Abstract 2928: Functional effects of SNPs in non-coding RNAs at the 3q25 ovarian cancer susceptibility locus

Abstract

Abstract Genome-wide association studies have identified six common, low penetrance susceptibility loci for epithelial ovarian cancer (EOC). The functional mechanisms underlying disease susceptibility at these loci are unknown. SNPs at the 3q25 locus confer a 1.2-fold risk of serous EOC (P=7.1x10-8). To identify causal SNP(s) we used the 1000 genomes project to identify 952 correlated SNPs within a 884kb linkage disequilibrium region. Of these, 786 tagging SNPs were genotyped in ∼20,000 invasive EOC cases and 20,000 controls. Fine mapping limited the size of the risk region to ∼130kb and 43 possible causal SNPs. This region contains four protein-coding genes; none of the 43 SNPs was predicted to have a functional effect on the translated protein. Genomic profiling data from The Cancer Genome Atlas analysis of ∼500 serous EOCs indicated one gene in the region (TIPARP) was significantly down regulated compared to normal tissues (P=0.005). Using RT-PCR expression analysis, we found significantly lower TIPARP expression in EOC cell lines (N=37) compared to normal ovarian surface epithelial (OSE) and fallopian tube epithelial (FTE) cell lines (N=39) (P=1.57x10-8). TIPARP was also significantly downregulated in an in vitro model of early stage transformation of OSE cells (P=0.0003). Collectively, these data suggest that TIPARP is a novel tumor suppressor gene (TSG) in EOC development. TIPARP is a poly(ADP-ribose) polymerase known to catalyze histone ribosylation, an epigenetic modification that plays a role in DNA repair. We evaluated its functional significance in more detail using stable lentiviral-mediated shRNA knockdown of TIPARP in normal OSE/FTE cells. TIPARP knockdown induced cellular senescence similar to the effects of many classical TSG knockdowns in normal cells. We hypothesized that TIPARP may be involved in cellular response to DNA damage; when normal OSE/FTE cells were exposed to α-irradiation TIPARP was significantly upregulated in a dose-dependent manner. We examined the functional link between susceptibility SNPs and TiPARP expression. Of 43 putative causal SNPs, four fell within two non-coding RNAs- LOC730091 and TIPARP-AS1. There is no homology between LOC730091 and genes at the 3q25 locus; but TIPARP-ASI is a non-coding RNA with antisense sequence similarity to exon 1 of TIPARP. LOC730091 and TIPARP-AS1 were variably expressed in both normal ovarian and EOC cell lines. We found no correlation between LOC730091 and TIPARP expression; but cell lines expressing TIPARP-AS1 showed significantly reduced TIPARP expression. In summary, fine mapping at the 3q25 has identified multiple possible causal SNPs. Functional studies implicate TIPARP as a plausible target TSG in this region. EOC risk associated SNPs were located in a non-coding RNA that causes reduced expression of TIPARP suggesting a mechanism by which susceptibility SNPs may confer neoplastic transformation of ovarian epithelial cells. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2928. doi:1538-7445.AM2012-2928