Introduction Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal disease due to its aggressive nature. Patients typically present with distant metastases, at which point cytotoxic agents can extend life expectancy by several months at most. Large-scale phosphoproteomics complements our knowledge obtained from genomics and transcriptomics as it provides information on which proteins and kinases are phosphorylated, thereby implicating pathways that are activated. This approach in cancer research may lead to improved patient selection for treatment with tyrosine kinase inhibitors (TKI). This study is the first to employ phosphotyrosine-based phosphoproteomics on three different preclinical PDAC models as well as patient tumor tissues to understand the aggressive nature of this disease and identify new drug targets.

Approach We performed phosphoproteomics on a panel of 11 PDAC cell lines, 7 primary cell cultures, 10 patient-derived xenografts (PDX) and 16 fresh frozen human tumor tissues. Tyrosine phosphopeptides were enriched via immunoprecipitation and phosphopeptides were analyzed by high-resolution nano-LC mass spectrometry.

Results Using phosphotyrosine-based phosphoproteomics, we identified a total of 1723 tyrosine phosphorylated proteins and 138 phosphorylated kinases, representing 27% of the kinome. The reproducibility of our workflow was very high, with Pearson correlation coefficients of r = 0.937 for technical replicates of cell lines and r = 0.876 for biological replicates of tumors. In our cell line panels, multiple kinases were commonly highly phosphorylated (e.g. PTK2, EPHA2, EGFR and MET). Functional testing of PTK2 by using TKI defactinib in primary cell lines with high phosphorylation resulted in inhibition of proliferation and migration in vitro. Inhibition of EPHA2 by shRNAs resulted in reduced proliferation in vitro. To validate the relevance of these candidate target proteins in vivo, the tyrosine phosphoproteome of PDXs and human tumors was analyzed. In these tumors, kinase activity analysis based on kinase phosphorylation levels and kinase-substrate networks validated these common active nodes in the majority of these tumors.

Conclusion Our extensive tyrosine phosphoproteome analysis spanning a wide range of PDAC models revealed high phosphorylation levels of multiple kinases. Interestingly, the phosphorylated kinase profiles of tumors and cell lines did not show as much heterogeneity as expected, taken into account the existence of biological subtypes in PDAC identified by others via transcriptomics. The aggressive biology of this disease may be correlated with the consistent activation of multiple pathways, some of which we have shown to be targetable in vitro. This study prompts further validation and prognostic evaluation of the identified active kinases to improve treatment of PDAC.

Citation Format: Tessa Y. Le Large, Maarten F. Bijlsma, Btissame El Hassouni, Nicolla Funel, Nicole C. van Grieken, Helene Damhofer, Jaco C. Knol, Sander R. Piersma, Thang V. Pham, Henk M. Verheul, Hanneke W. van Laarhoven, Geert Kazemier, Elisa Giovannetti, Connie R. Jimenez. Phosphoproteome networks display consistent hyperactive kinase activity in pancreatic cancer: evidence for new therapeutic options [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4141. doi:10.1158/1538-7445.AM2017-4141