Abstract A62: Assessment of genetic diversity of EphB2 and its role as a genetic risk factor for prostate cancer among African American men

Our group was the first to implicate the EphB2 tyrosine kinase receptor as a tumor suppressor in prostate cancer (PC), where we reported somatic inactivating mutations occurring in 5%–10% of sporadic tumors. With its role in maintaining normal epithelial architecture and functional data suggestive of a tumor suppressor in both prostate and colon cancer, EphB2 is an attractive candidate genetic risk factor for PC. First, in a screen of EphB2 coding variants in African American familial and sporadic cases and controls, we showed that the frequency of a common germline nonsense mutation (K1019X) was significantly higher in African American men with a positive family history of prostate cancer as compared to matched controls. Following on these findings, we hypothesized that there might be other common variants in EphB2 that confer risk of sporadic prostate cancer among African American men. To test this hypothesis, we genotyped 490 African American cases and 567 matched controls with 335 genetic variants mapping within the EphB2 locus. SNPs genotyped included all haplotype tagging SNPs based on the Yoruban population of Ibadan, Nigeria, International HapMap data, and several novel coding variants discovered by Sanger sequencing. Logistic regression revealed 10 variants showing statistically significant association with prostate cancer risk after correction for multiple testing, including rs4654822 (Odds Ratio = 0.59 (CI=0.47–0.74; P=4.0X10-06). Furthermore, to supplement these data, we used Next Generation Sequencing to interrogate the entire EphB2 genomic locus in a pool of 400 African American cases and controls for discovery of all EphB2 common variants in our cohort. This represents ∼200,000 basepairs of total genomic sequence in 800 haploid genomes or 160,000,000 bases of DNA sequence. Using the Applied Biosystems SOLiD system, we generated >500,000,000 bases of sequence in a single run. We are currently analyzing these data using bioinformatics tools to discover novel common and rare variants in the EphB2 gene, with a focus on regions encompassing SNPs showing association to prostate cancer. We hope to further validate SNPs showing association to prostate cancer and any nearby novel SNPs identified by our sequencing efforts in an independent cohort of men of recent African descent. These data form the basis of an extremely thorough analysis of EphB2 as a genetic risk factor for prostate cancer among men of recent African descent.