Cell Transplantation, Vol. 16, pp. 639–647, 2007 0963-6897/07 $90.00 + .

00
Printed in the USA. All rights reserved. E-ISSN 1555-3892
Copyright  2007 Cognizant Comm. Corp. www.cognizantcommunication.com

Cryopreservation-Induced Nonattachment of Human Hepatocytes:

Role of Adhesion Molecules
Claire Terry, Robin D. Hughes, Ragai R. Mitry, Sharon C. Lehec, and Anil Dhawan

Institute of Liver Studies, King’s College London School of Medicine at King’s College Hospital, London, UK

Good quality cryopreserved human hepatocytes are becoming an important source for clinical hepatocyte
transplantation. However, the process of cryopreservation leads to both structural and functional impairment
of hepatocytes. The aim of this study was to investigate the mechanisms of cryopreservation-induced nonat-
tachment in human hepatocytes. Hepatocytes were cryopreserved after isolation from unused donor liver
tissue. Cell attachment to collagen-coated plates was measured. A cDNA gene array system for 96 cell
adhesion-related molecules was used to determine mRNA expression in fresh and cryopreserved hepatocytes.
Two cell adhesion molecule proteins were investigated further: β1-integrin, a cell-matrix adhesion molecule,
and E-cadherin, a cell–cell adhesion molecule. Attachment efficiency was significantly decreased after cryo-
preservation of human hepatocytes. Twenty-two genes were downregulated after cryopreservation including
integrins, cadherins, catenins, and matrix metalloproteinases (MMPs). β1-Integrin gene and protein expres-
sion were significantly decreased in cultured cryopreserved hepatocytes compared to fresh hepatocytes.
There was a significant correlation between loss of β1-integrin and attachment in cryopreserved cells. Degra-
dation of E-cadherin was increased in cryopreserved hepatocytes. The process of cryopreservation leads to
downregulation of cell adhesion molecules at the gene and the cellular level. New cryopreservation protocols
are needed to prevent these effects on cell attachment.

Key words: Hepatocyte; Cryopreservation; Attachment; β1-Integrin; E-cadherin

INTRODUCTION nisms involved in CINA may lead to cryopreservation

Human hepatocytes are being used in clinical cell
transplantation for the treatment of liver disease. Cryo-
preserved hepatocytes are an important resource for both
clinical and experimental usage. It is well known that
cryopreservation reduces hepatocyte viability and meta-
bolic function with the cell attachment efficiency af-
fected by cryopreservation to a greater extent than other
cell metabolic functions (3,4,7,10,16,23,26). This effect
has been described as cryopreservation-induced nonat-
tachment (CINA). The ability of hepatocytes to attach
after cryopreservation will be important for hepatocyte
transplantation, so that when cells are infused into a pa-
tient’s liver, they will adhere to the sinusoidal endothe-
lium and subsequently become engrafted.
The attachment of hepatocytes requires reestablish-
ment of cellular contacts, which are severed on their iso-
lation from the liver tissue (1). The addition of a further
damaging step (i.e., cryopreservation) to the process be-
fore culture understandably puts additional stress upon
the cellular systems. Better understanding of the mecha-
protocols that can specifically maintain the attachment
efficiency of fresh hepatocytes. Significantly higher sin-
gle and double DNA strand breaks were observed in
cryopreserved rat hepatocytes compared to fresh hepato-
cytes (21). This damage could affect gene expression
related to cell attachment molecules and thus be a possi-
ble cause for CINA.
Integrins are heterodimers consisting of α and β sub-
units, divided into subfamilies that share the β subunit
(8,9). β1-Integrin is a cell substrate adhesion molecule.
Integrins are expressed as constitutively active or inac-
tive receptors for extracellular matrix (ECM) ligands
such as collagen or fibronectin. Integrins are expressed
in a cell-type and stage-specific manner, with one group
of integrins associated with migration and proliferation
in various types of cells. Many integrins bind the Arg-
Gly-Asp (RGD) tripeptide attachment sequence but rec-
ognize that sequence differentially, such that some bind
primarily to fibronectin and others to vitronectin (25).
E-cadherin, a 120-kDa transmembrane glycoprotein,
is an example of a cell–cell adhesion molecule, and is

Received November 17, 2006; final acceptance March 23, 2007.

