In the December 2005 Issue of Liver Transplantation, the Following article: Terry C, Dhawan A, Mitry RR, Lehec SC, and Hughes RD. Preincubation of Rat and Human Hepatocytes with Cytoprotectants Prior to Cryopreservation Can Improve Viability and Function Upon Thawing. Liver Transpl 2005;11:1533–1540 was printed without its figures included. The following is the manuscript in its entirety with the figures present.
Successful cryopreservation of human hepatocytes is important for the treatment of liver disease by hepatocyte transplantation and also for the use of hepatocytes as an in vitro model in toxicity and drug metabolism testing. A number of human hepatocyte cryopreservation protocols are currently in use.1–10 A fundamental limitation to thawed cell function is often the quality of hepatocytes before cryopreservation, which usually determines the quality of hepatocytes obtained upon thawing. The initial quality is dependent upon two major factors. First is the nature of the tissue from which the hepatocytes are isolated,11 including factors such as age or condition of the donor, fat content, and cold and warm ischemia times involved in procuring the liver. Second is the isolation procedure, including factors such as the time taken, type or batch of collagenase used, and the level of aeration during isolation. Postisolation, the first stage of the cryopreservation procedure is amenable to treatments that could improve hepatocyte function after thawing such as preincubation of hepatocytes with protective agents to help the cells withstand the effects of freezing.
There are limited data on preincubation of hepatocytes prior to cryopreservation. Preincubation in culture media containing glucose at 37°C prior to cryopreservation was found to be beneficial to both rat and human hepatocytes. Upon thawing, preincubated rat hepatocytes exhibited improved viability and energy status.12 Human hepatocytes had viability and urea production similar to that of fresh hepatocytes and had significantly higher plating efficiency13 than hepatocytes cryopreserved without preincubation. It was thought that the mechanism for the beneficial effect of glucose was through boosting cellular adenosine triphosphate (ATP) levels before cryopreservation. Maintenance of ATP levels has been shown to be important during cryopreservation11 and also allows recovery of the hepatocytes from the “trauma” of isolation.
This study investigates the effect of preincubation of rat and human hepatocytes with cytoprotectants prior to cryopreservation on the subsequent viability and function of the cells after thawing.
ATP, adenosine triphosphate; LDH, lactate dehydrogenase; UW, University of Wisconsin; WEM, Williams' medium E; CYP1A1/2, cytochrome P4501A1/2; GSH, glutathione; SAMe, S-adenosyl-L-methionine; NAC, N-acetylcysteine; PTX, pentoxyfylline; PGE, prostaglandin; TUDCA, tauroursodeoxycholic acid.
MATERIALS AND METHODS
Rat liver tissue was obtained from male Sprague-Dawley rats (Harlan Olec, Bicester, UK). Animals were allowed to acclimatize to housing conditions (room temperature 21 ± 2°C; humidity 55 ± 10%) for at least 7 days before use. Rats weighing from 250 - 280 g were used for hepatocyte isolation. Rats were injected intraperitoneally with 500 U heparin solution (Multiparin®, Holder CP Pharmaceuticals Ltd., Wrexham, UK). Ten minutes after injection, the animals were killed by decapitation, the abdomen opened, and the liver immediately excised. The liver was kept in ice-cold Dulbecco's phosphate-buffered solution until the perfusion process began (maximum of 10 minutes' cold ischemia time). The culture media used were obtained from BioWhittaker (Berkshire, UK). Isolation of rat hepatocytes was carried out using an ex situ collagenase (type P, Roche Diagnostics Ltd., East Sussex, UK) perfusion method.14, 15 Viability was determined using the trypan blue exclusion method.
