optimization of the cryopreservation and thawing protocol for human hepatocytes for use in cell...

ORIGINAL ARTICLE

Optimization of the Cryopreservation andThawing Protocol for Human Hepatocytesfor Use in Cell TransplantationClaire Terry, Anil Dhawan, Ragai R. Mitry, Sharon C. Lehec, and Robin D. HughesInstitute of Liver Studies, King’s College London School of Medicine at King’s College Hospital,London, United Kingdom

Cryopreservation of human hepatocytes is important for their use in hepatocyte transplantation. On thawing, cryopreservedhepatocytes often have reduced viability and metabolic function in comparison with fresh cells. The aim of this study was tomodify the different steps in the standard cryopreservation procedure in an attempt to improve the overall outcome. Humanhepatocytes with a viability of 69% 6 SD 16% were isolated from donor livers with a collagenase perfusion technique. Dif-ferent cell densities, concentrations, rates, and methods of addition of dimethyl sulfoxide were tested for the freezing solu-tion. Modified controlled-rate freezer programs were tested to obtain a linear decrease in the temperature. Once they werefrozen, the storage time and thawing method for hepatocytes were investigated. The effects on thawed cell viability andattachment, lactate dehydrogenase release, cytochrome P450 1A1/2 activity, and albumin synthesis were determined. Theresults were used to produce an improved cryopreservation protocol suitable for good manufacturing practice conditions.With a cell density of 107 cells/mL in University of Wisconsin solution containing 300 mM glucose, 10% (vol/vol) dimethylsulfoxide was added dropwise over 5 minutes, and was immediately frozen. Thawing was done rapidly at 37�C, and dilutionwas performed with Eagle’s minimum essential medium containing 300 mM glucose and 4% human serum albumin. Hepa-tocytes could be stored at �140�C without significant further loss of function for up to 3 years. With this protocol, hepato-cytes had a viability of 52% 6 9%, an attachment efficiency of 48% 6 8%, and lactate dehydrogenase leakage of 17% 64%. This protocol is currently in use to cryopreserve hepatocytes for use in cell transplantation at our center. LiverTranspl 16:229-237, 2010. VC 2010 AASLD.

Received September 18, 2009; accepted November 3, 2009.

Cryopreservation of human hepatocytes is essentialfor their use in hepatocyte transplantation. Successfulcryopreservation allows hepatocytes to be availablefor the emergency treatment of acute liver failure andalso for planned and repeated treatments with cellsfrom the same donor for liver-based metabolic disor-ders. However, it is well known that cryopreservationaffects the viability and metabolic function of hepato-cytes on thawing, with the result that the thawed cellsare often not suitable for clinical use.

A considerable number of human hepatocyte cryo-preservation protocols have been reported for experi-mental use.1-10 Not surprisingly, one of the main deter-

minants of thawed cell function is the quality of thefresh hepatocytes before cryopreservation. This is oftenrelated to the nature of the liver tissue from which thehepatocytes are isolated, including factors such asthe age or condition of the donor, liver steatosis, andthe lengths of the cold and warm ischemia timesinvolved in procuring the liver.11 There are a number ofsteps in the cryopreservation process that can influ-ence the function of the thawed hepatocytes, such asthe cryoprotectant used, the method and media usedfor freezing, and finally the thawing of the cells.12

In a previous study, we found that, as a first step inthe cryopreservation process, pre-incubation of human

Abbreviations: CRF, controlled-rate freezer; CYP1A1/2, cytochrome P4501A1/2; DMSO, dimethyl sulfoxide; FCS, fetal calf serum;LDH, lactate dehydrogenase; UW, University of Wisconsin; WEM, William’s medium E.Claire Terry was supported by a grant from Merck Sharp & Dohme, Ltd.Address reprint requests to Robin D. Hughes, Institute of Liver Studies, King’s College London School of Medicine, Bessemer Road, London,SE5 9PJ, United Kingdom. Telephone: þ44-(0)20-3299-3137; FAX: þ44-(0)20-3299-3760; E-mail: [email protected]

DOI 10.1002/lt.21983Published online in Wiley InterScience (www.interscience.wiley.com).

LIVER TRANSPLANTATION 16:229-237, 2010

VC 2010 American Association for the Study of Liver Diseases.

hepatocytes at 4�C with the cytoprotective agents glu-cose, fructose, and lipoic acid improved hepatocytefunction on thawing.13 The aim of this study was toinvestigate systematically the different steps involvedin the hepatocyte cryopreservation and thawing pro-cess, including the cell density, addition of cryoprotec-tants, freezing temperature program, and thawingmethod. From this, an optimized protocol has beenproduced that can be carried out under good manufac-turing practice conditions with reagents that are ac-ceptable for use in clinical hepatocyte transplantation.

