Yap-Sox9 signaling determines hepatocyte plasticity and lineage-specific hepatocarcinogenesis

•Sox9 is required for Yap-induced hepatocyte plasticity during hepatocarcinogenesis in mice. •Sox9 removal in hepatocytes with Yap activation resulted in more aggressive HCC. •Sox9 expression is sufficient to promote cell fate plasticity in less differentiated hepatocytes in vivo. •YAP activity and SOX9 expression are strongly correlated with liver tumor plasticity in human patients. Background & Aims Primary liver tumors comprise distinct subtypes. A subset of intrahepatic cholangiocarcinoma (iCCA) can arise from cell fate reprogramming of mature hepatocytes in mouse models. However, the underpinning of cell fate plasticity during hepatocarcinogenesis is still poorly understood, hampering therapeutic development for primary liver cancer. As YAP activation induces liver tumor formation and cell fate plasticity, we investigated the role of Sox9, a transcription factor downstream of Yap activation that is expressed in biliary epithelial cells (BECs), in Yap-induced cell fate plasticity during hepatocarcinogenesis. Methods To evaluate the function of Sox9 in YAP-induced hepatocarcinogenesis in vivo, we used several genetic mouse models of inducible hepatocyte-specific YAP activation with simultaneous Sox9 removal. Cell fate reprogramming was determined by lineage tracing and immunohistochemistry. The molecular mechanism underlying Yap and Sox9 function in hepatocyte plasticity was investigated by transcription and transcriptomic analyses of mouse and human liver tumors. Results Sox9, a marker of liver progenitor cells (LPCs) and BECs, is differentially required in YAP-induced stepwise hepatocyte programming. While Sox9 has a limited role in hepatocyte dedifferentiation to LPCs, it is required for BEC differentiation from LPCs. YAP activation in Sox9-deficient hepatocytes resulted in more aggressive HCC with enhanced Yap activity at the expense of iCCA-like tumors. Furthermore, we showed that 20% of primary human liver tumors were associated with a YAP activation signature, and tumor plasticity is highly correlated with YAP activation and SOX9 expression. Conclusion Our data demonstrated that Yap-Sox9 signaling determines hepatocyte plasticity and tumor heterogeneity in hepatocarcinogenesis in both mouse and human liver tumors. We identified Sox9 as a critical transcription factor required for Yap-induced hepatocyte cell fate reprogramming during hepatocarcinogenesis. Lay summary Sox9, a marker of liver progenitor cells and bile duct lining cells, is a downstream target of YAP protein activation. Herein, we found that YAP activation in hepatocytes leads to a transition from mature hepatocytes to liver progenitor cells and then to bile duct lining cells. Sox9 is required in the second step during mouse hepatocarcinogenesis. We also found that human YAP and SOX9 may play similar roles in liver cancers. Primary liver tumors comprise distinct subtypes. A subset of intrahepatic cholangiocarcinoma (iCCA) can arise from cell fate reprogramming of mature hepatocytes in mouse models. However, the underpinning of cell fate plasticity during hepatocarcinogenesis is still poorly understood, hampering therapeutic development for primary liver cancer. As YAP activation induces liver tumor formation and cell fate plasticity, we investigated the role of Sox9, a transcription factor downstream of Yap activation that is expressed in biliary epithelial cells (BECs), in Yap-induced cell fate plasticity during hepatocarcinogenesis. To evaluate the function of Sox9 in YAP-induced hepatocarcinogenesis in vivo, we used several genetic mouse models of inducible hepatocyte-specific YAP activation with simultaneous Sox9 removal. Cell fate reprogramming was determined by lineage tracing and immunohistochemistry. The molecular mechanism underlying Yap and Sox9 function in hepatocyte plasticity was investigated by transcription and transcriptomic analyses of mouse and human liver tumors. Sox9, a marker of liver progenitor cells (LPCs) and BECs, is differentially required in YAP-induced stepwise hepatocyte programming. While Sox9 has a limited role in hepatocyte dedifferentiation to LPCs, it is required for BEC differentiation from LPCs. YAP activation in Sox9-deficient hepatocytes resulted in more aggressive HCC with enhanced Yap activity at the expense of iCCA-like tumors. Furthermore, we showed that 20% of primary human liver tumors were associated with a YAP activation signature, and tumor plasticity is highly correlated with YAP activation and SOX9 expression. Our data demonstrated that Yap-Sox9 signaling determines hepatocyte plasticity and tumor heterogeneity in hepatocarcinogenesis in both mouse and human liver tumors. We identified Sox9 as a critical transcription factor required for Yap-induced hepatocyte cell fate reprogramming during hepatocarcinogenesis.