Characterization of tumorigenesis in the PP2A B'delta knockout mouse: A model for hepatocarcinogenesis

To fully transform immortalized human cells into malignant cells, Protein Phosphatase type 2A (PP2A) needs to be inactivated, identifying PP2A as a major cellular tumor suppressor. It remained however an open question, whether in vivo loss of PP2A might also predispose mice tocancer development. In this work, we found that Ppp2r5d KO mice, deficient for the ubiquitously expressed PP2A B'delta subunit, spontaneously develop several primary tumors, with a high incidence of Hepatocellular Carcinoma (HCC) and lymphoma. Hence, we demonstrated for the first time that loss of one specific PP2A holoenzyme, is without doubt, sufficient to promote tumorigenesis in vivo. Because of the high incidence of HCC and the rarity of spontaneous HCC development in mice, we focused on the further characterization of this phenotype. B'delta-null HCCs feature all histopathological stages seen in human HCC, rangingfrom the most benign dysplastic nodules to the most malignant, poorly-differentiated HCCs, without any correlation with increased inflammation or steatosis. Furthermore, we observed intracellular fibrinogen deposits, another feature of human HCC, and an unbiased 2D DIGE analysis confirmed differential expression of several proteins/markers of human HCC. In order to define the exact role of B'delta in hepatocarcinogenesis, we performed targeted western blot studies in protein lysates of several independent samples of HCCs, healthy surrounding liver tissues and WT liversto analyze potential activation of signaling pathways known to be disturbed in HCC and to be regulated by PP2A. In general, these results reflected the enormous diversity, so notoriously associated with HCC. However, amidst all this heterogeneity, two consistent alterations were observed in all HCC samples tested: an increase in S62 phosphorylation of the c-myc oncogene and an increase in S9 phosphorylation of GSK-3, the c-myc T58 kinase. These findings are consistent with increased c-myc oncogenicactivity. In accordance, we found increased expression of several established c-myc target genes. Because GSK-3 S9 hyperphosphorylation is already present in non-malignant KO liver, the lack of direct GSK-3 dephosphorylation is an integral part of the tumor predisposing mechanism in B'delta-null mice. In addition, we hypothesize that PP2A-B'delta may be themajor c-myc S62 phosphatase in liver, since expression of B'alpha, the established PP2A c-myc-targeting subunit, is barely detectable in this tissue. Interestingly, we identified several direct and indirect mechanisms, resulting in PP2A-B'delta malfunction in human cancers. Together, our data identify B'delta as a novel candidate tumor suppressor in HCC, and B'delta KO mice as a valuable novel animal model to study hepatocarcinogenesis.

Publication year:2013

Accessibility:Closed