MATERIALS AND METHODS

Hepatocyte Isolation

Human liver tissue was obtained from donor tissuethat was rejected or unused for orthotopic liver trans-plantation at King’s College Hospital. All donor tissuewas delivered to the laboratory, flushed with Univer-sity of Wisconsin (UW) solution (Bristol-Myers SquibbPharma, Ltd., Hounslow, United Kingdom), and main-tained in this solution on ice. Consent was obtainedto use all the tissue for research in accordance withthe Research Ethics Committee of King’s College Hos-pital. Isolation of human hepatocytes was carried outwith a collagenase perfusion method with cell purifi-cation by slow-speed centrifugation14 with some mod-ifications.15 The obtained hepatocytes (fresh viability>50%) were used immediately in the cryopreservationexperiments described in this article.

Standard Hepatocyte Cryopreservation Protocol

This protocol was based on the literature published atthe starting time of this study in the context of clinicalapplication. It was the baseline against which furthermodifications were tested. The freezing medium con-sisted of UW solution containing 10% (vol/vol) di-methyl sulfoxide (DMSO; Sigma-Aldrich, Gillingham,United Kingdom). Hepatocytes in suspension werecentrifuged at 50g for 5 minutes at 4�C in a 2- or 5-mL cryovial, and the supernatant was removed. Thefreezing density was 3 � 106 viable hepatocytes/mL offreezing medium. Ice-cold UW solution was added tothe cell pellet to give a total volume (cells and freezingmedium) of either 1.8 (for 2-mL cryovials) or 4.5 mL(for 5-mL cryovials). DMSO was then added in a drop-wise manner to the cell suspension/freezing mediummixture to give a final concentration of 10% (vol/vol).The suspension was then kept on ice for a maximumof 5 minutes before the freezing process began. A con-trolled-rate freezer (CRF; Kryo 10, series III, PlanerProducts, Ltd., Middlesex, United Kingdom) was used.The chosen freezing protocol was adapted from that ofDiener et al.16 When the freezing protocol was fin-ished, the cryovials were immediately transferred to a�140�C freezer (Lab Impex Research, Ltd., East Sus-sex, United Kingdom) and stored until thawing.

The thawing protocol was based on that of Stein-berg et al.7 After 2 weeks of storage at �140�C, the

frozen cell suspension was rapidly thawed in a 37�Cwater bath with gentle agitation and diluted with amedium.

In Vitro Cell Viability and Function Assays

Cell viability was determined with the trypan blueexclusion method. Unless stated otherwise, all func-tional assays were performed on 30,000 viable cellsper well (8 wells per sample) in 96-well, flat-bottomed,collagen-coated plates (BiocoatTM, BD Bioscience,Oxford, United Kingdom) after 24 hours of culture.Hepatocytes were cultured in William’s medium E(WEM) containing 10% fetal calf serum (FCS), penicil-lin (50 U/mL) and streptomycin (50 lg/mL), and L-glutamine (2 mM) at 37�C in 95% O2/5% CO2. Theattachment efficiency was determined by the mea-surement of the protein content17 of attached cellsand the initial number of cells (30,000 total cells perwell). The lactate dehydrogenase (LDH) concentrationwas measured in supernatant and cell lysate sampleswith a CytoTox 96 nonradioactive cytotoxicity assaykit (Promega, Southampton, United Kingdom), whichallowed LDH leakage to be calculated as a percentageof the total. Cytochrome P4501A1/2 (CYP1A1/2) ac-tivity was determined with the ethoxyresorufin O-de-ethylase method.18 This assay was carried out on he-patocytes plated at a density of 150,000 viable cellsper well (a 24-well, collagen-coated plate) after 24hours of culture. A resorufin standard curve (0-800pmol) was used to calculate CYP1A1/2 activity, whichwas expressed as picomoles of resorufin produced perminute per milligram of cell protein. Albumin produc-tion was determined with a human albumin enzyme-linked immunosorbent assay quantification kit (BethylLaboratories, Inc., Biognosis, Hailsham, United King-dom). The albumin production was expressed as micro-grams of albumin produced per hour per milligram ofcell protein.

Modified Cryopreservation Protocols

For each experiment, one stage/parameter of thestandard protocol was modified at a time, and thethawed hepatocytes obtained were compared to thosefrom the standard protocol. Eight batches of humanhepatocytes were used with at least 3 replicates perexperiment.

Cell Density

Hepatocytes were resuspended at cell densities of 1, 2,3, 4, 5, and 10 � 106 viable cells/mL in UW solutioncontaining 10% DMSO in 2-mL cryovials [hepatocyteswere suspended in UW solution to a total volume of1.8 mL, and then 0.2 mL of DMSO was added dropwiseto each cryovial to give a final DMSO concentration of10% (vol/vol)]. UW solution was used as it was found tobe superior to Optimem, WEM, and L-15 media in pre-liminary experiments. DMSO was used as initial stud-ies investigated a range of permeating cryoprotectants

230 TERRY ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

(DMSO, 2,3-butanediol, glycerol, acetamide, and form-amide) and nonpermeating cryoprotectants (lactose, D-mannitol, raffinose, sucrose, and trehalose) and foundthat none of these alone were superior to UW solution(unpublished data).

