Abstract 3500: A pipeline to identify driver mutations

Abstract Introduction: Cancer tissues harbor thousands of mutations while the current list of clinically validated actionable variants contains only about a dozen genetic markers. Most of the somatic mutations in cancer are expected to be inconsequential passenger mutations that reflect the general instability of the tumors. The discovery of most of the currently known driver mutations has been facilitated by their accumulation in mutation hot-spots within their respective genes. However, a great majority of mutations in cancer tissues are rare and no information is currently available about their functional significance. Several lines of in vitro and in vivo clinical evidence also indicate that there is a significant number of, as yet unidentified, activating driver mutations that could serve as predictive markers for oncology. Method: To identify functional driver mutations, we have established a functional genetics screen based on expression of random gene variants and the ability of driver mutations to promote a growth-advantage in vitro, as compared to passenger mutations. The screen was set up using a randomly mutated expression library encoding thousands of variants of epidermal growth factor receptor (EGFR), a well-known oncogene, as a model. The library was retrovirally introduced into murine lymphoid Ba/F3 cells, that normally require interleukin-3 (IL-3) for survival but can exploit ectopic expression of activated variants of oncogenic kinases to compensate for the deficiency of exogenous IL-3. While expression of wild-type EGFR did not promote IL3-independent survival of the Ba/F3 cells, as expected, transduction of the mutant EGFR library did, indicating that the surviving Ba/F3 pool included EGFR mutations with ligand-independent activity. To identify and quantify the frequency of these mutations, targeted next-generation sequencing of the EGFR inserts was carried out. Results: Using the method we were able to identify approximately 20 candidate activating mutations out of the 7000 random EGFR mutations in the original library. The 20 candidates included the well-known activating EGFR mutation, L858R, validating the pipeline. A previously unidentified activating mutation was also found, and the growth promoted by the mutant was susceptible to inhibition by clinically available EGFR inhibitors. Citation Format: Deepankar Chakroborty, Kari Kurppa, Ilkka Paatero, Laura Elo, Klaus Elenius. A pipeline to identify driver mutations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3500.