Factors determining successful engraftment of hepatocytes and susceptibility to hepatitis B and C virus infection in uPA-SCID mice

Research Article

Factors determining successful engraftment of hepatocytes and susceptibility to hepatitis B and C virus infection in uPA-SCID mice Thomas Vanwolleghem1, Louis Libbrecht2, Bettina E. Hansen3, Isabelle Desombere1, Tania Roskams4, Philip Meuleman1, Geert Leroux-Roels1,* 1

Center for Vaccinology, Ghent University and Hospital, Gent, Belgium; 2Department of Pathology, Ghent University and Hospital, Gent, Belgium; 3 Erasmus MC University Medical Center, Department of Gastroenterology & Hepatology, and Biostatistics, Rotterdam, The Netherlands; 4 Department of Morphology and Molecular Pathology, University Hospital Leuven, Leuven, Belgium

See Editorial, pages 421–423

Background & Aims: The human liver-uPA+/+-SCID mouse is currently the best small animal model available for viral hepatitis infection studies. Methods: We identify critical factors affecting animal survival, engraftment efficacy, kinetics of liver repopulation and virological outcome by analysing the data from 400 human hepatocyte transplantations and 115 subsequent HBV and/or HCV inoculations in this mouse model. Results: Almost one third of animals succumbed during the first week after hepatocyte transplantation. Only during this critical period, liver necrosis due to embolization of donor cells in the portal vein was observed. This may have caused a fatal acute liver failure that complicated the pre-existing chronic liver disease. From the second week onwards, confluent hepatocyte clusters repopulated the liver and restored its synthetic functions as evidenced by increasing human albumin levels in plasma. Xenogenic repopulation by human cells proceeded approximately 4-times slower compared to allogenic mouse hepatocytes. All HBV inoculations were successful, even in animals with low graft take. HCV infection rate varied substantially, although every donor cell type yielded infectable animals. A reproducible 100% HCV infectivity was reached with high quality inocula in animals with human albumin plasma levels >1 mg/ml. Superior animal survival, adequate liver engraftment and a high viral infection rate were observed after transplanting cryopreserved commercial human hepatocytes.

Keywords: Hepatocyte transplantation; Human hepatocyte; Mouse hepatocyte; Repopulation; uPA-SCID mouse; HCV; HBV; Viral kinetics; Liver necrosis. Received 8 November 2009; received in revised form 25 March 2010; accepted 29 March 2010; available online 31 May 2010 q DOI of original article: 10.1016/j.jhep.2010.05.001. * Corresponding author. Address: Center for Vaccinology, Ghent University and Hospital, 1 Blok A, De Pintelaan 185, B-9000 Gent, Belgium. Tel.: +32 9 3323422; fax: +32 9 3326311. E-mail address: [email protected] (G. Leroux-Roels). Abbreviations: uPA, urokinase-type plasminogen activator; SCID, severe combined immune deficient; HBV, hepatitis B virus; HCV, hepatitis C virus; B6 mice, C57BL/ 6J mice; F, fresh hepatocytes; P1, Provider 1; P2, Provider 2; P3, Provider 3; IU, international units.

Conclusions: Our findings favour the use of commercially available, cryopreserved human hepatocytes for the humanization of the uPA+/+-SCID liver. While HBV infectivity criteria are less stringent, human albumin plasma levels exceeding 1 mg/ ml are required for a consistent HCV infection in chimeric mice. Ó 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.

Introduction Liver cell transplantation is a long awaited strategy to treat a variety of liver diseases. Current clinical successes are limited to some inherited metabolic liver diseases and in most cases only bridge the time to liver transplantation. This experience however shows that human hepatocytes can be safely injected in the portal circulation of patients [1]. The major hurdle for widespread clinical application of liver cell transplantation is the need for repeated infusion of large numbers of hepatocytes. Rodent models have shown that the underlying lack of liver repopulation can be overcome by inducing a selective proliferative advantage to the transplanted cells over the endogenous hepatocytes [2,3]. The transplantation of human hepatocytes in the uPA-SCID mouse has contributed to the understanding of hepatocyte engraftment into the liver [4], provided insight in human-type metabolization of steroids [5] and allowed for the study of human hepatotropic viruses, like hepatitis B virus (HBV) and hepatitis C virus (HCV) [4,6–11]. Widespread use of these chimeric mice is hampered by several logistic and technical difficulties. Optimization of this procedure is therefore of much interest. Analysis of the data collected from 400 consecutive human hepatocyte transplantations in uPA+/+-SCID mice and subsequent HBV and HCV inoculations, allowed us to discriminate different factors that have an impact on survival, engraftment efficacy, infection rate and kinetics of repopulation.

Journal of Hepatology 2010 vol. 53 j 468–476

JOURNAL OF HEPATOLOGY Materials and methods

Statistical analysis

Animals

Statistical analysis was performed with SPSS software for Windows version 15.0. Data are expressed as means ± standard deviation (SD) or median with range where appropriate. Survival analysis was performed using a Kaplan–Meier or Cox-Regression method. Statistical significance of differences between groups was tested with the Log Rank test, the Fisher’s Exact, the Mann–Whitney U and Kruskal–Wallis tests or one-way ANOVA where appropriate. Bonferroni correction was applied for multiple comparisons. Pearson’s correlation method and linear regression were performed to assess correlations between variables. Overall a probability value of p < 0.05 was considered significant.