DMSO Concentration

Human hepatocytes were cryopreserved in DMSOwith 1 of 6 concentrations [7.5, 10, 12.5, 15, 17.5, or20% (vol/vol)]. Hepatocytes were suspended in an ice-cold UW solution with 3 � 106 viable hepatocytes/mLof a 2-mL cryovial. DMSO was then added dropwise toeach cryovial to give the appropriate final concentra-tion for testing.

Rate of Addition of DMSO

DMSO (10%) was added to the cell/freezing mediumsuspension (3 � 106 viable cells/mL in UW solution)by 1 of 5 different methods. Cryopreservation was car-ried with the standard freezing protocol in the CRF:

1. Standard method. DMSO was dropwise added tothe cell suspension, and after 5 minutes on ice,the cells were cryopreserved.

2. Method of Utesch et al.19 Hepatocytes wereresuspended at a cell density of 3 � 106 viablecells/mL in UW solution. The suspension wasthen centrifuged (50g for 5 minutes at 4�C), anda volume equal to two-thirds of the original vol-ume (5 mL) of the supernatant was carefully dis-carded (ie, 3.33 mL). The cell pellet wasresuspended in the remaining UW solution, andto the suspension, an ice-cold UW solution con-taining 12% (vol/vol) DMSO was added up to50% of the original volume (ie, 0.84 mL wasadded). This resulted in a DMSO concentrationof 4%. After 5 minutes on ice, an ice-cold UW so-lution containing 16% (vol/vol) DMSO was addedup to the volume of the original cell suspension(ie, 2.5 mL was added). This resulted in a finalDMSO concentration of 10% (vol/vol). After 5minutes on ice, the cells were cryopreserved.

3. Ten-minute method. DMSO was dropwise addedto the cell suspension, and after 10 minutes onice, the cells were cryopreserved.

4. Twenty-minute method. DMSO was dropwiseadded to the cell suspension, and after 20 minuteson ice, the cells were cryopreserved.

5. Thirty-minutemethod. DMSOwas dropwise addedto the cell suspension, and after 30minutes on ice,the cells were cryopreserved.

CRF Protocols

Cell preparation was carried out according to thestandard method, and then 1 of 3 modified freezingprotocols was tested. Slight modifications were madeto the standard 9-step temperature change protocolwith changes to the shock cooling step and subse-quent cooling regimen with the aim of achieving a lin-

ear decrease in the cell suspension temperature. Spe-cifically, in step 4, �8�C to �28�C was tested atcooling rates of �35, �40, and �50�C/minute (stand-ard); in step 7, �28�C to �60�C was tested at �1 and�2�C/minute (standard); and in step 8, �5 and�10�C/minute (standard) were tested. The tempera-ture in the cell suspension was measured continu-ously with the machine’s temperature probe.

Storage

Five human hepatocyte batches were used to investi-gate the effects of the storage time at �140�C on cryo-preserved hepatocytes. Hepatocytes were cryopre-served by the standard method and stored in a �140�Cultralow-temperature freezer. A cryovial of each hepato-cyte batch was thawed after 1 week, 2 weeks, 1 month,3 months, 1 year, and 3 years of storage at �140�C. He-patocytes were thawed rapidly at 37�C, and the freezingmedium was diluted according to the standard proto-col. Viability and cell number measurements weremade, and hepatocytes were plated on collagen-coatedplates for functional assays after 24 hours.

Modified Thawing Protocol

Hepatocytes were cryopreserved with the standardprotocol. After 2 weeks of storage at �140�C, cryovialswere thawed rapidly in a 37�C water bath, and dilu-tion of the cell suspension was carried out with 1 of 4different protocols:

1. Immediate method. The thawed cell suspensionwas transferred to a tube containing 25 mL ofice-cold WEM containing 10% FCS (this took lessthan 1 minute) and mixed by inversion.

2. Constant method: The thawed cell suspensionwas transferred to an ice-cold tube, and 5 mL ofice-cold WEM containing 10% FCS was added tothe cell suspension and mixed by inversion. After5 minutes on ice, another 5 mL of ice-cold WEMcontaining 10% FCS was added to the suspen-sion and mixed by inversion. This was repeated3 times for a total added volume of 25 mL ofWEM containing 10% FCS (this took about20 minutes).

3. Dropwise method. The thawed cell suspensionwas transferred to an ice-cold tube, and 25 mLof WEM containing 10% FCS was added to thecell suspension in a dropwise manner (this tookabout 3 minutes).

4. Standard method. The thawed cell suspensionwas transferred to an ice-cold tube and dilutedaccording to the standard method of thawing.7

For every 1 mL of cell suspension thawed, thefollowing volume of thawing medium was addeddropwise with 5 minutes on ice between eachaddition: 0.5, 1, 2, 3, and 6 mL.

