Abstract 2568: Genetic association of POU5F1B at 8q24 with prostate cancer.

Numerous genetic variants at 8q24 have been associated with risk for multiple common cancers, most prominently for prostate cancer. POU5F1B is a retrotransposed copy of the master stem cell factor OCT4, and one of few known genes within the 8q24 interval. While OCT4 is expressed selectively in embryonic stem cells, POU5F1B retains its open reading frame and has been shown to be expressed in cancers. Ectopic expression of OCT4 blocks progenitor cell differentiation and causes dysplasia, raising the possibility that aberrant expression of POU5F1B might have analogous effects. Because POU5F1B sequence is non-unique, a potential contribution of POU5F1B genetic variation to cancer risk has not been investigated in prior GWAS. We evaluated genetic variants of POU5F1B for association with prostate cancer within a case-control study of familial prostate cancer at Vanderbilt University that included 830 familial prostate cancer probands, and 830 age-matched control probands without a personal or family history of prostate cancer. Our study identified four missense variants within POU5F1B, and eight haplotypes of > 1% frequency. G176E was associated with prostate cancer (P = 3.1 x 10−4) and was in LD with GWAS SNP rs6983267 (r2 = 0.69). The minor allele (encoding glutamate) uniquely marked a single haplotype (the most common) with an inverse association with prostate cancer (P = 5.7 x 10−4, OR = 0.78 (95%CI 0.7-0.9)). As expected, the haplotype association was correlated with that of GWAS SNP rs6983267. A second haplotype differing from the first only at G176E (encoding glycine) was independently associated with risk of prostate cancer (P = 5.5 x 10−7, OR = 1.91 (95%CI 1.5-2.5)). However, glycine would also be encoded at this position by all other haplotypes, none carrying a risk effect. The risk haplotype was tagged by the pair of missense variants, G176E (glycine) and E238Q (glutamine). In a single allele analysis, the latter variant had no apparent association with prostate cancer. The risk effect of the former adjusted for the latter was OR = 2.12 (95%CI 1.6-2.7), P = 1.7 x 10−8; the risk effect of the latter adjusted for the former was OR = 1.79 (95%CI 1.4-2.3), P = 9.4 x 10−6. The two missense variants would alter each of the DNA-binding domains: one altering a highly conserved site with a non-conservative substitution, and the other abrogating a kinase site known to physiologically regulate other POU transcription factors. Our data suggest that combinations of these variants may be associated with risk of prostate cancer and warrant further investigation of the potential role of POU5F1B in the 8q24 cancer associations. Citation Format: Joan P. Breyer, Jeffrey R. Smith. Genetic association of POU5F1B at 8q24 with prostate cancer. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2568. doi:10.1158/1538-7445.AM2013-2568