Crosslinking of collagen gels by transglutaminase

Janine M. Orban,1 Lorri B. Wilson,1 Jessica A. Kofroth,2 Mohammed S. El-Kurdi,1 Timothy M. Maul,2
David A. Vorp1,2
1
Department of Surgery, University of Pittsburgh, Room 236, Cellomics Bldg., McGowan Institute for Regenerative
Medicine, 100 Technology Drive, Pittsburgh, Pennsylvania 15219
2
Department of Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15219
Received 21 February 2002; revised 14 October 2003; accepted 15 October 2003

Abstract: Collagen is commonly used as a tissue-engineering scaffold, yet its in vivo applications are limited by a
deciency in mechanical strength. The purpose of this work
was to explore the utilization of a unique enzymatic
crosslinking procedure aimed at improving the mechanical
properties of collagen-based scaffold materials. Type I bovine collagen gel was crosslinked by transglutaminase,
which selectively mediates the chemical reaction between
glutamine and lysine residues on adjacent protein bers,
thus providing covalent amide bonds that serve to reinforce
the three-dimensional matrix. The degree of crosslinking
was veried by thermal analysis and amine group content.
The denaturation temperature of crosslinked collagen
reached a maximum of 66 1C. The chemical reaction was

conrmed to be noncytotoxic with respect to bone marrow

stromal cells acquired from New Zealand White rabbits.
Tube-shaped cellular constructs fashioned from crosslinked
collagen and bone marrow stromal cells were found to have
burst pressures signicantly higher than their noncrosslinked analogs (71 4 mmHg vs. 46 3 mmHg; p
0.01). Thus, the transglutaminase mediated reaction served
to successfully strengthen collagen gels while remaining
benign toward cells. 2004 Wiley Periodicals, Inc. J Biomed
Mater Res 68A: 756 762, 2004

INTRODUCTION

imide,18 dimethylsuberimidate,19 and nordihydroguaiaretic acid20 impart some degree of cytotoxicity when
prepared prior to cell seeding. Furthermore, the chemical reaction that occurs between amine or carboxylic acid
groups is characteristic of xing techniques, and must
be averted in the case of in situ crosslinking of cellseeded gels. To date, the only recognized mechanisms
for strengthening collagen constructs in the presence of
cells are nonenzymatic glycation10 and enzyme-mediated crosslinking techniques.22,24 We postulate that enzymes can be used to crosslink cell-seeded collagen gels
in situ, thereby enhancing mechanical strength while
remaining benign toward the cells.
Transglutaminases (TG), calcium-dependent enzymes distributed intra- and extracellularly throughout the body, are responsible for tissue stabilization
through formation of high molecular weight complexes via chemical crosslinking.21 The reaction has
been identied to catalyze the formation of the amide
crosslink from -carboxamide and primary amine
functionalities.22,23 In vivo, this reaction characteristically takes place between the glutamine and lysine
residues of collagen (Fig. 1). In addition to TG, the
lysyl oxidase family of enzymes is also known to
catalyze the crosslinking of elastin and collagen bers.

Collagens comprise the majority of proteins in connective tissue such as skin, bone, cartilage, and tendons, and are therefore popular candidates for biomedical materials. However, the effectiveness of
collagen-based tissue engineered materials has been
severely limited by their lack of mechanical strength.
To contend with this issue, collagen-based materials
can be strengthened utilizing chemical crosslinking
agents such as glutaraldehyde,1 4 acyl azides,5 glycidyl ethers,6 diisocyanates,7 and other methods.8 14
Such crosslinking treatments have also been shown to
reduce biodegradation.4,13,14 Traditional collagen
crosslinking reagents, for example, glutaraldehyde,15
hexamethylenediisocyanate,16 bis-epoxide,17 carbodiCorrespondence to: D.A. Vorp; e-mail: VorpDA@upmc.edu
Contract grant sponsor: the Pittsburgh Tissue Engineering
Initiative (PTEI; to D.A.V.).
Contract grant sponsor: a PTEI summer internship (to
L.B.W.).
Contract grant sponsor: NIH; contract grant number: R01
HL069368-01A1
2004 Wiley Periodicals, Inc.

