A Novel Mouse Model of Combined Hepatocellular-Cholangiocarcinoma Induced by Diethylnitrosamine and Loss of Ppp2r5d

SIMPLE SUMMARY: Primary liver cancer (PLC) is among the leading causes of cancer-related deaths worldwide. PLC can be classified in hepatocellular carcinoma (HCC), cholangiocarcinoma (CCA), and the less common combined HCC-CCA (cHCC-CCA) based on histological features. The underlying mechanisms for PLC development/progression are still unknown, hampering the development of targeted PLC therapeutics. Protein Phosphatase 2A-B56δ (PP2A-B56δ) has been shown to have a tumor-suppressive role in mouse liver. We found that depletion of Ppp2r5d in mice accelerated HCC development induced by diethylnitrosamine (DEN), a well-established liver carcinogen, but unexpectedly also resulted in cHCC-CCA development. We also observed that Ppp2r5d is upregulated in tumors from wildtype and heterozygous mice, and that loss of Ppp2r5d alters specific oncogenes and signaling pathways in pre-tumor and tumor tissues. Our study highlights that mouse PP2A-B56δ has a tumor-suppressive role not only in HCC, but also in cHCC-CCA, which may have further implications for human PLC development and targeted treatment. ABSTRACT: Primary liver cancer (PLC) can be classified in hepatocellular (HCC), cholangiocarcinoma (CCA), and combined hepatocellular-cholangiocarcinoma (cHCC-CCA). The molecular mechanisms involved in PLC development and phenotype decision are still not well understood. Complete deletion of Ppp2r5d, encoding the B56δ subunit of Protein Phosphatase 2A (PP2A), results in spontaneous HCC development in mice via a c-MYC-dependent mechanism. In the present study, we aimed to examine the role of Ppp2r5d in an independent mouse model of diethylnitrosamine (DEN)-induced hepatocarcinogenesis. Ppp2r5d deletion (heterozygous and homozygous) accelerated HCC development, corroborating its tumor-suppressive function in liver and suggesting Ppp2r5d may be haploinsufficient. Ppp2r5d-deficient HCCs stained positively for c-MYC, consistent with increased AKT activation in pre-malignant and tumor tissues of Ppp2r5d-deficient mice. We also found increased YAP activation in Ppp2r5d-deficient tumors. Remarkably, in older mice, Ppp2r5d deletion resulted in cHCC-CCA development in this model, with the CCA component showing increased expression of progenitor markers (SOX9 and EpCAM). Finally, we observed an upregulation of Ppp2r5d in tumors from wildtype and heterozygous mice, revealing a tumor-specific control mechanism of Ppp2r5d expression, and suggestive of the involvement of Ppp2r5d in a negative feedback regulation restricting tumor growth. Our study highlights the tumor-suppressive role of mouse PP2A-B56δ in both HCC and cHCC-CCA, which may have important implications for human PLC development and targeted treatment.