Human liver tissue was obtained from donor tissue rejected or unused for orthotopic liver transplantation at King's College Hospital. All donor tissues were delivered to the laboratory flushed with University of Wisconsin (UW) solution (Bristol-Myers Squibb Pharma Ltd., Hounslow, UK) and maintained in this solution on ice. All tissue was consented for research in accordance with the Research Ethics Committee of King's College Hospital. Isolation of human hepatocytes was carried out using a collagenase (type P) perfusion method15 with some modifications.16
Rat hepatocytes were used to screen 10 compounds (all obtained from Sigma-Aldrich Company Ltd., Dorset, UK), each at three concentrations and at both 4°C and 37°C, for their efficacy in preincubation experiments (Table 1). Human hepatocytes were then used to further investigate 5 of the original 10 compounds that had given significant effects in rat hepatocytes, each at both 4°C and 37°C. All compounds were tested at 4°C because this is the standard hypothermic storage temperature, and although metabolism is slowed at this temperature, it is not stopped, so compounds could still enter the hepatocytes and exert a beneficial effect. Any beneficial effects seen at 4°C also may enable extension of the time hepatocytes can currently be stored at 4°C before use or before cryopreservation. All compounds also were tested at 37°C because this is the standard culture temperature and hepatocytes should be able to metabolize and utilize any beneficial compounds. Control hepatocytes (3 × 106 viable cells/ml) were immediately cryopreserved in UW solution with 10% DMSO (v/v final concentration) after isolation (standard technique). Preincubated hepatocytes were suspended in the appropriate preincubation media (serum-free Williams' medium E [WEM] plus test compound) and were incubated at either 4°C or 37°C. A serum-free, WEM-only preincubation control also was included in each experiment. Each treatment consisted of 3 × 106 viable hepatocytes per ml preincubation media (1.5 × 107 cells in 5 ml preincubation media). All preincubation media were freshly prepared immediately prior to use and pH checked and adjusted to pH 7.4 if necessary. After 2 hours of incubation, the tubes were centrifuged at 50g for 5 minutes at 4°C to pellet the hepatocytes. The supernatant was then removed, replaced with ice-cold UW solution to the total volume (cells and UW solution) of 4.5 ml, and transferred to ice-cold 5 ml cryovials. An amount of 0.5 ml DMSO was added drop-wise to each cryovial to give a final DMSO concentration of 10% (v/v), and hepatocytes were then immediately cryopreserved. For both standard and preincubated hepatocytes, cryopreservation was undertaken using a controlled-rate freezer (Kryo 10, Series III, Planer Products Ltd., Middlesex, UK) and a step-wise freezing protocol4 with some modifications. Frozen cells were stored at −140°C for 2 weeks until thawing17 using WEM containing 10% FCS as the thawing media. The hepatocytes were then pelleted by centrifugation at 50g at 4°C for 5 minutes, and the pellet was resuspended in a known volume of WEM.
Table 1. Compounds Used for Preincubation of Rat and Human Hepatocytes at 4°C and 37°C
Denotes compounds used with both rat and human hepatocytes.
Cell viability was determined using the trypan blue exclusion method. Unless stated, all functional assays were performed on 30,000 viable cells/well (8 wells per sample) in 96-well flat-bottomed collagen-coated plates (Biocoat™, BD Bioscience, Oxford, UK) after 24 hours of culture. Hepatocytes were cultured in WEM 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.Attachment efficiency was determined by measuring the protein content18 of attached cells and that of the initial number of cells (30,000 total cells/well). LDH concentration was measured in supernatant and cell lysate samples using a CytoTox 96® Non-Radioactive Cytotoxicity Assay Kit (Promega, Southampton, UK), allowing LDH leakage to be calculated as a percentage of the total. Cytochrome P4501A1/2 (CYP1A1/2) activity was determined using the ethoxyresorufin O-deethylase method.19 This assay was carried out on hepatocytes plated at a density of 150,000 viable cells/well (24-well collagen-coated plate) after 24 hours of culture. A resorufin standard curve (0 - 800 pmol) was used to calculate CYP1A1/2 activity, expressed as pmol resorufin produced/min/mg cell protein. Albumin production was determined (in human hepatocytes only) using a Human Albumin ELISA Quantification Kit (Bethyl Laboratories, Inc., Biognosis, Hailsham, UK). An albumin standard curve (0 - 1,500 ng/ml) was used to calculate the albumin production expressed as μg albumin produced/h/mg cell protein.
All results are presented as mean ± standard deviation. Statistical analysis of data was performed by comparing means using ANOVA with repeated measurements.
Hepatocytes were isolated from 15 rat tissue samples yielding cells with a mean fresh viability of 87 ± 10%, mean attachment efficiency of 80 ± 13%, mean CYP1A1/2 activity of 16.1 ± 1.5 pmol/min/mg protein, and mean LDH leakage of 20 ± 4%.
There was no significant difference between standard rat hepatocytes and those preincubated at either 4°C or 37°C in serum-free WEM only.
