Unexpected role of nuclear receptor CAR in cholangiocarcinoma development

Cholangiocarcinoma (CCA), the second most common primary liver cancer, arises from the biliary epithelium, either within the liver parenchyma or in the hepatoduodenal ligament. Several cholestatic liver diseases, such as primary sclerosing cholangitis, developmental bile duct malformation, and progressive familial intrahepatic cholestasis (PFIC), have been implicated as CCA risk factors. Cholestatic conditions exhibit a cocarcinogenic effect when combined with chronic liver injuries, as supported by a recently reported CCA animal model in which partial bile duct ligation is combined with repeated treatments of the carcinogen diethylnitrosamine. The nuclear receptor Constitutive Androstane Receptor (CAR, NR1I3), a central regulator of xenobiotic metabolism, is known to drive hepatocellular carcinoma (HCC) when chronically activated, while CAR ablation prevents it in multiple liver cancer models. We have recently characterized the impact of endogenous CAR activation in a mouse model of human PFIC5 disease ( i.e. , Fxr;Shp double knockout, FS-DKO) that exhibits cholestatic liver injury and develops spontaneous HCC at 1 year of age. In this study, we demonstrated the role of endogenous CAR activation in cholestasis-driven liver cancer development. Based on the strong correlation between CAR activation and liver cancer development in FS-DKO liver, we initially hypothesized that chronic CAR activation mediates hepatic carcinogenesis in FS-DKO liver. However, in contrast to our expectations, further deletion of CAR in Fxr;Shp;Car triple knockouts (FSC-TKO) did not reduce tumor incidence compared to FS-DKO. Instead, highly hypertrophic as well as hyperplastic cholangiocytes became much more abundant in FSC-TKO livers than in FS-DKO livers. Accordingly, about half of the visible FSC-TKO tumor nodules were CCA or combined HCC-CCA types, while most tumors in FS-DKO livers were hepatocellular types. Using RNA-seq and TRRUST transcription factor analysis, we found that CAR/PXR targets were significantly depleted in FSC-TKO, while ETS1 targets were significantly enriched, consistent with emerging evidence that ETS1 biases liver tumor fate toward CCA lineage. These findings suggest that, under cholestatic stress, loss of CAR unexpectedly promotes CCA development by unleashing a microenvironment that favors CCA. Targeting the CAR-ETS1 axis may offer new opportunities to modulate HCC-CCA lineage specification and improve therapeutic strategies.