at SciVerse ScienceDirect

Biomaterials 33 (2012) 829e836

Contents lists available

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

Influence of cell culture configuration on the post-cryopreservation viabilityof primary rat hepatocytes

Raquel Magalhães a,b, Bramasta Nugraha c,d,e, Shazib Pervaiz b,d,f,g, Hanry Yu b,c,d,e,g,h,**,Lilia L. Kuleshova a,i,*

a Low Temperature Preservation Unit, National University Medical Institutes Yong Loo Lin School of Medicine, National University of Singapore, Block MD11 #03-01C,10 Medical Drive, Singapore 117597, SingaporebDepartment of Physiology, Yong Loo Lin School of Medicine, National University of Singapore Health System, MD9-03-03, 2 Medical Drive, Singapore 117597, SingaporecMechanobiology Institute, T-Labs, #05-01, 5A Engineering Drive 1, Singapore 117411, SingaporedNUS Graduate School for Integrative Sciences and Engineering, Center for Life Sciences, #05-01, 28 Medical Drive, Singapore 117576, Singaporee Institute of Bioengineering and Nanotechnology, A*STAR, The Nanos, #04-01, 31 Biopolis Way, Singapore 138669, SingaporefDuke-NUS Graduate Medical School, 8 College Road, Singapore 169857, Singaporeg Singapore-MIT Alliance for Research and Technology, 3 Science Drive 2, S16-05-08, Singapore 117543, Singaporeh Singapore-MIT Alliance, Computational and System Biology Program, E4-04-10, 4 Engineering Drive 3, Singapore 117576, Singaporei Institute for Multiphase Processes, Leibniz Universität Hannover, Callinstraße 36, 30167 Hannover, Germany

a r t i c l e i n f o

Article history:Received 16 September 2011Accepted 10 October 2011Available online 9 November 2011

Keywords:Primary hepatocytesCryopreservationTissue engineered constructsFreezingVitrification

* Corresponding author. Low Temperature PreservaMedical Institutes Yong Loo Lin School of MedicSingapore, Block MD11 #03-01C, 10 Medical Drive, Si** Corresponding author. Department of PhysioloMedicine, National University of Singapore Health SyDrive, Singapore 117597, Singapore.

E-mail addresses: medyuh@nus.edu.sg (H. Yu), kul(L.L. Kuleshova).

0142-9612/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.biomaterials.2011.10.015

a b s t r a c t

Cryopreservation has been identified as a necessary barrier to overcome in the production of tissueengineered products for clinical application. Liver engineering and bioartificial liver assisting devices areon the forefront of tissue engineering research due to its high demand and clinical potential. In this studywe propose that the cryopreservation of primary mammalian hepatocytes yields better results whenthese cells are in a tissue-like culture configuration since cell attachment is essential for cell survival inthis cell type. We used two different tissue-engineered culture configurations: monolayers and spheroidculture; and two different concepts of cryopreservation, namely vitrification and freezing. Cell suspen-sions were also cryopreserved using both approaches and results were compared to the engineeredcultures. Both engineered configurations and suspension were cryopreserved using both conventionalfreezing (cooling at 1 �C/minute using 10% DMSO in foetal calf serum) and vitrification (using 40%ethylene glycol 0.6 M sucrose supplemented with 9% Ficoll). These two approaches differ on the degree ofmechanical stress they inflict on the material to be cryopreserved. The maintenance of cell-to-cell andthe integrity of the actin cytoskeleton were assessed using scanning electron microscopy andimmunohistochemistry respectively. Results showed that while there was no significant differencebetween the degree of integrity shown between vitrified and control engineered cultures, the same didnot happen to the frozen engineered constructs. The disruption of the cytoskeletal structure correlatedwith increased levels of apoptotic markers. With cryopreserved suspensions there was evidence ofdisruption of the cytoskeletal structure. This study concluded that cell-to-cell contact is beneficial in themaintenance of viability post-cryopreservation and that the vitrification approach was far superior tothose of conventional freezing when applied to 2D and 3D hepatocyte based engineered cultures.

� 2011 Elsevier Ltd. All rights reserved.

tion Unit, National Universityine, National University ofngapore 117597, Singapore.gy, Yong Loo Lin School ofstem, MD9-03-03, 2 Medical

eshova@imp.uni-hannover.de

All rights reserved.

1. Introduction

Bioengineered liver support requires consistent and reliablepool of functional hepatocytes tomake it into a clinically significantalternative to liver transplant. As these cells are prone to losingtheir functional ability in vitro [1e6] the development of aneffectivelow temperature storage has immense potential.