Address correspondence to Professor Anil Dhawan, Paediatric Liver Service, King’s College Hospital, Denmark Hill, London SE5 9RS, UK. Tel:
+44 (0)20-3299-3578; Fax: +44 (0)20-3299-4228; E-mail: anil.dhawan@kcl.ac.uk

639
640
the most predominant cadherin in the liver epithelium.
The cytoplasmic domain is directly associated with α-
and β-catenin and through these proteins it is indirectly
associated with proteins that link with the transmem-
brane E-cadherin to the actin cytoskeleton (18). The
degradation of E-cadherin can have important conse-
quences for the cells. The extracellular domain of E-
cadherin is proteolytically cleaved from some cancer
cells promoting metastasis, by a membrane-bound met-
alloproteinase to yield the soluble form (11).
In the present study, human hepatocytes were used to
investigate CINA using a cDNA gene array system, and
the two cell adhesion molecules (β1-integrin and E-
cadherin) were selected for preliminary investigation at
the cellular level in fresh and cryopreserved hepatocytes.
MATERIALS AND METHODS
Hepatocyte Isolation
Human liver tissue was obtained from donor tissue
rejected or unused for orthotopic liver transplantation at
King’s College Hospital. All tissue was consented for
research in accordance with the Research Ethics Com-
mittee of King’s College Hospital. Isolation of human
hepatocytes was carried out using a collagenase (Type
P, Roche Diagnostics Ltd, East Sussex, UK) perfusion
method (29) with some modifications (19).
Hepatocyte Cryopreservation
Hepatocytes (3 × 106 viable cells/ml) were cryopre-
served in UW solution (Bristol-Myers Squibb Pharma
Ltd, Hounslow, UK) containing 10% DMSO (v/v final
concentration) immediately after isolation. Freezing was
carried out in a controlled rate freezer (Kryo 10, Series
III, Planer Products Ltd, Middlesex, UK) using a step-
wise freezing protocol based on that of Diener et al. (5)
with some modifications (30). Frozen cells were stored
at −140°C for 2 weeks until thawing, by the method of
Steinberg et al. (28) using Williams E medium contain-
ing 10% FCS as the thawing medium.
Cell Function and Attachment Assays
Cell viability was determined using the trypan blue
exclusion method. Hepatocytes were plated (30,000 via-
ble cells/well) in 96-well flat-bottomed collagen-coated
plates (BiocoatTM, BD Bioscience, Oxford, UK). Culture
media consisted of Williams E medium containing 10%
fetal calf serum, penicillin (50 U/ml) and streptomycin
(50 µg/ml), and L-glutamine (2 mM) at 37°C in 95%
O2/5% CO2. After 24 h of culture the assays were per-
formed. LDH leakage from the hepatocytes was mea-
sured using a CytoTox 96 Non-Radioactive Cytotoxic-
ity Assay Kit (Promega, Southampton, UK). The LDH
contents of supernatant and cell lysate samples were de-
termined and LDH leakage calculated as a percentage of
TERRY ET AL.
the total. Albumin production was determined using a
Bethyl Human Albumin ELISA Quantification Kit
(BioGnosis, Hailsham, UK). Attachment efficiency was
determined from the protein content (17) of attached
cells relative to that of the initial number of cells plated.
The degree of cell attachment was also followed serially
and determined at 24, 48, 72, and 96h of culture for the
corresponding fresh and cryopreserved hepatocytes.
Total RNA Preparation
Total RNA samples were prepared from liver tissue
before isolation of hepatocytes, freshly isolated hepato-
cytes, cryopreserved hepatocytes 30 min after thawing
being left at room temperature, and after culture for 24
h. RNA was prepared from 5 × 106 viable fresh or cryo-
preserved hepatocytes using the RNeasy Mini Kit pro-
tocol for isolation of total RNA from animal cells (Qia-
gen, Crawley, West Sussex, UK). The RNA samples
were stored at −80°C until analysis. RNA preparations
from three hepatocyte batches both fresh and after cryo-
preservation were pooled to give the final fresh and
cryopreserved RNA samples for analysis. RNA purity
was assessed by the absorbance ratio at 260 nm/280 nm
(values 2.0 ± 0.1) and the presence of discrete bands
when run on an agarose gel.
Gene Array
A SuperArray cDNA GEArrayTM Q Series Human
Extracellular Matrix and Adhesion Molecules Gene
Array (obtained from tebu-bio, Peterborough, Cam-
bridgeshire, UK) for 96 adhesion-related genes was used
to compare the gene expression of fresh and cryopre-
served human hepatocytes. Total RNA was reverse-tran-
scribed to prepare 32P-labeled cDNA using the Super-
array kit. Briefly, samples were prepared by adding 5
µl RNA sample, 3 µl buffer A, and 2 µl dH2O. The