Cell suspensions were then centrifuged at 50g for 5minutes at 4�C, and the pellet was resuspended in aknown volume of WEM. Viability and cell number

CRYOPRESERVATION OF HUMAN HEPATOCYTES 231


measurements were made, and hepatocytes wereplated on collagen-coated plates for functional assaysafter 24 hours.

Cytoprotection During Thawing

Glucose (0-300 mM), fructose (0-300 mM), or a-lipoicacid (0-5 mM) was added to the thawing medium ofthe standard thawing method, and the effects on thethawed cell function were determined.

Optimized Cryopreservation and Thawing

Protocol for Clinical Use

The protocol was based on the combined results ofthe aforementioned experiments. Hepatocytes werecryopreserved at 1 � 107 cells/mL in UW solutioncontaining 300 mM glucose (50% dextrose injection)with 10% (vol/vol) DMSO added by the standarddropwise method. Blood cryobags (50 mL; BaxterHealth Care Corp., Deerfield, IL) filled with 10-mL cellsuspensions were used instead of cryovials as a resultof the larger number of cells to be cryopreserved.

The modified stepwise cooling program (with respectto the standard), as shown in Table 1, was used. Thethawing medium consisted of Eagle’s minimum essen-tial medium with the addition of 300 mM glucose and4% human serum albumin. The rapid thawing proto-col was based on that of Steinberg et al.7 Thawed he-patocytes were assessed for viability and plated for invitro assays of function.

Statistics

All results are presented as means and standard devi-ations. A statistical analysis of the data was per-formed by a comparison of means with an analysis ofvariance with repeated measurements.

RESULTS

Cell Density

Human hepatocytes cryopreserved at cell densities of1 and 2 � 106 viable cells/mL had significantly lower

viability (1 � 106 ¼ 30% 6 6%, P ¼ 0.01; 2 � 106 ¼35% 6 5%, P ¼ 0.02) and attachment efficiency (1 �106 ¼ 35% 6 6%, P ¼ 0.02; 2 � 106 ¼ 36% 6 6%, P ¼0.03) and significantly higher LDH leakage (1 � 106 ¼30% 6 3%, P ¼ 0.02; 2 � 106 ¼ 29% 6 3%, P ¼ 0.02)than standard hepatocytes (for those cryopreserved ata cell density of 3 � 106 cells/mL, viability ¼ 46% 65%, attachment efficiency ¼ 48% 6 5%, and LDHleakage ¼ 15% 6 2%). There was no effect of cell den-sity on the CYP1A1/2 activity or albumin productionof cryopreserved hepatocytes. There was no significantdifference between the viability and function of hepa-tocytes cryopreserved at cell densities of 4, 5 or 10 �106 viable cells/mL and those cryopreserved at a celldensity of 3 � 106 cells/mL.

DMSO Concentration

The concentration of the cryoprotectant added to thefreezing solution affected the viability and function ofhuman hepatocytes after cryopreservation (Fig. 1). All

TABLE 1. Optimized Controlled-Rate Freezer Protocol

for the Cryopreservation of Human Hepatocytes

Step

Starting

Temperature

(�C)

Rate

(�C/minute)

Time

(minutes)

End

Temperature

(�C)

1 8 �1 8 02 0 Hold 8 03 0 �2 4 �84 �8 �35 0.55 �285 �28 �2.5 2 �336 �33 þ2.5 2 �287 �28 �1 32 �608 �60 �10 4 �1009 �100 �20 2 �140

Figure 1. Effect of the cryoprotectant concentration(DMSO) on the viability and function of human hepatocytes.Human hepatocytes were cryopreserved in University ofWisconsin solution containing 7.5, 10 (standard), 12.5, 15,17.5, or 20% DMSO. Results are presented as means andstandard deviations of 3 experiments. Values are comparedto those of the standard (10% DMSO). Significance: *P <0.05 and ***P < 0.001.