This study aims to illustrate the options available to the cryo-preservation hepatocytes. Two approaches to cryopreservationwere used, conventional freezing and vitrification. The freezing

R. Magalhães et al. / Biomaterials 33 (2012) 829e836830

approach has been shown to cause disruption in the integrity of cellaggregates, contrary to vitrificationwhere no visible disruptionwasdetected [7]. This will be the base of our hypothesis that thecryopreservation of tissue engineered constructs is only possible bymeans of vitrification since the maintenance of cell-to-cell contact(in a tissue-like structure) is essential for the maintenance of cellviability. Hepatocytes are notably sensitive to loss of viability aftertissue isolation. In fact, several studies have reported that the maincause for the decrease in cell viability after isolation is due to anapoptotic pathway triggered by cell detachment known as ‘anoikis’,a key contributor to the rapid decrease in viability of primarymammalian hepatocytes after isolation [8e12]. In the context ofmammalian hepatocyte suspension, the great majority of thenumerous studies on cryopreservation uses the freezing approachbut with limited results [13e17]. Most common hindrance, apartfrom the low post-thaw viabilities which have been reported as lowas 10% [18] is due to DNA damage. In the light of the inability tocryopreserve primary hepatocyte suspensions several tissueculture configurations have been explored.

This study explored the influence of cell attachment and cyto-skeletal disruption on the efficiency of cryopreservation of primaryrat hepatocytes [19e21]. The maintenance of the tissue constructintegrity and the profile of actin distribution post-cryopreservationwere assessed. It has been shown that actin filaments play a pivotalrole in regulating the extent to which mammalian cells cope withchanges in cell volume induced by extracellular stress [21e27].Typically, in resting hepatocytes, F-actin has been shown toconcentrate on the peripheral region of the cell. By measuring theintensity F-actin of a cell through its equatorial plane using a fluo-rescent stain, phalloidin, an actin ring was shown as bright andthick underlying the cell membrane [27]. In tissue, especially wherecell junctions are present, a similar phenomenon occurs [23,28,29].Although there is the involvement of numerous factors and path-ways, downstream from cell detachment there is the activation ofthe caspase and apoptosis signalling. For this, both caspase levelsand cell trans-membrane potential transition (DJ) (for earlyapoptotic event) were quantitated in this study.

To further demonstrate the importance of cell attachment in thecryopreservation of hepatocytes, cryopreservation was also per-formed on primary hepatocyte cell suspension.

2. Materials and methods

2.1. Materials

Chemicals and reagents used in all experiments were purchased from Sigma, St.Louis, MA, USA, unless stated otherwise stated.

2.2. Rat hepatocyte isolation and culture

Hepatocytes were harvested from the livers of male Wistar rats, weighing250e300 g, by a modified two-step in situ collagenase perfusion technique [30].Viability of hepatocyte suspension was determined by Trypan blue exclusion test;survival rate of 90% was taken as the minimal level of acceptance for experimen-tation. Yields were approximately 108 cells per rat. All animal experiments wereapproved by the Institutional Animal Care and Use Committee of the NationalUniversity of Singapore and were conducted in accordance with the Guide for theCare and Use of Laboratory Animals, National Institutes of Health, USA.Cells weremaintained with Williams’ E medium supplemented with 10 mM NaHCO3, 1 mg/mLBSA, 10 ng/mL of EGF, 0.5 mg/mL of insulin, 5 nM dexamethasone, 50 ng/mL linoleicacid, 100 units/mL penicillin, and 100 mg/mL streptomycin and were incubated with5% CO2 at 37 �C and 95% humidity.

2.3. Formation, culture and cryopreservation of tissue engineered constructs

Hepatocytes were allowed to aggregate into spheroids in spheroid-formingmedia as previously described elsewhere [31]. Hepatocyte monolayers cultured oncollagen-coated polyethylene terephthalate membrane were assembled andcultured as described previously in [32]. These were henceforth denoted as HS andPETh, respectively. Both HSs and PETh were cryopreserved within 24 h of assembly.

2.3.1. Cryopreservation of tissue engineered constructsVitrification of HS and of PETh were performed as described in [31] and [32],

respectively.For freezing, the PETh was cut into strips to (0.7 � 1.0 cm) and HS were re-

suspended in freezing media and frozen in 2.0 ml cryovial. Freezing media con-sisted of 10% (v/v) DMSO in foetal calf serum. Samples were removed from cultureand exposed to cryoprotectants for 3 min at room temperature before transferto�80 �C freezer on a ‘Frosty boy’ (Nalgene, USA) apparatus, at cooling 1 �C/ minute.For thawing, cryovials were immersed in a 37 �C water-bath until no visible ice wasdetected. Samples were then removed from the DMSO solution and washed in freshHepatozyme Serum Free Medium (Gibco, Invitrogen, USA) equilibrated to 37 �C, andkept in culture in the growth chamber at 37 �C, 5% CO2.

Beforemetabolic assessment, each PETh in culturewas cut into smaller pieces sothat it would fit in 96-well NUNC plates. This was done in order to reduce samplingvariability, each frozen or vitrified PETh was followed through both caspase andtrans-membrane potential assessment.

2.3.2. Vitrification of primary rat hepatocyte suspensionPrimary rat hepatocyte suspension was vitrified within 5 h of isolation. Vitrifi-

cation procedure followed the vitrification of HSs [31] where straw-in-straw wasused for cooling. After warming, hepatocytes were seeded onto collagen-coated PETfilms and allowed to attach under the same conditions of control hepatocytes.

2.3.3. Freezing of primary rat hepatocyte suspensionFreezing was performed similarly as it has been described previously for

primary rat hepatocytes seeded onto PET films [32]After thawing, hepatocytes wereseeded onto collagen-coated PET films and allowed to attach under the sameconditions of control hepatocytes.