probe mix was prepared by adding 8 µl 5× GEA labeling
buffer, 6 µl [α-32P]dCTP (10 µCi/µl, Amersham Interna-
tional, Buckinghamshire, UK), 2 ml RNase inhibitor, 2
µl reverse transcriptase (50 U/µl), and 2 µl RNase-free
H2O to a tube and mixed. The sample mix was heated
to 70°C for 3 min and then cooled to 42°C (Teche PHC-
3 Thermal Cycler, Princeton, NJ, USA). Then 10 µl
probe mix (42°C) was added to each sample and incu-
bated for 25 min. The labeling reaction was stopped by
adding 2µl of 10× Stop Solution and the 32P-labeled
cDNA was denatured by heating to 94°C for 5 min and
then chilled quickly on ice. To each rehybridized array
tube (one for fresh and one for cryopreserved sample)
750 µl hybridization solution and 100 µl sample mix
was added. The array tubes were mixed at 60°C over-
night. The arrays were washed twice with prewarmed
(60°C) 2× SSC + 1% SDS for 5 min each and then twice
with prewarmed (60°C) 0.1× SSC + 0.5% SDS for 5
ATTACHMENT OF CRYOPRESERVED HEPATOCYTES
min each. The arrays were exposed on photographic film
at −140°C for varying times (2–48 h) to give appropriate
exposure.
Arrays were analyzed using Scanlyze (www.micro
arrays.org/software.html). The array image was recorded
and converted into TIFF-type image files of the nega-
tives of the original scans. A grid was laid over the im-
age with each cell of the grid covering one spot of the
original array. The density information for each gene
was then extracted from the array image. The back-
ground intensity was subtracted from the signal inten-
sity, giving an expression value for each spot. Any re-
sulting expression value less than 1500 was discounted.
Normalization was carried out using the housekeeping
genes provided on the arrays [glyceraldehyde-3-phos-
phate dehydrogenase (GAPDH), peptidylprolyl isomer-
ase A (cyclophilin A), ribosomal protein L13a, and β-
actin].
Northern Blotting
Samples of total RNA (10 µg) were loaded on to a
1.5% agarose, 6.66% formaldehyde gel in MOPS/EDTA
buffer and run in the electrophoresis tank. The RNA on
the gel was transferred to a Hybond N+ (Amersham In-
ternational) membrane. The membrane was washed and
left to dry at room temperature. The RNA was then fixed
to the membrane by UV cross-linking. The membrane
was transferred to a small glass hybridization tube and
was prehybridized in 10 ml of hybridization buffer con-
taining 100 µg denatured single-stranded DNA. The
probe for β1-integrin (458 bp) was generated using a
Human Integrin-II Multigene-12TM RT-PCR Profiling
Kit (SuperArray) according to the manufacturer’s proto-
col and was then radioactively labeled with [32P]dCTP
(1.85 MBq). The denatured probe was added to the hy-
bridization solution containing the membrane and left
overnight at 42°C. The membrane was washed and left
with photographic film for 24-h exposure at −80°C.
Soluble E-Cadherin Release
Soluble E-cadherin released into the 24-h culture su-
pernatant from fresh and cryopreserved human hepato-
cytes was measured using a colorimetric ELISA kit
(Zymed Laboratories, Inc, Cambridge, UK). Fresh and
cryopreserved hepatocytes were plated as in the attach-
ment assays. After 24 h, plates were centrifuged to pellet
unattached cells and supernatant samples were taken for
soluble E-cadherin analysis. In other experiments, the
attached and unattached cells (for both fresh and cryo-
preserved cells) were collected separately. After further
culture, supernatant samples were then taken for soluble
E-cadherin analysis. The absorbance of each well was
measured at 450 nm using a Dynex MRX Microplate
Reader (Dynex Technologies Ltd, Worthing, West Sus-
641
sex, UK). The concentration of soluble E-cadherin was
calculated from a standard curve and expressed as ng/h/
mg cell protein.
β1-Integrin Protein Expression
The expression of β1-integrin by fresh and cryopre-
served human hepatocytes was measured using a colori-
metric β1-integrin-mediated cell adhesion kit (Chemicon
International Inc., Hampshire, UK). Monoclonal anti-
bodies against human β1-integrin are immobilized on a
microplate to capture cells expressing β1-integrin on the
cell surface and this is quantitated. The absorbance val-
ues of both control and test plates were measured at 540
nm using a Dynex MRX Microplate Reader. By sub-
tracting the results of the control plate from the test
plate, the relative amount of β1-integrin expression
could be determined.
Fresh and cryopreserved human hepatocytes were
also stained for β1-integrin. Hepatocytes (1 × 105) were
