sign of forebrain neuroanatomical specificity was found in the correlations
measured, and in particular no significant differences between hippocampal and
extra-hippocampal areas could be detected. However these results may be related
to the high variability of pre- and post-novelty correlation profiles in all animals,
which typically showed predominant effects in a different subset of areas for each
animal. For example, rat #4 showed marked reverberation in the HP but small
changes in the CX, and rat #5 showed just the opposite. The most widespread
reverberation was seen in rat #1, which showed sustained reverberation in the CX
and decaying reverberation in the HP and TH (Figure 12.3 C). A somewhat similar
pattern was seen in rat #5, but rat #3 showed strong reverberation only in the PU
(Figure 12.3 C), and rat #4 in the HP and TH. Rat #2 showed the least reverberation
of all, with somewhat stronger effects in the PU.106 At present it is unclear whether
such differences reflect real interanimal differences in the neuroanatomical distri-
bution and dynamics of neuronal reverberation, or rather a variation from animal
to animal due to possible differences in chronic electrode implants or other exper-
imental variables.
Taken together our results indicate that the tactile, gustatory, olfactory, spatial,
and motor activity produced by the free exploration of the experimental objects
engaged multiple forebrain structures in widespread neuronal reverberation. Our
results also suggest that enhanced neuronal reverberation (post- >precorrelations)
is not the only kind of experience-dependent change possible.106 Although post-
novelty traces (red) run above pre-novelty traces (black) for the majority of the
recorded sites, some animals showed significant long-lasting anti-reverberation (pre-
>post-correlations); i.e., patterns of activity that were statistically more dissimilar
from novel stimulation templates than expected by chance. Anti-reverberation
occurred in the HP (1 of 4 rats), PU (2 of 4 rats) and TH (1 of 5 rats) but not in
the CX. In principle the novelty-induced reverberation and anti-reverberation of
neuronal firing patterns could play balancing roles in the delineation of new memory
traces, embossing high- and low relieves in the vast synaptic landscape where
memories reside.

NOVELTY-INDUCED NEURONAL REVERBERATION IS

CONTEXT-DEPENDENT
Our results indicate that large-scale neuronal firing patterns generated during the
exploration of the experimental objects can recur for several hours after the reference
experience throughout the forebrain, but firing patterns associated with the walls of
the recording box are substantially less detectable over time. The sensory stimulation
provided by the free exploration of the experimental objects is at once novel and
complex. In contrast, exploration of the smooth walls of the recording box to which
animals were habituated elicited stimulation that was both familiar and simple.
Therefore the enhanced neuronal reverberation detected after object exploration
could in principle be related not to object novelty but rather to object complexity.
In order to disambiguate these effects, we scrutinized the neuronal reverberation
associated with templates of neuronal activity sampled during alert WK within the
novel stimulation 1-h period but excluding moments of contact between whiskers
and objects. Surprisingly no-contact templates yielded correlation profiles that were
almost indistinguishable from those obtained when animals had tactile contact with
the novel objects (Figure 12.3 D). Equivalent results were obtained for CX, HP, TH,
and PU.106
The exploration of a novel environment enhanced the reverberation of all the
neuronal activity patterns concomitant with that experience and not just of those
corresponding to moments in which animals received tactile inputs from the objects.
This rules out the possibility that stimulus complexity, rather than novelty, was the
underlying cause of the enhanced neuronal reverberation observed after exploration
of the objects. It also indicates that the kind of experience-dependent neuronal
reverberation detected by the neuronal ensemble correlation method79 does not reflect
the specific features of the stimuli but is related to the overall behavioral salience
of the novel stimulation period; i.e., context- rather than stimulus-specific. In prin-
ciple these results are compatible with a slow and progressive process of memory
consolidation,67 proportional to the novelty of the experience, and able to bind
together a multitude of contextual cues related to its core sensory elements.113
NEURONAL REVERBERATION IS STATE-DEPENDENT

AND PEAKS DURING SLOW-WAVE SLEEP
Results observed in single forebrain areas indicated that neuronal ensemble corre-
lations peak during discrete epochs that may last a few hours, causing marked
oscillations of the correlation trace. These observations suggested that some under-
lying biological process, with slow evolution but with sharp phase transitions, gov-
erns the long-term reverberation of neuronal firing patterns. To test whether transi-
tions in the wake-sleep cycle could amount for these effects, we investigated how
experience-dependent changes in neuronal ensemble correlations varied across the
three major rat behavioral states, which were coded as WK, SW, and REM sleep
based on a spectral analysis of LFP and visual inspection of videotaped behaviors,
according to previously described criteria.85

A comparison across states of post- and pre-correlations ratios calculated from

averages of entire recordings indicated a significant state-specific effect (ANOVA, F
= 9.289, d.f. 2, P = 0.0004), with SW ratios being significantly higher than those of
both WK (Bonferroni, P <0.05) and REM (Bonferroni, P <0.003). Indeed, significant
state-specific differences in post-/pre-correlation ratios were individually detected in
4 out of 5 animals (ANOVAs, d.f. 2: rat #2, PU, F = 4.13, P = 0.039; rat #3, CX, F
= 6.45, P = 0.026; rat #4, HP, F = 3.99, P = 0.029; rat #5, CX, F = 13.81, P <0.0001).
The mean correlation values found in those recording sites for the three behavioral
states revealed that SW correlations were systematically larger than WK correlations;
several other recorded sites displayed similar but nonsignificant trends.106 Meanwhile
the REM correlations measured were variable and could not be consistently ranked
in relation to WK and SW sleep. Comparable neuronal reverberation between SW
and REM sleep was observed in only one animal (rat #5 CX). A major effect of SW
sleep on neuronal reverberation was corroborated by the temporal evolution of
successive state-specific Bonferroni P-values calculated for pre- and post-novelty 4-h
average correlations across all animals and brain areas studied.106
Altogether our results indicate that neuronal reverberation was consistently
stronger during SW sleep and decreasing during WK. This is remarkably well-
illustrated by a superimposition of behavioral state classification and neuronal
ensemble correlations (Figure 12.4 A), which reveals an exquisite long-term temporal
match between SW episodes (red) and epochs of increased neuronal ensemble
correlations in the CX. Likewise neuronal ensemble correlation troughs show a tight
correspondence with WK episodes (blue). This characteristic state-dependency per-
sisted throughout the 45 h of post-novelty recording (Figure 12.4 B). Notice that
REM sleep only showed SW-like results in one out of five animals (Rat #5 CX,
depicted in Figure 12.4). In the remaining animals, REM correlations were either
closer to WK levels than to SW levels or in between.106 Given this marked variability
and the very short duration of total REM sleep in comparison with total SW sleep
(WK 52%, SW 40%, and REM 8% of total recording time for five animals), one
must conclude that REM sleep plays a minor role in neuronal reverberation.106
Therefore the function of experience-dependent brain reactivation during REM
sleep71,79,83,103 remains to be explained. One attractive possibility yet to be tested is
that neuronal reverberation during REM sleep, being noisier than that of SW sleep,
may facilitate memory-trace restructuring and the generation of insights.114
Interestingly a comparison of the correlation temporal profile (Figure 12.4 C,
top panel) with the concurrent neuronal firing record (Figure 12.4 C, bottom panel)
reveals that SW correlation peaks correspond to periods of decreased firing rate, and
WK correlation troughs match epochs of increased neuronal activity. Figure 12.4 C
depicts data segments approximately 2 h long comprising the three major behavioral
states studied. The first segment (*) corresponds to the 60¢ of novel stimulation and
the immediately ensuing sleep-wake cycles, and the second segment (**) illustrates
sleep-wake episodes occurring ~15 h after the original experience. Although novel
stimulation templates of neuronal activity were selected from WK episodes charac-
terized by high firing rates, ensuing reverberation of these neuronal firing patterns
was most pronounced during SW sleep under lower firing rates. Neuronal ensemble
reverberation decreased but did not disappear during WK (Figure 12.4 D), in agree-

ment with the original findings of post-stimulus changes in hippocampal firing rates
(Figure 12.1 A),71 and a more recent investigation of the same issue.78 Taken together
these data indicate that novel experience causes sustained neuronal reverberation70
rather than discrete reactivation,74,105 in the sense that traces of a given salient
experience are continuously detectable during subsequent periods across all behav-
ioral states in a state-dependent manner.
A
* **
Time (minutes)
B
Time (minutes)
C D 0.3
WK
Pre
Cortical Neurons
Post
0.2
0.1
0.0
EXP -0.1
0 12 24 36 48
Time (hours)
FIGURE 12.4 (See color insert.)

The inverse correlation between neuronal ensemble correlations and concurrent

firing rates suggests that reverberating patterns of neuronal activity associated with
past novel experience are largely — but not completely — masked during WK by
incoming sensory inputs unrelated to the reference experience. By the same token,
peak neuronal ensemble correlations arise during SW sleep, when sensory interfer-
ence ceases. These observations corroborate the notion that the importance of sleep
for memory consolidation stems from the off-line processing of memory traces; i.e.,
from the absence of sensory interference.10,69,115 The consistent increase in neuronal
reverberation during SW sleep, the high interanimal variability of neuronal reverber-
ation during REM sleep, and the small contribution of REM sleep to total sleep time
suggest a major role for SW sleep in the post-acquisition recall of new memory traces.
FOREBRAIN REVERBERATION CONSISTS OF

LOW-FIDELITY REPLAY AT PHYSIOLOGICAL SPEEDS
One interesting aspect of the long-term assessment of forebrain neuronal ensemble
correlations is that the measurements obtained were typically small (on the order of
0.1 to 0.3), in complete agreement with values previously reported for
pairwise74,75,81,82 or many-neuron correlations.79 Qualitatively similar results were
observed for bins ranging from 5 ms to 1000 msec, with larger bin sizes correspond-
ing to higher correlation values, due to an averaging effect.106 This suggests that
neurons of multiple forebrain areas, once exposed to novel experience, do not
accurately replay prior WK activity patterns longer than 5 msec but show instead a
mild but long-lasting bias toward (or against) the reference activity patterns. High-
fidelity replay of neuronal firing patterns was not observed when single areas were
considered: Not a single template-to-target match (out of 979,200 matches sampled)
yielded correlation values higher than 0.45, indicating that novelty-induced neuronal
reverberation occurs at low-fidelity.
FIGURE 12.4 (See facing page.) (A) Neuronal reverberation is strongest during SW sleep.
The superimposition of successive neuronal ensemble correlations and concurrent behavioral
states for the CX neurons of rat #5 dramatically illustrate the state-dependency of neuronal
ensemble correlations, which are strongly increased by SW sleep but readily decreased by
WK. Nearly all correlation peaks correspond to SW episodes, and almost all troughs match
WK epochs. (B) State-dependent neuronal reverberation was sustained throughout the record-
ing period, as shown by segments representing the beginning (3200–3300¢), middle
(4700–4800¢) and end (5200–5250¢) of the experimental record. On the left panel, notice the
progressive increase of neuronal correlations across a single SW sleep episode (white arrows),
suggesting a progressive amplification of the memory trace. (C) Blow-up of two selected data
segments indicated by asterisks in (A). Despite having being sampled from moments of high
neuronal firing rates,* novel stimulation templates reverberate most strongly during SW sleep
when firing rates are low.*,** The high firing rates that characterize WK correspond to
decreased neuronal reverberation, probably due to sensory interference. (D) Post-novelty
neuronal ensemble correlations decrease during WK but do not reach pre-novelty levels,
indicating the occurrence of post-stimulus neuronal reverberation, and not reactivation.