Key words: crosslinking; collagen; enzyme; transglutaminase; tissue engineering

CROSSLINKING OF COLLAGEN GELS

757

type I collagen (CELLGEN), (0.5%) was purchased from

Kohken Corp. (Japan). Porcine transglutaminase was aliquotted into a storage buffer containing 10 mM Tris-acetate, 160
mM KCl, 1.0 mM EDTA, 2.0 mM DTT, pH 8.0, and kept at
120C until use.

Cell culture

Figure 1. Schematic of the crosslinking reaction of collagen

by TG.

Lysyl oxidase catalyzes the oxidation of lysine to

-aminoadipic--semialdehyde, a reactive precursor
moiety,24 and has been reviewed elsewhere.2527
Experimentally, TG has primarily been exploited for
crosslinking various peptides to extracellular matrix
components, to alter the biological characteristics of
three-dimensional matrices in vitro.28 32 TG also has
been utilized in the preparation of synthetic polymeric
hydrogels.33 The objective of this work was to prepare
and characterize this reaction, evaluate the cytotoxicity, and quantify the improvement in mechanical
strength of a cell-seeded collagen gel. We believe that
this technique will lead to structurally robust collagen
matrices that can be used reliably for tissue engineering applications. We report herein that TG successfully crosslinks acid solubilized bovine type I collagen
through a transamidation reaction that produces viable cell-seeded crosslinked gels with signicantly enhanced mechanical strength.

MATERIALS AND METHODS

Materials
Tris-acetate, Tris-HCl, potassium chloride (KCl), ethylenediamine tetraacetic acid (EDTA), DL-dithiothreitol
(DTT), 3-(4,5-dimethylazol-2-yl)-2,5-diphenyl-2H-tetrazolium
bromide (MTT), calcium chloride (CaCl2), 2,4,6-trinitrobenzenesulfonic acid (TNBS); dimethyl formamide (DMF), sodium
dodecylsulfate (SDS), Dulbeccos modied eagles medium
(DMEM), penicillin, streptomycin, fungizone, and fetal calf
serum (FCS) were used as received. Acid solubilized bovine

Bone marrow stromal cells (BMSCs) are currently used

within collagen gels in our laboratory for tissue engineered
constructs.34 Therefore, we employed this cell type for the
investigation of transglutaminase cytotoxicity. BMSC culture and isolation was performed via the Dexter culture
methods. Briey, BMSCs were collected from New Zealand
White rabbits obtained from unrelated studies, and the femora excised, cleaned of soft tissue, cracked open, and the
bone marrow collected by aspiration with DMEM. In all
cases, DMEM was supplemented with sodium bicarbonate
(2.3 mg/mL), penicillin (50 IU/mL), streptomycin (50 g/
m), fungizone (1.5 g/mL), and 15% FCS. The cell suspension was centrifuged twice at 1000 rpm, resuspending the
pellet in fresh DMEM each time. The cells were cultured no
higher than passage 6 and maintained under standard culture conditions, i.e., a sterile, humidied, 37C, 5% CO2/95%
air environment.

Collagen gel preparation

Gels were prepared by mixing collagen (0.5%), Tris-HCl
(pH 8.0, 50 mM), CaCl2 (2.5 mM), DTT (1 mM), and the
designated ratio (w/w) TG at 4C in a 15 mL centrifuge tube.
Samples for thermal and amine group analysis were prepared using the (w/w) collagen:enzyme ratios 50,000:1,
5,000:1, 500:1, 50:1, and noncrosslinked. Each solution was
aliquotted into a 24-well cluster (0.5 mL per well), incubated
for a gelation period of 30 min at 37C, and then maintained
in a CO2 incubator at 37C in PBS until analysis. For thermal
analysis N 3, and for amine group quantication N 2.
Gels containing BMSCs for viability and burst pressure experiments were prepared by mixing a 1:1 (v/v) suspension
of BMSC (2 106 cells/mL in DMEM) and collagen at 4C,
with the addition of Tris-HCl (pH 8.0, 50 mM), CaCl2 (2.5
mM), and DTT (1 mM). DMEM was supplemented with
sodium bicarbonate (2.3 mg/mL), penicillin (50 IU/mL),
streptomycin (50 g/mL), fungizone (1.5 g/mL), and 15%
FCS. The suspension was aliquotted (0.5 mL) into two columns of a 24-well cluster (N 4). One column was used as
the noncrosslinked control, and to the other was added TG
(5000:1 collagen:TG w/w). A separate plate was prepared
for each time point. Samples were maintained under standard culture conditions, performing media changes every 3
days. The time zero samples were analyzed immediately
(within 15 min) after the 30-min incubation.