pelleted by centrifugation at 1400 rpm for 5 min at 4°C
and the supernatant discarded. The pellets were washed
with 200 µl PBS containing 5% FCS by centrifugation
at 1400 rpm for 5 min at 4°C. This pellet was resus-
pended in a small amount of the supernatant (⬃100 µl).
One microliter of mouse anti-human CD29 (β1-inte-
grin)-RPE (Serotec Ltd, Oxford, UK) was added and left
to incubate in the dark for 30 min on ice. For control
staining, Mouse IgG1 Negative Control: RPE (Serotec
Ltd) was used. Thirty microliters of VectaShield
Mounting Medium (Vector Laboratories Ltd, Peterbor-
ough, UK) was added to each tube and samples were
kept in the dark on ice until microscopy. Hepatocytes
were then visualized using a fluorescence microscope
(Leica Microsystems Ltd, Milton Keynes, UK) using the
appropriate filter (RPE).
Statistical Analysis
Results are presented as mean ± SD. Statistical analy-
sis of data was performed by ANOVA with repeated
measurements and Pearson’s Correlation Test.
RESULTS
The viability of hepatocytes was significantly de-
creased after cryopreservation (fresh = 74 ± 13%; cryo-
preserved = 52 ± 10%, p < 0.001, n = 8). The leakage of
LDH was similar after cryopreservation (fresh = 17.9 ±
3.6%; cryopreserved = 20.7 ± 3.1%). Albumin secretion
was significantly reduced by cryopreservation (fresh =
10.4 ± 1.1 µg/h mg protein; cryopreserved = 4.5 ± 1.7
µg/h mg protein, p < 0.001). The attachment efficiency
was significantly decreased after cryopreservation (fresh =
53 ± 15%; cryopreserved = 39 ± 14%, p = 0.03). Mea-
surements of cell attachment at 24, 48, 72, and 96h after
plating showed that with both fresh and cryopreserved
642
hepatocytes there was a decline in attachment with time
(Fig. 1). The extent of loss of attachment was greater
and more rapid in cryopreserved hepatocytes compared
to fresh hepatocytes. The attachment of hepatocytes at
48 h (30 ± 5%, p = 0.03), 72 (24 ± 5%, p = 0.01), and
96 h (5 ± 1%, p = 0.002) postplating was significantly
lower than at 24 h (40 ± 9%) postplating.
Gene Microarray
From the microarrays, 22 of the 96 genes tested were
found to be downregulated after cryopreservation of hu-
man hepatocytes relative to expression in fresh hepato-
cytes (Table 1). There were no genes detected as signifi-
cantly upregulated. There was no significant difference
in the expression of the four housekeeping genes in-
cluded on the array between fresh and cryopreserved he-
patocytes (84–114% of fresh expression). Gene expres-
sion was considered to be significantly different when
the reciprocal ratio was >2 (27). The genes downregu-
lated included a number from the same molecular fami-
lies: catenins (α1, β1, δ1); cathepsins (D, L); integrins
(α2b, α3, α6, α9, β1); and metalloproteinases (1 and 26).
β1-Integrin Expression
The preliminary results of Northern blotting showed
that mRNA expression of β1-integrin was ⬃50% lower
Figure 1. Time course of attachment of fresh (filled circles) and cryopreserved (open circles)
human hepatocytes after plating on collagen-coated tissue culture plates. Mean ± SD of three ex-
periments. Values at 48, 72, and 96 h were compared to the corresponding 24-h attachment values.
Significance: *p < 0.05; **p < 0.01.
TERRY ET AL.
in cryopreserved hepatocytes compared to fresh hepato-
cytes (Fig. 2). This was reduced further after 24 h in
culture. The mRNA expression was lower in the original
liver tissue prior to isolation of hepatocytes.
The β1-integrin protein-mediated cell adhesion was
significantly lower in cryopreserved hepatocytes (0.085 ±
0.034 OD) compared to fresh hepatocytes (0.198 ± 0.022,
p < 0.001). There was a significant correlation between
β1-integrin expression and the attachment efficiency of
both fresh and cryopreserved hepatocytes (Fig. 3).
Staining of human hepatocytes with β1-integrin anti-
body showed that fresh human hepatocytes had a greater
and more uniform expression of β1-integrin than cryo-
preserved hepatocytes, particularly on the cell surface
(Fig. 4).
Soluble E-Cadherin Release
There was a significantly higher concentration of sol-
uble E-cadherin released into the culture media from
cryopreserved human hepatocytes (27.8 ± 4.8 ng/h/mg
protein, n = 27) compared to fresh hepatocytes (16.7 ±
5.6 ng/h/mg protein, n = 8, p < 0.001) in the first 24 h
of culture. Subsequently there was a significantly higher
concentration of soluble E-cadherin released into the su-
pernatant from both fresh (24.3 ± 7.6 ng/h/mg protein)
and cryopreserved (23.1 ± 7.0 ng/h/mg protein) non-
ATTACHMENT OF CRYOPRESERVED HEPATOCYTES 643

