Establishment of hyperoxic cell culture system for predicting drug-induced liver injury: reducing accumulated lipids in hepatocytes derived from chimeric mice with humanized liver.

Drug-induced liver injury (DILI) is a major adverse reaction. Species-specific differences between humans and laboratory animals make it difficult to establish evaluation models that can accurately predict DILI in the preclinical phase. Chimeric mice with humanized liver are potential predictive models for understanding DILI. Chimeric mice generated by transplanting human hepatocytes into urokinase-type plasminogen activator/severe combined immunodeficient mice are known to develop fatty liver and show lipid accumulation in isolated hepatocytes. It is speculated that the lipids accumulated in hepatocytes may interfere with DILI assessment. It is known that normal 20% oxygen culture conditions do not meet oxygen demand because oxygen consumption rate is higher than the oxygen supply rate. Therefore, we predicted that hyperoxic cultures could induce hepatocyte function and reduce accumulated lipids. A culture of chimeric mouse hepatocytes in 40% oxygen showed reduced intracellular lipid and triglyceride levels compared to those cultured in 20% oxygen on days 7 and 10. In addition, fatty acid ß-oxidation (FAO) activity increased from day 7 under 40% oxygen conditions. On the other hand, FAO activity increased on day 10 under 20% conditions. Microarray and Ingenuity Pathway Analysis showed that lipid metabolism-related pathways were downregulated under 40% oxygen conditions for 7 days, suggesting the involvement of several mechanisms in decreasing lipid levels and increasing FAO. Furthermore, some pathways related to cellular function and maintenance were upregulated under 40% oxygen conditions for 7 days. In conclusion, chimeric mouse hepatocytes cultured under hyperoxic conditions may be useful for predicting DILI.

Culture density contributes to hepatic functions of fresh human hepatocytes isolated from chimeric mice with humanized livers: Novel, long-term, functional two-dimensional in vitro tool for developing new drugs.

Chimeric mice with humanized livers are considered a useful animal model for predicting human (h-) drug metabolism and toxicity. In this study, the characteristics of fresh h-hepatocytes (cFHHs, PXB-cells®) isolated from chimeric mice (PXB-mice®) were evaluated in vitro to confirm their utility for drug development. cFHHs cultured at high density (2.13 × 105 cells/cm2) displayed stable production of h-albumin and cytochrome P450 (CYP) 3A activities for at least 21 days. The mRNA expression levels of 10 of 13 CYP, UDP-glucuronosyltransferase (UGT), and transporters were maintained at >10% of the levels of freshly isolated cFHHs after 21 days. From 1 week, many bile canaliculi were observed between cFHHs, and the accumulation of the multidrug resistance-associated protein and bile salt export pump substrates in these bile canaliculi was clearly inhibited by cyclosporin A. Microarray analysis of cFHHs cultured at high density and at low density (0.53 × 105 cells/cm2) revealed that high density culture maintained high expressions of some transcription factors (HNF4α, PXR, and FXR) perhaps involved in the high CYP, UGT and transporter gene expressions of cFHHs. These results strongly suggest that cFHHs could be a novel in vitro tool for drug development studies.

Detection of acute toxicity of aflatoxin B1 to human hepatocytes in vitro and in vivo using chimeric mice with humanized livers.

Aflatoxin B1 (AFB1), a mycotoxin, is acutely hepatotoxic to many animals including humans. However, there are marked interspecies differences in sensitivity to AFB1-induced toxicity depending on bioactivation by cytochrome P450s (CYPs). In the present study, we examined the applicability of chimeric mice with humanized livers and derived fresh human hepatocytes for in vivo and vitro studies on AFB1 cytotoxicity to human hepatocytes. Chimeric mice with highly humanized livers and SCID mice received daily injections of vehicle (corn oil), AFB1 (3 mg/kg), and carbon tetrachloride (50 mg/kg) for 2 days. Histological analysis revealed that AFB1 promoted hepatocyte vacuolation and inflammatory cell infiltration in the area containing human hepatocytes. A novel human alanine aminotransferase 1 specific enzyme-linked immunosorbent assay demonstrated the acute toxicity of AFB1 to human hepatocytes in the chimeric mouse livers. The sensitivity of cultured fresh human hepatocytes isolated from the humanized liver mice for AFB1 cytotoxicity was comparable to that of primary human hepatocytes. Long-term exposure to AFB1 (6 or 14 days) produced a more severe cytotoxicity. The half-maximal lethal concentration was 10 times lower in the 2-week treatment than after 2 days of exposure. Lastly, the significant reduction of AFB1 cytotoxicity by a pan-CYP inhibitor or transfection with CYP3A4 specific siRNA clearly suggested that bioactivation of AFB1 catalyzed by CYPs was essential for AFB1 cytotoxicity to the human hepatocytes in our mouse model. Collectively, our results implicate the humanized liver mice and derived fresh human hepatocytes are useful models for studies of AFB1 cytotoxicity to human hepatocytes.

