Modeling Drug Induced Liver Injury with Patient-Derived Liver Organoids and Microfluidic Chips

University Of Michigan At Ann Arbor, Ann Arbor MI

Investigators

Abstract

Project Summary/Abstract Drug-induced liver injury (DILI) is a significant concern due to its potentially life-threatening consequences and its contribution to clinical trial failures and market withdrawals. Current models cannot accurately predict DILI risk, necessitating the development of more physiologically relevant systems. We have previously demonstrated the utility of patient-derived induced pluripotent stem cells (iPSCs) to produce human liver organoids (HLOs) in a high-throughput, high-content microscopy-based screening for DILI. However, optimizations are needed to enhance assay robustness and specificity mainly by ensuring that the HLOs assayed are of high quality. We aim to optimize the assay system by implementing multiple quality control measures throughout the differentiation process of HLOs. These measures include increasing the passage number of iPSCs, confirming pluripotency marker expression, and validating lineage-specific markers at various stages of differentiation. Additionally, we will utilize a comprehensive set of dyes to characterize cellular populations within HLOs and gain insights into the mechanisms of DILI. We will screen a set of control compounds known to cause DILI using our improved assay system and employ single-cell analysis techniques to confirm known phenotypes. Furthermore, we will develop machine learning models to predict DILI phenotypes based on cellular features, facilitating the identification of potential hepatotoxic compounds. This approach will enhance our understanding of DILI mechanisms and enable the prediction of DILI risk in pre- clinical drug development. Herbal dietary supplements (HDS) represent a significant but poorly understood cause of DILI, with high variability in composition and limited regulation. The complexity of HDS-induced liver injury (HILI) presents challenges for modeling and requires robust systems to elucidate underlying mechanisms. I aim to predict HILI and identify specific molecules responsible for liver injury in HDS by utilizing our established HLO assay system to screen a diverse collection of 23 HDS extracts obtained from patients who have experienced HILI and 25 isolated natural products from these extracts for hepatotoxicity risk. Mass spectral molecular networking will be employed to characterize the molecular composition of HDS extracts and identify potential hepatotoxic compounds. Subsequent screening of purified compounds will confirm their hepatotoxic effects, allowing us to characterize the mechanisms of HILI through transcriptomic analysis. This approach will enable the identification of intrinsic hepatotoxicants and provide insights into the complex mechanism of HILI. By combining high-content microscopy techniques with comprehensive molecular analysis, we aim to improve our understanding of HILI and facilitate the development of predictive models for HILI. Overall, DILI and HILI present a significant unmet medical need, and I plan to use HLOs and high-content imaging to develop a superior pre-clinical model to predict HILI risk in HDS.