Nonetheless studies of hippocampal place cells have proposed that a high-fidelity

replay of neuronal firing patterns during sleep may be achieved, assuming that
replayed patterns can undergo time compression and expansion76,79,80; i.e., that the
experience-dependent reverberation of neuronal firing patterns during sleep can be
slower (REM)79 or faster (SW)76,80 than during WK. To investigate this possibility,
we obtained template-to-target matches at different speed factors by comparing 250
msec-binned templates with targets binned within a range of different bin sizes (12.5
msec to 500 msec), temporally compressing and expanding target spike records
before matching them to templates. This procedure allowed us to determine the
magnitude of neuronal ensemble correlations for speed factors ranging from 0.5 to
20 times the physiological WK processing speed, which covers the reported optimum
speed ranges for SW76,80 and REM sleep.79
We found a predominance of neuronal reverberation during SW sleep for all
speed factors, as indicated by Bonferroni comparisons.106 However, no significant
differences were seen when post- and pre-correlations ratios (calculated from aver-
ages of entire recordings) were compared across different speed factors (ANOVA,
F = 1.496, d.f. 5, P = 0.19). Within any given state or area, neuronal reverberation
did not vary systematically with speed factor, and the temporal distribution of
neuronal ensemble correlations was largely insensitive to speed factor.106 We found
no evidence that forebrain neuronal reverberation can be optimized, assuming replay
speeds different from the WK normal rate. A subtle but consistent decrease of P-
values can be observed for speed factors 10¥ and 20¥ faster than normal WK rates,
while speed factors near the physiological range (2¥ to 0.5¥) show stronger and
similar effects.106 This was the case even in the HP, in contrast with previous findings
in hippocampal place cells recorded in over-trained animals performing a spatial
navigation task.76,79,80 At present it is unclear whether this discrepancy reflects dif-
ferences in stimulus familiarity (novel versus habitual), stimulation modality (tactile
exploration versus spatial navigation), the very low representation of place cells in
our hippocampal samples (<5%), or possible analytical artifacts in previous studies
based on the statistical boot-strapping of relatively small datasets.76,79,80
In the face of consistently low neuronal ensemble correlation values under a broad
range of putative processing speeds, we conclude that for most of the rat forebrain
the high-fidelity replay hypothesis should be rejected, at least for time periods longer
than 5 msec. This is at variance with the compelling example of accurate neuronal
replay found in the motor nucleus RA of zebra finches84 (Figure 12.1 B). Such
remarkable case of high-fidelity replay likely derives from the very specialized neural
processing carried out by the nucleus RA,116–119 namely the descending motor gen-
eration of a single, unique, and invariant sequential object: the bird’s own song.
In contrast our experiment involved the free exploration of four novel and
complex objects placed in four well-separated places and including the presence of
novel food. Exposure to these objects is not at all stimulus-invariant because the
animals recursively explore the different objects at different angles and moments.
Compared to the performance of birdsong, our paradigm is much more representative
of the sensory stimulation encountered by higher vertebrates in their routine inter-
actions with the natural environment. It involves a plethora of tactile, gustatory,
olfactory, spatial, and motor cues, likely producing distributed and partially over-

lapping memory traces that may directly account for the fact that neuronal replay
occurs at low fidelity. An equivalent argument can be made for the performance of
spatial and sensory-motor tasks previously used in rats74–76,79–81 and nonhuman pri-
mates,82 suggesting that the low-fidelity replay of neuronal firing patterns is the rule,
not the exception, in the brains of higher vertebrates.
A MODEL FOR THE COMPLEMENTARY ROLES OF SW

AND REM SLEEP IN MEMORY CONSOLIDATION
The results discussed in this chapter indicate that novelty-induced neuronal rever-
beration consists of a low-fidelity recurrence of previously salient neuronal firing
patterns replayed at physiological speeds. Such recurrence is sustained and long
lasting, occurs in multiple forebrain areas in a context-dependent manner, and peaks
during SW sleep in inverse correlation with firing rates. Neuronal reverberation is
sustained for long epochs during SW sleep, which suggests that unconsolidated
synaptic changes may not only be recalled but also amplified over time during SW
sleep. Indeed a progressive increase of neuronal ensemble correlations across single
SW sleep episodes was often observed (Figure 12.4 B, white arrows). We have
proposed in this regard106 that intrinsic brain activation during SW sleep, free of
sensory interference, would be biased toward previously potentiated synapses; hence
neuronal firing patterns originally produced during novel WK experience would
reverberate during SW sleep significantly above chance levels. The periodic activa-
tion of calcium-dependent second-messenger cascades120–124 by the large-amplitude
SW oscillations125,126 would then result in the progressive amplification of the syn-
aptic changes that encode the novel memory trace.106
Our view resonates to a degree with a previous suggestion that the neuronal
reverberation of newly acquired synaptic changes during SW sleep may lead to the
recall and storage of new memories by way of “calcium-mediated intracellular
cascades” capable of opening the “molecular gates to plasticity.”127 This hypothesis
is partially contradicted by evidence that calcium-dependent gene expression related
to synaptic plasticity is shut down during SW sleep.85,86,128 Based on the current
literature, we have proposed instead106 that SW and REM sleep play distinct and
complementary roles on memory consolidation: Initial neuronal reverberation
depends mainly on SW sleep episodes,71,74,106 but transcriptional events able to effect
durable neuronal plasticity129 are only triggered during ensuing REM sleep.85,86
This functional dissociation with regard to memory consolidation implies that
the two main sleep phases separately satisfy Hebb’s two learning postulates,70 with
memory recall occurring during SW sleep and memory storage taking place during
REM sleep (Figure 12.5). According to this view, the deleterious effects of sleep
deprivation on memory consolidation would be a consequence of the disruption of
the underlying neuronal reverberation and gene expression during SW and REM
sleep, respectively. Such a scheme fulfills earlier conceptual notions of a two-step
process for memory consolidation during sleep130–132 and is in line with evidence
that SW and REM sleep have synergistic effects on human procedural learning.26,133
The postsynaptic nature of the zif-268 response,98,134 its putative consequences
on synaptic strengthening,101,135 and the hippocampofugal pattern of zif-268 expres-

Labile Consolidated
WK SWS REM WK SWS REM WK
*
Memory Trace Strength
*
GE GE GE
Encoding Time
FIGURE 12.5 We have proposed106 that SW and REM sleep play distinct and complementary
roles in the processing of new memory traces, with memory recall (reverberation) occurring
during SW sleep and memory storage (plasticity-related gene expression, GE) taking place
during REM sleep. Such functional dissociation implies that memory processing progresses
in cycles of pretranscriptional amplification of labile traces during SW sleep and transcrip-
tional trace consolidation triggered by REM sleep. According to this scheme, the combined
action of SW and REM sleep would cause a progressive increase in the strength and consol-
idation level of memories produced over several hours via plasticity-related protein synthesis.
The model also predicts that after some time in WK, such effects would be completed and
memory strength would then start to decrease, due to sensory interference (indicated by
asterisks). The recurrence of sleep would therefore prevent sensory interference from further
degrading the strength of recently acquired memory traces.
sion during REM sleep86 led us to propose that the cyclical reiteration of trace
amplification during SW sleep and trace storage during REM sleep promotes a
postsynaptic propagation of memory traces.106,136 Potentially this propagation could
cause memory traces to progressively reach farther and farther away from the original
synaptic trajectory activated at initial encoding. Over time this sleep-dependent
propagation could lead to deeper memory encoding within the cerebral cortex,137–139
cumulative learning after memory trace acquisition,133 and progressive hippocampal
disengagement.54,57,59–61,63,65,66,140 In conclusion sustained neuronal reverberation dur-
ing SW sleep, immediately followed by plasticity-related gene expression during
REM sleep, may be sufficient to explain the beneficial role of sleep on the consol-
idation of new memories.
ACKNOWLEDGMENTS
We thank Jonathan Winson, Robert Stickgold, Ivan de Araújo, and David Schwartz
for fruitful discussions of the views exposed here, and Susan Halkiotis for secretarial

help. This work was supported by a fellowship from the Pew Latin American
Program in Biomedical Sciences to SR, by an INSERM fellowship to DG, and by
NIH 5 R01, DE11451, and 5 R01 DE13810 grants to MALN.
REFERENCES
1. Aserinsky, E. and Kleitman, N., Regularly occurring periods of eye motility and
concomitant phenomena during sleep, Science, 118, 273–274, 1953.
2. Dement, W.C. and Kleitman, N., The relation of eye movements during sleep to
dream activity: An objective method for the study of dreaming, J. Exp. Psychol., 53,
339–346, 1957.
3. Dement, W. and Kleitman, N., Cyclic variations in EEG during sleep and their relation
to eye movements, body motility, and dreaming, Electroencephalogr. Clin. Neuro-
physiol. Suppl., 9 (4), 673–690, 1957.
4. Jouvet, M., Michel, F., and Courjon, J., Sur un stade d’activité électrique cérébrale
rapide au cours du sommeil physiologique, C.R. Soc. Biol. (Paris), 153, 1024–1028,
1959.
5. Dement, W.C., The occurrence of low voltage, fast, electroencephalogram patterns
during behavioral sleep in the cat, Electroenceph. Clin. Neurophysiol., 10, 291–296,
1958.
6. Roffwarg, H. P., Dement, W. C., Muzio, J. N., and Fisher, C., Dream imagery:
relationship to rapid eye movements of sleep, Arch. Gen. Psychiatry, 7, 235–258,
1962.
7. Tradardi, V., Sleep in the pigeon, Arch. Ital. Biol., 104, 516–521, 1966.
8. Jouvet, M., The states of sleep, Sci. Am. 216, 62–68, 1967.
9. Rechtschaffen, A. and Kales, A., Report No. 204, 1968.
10. Jenkins, J. B. and Dallenbach, K M., Oblivescence during sleep and waking, Am. J.
Psychology, 35, 605–612, 1924.
11. Pearlman, C. A., Effect of rapid eye movement (dreaming) sleep deprivation on
retention of avoidance learning in rats, Report US Naval Submarine Medical Center,
22, 1–4, 1969.
12. Leconte, P. and Bloch, V., Déficit de la rétention d’un conditionnement après privation
de sommeil paradoxal chez le rat, Comptes Rendus de l’Académie des Sciences
(Paris), 271D, 226–229, 1970.
13. Fishbein, W., Disruptive effects of rapid eye movement sleep deprivation on long-
term memory, Physiol. Behav., 6, 279–282, 1971.
14. Pearlman, C., REM sleep deprivation impairs latent extinction in rats, Physiol. Behav.,
11, 233–237, 1973.
15. Pearlman, C. and Becker, M., REM sleep deprivation impairs bar-press acquisition
in rats, Physiol. Behav, 13, 813–817, 1974.
16. Linden, E. R., Bern, D., and Fishbein, W., Retrograde amnesia: prolonging the fixation
phase of memory consolidation by paradoxical sleep deprivation, Physiol. Behav.,
14, 409–412, 1975.
17. Shiromani, P., Gutwein, B. M., and Fishbein, W., Development of learning and
memory in mice after brief paradoxical sleep deprivation, Physiol. Behav., 22,
971–978, 1979.
18. Smith, C. and Butler, S., Paradoxical sleep at selective times following training is
necessary for learning, Physiol. Behav., 29, 469–473, 1982.