Characterization and applications of chimeric mice with humanized livers for preclinical drug development.

We have succeeded in stable mass production of chimeric PXB-mice, whose liver is repopulated by human hepatocytes at a ratio of more than 70%, and we are providing these mice to academia and pharmaceutical companies to support the development of new drugs or studies of liver function. Furthermore, we isolated human hepatocytes, called PXB-cells, from the chimeric mice, and provide them for clients weekly for in vitro studies. In this review, we summarize the existing characterizations of PXB-mice and PXB-cells and their present and future applications.

Generation of Novel Chimeric Mice with Humanized Livers by Using Hemizygous cDNA-uPA/SCID Mice.

We have used homozygous albumin enhancer/promoter-driven urokinase-type plasminogen activator/severe combined immunodeficient (uPA/SCID) mice as hosts for chimeric mice with humanized livers. However, uPA/SCID mice show four disadvantages: the human hepatocytes (h-heps) replacement index in mouse liver is decreased due to deletion of uPA transgene by homologous recombination, kidney disorders are likely to develop, body size is small, and hemizygotes cannot be used as hosts as more frequent homologous recombination than homozygotes. To solve these disadvantages, we have established a novel host strain that has a transgene containing albumin promoter/enhancer and urokinase-type plasminogen activator cDNA and has a SCID background (cDNA-uPA/SCID). We applied the embryonic stem cell technique to simultaneously generate a number of transgenic lines, and found the line with the most appropriate levels of uPA expression-not detrimental but with a sufficiently damaged liver. We transplanted h-heps into homozygous and hemizygous cDNA-uPA/SCID mice via the spleen, and monitored their human albumin (h-alb) levels and body weight. Blood h-alb levels and body weight gradually increased in the hemizygous cDNA-uPA/SCID mice and were maintained until they were approximately 30 weeks old. By contrast, blood h-alb levels and body weight in uPA/SCID chimeric mice decreased from 16 weeks of age onwards. A similar decrease in body weight was observed in the homozygous cDNA-uPA/SCID genotype, but h-alb levels were maintained until they were approximately 30 weeks old. Microarray analyses revealed identical h-heps gene expression profiles in homozygous and hemizygous cDNA-uPA/SCID mice were identical to that observed in the uPA/SCID mice. Furthermore, like uPA/SCID chimeric mice, homozygous and hemizygous cDNA-uPA/SCID chimeric mice were successfully infected with hepatitis B virus and C virus. These results indicate that hemizygous cDNA-uPA/SCID mice may be novel and useful hosts for producing chimeric mice for use in future long-term studies, including hepatitis virus infection analysis or drug toxicity studies.

Polymorphic CAG repeat numbers in the androgen receptor gene of female pattern hair loss patients.

Female pattern hair loss (FPHL) is frequently referred to as female androgenetic alopecia (FAGA). However, the role of androgen in this type of hair loss remains uncertain. We previously reported greater therapeutic efficacy of finasteride in Japanese male androgenetic alopecia (MAGA) patients in cases where the CAG repeats of the androgen receptor (AR) gene were short. To examine the correlation between CAG repeat numbers and the therapeutic efficacy of finasteride in FPHL patients, the efficacy of finasteride (1 mg/day) was evaluated macroscopically. Because women have two X-chromosomes, the shorter and longer CAG repeat numbers were analyzed in 37 Japanese FPHL patients, then the correlation of these factors was statistically analyzed by anova. No statistical significance in terms of the differences in CAG repeat numbers was detected among the four groups classified on the basis of the efficacy of finasteride. From these results, it may be concluded that the efficacy of this medicine in each FPHL patient cannot be predicted by the CAG repeat numbers in the AR gene.