2.4. Morphology assays

2.4.1. Morphological study by scanning electron microscopy (SEM)SEM was used to assess ultra-structural effects of cryopreservation on

HSs.Briefly, the samples were fixed with 2.5% (w/v) glutaldehyde in PBS overnight at4 �C, incubated in 1% (w/v) OsO4 in de-ionized water for 1 h at room temperature,and sequentially dehydrated with ethanol series (25, 50, 75, 95 and 100% (v/v)) for10 min each. All processes were done with gentle pipetting to prevent damage toHSs. The HSs were then pasted on glass cover slide with attached carbon tape, driedat 50 �C to remove moisture, platinum-sputtered (JFC-1600, JEOL, Tokyo, Japan) andexamined by Field Emission Scanning Electron Microscope (FESEM, JSM-7400F,JEOL, Tokyo, Japan). The FESEM viewing was done using 10 kV.

2.4.2. Cytoskeleton integrityCytoskeleton integrity of hepatocytes was assessed in the in PETh configuration,

i.e. monolayers. All samples were fixed and kept in 4% formaldehyde before staining.To visualise the actin cytoskeleton, cells were permeabilised with 0.1% Triton X-

100 inPBS for 10min, andstained for1h at roomtemperaturewith a 15mg/ml solutionof Alexa Fluor 488-phalloidin (Molecular Probes Europe, BV). Samples were thenwashed three timesby immersion in 1x PBS. Samples were fixed with VectashieldMountingMediacontaining1.5mg/mlDAPI (Vector Laboratories,Burlingame,CA,USA).

Images of stained samples were captured with a confocal laser microscope(Olympus FV1000, Tokyo, Japan). Excitation wavelengths of 488 nm and 780 nmwere used for the phalloidin-labelled actin and the DAPI-stained nuclear material,respectively. The images produced by sequential scans via the different colourchannels were then merged and recorded in digital format. Subsequently, theimages were displayed and analysed using ImageJ 1.42q software (National insti-tutes of Health, USA).

The distribution of the phalloidin-labelled stain on the hepatocytes within themonolayer was established upon measuring the gray intensity values after acrossthe cell diameter. The thickness of the actin band underlying the cellular membranewas measured as defined in Fig. 1.

The ratio (Iperipheral � Icentral)/(Iperipheral þ Icentral) was calculated as previouslydescribed [22,33]. By dividing the difference between the peripheral and centralgray intensity, by the total intensity, the ratio value measured for each singlehepatocyte is normalized and therefore comparable between different samples.Gray intensity values above 1500 were generally considered as brightest and cor-responding to a higher stain concentration.

The thickness of the actin band underlying the membrane was determined byaveraging the length (in mm) of the peaks corresponding to the higher gray intensity,as indicated in Fig. 1.

2.5. Apoptotic profile

2.5.1. Caspase levelsLive PETh cultures were assessed for caspase expression levels using the APO

LOGIX SulforhodamineCaspase Detection Kit � (Cell Technology, Cat# SR100-1, CA,USA). The kit contains a generic caspase inhibitor labelled with sulforhodamine-peptide-fluoromethyl ketone) to which the cell membrane is permeable. BrieflyeachPEThfilmwascut into0.5cmdiameterdiscs tofit into thebottomof96-wellblack

Fig. 1. Schematic representation of the method of establishing actin distribution across the cells within the monolayer, and also of the establishment of the thickness of the actinband underlying the cellular membrane. The graph in the scheme represents the distribution of gray intensity across the length of measurement.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836 831

NUNC plates. For each cryopreservation, and control samples, caspase levels werequantitated immediately after warming (day 0); samples were kept in continuousculture at 37 �C, 5%CO2 for2days and4daysafter cryopreservation. Theprocedurewasperformed according to the detailedprotocols attached to the kit and thefluorescenceintensitywas readwithaTECANfluorescenceplate reader (TecanTechnologies, Japan)with excitation source at 550 nm and measuring emission at 600 nm.

2.5.2. Detection of the mitochondrial trans-membrane potentialLive PETh cultures were assessed for trans-membrane potential transition using

the APO LOGIX JC-1 Assay Kit� (Cell Technology, Cat# JC-100, CA, USA). The kitmeasures the change in the mitochondrial membrane potential (DJ). Loss ofmitochondrial DJ is indicative of apoptosis and is detected by a unique fluorescentcationic dye, 5,50 ,6,60-tetrachloro-1,10,3,30-tetraethyl-benzamidazolocarbocyaniniodide, commonly known as JC-1. Briefly each PETh film was cut into 0.5 cmdiameter discs to fit into the bottom of 96-well black NUNC plates. For each cryo-preservation and control samples, mitochondrial trans-membrane potential wasquantitated immediately after warming (day 0); samples were kept in continuousculture at 37 �C, 5% CO2 for 2 and 4 days after cryopreservation. The procedure wasperformed according to the detailed protocols attached to the kit and the fluores-cence intensity was read with a TECAN fluorescence plate reader (Tecan technolo-gies, Japan) with excitation source at 550 nm and measuring emission at 600 nm(non-apoptotic cells e stained red) and excitation 485 nm and measuring emissionat 535 nm (apoptotice stained green). The result on the number of cells undergoingan early apoptotic event is given by ratio of red/green fluorescence.