Our monogamous uPA-heterozygous breeding strategy yields reproductive matings in a mean 55.6% of cases with on average 4 viable pups over a period of 6 weeks. Screening and genotyping of neonatal uPA-SCID mice identifies around 25% of litter as being homozygous for the uPA transgene [4]. For the 400 homozygous uPA-transgenic SCID pups used in this study, an average of 719 monogamous, uPA-heterozygous breeding couples were needed. C57BL/6J (B6) mice were originally purchased from Harlan Netherlands (Zeist, The Netherlands). All strains are bred and kept in our pathogen-free animal facility. Male B6 hepatocyte donors were between 3 and 5 months of age. The study protocol was approved by the Animal Ethics Committee of the Faculty of Medicine of the Ghent University.

Results

Liver samples and liver parenchymal cell preparation Fresh adult human hepatocytes (further called Fresh or F) obtained from patients undergoing a partial hepatectomy for liver metastasis of colorectal carcinoma, were isolated as described before [4]. These hepatocyte donors all gave written, informed consent and the experiments were approved by the Ethical Committee of the Ghent University Hospital. Cryopreserved human hepatocytes were purchased from Lonza (Verviers, Belgium; further called Provider 1 or P1), from BD Biosciences (Erembodegem, Belgium; further called Provider 2 or P2), or were obtained from an academic collaborator (further called Provider 3 or P3) [12]. For most transplantations with cells from P2, Percoll centrifugation was performed after thawing according to instructions of the vendor [13]. Murine liver cell suspensions were prepared via an in situ two-step collagenase perfusion through the portal vein, using Liver Perfusion Medium and Liver Digest Medium sequentially (both from Invitrogen) [14]. After low speed centrifugation the cell viability generally was around 40% as determined by Trypan Blue exclusion. The viability was improved to at least 64% by applying a one-step Percoll centrifugation as described [13]. Experimental design Donor liver fragments were kept in Viaspan solution (Bristol Myers Squibb) at 4 °C and used within 24 h after the surgical procedure. One to two week old uPA+/+-SCID mice were sedated using isoflurane anesthesia. After local asepsis with 70% ethanol, a minimal left flank incision through skin, abdominal wall and peritoneum were performed sequentially. The spleen was exposed and immobilized by applying gentle traction on its mesentery with an atraumatic thumb forceps. Between 0.5 and 1  106 viable cells were resuspended in 20 ll PBS and slowly injected with a 0.5 ml single-use insulin syringe in the lower pole of the spleen. After withdrawal of the syringe, local hemostasis was achieved by applying gentle pressure using a sterile cloth for at least 30 s, before both the abdominal wall and skin were closed separately with non-absorbable sutures. Transplanted animals were covered by their own nesting material and allowed to recover completely before the nursing mother was reintroduced in the cage. To minimize cannibalism, nursing mothers were sedated with 5 mg/kg diazepam intraperitoneally. Since transplanted uPA+/+ mice are weaker compared to nontransgenic littermates, weaning is not performed before 5–6 weeks of age, i.e. 3–4 weeks after transplantation. After weaning one retro-orbital blood sample was obtained for quantification of human albumin levels as described [4]. Depending on the graft take, surviving animals were used in different studies of either HBV or HCV infection and treatment, or pharmacokinetics and toxicology of selected compounds, as described [4,5,8,11]. In a separate experimental set-up the kinetics of the liver repopulation by human and murine hepatocytes were compared. Therefore animals were sacrificed at indicated time points early after transplantation to quantify the fraction of liver parenchyma occupied by transplanted human or murine hepatocytes. In addition, body weight was measured and albumin levels in mouse plasma were determined. These measurements were correlated with morphology. Control uPA+/+-SCID mice were injected with D-PBS (Invitrogen) and monitored until spontaneous death occurred or cachectic state required mercy killing. Liver tissue specimens, histochemistry and immunohistochemistry Histological analysis of liver samples including standard H&E staining, immunohistochemistry and quantification of the fraction of liver parenchyma occupied by donor hepatocytes is described in detail in the on-line supplement.

Transplantation of cryopreserved human hepatocytes results in a better survival rate of transplant recipients than that of fresh human hepatocytes In 44 consecutive sessions 400 uPA+/+-SCID mice were transplanted with human hepatocytes from 14 different human donors aged between 4 days and 81 years (Fig. 1, Table 1). Via the intra-splenic route mice were injected with fresh hepatocytes (F, n = 73), commercially available cryopreserved hepatocytes from P1 (n = 47) and P2 (n = 223) or with cryopreserved hepatocytes obtained from an academic collaborator (P3, n = 57). Within a 3-day period following transplantation a total of 51 animals suffered from excessive bleeding, rupture of the spleen, nursing problems or cannibalism by the mother or siblings elicited by the incision wound. This occurred independent of the type of donor cells (Table 1; p = 0.270) and was associated with an almost 2-fold increased risk of precocious death within 6 weeks after transplantation (hazard ratio = 1.7; p = 0.017). Overall 256 transplanted animals survived an extended fostering period of 3–4 weeks after transplantation, after which these animals were weaned and retro-orbital blood sampling was performed. Only three animals were sacrificed without taking blood because of their poor general condition (Fig 1). We and others have previously shown that human albumin levels in mouse plasma can be used as a marker for graft take and expansion [4,15]. Preliminary experiments indicated that graft expansion or a significant increase in human albumin levels is highly unlikely in animals with less than 60 lg/ml human albumin levels in their plasma. A total of 57 animals were therefore sacrificed immediately after blood sampling and human albumin quantification (Fig. 1, Supplementary Fig. 1). The sacrifice of these animals led to acute drops on the Kaplan–Meier survival curve around blood sampling time points, which was performed at day 21, day 28, day 32, day 35, day 38 and day 42 after transplantation. Overall, 165 animals survived up to 6 weeks after transplantation, while 60 were sacrificed and 175 died spontaneously during follow-up (Fig. 2A, Kaplan–Meier curve overall survival). As almost all animals were sacrificed because of unsuccessful engraftment, sacrifice cannot be regarded as an outcome of the cell transplantation procedure itself. Therefore, a ‘‘true” survival analysis censors all sacrificed animals at time of death (Fig. 2B Kaplan–Meier curve, censored animals indicated with ‘‘+”). This intervention did not change the overall survival curve during the first 3 weeks (Fig. 2A and B). Indeed, almost one third (28.3%) of the animals succumbed spontaneously within the first 7 days, including a 4% loss in the 2 h immediately following