19. Smith, C. and Kelly, G., Paradoxical sleep deprivation applied two days after end of
training retards learning, Physiol. Behav., 43, 213–216, 1988.
20. Smith, C. and MacNeill, C., A paradoxical sleep-dependent window for memory
53–56 H after the end of avoidance training, Psychobiology, 21, 109–112, 1993.
21. Karni, A., Tanne, D., Rubenstein, B.S., Askenasy, J.J., and Sagi, D., Dependence on
REM sleep of overnight improvement of a perceptual skill, Science, 265, 679–682,
1994.
22. Smith, C. and Rose, G.M., Evidence for a paradoxical sleep window for place learning
in the Morris water maze, Physiol. Behav., 59, 93–97, 1996.
23. Stickgold, R., James, L., and Hobson, J.A., Visual discrimination learning requires
sleep after training, Nat. Neurosci., 3, 1237–1238, 2000.
24. Walker, M.P., Brakefield, T., Morgan, A., Hobson, J.A., and Stickgold, R., Practice
with sleep makes perfect: sleep-dependent motor skill learning, Neuron, 35, 205–211,
2002.
25. Maquet, P., Schwartz, S., Passingham, R., and Frith, C., Sleep-related consolidation
of a visuomotor skill: brain mechanisms as assessed by functional magnetic resonance
imaging, J. Neurosci., 23, 1432–1440, 2003.
26. Mednick, S.C., Nakayama, K., and Stickgold, R., Sleep-dependent learning: a nap is
as good as a night, Nat. Neurosci., 6, 697–698, 2003.
27. Wetzel, W., Wagner, T., and Balschun, D., REM sleep enhancement induced by
different procedures improves memory retention in rats, Eur. J. Neuro., 18,
2611–2617, 2003.
28. Lucero, M.A., Lengthening of REM sleep duration consecutive to learning in the rat,
Brain Res., 20, 319–322, 1970.
29. Leconte, P. and Hennevin, E., Augmentation de la durée de sommeil paradoxal
consécutive à un apprentissage chez le rat, C.R. Acad. Sc. (Paris), 273, 86–88, 1971.
30. Fishbein, W., Kastaniotis, C., and Chattman, D., Paradoxical sleep: prolonged aug-
mentation following learning, Brain Res., 79, 61–75, 1974.
31. Smith, C., Kitahama, K., Valatx, J.L., and Jouvet, M., Increased paradoxical sleep in
mice during acquisition of a shock avoidance task, Brain Res., 77, 221–230, 1974.
32. Smith, C., Young, J., and Young, W., Prolonged increases in paradoxical sleep during
and after avoidance-task acquisition, Sleep, 3, 67–68, 1980.
33. Smith, C. and Lapp, L., Prolonged increases in both PS and number of REMS
following a shuttle avoidance task, Physiol. Behav., 36, 1053–1057, 1986.
34. Smith, C. and Wong, P.T.P., Paradoxical sleep increases predict successful learning
in a complex operant task, Behav. Neurosci., 105, 282–288, 1991.
35. Smith, C. and Lapp, L., Increases in number of REMS and REM density in humans
following an intensive learning period, Sleep, 14, 325–330, 1991.
36. Adey, W.R., Dunlop, C.W., and Hendrix, C.E., Hippocampal slow waves: distribution
and phase relationship in the course of approach learning, Arch. Neurol., 3, 74–90,
1960.
37. Elazar, Z. and Adey, W.R., Electroencephalographic correlates of learning in subcor-
tical and cortical structures, Electroencephalogr. Clin. Neurophysiol., 23, 306–319,
1967.
38. Landfield, P.W., McGaugh, J.L., and Tusa, R.J., Theta rythm: a temporal correlate of
memory storage processes in the rat, Science, 87–89, 1972.
39. Bennett, T.L., The effects of centrally blocking hippocampal theta activity on learning
and retention, Behav. Biol., 9, 541–552, 1973.
40. Bennett, T.L., Nunn Hebert, P., and Moss, D.E., Hippocampal theta activity and the
attention component of discrimination learning, Behav. Biol., 8, 173–181, 1973.

41. Winson, J., Loss of hippocampal theta rhythm results in spatial memory deficit in
the rat, Science, 201, 160–163, 1978.
42. Sederberg, P.B., Kahana, M.J., Howard, M.W., Donner, E.J., and Madsen, J.R., Theta
and gamma oscillations during encoding predict subsequent recall, J. Neurosci., 23,
10809–10814, 2003.
43. Green, J.D. and Arduini, A.A., Hippocampal electrical activity in arousal, J. Neuro-
physiol., 17, 533–557, 1954.
44. Brown, B. B., Frequency and phase of hippocampal theta activity in the spontaneously
behaving cat, Electroencephalogr. Clin. Neurophysiol., 24, 53–62, 1968.
45. Sainsbury, R.S., Hippocampal activity during natural behavior in the guinea pig,
Physiol. Behav., 5, 317–324, 1970.
46. Harper, R.M., Frequency changes in hippocampal electrical activity during movement
and tonic immobility, Physiol. Behav., 7, 55–58, 1971.
47. Arnolds, D.E., Lopes da Silva, F.H., Aitink, J. W., Kamp, A., and Boeijinga, P., The
spectral properties of hippocampal EEG related to behaviour in man, Electroenceph-
alogr. Clin. Neurophysiol., 50, 324–328, 1980.
48. Stewart, M. and Fox, S.E., Hippocampal theta activity in monkeys, Brain Res., 538,
59–63, 1991.
49. Kahana, M.J., Sekuler, R., Caplan, J.B., Kirschen, M., and Madsen, J.R., Human
theta oscillations exhibit task dependence during virtual maze navigation, Nature,
399, 781–784, 1999.
50. Vanderwolf, C.H., Hippocampal electrical activity and voluntary movement in the
rat, Electroenceph. Clin. Neurophysiol., 26, 407–418, 1969.
51. Timo-Iaria, C., Negrao, N., Schmidek, W.R., Hoshino, K., Lobato de Menezes, C.E.,
and Leme da Rocha, T., Phases and states of sleep in the rat, Physiol. Behav., 5,
1057–1062, 1970.
52. Winson, J., Patterns of hippocampal theta rhythm in the freely moving rat, Electro-
enceph. Clin. Neurophysiol., 36, 291–301, 1974.
53. Cantero, J.L., Atienza, M., Stickgold, R., Kahana, M.J., Madsen, J.R., and Kocsis,
B., Sleep-dependent theta oscillations in the human hippocampus and neocortex., J.
Neurosci., 23, 10897–10903, 2003.
54. Scoville, W.B. and Milner, B., Loss of recent memory after bilateral hippocampal
lesions., J Neurol. Neurosurg. Psych., 20, 11–21, 1957.
55. Mishkin, M., Memory in monkeys severely impaired by combined but not by separate
removal of amygdala and hippocampus, Nature, 273, 297–298, 1978.
56. Kesner, R.P. and Novak, J.M., Serial position curve in rats: role of the dorsal hippo-
campus, Science, 218, 173–175, 1982.
57. Zola-Morgan, S.M. and Squire, L.R., The primate hippocampal formation: evidence
for a time-limited role in memory storage, Science, 250, 288–290, 1990.
58. Buzsaki, G., Chen, L.S., and Gage, F.H., Spatial organization of physiological activity
in the hippocampal region: relevance to memory formation, Prog. Brain Res., 83,
257–268, 1990.
59. Squire, L.R., Memory and the hippocampus: a synthesis from findings with rats,
monkeys, and humans, Psychol. Rev., 99, 195–231, 1992.
60. Kim, J.J., Clark, R.E., and Thompson, R.F., Hippocampectomy impairs the memory
of recently, but not remotely, acquired trace eyeblink conditioned responses, Behav.
Neurosci., 109, 195–203, 1995.
61. Izquierdo, I. and Medina, J.H., Memory formation: the sequence of biochemical
events in the hippocampus and its connection to activity in other brain structures,
Neurobiol. Learn. Mem., 68, 285–316, 1997.

