Type 2 Diabetes-Related Variants Influence the Risk of Developing Prostate Cancer: A Population-Based Case-Control Study and Meta-Analysis

SIMPLE SUMMARY: We investigated the influence of GWAS-identified variants for T2D in modulating prostate cancer (PCa) risk through a meta-analysis of our data with those from the UKBiobank and FinnGEn cohorts and four large European cohorts. We found that genetic variants within the FTO, HNF1B, and JAZF1 loci were associated with PCa risk. Our results also suggested, for the first time, a potentially interesting association of SNPs within NOTCH2 and RBMS1 genes that need to be further explored and validated. This study also shed some light onto the functional mechanisms behind the observed associations, and demonstrated that the HNF1B(rs7501939) polymorphism correlated with lower levels of SULT1A1, an enzyme responsible for the sulfate conjugation of multiple endogenous and exogenous compounds. Furthermore, we found that SNPs within the HFN1B, NOTCH2, and RBMS1 genes impacted PCa risk through the modulation of mRNA gene expression levels of their respective genes. However, given the healthy nature of the subjects included in the cohort used for functional experiments, the link between the HNF1B locus and SULT1A1 should be considered still speculative and, therefore, requires further validation. ABSTRACT: In this study, we have evaluated whether 57 genome-wide association studies (GWAS)-identified common variants for type 2 diabetes (T2D) influence the risk of developing prostate cancer (PCa) in a population of 304 Caucasian PCa patients and 686 controls. The association of selected single nucleotide polymorphisms (SNPs) with the risk of PCa was validated through meta-analysis of our data with those from the UKBiobank and FinnGen cohorts, but also previously published genetic studies. We also evaluated whether T2D SNPs associated with PCa risk could influence host immune responses by analysing their correlation with absolute numbers of 91 blood-derived cell populations and circulating levels of 103 immunological proteins and 7 steroid hormones. We also investigated the correlation of the most interesting SNPs with cytokine levels after in vitro stimulation of whole blood, peripheral mononuclear cells (PBMCs), and monocyte-derived macrophages with LPS, PHA, Pam3Cys, and Staphylococcus Aureus. The meta-analysis of our data with those from six large cohorts confirmed that each copy of the FTO(rs9939609A), HNF1B(rs7501939T), HNF1B(rs757210T), HNF1B(rs4430796G), and JAZF1(rs10486567A) alleles significantly decreased risk of developing PCa (p = 3.70 × 10(−5), p = 9.39 × 10(−54), p = 5.04 × 10(−54), p = 1.19 × 10(−71), and p = 1.66 × 10(−18), respectively). Although it was not statistically significant after correction for multiple testing, we also found that the NOTCH2(rs10923931T) and RBMS1(rs7593730) SNPs associated with the risk of developing PCa (p = 8.49 × 10(−4) and 0.004). Interestingly, we found that the protective effect attributed to the HFN1B locus could be mediated by the SULT1A1 protein (p = 0.00030), an arylsulfotransferase that catalyzes the sulfate conjugation of many hormones, neurotransmitters, drugs, and xenobiotic compounds. In addition to these results, eQTL analysis revealed that the HNF1B(rs7501939), HNF1B(rs757210), HNF1B(rs4430796), NOTCH2(rs10923931), and RBMS1(rs7593730) SNPs influence the risk of PCa through the modulation of mRNA levels of their respective genes in whole blood and/or liver. These results confirm that functional TD2-related variants influence the risk of developing PCa, but also highlight the need of additional experiments to validate our functional results in a tumoral tissue context.