Association Of High-Throughput Rnai And Drug Screening With Candidate Novel Therapeutic Targets In Esophageal Carcinoma

31 Background: Oesophageal cancer is the sixth most commonly diagnosed cancer worldwide and carries a poor prognosis. Targeted therapeutic strategies have relied on the identification of genes which are amplified and overexpressed but the discrimination between genes which drive cancer and those that are coincidental has not yet been fully achieved. Methods: We carried out a functional genetic screen of 714 kinases in 18 tumour cell line models of oesophageal adenocarcinoma (EAC) and squamous cell carcinoma (SCC) to identify genes critical to the survival of specific oesophageal subtypes. High throughput drug screening of 80 compounds, largely comprising those used in clinical trials or in routine clinical practice, was undertaken in parallel to reveal druggable targets and to validate functional screening results. Results: We show proof of principle that the integration of functional and drug profiling results with molecular profiling data is a valid approach for identifying druggable targets by correlating the effects of ERBB2, EGFR and CDK6 gene knockdown with sensitivity to drugs targeting their respective proteins. We have characterised new genetic dependencies for EAC involving TGF beta signalling and the JAK/STAT pathway. Decreased cell viability associated with silencing of ACVR2 and ACVR1C (activin receptors involved in TGF beta signalling) was consistent with sensitivity to nilotinib, a cABL inhibitor that modulates TGF beta signalling. Furthermore, decreased cell viability in EAC models was observed with silencing of JAK2. This strongly correlated with sensitivity to PF-04691502 and BEZ 235, both potent inhibitors of PI3K/mTOR which signal downstream of JAK2. We have validated novel sensitivity to small molecule tankyrase inhibitors in a subgroup of oesophageal models, suggesting that a subset of Wnt dependent oesophageal cancers could be targeted with these agents. Conclusions: The molecular and functional genetics of oesophageal cancer is poorly understood. By generating functional viability and drug sensitivity data for a panel of oesophageal tumour cell lines, we have identified new genetic dependencies and candidate drug targets for this aggressive disease.