62. Corkin, S., Amaral, D.G., Gonzalez, R.G., Johnson, K.A., and Hyman, B.T., H.M.’s
medial temporal lobe lesion: findings from magnetic resonance imaging, J. Neurosci.,
17, 3964–3979, 1997.
63. Bontempi, B., Laurent-Demir, C., Destrade, C., and Jaffard, R., Time-dependent
reorganization of brain circuitry underlying long-term memory storage, Nature, 400,
671–675, 1999.
64. Lavenex, P. and Amaral, D.G., Hippocampal-neocortical interaction: a hierarchy of
associativity, Hippocampus, 10, 420–430, 2000.
65. Haist, F., Bowden Gore, J., and Mao, H., Consolidation of human memory over
decades revealed by functional magnetic resonance imaging, Nat. Neurosci., 4,
1139–1145, 2001.
66. Winocur, G., McDonald, R.M., and Moscovitch, M., Anterograde and retrograde
amnesia in rats with large hippocampal lesions, Hippocampus, 11, 18–26, 2001.
67. Bryson, D. and Schacher, S., Behavioral analysis of mammalian sleep and learning,
Perspectives in Biology & Medicine, 13, 71–79, 1969.
68. Winson, J., Interspecies differences in the occurrence of theta., Behavioral Biology,
7, 479–487, 1972.
69. Winson, J., Brain and Psyche, Anchor Press, New York, 1985.
70. Hebb, D.O., The Organization of Behavior: A Neuropsychological Theory, John Wiley
& Sons, New York, 1949.
71. Pavlides, C. and Winson, J., Influences of hippocampal place cell firing in the awake
state on the activity of these cells during subsequent sleep episodes, J. Neurosci., 9,
2907–2918, 1989.
72. O’Keefe, J., Place units in the hippocampus of the freely moving rat., Exp. Neurol.,
51, 78–109, 1976.
73. O’Keefe, J., Hippocampus, theta, and spatial memory. [Review] [45 refs], Curr. Op.
Neurobiol., 3, 917–924, 1993.
ories during sleep, Science, 265, 676–679, 1994.
75. Skaggs, W.E. and McNaughton, B.L., Replay of neuronal firing sequences in rat hip-
pocampus during sleep following spatial experience, Science, 271, 1870–1873, 1996.
76. Nadasdy, Z., Hirase, H., Czurko, A., Csicsvari, J., and Buzsaki, G., Replay and time
compression of recurring spike sequences in the hippocampus, J. Neurosci., 19,
9497–9507, 1999.
77. Poe, G.R., Nitz, D.A., McNaughton, B.L., and Barnes, C.A., Experience-dependent
phase-reversal of hippocampal neuron firing during REM sleep, Brain Res., 855,
176–180, 2000.
78. Hirase, H., Leinekugel, X., Czurko, A., Csicsvari, J., and Buzsaki, G., Firing rates
of hippocampal neurons are preserved during subsequent sleep episodes and modified
by novel awake experience, Proc. Natl. Acad. Sci. USA, 98, 9386–9390, 2001.
79. Louie, K. and Wilson, M.A., Temporally structured replay of awake hippocampal
ensemble activity during rapid eye movement sleep, Neuron, 29, 145–156, 2001.
80. Lee, A.K. and Wilson, M.A., Memory of sequential experience in the hippocampus
during slow-wave sleep, Neuron, 36, 1183–1194, 2002.
81. Qin, Y.L., McNaughton, B.L., Skaggs, W.E., and Barnes, C.A., Memory reprocessing
in corticocortical and hippocampocortical neuronal ensembles, Philos. Trans. R. Soc.
Lond. B. Biol. Sci., 352, 1525–1533, 1997.
82. Hoffman, K.L. and McNaughton, B., Coordinated reactivation of distributed memory
traces in primate neocortex, Science, 297, 2070–2073, 2002.

83. Maquet, P., Laureys, S., Peigneux, P., Fuchs, S., Petiau, C., Phillips, C., Aerts, J., Del
Fiore, G., Degueldre, C., Meulemans, T., Luxen, A., Franck, G., Van Der Linden, M.,
Smith, C., and Cleeremans, A., Experience-dependent changes in cerebral activation
during human REM sleep, Nat. Neurosci., 3, 831–836, 2000.
84. Dave, A.S. and Margoliash, D., Song replay during sleep and computational rules
for sensorimotor vocal learning, Science, 290, 812–816, 2000.
85. Ribeiro, S., Goyal, V., Mello, C.V., and Pavlides, C., Brain gene expression during
REM sleep depends on prior waking experience, Learn. Mem., 6, 500–508, 1999.
86. Ribeiro, S., Mello, C.V., Velho, T., Gardner, T.J., Jarvis, E.D., and Pavlides, C.,
Induction of hippocampal long-term potentiation during waking leads to increased
extrahippocampal zif-268 expression during ensuing rapid-eye-movement sleep,
J. Neurosci., 22, 10914–10923, 2002.
87. Agranoff, B.W., Davis, R.E., and Brink, J.J., Chemical studies on memory fixation
in goldfish., Brain. Res., 1, 303–309, 1966.
88. Milbrandt, J., A nerve growth factor-induced gene encodes a possible transcriptional
regulatory factor, Science, 238, 797–799, 1987.
89. Sukhatme, V.P., Cao, X.M., Chang, L.C., Tsai-Morris, C.H., Stamenkovich, D., Fer-
reira, P.C., Cohen, D.R., Edwards, S.A., Shows, T.B., Curran, T. et al., A zinc finger-
encoding gene coregulated with c-fos during growth and differentiation, and after
cellular depolarization, Cell, 53, 37–43, 1988.
90. Christy, B. and Nathans, D., DNA binding site of the growth factor-inducible protein
Zif268, Proc Natl Acad Sci USA, 86, 8737–8741, 1989.
91. Morgan, J.I. and Curran, T., Stimulus-transcription coupling in neurons: role of
cellular immediate-early genes., Trends Neurosci., 12, 459–462, 1989.
92. Cole, A.J., Saffen, D.W., Baraban, J.M., and Worley, P.F., Rapid increase of an
immediate early gene messenger RNA in hippocampal neurons by synaptic NMDA
receptor activation, Nature, 340, 474–476, 1989.
93. Wisden, W., Errington, M.L., Williams, S., Dunnett, S.B., Waters, C., Hitchcock, D.,
Evan, G., Bliss, T.V., and Hunt, S.P., Differential expression of immediate early genes
in the hippocampus and spinal cord, Neuron, 4, 603–914, 1990.
94. Madison, D.V., Malenka, R.C., and Nicoll, R.A., Mechanisms underlying long-term
potentiation of synaptic transmission, Annu. Rev. Neurosci., 14, 379–397, 1991.
95. Nikolaev, E., Kaminska, B., Tischmeyer, W., Matthies, H., and Kaczmarek, L., Induc-
tion of expression of genes encoding transcription factors in the rat brain elicited by
behavioral training, Brain Res. Bull., 28, 479–384, 1992.
96. Mello, C.V., Vicario, D.S., and Clayton, D.F., Song presentation induces gene expres-
sion in the songbird forebrain, Proc. Natl. Acad. Sci. USA, 89, 6818–6822, 1992.
97. Bliss, T.V. and Collingridge, G.L., A synaptic model of memory: long-term potenti-
ation in the hippocampus, Nature, 361, 31–39, 1993.
98. Thiel, G., Schoch, S., and Petersohn, D., Regulation of synapsin I gene expression
by the zinc finger transcription factor zif268/egr-1, J. Biol. Chem., 269, 15294–15301,
1994.
99. Petersohn, D., Schoch, S., Brinkmann, D.R., and Thiel, G., The human synapsin II
gene promoter. Possible role for the transcription factor zif268/egr-1, polyoma
enhancer activator 3, and AP2, J. Biol. Chem., 270, 24361–24369, 1995.
100. Wallace, C.S., Withers, G.S., Weiler, I.J., George, J.M., Clayton, D.F., and Greenough,
W. T., Correspondence between sites of NGFI-A induction and sites of morphological
plasticity following exposure to environmental complexity, Brain Res. Mol. Brain
Res., 32, 211–220, 1995.

101. Jones, M.W., Errington, M.L., French, P.J., Fine, A., Bliss, T.V., Garel, S., Charnay,
P., Bozon, B., Laroche, S., and Davis, S., A requirement for the immediate early gene
Zif268 in the expression of late LTP and long-term memories, Nat. Neurosci., 4,
289–296, 2001.
102. Gerrard, J.L., Reactivation of Hippocampal Ensemble Activity Patterns in the Aging
Rat: Insights into Memory Consolidation within the Aged Brain, University of Ari-
zona, 2002.
103. Peigneux, P., Laureys, S., Fuchs, S., Destrebecqz, A., Collette, F., Delbeuck, X.,
Phillips, C., Aerts, J., Del Fiore, G., Degueldre, C., Luxen, A., Cleeremans, A., and
Maquet, P., Learned material content and acquisition level modulate cerebral reacti-
vation during posttraining rapid-eye-movements sleep., Neuroimage, 20, 125–134,
2003.
104. Datta, S., Avoidance task training potentiates phasic pontine-wave density in the rat:
A mechanism for sleep-dependent plasticity, J. Neurosci., 20, 8607–8613, 2000.
105. Kudrimoti, H.S., Barnes, C.A., and McNaughton, B.L., Reactivation of hippocampal
cell assemblies: effects of behavioral state, experience, and EEG dynamics, J. Neu-
rosci., 19, 4090–4101, 1999.
106. Ribeiro, S., Gervasoni, D., Soares, E.S., Zhou, Y., Lin, S.C., Pantoja, P., Lavine, M.,
and Nicolelis, M., Long-lasting novelty-induced neuronal reverberation during slow-
wave sleep in multiple forebrain areas, PLoS Biology, 2, 126–137, 2004.
107. Nowak, R.M., Walker’s Mammals of the World, 6th ed., Johns Hopkins University
Press, Baltimore, MD, 1999.
108. Simons, D.J., Response properties of vibrissa units in rat SI somatosensory neocortex,
J. Neurophysiol., 41, 798–820, 1978.
109. Ghazanfar, A.A., Stambaugh, C.R., and Nicolelis, M.A., Encoding of tactile stimulus
location by somatosensory thalamocortical ensembles, J. Neurosci., 20, 3761–3775,
2000.
110. Dieckmann, G. and Hasser, R., Stimulation experiments on the function of putamen
in the cat, J. Hirnforsch., 10, 187–225, 1968.
111. Jog, M.S., Kubota, Y., Connolly, C.I., Hillegaart, V., and Graybiel, A.M., Building
neural representations of habits, Science, 286, 1745–1749, 1999.
112. Fenn, K.M., Nusbaum, H.C., and Margoliash, D., Consolidation during sleep of
perceptual learning of spoken language, Nature, 425, 614–616, 2003.
113. Kohler, W., Gestalt Psychology: An Introduction to New Concepts in Modern Psy-
chology, Reissue ed., Liveright, New York, 1947.
114. Wagner, U., Gais, S., Haider, H., Verleger, R., and Born, J., Sleep inspires insight,
Nature, 427, 352–355, 2004.
115. Melton, A.W. and Irwin, J.M., The influence of degree of interpolated learning on
retroactive inhibition and the overt transfer of specific responses., Am. J. Psychol.,
53, 173–203, 1940.
116. Nottebohm, F., Stokes, T.M., and Leonard, C.M., Central control of song in the canary,
Serinus canarius, J. Comp. Neurol., 165, 457–486, 1976.
117. Doupe, A.J. and Konishi, M., Song-selective auditory circuits in the vocal control
system of the zebra finch, Proc. Natl. Acad. Sci. USA, 88, 11339–11343, 1991.
118. Vicario, D.S. and Yohay, K.H., Song-selective auditory input to a forebrain vocal
control nucleus in the zebra finch, J. Neurobiol. 24, 488–505, 1993.
119. Margoliash, D., Functional organization of forebrain pathways for song production
and perception, J. Neurobiol., 33, 671–693, 1997.
120. Dingledine, R., N-methyl-D-aspartate activates voltage-dependent calcium conduc-
tance in rat hippocampal pyramidal cells, J. Physiol., 343, 385–405, 1983.