2.6. Statistical analysis

Analysis was performed with values obtained for vitrified, control and frozensamples and compared within treatments and within different days of culture with3-way ANOVA (SPSS version 17, SPSS Inc., Chicago, IL, USA). p < 0.05 was consideredto be statistically significant. All data are reported as means � standard error of themean (s.e.m).

3. Results

3.1. Effect of cryopreservation in maintaining the morphology ofHSs

The morphology of HSs examined for control, vitrification andfreezing using scanning electron microscopy (SEM) are shown in

Fig. 2. The maintenance of structural integrity of the bioengineeredconstruct after cryopreservation is essential for the maintenance ofpost-thaw function. The use of HS configuration for this morpho-logical study illustrates well howmechanically more damaging theprocess of freezing is relative to vitrification.

Fig. 2 also shows SEM images of HSs which were maintained instatic culture in Primaria� plates for 5 days after cryopreservationtogether with an HS size distribution chart. SEM images showregular surface with well defined cell morphology for both controland vitrified HS, throughout the 5 days culture. The frozen HSs, inthe other hand have show a consistently roughed surfacethroughout the 5 days static culture, which may indicate that mostcells on the surface of the HSs were either dead or senescentimmediately after cryopreservation throughout the continuousculture. An average surface distribution chart was calculated foreach of the experimental groups. The surface areas in mm2 wereestimated from the SEM images using imaging software bymeasuring the longest length andwidth (n¼ 16e20). No significantdifference on the average surface area was observed between thecontrol and vitrified HSs throughout the 5 day culture, which is anindication that the process of vitrification maintained the integrityof the HS. On the other hand, in the freezing treatment group, theprocess of cryopreservation seems to have led to the disintegrationof the overall size of the structures to sizes that were up to 85%smaller than the control (on day 3) and 86% smaller than thevitrified (day 3 for vitrification).

3.2. Effect of cryopreservation on cellular actin distribution

Actin stain distribution on vitrified and frozen PETh (upperpanels), vitrified and frozen hepatocytes suspension and controlPETh(lower panels) are shown on Fig. 3. For the time point 48 hafter restoring samples to 37 �C, no qualitative difference can beobserved between vitrification and control PEThs. On the other

Fig. 2. SEM images of control, vitrified and frozen HSs on day 1, 3 and 5. The plotted graph indicates the average surface area of the HS estimated from SEM images. Results arepresented as means � s.e.m. (x indicates significant difference with the control and * indicates significant difference with the vitrified). Scale bars denote 100 mm.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836832

hand, in the frozen PETh, a ‘diffuse’ pattern of the actin distributioncan be observed, implying that the accumulation of the stain hadmoved from the peripheral to the central plane of the cells.

For the hepatocytes suspension cryopreserved either byfreezing or vitrification, there was no apparent qualitative differ-ence between the concentrations of fibres on the peripheral or on

the central area of the cells. Furthermore, the tissue like structurewas also not apparent suggesting that the ability of the hepato-cytes to aggregate in a tissue like structure was low. The resultsshown in the confocal images were quantified by the methoddescribed by Erickson and collaborators and it is shown on thegraph of Fig. 4 [33].

Fig. 3. F-actin confocal images of vitrified PETh, control, frozen PETh, vitrified suspension, and frozen suspension at 48 h post-seeding.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836 833

Peripheral and central brightness of gray images from confocalacquired files were quantified and the ratio of (Iperipheral e Icentral)/(Iperipheral þ Icentral) was calculated. The lower the ratio the morepolarized the cell would be, i.e. the higher peripheral concentrationof actin would be found in the peripheral area. Results show thatthere were no significant difference between the patter of actindistribution in control and vitrified PETh, except on the last timepoint where vitrified PETh appeared to be slightly more polarizedthan control. On the frozen PETh, a dramatic decrease in the ratiowas observed, which suggests that the cellular actin cytosketetonhad changed pattern to a diffuse distribution of the fibres, leadingto cell disaggregation [20,27,34e36]. The importance of cellattachment prior to cryopreservation in maintaining high viability

Fig. 4. Cellular actin distribution described by the ratio (I peripheral - I central)/(Iperipheral þ I central). Statistical differences are marked with an asterisk (*, p < 0.05,control) and a section sign (x, p < 0.05, vitrification). n ¼ 14e60. Error bars represents.e.m.

was also confirmed by comparing actin distribution of cellsuspension after cryopreservationwith cellular actin distribution ofcryopreserved PETh.

The thickness of the cellular membrane was also analysed andis shown in Fig. 5. As it is expected, there was a thinning of theactin band underlying the cellular membrane concomitantly withthe lowering of the peripheral phalloidin stain intensity. This wasclearly observed for the hepatocytes on the frozen PETh.However, this was less steep than the decrease in its intensity. Infact, immediately after cryopreservation, a significantly lower cellmembrane thickness was observed in hepatocytes in the frozenPETh, which was 34 and 40% lower than its vitrified and controlcounterparts, respectively. This suggests that a thinning of the

Fig. 5. Thickness of the actin band underlying the cellular membrane. Statisticaldifferences are marked with an asterisk (*, p < 0.05, control) and a section sign (x,p < 0.05, vitrification). n ¼ 14e60. Error bars represent s.e.m.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836834

membrane occurred prior to the actin migration to the core ofthe cell.