Journal of Hepatology 2010 vol. 53 j 468–476

469

Research Article 400 uPA+/+SCID mice transplanted with human hepatocytes

n = 223 Commercial Cryo hepatocytes (Provider 2)

n = 47 Commercial Cryo hepatocytes (Provider 1)

n = 73 Fresh hepatocytes

•

3 peri-operative problems • None sacrificed

•

9 perioperative problems • None sacrificed

43 Survived until blood sampling

32 Survived until blood sampling

9 Sacrificed

28 peri-operative problems • 3 Sacrificed

130 Survived until blood sampling

24 Alive at wk 6 after Tx

101 Alive at wk 6 after Tx

• 14 Inoculated with HCV • 0 Inoculated with HBV

• 12 Inoculated with HCV • 2 Inoculated with HBV

• 70 Inoculated with HCV • 13 Inoculated with HBV

• 3 low albumin levels • 0 precocious death • 2 low quality inoculum

7 HCV+ 2 HBV+

48 Survived until blood sampling

30 Sacrificed

14 Sacrificed

26 Alive at wk 6 after Tx

4 HCV+

• 11 peri-operative problems • None sacrificed

•

4 Sacrificed

• 6 low albumin levels • 2 precocious death • 2 low quality inoculum

n = 57 Cryo hepatocytes from non-commercial source (Provider 3)

14 Alive at wk 6 after Tx • 4 Inoculated with HCV • 0 Inoculated with HBV

• 3 low albumin levels • 3 precocious death • 4 low quality inoculum • 7 neutralization study

53 HCV+ 13 HBV+

No Infectivity Problems

4 HCV+

Fig. 1. Critical points during clinical follow-up of uPA+/+-SCID mice after intra-splenic transplantation with human hepatocytes. Immediately following transplantation several animals experience peri-operative problems associated with worse outcome. During follow-up only three animals were sacrificed because of a bad general condition, before weaning and blood sampling performed at days 21, 28, 32, 35, 38, and 42 after transplantation. Thereafter, another cohort of animals was sacrificed because of low or undetectable human albumin plasma levels. Finally, depending on graft take, remaining animals were used in HBV and HCV infection studies. (Tx, transplantation. Wk, week).

surgery. During weeks 2 and 3 after transplantation, while the animals were further fostered, survival remained fairly stable. Thereafter the stressful events of weaning and blood sampling were associated with a progressive loss of transplanted animals, resulting in a true survival rate of 54% at week 6 after transplantation (Fig. 2B). A Cox-Regression analysis identified the peri-operative problems discussed above (p = 0.017), as well as the donor cell type as independent risk factors for survival (overall, p = 0.009). When the animals that suffered from peri-operative problems were excluded from the analysis, it became clear that animals transplanted with cryopreserved hepatocytes survived on average 5– 10 days longer than mice receiving fresh hepatocytes (Table 1, p = 0.023 for F versus P1; p = 0.046 for F versus P2; p = 0.005 for F versus P3) (Fig. 3). This difference in survival was already apparent from the 1st day after transplantation and remained so during the 42-day follow-up period. The observed differences in survival between the cryopreserved cells (P1, P2 and P3) were not significant, but there was a trend towards a worse survival for 470

animals receiving P2 cells (p = 0.064 for P2 versus P3; p = 0.279 for P2 versus P1). Efficacy of hepatocyte engraftment in uPA+/+-SCID mice To qualitatively estimate the overall engraftment of a certain donor cell type, the proportion of animals with albumin levels above the lower limit of detection of our ELISA assay (>60 lg/ ml) was determined. Transplantation with P3 cells yielded a much lower overall engraftment (Table 1, p <0.001), than all other donor cell types for which overall engraftment did not differ significantly (p >0.29 for all comparisons). We previously demonstrated that ongoing repopulation of the murine liver by fresh human hepatocytes is reflected by increasing human albumin plasma levels [4]. In the current transplantation series, a similar expansion with time is observed for cryopreserved cells (P2 cell donor, r = 0.501, p <0.001; Supplementary Fig. 2). In order to compare human albumin levels quantitatively, irrespective of the time of blood sampling, normalized

Journal of Hepatology 2010 vol. 53 j 468–476

JOURNAL OF HEPATOLOGY Table 1. Clinical characteristics and outcomes after human hepatocyte transplantation and subsequent HCV and/or HBV inoculation in uPA+/+-SCID mice. Fresh

