Alternative splicing regulation by GWAS risk loci for breast cancer
ABSTRACTRecent genome-wide association studies (GWAS) have revealed the association of hundreds of single nucleotide polymorphisms (SNPs) with breast cancer (BC) risk, which mostly locate in non-coding regions, suggesting regulatory roles to the causal variants. Functional characterisation of GWAS loci has been biased towards the effect of regulatory SNPs on transcription factor binding. Here we set out to determine the extent of the contribution of breast cancer risk-associated SNPs to alternative splicing (AS).We screened genome-wide significant (P ≤ 5 × 10−8) BC risk SNPs for association with AS, using expression and genotype data from normal breast samples, from the GTEx project. We identified four splicing quantitative trait loci (sQTL). In locus 6p22.1, rs6456883 is a significant cis-sQTL for the expression of ZNF311 gene isoforms. Three SNPs in locus 8p23.3, rs6682326, rs3008282 and rs2906324, were also identified as significant cis-sQTLs/svQTLs for the expression of RPL23AP53 gene isoforms. In-silico functional analysis revealed that these variants can potentially alter enhancer splicing elements within the target genes.Our work shows that BC risk-associated variants at two loci are associated with AS isoforms in normal breast tissue, thus demonstrating that AS plays an important role in breast cancer susceptibility. Furthermore, it supports that all cis-regulatory mechanisms should be considered in the functional characterisation of risk loci.