Recurrent copy number gains drive PKCι expression and PKCι-dependent oncogenic signaling in human cancers.

Abstract

PRKCI is frequently overexpressed in multiple human cancers, and PKCι expression is often prognostic for poor patient survival, indicating that elevated PKCι broadly plays an oncogenic role in the cancer phenotype. PKCι drives multiple oncogenic signaling pathways involved in transformed growth, and transgenic mouse models have revealed that PKCι is a critical oncogenic driver in both lung and ovarian cancers. We now report that recurrent 3q26 copy number gain (CNG) is the predominant genetic driver of PRKCI mRNA expression in all major human cancer types exhibiting such CNGs. In addition to PRKCI, CNG at 3q26 leads to coordinate CNGs of ECT2 and SOX2, two additional 3q26 genes that collaborate with PRKCI to drive oncogenic signaling and tumor initiation in lung squamous cell carcinoma. Interestingly however, whereas 3q26 CNG is a strong driver of PRKCI mRNA expression across all tumor types examined, it has differential effects on ECT2 and SOX2 mRNA expression. In some tumors types, particularly those with squamous histology, all three 3q26 oncogenes are coordinately overexpressed as a consequence of 3q26 CNG, whereas in other cancers only PRKCI and ECT2 mRNA are coordinately overexpressed. This distinct pattern of expression of 3q26 genes corresponds to differences in genomic signatures reflective of activation of specific PKCι oncogenic signaling pathways. In addition to highly prevalent CNG, some tumor types exhibit monoallelic loss of PRKCI. Interestingly, many tumors harboring monoallelic loss of PRKCI express significantly lower PRKCI mRNA and exhibit evidence of WNT/β-catenin signaling pathway activation, which we previously characterized as a major oncogenic pathway in a newly described, PKCι-independent molecular subtype of lung adenocarcinoma. Finally, we show that CNG-driven activation of PKCι oncogenic signaling predicts poor patient survival in many major cancer types. We conclude that CNG and monoallelic loss are the major determinants of tumor PRKCI mRNA expression across virtually all tumor types, but that tumor-type specific mechanisms determine whether these copy number alterations also drive expression of the collaborating 3q26 oncogenes ECT2 and SOX2, and the oncogenic PKCι signaling pathways activated through the collaborative action of these genes. Our analysis may be useful in identifying tumor-specific predictive biomarkers and effective PKCι-targeted therapeutic strategies in the multitude of human cancers harboring genetic activation of PRKCI.