A population-specific low-frequency variant of SLC22A12 (p.W258*) explains nearby genome-wide association signals for serum uric acid concentrations among Koreans

Abstract

Prolonged hyperuricemia is a cause of gout and an independent risk factor for chronic health conditions including diabetes and chronic kidney diseases. Genome-wide association studies (GWASs) for serum uric acid (SUA) concentrations have repeatedly confirmed genetic loci including those encoding uric acid transporters such as ABCG2 and SLC9A2. However, many single nucleotide polymorphisms (SNPs) found in GWASs have been common variants with small effects and unknown functions. In addition, there is still much heritability to be explained. To identify the causative genetic variants for SUA concentrations in Korean subjects, we conducted a GWAS (1902 males) and validation study (2912 males and females) and found four genetic loci reaching genome-wide significance on chromosomes 4 (ABCG2) and 11 (FRMD8, EIF1AD and SLC22A12-NRXN2). Three loci on chromosome 11 were distributed within a distance of 1.3 megabases and they were in weak or moderate linkage disequilibrium (LD) states (r2 = 0.02-0.68). In a subsequent association analysis on the GWAS loci of chromosome 11 using closely positioned markers derived from whole genome sequencing data, the most significant variant to be linked with the nearby GWAS signal was rs121907892 (c.774G>A, p.W258*) of the SLC22A12 gene. This variant, and each of the three GWAS SNPs, were in LD (r2 = 0.06-0.32). The strength of association of SNPs with SUA concentration (negative logarithm of P-values) and the genetic distance (r2 of LD) between rs121907892 and the surrounding SNPs showed a quantitative correlation. This variant has been found only in Korean and Japanese subjects and is known to lower the SUA concentration in the general population. Thus, this low-frequency variant, rs121907892, known to regulate SUA concentrations in previous studies, is responsible for the nearby GWAS signals.

Fig 1

(A) Manhattan plot of the GWAS results for SUA concentration. The horizontal red line represents the genome-wide significance level. Regional plots of the GWAS signals are shown for chromosome 11, 65–66 Mb (B), 64–65 Mb (C) and chromosome 4, 89 Mb (D). The red circle indicates the most significant SNP in the locus, and the circle size is proportional to the strength of LD (r²) with the most significant SNP. (B) Rs184521656 and rs117625825 are in a moderate LD relationship (r² = 0.68).

Fig 2. The role of rs121907892 in the GWAS signal on chromosome 11: 64–66 Mb region.

All dots in the figure are SNPs showing a genetic association with rs121907892, with r² values > 0.01. Green circles are representative SNPs of each locus that reached genome-wide significance in the GWAS. The circle size is proportional to the strength of LD (r²) with rs121907892 (red circle). (A) Correlation between LD value (r²) with rs121907892 and the significance of association (–log₁₀ p) of SNPs with serum uric acid (SUA) concentrations. (B) Regional plot for the results of the GWAS for SUA concentration. (C, D) Regional plots for the results of the association analysis for SUA concentration using SNPs derived from whole genome sequencing data before (C) and after (D) the conditional analysis incorporating rs121907892.

Fig 3. Effect of the p.W258* genotype on serum uric acid concentration in this study group.