232 TERRY ET AL.


concentrations were compared to the standard of 10%DMSO. The viability and attachment efficiency with12.5% DMSO (viability ¼ 49% 6 6%, attachment effi-ciency ¼ 39% 6 3%) and 15% DMSO (viability ¼ 45%6 4%, attachment efficiency ¼ 35% 6 4%) were notsignificantly different from the standard (viability ¼50% 6 8%, attachment efficiency ¼ 40% 6 4%). How-ever, the DMSO concentrations of 7.5% (P ¼ 0.02),17.5% (P ¼ 0.03), and 20% (P ¼ 0.01) yielded signifi-cantly lower viability than the standard. DMSO con-centrations of 7.5% (P ¼ 0.03), 17.5% (P ¼ 0.03), and20% (P ¼ 0.01) also yielded significantly lower attach-ment efficiency than the standard. LDH leakage wassignificantly higher in hepatocytes cryopreserved with7.5%, 15%, 17.5%, or 20% DMSO versus the stand-ard (30% 6 3%, P ¼ 0.02; 34% 6 4%, P ¼ 0.01; 40%6 3%, P < 0.001; and 46% 6 3%, P < 0.001, respec-tively). LDH leakage was similar in the standard (21%6 3%) and 12.5% DMSO hepatocytes (20% 6 4%).CYP1A1/2 activity was significantly lower only in he-patocytes cryopreserved with 7.5% DMSO (7.2 6 0.9pmol/minute/mg of protein, P ¼ 0.04) and 20%DMSO (7.0 6 0.8 pmol/minute/mg of protein, P <0.001) in comparison with the standard (8.2 6 1.1pmol/minute/mg of protein).

Rate of Addition of DMSO

There was no significant difference between the viabil-ity, attachment efficiency, LDH leakage, or CYP1A1/2activity of hepatocytes with respect to the standard(dropwise) method, Utesch et al.’s method, or the 10-minute acclimatization method in which DMSO wasadded to the freezing suspension. However, theattachment efficiency was significantly lower after the30-minute acclimatization method (41% 6 8%, P ¼0.009) in comparison with the standard method (60%6 9%). LDH leakage was significantly higher after the20-minute acclimatization method (25% 6 6%, P ¼0.04) and 30-minute acclimatization method (27% 66%, P ¼ 0.02) in comparison with the standardmethod (21% 6 5%).

CRF Protocols

The freezing temperature protocol shown in Table 1was closer to a linear decrease in the cell suspensiontemperature than the other protocols tested. This pro-tocol also gave the best cell viability and attachmentand lowest LDH release, although this was not statis-tically significant.

Storage

Human hepatocytes were stored for up to 3 years in a�140�C freezer with no detrimental effect upon the vi-ability or cell attachment on thawing (Fig. 2). Similarresults were found for LDH release, CYP1A1/2 activ-ity, and albumin synthesis.

Modified Thawing Protocol

Human hepatocytes thawed with the immediate andconstant methods had significantly lower viability (im-mediate: 20% 6 5%, P ¼ 0.01; constant: 45% 6 7%,P ¼ 0.04), attachment efficiency (immediate: 29% 64%, P < 0.001; constant: 42% 6 6%, P ¼ 0.07), andCYP1A1/2 activity (immediate: 3.1 6 1.0 pmol/min-ute/mg of protein, P ¼ 0.005; constant: 5.4 6 1.3pmol/minute/mg of protein, P ¼ 0.02) in comparisonwith the standard method of thawing (viability ¼ 60%6 11%, attachment efficiency ¼ 54% 6 10%,CYP1A1/2 activity ¼ 6.8 pmol/minute/mg of protein;Fig. 3). LDH leakage was significantly higher with theimmediate method of thawing hepatocytes (43% 66%, P ¼ 0.004) in comparison with the standard he-patocytes (25% 6 4%). There was no significant differ-ence between the viability and functional assays ofhepatocytes thawed by the dropwise method and thestandard method.

Figure 2. Effect of the storage time at 2140�C on humanhepatocytes. Human hepatocyte batches were cryopreservedand then thawed at regular times after freezing (1 week, 2weeks, 1 month, 3 months, 1 year, and 3 years). Viabilityand functional measurements were made after thawing.Results are presented as means and standard deviations of5 experiments. Values are compared to those of hepatocytesat 1 week.



Cytoprotection During Thawing

Thawing with fructose did not improve cell viabilitybut had significant effects on the other parametersmeasured. Hepatocytes thawed with 200 or 300 mMfructose had significantly higher attachment efficiency(200 mM: 54% 6 5%, P ¼ 0.03; 300 mM: 59% 6 5%,P ¼ 0.02) and significantly lower LDH leakage (200mM: 19% 6 6%, P ¼ 0.04; 300 mM: 16% 6 5%, P ¼0.03) than standard hepatocytes without fructose inthe medium (attachment efficiency ¼ 43% 6 6%, LDHleakage ¼ 27% 6 7%; Fig. 4). Fructose (200 mM) inthe thawing medium yielded hepatocytes with signifi-cantly higher albumin production (6.3 6 0.5 lg/hour/mg of protein, P ¼ 0.002) than standard hepato-cytes (5.0 6 0.6 lg/hour/mg of protein). Hepatocytesthawed with 5 mM a-lipoic acid in the thawing

medium had significantly higher viability (58% 6 7%,P ¼ 0.01), attachment efficiency (55% 6 5%, P ¼0.03), and albumin production (5.7 6 0.7 lg/hour/mg of protein, P ¼ 0.008) than standard hepatocytes(viability ¼ 47% 6 7%, attachment efficiency ¼ 45% 66%, albumin production ¼ 4.3 6 0.8 lg/hour/mg ofprotein). Human hepatocytes thawed with 300 mM glu-cose in the thawing medium had significantly higher vi-ability (59%6 8%, P ¼ 0.03) than standard hepatocyteswithout glucose in the medium (49%6 6%).