# donors # transplantations donor age, median (range), % donor cell viability, median (range) # animals transplanted acceptor age, mean (SD) days peri-operative problems, % sacrificed, % survival time, mean (SD) dayso 6-weeks overall survival rate, % # animals blood sampled efficacy of engraftment  % overall efficacy of engraftmentà, % normalized hu alb level, mean (SD) ug/ml # animals inoculated with HCV and/or HBV§ HCV infection rate, % HBV infection rate, % HBV/HCV coinfection rate, % overall efficacy of infection, %

n=6 n=6 63.5 yrs (58–81) yrs 82.9 (62.5–97.1) n = 73 13.4 (3.7) 12.3 (n = 9) 12.3 (n = 9) 24.6 (2.4) 35.6 (n = 26) n = 43 79.1 (n = 34) 46.6 (n = 34) 166.8 (180.7) n = 14 + 0 28.6 (n = 4) 0 0 5.5 (n = 4)

Cryopreserved

p value

Provider 1 (com.)

Provider 2 (com.)

Provider 3 (non-com.)

n=2 n=6 4 months (4 days–4 months) 85.2 (61.1–92.3) n = 47 12.8 (2.5) 6.4 (n = 3) 8.5 (n = 4) 34.9 (2.0) 51.1 (n = 24) n = 32 78.1 (n = 25) 53.2 (n = 25) 146.5 (135.0) n = 12 + 2 58.3 (n = 7) 100.0 (n = 2) 0 19.1 (n = 9)

n=1 n = 25 27 yrs (single donor) 87.2 (61.8–96.2) n = 223 13.8 (4.4) 12.6 (n = 28) 7.6 (n = 17) 29.9 (1.2) 45.3 (n = 101) n = 130 91.5 (n = 119) 53.4 (n = 119) 457.2 (419.0) n = 70 + 13 74.7 (n = 53) 100.0 (n = 13) 100.0 (n = 3) 28.3 (n = 63)

n=5 n=7 18 yrs (16–44) yrs 64.2 (48.5–77.0) n = 57 14.2 (3.2) 19.3 (n = 11) 52.6 (n = 30) 34.4 (2.1) 24.6 (n = 14) n = 48 25.0 (n = 12) 21.1 (n = 12) 16.3 (32.2) n=4+0 100.0 (n = 4) 0 0 7.0 (n = 4)

p = 0.010*

p = 0.003* p = 0.276** p = 0.270* p <0.001* p = 0.013* p = 0.013* p <0.001* p <0.001* p <0.001** p <0.001*

p <0.001*

yrs, years. com., commercial. non-com., non-commercial. * Kruskal–Wallis test or Log Rank test. ** One-way ANOVA. # Indicates number. ° After exclusion of animals with peri-operative problems.   Fraction of blood sampled animals with albumin levels >60 lg/ml (ELISA detection limit). à Fraction of all transplanted animals with albumin levels >60 lg/ml (ELISA detection limit). § Animals inoculated with HCV and/or HBV: n indicates the number of animals inoculated with HCV or HBV at both sides of the plus sign, respectively.

human albumin levels were calculated. Thus, human albumin values were divided by the time point (expressed in weeks after transplantation) at which the sample was obtained. Transplantation with P2 cells resulted in significantly higher normalized albumin levels compared to all other cell types (p <0.001; Table 1, Supplementary Fig. 1). This difference persisted after repeating the analysis with a random selection of 40–50 animals transplanted with P2 cells, which excludes a selection bias. HBV and HCV infection studies in surviving animals Apart from a high albumin production, the optimal donor cell type should yield a high proportion of transplanted animals that can be used for further experimentation. Remarkably, even donor cell types with low engraftment efficacy yielded animals that could be infected with HCV (Table 1). The highest overall infectivity rates were obtained in animals transplanted with P1 and P2 cells, of which roughly 20% and 30% could be infected with HCV and/or HBV, respectively (overall p <0.001; p = 0.20 for P1 versus P2, Table 1). Of 100 animals inoculated with different genotypes and strains of HCV, a total of 68 became HCV RNA positive within 3 weeks. Analysis of the remaining 32 uninfected animals showed that five animals died before a blood sample could be drawn to measure viremia (Fig. 1). Seven of these 32 animals were regarded as protected from infection during neutralization experiments [9,10]. Another 12 animals had low human albumin

levels of less than 1 mg/ml either on inoculation or during the course of the infection experiment. The remaining eight uninfected animals had good infectable hepatocyte grafts but received either low quality and/or low genome titered inocula: five animals were injected with 1.4 to 5.0  105 IU of a HC-J4 genotype 1b viral pool that underwent at least three freeze–thaw cycles (a similar HC-J4 pool with only one freeze–thaw cycle was 100% infectious at 5  103 IU in 2 animals); one of two animals inoculated with 1.6  104 IU of the ED43 genotype 4a strain became HCV RNA positive (a minimal higher dose of 2.0  104 IU ED43 was 100% infectious for four animals and in later neutralization experiments [10]); and finally, two animals were injected with only 800 IU of the patient H derived acute phase H77 genotype 1a virus that underwent at least two freeze–thaw cycles and was kept frozen for more than 25 years [16]. Based on these observations we applied the following quality criteria for HCV infectivity in further experimentation with a 100% success rate in positive control animals [9–11]: (1) Human albumin level well above 1 mg/ml on inoculation and during the course of the experiment. (2) Inoculum of at least 1  104 IU HCV RNA. (3) Inoculum undergoing no more than one freeze–thaw cycle. The short term viral kinetics in HCV-infected transplanted animals is characterized by a steep increase in viremia during the first