Thermal analysis
Thermal analysis was exploited in the assessment of collagen gel strength, by measuring the denaturation temper-

758

ature (Td). A rise in Td indicates augmented collagen gel

strength, and therefore an increase in crosslinking in the case
of these collagen matrices. Thermal transitions of samples
within each collagen:TG ratio group were measured using a
PerkinElmer DSC-7 differential scanning calorimeter over a
temperature range of 25 to 80C at a heating rate of 10C/
min. The denaturation temperature (Td) was recorded at
maximum peak height of the denaturation endotherm.

Amine group content

End group analysis was used to assay the extent of reaction via the occurrence of transamidation. The number of
free amine groups was quantied by trinitrobenzenesulfonic
acid (TNBS) absorption.35 After the 30-min gelation period,
a solution containing 4% NaHCO3 and 0.5% TNBS (200 L)
was added to each well, and the plates kept in the dark at
37C for 2 h. Subsequently, 6 M HCl (200 L) was added and
the temperature maintained at 37C until solubilization, typically less than 60 min. The resulting solutions were aliquotted into four wells of a 96-well cluster and the absorbance
measured using a microplate spectrophotometer (Molecular
Devices) at 345 nm. The blank was prepared from a noncrosslinked sample in the same manner, with the exception
that solubilization was performed prior to addition of the
chromophore. The absorbance values exhibited graphically
represent corrected optical density (O.D.) values.

Cell viability
The toxicity of TG crosslinking technique with regard to
BMSCs was measured using the MTT assay. The number of
viable cells was assayed over a period of 14 days, assuming
that any toxic effects would be manifested within that period
of time. The MTT chromophore represents the number of
viable cells by measuring the amount of formazan generated
by the mitochondrial enzymes of metabolically active cells.36
Thus, the number of viable cells can be correlated with the
absorbance at 540 nm (max). To quantify BMSC viability,
MTT (50 L, 5 mg/mL PBS) was added to each well and the
plate incubated for 3 h at 37C. After the MTT treatment, the
plates were centrifuged at 800 g for 5 min, the supernatant
removed, and the pellet washed once with PBS. To each well
was added extraction buffer (250 L, 20% SDS/50% DMF,
pH 7.4) and the gels kept at 37C overnight. After solubilization, each well was diluted 10-fold, and aliquotted into
one column of 96-well plate. The absorbance was measured
at 0, 2, 6, 9, and 14 days using a microplate spectrophotometer (Molecular Devices) at 550 nm, with the extraction
buffer as the blank.

Burst pressure
Double-layer tubular constructs were fashioned as described34 from crosslinked and noncrosslinked BMSC-con-

ORBAN ET AL.

taining collagen gels for burst strength analysis. The samples

remained in static culture in DMEM on a supporting mandrel for 7 days, after which they were removed from their
mandrel for burst pressure measurement as follows. A
steady-ow centrifugal pump and ow loop were primed
with PBS at 37C. Flow was halted momentarily while the
tubular construct was attached via purse-string suture to
two perfusion tees within a housing chamber (also lled
with circulating PBS at 37C). Flow was then reinitiated and
adjusted to a mean rate of 40 mL/min with a uniform
pressure of 15 mmHg. The construct was exposed to these
conditions for 15 min to accelerate equilibration to physiologic conditions (i.e., temperature and blood gases). Following this, continuous (30 Hz) measurements of intraluminal
pressure were recorded using a strain gauge pressure transducer. Flow was then halted and a syringe pump switched
into the system proximal to the construct. The tubing distal
to the specimen was clamped off and PBS infused at a rate of
2 mL/min. The intraluminal pressure was measured continuously until failure. Burst pressure was recorded as the
maximum pressure attained prior to gross rupture of the
construct.

Data analysis
Statistical analysis was performed using Statistica 98
software (Statsoft, Inc., Tulsa, OK). All data is presented as
mean SEM. Assessment of burst strength was resolved
using one-tailed Students t-test assuming two samples of
unequal variance. For all other experiments, one-way analysis of variance (ANOVA) was executed to determine differences among collagen:enzyme treatment groups, assuming a 95% condence interval. Signicance was assumed for
p 0.05. When signicance was observed, the Sheffe post
hoc multivariable test was used to determine variance
among groups.