Rat hepatocytes preincubated with glucose at 4°C had significantly higher viability (100 mM = 59 ± 11%, P = 0.003; 200 mM = 60 ± 9%, P = 0.002; 300 mM = 59 ± 11%, P = 0.002) and attachment efficiency (100 mM = 69 ± 12%, P = 0.003; 200 mM = 68 ± 5%, P = 0.004; 300 mM = 69 ± 7%, P < 0.001) than standard hepatocytes (viability = 50 ± 8%; attachment efficiency = 51 ± 5%, Fig. 1) after cryopreservation. There was no significant concentration-dependent effect of glucose at 4°C. In addition, there was no significant effect of glucose preincubation at 37°C with any concentration used.
Rat hepatocytes preincubated with fructose at 4°C had significantly higher viability (100 mM = 53 ± 3%, P = 0.005; 200 mM = 55 ± 4%, P < 0.001; 300 mM = 53 ± 4%, P < 0.001) and attachment efficiency (100 mM = 62 ± 11%, P < 0.001; 200 mM = 66 ± 12%, P < 0.001; 300 mM = 65 ± 10%, P < 0.001) than standard hepatocytes (viability = 43 ± 6%; attachment efficiency = 49 ± 12%, Fig. 2) after cryopreservation. There was no significant concentration-dependent effect of fructose at 4°C, nor was there any significant effect of fructose preincubation at 37°C with any concentration used.
Rat hepatocytes preincubated with α-lipoic acid at 4°C had significantly higher viability (0.5 mM = 64 ± 9%, P < 0.001; 2.5 mM = 61 ± 11%, P < 0.001; 5 mM = 61 ± 10%, P < 0.001) and attachment efficiency (0.5 mM = 71 ± 7%, P = 0.004; 2.5 mM = 71 ± 9%, P = 0.003; 5 mM = 70 ± 10%, P = 0.004) than standard hepatocytes (viability = 51 ± 7%; attachment efficiency = 62 ± 7%, Fig. 3) after cryopreservation. Also, hepatocytes preincubated with α-lipoic acid at 37°C had significantly higher viability (0.5 mM = 61 ± 8%, P < 0.001; 2.5 mM = 63 ± 8%, P < 0.001; 5 mM = 61 ± 7%, P < 0.001) and attachment efficiency (0.5 mM = 67 ± 10, P = 0.005; 2.5 mM = 73 ± 12%, P < 0.001; 5 mM = 69 ± 10%, P < 0.001) than standard hepatocytes after cryopreservation. There was no significant concentration-dependent increase in the positive effects at either 4°C or 37°C.
Rat hepatocytes preincubated with pentoxifylline at both 4°C and 37°C had significantly lower viability and significantly higher LDH leakage than standard hepatocytes after cryopreservation, as well as lower attachment efficiency and CYP1A1/2 activity when incubated at 37°C (Table 2). Rat hepatocytes preincubated with 3 × 10−5 mM PGE1 at 37°C had significantly higher LDH leakage than standard hepatocytes after cryopreservation (Table 2).
Table 2. Effect on Function of Rat Hepatocytes of Preincubation with Compounds Prior to Cryopreservation
Glutathione, SAMe, ascorbic acid, NAC, and TUDCA preincubation had no significant effects on the viability or functional measurements in rat hepatocytes after cryopreservation (Table 2). There was no significant effect of any compound on the LDH leakage or CYP1A1/2 activity of rat hepatocytes after preincubation.
Hepatocytes were isolated from 10 human tissue samples yielding cells with a mean fresh viability of 67 ± 10%, mean attachment efficiency of 55 ± 15%, mean LDH leakage of 19 ± 5%, mean CYP1A1/2 activity of 15.9 ± 2.4 pmol/min/mg protein, and mean albumin production of 12.5 ± 1.9 μg/h/mg protein.
Human hepatocytes preincubated with glucose at 4°C had significantly higher viability (100 mM = 55 ± 6%, P = 0.004; 200 mM = 57 ± 8%, P = 0.006; 300 mM = 61 ± 5%, P = 0.003) and significantly lower LDH leakage (100 mM = 19 ± 4%, P = 0.002; 200 mM = 15 ± 3%, P = 0.001; 300 mM = 18 ± 4%, P = 0.003) than standard hepatocytes (viability = 45 ± 6%; LDH leakage = 26 ± 3%) after cryopreservation (Fig. 4). There was a significant (P = 0.02) correlation between the concentrations of glucose used at 4°C and the increase in viability upon thawing. There was no significant effect of glucose preincubation at 37°C with any concentration used.