A relative thinning of the cell membrane was also observed inhepatocytes on the vitrified and control PETh. Although veryprominent, the relative thinning was the highest immediately aftercryopreservation (27%), but recovered to an average of 10%, duringthe culture period 24e96 h. A similar pattern was found whencomparing the membrane thickness of hepatocytes on frozen andcontrol PETh where the highest decrease in membrane thicknesswas found at the time point 0 h (40%).

The results presented hitherto suggest that the cryopreservationof primary hepatocyte suspension is less efficacious than cryo-preservation in tissue-like configuration. On the other hand, vitri-fication shows a considerable advantage in the cryopreservation ofPETh compared to freezing.

3.3. Effect of cryopreservation on cellular apoptosis

The induction of early apoptotic events was investigated infrozen and vitrified PETh for an early period upon restoration to37 �C (day 0) up to 4 days after restoration to physiologicaltemperature. A decline in mitochondrial membrane potential (DJ)is a hallmark indication of apoptosis, leading to the collapse of theorganelle and release of cytocrome c to the cytoplasm and theactivation of the apoptotic cascade [37e40].

The ratio of non-apoptotic to apoptotic and necrotic cells ispresented in Fig. 6. Except for the ratio comparison between controland vitrification on day 0, all other experimental groups exhibitedsignificant statistical difference when compared within the sameday. However, while significant statistical difference betweencontrol and vitrified was not too accentuated, the value of frozenPETh lay well below control levels for all days of observation. Thisabyssal significant difference was also observed between vitrifiedand frozen PETh.

Both control and vitrified PETh sustained the low levels ofearly apoptosisfor the first two days of culture showing nosignificant statistical difference between the ratio non-apoptotic/apoptotic cells for day 0 and day 2. Similarly, for both control andvitrified PETh, a level of significant difference was observed forthe ratio expressed for either group between days 2 and 4, witha marginally more significant decay for the vitrified sample.Frozen PETh showed a steady decrease of the ratio of non-apoptotic to apoptotic and necrotic cells throughout the cultureperiod. The results suggest that the approach to cryopreservation

Fig. 6. Quantitation of DJ and calculation ratio of non-apoptotic to apoptotic andnecrotic cells on day 0, 2 and 4. No statistical difference was observed between controland vitrified PETh on day 0 observation (p > 0.05). In all other experimental groupsmarked with an asterisk (*, p < 0.05, control) and those marked with a section sign (x,p < 0.05, vitrification), a significant statistical difference was observed. Error barsrepresent s.e.m.

have an effect on the occurrence of early apoptotic events onprimary rat hepatocytes. Furthermore, they show that in hepa-tocytes vitrified as PETh the occurrence of early apoptosis doesnot increase significantly within 48 h of culture (equally observedin control PETh).The increase in early apoptotic event is signifi-cant for hepatocytes frozen as PETh within the first 48 h ofculture. The similarities between the behaviour of control andvitrified PETh throughout the 4 days of culture lead us to assumethat apoptosis in cells that are vitrified in this configuration occurmostly immediately after restoration to physiological tempera-ture, and that the surviving cells are able to sustain in culture,which is similar as native untreated cells.

The induction of early apoptotic events was also studied infrozen and vitrified hepatocyte suspensions. Analysis of fluores-cence emission which provides a measure of the number of cellsassayed lead us to assume that some cells did not attach to thecollagen-coated PET film. The low values for 535 nm and 600 nmemissions for the primary rat hepatocytes cryopreserved assuspensionwere consistently observed in all 18 samples for each ofthe cryopreserved groups throughout the 4 days of continuousculture. Therefore, this indicates that they are not artefacts of nonattached cells due to insufficient culture time, and thus strength-enour hypothesis that attachment of primary rat hepatocytesbefore cryopreservation is an essential feature for cell survival post-cryopreservation.

Apoptotic events on cryopreserved PETh were also quantifiedusing a commercial general caspase (caspase 1e9) detection kit andthe results are shown in Fig. 7. Caspases are proteolytic enzymeswhich participate in a cascade of protein cleavage reaction inresponse to a cell death signal. The response, pro-apoptotic signals,will ultimately result in the disassembly of the cell structure [41].

Caspase expression levels on primary rat hepatocytes seeded oncollagen-coated PETh, expressed as RFU (relative fluorescenceunits) which was pre-normalized to cell numbers, did not show anysignificant difference between control and vitrification groups,except on day 4 (p ¼ 0.02).

Significant statistical difference was prominently observedbetween both control and vitrified PETh and frozen PETh, wherethe number of cells expressing a wide array of caspases 1 to 9increased by 2.5 folds from day 0 to day 4. On day 4,caspaseexpression was over 100% higher in the frozen PETh than on eithercontrol or vitrified PETh. Caspase expressions level increase forvitrified PETh and control are 1.7 and 1.4 fold, respectively.

