Genetic Variants in Selenoprotein Genes Selenos, GPX4, and SEPP1, but not Selenoprotein Levels, are Associated with the Development of Peripheral Artery Disease and the Inter-individual Variation in the Ankle-Brachial Index

Abstract

Introduction - Selenium (Se) is an essential micronutrient with important functions in human health and relevance to several pathophysiological conditions. The biological effects of Se are largely mediated by Se-containing proteins, referred to as selenoproteins. To date, a total of 25 selenoproteins have been identified in humans. The concentration of selenoproteins has been documented to correlate with a range of inflammatory markers and glucose homeostasis, which suggests their role in the pathogenesis of atherosclerotic peripheral arterial disease (PAD). In this study we analyze the relationships of single nucleotide polymorphisms (SNPs) in SEPP1, SELENOS, TXNRD1, TXNRD2, GPX4, and SOD2 genes as well as selenoprotein P (SeP) and thioredoxin (Trx) concentrations with the development of PAD and the values of ankle-brachial pressure index (ABI), a noninvasive marker of this disease. Methods - The two study samples were evaluated: 503 PAD patients and 594 controls in the discovery phase, and 408 PAD patients and 933 controls in the replication phase. All individuals were of Caucasian origin. SNPs were ascertained using pre-designed TaqMan SNP assays, and plasma selenoprotein levels with the ELISA immunoassay method. The main effects of each SNP and SNP-covariate interactions were assessed. Results - The SELENOS rs34713741T (recessive model: OR=1.49, 95%CI: 1.01-2.18, P=.043) and GPX4 rs713041T (dominant model: OR=1.31, 95%CI: 1.02-1.68; P=.035) alleles were associated with PAD, with a stronger effect in type 2 diabetes mellitus (T2DM; respective ORs: 2.77 95%CI: 1.00-7.68, P=.043 and 2.05, 95%CI: 1.19-3.53; P=.009). T2DM was related to 2.09-fold (95%CI: 1.56-2.78; P<.0001) increase in PAD risk. The association of the GPX4 SNP with PAD was confirmed in both a replication (P=.049) and a combined analysis (the discovery plus replication samples; P=.002), while the SELENOS SNP in a combined analysis only (P=.039). The influence of these SNPs on PAD risk suggested SNP-T2DM interaction (SELENOS x T2DM: ORObserved=5.47 > ORExpected=2.32; GPX4 x T2DM: ORObserved =2.87 > ORExpected= 1.53). The simultaneous presence of both risk genotypes with T2DM resulted in a 9.71-fold increase in PAD risk (95% CI: 2.21-42.45; P=.0002). Additionally, the SEPP1 rs7579A allele was associated with the presence of T2DM in PAD. All these associations, were independent of the influence of conventional cardiovascular risk factors distribution, except for the influence of GPX4 SNP on PAD, as this effect became nonsignificant after adjustment for T2DM. The SELENOS rs34713741T and SEPP1 rs7579A alleles, were the independent predictors of ABI in PAD in a multivariable analysis (allele dose: β= - 0.250, P=.003 and β= - 0.232, P=.006, respectively), explaining a total of 12.4% of the inter-individual variation in ABI. SeP concentration was decreased in T2DM (5.20 ±2.45 mg/ml, N=50 vs 3.99 ±2.51 mg/ml, N=26; P=.049), while being positively correlated with blood cholesterol levels, especially HDLC (r= 0.422, P=.0001). Conclusion - We found a link between the SeP levels and the occurrence of a more favorable profile of cardiometabolic risk factors in PAD. The SELENOS, SEPP1 SNPs contributed to the inter-individual variation in ABI, while the SELENOS, GPX4, SEPP1 SNPs were associated with PAD related to T2DM. These results confirmed the complex genetic architecture of PAD and ABI, in which candidate gene SNP main effects and SNP-covariate interactions contribute to the risk.