Activating β‐catenin mutations and PI3KCA synergize to promote lipogenic liver tumors in mice

Abstract

In a previous study, we showed co‐expression of hMet and mutant beta‐catenin (S33Y, S45Y) in mouse liver using the sleeping beauty transposon/transposase (SB) delivered via hydrodynamic tail vein injection (HTVI) led to Hepatocellular cancer (HCC) which mimicked 10% of HCC patients by transcriptomic analysis. Since PI3KCA activation can occur downstream of hMet signaling, we co‐expressed H1047R‐mutant‐PI3KCA and β‐catenin mutants (S33Y or S45Y) using SB and HTVI. HCC was evident at 15 weeks post HTVI in the H1047R‐PI3KCA/S45Y‐β‐catenin group and at 22 weeks in the H1047R‐PI3KCA/S33Y‐β‐catenin group as compared to 6–7 weeks in hMet‐mutant‐β‐catenin studies. This delay in tumor development suggests that PI3KCA activation downstream of hMet may not be an important contributor to the hMet/beta‐catenin induced HCC. Intriguingly, all tumor cells in the PI3KCA/mutant‐β‐catenin mice displayed notable lipid accumulation. This suggests some unique features of PI3KCA and β‐catenin cooperation in HCC development. Further, introduction of dominant‐negative (dn) TCF4 at the time of PI3KCA/mutant‐β‐catenin HTVI prevented tumorigenesis. Intriguingly, co‐injection of dnTEAD2 prevented HCC in H1047R‐PI3KCA/S33Y‐β‐catenin and H1047R‐PI3KCA/S45Y‐β‐catenin model but no effect on hMet and mutant beta‐catenin model. Thus, PI3Kinase and β‐catenin cooperation results in HCC with unique characteristics such as lipogenic features and signaling involving Yap signaling. Also, the rate of tumorigenesis is notably slower than hMet‐β‐catenin model. All these findings suggest that PI3KCA is not downstream of hMet in cooperating with β‐catenin to induce HCC.