121. MacDermott, A.B., Meyer, M. L., Westbrook, G. L., Smith, J., and Baker, J. L.,
NMDA-receptor activation increases cytoplasmic calcium concentration in cultured
spinal cord neurones, Nature, 321, 519–522, 1986.
122. Fields, R.D., Yu, C., and Nelson, P.G., Calcium, network activity, and the role of
NMDA channels in synaptic plasticity invitro, J. Neurosci., 11, 134–146, 1991.
123. Collingridge, G.L., Randall, A.D., Davies, C.H., and Alford, S., The synaptic activa-
tion of NMDA receptors and Ca2+ signalling in neurons, Ciba Found. Symp., 164,
162–171, 1992.
124. Augustine, G. J. and Neher, E., Neuronal Ca2+ signalling takes the local route, Curr.
Opin. Neurobiol., 2, 302–307, 1992.
125. Steriade, M. and McCarley, R. W., Brainstem Control of Wakefulness and Sleep,
Plenum Press, New York, 1990.
126. Steriade, M., McCormick, D. A., and Sejnowski, T. J., Thalamocortical oscillations
in the sleeping and aroused brain, Science, 262, 679–685, 1993.
127. Sejnowski, T. J. and Destexhe, A., Why do we sleep?, Brain Res., 886, 208–223, 2000.
128. Pompeiano, M., Cirelli, C., and Tononi, G., Immediate-early genes in spontaneous
wakefulness and sleep: expression of c-fos and NGIF-A mRNA protein, J., Sleep
Res., 3, 80–96, 1994.
129. Gutwein, B.M., Shiromani, P.J., and Fishbein, W., Paradoxical sleep and memory:
long-term disruptive effects of Anisomycin, Pharmacol. Biochem. Behav., 12,
377–384, 1980.
130. Giuditta, A., A sequential hypothesis for the function of sleep, in Sleep in ’84, Koella,
W.P., Ruther, E., and Schultz, H., Eds., Fisher-Verlag, Stuttgart, 1985, pp. 222–224.
131. Giuditta, A., Ambrosini, M.V., Montagnese, P., Mandile, P., Cotugno, M., Zucconi,
G.G., and Vescia, S., The sequential hypothesis of the function of sleep, Behav. Brain
Res., 69, 157–166, 1995.
132. Stickgold, R., Sleep: off-line memory reprocessing, Trends Cogn. Sci., 2, 484–492, 1998.
133. Stickgold, R., Whidbee, D., Schirmer, B., Patel, V., and Hobson, J. A., Visual dis-
crimination task improvement: A multi-step process occurring during sleep, J. Cogn.
Neurosci., 12, 246–254, 2000.
134. Lemaire, P., Vesque, C., Schmitt, J., Stunnenberg, H., Frank, R., and Charnay, P., The
serum-inducible mouse gene Krox-24 encodes a sequence-specific transcriptional
activator, Mol. Cell Biol., 10, 3456–3467, 1990.
135. Bozon, B., Kelly, A., Josselyn, S.A., Silva, A.J., Davis, S., and Laroche, S., MAPK,
CREB, and zif268 are all required for the consolidation of recognition memory,
Philos. Trans. R. Soc. Lond. B. Biol. Sci., 358, 805–814, 2003.
136. Pavlides, C. and Ribeiro, S., Recent evidence of memory processing in sleep, in Sleep
and Brain Plasticity, Maquet, P., Smith, C., and Stickgold, R., Eds., Oxford University
Press, Oxford, 2003, pp. 327–362.
137. Hebb, D.O., The effects of early and late brain injury upon test scores, and the nature
of normal adult intelligence, Proc. Am. Philosophical Soc., 85, 275–292, 1942.
138. Craik, F. and Lockhart, R., Levels of processing: A framework for memory research.,
J. Verb. Learn. Verb. Behav., 11, 671–684, 1972.
139. Cermak, L. and Craik, F., Levels of Processing in Human Memory, John Wiley &
Sons, New York, 1979.
140. Frankland, P.W., O’Brien, C., Ohno, M., Kirkwood, A., and Silva, A.J., Alpha-
CaMKII-dependent plasticity in the cortex is required for permanent memory, Nature,
411, 309–313, 2001.
141. Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Compact 3rd
ed., Academic Press, San Diego, 1997.

13 Cerebral Functional
Segregation and
Integration during
Human Sleep
Pierre Maquet, Fabien Perrin, Steven Laureys,
Tahn Dang-Vu, Martin Desseilles, Mélanie Boly,
and Philippe Peigneux
CONTENTS
Introduction
Two Sleep Types, Two Different Distributions of Regional Brain Activity
NREM Sleep
REM Sleep
Brain Responses to External Stimuli during Sleep
Brain Responses to Internal Stimuli during Sleep: Does the PGO Activity
Exist in Humans?
Experience-Dependent Changes in Functional Connectivity during
Post-Training Sleep
Conclusion
Acknowledgments
References
INTRODUCTION
A comprehensive understanding of human brain function requires the characteriza-
tion of both cerebral segregation and connectivity.1 Functional segregation pertains
to the involvement of certain cerebral areas and networks in specific cerebral func-
tions. For instance, Broca’s and Wernicke’s areas are known to participate in lan-
guage. On the other hand, functional integration reflects how different regions
interact to mediate a specific function. At the level of macroscopic systems, func-
tional neuroimaging using positron emission tomography (PET) or functional mag-
netic resonance imaging (fMRI) can probe in vivo the segregation and integration
of the human brain function during sleep and wakefulness.
0-8493-1519-0/05/$0.00+$1.50

Early studies described the functional anatomy of normal human sleep. They
showed that the distribution of brain activity was specific for each type of sleep and
differed from the waking pattern of brain activity. While the activity of subcortical
structures was easily explained by the mechanisms that generate rapid eye movement
(REM) sleep and non-REM (NREM) sleep in animals, the distribution of the activity
within the cortex was harder to explain and its origin remains speculative.
In order to better understand how cortical function is organized during sleep,
regional cerebral responses have been explored in three different situations:
1. In response to external auditory stimulations in NREM sleep

2. In response to the internal activation due to putative PGO activity in
humans during REM sleep
3. In relation to previous waking experience
This chapter reviews these three issues after a short account of the functional
neuroanatomy of NREM and REM sleep.
TWO SLEEP TYPES, TWO DIFFERENT DISTRIBUTIONS

OF REGIONAL BRAIN ACTIVITY
NREM SLEEP
In mammals the neuronal activity observed during NREM sleep oscillations (spin-
dles and slow rhythms) is characterized by bursting patterns that alternate short
bursts of firing with long periods of hyperpolarization.2 The latter have a major
impact on the regional blood flow, which on the average decreases in the areas where
these oscillations are expressed. Accordingly the average cerebral metabolism and
blood flow begin to decrease in light (stage 2) NREM sleep,3,4 and their nadir is
observed in deep (stage 3 and 4) NREM sleep or slow-wave sleep (SWS).5,6
The cascade of events that underpin the NREM sleep oscillations in the thalamo-
neocortical networks is conditional upon a decreased firing in the activating struc-
tures of the brainstem tegmentum. In humans the brainstem blood flow is decreased
during light NREM sleep7 as during SWS.7–9 In light NREM sleep, the pontine
tegmentum is specifically deactivated, whereas the mesencephalon seems to retain
an activity that is not significantly different from wakefulness.7 In SWS both pontine
and mesencephalic tegmenta are deactivated.
The thalamus plays a central role in the generation of NREM sleep rhythms due
to the intrinsic properties of its neurones and to the intrathalamic and thalamo-
cortico-thalamic connectivity. In humans the thalamus is deactivated during both
light and deep NREM sleep7–9 in proportion to the power density in the spindle and
delta frequency range,10 respectively.
The role of the cortex in the generation of NREM sleep oscillations is equally
important and begins to be better understood,2 but the respective contribution of the
different parts of the neocortex in NREM sleep rhythms generation is still unknown
at the microscopic level. In humans the deactivation of the cortex is not homogeneous.
The most deactivated areas are located in associative cortices of the frontal,

FIGURE 13.1 (See color insert following page 108.) Functional neuroimaging of REM
sleep. Schematic representation of the relative increases and decreases in neural activity
associated with REM sleep. Left panel: lateral view; middle panel: ventral view; right panel:
medial view. A, H = amygdala and hypothalamus; B = basal forebrain; Ca = anterior cingulate
gyrus; Cp = posterior cingulate gyrus and precuneus; F = prefrontal cortex; M = motor cortex;
P = parietal supramarginal cortex; PH = parahippocampic gyrus; PT = pontine tegmentum;
O = occipital-lateral cortex; Th = thalamus; T-O = temporo-occipital extrastriate cor-
tex.(Adapted from Schwartz, S. and Maquet, P., Sleep imaging and the neuro-psychological
assessment of dreams, Trends Cogn. Sci., 6, 23, 2002.)
parietal — And to a lesser extent temporal and insular lobes7–9,11 — while the primary
cortices are the least deactivated. This observation suggests that the first cortical relay
areas for exteroceptive stimuli remain relatively active during SWS. Although attrac-
tive, this hypothesis is challenged by another interpretation of the data. It should be
emphasized that polymodal association cortices are the most active cerebral areas
during wakefulness. Because of this high waking activity, they might be more pro-
foundly influenced by SWS rhythms than primary cortices.12 This suggestion supports
the view that sleep intensity is targeted disproportionately to areas of the brain
intensely used during prior waking.13 Accordingly, in cats involved for some time in
an active visual task, neurones in the associative visual cortex can adopt a bursting
pattern typical for the sleeping cortex and become less responsive to visual stimu-
lation, while the primary visual areas maintain a normal response to visual inputs.14
REM SLEEP
In mammals neuronal populations in the mesopontine tegmentum are the source of
a major activating input to the thalamic nuclei during REM sleep.15 The thalamus
forwards this activation to the entire forebrain. In humans the activation of mesopon-
tine tegmentum and thalamic nuclei has been systematically reported during REM
sleep8,16,17 (Figure 13.1). In the forebrain PET data showed that limbic and paralimbic
areas (amygdala, hippocampal formation, anterior cingulate, orbito-frontal, and insu-
lar cortices) were among the most active areas in REM sleep. Temporal and occipital
cortices were also shown to be very active,8 although this result is less reproducible.16
The functional integration is modified during human REM sleep. The functional
relationship between striate and extrastriate cortices, usually excitatory, is inverted
during REM sleep.18 Likewise, the functional relationship between the amygdala

and the temporal and occipital cortices is different during REM sleep than during
wakefulness or SWS.19 The reasons for these changes in the cerebral activity patterns
remain unclear. One possibility is that the brainstem structures influence the forebrain
activity in a regionally specific way through aminergic modulation or direct excita-
tory activities such as pontine waves.
BRAIN RESPONSES TO EXTERNAL STIMULI