Table 1. Attachment-Related Genes Downregulated oikis” occurs in cells detached from their matrix (31)
in Human Hepatocytes After Cryopreservation and this may be activated during cell isolation, and thus
before cryopreservation. Loss of hepatocyte function
% of Fresh Reciprocal
Gene Expression Ratio
due to dedifferentiation in culture may also be involved
in the loss of attachment.
E-cadherin (epithelial) 16 6.41 The present study is the first investigation into the
Carcinoembryonic antigen-related effects of cryopreservation on the gene expression of
cell adhesion molecule 5 38 2.66 adhesion-related molecules in hepatocytes. Cryopreser-
Catenin α1 3 28.90 vation stress has been investigated previously using a
Catenin β1 53 1.88 DNA microarray system in yeast cells (22). In the pres-
Catenin δ1 47 2.11 ent study preliminary investigation of the mechanisms
Cathepsin D 30 3.33
of CINA using a specific gene array system identified
Cathepsin L 16 6.38
some target genes/molecules for further investigation. A
Fibrinogen B (β polypeptide) 25 3.93
Fibronectin 1 32 3.12 total of 22 genes (from a panel of 96 adhesion-related
Integrin α2b 36 2.75 genes) were downregulated after cryopreservation of hu-
Integrin α3 39 2.56 man hepatocytes. None of the genes tested were upregu-
Integrin α6 8 12.78 lated after cryopreservation. From these results two ad-
Integrin α9 2 53.84 hesion molecules (β1-integrin and E-cadherin) were
Integrin β1 48 2.07 selected for further investigation, particularly as reagents
Matrix metalloproteinase 1 60 1.67 were available for both the gene and cellular level.
Matrix metalloproteinase 26 60 1.68 The downregulation of the expression of β1-integrin
Platelet/endothelial cell adhesion gene found using the microarrays was confirmed in
molecule 13 7.83
Northern blots, which showed that cryopreserved hepa-
Secreted protein, acidic, cysteine
tocyte suspensions have lower expression of β1-integrin
rich 40 2.49
Secreted phosphoprotein 1 55 1.83 RNA than fresh hepatocyte suspensions. In addition, β1-
Thrombospondin 1 14 7.04 integrin expression was even lower in cryopreserved
Tissue inhibitor of hepatocytes after 24 h of culture. Thus, the loss of β1-
metalloproteinase 3 54 1.84 integrin continues even when the hepatocytes are re-
Vitronectin 16 6.11 attached to collagen-coated plates in culture. The culture
conditions of cells can influence the expression of β1-
Results are presented as a percentage of the fresh value. A reciprocal
ratio of >2 is considered significant (27). integrin (20). Specific β1-integrin-mediated attachment
to type I collagen was involved in induction and mainte-
nance of liver-specific functions of collagen-sandwiched
attached hepatocytes compared to fresh (4.1 ± 2.4 ng/h/ hepatocytes. Inclusion of additional culture media com-
mg protein, p < 0.01) and cryopreserved (6.1 ± 3.7 ng/ ponents or different attachment substrates may be able
h/mg protein, p < 0.01) attached hepatocytes (Fig. 5). to limit the loss of β1-integrin and therefore improve the
There was no correlation between the E-cadherin con- attachment of cryopreserved hepatocytes in culture.
centration released and the attachment efficiency of ei- In the Northern blotting experiments, the expression
ther fresh or cryopreserved hepatocytes. of β1-integrin in the original liver tissue gave the lowest
relative hybridization signal compared to the cell sam-
DISCUSSION ples. The higher level of expression in isolated cells may
The attachment efficiency of cryopreserved human be a response to the effects of the isolation and cryopres-
hepatocytes isolated from unused donor liver was found ervation processes. In addition, staining of hepatocytes
to be significantly reduced compared to the correspond- for β1-integrin gave a visual indication of the loss of β1-
ing fresh cells. These results are consistent with the find- integrin from the cell surface with greater intracellular
ings of many authors including those of Zvibel et al. localization after cryopreservation. β1-Integrin staining
(31), who showed that human hepatocytes can have a was much more diffuse and patchy in cryopreserved he-
high viability (78–99%) yet the attachment efficiency to patocytes than seen in fresh hepatocytes. The highly sig-
collagen-coated plates is low (5–35%). Cryopreserved nificant correlation between β1-integrin protein expres-
hepatocytes initially attached but then detached after sion on hepatocytes and cell attachment gives strong
overnight culture, as found by others (15). One likely evidence for a key role of β1-integrin in adherance to
contribution to this loss of hepatocyte function is cell collagen containing matrices. This is consistent with the
death by apoptosis, as has been reported after cryopres- recent in vitro results where anti-β1-integrin antibodies
ervation (6). A specific form of apoptosis, termed “an- blocked the attachment of rat hepatocytes to collagen-
644 TERRY ET AL.