Optimized Cryopreservation and

Thawing Protocol for Clinical Use

Human hepatocytes were cryopreserved with the opti-mized protocol for clinical application. Glucose (300mM) was used as it was previously found to providethe highest viability on thawing. Hepatocytes had sig-nificantly higher viability (52% 6 9%, P ¼ 0.02) andattachment efficiency (48% 6 8%, P ¼ 0.02) than he-patocytes cryopreserved with the standard protocol(viability ¼ 40% 6 7%, attachment efficiency ¼ 39% 66%; Fig. 5). LDH leakage was significantly lower (17%64%, P ¼ 0.03) in comparison with hepatocytes (25% 66%) with the standard protocol.

Figure 4. Effect of fructose in the thawing medium ofcryopreserved human hepatocytes. Human hepatocytes werethawed with WEM containing 0 to 300 mM fructose.Viability and functional measurements were made afterthawing. Results are presented as means and standarddeviations of 3 experiments. Values are compared to thoseof the standard (WEM) hepatocytes without fructose in themedium. Significance: *P < 0.05 and **P < 0.01.Figure 3. Effects of different thawing protocols on human

hepatocytes. Human hepatocytes were cryopreserved andthen thawed with 1 of 4 different thawing methods: (1)immediate (diluted in 25 mL of the thawing medium), (2)constant (5 mL of the thawing medium added every5 minutes for a total of 25 mL), (3) dropwise (25 mL of thethawing medium added in a dropwise manner), and (4)standard (based on Steinberg et al.7). Viability andfunctional measurements were made after thawing. Resultsare presented as means and standard deviations of 3experiments. Values are compared to those of the standardhepatocytes. Significance: *P < 0.05, **P < 0.01, and ***P <0.001.

234 TERRY ET AL.


DISCUSSION

In this study, we have attempted to define the optimumconditions for the cryopreservation of human hepato-cytes. A cell density >3 � 106 cells/mL was found to besuitable for cryopreservation, with cell densities of upto 1 � 107 cells/mL generally yielding good results,whereas low-density freezing (<3 � 106 cells/mL)yielded significantly lower viability and functionalmeasurements on thawing. A low density may producepoor results because the cell pellet to freezing medium/cryoprotectant ratio is decreased (ie, cells are diluted ina higher volume of the freezing medium), and this mayincrease the osmotic stresses to which the cells areexposed. A higher cell density is convenient for thawingthe large numbers of cells needed for hepatocyte trans-plantation, with up to 5 � 108 cells used per infusion inchildren with liver-based metabolic disease.20

In preliminary studies, the choice of the freezingmedium did not appear to be a major determinant ofthe outcome of cryopreservation. A comparison of 10potential freezing media showed that only FCS wassuperior to the standard UW solution. The cryoprotec-tant thus has a greater impact on hepatocyte functionthan the medium used. FCS is unsuitable for the clin-ical cryopreservation of human hepatocytes becauseof the lack of a reproducible source and the risk ofcontamination with animal viruses. Human serumalbumin should be a suitable alternative to FCS asone of the protective effects of FCS could be its highlevels of albumin. The use of UW solution, which wasdesigned to protect cells from the deleterious effects ofhypothermia, as the freezing medium seems far morelogical than the use of culture medium, which wasdesigned for the culture of cells at 37�C. Additionally,from a practical point of view, it is a clinicallyapproved solution so it can be used for the cryopre-servation of human hepatocytes for use in hepatocytetransplantation. An alternative could be CryoStor, a

cryopreservation solution specifically developed forhuman use.21 This new cryopreservation solutionneeds to be fully evaluated with human hepatocytes.

The choice of the cryoprotectant for the cryopreserva-tion of human hepatocytes is critical. DMSO is the cryo-protectant of choice for hepatocytes because it gives op-timum viability and functional measurements incomparison with other cryoprotectants tested in prelim-inary studies. DMSO is able to enter the cells and reduceinjury by moderating the increase in the solute concen-tration during freezing. However, it may be that combi-nations of nonpermeating agents with DMSOmight givebetter results. In a recent study, the addition of the di-saccharide trehalose produced a significant improve-ment in the thawed cell function of human hepato-cytes.22,23 In terms of the concentration of DMSO forhepatocyte cryopreservation, there was little differencebetween the effects of 10%, 12.5%, and 15% DMSO inthe freezing medium on thawed hepatocytes. A concen-tration of 7.5% did not provide protection and thus wasnot sufficient to permeate hepatocytes adequately. Con-versely, 17.5% and 20% DMSO appeared to be toxic tothe hepatocytes and yielded significantly lower attach-ment efficiencies and CYP1A1/2 activities and signifi-cantly higher LDH leakage than the standard of 10%DMSO. Generally, DMSO is used at a concentration of10% for hepatocyte cryopreservation.1,4,7,10

The rate of addition of DMSO to the cell suspensionfor freezing can also affect the cryopreservation of he-patocytes. It is important to freeze the solution assoon as possible and certainly within 10 minutes afterthe addition of the cryoprotectant as DMSO rapidlyenters cells and can have toxic effects. There did notappear to be an advantage with the complicated cryo-protectant addition protocol developed by Uteschet al.19 However, slow addition of DMSO did seem tobe essential. Diener et al.16 also found that the single-step addition of DMSO produced significantly lowerviability than gradual addition.

