Establishment of a variant-to-function framework for cholestatic diseases using human liver in vitro models

Hepatic cholestatic diseases are characterised by impaired bile flow and harmful accumulation of bile acids in the liver. Genetics is a key contributing factor to cholestatic disease risk, with rare mutations in transporter genes being well-established causes of cholestasis. Recent studies have highlighted that common genetic variants within liver transcriptional enhancers also contribute to cholestasis. However, unlike coding sequences, enhancers are poorly conserved in other species, making the usefulness of animal models for the study of cholestasis genetic factors limited at best.

A major hurdle to replace animal models for the study of cholestasis risk variants with human model systems is the lack of appropriately characterised human liver in vitro models. Human induced pluripotent stem cell (iPSC)-derived hepatocytes (iHeps) are an attractive model to study the effects of human genetic variants on hepatic transcriptional programs and ultimately disease-relevant cellular outputs. In particular, iHeps cultured under polarisation-inducing conditions show enhanced functionality. We envision that a better characterisation of the molecular features of iPSC-derived liver in vitro systems will facilitate their use to investigate liver disease genetics and enable the replacement of animal models in this and future studies.

In this project, a multidisciplinary team joining expertise in human cholestasis, regulatory genomics, stem cell-based modelling and lipidomics, will investigate two culture systems (classic 2D vs. new sandwich model optimised for cholestasis modelling) generating global expression and chromatin activity maps. These systems will be benchmarked against human liver datasets to identify which system (1) better recapitulates relevant features of human hepatocytes, and (2) is better suited to functionally investigate genetic loci associated with human cholestasis. These data will also be systematically compared against data obtained from rodents, highlighting loci for which rodents are not suitable models. We will then carry out a CRISPR-activation screen of cholestasis loci to identify the target genes of noncoding variants, as enhancers can interact with genes at varying distances. Finally, we will focus on one locus, which will be modelled by combining genome editing in iPSC with the cholestasis-optimised iHeps protocol. Edited iHeps will then be used to investigate cholestasis-relevant cellular outputs, including formation of canaliculi and lipidomic characterisation. We expect to demonstrate the superior suitability of iHeps to study noncoding variants associated with cholestasis, providing a framework for variant-to-function investigation of risk loci in iHeps. More broadly, these datasets and analysis framework will be applicable to investigate risk loci for other liver diseases in future studies.