SNP interactions in the genetic architecture of blood pressure.

摘要

Evidence suggests that there is a substantial genetic component to the regulation of blood pressure (BP) but most BP variance remains unattributed to specific genetic variants. Candidate gene studies show linkage to rare Mendelian disorders but inconsistent association with BP. Massive consortia of genome-wide association studies have identified several dozen loci explaining less than 3% of BP variance. Many explanations have been offered for these observations, of which the current study addresses two: the focus on SNP main effects to the exclusion of interactions, and the low statistical power resulting from the strong multiple-testing burden necessarily imposed when analyzing millions of variants. This study examined the role of interactions among a limited set of variants (818 SNPs in 70 genes) drawn from two broad signaling pathways: renal ion balance and inflammation. Exhaustive SNP-SNP interaction testing was performed, as was analysis of SNP-sex, SNP-BMI, and SNP-hypertension interactions. In hypertensive subjects only, rs12821401 in ADIPOR2 was significantly associated with SBP (p=6.1e-5). A set of three SNP-SNP interactions among 6 inflammation genes (including ADIPOR2) showed significant evidence of containing one or more true association signals (FDR=4.2%), though no single interaction reached experiment-wise significance. These results provide evidence of a role for SNP-SNP and other SNP interactions and implicate an inflammation gene, ADIPOR2, not previously associated with hypertension.