Journal of Hepatology 2010 vol. 53 j 468–476

471

Research Article

Cumulative survival (%)

100

80

60

40

20

0 0

7

14

21

28

35

42

Post Tx (days) Fig. 3. Kaplan–Meier curve comparing true survival by donor type after exclusion of peri-operative problems. Within 1 day after transplantation, a quarter of animals transplanted with Fresh human hepatocytes already succumbed. At that time, only a limited loss is observed in the group of animals transplanted with any type of cryopreserved cells (Provider 1, 2, and 3 cells). The observed survival benefit for cryopreserved cells remains during the 42-day posttransplantation follow-up period (p = 0.013) (Vertical line at day 1 after transplantation. Censored animals indicated with ‘‘+” in respective color. Post Tx, after transplantation).

Cumulative survival (%)

100

80

60

HCV infection, largely divergent HBV viremia ranging between 2.33 and 6.05 log10 IU/ml were measured 4 weeks after injection with similar high-titered HBV inocula. Furthermore, HBV viremia did not correlate with measured human albumin levels (r = 0.356; p = 0.233), which confirms our previous observation [17].

40

20

0 0

7

14

21

28

35

42

Which donor hepatocyte qualities and acceptor mice characteristics determine survival and engraftment efficacy?

Post Tx (days) Fig. 2. Kaplan–Meier curve for overall and true survival of uPA+/+-SCID mice after receiving human hepatocytes transplants. (A) Overall survival: deaths due to sacrifice are clustered around bleeding timepoints (at days 21, 28, 32, 35, 38, and 42 after transplantation), resulting in acute drops on the survival curve. (B) True survival: sacrificed animals are considered as lost to follow-up (‘‘censored”, indicated with ‘‘+”). On both curves three phases can be discerned: a rapid spontaneous loss of animals during the first week after transplantation. The next 2 weeks (from week 2 to 3 after transplantation) are characterized by a stable situation during further fostering of transplanted animals. From day 21 after transplantation onwards, a more steady decline in survival is observed coinciding with weaning and blood sampling, resulting in a 6-week post-transplantation ‘‘overall” and ‘‘true” survival rate of 41.3% and 54.0%, respectively. (Post Tx, after transplantation).

two weeks, that is leveling-off thereafter at 6–7 log10 IU/ml [9–11]. In the long term, HCV RNA levels dropped steadily over the course of several months until animals died spontaneously or required mercy killing (Fig. 4). This drop correlated well with slowly decreasing human albumin levels (r = 0.559; p = 0.001) and has led to graft loss and resolution of J6-JFH-1 infection in one animal after at least 21 weeks of high-titered viremia (data not shown). All HBV inoculated animals became HBV DNA positive within 3 weeks after infection. Interestingly, the graft requirements for HBV infectivity are less stringent, as one animal with human albumin levels of 0.5 mg/ml was readily infected. Furthermore, inoculation of three animals with a long standing HCV infection led to HBV co-infection (Table 1). Compared to mono-infected animals with identical strains, co-infection did neither influence hepatitis B or C viremia, nor graft function during a 3-week follow-up period. In contrast to the reproducible fast viral kinetics seen after 472

Donor cell type not only determined animal survival but also engraftment after transplantation of human hepatocytes in uPA+/+-SCID mice. While survival was not influenced by the type of cryopreserved cells, engraftment efficacy varied substantially. Identification of donor characteristics associated with this different outcome would allow for the selection of the best donor hepatocytes and optimization of the chimeric model. Table 1 shows the significant differences in age of the hepatocyte donors (p = 0.010, Table 1). The most successful transplantations were performed with commercial hepatocytes from a 27-year-old donor (P2). This age fits to the age range of P3 cell donors, indicating that age is not an independent predictor of successful engraftment. The lower cell viability of P3 cells (p <0.032 for all comparisons) may be responsible for the lower engraftment of these cells. Percoll was used to improve cell viability after thawing for all but 17 animals receiving P2 cells. We compared cell viability and transplantation outcome of these 17 animals to exclude a Percoll-effect. The median cell viability of these thawed Percoll-free P2 cells was not significantly different from P3 cells (63.1% vs. 64.2%; p = 1.00). However, of the 17 animals transplanted, nine survived until blood sampling and eight had detectable human albumin levels. Engraftment in this transplantation set was hence significantly higher than P3 cell transplantations (88.9% vs. 25.0%, p <0.001). This suggests that cell viability as determined by Trypan Blue exclusion cannot be used as an independent predictor of engraftment. At transplantation, mice were between 7 and 29 days old. Although a trend could be observed toward higher human albu-

Journal of Hepatology 2010 vol. 53 j 468–476

JOURNAL OF HEPATOLOGY HCV RNA (IU/ml) hu alb (mg/ml)

H77C

J6/JFH-1

2 1 0 0

4

8 12 16 20 24 28 32 36

5 4 3 2 1 0

Weeks after inoculation

4

8

0 12 16 20 24 28

Weeks after inoculation

10 10 7 10 6 10 5 10 4 10 3 10 2 10 1 10 0

5 4 3 2 1

hu alb (mg/ml)

3

10 10 7 10 6 10 5 10 4 10 3 10 2 10 1 10 0

hu alb (mg/ml)