Fig. 7. Quantitation of caspase expression level in PETh cultures normalized with cellnumbers on day 0, 2 and 4. No statistical difference was observed between control andvitrified PETh on day 0 and 2 observation (p > 0.05). In all other experimental groupsmarked with an asterisk (*, p < 0.05, control) and those marked with a section sign (x,p < 0.05, vitrification), a significant statistical difference was observed. Error barsrepresent s.e.m.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836 835

These results come in the same line as those shown by theoccurrence of a raise in the mitochondrial membrane transitionpotential in the frozen batch of PETh, which confirms that occur-rence of apoptosis in primary rat hepatocytes is directly linked tothe approach used for cryopreservation.

4. Discussion

Two different culture configurations were used in this study inorder to provide easier access to the interpretation of results. Theuse of hepatocytes self aggregation into spheroids allowed the useof SEM to survey the degree of integrity maintenance aftera complete cryopreservation study; use of hepatocyte monolayerseeded on PET films provided the ideal base for histochemicalstudies where the polarity of the cell and even the cytoskeletalintegrity could be easily visualised with the aid of microscopictechniques. To test whether cell attachment has an effect thatoverrides the cryopreservation approach followed, in terms of cellviability, for cryopreservation, the same set of experiments wereperformed, in turn using cell suspensions that were cryopreservedby either vitrification or freezing and then seeded on collagen-coated PET.

Vitrification employing the solution consisting of 40% ethyleneglycol (v/v) 0.6M sucrose was validated for the cryopreservation ofprimary rat hepatocytes supported by high cell viability andmaintenance of structural polarization and cell metabolic function[31,32,42,43]. Here, the hypothesis that vitrification is superior inthe ability to cryopreserve 3D tissue-like cultures (as demonstratedby SEM images) when compared to freezing has successfully sup-ported for rat hepatocyte-based constructs as it had been previ-ously suggested for neuronal progenitor cells aggregates [7]. Thequestion remained on whether the success of vitrification laid onthe fact that it is less mechanically stressful on the biologicalmaterial or the 3D configuration itself and the maintenance of cell-to-cell junction and tissue like configuration was effectively pro-tecting the cells against damage. The answer could lay in the factthat while there is a disruption of the tissue-like structure duringfreezing the same does not occur in vitrification to the same extent.To answer the question on whether the type of cryopreservationapproach is an essential factor in the maintenance of cell viabilityand metabolic function after cryopreservation cellular actindistribution and actin ring thinning post-cryopreservation (for bothvitrification and freezing) and compared the results to control ina 2D culture configuration.

Vitrification has an advantage over freezing which is theabsence of ice formation, thus less mechanical stress to the bio-logical material. Viewing the cell membrane and the actin cyto-skeleton as two of the sub-cellular organelles that would be mostgreatly affected bymechanical damage, the effect that both freezingand vitrification have on the cell actin distribution and membranethinning were observed. The results presented here have clearlyshown that the detrimental effects of cryopreservation by freezingare much more prominent than those incurred by the hepatocytespreserved using the vitrification approach, which is in line with thehypothesis we had proposed in introduction.

Additionally we have observed whether cell attachment prior tocryopreservation had an effect of the effectiveness of the processthat would override the use of either approach. The observedresults regarding the damage incurred by the hepatocytes cry-opreserved in suspension using either approach leads us to believethat the cryopreservation of primary cell hepatocytes followingisolation might yield lower efficacy. These results had been ex-pected as the cryopreservation of rat hepatocytes by freezing hasyielded less desirable results in other application studies of cryo-preservation of hepatocytes aggregates in 2D and 3D cultures. All

these studies on vitrification have been very successful in terms ofviability and cell metabolic function [31,32,42,43]. We have alsocomprehensively shown that the success of these studies had beenconnected to the maintenance of their cell aggregate configurationand thereby of the cellecell interaction and that anoikis is a keyplayer on the induction of cell death on mammalian hepatocytes.Finally the results on the disruption of cytoskeleton have beenpositively correlated to the induction of cell death by apoptosis byinvestigating the general caspase [1e9] expression and also theshift in mitochondrial membrane potential.

5. Conclusion

The understanding of not only the cause but also the mecha-nisms promoting cell linked to cryopreservation would pave theway to the development of new cell culture strategies which wouldimprove the quality of cell based products. Comprehensive studieson a wider array of cell-biomaterial constructs and on otherprimary cells would support accuracy of our novel findings to addto a our contribution on the study of the vitrification concept usingprimary hepatocyte based self-assembled cell aggregates and cellmatrix system.

Acknowledgement

This research was supported by a grant from the BiomedicalResearch Council (BMRC) of Singapore (Grant No.: 04/1/21/19/317)to Lilia Kuleshova; and supported in part by funding from theInstitute of Bioengineering and Nanotechnology, BMRC, A*STAR ofSingapore, and grants from Janssen Cilag (R-185-000-182-592),Singapore-MIT Alliance Computational and Systems Biology (C-382-641- 001-091), SMART BioSyM and Mechanobiology Instituteof Singapore (R-714-001-003-271) to Hanry Yu.

