Classification of hepatotoxicity due to cholestasis and necrosis using transcriptomics on human precision-cut liver slices

Human toxicity screening is an important stage in the development of safe drug candidates. Hepatotoxicity is one of the major reasons for withdrawal of drugs from the market due to the fact that the liver is the major organ involved in drug metabolism and it can generate toxic metabolites. There is a need to screen the molecules for drug-induced hepatotoxicity in human at an earlier stage. Transcriptomics is a technique widely used to screen molecules for toxicity and to unravel toxicity mechanisms. The aim of this study was to classify known hepatotoxicants on their phenotype of toxicity using gene expression profiles ex vivo in human precision-cut liver slices (PCLS). Hepatotoxicants which are known to induce either necrosis (n=5) or cholestasis (n=5) were used at concentrations inducing low ( <20%) and medium (30-50%) toxicity, based on ATP or LDH. Random Forest and Support Vector Machine algorithms were used to classify hepatotoxicants using leave one-compound-out cross-validation method. Optimized biomarkers sets for each of the cross-validation steps were used to derive a consensus list of markers. In this approach, the classification correctly predicts the phenotype of toxicity with an accuracy of 70-80%. The classification is slightly better for the low than for the medium toxicity. The consensus list of markers includes bile acid transporter (SLC10A7), cholesterol and lipid metabolism and transport (HACL1 and SORL1), heat shock and other stress response genes. This study shows that human PCLS are a useful model to predict the phenotype of drug-induced hepatotoxicity. Additional compounds should be included to confirm the consensus list of markers, which could then be used to develop a biomarker PCR-array for hepatotoxicity screening.