4

ED43 8

HCV RNA (IU/ml)

5

10 10 7 10 6 10 5 10 4 10 3 10 2 10 1 10 0

HCV RNA (IU/ml)

8

hu alb (mg/ml)

HCV RNA (IU/ml)

8

0 0

2

4

6

8

10

12

Weeks after inoculation

Fig. 4. Long Term HCV infection in chimeric mice. Chimeric animals were infected with HCV viruses of (A) genotype 1a H77C, n = 4, (B) genotype 2 chimeric full-length J6/ JFH-1, n = 5, and (C) genotype 4a ED43, n = 3. At indicated timepoints plasma was obtained for quantification of human albumin levels and HCV RNA as described. Animals were monitored until spontaneous death or mercy killing. Last samples were obtained at (A) week 32 (n = 1); (B) week 25 (n = 1), and (C) week 10 (n = 3). In all long term HCV-infected animals a steady drop in viremia was observed, coinciding with gradually falling human albumin levels over months (r = 0.559; p = 0.001). The HCV RNA detection limit is 750 IU/ml, indicated with a striped line on graph C.

min plasma levels following transplantation of the youngest animals, no significant correlation with acceptor age was found (data not shown). The acceptor age did not significantly differ between donor cell types (Table 1, p = 0.276) which excludes any accidental bias.

diseased parenchyma and red foci hepatocytes at all examined time points (Table 2). Up to 6 days after transplantation and irrespective of the donor species, scattered singular donor hepatocytes were observed in the liver parenchyma (Supplementary Fig. 3). No clusters of donor hepatocytes were found. Likewise, biochemical

Kinetics of the repopulation of uPA+/+-SCID liver by human versus mouse hepatocytes Survival analysis identified the first 7 days after transplantation as the most critical period (Fig. 2). Furthermore, the survival benefit of cryopreserved cells was already apparent within the first day (Fig. 3). Previous studies in other models have demonstrated that migration, entry and integration of transplanted rodent hepatocytes into the liver parenchyma occurs during this crucial phase [18,19]. We examined whether the kinetics of early human hepatocyte engraftment was similar to that of rodent hepatocytes by comparing the histology of repopulation with fresh human hepatocytes to that of murine B6 hepatocytes in uPA+/+-SCID mice. Using criteria described in on-line supplementary data, donor mouse and human hepatocytes could be discerned from Table 2. Liver repopulation and necrosis extent after murine versus human hepatocyte transplantation. Days post Tx

1 3 6 9 15 23 35

Repopulation%

Necrosis extent

Murine

Human

Murine

Human

NA 0 0; 0; 0; 0 3.6; 7.4 11.4; 25.1 NA 99.0; 100

0; 0 0; 0 NA 0.4; 0.3 7.1; 1.1 15.9; 18.9 25.3

NA 2 0; 0; 1; 1 0; 0 0; 0 NA 0; 0

2; 2 0; 2 NA 0; 2 0; 0 0; 0 NA*

The percentage of liver parenchyma occupied by donor hepatocytes and the grade of liver necrosis were analysed histologically. Both parameters are indicated with one single score per animal and timepoint. The extent of necrosis is semi-quantitatively scored with 0 = absent, 1 = focal, 2 = diffuse. Generally, two animals per timepoint were sacrificed for histological analysis. At day 3 and 6, one and 4 mice transplanted with murine hepatocytes were analysed, respectively. NA, not available; NA*, all liver tissue used for extensive immunohistochemistry and repopulation analysis; murine, animals transplanted with mouse hepatocytes; human, animals transplanted with human hepatocytes.

Fig. 5. Repopulation kinetics of human and murine hepatocytes after transplantation in uPA+/+-SCID mice. (A) Liver repopulation was examined histologically by calculating the fraction of murine liver parenchyma occupied by donor hepatocytes at each indicated time point. Human albumin levels were quantified as described and expressed as lg/ml on a log scale (left axis, detection limit 2,4 lg/ml). Regression analysis shows a linear expansion of engrafted cells from day 6 after Tx onwards. (r2 > 0.9; p <0.001). The regression line formula for both donor species is indicated in boxes. (B) Weight (Daily weight gain) of B6 hepatocytes transplanted animals (n = 6) follows histological engraftment phases. Daily weight gain used as the absolute weight of litter of similar age can vary substantially. Therefore, the % of weight gain on a daily basis compared to the weight at study start was calculated and compared to animals injected with PBS only (n = 3) or non-transgenic Balb/c control mice of similar age (n = 2).

Journal of Hepatology 2010 vol. 53 j 468–476

473

Research Article

Fig. 6. Three days after transplantation of human (left) and mouse hepatocytes (right), sharply demarcated areas of necrosis (N) of recipient parenchyma are centered around portal vein branches. Donor hepatocytes are plugged in these portal vein branches and undergo coagulative necrosis and calcification: only the a-nuclear shadows or ‘ghost’ of the donor hepatocytes remain (arrows). Original magnifications 200 (upper part) and 400 (lower part).