References

[1] Baccarani U, Sanna A, Cariani A, Sainz M, Adani GL, Lorenzin D, et al. Cry-opreserved human hepatocytes from cell bank: in vitro function and clinicalapplication. Transplant Proc 2005;37(1):256e9.

[2] Terry C, Mitry RR, Lehec SC, Muiesan P, Rela M, Heaton ND, et al. The effects ofcryopreservation on human hepatocytes obtained from different sources ofliver tissue. Cell Transplant 2005;14(8):585e94.

[3] Dvir-Ginzberg M, Elkayam T, Aflalo E, Agbaria R, Cohen S. Ultrastructural andfunctional investigations of adult hepatocyte spheroids during in vitro culti-vation. Tissue Eng 2004;10(11âV“12):1806e17.

[4] Hewitt NJ, Utesch D. Cryopreserved rat, dog and monkey hepatocytes:measurement of drug metabolizing enzymes in suspensions and cultures.Hum Exp Toxicol 2004;23(6):307e16.

[5] Darr TB, Hubel A. Postthaw viability of precultured hepatocytes. Cryobiology2001;42(1):11e20.

[6] Kravchenko LP, Petrenko AY, Somov AY, Grischenko VI, Fuller BJ. Respiratoryactivity of isolated rat hepatocytes following cold storage and subsequentrewarming: a comparison of sucrose-based and University of Wisconsinsolutions. Cryobiology 2001;42(3):218e21.

[7] Tan FCK, Lee KH, Gouk SS, Magalhaes R, Poonepalli A, Hande MP, et al. Opti-mization of cryopreservation of stem cells cultured as neurospheres:comparison between vitrification, slow-cooling and rapid cooling "freezing"protocols. Cryoletters 2007;28(6):445e60.

[8] Hoshiba T, Nagahara H, Cho CS, Tagawa Y, Akaike T. Primary hepatocytesurvival on non-integrin-recognizable matrices without the activation of Aktsignaling. Biomaterials 2007;28(6):1093e104.

[9] Pinkse GG, Voorhoeve MP, Noteborn M, Terpstra OT, Bruijn JA, De Heer E.Hepatocyte survival depends on beta1-integrin-mediated attachment ofhepatocytes to hepatic extracellular matrix. Liver Int 2004;24(3):218e26.

[10] Smets FN, Chen Y, Wang LJ, Soriano HE. Loss of cell anchorage triggersapoptosis (anoikis) in primary mouse hepatocytes. Mol Genet Metab 2002;75(4):344e52.

[11] Thomas FT, Contreras JL, Bilbao G, Ricordi C, Curiel D, Thomas JM. Anoikis,extracellular matrix, and apoptosis factors in isolated cell transplantation.Surgery 1999;126(2):299e304.

[12] Zvibel I, Smets F, Soriano H. Anoikis: roadblock to cell transplantation? CellTransplant 2002;11(7):621e30.

R. Magalhães et al. / Biomaterials 33 (2012) 829e836836

[13] Maganto P, Cienfuegos JA, Santamaria L, Eroles G, Andres S, Castillo-Olivares JL, et al. Cryopreservation and transplantation of hepatocytes: anapproach for culture and clinical application. Cryobiology 1988;25(4):311e22.

[14] Guillouzo A, Rialland L, Fautrel A, Guyomard C. Survival and function of iso-lated hepatocytes after cryopreservation. Chem Biol Interact 1999;121(1):7e16.

[15] Skett P, Roberts P, Khan S. Maintenance of steroid metabolism and hormoneresponsiveness in cryopreserved dog, monkey and human hepatocytes. ChemBiol Interact 1999;121(1):65e76.

[16] Westerink WM, Schoonen WG. Cytochrome P450 enzyme levels in HepG2cells and cryopreserved primary human hepatocytes and their induction inHepG2 cells. Toxicol in Vitro 2007;21(8):1581e91.

[17] Westerink WM, Schoonen WG. Phase II enzyme levels in HepG2 cells andcryopreserved primary human hepatocytes and their induction in HepG2cells. Toxicol in Vitro 2007;21(8):1592e602.

[18] Muller P, Aurich H, Wenkel R, Schaffner I, Wolff I, Walldorf J, et al. Serum-freecryopreservation of porcine hepatocytes. Cell Tissue Res 2004;317(1):45e56.

[19] Naramoto A, Ohno S, Furuta K, Itoh N, Nakazawa K, Nakano M, et al. Ultra-structural studies of hepatocyte cytoskeletons of phalloidin-treated rats byquick-freezing and deep-etching method. Hepatology 1991;13(2):222e9.

[20] Poyck PP, Hoekstra R, Chhatta A, Bloemendaal LT, van Wijk AC, Galavotti D,et al. Time-related analysis of metabolic liver functions, cellular morphology,and gene expression of hepatocytes cultured in the bioartificial liver of theAcademic Medical Center in Amsterdam (AMC-BAL). Tissue Eng 2007;13(6):1235e46.

[21] Stefanovich P, Ezzell RM, Sheehan SJ, Tompkins RG, Yarmush ML, Toner M.Effects of hypothermia on the function, membrane integrity, and cytoskeletalstructure of hepatocytes. Cryobiology 1995;32(4):389e403.