RESULTS
Extent of crosslinking
The extent of crosslinking was measured both by
amine group quantication and by thermal analysis.
As the ratio of collagen:enzyme was varied between
50,000:1, 5000:1, 500:1, 50:1 (w/w), including a noncrosslinked control, the TNBS absorbance decreased
thereby signifying a decrease in amine groups present
with increasing enzyme concentration (Fig. 2). Analyses indicate a signicant difference in number of
amine groups present between the noncrosslinked
and all of the crosslinked samples. However, there
was not a signicant change in the number of amine
groups as enzyme concentration was varied between
5000:1 and 50:1 (w/w).
The denaturation temperature (Td) was used as a
relative measure of tissue strength. In the case of col-

CROSSLINKING OF COLLAGEN GELS

Figure 2. Amine group content of crosslinked collagen gels

(N 8 for each group). Average TNBS absorption (345 nm)
plotted against TG concentration. The value at 0 collagen:
enzyme represents the noncrosslinked control. Error bars
stand for SEM. p 0.01 between all crosslinked samples and
the control.

759

Figure 4. Cell viability as measured by average MTT absorbance (SEM) over a 14-day culture period for noncrosslinked and crosslinked (5000:1 collagen:enzyme) samples (N 4 for each data point). *p 0.05 for comparison of
crosslinked and noncrosslinked samples. There is no statistical difference at any other time points.

Cytotoxicity
lagen gels, as the sample becomes progressively more
crosslinked, and accordingly stronger, the temperature required to denature the gel, increased. This trend
is evident in Figure 3 as the mean Td increases from
38 1C to 66 1C between the control and foremost crosslinked sample. Statistical analysis revealed
a signicant increase in average denaturation temperature between the noncrosslinked and each
crosslinked group (p 0.05). In addition, the denaturation temperature of samples crosslinked with
50,000:1 was signicantly different (p 0.05) from
5000:1, 500:1, and 50:1.

The viability of BMSCs within the TG-crosslinked

(5000:1 collagen:TG) gels was measured over a 14-day
period. Crosslinked and noncrosslinked gels were
found to contain a similar number of viable cells at 2,
6, and 9 days. A statistical difference (p 0.05) exists
at 0 and 14 days, with the crosslinked gels having a
greater number of viable cells (Fig. 4). Further comparison indicates no signicant difference (p 0.05) in
cell viability between crosslinked and noncrosslinked
samples at any other time point.

Burst strength
During in vitro culture, the construct underwent
compaction (remodeling) as evidenced by longitudinal contraction of the gel on the mandrel. This contraction was not quantied here, nor was the effect of
crosslinking. However, controls typically contract
30% in length at 7 days, and we did not note (qualitatively) any differences between the controls and
crosslinked constructs. The average burst pressures
between crosslinked (5000:1 collagen:TG) and noncrosslinked samples were determined to be 71 4 and
46 3 mmHg, respectively (p 0.01, Fig. 5). The
mode of failure for all samples was a pinhole rupture
in the center of the construct.
Figure 3. Mean denaturation temperature (SEM) in C of
crosslinked collagen gels measured by differential scanning
calorimetry (N 3 for each group). The value at 0 collagen:
enzyme represents the noncrosslinked control. A signicant
difference (p 0.05) is noted between all crosslinked samples and the control, and between 50,000:1 collagen:enzyme
and higher enzyme concentrations.

DISCUSSION
In this article, structurally robust collagen matrices
were prepared via enzyme-mediated crosslinking us-

760

Figure 5. Mean burst pressure of tube-shaped cellular constructs (SEM) as measured by peak pressure (mmHg) attained prior to specimen failure. p 0.01 between control
(N 8) and samples crosslinked with 5000:1 collagen:TG
(N 10).