DURING SLEEP
Sleep is not a state of complete unresponsiveness to external stimuli. Although
animal studies have suggested a decreased processing of sensory information during
sleep,20 human behavioral and physiological studies have shown that stimuli can be
integrated even into the sleeper’s mental or oneiric activity.21 External stimuli can
also induce an autonomic or electrophysiological response, in particular after a
relevant stimulus presentation.22 Event-related potentials (ERPs) studies have dem-
onstrated that external information is efficiently processed during sleep. The brain-
stem auditory-evoked potentials are not modulated by the vigilance state but rather
by the circadian variations of the body temperature, whereas the middle latency-
evoked potentials are found to be reduced during deep sleep.23
Long-latency components are also observed during sleep but are modulated by
the sleep stage. During NREM sleep (and especially in stage 2 sleep) ERPs corre-
spond to K-complexes, which are differently affected by the characteristics of the
stimulus, the early ones being more connected to the stimulus physical attributes
and the latter ones to its intrinsic significance.24 In contrast during REM sleep, the
morphology of long-latency components was very comparable to that observed in
wakefulness. Notably it has been shown that N100, mismatch negativity (MMN),
P300, and N400 could be recorded during this sleep stage. This implies that during
PS, subjects may automatically detect stimulus occurrence and discernible changes
in environment,25 may discriminate a deviant tones26,27 as well as her or his own first
name28 and may detect the presence of a linguistic incongruence.29,30
As indicated earlier, the preserved capacity to evaluate salient stimulus features
during SWS might be related to the relative preservation of cerebral activity in
unimodal sensory cortical regions.7,8 Accordingly the presentation of auditory stimuli
activates bilaterally the thalamus and the auditory cortex during NREM sleep as
well as during wakefulness;31 however hearing one’s own name (as compared to
hearing a neutral pure tone) additionally activates the left amygdala and prefrontal
(associative) cortex. These results suggest that the processing of external stimuli can
go beyond the primary cortices during NREM sleep. The mechanisms by which
salient stimuli can recruit associative cerebral areas during sleep remain unclear.
BRAIN RESPONSES TO INTERNAL STIMULI DURING

SLEEP: DOES THE PGO ACTIVITY EXIST IN HUMANS?
Ponto-geniculo-occipital (PGO) waves are prominent phasic bioelectrical potentials
that occur in isolation or in bursts just before and during REM sleep.32 In several

mammal species, including nonhuman primates, PGO waves seem to represent a

fundamental process of REM sleep, at least in its phasic aspects.33 PGO waves are
closely related to the generation of ocular saccades,34 therefore during REM sleep
saccades might be also generated in humans by mechanisms similar to PGO waves
in cats.35–37 This hypothesis implies that the neural activity of the brain regions from
which PGO are the most easily recorded in cats (i.e., the mesopontine tegmentum,38
the lateral geniculate bodies,39 and the occipital cortex32) should be more closely
related to spontaneous ocular movements during REM sleep than during wakefulness.
According to this prediction, regional blood flow changes in the lateral geniculate
bodies and in the striate cortex are significantly more correlated to ocular movement
density during REM sleep than during wakefulness (Figure 13.2).40 Hence cerebral
mechanisms for spontaneous ocular movement generation differ during REM sleep
and during wakefulness in humans, and brain regions known to be involved in the
generation of PGO waves in animals are involved in this phenomenon.
This finding is potentially important because PGO waves have been implicated
in various nonexclusive processes, such as the alerting reaction to external stimuli
or internal signals,41 sensorimotor integration through the transmission of an efferent
copy of ocular movements to the visual system,33 and facilitation of brain plasticity.42
EXPERIENCE-DEPENDENT CHANGES IN
FUNCTIONAL CONNECTIVITY DURING
POST-TRAINING SLEEP
Sleep is believed to participate in the consolidation of memory traces.43,44 Although
the processes of this consolidation remain unknown, the reactivation during sleep
of neuronal ensembles activated during learning appears as a possible mechanism
for the off-line memory processing. Such a reactivation has been reported in at least
two experimental situations: in the rat hippocampus45–50 and in the song area of
young zebra finches.51 This suggests the generality of the reactivation in the pro-
cessing of memory traces during sleep.
In order to observe the reactivation of brain areas during post-training sleep in
humans, we designed a multi-group experiment.52 A first group of subjects (group
1) were trained on a probabilistic serial reaction time (SRT) task* in the afternoon,
* In this task six permanent position markers are displayed on a computer screen above six spatially
compatible response keys. On each trial a black circle appears below one of the position markers, and
the task consists of pressing as fast and as accurately as possible on the corresponding key. The next
stimulus is displayed at another location after a 200-ms response-stimulus interval. Unknown to the
subjects, the sequential structure of the material is manipulated by generating series of stimuli based on
a probabilistic finite-state grammar that defines legal transitions between successive trials. To assess
learning of the probabilistic rules of the grammar, there is a 15% chance on each trial that the stimulus
generated based on the grammar (grammatical stimuli; G) is replaced by a nongrammatical (NG), random
stimulus. Assuming that response preparation is facilitated by high predictability, predictable G stimuli
should thus elicit faster responses than NG stimuli, but only if the context in which stimuli may occur
has been encoded by participants. In this task contextual sensitivity emerges through practice as a
gradually increasing difference between the reaction times (RTs) elicited by G and NG stimuli occurring
in specific contexts set by 2 to 3 previous trials at most.53

FIGURE 13.2 (See color insert.) Cerebral areas more active responding proportionally more
in relation to saccades during REM sleep than during wakefulness. Upper panel: transverse
sections from –4 mm to 0 mm from the bi-commissural plane. The functional data are
displayed at p < 0.001 uncorrected, superimposed on the average MRI of the subjects,
coregistered to the same reference space. Bottom panel: plot of the regional adjusted CBF
(arbitrary units) in the right geniculate body in relation to the rapid eye movement (REMs)
counts. The geniculate CBF is correlated to the rapid eye movement counts more during REM
sleep (in red) than during wakefulness (in green). (Adapted from Peigneux, P. et al., Generation
of rapid eye movements during paradoxical sleep in humans, Neuroimage, 14, 701, 2001.)
then scanned during the post-training night, both during waking and in various sleep
stages (i.e., SWS, stage 2, and REM sleep). A postsleep training session verified
that learning had occurred overnight. The analysis of PET data identified the brain
areas more active in REM sleep than during resting wakefulness.
To ensure that the post-training REM sleep rCBF distribution differed from the
pattern of typical REM sleep, a second group of subjects (group 2), not trained to

the task, were similarly scanned at night, both awake and during sleep. The analysis
was aimed at detecting the brain areas that would be more active in trained than in
nontrained subjects and during REM sleep as compared to resting wakefulness. And
finally, to formally test that these brain regions, possibly reactivated during REM
sleep, would be among the structures that had been engaged by executing and
learning the task, a third group of subjects (group 3) were scanned during wakeful-
ness both while they were performing the SRT task and at rest. The comparison
described the brain areas that are activated during the execution of the SRT task.
A conjunction analysis identified the regions that would be both more active
during REM sleep in the trained subjects (group 1) compared to the nontrained
subjects (group 2) and activated during the execution of the task during waking
(group 3); i.e., the regions reactivated in post-training REM sleep. Our results (Figure
13.3) showed that the bilateral cuneus and the adjacent striate cortex, the mesen-
cephalon, and the left premotor cortex were both activated during the practice of
the SRT task and during post-training REM sleep in subjects previously trained on
the task, significantly more than in control subjects without prior training, suggesting
a reactivation process that may have contributed to overnight performance improve-
ment in the SRT task.
In addition we reasoned that if the reactivated regions participate in the pro-
cessing of memory traces during REM sleep, they should establish or reinforce
functional connections between parts of the network activated during the task.
Consequently such connections should be stronger, and the synaptic trafficking
between network components more intense, during post-training REM sleep than
during the typical REM sleep of nontrained subjects. Accordingly we found that
among the reactivated regions, the rCBF in the left premotor cortex was significantly
more correlated with the activity of the pre-SMA and posterior parietal cortex during
post-training REM sleep than during REM sleep in subjects without any prior
experience with the task54 (Figure 13.3). The demonstration of a differential func-
tional connectivity during REM sleep between remote brain areas engaged in the
practice of a previously experienced visuo-motor task gave further support to the
hypothesis that memory traces are replayed in the cortical network and contributes
to the optimization of the performance.
It should be stressed that in this first experiment our conclusions were limited
by the fact that we could not specify whether the experience-dependent reactivation
during REM sleep was related to the simple optimization of a visuo-motor skill or
to the high-order acquisition of the probabilistic structure of the learned material, or
both. To test the hypothesis that the cerebral reactivation during post-training REM
sleep reflects the reprocessing of high-order information about the sequential struc-
ture of the material to be learned, a new group of subjects (group 4) was scanned
during sleep after practice on the same SRT task but using a completely random
sequence.55 The experimental protocol was identical in all respects to the trained
group in our original study,52 except for the absence of sequential rules. Therefore
post-training regional cerebral blood flow differences during REM sleep between the
subjects trained respectively to the probabilistic SRT or to its random version should
be related specifically to the reprocessing of the high-order sequential information.