evidence of engraftment remained limited since human albumin levels did not exceed 100 lg/ml (Fig. 5A). Therefore, cells from both species showed qualitatively similar engraftment patterns within the first week and displayed only marginal biochemical or histological evidence of repopulation. Portal vein branches and dilated periportal sinusoids plugged by necrotic and sometimes calcified donor hepatocytes were frequently seen. This phenomenon was associated with zonal ischemic necrosis of the surrounding parenchyma (Fig. 6). The extent of necrosis was scored as absent, focal or extensive (Table 2). Extensive necrosis was more frequent in mice transplanted with human hepatocytes, but statistical evaluation was not possible due to the low number of cases. Necrosis was never seen later than 9 days after transplantation (p = 0.007). There was no evidence of scar tissue, indicating that these areas were quickly replaced by parenchyma. From 9 days after transplantation onwards, small clusters of donor hepatocytes were observed (Supplementary Fig. 3), which progressively occupied the host’s liver parenchyma. The repopulation process of B6 murine hepatocytes proceeded approximately 4-times faster compared to human cells (p <0.0001, Fig. 5A) and resulted in a complete reconstitution of the host liver in about one month. After one month only, 25% 474

of liver parenchyma was occupied by human hepatocytes (Table 2, Fig. 5A). As demonstrated previously, expanding clusters were functionally intact and produced increasing amounts of human albumin (Fig. 5A) [4]. Likewise, transplanted B6 murine hepatocytes restored liver function as demonstrated by the increasing body weight of the recipient [8]. Body weight evolution also followed early engraftment phases (Fig. 5B): Early after transplantation of B6 hepatocytes, animals experienced a negative growth possibly related to worsening of the liver function by disturbed portal flow. This initial episode was followed by a lag phase, in which body weight rose to baseline levels. Finally, a rapid increase in body weight, similar to that of wild-type non-transgenic control animals, was observed during the expansion of hepatocyte clusters. At that time, animals injected with saline were losing weight, and the deterioration of their general status necessitated their sacrifice.

Discussion The human liver-uPA+/+-SCID mouse is currently the best available small animal model to study infections with hepatotropic

Journal of Hepatology 2010 vol. 53 j 468–476

JOURNAL OF HEPATOLOGY viruses. We examined the factors affecting animal survival, engraftment efficacy, kinetics of repopulation and susceptibility to infection with HBV/HCV after transplantation of human hepatocytes in this model. Engrafted human hepatocytes in the uPAtransgenic liver have a proliferative advantage over the sick endogenous murine hepatocytes which leads to progressive humanization of the rodent liver [4,6,7]. Donor cell type determines not only animal survival but also engraftment efficacy and susceptibility of the chimeric liver to infection with HBV/HCV. From the first day onwards, animals receiving fresh hepatocytes display a substantially lower survival rate than animals given thawed cryopreserved cells. Since our histological techniques do not allow single cell tracking, we were unable to demonstrate any quantitative differences during this early phase. However, transplantation of large numbers of hepatocytes has resulted in early mortality in several models [18,20]. Hepatocyte translocation to the lungs causing hypoxemia and ischemic liver injury have both been put forward as underlying causes of death [20–22]. The histological liver necrosis we documented supports the view that a fatal acute liver failure may have complicated the pre-existing chronic liver disease in alb-uPA transgenic animals. Engraftment success was comparable for fresh and both types of commercial cryopreserved cell donors (P1 and P2), but significantly higher normalized human albumin levels were obtained with P2 donor cells, even after controlling for selection bias. Similarly, P1 and P2 cell transplantations led to the highest fraction of animals that could be used for viral hepatitis infection studies. Although up to 30% of all transplanted animals could be infected, we reproducibly achieved a 100% successful HCV infection by applying a set of graft (human albumin >1 mg/ml) and inoculum (>104 IU HCV RNA, no more than one freeze–thaw cycle) criteria [9–11]. Interestingly, graft requirements for HBV infectivity are less stringent, leading to a 100% infection rate even in animals with human albumin levels of less than 1 mg/ml. Furthermore, in a set of 3 HBV and HCV co-infected animals no apparent viral interference was noted. Both observations are important and independent confirmations of recently published data [23,24]. In contrast to the reproducible rapid viral kinetics after HCV inoculation, slower and divergent HBV viral kinetics were observed after HBV infection. A similar lag phase before HBV reaches peak titers has been described after experimental HBV inoculations in chimpanzees, which is thought to reflect the slower spread of HBV throughout the liver [25]. The superior survival rate, the higher engraftment efficacies and infection rates and the logistic constraints posed by isolating fresh hepatocytes, are all in favour of the use of commercially available cryopreserved cells for the humanization of the uPAtransgenic mouse liver. Neither donor age, nor donor cell viability turn out to be independent predictors of engraftment potential in this transplantation series. Apart from age, the underlying medical condition of the donor may have an impact on the proliferative capacity of donor cells. Furthermore, the freezing protocol itself influences the migratory and proliferative capacities of cryopreserved cells [13]. Full disclosure of both the medical chart of the donor and the cryopreservation protocol will be needed to better appreciate the parameters that determine repopulating potential. In summary, it is shown here that commercially available cryopreserved human hepatocytes are perfectly suited for the efficient humanization of the uPA-SCID liver and superior to

freshly isolated human hepatocytes. Hepatocytes migrate and integrate in the liver plates during the first week after transplantation. This process determines both the survival of the recipient mice and the successful repopulation of their diseased liver. Up to 30% of transplanted animals can be used for further studies of HBV and/or HCV infections. However 100% successful HCV infection can be achieved with good quality inocula of different HCV genotypes in animals with plasma levels of human albumin of at least 1 mg/ml. The requirements to successfully infect a chimeric mouse with HBV are clearly less stringent.

