Systematic comparison of experimental and human obstructive cholestasis reveals conservation of canonical pathway activation and biomarkers relevant for cholestatic liver disease [human data set]

Cholelithiasis-induced cholestasis is one of the most common causes of hospitalization due to gastrointestinal disease, yet considerable knowledge gaps exist in the pathogenesis of this disease. This can partially be explained by inadequate characterization of experimental cholestasis models. Therefore, we compared the transcriptional profile of commonly used mouse models for obstructive cholestasis and benchmarked them to human disease to identify the model(s) best suited for cholelithiasis-induced cholestasis research and to uncover conserved mechanisms involved in human and murine cholestasis. Selected mouse models included bile duct ligation (BDL) surgery, 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) diet, associated with extra- and intrahepatic obstructive cholestasis, respectively, and a drug-induced cholestasis model relying on cyclosporin A (CsA), in an acute and chronic setting. Human samples were collected from patients with cholelithiasis-induced cholestasis of an acute and recurrent nature. RNA sequencing was performed on mouse and human liver tissue. Both the BDL and DDC models, but not the CsA model, were shown to be applicable for studying cholelithiasis-induced cholestasis, with transcriptomic profiles that highly correspond to acute cholestasis in human patients. In particular, the conservation of canonical pathways related to the inflammatory response and cytoskeleton organization, in which the Rho family GTPase is involved, were identified. This study furthermore revealed some promising mechanistic-based transcriptomic biomarkers relevant for murine and human cholestasis, which could potentially be useful for robust prediction and detection of diverse types of cholestatic liver disease. Overall design: Fresh liver tissue was collected from 3 acute (HUMAC) and 4 recurrent (HUMREC) cholestasis patients. The respective control group consists of 3 human patients (HUMN) that were scheduled for a gastric bypass (without confirmed non-alcoholic fatty liver disease) or liver resection without clinical signs of cholestasis. RNA was extracted from the fresh liver tissue of cholestasis patients and paraffin embedded liver tissue from the patients without cholestasis.