Figure 2. The relative intensity of hybridization signals of β1-integrin mRNA in human liver
tissue, fresh hepatocytes, and cryopreserved hepatocytes postthaw and after 24 h in culture.

Figure 3. Correlation between β1-integrin protein expression and attachment efficiency of fresh
(filled circles) and cryopreserved (open circles) human hepatocytes. Significance: **p < 0.01.
ATTACHMENT OF CRYOPRESERVED HEPATOCYTES 645

Figure 4. Micrograph of fresh (a) and cryopreserved (b) human hepatocytes stained with β1-
integrin antibody. β1-Integrin was detected in human hepatocytes using RPE-conjugated anti-β1-
integrin antibodies. Original magnification 100×.

coated surfaces and also reduced apoptosis (14). In the
context of hepatocyte transplantation infusion of colla-
gen into the liver prior to cell infusion enhanced cell
engraftment, indicating the integrin-dependent nature of
this process (24). It is interesting that different regres-
sion lines were obtained with fresh and cryopreserved
hepatocytes. This at first seems difficult to explain but
probably suggests that β1-integrin, in the form of hetero-
dimers containing different α integrins that are not in-
volved in cell adhesion to collagen are lost after cryo-
preservation of fresh cells. The antibody in the kit used
detects any α heterodimer with β1-integrin, thus, sup-
porting this possibility.
Although integrin expression on hepatocytes would
646
Figure 5. Release of soluble E-cadherin from attached and unattached, fresh and cryopreserved
human hepatocytes. Mean ± SD (n = 8). Value significantly different from corresponding attached
hepatocytes: **p < 0.01.
appear to be important for good attachment and engraft-
ment after hepatocyte transplantation, the situation is
more complex. The formation of hepatocyte aggregates
as a result of cell to cell adhesion leads to thrombi in
portal venules and loss of cellular function due to hy-
poxia (12). Blocking of α1β1-integrin reduced aggre-
gate formation and improved the survival of hepato-
cytes, but also blocked hepatocyte attachment to the
liver parenchyma (13), which is likely to limit eventual
cell engraftment. Other mechanisms may be involved,
and injected hepatocytes may die through apoptosis as a
result of lack of anchorage to a matrix and thus ligation
of α1- and β1-integrin may prevent this. Further investi-
gation of the specific forms of β1-integrin involved is
required.
The expression of E-cadherin was downregulated at
the gene level together with increased cleavage of the
cytoplasmic domain of the cellular protein in cryopre-
served hepatocytes. Thus, cryopreservation may also in-
crease degradation of adhesion proteins. However, no
correlation between E-cadherin and hepatocyte attach-
ment was found. The loss of E-cadherin-related cell–cell
adhesion could suggest that although cells are able to
attach to the matrix initially, they are unable to form
cell–cell attachments and as a result undergo apoptosis
or, more specifically, anoikis (31). On the other hand,
loss of E-cadherin could reduce the formation of intra-
vascular hepatocyte aggregates, which is detrimental to
cell engraftment after transplantation (12). Hepatocytes
TERRY ET AL.
cocultured with E-cadherin-presenting chaperone cells
secreted higher levels of albumin and urea than those
cocultured with cadherin-deficient chaperones, demon-
strating the importance of E-cadherin-related cell–cell
contact on cell function (2). The role of E-cadherin in
hepatocyte engraftment is not clear and needs further