Financial disclosure T.V. (Ph.D.-grant) and P.M. (postdoctoral fellowship) are both supported by the Research Foundation – Flanders (FWO-Vlaanderen). This study was financially supported by the Belgian State (IUAP P6/36–HEPRO) and the Ghent University via a Concerted Action Grant (01G00507).

Conflict of interest The authors who have taken part in this study declared that they do not have anything to disclose regarding or conflict of interest with respect to this manuscript.

Acknowledgements The authors thank Manuela Cousin, Petra Premereur, and Lieven Verhoye for excellent technical assistance; Prof. Dr. Troisi and Prof. Dr. de Hemptinne for the supply of fresh human liver fragments; Dr. R. Purcell and Dr. J. Bukh, as well as Dr. C. Rice for their kind provision of infectious HCV virus of different genotypes.

Supplementary data Supplementary data associated with this article can be found, in the on-line version, at doi:10.1016/j.jhep.2010.03.024.

References [1] Nussler A et al. Present status and perspectives of cell-based therapies for liver diseases. J Hepatol 2006;45 (1):144–159. [2] Rhim JA et al. Replacement of diseased mouse liver by hepatic cell transplantation. Science 1994;263 (5150):1149–1152. [3] Overturf K et al. Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinaemia type I. Nat Genet 1996;12 (3):266–273. [4] Meuleman P et al. Morphological and biochemical characterization of a human liver in a uPA-SCID mouse chimera. Hepatology 2005;41 (4):847–856. [5] Lootens L et al. UPA+/+-SCID mouse with humanized liver as a model for in vivo metabolism of exogenous steroids: methandienone as a case study. Clin Chem 2009;55 (10):1783–1793. [6] Mercer DF et al. Hepatitis C virus replication in mice with chimeric human livers. Nat Med 2001;7 (8):927–933. [7] Dandri M et al. Repopulation of mouse liver with human hepatocytes and in vivo infection with hepatitis B virus. Hepatology 2001;33 (4):981–988.

Journal of Hepatology 2010 vol. 53 j 468–476

475

Research Article [8] Vanwolleghem T et al. Ultra-rapid cardiotoxicity of the hepatitis C virus protease inhibitor BILN 2061 in the urokinase-type plasminogen activator mouse. Gastroenterology 2007;133 (4):1144–1155. [9] Vanwolleghem T et al. Polyclonal immunoglobulins from a chronic hepatitis C virus patient protect human liver-chimeric mice from infection with a homologous hepatitis C virus strain. Hepatology 2008;47 (6):1846–1855. [10] Meuleman P et al. Anti-CD81 antibodies can prevent a hepatitis C virus infection in vivo. Hepatology 2008;48 (6):1761–1768. [11] Lindenbach BD et al. Cell culture-grown hepatitis C virus is infectious in vivo and can be recultured in vitro. Proc Natl Acad Sci USA 2006;103 (10):3805–3809. [12] Stephenne X et al. Cryopreservation of human hepatocytes alters the mitochondrial respiratory chain complex 1. Cell Transplant 2007;16 (4):409–419. [13] Hengstler JG et al. Cryopreserved primary hepatocytes as a constantly available in vitro model for the evaluation of human and animal drug metabolism and enzyme induction. Drug Metab Rev 2000;32 (1):81–118. [14] Klaunig JE et al. Mouse liver cell culture. I. Hepatocyte isolation. In Vitro 1981;17 (10):913–925. [15] Tateno C et al. Near completely humanized liver in mice shows human-type metabolic responses to drugs. Am J Pathol 2004;165 (3):901–912. [16] Feinstone SM et al. Non-A, non-B hepatitis in chimpanzees and marmosets. J Infect Dis 1981;144 (6):588–598.

476

[17] Meuleman P et al. Immune suppression uncovers endogenous cytopathic effects of the hepatitis B virus. J Virol 2006;80 (6):2797–2807. [18] Rajvanshi P et al. Studies of liver repopulation using the dipeptidyl peptidase IV-deficient rat and other rodent recipients: cell size and structure relationships regulate capacity for increased transplanted hepatocyte mass in the liver lobule. Hepatology 1996;23 (3):482–496. [19] Gupta S et al. Entry and integration of transplanted hepatocytes in rat liver plates occur by disruption of hepatic sinusoidal endothelium. Hepatology 1999;29 (2):509–519. [20] Kocken JM et al. Acute death after intraportal hepatocyte transplantation in an allogeneic rat strain combination: a possible role for complement activation. Transplant Proc 1997;29 (4):2067–2068. [21] Muraca M et al. Intraportal hepatocyte transplantation in the pig: hemodynamic and histopathological study. Transplantation 2002;73 (6):890–896. [22] Gupta S et al. Cell transplantation causes loss of gap junctions and activates GGT expression permanently in host liver. Am J Physiol Gastrointest Liver Physiol 2000;279 (4):G815–G826. [23] Tsuge M et al. Infection of human hepatocyte chimeric mouse with genetically engineered hepatitis B virus. Hepatology 2005;42 (5):1046–1054. [24] Hiraga N et al. Absence of viral interference and different susceptibility to interferon between hepatitis B virus and hepatitis C virus in human hepatocyte chimeric mice. J Hepatol 2009;51 (6):1046–1054. [25] Wieland SF, Chisari FV. Stealth and cunning: hepatitis B and hepatitis C viruses. J Virol 2005;79 (15):9369–9380.

Journal of Hepatology 2010 vol. 53 j 468–476