Figure 5. Effect of the optimized cryopreservation and thawing protocol on the viability and function of human hepatocytes.Human hepatocytes were cryopreserved with the optimized cryopreservation protocol or the standard protocol. Viability andfunctional measurements were made after thawing. Results are presented as means and standard deviations of 3 experiments.Values are compared to those of standard hepatocytes without glucose in the freezing and thawing media. Significance: *P <0.05.



The thawing protocol for cryopreserved hepatocytesis the final stage in the protocol. However, if thawing issuboptimal, then all the care taken with the precedingstages to preserve hepatocyte function will be lost.Rapid thawing at 37�C was carried out in the presentstudy. Generally, it is accepted that slow dilution ofthe cryoprotectant in the thawed cell suspension isoptimal for hepatocytes. This allows the cryoprotectantto move out of the hepatocytes slowly and thus pre-vents osmotic shock. Although there was no statisti-cally significant difference between the dropwisemethod and the standard method of thawed hepato-cyte dilution, the standard method gave optimalresults with respect to viability and function. Thethawing medium used was also investigated in prelimi-nary experiments with different culture media and UWsolution, but this did not have any significant effect onthawed hepatocyte viability and function. This may bebecause the osmolalities of the media tested are simi-lar and so do not affect the dilution process. Glucoseand fructose both produced significant improvementsin function when they were added to the thawing me-dium. The effect of glucose and fructose during thaw-ing could be due to the hyperosmotic concentration ofthe solutes. The antioxidant a-lipoic acid also had posi-tive effects on the viability and function of hepatocyteswhen it was included in the thawing medium. Glucose,readily available as 50% dextrose for intravenous use,was most convenient for clinical application and wasthus chosen for use in the optimized protocol.

In the present study, pre-incubation was not usedin the clinical protocol as it might prove difficult toperform with a large number of cells in a good manu-facturing practice environment and also adds to theoverall time of liver processing. Glucose was added toboth the freezing and thawing media, and this wasshown to provide a significant benefit both duringthawing in the first part of the study and overall inthe clinical protocol. Glucose (300 mM) in the freezingand thawing media may have boosted adenosine tri-phosphate levels in hepatocytes before freezing withbeneficial effects on hepatocyte mitochondrial func-tion.24,25 Clearly, there is potential for further studieson the addition of cytoprotectants during cryopreser-vation with suitable agents or combinations of agents.

In conclusion, an improved cryopreservation proto-col has been developed for human hepatocytes and isnow in use for cells being cryopreserved and stored inour cell bank for clinical transplantation.

ACKNOWLEDGMENTSThe authors thank the liver transplant surgeons andcoordinators at King’s College Hospital for their helpin obtaining the donor liver tissue used.

REFERENCES

1. Rijintes PJM, Moshage HJ, Van Gemert PJL, De Waal R,Yap SH. Cryopreservation of adult human hepatocytes:the influence of deep freezing storage on the viability,

cell seeding, survival, fine structures and albumin syn-thesis in primary cultures. J Hepatol 1986;3:7-18.

2. Loretz LJ, Li AL, Flye MW, Wilson AGE. Optimization ofcryopreservation procedures for rat and human hepato-cytes. Xenobiotica 1989;19:489-498.

3. Dou M, De Sousa G, Lacarelle B, Placidi M, De La PortePL, Domingo M, et al. Thawed human hepatocytes in pri-mary culture. Cryobiology 1992;29:454-469.

4. Diener B, Traiser M, Arand M, Leissner J, Wirsch U,Hohenfellner R, et al. Xenobiotic metabolising enzymeactivities in isolated and cryopreserved human liverparenchymal cells. Toxicol In Vitro 1994;8:1161-1166.

5. Adams RM, Wang M, Crane AM, Brown B, DarlingtonGJ, Ledley FD. Effective cryopreservation and long-termstorage of primary human hepatocytes with recovery ofviability, differentiation, and replicative potential. CellTransplant 1995;4:570-586.

6. Li AP, Lu C, Brent JA, Pham C, Fackett A, Ruegg CE,Silber PM. Cryopreserved human hepatocytes: character-ization of drug-metabolising enzyme activities and appli-cations in higher throughput screening assays forhepatotoxicity, metabolic stability and drug-drug interac-tion potential. Chem Biol Interact 1999;121:17-35.