ing TG. The chemical reaction (Fig. 1) was characterized by amine group quantication, which conrmed
the presence of amide functionality. Cellular viability
was revealed to be unaffected by the crosslinking reaction over a 14-day period. Finally, tube-shaped cellular constructs fashioned from crosslinked gels possessed signicantly higher burst strengths than
noncrosslinked counterparts.
Evaluation of chemical crosslinking is commonly
performed by end group analysis because it probes
the consumption of functional groups. When collagen
is used as the substrate, lysine residues participate in
the reaction to produce the amide bond, and therefore
a decrease in the number of amine groups is anticipated. The TNBS chromophore (max 405 nm) is
commonly used in such analyses, and has indeed been
utilized to determine the extent of collagen crosslinking.6,35 Figure 2 reveals that after 30 min, the extent of
reaction reaches a plateau at 5000:1 collagen:TG. From
this data, it is established that 5000:1 is the optimal
enzyme concentration for such conditions, and was
employed as the standard enzyme concentration in
this investigation.
Thermal transitions have been documented for various types of tissues and gels, providing adequate
reference for the strength of TG-crosslinked collagen
gels. For example, bovine dermal collagen denatures
at 44C,5 while diphenylphosphorylazide-crosslinked
collagen exhibits a denaturation temperature as high
as 62C.5 This value approaches the value reported for
calf pericardium (68C),37 and is direct evidence of a
highly strengthened collagen-based construct. Our
thermal analyses of collagen gels crosslinked with TG
are consistent with these literature values (Fig. 3).
The presence of cells within a collagen matrix in
some applications, such as tissue engineering, requires
that the crosslinking techniques be noncytotoxic. In
situ preparation of engineered tissue is benecial because it eliminates complexity in the fabrication se-

ORBAN ET AL.

quence and provides a homogeneously seeded, threedimensional construct when compared with
conventionally seeded scaffolds. The MTT assay was
chosen for cell viability analysis because it is frequently used for this purpose and because it has been
found to correspond closely to values obtained using
tritiated thymidine uptake.38,39 Our results suggest
that, unlike various chemical methods, the enzymemediated reaction is benign and may be successfully
utilized in applications such as tissue engineering (Fig.
4). Although a statistical difference does exist in the
number of viable cells present in the crosslinked and
noncrosslinked gels at the rst and last time points,
the data suggests that crosslinked gels promote cell
viability, if anything. The statistical difference in viable cells between crosslinked and noncrosslinked gels
at t 0 is likely due to experimental error in preparing
a homogeneous collagen-cell suspension.
Although the burst pressures of the tissue engineered tubular constructs are remarkably low (46 and
71 mmHg), compared with the saphenous vein (1680
mmHg),40 and other constructs with polymer scaffolding,41 the statistically signicant improvement in burst
pressure due to crosslinking (Fig. 5) suggests an advancement provided by this technique. These results
might have implications for other tissue engineering
approaches that result in insufcient tissue strength.
For example, Hirai and Matsuda reported burst pressures of just over 100 mmHg for collagen/smooth
muscle cell (SMC) tubular constructs incubated for 7
days.42
There are certain limitations to this work that
should be kept in mind. For example, preparation of
cell seeded gels was performed in serum containing
media (15% FCS). The presence of serum proteins
introduces a potentially reactive species, capable of
quenching the crosslinking reaction. The rationale for
including serum was that maintaining a positive cellular environment took precedence over elimination of
small quantities of reactant moieties. In other words, a
higher degree of crosslinking may be established
through the exclusion of serum, although the burst
strength data indicates a signicant difference without
compromising cellular metabolism. Additionally, the
measurement of a true burst strength should be
made with consideration of construct dimensions
(e.g., thickness, inner diameter). However, the burst
pressure measurement made in this work was intended to serve as a simple comparison between the
strength of crosslinked and noncrosslinked constructs
examined in this study. This is a reasonable comparison because the tubular constructs were made in the
same fashion, in like molds, and were of similar dimensions. Uniaxial tensile testing was attempted on
the constructs, but such tests were deemed unsuccessful due to difculty in gripping the small, fragile,
circumferentially aligned tissue specimens. Nonethe-

CROSSLINKING OF COLLAGEN GELS

less, it is our belief that burst pressure is a sufcient

biomechanical end point for the purposes of this report.
In conclusion, TG was identied to crosslink bovine
collagen by providing the covalent bonds necessary
for strength augmentation. The reaction was found to
be noncytotoxic during in situ crosslinking, which imparts ease of fabrication during construct preparation.
Thus, enzyme-mediated crosslinking has an abundance of potential applications in tissue engineering
research.

761

13.

14.

15.

16.

17.

The authors express gratitude to Dr. Phil Campbell for

helpful discussions and David Wang for statistical support.
No benet of any kind will be received either directly or
indirectly by the authors.

18.

19.