Human hepatocytes preincubated with fructose at 4°C had significantly higher attachment efficiency (100 mM = 57 ± 7%, P = 0.04; 200 mM = 61 ± 7%, P = 0.03; 300 mM = 64 ± 8%, P = 0.02) than standard hepatocytes (45 ± 6%) after cryopreservation (Fig. 5). Human hepatocytes preincubated with fructose at 37°C also had significantly higher attachment efficiency (100 mM = 54 ± 6%, P = 0.04; 200 mM = 59 ± 8%, P = 0.02; 300 mM = 60 ± 7%, P = 0.04) than standard hepatocytes after cryopreservation. Preincubation with 200 and 300 mM fructose at 4°C also gave significantly higher albumin production (200 mM = 5.2 ± 0.6 μg/h/mg protein, P = 0.03; 300 mM = 5.6 ± 0.7 μg/h/mg protein, P = 0.01) than standard hepatocytes (4.2 ± 0.4 μg/h/mg protein) after cryopreservation.
Human hepatocytes preincubated with α-lipoic acid at 4°C had significantly higher viability (0.5 mM = 52 ± 5, P = 0.04; 2.5 mM = 59 ± 8%, P = 0.01; 5 mM = 60 ± 7%, P = 0.007) than standard hepatocytes (46 ± 5%, Fig. 6). Preincubation with 0.5, 2.5, and 5 mM at 4°C and 0.5 and 2.5 mM at 37°C also gave significantly higher attachment efficiency (4°C: 0.5 mM = 56 ± 5%, P = 0.002; 2.5 mM = 60 ± 8%, P = 0.001; 5 mM = 68 ± 7%, P < 0.0001; and 37°C: 0.5 mM 56 ± 6%, P = 0.01; 2.5 mM = 60 ± 7%, P = 0.002) than standard hepatocytes (42 ± 6%). Albumin production was significantly higher in human hepatocytes preincubated with 2.5 mM α-lipoic acid at 4°C (5.9 ± 0.4 μg/h/mg protein, P = 0.02) than standard hepatocytes (4.9 ± 0.5 μg/h/mg protein). In this experiment, human hepatocytes preincubated in WEM alone at 4°C also had significantly higher attachment efficiency (50 ± 7%, P = 0.02).
There was no significant difference between standard human hepatocytes and those preincubated for 2 hours at either 4°C or 37°C in serum-free WEM. The preincubation of human hepatocytes with SAMe and pentoxifylline in serum-free WEM had no significant effect on viability or functional measurements made after cryopreservation (Table 3).
Table 3. Effect on Function of Human Hepatocytes of Preincubation with Compounds Prior to Cryopreservation
Preincubation of both rat and human hepatocytes with either glucose, fructose, or α-lipoic acid at 4°C prior to cryopreservation had beneficial effects on the function of the cells upon thawing. The main effect of glucose preincubation on cryopreserved hepatocytes was improved viability and LDH retention, which indicates preservation of membrane integrity. This could be through preserved cellular structural and functional integrity by boosted cellular ATP content. The hypothesis that hepatocyte ATP content should be at the highest possible level before cryopreservation to allow preservation of cellular structural and functional integrity after thawing has been tested.12 The authors found that 15 mM glucose preincubation for 20 minutes brought cellular ATP levels to an optimum that is usually reached only after 1 hour of culture. Subsequently, the cellular contents of ATP, adenosine diphosphate (ADP), and adenosine monophosphate (AMP) were not significantly changed by cryopreservation. It seems likely therefore that glucose preincubation can boost the intracellular levels of ATP before freezing and allow hepatocytes to better withstand the cryopreservation process.
The effects seen also could be due to glucose reducing osmotic stress on the hepatocytes prior to, and during, cryopreservation. This could be due to glucose increasing the osmolality (×2) of the preincubation media, hence the hepatocytes may have begun to lose water through exosmosis. It is known that reducing the water content of cells leads to less intracellular ice-crystal formation and hence less damage to the hepatocytes upon thawing.
The beneficial effects of glucose preincubation at the concentrations used were restricted to 4°C. Indeed, although the bioavailability of components of preservation solutions at 4°C is lower than at physiological temperatures,20 indicating that the application of cytoprotectants at 37°C to organ donors might improve cold-storage preservation of livers, this was not found in our study. Succinic acid has been reported to be a gluconeogenic precursor and ATP-generating nutrient at 4°C, and thus it appears that temperature does not detract from this proposed mechanism.21 In our study, the effective glucose concentrations were higher than used in previous studies.12 The higher concentration used here may have allowed a preloading of the hepatocytes with glucose at low temperature, which may provide a source for energy metabolism on rewarming of the cells.
