Abstract 1275: Identification of cancer-specific synthetic lethal genes as novel therapeutic targets

Abstract

Abstract Introduction Genomic level data has significantly increased the depth of our understanding of genetic and epigenetic changes in cancer. However, using this complex data to improve patient management is challenging. Synthetic lethal (SL) genes (which are only required for growth/survival of cancer cells and not normal cells), represent ideal targets for the development of more targeted, less toxic and more effective cancer treatments. Here we report a novel bioinformatics approach that exploits the complexity of DNA methylation/expression changes to identify cancer subtype specific SL genes. Materials and Method Publicly available genome wide DNA methylation and expression data was obtained for ALL (n=517), medulloblastoma (n=763) and neuroblastoma (n=213). For neuroblastoma, which lacks established subgroups, novel subgroups were defined based on differential methylation (using the t-SNE/dbSCAN method) SL gene candidates were derived from genomic methylation/expression data using our in-house developed bioinformatic pipeline. Functional validation of identified candidate SL genes was performed using siRNA-mediated knockdown, coupled with analysis of cell proliferation (MTT assay) and apoptosis (annexin-V staining/caspase activity). Results 22 candidate SL genes were identified across six ALL genetic subgroups, ranging from nine in the TCF3-PBX1 to one in the iAMP21 subgroup. The identified genes lack any apparent genetic/epigenetic alterations, and thus most of the candidates have not previously been implicated in cancer. siRNA-mediated partial knockdown of ETV6/RUNX1 and TCF3-PBX1-specific candidate SL genes TUSC3 and FAT1 respectively resulted in a 40-60% reduction in cell proliferation (p <0.01) and induction of apoptosis (p <0.01), which was specific to the subtype in which the genes were predicted to be synthetically lethal. In medulloblastoma, seven candidate SL genes each were identified in the well-defined WNT and SHH subgroups, while identification of SL genes was limited in group 3 (0) and group 4 (1) tumours, which lack known group-defining molecular defects. However, using the recently defined further subtyping of Group3/Group4 tumours, nine additional candidates were identified across three of the novel subtypes. siRNA mediated knockdown of 5 candidate genes tested to date in the SHH subtype led to significant inhibition of cell growth (p<0.001) and induction of apoptosis (p<0.001). In neuroblastoma, we identified five novel methylation-dependent subgroups, which correlate with known molecular and clinical data. Six candidate SL genes were identified in methylation cluster 3, which consists almost exclusively of high-risk MYCN amplified cases. Conclusion We have demonstrated that SL genes, specific for defined molecular subtypes of cancer, can be identified utilising a novel approach combining methylation/expression data in multiple cancer types. Citation Format: Lalchung Nunga, Ed Schwalbe, Fadhel Lafta, Deborah Tweddle, John Maris, Timothy Barrow, Gordon Strathdee. Identification of cancer-specific synthetic lethal genes as novel therapeutic targets [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1275.