Hippo signaling drives mesenchymal cell fate and heterogeneity in liver cancer

Liver cancer comprises a spectrum of cancers that include pure hepatocellular carcinoma, cholangiocarcinoma, and tumors with varying degrees of mixed hepatocytic, cholangiocytic, and mesenchymal traits. Such heterogeneity can elevate cancer aggressiveness and induce therapy resistance. However, mechanisms that promote cellular plasticity in cancer remain poorly understood. To study mechanisms driving heterogeneity in liver cancer, we developed a mouse model in which cholangiocarcinoma cells spontaneously acquire mesenchymal characteristics.

We induced liver cancer in mice by hydrodynamic tail vein injection of plasmids expressing activated versions of Notch, Akt, and KRas. We performed lineage tracing, single nucleus RNA-sequencing, and gain- and loss-of-function experiments of Hippo signaling to determine the mechanisms inducing tumor heterogeneity.

Co-expression of activated Notch, Akt, and KRas led to the formation of aggressive intrahepatic cholangiocarcinoma (ICC) that presented a strong desmoplastic reaction and a heterogeneous population of cancer cells with varying levels of epithelial and mesenchymal features. Single-cell analysis showed that these tumors comprised cancer cells with cholangiocyte-like epithelial (N = 4006 cells) and mesenchymal (N = 605 cells) characters. Lineage tracing revealed that the mesenchymal cancer cells originated from the transdifferentiation of epithelial cancer cells. Mechanistically, we identified the Hippo pathway effectors Yap and Taz as drivers of this transition: high Yap/Taz levels correlated with the acquisition of mesenchymal features, Yap/Taz double mutant tumor cells were unable to transdifferentiate into mesenchymal tumor cells (N = 5, p = 0.007), and overexpression of Yap (N = 5, p = 0.008) or Taz (N = 5, p = 0.025) in pure ICC tumors driven by Notch and Akt led to the emergence of cancer cells with mesenchymal features.

Our data show that the Hippo pathway effectors Yap and Taz drive plasticity in liver cancer cells and promote the transdifferentiation of epithelial cholangiocarcinoma cells into tumor cells with mesenchymal characteristics.

Our study provides a mechanistic understanding of how cellular heterogeneity, a major challenge for treatment, develops in intrahepatic cholangiocarcinoma. We show that the Hippo pathway effectors Yap and Taz drive the transition of epithelial liver cancer cells into a mesenchymal state. These findings are important because they identify molecular mechanisms that promote cholangiocarcinoma plasticity and imply that therapeutic targeting of YAP/TAZ activity may inhibit cancer cell plasticity and enhance treatment efficacy in patients with cholangiocarcinoma and other cancers with similar molecular features. However, given the complexity of tumor signaling and the preclinical nature of our mouse model, additional validation is needed to define its clinical relevance.