References
20.
1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

Jayakrishnan A, Jameela SR. Glutaraldehyde as a xative in

bioprostheses and drug delivery matrices. Biomaterials 1996;
17:471 484.
Olde Damink LLH, Dijkstra PJ, van Luyn MJ, van Wachem PB,
Nieuwenhuis P, Feijen J. Glutaraldehyde as a crosslinking
agent for collagen based biomaterials. J Mater Sci Mater Med
1995;6:460 472.
Kanamori T, Habu T, Shinbo T, Sakai K. Difference in solute
diffusivity in crosslinked collagen gels prepared under various
conditions. Mater Sci Eng C 2000;13:85 89.
Goissis G, Yoshioka SA, Braile DM, Ramirez VD. The chemical
protecting group concept applied in crosslinking of natural
tissues with glutaraldehyde acetals. Artif Organs 1998;22:210
214.
Chevallay B, Abdul-Malak N, Herbage D. Mouse broblasts in
long-term culture within collagen three-dimensional scaffolds:
Inuence of crosslinking with diphenylphosphorylazide on
matrix reorganization, growth, and biosynthetic and proteolytic activities. J Biomed Mater Res 2000;49:448 459.
Zeeman R, Dijkstra PJ, van Wachem PB, van Luyn MJ, Hendriks M, Cahalan PT, et al. Crosslinking and modication of
dermal sheep collagen using 1,4-butanediol diglycidyl ether.
J Biomed Mater Res 1999;46:424 433.
Olde Damink L, Dijkstra PJ, van Luyn MJ, van Wachem PB,
Nieuwenhuis P, Feijen J. Crosslinking of dermal sheep collagen
using hexamethylene diisocyanate. J Mater Sci Mater Med
1995;6:429 434.
Wissink MJ, van Luyn MJ, Beernink R, Dijk F, Poot AA, Engbers GH, et al. Endothelial cell seeding on crosslinked collagen:
Effects of crosslinking on endothelial cell proliferation and
functional parameters. Thromb Haemost 2000;84:325331.
Pieper JS, van Wachem PB, van Luyn MJ, Brouwer LA, Hafmans T, Veerkamp JH, et al. Attachment of glycosaminoglycans to collagenous matrices modulates the tissue response in
rats. Biomaterials 2000;21:1689 1699.
Girton TS, Oegema TR, Tranquillo RT. Exploiting glycation to
stiffen and strengthen tissue equivalents for tissue engineering.
J Biomed Mater Res 1999;46:8792.
Noishiki Y, Ma XH, Yamane Y, Satoh S, Okoshi T, Takahashi K,
et al. Succinylated collagen crosslinked by thermal treatment
for coating vascular prostheses. Artif Organs 1998;22:672 680.
Bellincampi LD, Dunn MG. Effect of crosslinking method on
collagen berbroblast interactions. J Appl Polym Sci 1998;
63:14231428.

21.

22.

23.
24.
25.
26.

27.
28.

29.

30.

31.

32.

33.

34.