FIGURE 13.3 (See color insert.) Experience-dependent reactivations during human REM
sleep. (A) Brain regions that are both activated in subjects scanned while performing the task
during wakefulness and more active in trained than in nontrained subjects scanned during
REM sleep. The SPM is displayed thresholded at P < 0.001 (uncorrected). (Data from Maquet,
P. et al., Experience-dependent changes in cerebral activation during human REM sleep, Nat.
Neurosci., 3, 831, 2000. Reproduced with permission from Nature Neuroscience.) (B) Sig-
nificant group (trained versus nontrained) by left premotor rCBF interaction in the posterior
parietal cortex (upper image) and the supplementary motor area (lower image). The red arrow
in panel A indicates the left premotor cortex. The SPM is displayed at P < 0.001 (uncorrected).
On the right-hand side, plots of the adjusted and centered rCBF of the left premotor cortex
(abscissa) and, respectively, the posterior parietal cortex and the supplementary motor area
(ordinate). The functional relationships between these two areas are significantly different in
trained subjects (red) than in nontrained subjects (green). (Adapted from Laureys, S, et al.,
Experience-dependent changes in cerebral functional connectivity during human rapid eye
movement sleep, Neuroscience, 105, 521, 2001.)
During post-training REM sleep, blood flow in left and right cuneus increased
more in subjects previously trained to a probabilistic sequence of stimuli than to a
random one (Figure 13.3 B). Because both groups were exposed prior to sleep to
identical SRT practice that differed only in the sequential structure of the stimuli,
our result suggests that reactivation of neural activity in the cuneus during post-
training REM sleep is not merely due to the acquisition of basic visuo-motor skills,

but rather it corresponds to the reprocessing of elaborated information about the

sequential contingencies contained in the learned material.
If the material does not contain any structure, as it is the case in the random
SRT task, post-training REM sleep reactivation does not occur, or it occurs at a
significantly lesser extent. These results are reminiscent of previous experiments. At
the behavioral level, increase in REM sleep duration was observed in rats following
aversive conditioning in which a tone is paired with a foot-shock, but not after
pseudo-conditioning in which the tone and the foot-shock were not paired.56 Using
a similar procedure at the systems level, tone-evoked responses were obtained in
the medial geniculate nucleus57 during REM sleep after a conditioning procedure
initiated at wake, but not after pseudo-conditioning. Likewise in humans REM sleep
percentage increased after learning textbook passages, but only when they were
meaningful.58 A similar situation occurred when the material to learn was so complex
that its underlying structure could not be extracted through practice.
Consequently, during REM sleep, functional connections should be reinforced
between the reactivated areas and cerebral structures specifically involved in
sequence learning only after the practice of the probabilistic version of the task. As
compared to the practice of the random sequence, we observed that the cuneus
establishes or reinforces functional connections with the caudate nucleus during
REM sleep following probabilistic SRT practice (Figure 13.3 C). The cuneus, which
participates in the processing of the probabilistic sequence both during SRT practice
and post-training REM sleep, has been shown to be activated during sequential
information processing in the waking state.59 On the other hand, the striatum is
known to play a main role in implicit sequence learning60 and specifically in the
encoding of the temporal context set by the previous stimulus in the probabilistic
SRT task.61 The finding that the strength of the functional connections between
cuneus and striatum is increased during post-training REM sleep suggests the
involvement of the basal ganglia (Figure 13.3 D) in the off-line reprocessing of
implicitly acquired high-order sequential information.
Finally, a direct relationship between the presleep learning performance and
regional blood flow was found in the cuneus. In this region the regional blood flow
during post-training REM sleep is modulated by the level of high-order, but not low-
order, learning attained prior to sleep (Figure 13.3 E). In other words, the neural
activity recorded during REM sleep in brain areas already engaged in the learning
process during wakefulness is related to the amount of high-order learning achieved
prior to sleep. This latter result further supports the hypothesis that sleep is actively
involved in the processing of recent memory traces.
CONCLUSION
As compared to wakefulness, segregated patterns of regional CBF activity are
observed during NREM and REM sleep in humans. The cortical activity is not only
influenced by the processes that lead to the generation of specific sleep patterns but
remains responsive to external stimuli. Moreover the neural populations recently
challenged by a new experience are reactivated and increase their functional con-

nectivity during the post-training sleep episodes, suggesting the off-line processing
of recent memory traces in sleep.
ACKNOWLEDGMENTS
The work summarized in this paper was supported by the Fonds National de la
Recherche Scientifique – Belgique (FNRS), the Fondation Médicale Reine Elisabeth,
the Research Fund of ULg, PAI/IAP Interuniversity Pole of Attraction P4/22, and
the Welcome Trust. PM and SL are supported by the FNRS.
FIGURE 13.4 (See color insert.)

REFERENCES
1. Friston, K.J., Imaging neuroscience: principles or maps? Proc. Natl. Acad. Sci. USA,
95, 796, 1998.
2. Steriade, M. and Amzica, F., Coalescence of sleep rhythms and their chronology in
corticothalamic networks, Sleep Research Online, 1, 1, 1998.
3. Madsen, P.L. et al., Cerebral oxygen metabolism and cerebral blood flow in man
during light sleep (stage 2), Brain Res., 557, 217, 1991.
4. Maquet, P. et al., Cerebral glucose utilization during stage 2 sleep in man, Brain Res.,
571, 149, 1992.
5. Madsen, P.L. et al., Cerebral O2 metabolism and cerebral blood flow in humans during
deep and rapid-eye-movement sleep, J. Appl. Physiol., 70, 2597, 1991.
6. Maquet, P. et al., Cerebral glucose utilization during sleep-wake cycle in man deter-
mined by positron emission tomography and [18F]2-fluoro-2-deoxy-D-glucose
method, Brain Res., 513, 136, 1990.
7. Kajimura, N. et al., Activity of midbrain reticular formation and neocortex during
the progression of human non-rapid eye movement sleep, J. Neurosci., 19, 10065,
1999.
8. Braun, A.R. et al., Regional cerebral blood flow throughout the sleep-wake cycle. An
H2(15)O PET study, Brain, 120, 1173, 1997.
9. Maquet, P. et al., Functional neuroanatomy of human slow wave sleep, J, Neurosci.,
17, 2807, 1997.
FIGURE 13.4 (See facing page.) Probabilistic versus random serial reaction time task. Data
from Peigneux, P. et al., Learned material content and acquisition level modulate cerebral
reactivations during post-training REM sleep (submitted). (A) Average reaction times (and
standard errors) per block for grammatical (G; red lines) and nongrammatical (NG; blue lines)
stimuli during pre- and postsleep sessions in Probabilistic (left-hand side) and Random (right-
hand side) groups. Subjects in the Random Group were exposed to the random sequence in
presleep sessions and to the probabilistic sequence in blocks 1–20 of postsleep sessions. In
contrast to the subjects in group 1 (trained to the probabilistic sequence, left panel), reaction
times for G and NG stimuli do not differ during the presleep training session for the subjects
of group 4 (trained to the random sequence, right panel). (B) Statistical parametric maps of
the brain regions that both activated during SRT practice (versus rest) and activated more
during REM sleep (versus wakefulness) in Probabilistic rather than Random group, superim-
posed on the coronal section of a subject’s normalized MRI at 68 and 70 mm behind the
anterior commissure. The SPM is displayed at p < 0.001, uncorrected. (C) The right caudate
nucleus, with which the right cuneus has a tighter functional connection in subjects trained
to the probabilistic SRT task than in subjects trained to the random SRT task. A similar
regression is observed between cuneus and caudate nucleus in the left hemisphere. The SPM
is displayed at p < 0.001, uncorrected. (D) Plot of the regression of centred CBF in the right
cuneus (32, –68, 12 mm) and right caudate nucleus (18, –12, 20 mm) during post-training
REM sleep in subjects trained to the probabilistic SRT task (red circles) and subjects trained
to the random SRT task (blue stars). (E) Regression of presleep high-order performance on
post-training REM sleep CBF (centred) in the right parieto-occipital fissure (coordinates 26,
–70, 24 mm in standard anatomical space), in Probabilistic SRT (circles) and Random SRT
(stars) subjects.

10. Hofle, N. et al., Regional cerebral blood flow changes as a function of delta and
spindle activity during slow wave sleep in humans, J. Neurosci., 17, 4800, 1997.
11. Andersson, J.L. et al., Brain networks affected by synchronized sleep visualized by
positron emission tomography, J. Cereb. Blood. Flow. Metab., 18, 701, 1998.
12. Maquet, P., Functional neuroimaging of normal human sleep by positron emission
tomography, J. Sleep. Res., 9, 207, 2000.
13. Krueger, J. et al., Sleep Modulation of the Expression of Plasticity Markers, in Sleep
and Brain Plasticity, Maquet, P., C., Smith, and Stickgold, R., Eds., 2003, Oxford
University Press, Oxford, (in press).
14. Pigarev, I.N., Nothdurft, H.C., and S. Kastner, Evidence for asynchronous develop-
ment of sleep in cortical areas, Neuroreport, 8, 2557, 1997.
15. Steriade, M. and McCarley, R.W., Brainstem Control of Wakefulness and Sleep, 1990,
Plenum Press, New York, p. 499.
16. Maquet, P. et al., Functional neuroanatomy of human rapid-eye-movement sleep and
dreaming, Nature, 383, 163, 1996.
17. Nofzinger, E.A. et al., Forebrain activation in REM sleep: an FDG PET study, Brain
Res., 770, 192, 1997.
18. Braun, A.R. et al., Dissociated pattern of activity in visual cortices and their projec-
tions during human rapid eye movement sleep, Science, 279, 291, 1998.
19. Maquet, P. and Phillips, C., Functional brain imaging of human sleep, J. Sleep Res.,
7, 42, 1998.
20. Pompeiano, O., Mechanisms of sensory-motor integration during sleep, Progr. Phys-
iol. Psychol., 3, 1, 1970.
21. Burton, S.A., Harsh, J.R., and Badia, P., Cognitive activity in sleep and responsiveness
to external stimuli, Sleep, 11, 61, 1988.
22. Bonnet, M., Performance during sleep, in Biological Rythms, Sleep, and Performance,
W. Webb, Ed., 1982, John Wiley & Sons, Chichester, p. 205.
23. Bastuji, H. and García-Larrea, L., Evoked potentials as a tool for the investigation
of human sleep, Sleep Medicine Rev., 3, 23, 1999.
24. Perrin, F. et al., Functional dissociation of the early and late portions of human K-
complexes, Neuroreport, 11, 1637, 2000.
25. Atienza, M., Cantero, J.L., and Escera, C., Auditory information processing during
human sleep as revealed by event-related brain potentials, Clin. Neurophysiol., 112,
2031, 2001.
26. Bastuji, H. et al., Brain processing of stimulus deviance during slow-wave and
paradoxical sleep: a study of human auditory evoked responses using the oddball
paradigm, J. Clin. Neurophysiol., 12, 155, 1995.
27. Niiyama, Y. et al., Endogenous components of event-related potential appearing
during NREM stage 1 and REM sleep in man, Intl. J. Psychophysiol., 17, 165, 1994.
28. Perrin, F. et al., A differential brain response to the subject’s own name persists during
sleep, Clin. Neurophysiol., 110, 2153, 1999.
29. Brualla, J. et al., Auditory event-related potentials to semantic priming during sleep,
Electroencephalogr. Clin. Neurophysiol., 108, 283, 1998.
30. Perrin, F., Bastuji, H., and Garcia-Larrea, L., Detection of verbal discordances during
sleep, Neuroreport, 13, 1345, 2002.
31. Portas, C.M. et al., Auditory processing across the sleep-wake cycle: simultaneous
EEG and fMRI monitoring in humans, Neuron, 28, 991, 2000.
32. Mouret, J., Jeannerod, M., and Jouvet, M., L’activité électrique du système visuel au
cours de la phase paradoxale du sommeil chez le chat, J. Physiol. (Paris), 55, 305,
1963.