[22] Pritchard S, Erickson GR, Guilak F. Hyperosmotically induced volume changeand calcium signaling in intervertebral disk cells: the role of the actin cyto-skeleton. Biophys J 2002;83(5):2502e10.

[23] Ohta M, Okanoue T, Takami S, Nagao Y, Mori T, Hori N, et al. Morphologicalalterations of gap junctions in phalloidin-treated rat livers. J Gastroenterol1994;29(2):172e9.

[24] Morisset C, Gazeau C, Hansz J, Dereuddre J. Is actin important for cryosurvival.Cryo-Letters 1994;15(4):215e22.

[25] Makarevich AV, Chrenek P, Olexikova L, Popelkova M, Turanova Z, Ostro A,et al. Post-thaw survival, cell death and actin cytoskeleton in gene-microinjected rabbit embryos after vitrification. Theriogenology 2008;70(4):675e81.

[26] Hosu BG, Mullen SF, Critser JK, Forgacs G. Reversible disassembly of the actincytoskeleton improves the survival rate and developmental competence ofcryopreserved mouse oocytes. PLoS ONE 2008;3(7):e2787.

[27] Ebner H, Cordas A, Pafundo D, Schwarzbaum P, Pelster B, Krumschnabel G.Importance of cytoskeletal elements in volume regulatory responses of trouthepatocytes. Am J Physiol Regul Integr Comp Physiol 2005;289(3):R877e90.

[28] Miettinen A, Virtanen I, Linder E. Cellular actin and junction formation duringreaggregation of adult rat hepatocytes into epithelial cell sheets. J Cell Sci1978;31:341e53.

[29] Storch W. [Immunohistochemical localization of actin and myosin in liver,kidney, stomach, heart and skeletal muscle: a reference to a cytoplasmic actinfibrillar network in liver cells (author’s transl)]. Acta Histochem 1981;68(2):208e17.

[30] Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol 1976;13:29e83.

[31] Magalhaes R, Wang XW, Gouk SS, Lee KH, Ten CM, Yu H, et al. Vitrificationsuccessfully preserves hepatocyte spheroids. Cell Transplant 2008;17(7):813e28.

[32] Magalhaes R, Anil Kumar PR, Wen F, Zhao X, Yu H, Kuleshova LL. The use ofvitrification to preserve primary rat hepatocyte monolayer on collagen-coatedpoly(ethylene-terephthalate) surfaces for a hybrid liver support system.Biomaterials 2009;30(25):4136e42.

[33] Erickson GR, Northrup DL, Guilak F. Hypo-osmotic stress induces calcium-dependent actin reorganization in articular chondrocytes. OsteoarthritisCartilage 2003;11(3):187e97.

[34] Eschbach E, Chatterjee SS, Noldner M, Gottwald E, Dertinger H, Weibezahn KF,et al. Microstructured scaffolds for liver tissue cultures of high cell density:morphological and biochemical characterization of tissue aggregates. J CellBiochem 2005;95(2):243e55.

[35] Liu BL, McGrath JJ. Effects of freezing on the cytoskeleton, focal adhesions andgap-junctionsin murine osteoblast cultures. Conf Proc IEEE Eng Med Biol Soc2005;5:4896e9.

[36] Lange K, Gartzke J. A critical comparison of the current view of Ca signalingwith the novel concept of F-actin-based Ca signaling. Crit Rev Eukaryot GeneExpr 2006;16(4):307e65.

[37] Abu-Qare A, Abou-Donia M. Biomarkers of apoptosis: release of cytochrome c,activation of caspase-3, induction of 8-hydroxy-2’-deoxyguanosine, increased3-nitrotyrosine, and alteration of p53 gene. J Toxicol Environ Health B Crit Rev2001;4(3):313e32.

[38] Pervaiz S, Seyed M, Hirpara J, Clément M, Loh K. Purified photoproducts ofmerocyanine 540 trigger cytochrome C release and caspase 8-dependentapoptosis in human leukemia and melanoma cells. Blood 1999;93(12):4096e108.

[39] Hirpara J, Seyed M, Loh K, Dong H, Kini R, Pervaiz S. Induction of mitochon-drial permeability transition and cytochrome C release in the absence ofcaspase activation is insufficient for effective apoptosis in human leukemiacells. Blood 2000;95(5):1773e80.

[40] Hanai A, Yang WL, Ravikumar TS. Induction of apoptosis in human coloncarcinoma cells HT29 by sublethal cryo-injury: Mediation by cytochrome Crelease. Int J Cancer 2001;93(4):526e33.

[41] Slee E, Adrain C, Martin S. Serial killers: ordering caspase activation events inapoptosis. Cell Death Differ 1999;6(11):1067e74.

[42] Kuleshova LL, Wang XW, Wu YN, Zhou Y, Yu H. Vitrification of encapsulatedhepatocytes with reduced cooling and warming rates. Cryo Lett 2004;25(4):241e54.

[43] Wu YN, Yu HR, Chang S, Magalhaes R, Kuleshova LL. Vitreous cryopreservationof cell-biomaterial constructs involving encapsulated hepatocytes. Tissue Eng2007;13(3):649e58.