Dense mapping of MYH9 localizes the strongest kidney disease associations to the region of introns 13 to 15

Abstract

Admixture mapping recently identified MYH9 as a susceptibility gene for idiopathic focal segmental glomerulosclerosis (FSGS), HIV-associated nephropathy (HIVAN) and end-stage kidney disease attributed to hypertension (H-ESKD) in African Americans (AA). MYH9 encodes the heavy chain of non-muscle myosin IIA, a cellular motor involved in motility. A haplotype and its tagging SNPs spanning introns 12-23 were most strongly associated with kidney disease (OR 2-7; P < 10(-8), recessive). To narrow the region of association and identify potential causal variation, we performed a dense-mapping study using 79 MYH9 SNPs in AA populations with FSGS, HIVAN and H-ESKD (typed for a subset of 46 SNPs), for a total of 2496 cases and controls. The strongest associations were for correlated SNPs rs5750250, rs2413396 and rs5750248 in introns 13, 14 and 15, a region of 5.6 kb. Rs5750250 showed OR 5.0, 8.0 and 2.8; P = 2 x 10(-17), 2 x 10(-10) and 3 x 10(-22), respectively, for FSGS, HIVAN and H-ESKD; OR 5.7; P = 9 x 10(-27) for combined FSGS and HIVAN, recessive. An independent association was observed for rs11912763 in intron 33. Neither the highly associated SNPs nor the results of resequencing MYH9 in 40 HIVAN or FSGS cases and controls revealed non-synonymous changes that could account for the disease associations. Rs2413396 and one of the highly associated SNPs in intron 23, rs4821480, are predicted splicing motif modifiers. Rs5750250 combined with rs11912763 had receiver operator characteristic (ROC) C statistics of 0.80, 0.73 and 0.65 for HIVAN, FSGS and H-ESKD, respectively, allowing prediction of genetic risk by typing two SNPs.

Figure 1.

Map of MYH9 showing location of exons, and of 80 SNPs, with the strength of association of the SNP (recessive model for the risk allele) with combined FSGS and HIVAN indicated by color. HapMap linkage disequilibrium (D′) for MYH9 is shown.

Figure 2.

Receiver operating characteristic C statistics of haplotype E-1, SNPs rs2413396 and rs5750250, and combined genotypes of rs5750250 and rs11912763 as predictors of HIVAN, FSGS and H-ESKD. Statistics were calculated by counting the frequency with which, in random draws of cases and controls, the case has a more susceptible genotype than the control. For the E-1 and the SNPs, susceptibility of the genotype was taken to be the number of copies of the risk allele. For combined rs5750250 and rs11912763, susceptibility of the genotype was taken from the following ordering, from least to most susceptible, for rs5750250/rs11912763 genotypes: AA/GG or AA/AG; AG/GG or AG/AG; GG/GG; GG/AG; AG/AA or GG/AA.