Rault I, Frei V, Herbage D, Abdul-Malak N, Huc A. Evaluation

of different chemical methods for crosslinking collagen gels,
lms, and sponges. J Mater Sci Mater Med 1996;7:215222.
Simmons DM, Kearney JN. Evaluation of collagen cross-linking techniques for the stabilization of tissue matrices. Biotechnol Appl Biochem 1993;17:2329.
Speer DP, Chvapil M, Eskelson CD, Ulreich J. Biological effects
of residual glutaraldehyde in glutaraldehyde-tanned collagen
biomaterials. J Biomed Mater Res 1980;14:753764.
van Luyn MJ, van Wachem PB, Damink LO, Dijkstra PJ, Feijen
J, Nieuwenhuis P. Relations between in vitro cytotoxicity and
crosslinked dermal sheep collagens. J Biomed Mater Res 1992;
26:10911110.
Huang LL, Sung HW, Tsai CC, Huang DM. Biocompatibility
study of a biological tissue xed with a naturally occurring
crosslinking reagent. J Biomed Mater Res 1998;42:568 576.
Nishi C, Nakajima N, Ikada Y. In vitro evaluation of cytotoxicity of diepoxy compounds used for biomaterial modication.
J Biomed Mater Res 1995;29:829 834.
Hey KB, Lachs CM, Raxworthy MJ, Wood EJ. Crosslinked
brous collagen for use as a dermal implant: Control of the
cytotoxic effects of glutaraldehyde and dimethylsuberimidate.
Biotechnol Appl Biochem 1990;12:8593.
Koob TJ, Willis TA, Hernandez DJ. Biocompatibility of NDGApolymerized collagen bers. I. Evaluation of cytotoxicity with
tendon broblasts in vitro. J Biomed Mater Res 2001;56:3139.
Greenberg CS, Birckbichler PJ, Rice RH. Transglutaminases:
Multifunctional cross-linking enzymes that stabilize tissues.
FASEB J 1991;5:30713077.
Aeschlimann D, Paulsson M. Transglutaminases: Protein
cross-linking enzymes in tissues and body uids. Thromb Haemost 1994;71:402 415.
Folk JE, Chung SI, editors. Transglutaminases. New York: Academic Press, Inc.; 1985.
Kagan HM, Trackman PC. Properties and function of lysyl
oxidase. Am J Respir Cell Mol Biol 1991;5:206 210.
Kagan HM. Lysyl oxidase: Mechanism, regulation and relationship to liver brosis. Pathol Res Pract 1994;190:910 919.
Smith-Mungo LI, Kagan HM. Lysyl oxidase: properties, regulation and multiple functions in biology. Matrix Biol 1998;16:
387398.
Krawetz SA. The origin of lysyl oxidase. Comp Biochem
Physiol Biochem Mol Biol 1994;108:117119.
Sakiyama-Elbert SE, Hubbell JA. Development of brin derivatives for controlled release of heparin-binding growth factors.
J Controlled Release 2000;65:389 402.
Mosher DF, Schad PE. Cross-linking of bronectin to collagen
by blood coagulation Factor XIIIa. J Clin Invest 1979;64:781
787.
Kaartinen MT, Pirhonen A, Linnala-Kankkunen A, Maenpaa
PH. Cross-linking of osteopontin by tissue transglutaminase
increases its collagen binding properties. J Biol Chem 1999;274:
1729 1735.
Sakiyama SE, Schense JC, Hubbell JA. Incorporation of heparin-binding peptides into brin gels enhances neurite extension: an example of designer matrices in tissue engineering.
FASEB J 1999;13:2214 2224.
Schense JC, Hubbell JA. Cross-linking exogenous bifunctional
peptides into brin gels with factor XIIIa. Bioconjug Chem
1999;10:75 81.
Sperinde JC, Grifth LG. Synthesis and characterization of
enzymatically-crosslinked poly(ethylene glycol) hydrogels.
Macromolecules 1997;30:52555264.
Vorp DA, Okano T, Schiro BJ, Campbell PG, Weiss LE. Effect of
construct constituents on the burst strength of a bone marrow
stromal cell-based tissue-engineered blood vessel. Cardiovasc
Pathol 2000;9:196.

762

35.

36.

37.

38.

ORBAN ET AL.

Bubnis WA, Ofner CM 3rd. The determination of epsilonamino groups in soluble and poorly soluble proteinaceous
materials by a spectrophotometric method using trinitrobenzenesulfonic acid. Anal Biochem 1992;207:129 133.
Mosmann T. Rapid colorimetric assay for cellular growth and
survival: Application to proliferation and cytotoxicity assays.
J Immunol Methods 1983;65:55 63.
Petite H, Duval JL, Frei V, Abdul-Malak N, Sigot-Luizard MF,
Herbage D. Cytocompatibility of calf pericardium treated by
glutaraldehyde and by the acyl azide methods in an organotypic culture model. Biomaterials 1995;16:10031008.
Sasaki CY, Passaniti A. Identication of anti-invasive but noncytotoxic chemotherapeutic agents using the tetrazolium dye
MTT to quantitate viable cells in Matrigel. Biotechniques 1998;
24:1038 1043.

39.

40.

41.

42.

Egawa M, Hisazumi H, Uchibayashi T, Tanaka M, Sasaki T.

Comparative study of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide and tritiated thymidine in a chemosensitivity test using collagen gel matrix. Urol Res 1993;21:83
88.
LHeureux N, Paquet S, Labbe R, Germain L, Auger FA. A
completely biological tissue-engineered human blood vessel.
FASEB J 1998;12:4756.
Niklason LE, Gao J, Abbott WM, Hirschi KK, Houser S, Marini
R, et al. Functional arteries grown in vitro. Science 1999;284:
489 493.
Hirai J, Matsuda T. Self-organized, tubular hybrid vascular
tissue composed of vascular cells and collagen for lowpressure-loaded venous system. Cell Transplant 1995;4:597
608.