Fructose preincubation also had beneficial effects with both rat and human hepatocytes. There are several possible mechanisms for fructose preincubation improving the viability and function of cryopreserved hepatocytes. The major effect in our study was on the attachment efficiency of hepatocytes. This may indicate a possible protective role of fructose for cell detachment and apoptosis, and studies have found fructose to be a simple antiapoptotic compound. Fructose can protect against apoptosis in rat hepatocytes cultured in hypoxic conditions by forming additional nicotinamide adenine dinucleotide phosphate (NADPH) for the regeneration of reduced glutathione (GSH) via stimulation of the pentose phosphate pathway and thereby reducing generation of reactive oxygen species (ROS).22 Hepatocyte damage after oxidative stress occurs in the presence of the physiological substrate (glucose). It has been suggested that during hypoxia, increased cytoprotection might be achieved by alternative substrates that are faster metabolized via anaerobic glycolysis (i.e., fructose), the only mechanism for ATP generation in the absence of oxygen.23 Fructose (30 g/l) has been included in the digestion buffer for isolation of human hepatocytes from tissue to “reduce cell loss,” but the authors did not suggest why fructose was beneficial to the isolation process.24 It has been suggested that adding substrates such as fructose during isolation of hepatocytes may improve the energy recovery of hepatocytes after the stress of ischemia reperfusion and cell isolation.25 Rat hepatocytes stored under cold hypoxia exhibited improved viability and LDH retention with 10 - 20 mM fructose in the UW solution.26 During cold hypoxia, metabolic requirements decrease, but upon rewarming of the hepatocytes, fructose-1-phosphate enters the glycolytic pathway at the level of triose phosphates, allowing a high degree of glycolytic activity so that cells remain more viable than cells receiving other carbohydrates. Therefore, it seems possible that in our study, an accumulation of fructose-1-phosphate due to fructose preincubation prior to cryopreservation allowed a “fast reconditioning” of metabolic activities after thawing.
Preincubation with α-lipoic acid also had beneficial effects for rat and human hepatocytes at both 4°C and 37°C. The improved viability and LDH retention of preincubated hepatocytes suggests that α-lipoic acid is able to protect hepatocyte membranes from potential oxidative damage before cryopreservation. The compound is taken up rapidly by the cells and reduced to dihydrolipoic acid, which is released by the cells. It is present as a cofactor in α-keto-acid dehydrogenases and in the glycine cleavage system. These enzyme complexes are involved in the metabolic pathways of pyruvate oxidation, the citric acid cycle, and amino acid degradation and biosynthesis.27 Free α-lipoic acid also can act at various levels in biochemical pathways, being an enzyme substrate, a protein modifier (redox dependent and independent action), a scavenger of free-radical species, a metal chelator, and an enhancer of cellular glutathione status.27 α-lipoic acid can increase the synthesis of cellular glutathione by the metabolic reduction of α-lipoic acid to dihydrolipoic acid.28 Dihydrolipoic acid can then reduce cysteine, which is readily taken up by the neutral amino acid transport systems and used for glutathione synthesis. Culture with 0.1 to 0.5 mM α-lipoic acid increased intra- and extracellular GSH levels in HeLa and HepG2 cells,29 so it seems feasible that improved GSH levels before freezing could be responsible for the reduction in membrane damage during cryopreservation. Reduced LDH leakage has been shown during rat liver perfusion when treated with α-lipoic acid30 as a result of increased ATP content and inhibition of release of proinflammatory mediators. This membrane protection also may extend to preservation of cell attachment molecules on the plasma membrane, which could explain the improved attachment efficiency seen in our study. In addition, there is some evidence from the present study that α-lipoic acid also can improve synthetic (i.e., albumin production) and enzyme (i.e., CYP1A1/2 activity) function of the hepatocytes.
In conclusion, this study has shown that adding the relatively simple step of preincubation prior to cryopreservation of rat and human hepatocytes can boost the viability and function of cells upon thawing. This is particularly important with human hepatocytes as researchers often have little or no control over the tissue quality before isolation of hepatocytes. Therefore, in addition to continuing to improve the cell isolation protocol, preincubation of hepatocytes before freezing is another method of improving the quality of hepatocytes after cryopreservation.