investigation.
In conclusion, cryopreservation of human hepato-
cytes reduces the expression of cell adhesion molecules
that could be involved in CINA. Further studies to in-
vestigate methods to modulate the effects of cryopreser-
vation on adhesion molecules would be worthwhile.
ACKNOWLEDGMENTS: We would like to acknowledge the
expertise and significant contribution of the late Dr Phillip
Murphy to the gene expression studies in this article. His un-
timely death was a great loss to our department. The study
was supported by Merck Sharp and Dohme Ltd (UK).
REFERENCES
1. Baker, T. K.; Carfagna, M. A.; Gao, H.; Dow, E. R.; Li,
Q.; Searfss, G. H.; Ryan, T. P. Temporal gene expression
analysis of monolayer cultured rat hepatocytes. Chem.
Res. Toxicol. 14:1218–1231; 2001.
2. Brieva, T. A.; Moghe, P. V. Functional engineering of he-
patocytes via heterocellular presentation of a homoadhe-
sive molecule, E-cadherin. Biotechnol. Bioeng. 76:295–
302; 2001.
3. Chesne, C.; Guyomard, C.; Fautrel, A.; Poullain, M. G.;
Fremond, B.; De Jong, H.; Guillouzo, A. Viability and
function in primary culture of adult hepatocytes from vari-
ous animal species and human beings after cryopreserva-
tion. Hepatology 18:406–414; 1993.
ATTACHMENT OF CRYOPRESERVED HEPATOCYTES
4. David, P.; Viollon, C.; Alexandre, E.; Azimzadeh, A.;
Nicod, L.; Wolf, P.; Jaeck, D.; Boudjema, K.; Richert, L.
Metabolic capacities in cultured human hepatocytes ob-
tained by a new isolating procedure from non-wedge
small liver biopsies. Human Exp. Toxicol. 17:544–553;
1988.
5. Diener, B.; Utesch, D.; Beer, N.; Durk, H.; Oesch, F. A
method for the cryopreservation of liver parenchymal cells
for studies of xenobiotics. Cryobiology 30:116–127;
1993.
6. Fu, T.; Guo, D.; Huang, X.; O’Gorman, M. R. G.; Huang,
L.; Crawford, S. E.; Soriano, H. E. Apoptosis occurs in
isolated and banked primary mouse hepatocytes. Cell
Transplant. 10:59–66; 2001.
7. Hengstler, J. G.; Utesch, D.; Steinberg, P.; Platt, K. L.;
Diener, B.; Ringel, M.; Swales, N.; Fischer, T.; Biefang,
K.; Gerl, M.; Bottger, T.; Oesch, F. Cryopreserved pri-
mary hepatocytes as a constantly available in vitro model
for the evaluation of human and animals drug metabolism
and enzyme induction. Drug Metab. Rev. 32:81–118;
2000.
8. Hynes, R. O. Integrins: Versatility, modulation, and sig-
naling in cell adhesion. Cell 69:11–25; 1992.
9. Hynes, R. O. Integrins: Bidirectional, allosteric signaling
machines. Cell 110:673–687; 2002.
10. Innes, G. K.; Fuller, B. J.; Hobbs, K. E. F. Functional
testing of hepatocytes following their recovery from cryo-
preservation. Cryobiology 25:23–30; 1988.
11. Ito, K.; Okamoto, I.; Araki, N.; Kawano, Y.; Nakao, M.;
Fujiyama, S.; Tomita, K.; Mimori, T.; Saya, H. Calcium
influx triggers the sequential proteolysis of extracellular
and cytoplasmic domains of e-cadherin, leading to loss of
beta-catenin from cell-cell contacts. Oncogene 18:7080–
7090; 1999.
12. Kocken, J. M.; Borel Rinkes, I. H.; Bijma, A. M.; de
Roos, W. K.; Bouwman, E.; Terpstra, O. T.; Sinaasappel,
M. Correction of an inborn error of metabolism by intra-
portal hepatocyte transplantation in a dog model. Trans-
plantation 15:358–364; 1996.
13. Kocken, J. M.; de Heer, E.; Borel Rinkes, I. H. M.; Sinaa-
sappel, M.; Terpstra, O. T.; Bruijn, J. A. Blocking of
alpha1 beta1 integrin strongly improves survival of hepa-
tocytes in allogeneic transplantation. Lab. Invest. 77:19–
28; 1997.
14. Kumaran, V.; Joseph, B.; Benten, D.; Gupta, S. Integrin