7. Steinberg P, Fischer T, Kiulies S, Biefang K, Platt KL,Oesch F, et al. Drug metabolizing capacity of cryopre-served human, rat, and mouse liver parenchymal cells insuspension. Drug Metab Dispos 1999;27:1415-1422.

8. Hengstler JG, Utesch D, Steinberg P, Platt KL, Diener B,Ringel M, et al. Cryopreserved primary hepatocytes as aconstantly available in vitro model for the evaluation ofhuman and animal drug metabolism and enzyme induc-tion. Drug Metab Rev 2000;32:81-118.

9. Ostrowska A, Bode DC, Pruss J, Bilir B, Smith GD, Zei-sloft S. Investigation of functional and morphological in-tegrity of freshly isolated and cryopreserved humanhepatocytes. Cell Tissue Bank 2000;1:55-68.

10. Alexandre E, Viollon-Abadie C, David P, Gandillet A,Coassolo P, Heyd B, et al. Cryopreservation of adulthuman hepatocytes obtained from resected liver biop-sies. Cryobiology 2002;44:103-113.

11. Terry C, Mitry RR, Lehec SC, Muiesan P, Rela M, HeatonND, et al. The effects of cryopreservation on human he-patocytes obtained from different sources of liver tissue.Cell Transplant 2005;14:585-594.

12. Terry C, Dhawan A, Mitry RR, Hughes RD. Cryopreserva-tion of isolated human hepatocytes for transplantation:state of the art. Cryobiology 2006;53:149-159.

13. Terry C, Dhawan A, Mitry RR, Lehec SC, Hughes RD. Pre-incubation of rat and human hepatocytes with cytoprotec-tants prior to cryopreservation can improve viability andfunction on thawing. Liver Transpl 2006;12:165-177.

14. Strom SC, Dorko K, Thompson MT, Pisarov LA, NusslerAK. Large scale isolation and culture of human hepato-cytes. In: Franco D, Boudjema K, Varet B, eds. Ilots deLangerhans et Hepatocytes. Paris, France: Les EditionsINSERM; 1982:195-205.

15. Mitry RR, Hughes RD, Aw MM, Terry C, Mieli-Vergani,G,Girlanda R, et al. Human hepatocyte isolation and rela-tionship of cell viability to early graft function. CellTransplant 2003;12:69-74.

16. Diener B, Utesch D, Beer N, Durk H, Oesch,F. A methodfor the cryopreservation of liver parenchymal cells forstudies of xenobiotics. Cryobiology 1993;30:116-127.

17. Lowry OH, Roseburgh NJ, Farr AL, Randall RJ. Proteinmeasurement with the Folin phenol reagent. J BiolChem 1951;193:265-275.

18. Donato MT, Gomez-Lechon MJ, Castell JV. A microassayfor measuring cytochrome P450IA1 and P450IIB1 activ-ities in intact human and rat hepatocytes cultured on96-well plates. Anal Biochem 1993;213:29-33.

236 TERRY ET AL.


19. Utesch D, Diener B, Molitor E, Oesch F, Platt KL. Char-acterization of cryopreserved rat liver parenchymal cellsby metabolism of diagnostic substrates and activities ofrelated enzymes. Biochem Pharmacol 1992;44:309-315.

20. Dhawan A, Mitry RR, Hughes RD. Hepatocyte transplan-tation for liver-based metabolic disorders. J InheritMetab Dis 2006;29:431-435.

21. Sosef MN, Baust JM, Sugimachi K, Fowler A, TompkinsRG, Toner M. Cryopreservation of isolated primary rathepatocytes: enhanced survival and long-term hepato-specific function. Ann Surg 2005;241:125-133.

22. Miyamoto Y, Suzuki S, Nomura K, Enosawa S. Improve-ment of hepatocyte viability after cryopreservation bysupplementation of long-chain oligosaccharide in the

freezing medium in rats and humans. Cell Transplant2006;15:911-909.

23. Katenz E, Vondran FW, Schwartlander R, Pless G, GongX, Cheng X, et al. Cryopreservation of primary humanhepatocytes: the benefit of trehalose as an additionalcryoprotective agent. Liver Transpl 2007;13:38-45.

24. Zaleski J, Richburg J, Kauffman FC. Preservation of therate and profile of xenobiotic metabolism in rat hepato-cytes stored in liquid nitrogen. Biochem Pharmacol1993;46:111-116.

25. Stephenne X, Najimi M, Ngoc DK, Smets F, Hue L, Gui-gas B, Sokal EM. Cryopreservation of human hepato-cytes alters the mitochondrial respiratory chain complex.Cell Transplant 2007;16:409-419.



optimization of the cryopreservation and thawing protocol for human hepatocytes for use in cell...

Documents