Biology Faculty Research

Title

the Expression Phenotype of SNPs Linked to the Risk for Prostate Cancer

Document Type

Conference Proceeding

Publication Date

Spring 4-5-2014

Abstract

Numerous Genome Wide Association (GWA) studies of large populations have provided limited biomarkers for cancer and few linkages to phenotype. We investigated whether SNPs that are known to be associated with prostate cancer (PCa) affect expression in prostate stroma. The stroma of prostate tumors provides genetically homogenous tissue type with hundreds of tumor-dependent expression changes [1] which favors linkage identification. We surveyed published GWA studies covering a total of ∼90,000 patients and selected the most significant 35 susceptibility loci for further linkage analysis. We also selected 4030 transcripts previously associated with PCa diagnosis and prognosis. We then investigated 47 PCa cases using genetic variations in nontumor DNA (Illumina Human1M-Duov3_B snp arrays) and correlating the genotypes with RNA gene expression data from the same patients (U133 plus 2.0 arrays). A novel eQTL (Expression Quantitative Trait Loci) analysis was carried out by a modified BAYES method to analyze the associations between the risk variants and expressed transcripts jointly in a single model. The model assumed that a risk variant can be associated with multiple RNA expression changes and vice versa. We observed 47 significant linkages between 8 risk variants and 46 significant expression changes. The false discovery rates of two resampling studies are 0.03 and 0.05, respectively, which indicated the 47 linkages are significant. All linkages are “trans' linkages indicating remote effects of key SNPs on expressing genes. This is the first study to identify trans-linkages between multiple SNPs and multiple significant expression changes identified in cancer stroma. The linkages also associated with trends in clinical phenotypes. The strongest trend is between linkage rs9623117(C)-FOXD1 and biochemical relapse (p=0.01). The study reveals links between SNPs associated with PCa risk and transcription, which may identify the subset of patients where these genotypes may affect relevant phenotypes.

Publication Title

105th Annual Meeting of the American-Association-for-Cancer-Research (AACR)