and extracellular matrix interactions regulate engraftment
of transplanted hepatocytes in the rat liver. Gastroenterol-
ogy 129:1643–1653; 2005.
15. Li, A. P. Human hepatocytes as an experimental system
for the evaluation of xenobiotics. In: Berry, M. N.;
Edwards, A. M., eds. The hepatocyte review. Netherlands:
Kluwer Academic Publishers; 2000:391–410.
16. Li, A. P.; Gorycki, P. D.; Hengstler, J. G.; Kedderis,
G. L.; Koebe, H. G.; Rahmani, R.; de Sousa, G.; Silva,
J. M.; Skett, P. Present status of the application of cryo-
preserved hepatocytes in the evaluation of xenobiotics:
Consensus of an international expert panel. Chem. Biol.
Interact. 121:117–123; 1999.
647
17. Lowry, O. H.; Roseburgh, N. J.; Farr, A. L.; Randall,
R. J. Protein measurement with the Folin phenol reagent.
J. Biol. Chem. 193:265–275; 1951.
18. Mitic, L. L.; Anderson, J. M. Molecular architecture of
tight junctions. Ann. Rev. Physiol. 60:121–142; 1998.
19. Mitry, R. R.; Hughes, R. D.; Aw, M. M.; Terry, C.; Mieli-
Vergani, G.; Girlanda, R.; Muiesan, P.; Rela, M.; Heaton,
N. D.; Dhawan, A. Human hepatocyte isolation and rela-
tionship of cell viability to early graft function. Cell
Transplant. 12:69–74; 2003.
20. Moghe, P. V.; Ezzell, R. M.; Toner, M.; Tompkins,
R. G.; Yarmush, M. L. Role of β1 integrin distribution in
morphology and function of collagen-sandwiched hepato-
cytes. Tissue Eng. 3:1–16; 1997.
21. Mueller, P.; Aurich, H.; Fleig, W.; Christ, B.; Luther, M.
DNA damage during cryopreservation of hepatocytes.
Hepatology 36:425A; 2002.
22. Odani, M.; Komatsu, Y.; Oka, S.; Iwahashi, H. Screening
of genes that respond to cryopreservation stress using
yeast DNA microarray. Cryobiology 47:155–164; 2003.
23. Ostrowska, A.; Bode, D. C.; Pruss, J.; Bilir, B.; Smith,
G. D.; Zeisloft, S. Investigation of functional and morpho-
logical integrity of freshly isolated and cryopreserved hu-
man hepatocytes. Cell Tissue Bank 1:55–68; 2000.
24. Pinkse, G. G.; Voorhoeve, M. P.; Noteborn, M.; Terpstra,
O. T.; Bruijn, J. A.; De Heer, E. Hepatocyte survival de-
pends on beta1-integrin-mediated attachment of hepato-
cytes to hepatic extracellular matrix. Liver Int. 24:218–
226; 2004.
25. Rouslahti, E. Integrins as signalling molecules and targets
for tumour therapy. Kidney Int. 51:1413–1417; 1997.
26. Ruegg, C. E.; Silber, P. M.; Mughal, R. A.; Ismail, J.; Lu,
C.; Bode, D. C.; Li, A. P. Cytochrome-P450 induction and
conjugated metabolism in primary human hepatocytes
after cryopreservation. In Vitro Toxicol. 10:217–222;
1997.
27. Shackel, N. A.; Gorrell, M. D.; McCaughan, G. W. Gene
array analysis and the liver. Hepatology 36:1313–1325;
2002.
28. Steinberg, P.; Fischer, T.; Kiulies, S.; Biefang, K.; Platt,
K. L.; Oesch, F.; Bottger, T.; Bulitta, C.; Kempf, P.; Hen-
gstler, J. Drug metabolizing capacity of cryopreserved hu-
man, rat, and mouse liver parenchymal cells in suspen-
sion. Drug Metab. Dispos. 27:1415–1422; 1999.
29. Strom, S. C.; Dorko, K.; Thompson, M. T.; Pisarov,
L. A.; Nussler, A. K. Large scale isolation and culture of
human hepatocytes. In: Franco, D.; Boudjema, K.; Varet,
B., eds. Ilots de Langerhans et Hepatocytes. Paris: Les
Editions INSERM; 1989:195–205.
30. Terry, C.; Mitry, R. R.; Lehec, S. C.; Muiesan, P.; Rela,
M.; Heaton, N. D.; Hughes, R. D.; Dhawan, A. The effects
of cryopreservation on human hepatocytes obtained from
different sources of liver tissue. Cell Transplant. 14:585–
594; 2005.
31. Zvibel, I.; Smets, F.; Soriano, H. Anoikis: Roadblock to
cell transplantation. Cell Transplant. 11:621–630; 2002.