33. Callaway, C.W. et al., Pontogeniculooccipital waves: spontaneous visual system activ-
ity during rapid eye movement sleep, Cell Mol. Neurobiol., 7, 105, 1987.
34. Datta, S., PGO wave generation: mechanism and functional significance, in Rapid
Eye Movement Sleep, Mallick, B.N. and Inoue, S., Eds., 1999, Narosa Publishing
House, New Delhi, p. 91.
35. Inoué, S., Saha, U.K., and Musha, T., Spatio-temporal ristribution of neuronal activ-
ities and REM sleep, in Rapid-Eye-Movement Sleep, B.N. Mallick and S. Inoue, Eds.,
1999, Narosa Publishing, New Delhi, p. 214.
36. McCarley, R.W., Winkelman, J.W., and Duffy, F.H., Human cerebral potentials asso-
ciated with REM sleep rapid eye movements: links to PGO waves and waking
potentials, Brain Res., 274, 359, 1983.
37. Salzarulo, P. et al., Direct depth recording of the striate cortex during REM sleep in
man: are there PGO potentials? EEG Clin. Neurophysiol., 38, 199, 1975.
38. Jouvet, M. and Michel, F., Corrélations électromyographiques du sommeil chez le
Chat décortiqué et mésencéphalique chronique, C.R. Soc. Biol. (Paris), 153, 422,
1959.
39. Mikiten, T.H., Niebyl, P.H., and Hendley, C.D., EEG desynchronization during behav-
ioral sleep associated with spike discharges from the thalamus of the cat, Fed. Proc.,
20, 327, 1961.
40. Peigneux, P. et al., Generation of rapid eye movements during paradoxical sleep in
humans, Neuroimage, 14, 701, 2001.
41. Bowker, R.M. and Morrison, A.R., The startle relfex and PGO spikes, Brain Res.,
102, 185, 1976.
42. Datta, S., A physiological substrate for sleep dependent memory processing, Sleep
Research Online, 2, 23, 1999.
43. Peigneux, P. et al., Sleeping brain, learning brain, the role of sleep for memory
systems, Neuroreport, 12, A111, 2001.
44. Maquet, P., The role of sleep in learning and memory, Science, 294, 1048, 2001.
45. Kudrimoti, H.S., C.A., Barnes, and McNaughton, B.L., Reactivation of hippocampal
cell assemblies: effects of behavioral state, experience, and EEG dynamics, J. Neu-
rosci., 19, 4090, 1999.
46. Lee, A.K. and Wilson, M.A., Memory of sequential experience in the hippocampus
during slow wave sleep, Neuron, 36, 1183, 2002.
47. Louie, K. and Wilson, M.A., Temporally structured replay of awake hippocampal
ensemble activity during rapid eye movement sleep, Neuron, 29, 145, 2001.
48. Nadasdy, Z. et al., Replay and time compression of recurring spike sequences in the
hippocampus, J. Neurosci., 19, 9497, 1999.
49. Pavlides, C. and Winson, J., Influences of hippocampal place cell firing in the awake
state on the activity of these cells during subsequent sleep episodes, J. Neurosci., 9,
2907, 1989.
ories during sleep, Science, 265, 676, 1994.
51. Dave, A.S. and Margoliash, D., Song replay during sleep and computational rules
for sensorimotor vocal learning, Science, 290, 812, 2000.
52. Maquet, P. et al., Experience-dependent changes in cerebral activation during human
REM sleep, Nat. Neurosci., 3, 831, 2000.
53. Cleeremans, A. and McClelland, J.L., Learning the structure of event sequences, J.
Exp. Psychol. General, 120, 235, 1991.
54. Laureys, S, et al., Experience-dependent changes in cerebral functional connectivity
during human rapid eye movement sleep, Neuroscience, 105, 521, 2001.

55. Peigneux, P. et al., Learned material content and acquisition level modulate cerebral
reactivations during post-training REM sleep (submitted).
56. Hennevin, E. and Leconte, P., The function of paradoxical sleep: facts and theories,
Annee. Psychol., 71, 489–519, 1971.
57. Hennevin, E. et al., Learning-induced plasticity in the medial geniculate nucleus is
expressed during paradoxical sleep, Behav. Neurosci., 107, 1018, 1993.
58. Verschoor, G. and Verschoor, H.T.L., REM bursts and REM sleep following visual
and auditory learning, S. Afr. J. Psychol., 14, 69, 1984.
59. Schubotz, R.I. and von Cramon, D.Y., Interval and ordinal properties of sequences
are associated with distinct premotor areas, Cerebral Cortex, 11, 210, 2001.
60. Rauch, S. et al., A PET investigation of implicit and explicit sequence learning,
Human Brain Mapping, 3, 271, 1995.
61. Peigneux, P. et al., Striatum forever, despite sequence learning variability: a random
effect analysis of PET data, Human Brain Mapping, 10, 179, 2000.
62. Schwartz, S. and Maquet, P., Sleep imaging and the neuro-psychological assessment
of dreams, Trends Cogn. Sci., 6, 23, 2002.

1519_CPgs.fm Page 225 Thursday, August 19, 2004 9:48 AM
COLOR FIGURE 3.6
W/SWS PeF, HLA PS PeF, HLA

Hcrt Hcrt
Pef/HLA Pef/HLA
MCH MCH
Thalamus Thalamus
EEG activation EEG activation
Glu Glu
SLD LC/DRN SLD LC/DRN

PAG, DPMe, PRN PAG, MRN, PRN
NA/5-HT NA/5-HT
Glu Glu
Glu Glu
DPMe, PRN, SLD
D DPMe, PRN, SLD
LDT/PPT GABA LDT/PPT GABA
Ach Ach
Mc Mc
Muscle atonia Muscle atonia
Gly Gly
vlPAG/DPGi vlPAG/DPGi
GABA GABA
PS-on PS-off PS-on PS-off
COLOR FIGURE 5.5

1519_CPgs.fm Page 226 Friday, August 27, 2004 3:43 PM
COLOR FIGURE 6.2 COLOR FIGURE 8.1
COLOR FIGURE 8.2 COLOR FIGURE 8.3

COLOR FIGURE 8.4
0.60
A Post-Nonexposure
Post-Exposure
B
0.50
*
SPIKES/SEC
0.40
0.30 SONG
*
0.20 SLEEP
0.10
100ms
0.00
WK SWS REM
C WK SWS REM D
4
Brain activation during REM sleep

Control
(adjusted CBF)
1
Enriched Environment
-1
-2
-3
Low High -20 0 20 40 60 80 100
Zif-268 Expression Learning level attained prior to sleep
COLOR FIGURE 12.1

A
* **
Time (minutes)
B
Time (minutes)
C D 0.3
WK
Pre
Cortical Neurons
Post
0.2
0.1
0.0
EXP -0.1
0 12 24 36 48
Time (hours)
COLOR FIGURE 12.4
COLOR FIGURE 13.1

COLOR FIGURE 13.2
COLOR FIGURE 13.3

COLOR FIGURE 13.4

Sleep Circuits and Functions PDF

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Sleep Circuits and Functions PDF

Caricato da

Copyright:

Formati disponibili

1519_Title 8/27/04 11:05 AM Page 1

Library of Congress Cataloging-in-Publication Data

Sleep : circuits & functions / edited by Pierre-Hervé Luppi.

Visit the CRC Press Web site at www.crcpress.com

© 2005 by CRC Press LLC

No claim to original U.S. Government works

Copyright © 2005 CRC Press LLC

Methods & New Frontiers

Sidney A. Simon, Ph.D.

Copyright © 2005 CRC Press LLC

Copyright © 2005 CRC Press LLC

• The use of cholera toxin A, which is equivalent to placing a stimulating

Copyright © 2005 CRC Press LLC

Chapter 2 Role of Basalo-Cortical System in Modulating Cortical Activity

Chapter 3 In Vitro Identification of the Presumed Sleep-Promoting Neurons

Chapter 4 Molecular Mechanisms of Sleep-Wake Regulation: A Role of

Chapter 5 The Network Responsible for Paradoxical Sleep Onset and

Chapter 6 Reverse Genetics and the Study of Sleep-Wake Cycle:

Chapter 7 Genetic Regulation of Sleep

Chapter 8 Searching for Sleep Mutants of Drosophila melanogaster

Copyright © 2005 CRC Press LLC

Chapter 9 Sleep Phylogeny: Clues to the Evolution and Function

Chapter 10 Sleep, Synaptic Plasticity, and the Developing Brain

Chapter 11 Changes in Brain Gene Expression between Sleep and

Chapter 12 Neuronal Reverberation and the Consolidation of New

Chapter 13 Cerebral Functional Segregation and Integration during

Copyright © 2005 CRC Press LLC

1 Sleep and Neuronal

Copyright © 2005 CRC Press LLC

SLEEP RHYTHMS: THEIR GROUPING BY CORTICAL

• Neuronal circuitry that underlies different sleep oscillations

Copyright © 2005 CRC Press LLC

FIGURE 1.1 Natural slow-wave sleep (SWS) characterized by prolonged hyperpolarizations

• Human studies that corroborate experimental work and together empha-

SPINDLES, A THALAMIC RHYTHM UNDER CORTICAL INFLUENCE

Copyright © 2005 CRC Press LLC

rhythmic inhibitory postsynaptic potentials (IPSPs) that de-inactivate the Ca2+-

Copyright © 2005 CRC Press LLC

The pacemaking role of thalamic reticular GABAergic neurons was demon-

Copyright © 2005 CRC Press LLC

isolated network between thalamocortical and thalamic reticular neurons oscillated

• The first part of a spindle sequence is generated in the pacemaker thalamic

Importantly, although generated within the thalamus, spindles can be triggered

Copyright © 2005 CRC Press LLC

CLOCK-LIKE THALAMIC DELTA, AN INTRINSIC CELL OSCILLATION

Copyright © 2005 CRC Press LLC

for the appearance of the clock-like rhythm is the hyperpolarization of thalamocortical

THE SLOW CORTICAL OSCILLATION AND ITS ACTIONS IN GROUPING OTHER

Copyright © 2005 CRC Press LLC

Several pieces of evidence support this conclusion:

• Neurons identified morphologically as basket (aspiny or sparsely spiny)

The disfacilitation is explained by a progressive depletion of [Ca2+]o during the

SLEEP RHYTHMS LEADING TO NEURONAL

Copyright © 2005 CRC Press LLC

on the behavioral state of vigilance, the mechanisms underlying plasticity include

Copyright © 2005 CRC Press LLC

or their experimental models to neuronal plasticity was investigated in thalamocor-

INTRATHALAMIC AND THALAMOCORTICAL NEURONAL CIRCUITS

Copyright © 2005 CRC Press LLC

The distinction between augmenting and recruiting responses is no longer necessary

NEURONAL PLASTICITY OUTLASTING AUGMENTING RESPONSES AND

Copyright © 2005 CRC Press LLC

presents evidence of memory processes in the more complex corticothalamic net-