Single nucleotide polymorphisms (SNPs) within a microRNA (miRNA) binding site of its target gene, referred to as miRSNPs, are known to have functional consequences for cancer risk. We investigated the association between 2,169 putative miRSNPs and PCa risk in a large population of 22,301 cases and 22,320 controls of European ancestry from 23 participating studies within the large PCa Consortium (PRACTICAL). We identified 22 SNPs to be associated with risk of PCa, seven of which have not been previously reported by GWAS studies. We then validated the functional role KLK3 rs1058205 (T>C) SNP. We showed that miR-3162-5p has specific affinity for the KLK3 (Prostate-specific antigen aka PSA) rs1058205 SNP T-allele.
Since PSA (KLK3) is a primary diagnostic biomarker for PCa, to understand this locus further, we undertook additional fine-mapping analysis of this locus and identified a coding variant, rs17632542: T>C (I161T codon-change), within the PSA gene to be significantly associated with PCa risk. We used allele-specific expression analysis and biochemical approaches to test the hypothesis that rs17632542 SNP regulates PSA/KLK3 gene expression, affects PSA antigenicity, protein function or complexing ability with serum inhibitors and thereby the free:total PSA ratio clinically. We showed that the rs17632542 SNP had profound effects on the proliferation, migration and bone metastasis of PCa PSA overexpression cell models. These effects in part were modulated by the fact that C allele of the rs17632542 SNP reduces stability and proteolytic activity of PSA enzyme as well as effects its mRNA splicing.
Our results provide evidence that PSA coding / 3’UTR variants have a functional effect that underpins its association with PCa. Incorporating the information on these two (and additional) SNP genotypes into the current PSA test may lead to the development of a personalised serum PSA test with implications for clinical decision-making at PCa diagnosis.