Prediction of Drug-Induced Liver Injury in Micropatterned Co-cultures Containing iPSC-Derived Human Hepatocytes

• Published in: Toxicological sciences Vol. 145; no. 2; pp. 252 - 262
• Main Authors: Ware, Brenton R, Berger, Dustin R, Khetani, Salman R
• Format: Journal Article
• Language: English
• Published: United States 01.06.2015
• Subjects: 3T3 Cells; Albumins - metabolism; Animals; Biomarkers - metabolism; Chemical and Drug Induced Liver Injury - etiology; Chemical and Drug Induced Liver Injury - metabolism; Chemical and Drug Induced Liver Injury - pathology; Coculture Techniques; Dose-Response Relationship, Drug; Female; Fibroblasts - drug effects; Fibroblasts - metabolism; Fibroblasts - pathology; Hepatocytes - drug effects; Hepatocytes - metabolism; Hepatocytes - pathology; High-Throughput Screening Assays; Humans; Induced Pluripotent Stem Cells - drug effects; Induced Pluripotent Stem Cells - metabolism; Induced Pluripotent Stem Cells - pathology; Male; Mice; Reproducibility of Results; Risk Assessment; Toxicity Tests - methods; Urea - metabolism; Xenobiotics - classification; Xenobiotics - toxicity; drug-induced liver injury; iPSC-HHs; iCell Hepatocytes; micropatterning; co-cultures; primary human hepatocytes
• ISSN: 1096-6080; 1096-0929
• DOI: 10.1093/toxsci/kfv048

Primary human hepatocytes (PHHs) are a limited resource for drug screening, their quality for in vitro use can vary considerably across different lots, and a lack of available donor diversity restricts our understanding of how human genetics affect drug-induced liver injury (DILI). Induced pluripotent stem cell-derived human hepatocyte-like cells (iPSC-HHs) could provide a complementary tool to PHHs for high-throughput drug screening, and ultimately enable personalized medicine. Here, we hypothesized that previously developed iPSC-HH-based micropatterned co-cultures (iMPCCs) with murine embryonic fibroblasts could be amenable to long-term drug toxicity assessment. iMPCCs, created in industry-standard 96-well plates, were treated for 6 days with a set of 47 drugs, and multiple functional endpoints (albumin, urea, ATP) were evaluated in dosed cultures against vehicle-only controls to enable binary toxicity decisions. We found that iMPCCs correctly classified 24 of 37 hepatotoxic drugs (65% sensitivity), while all 10 non-toxic drugs tested were classified as such in iMPCCs (100% specificity). On the other hand, conventional confluent cultures of iPSC-HHs failed to detect several liver toxins that were picked up in iMPCCs. Results for DILI detection in iMPCCs were remarkably similar to published data in PHH-MPCCs (65% versus 70% sensitivity) that were dosed with the same drugs. Furthermore, iMPCCs detected the relative hepatotoxicity of structural drug analogs and recapitulated known mechanisms of acetaminophen toxicity in vitro. In conclusion, iMPCCs could provide a robust tool to screen for DILI potential of large compound libraries in early stages of drug development using an abundant supply of commercially available iPSC-HHs.