ASSOCIATION ANALYSIS OF 18 RECENTLY DISCOVERED SERUM URATE LOCI WITH GOUT

Background: Elevated serum urate levels are causal for gout. The first genome-wide association studies associated common genetic variation in 10 loci of stronger effect with serum urate. Predictably, the majority of these loci (e.g. SLC2A9 and ABCG2) are dominated by uric acid transporters involved in renal and gut excretion of uric acid, and have been associated with gout in multiple ancestral groups. However, despite very strong evidence for association with serum urate levels, PDZK1 is unequivocally not associated with gout (OR=1.03) in a large European sample set. Therefore, it is important to test serum urate loci for association with gout. Lack of association could indicate pleiotropic effects (e.g. PDKZ1), and understanding the reasons for lack of association should yield insights into the etiology of gout. Recently, a further 18 serum urate loci have been identified by meta-analysis of individual GWAS, of which 9 were nominally associated with gout in >3,000 cases. These loci include genes involved in glycolysis, but no known uric acid transporters. Objectives: To test the 18 recently discovered loci for association with gout in New Zealand European and Polynesian (Maori and Pacific Island) case-control sample sets. Methods: All gout cases were ascertained by clinical examination according to the American Rheumatology Association criteria and comprised 648 European and 888 Polynesian cases. The 1550 European and 1095 Polynesian controls self-reported no gout. The 18 genetic variants were genotyped by Sequenom multiplex technology or Taqman SNP genotyping. Association analysis was performed using STATA. Age and sex were adjusted for in both sample sets, with European admixture adjusted for in Polynesian. Results: There was evidence for association with gout at six of the loci in the European sample set (TRIM46, SFMBT1, PRKAG2, A1CF, INHBB, IGF1GR). In all instances, the same allele associated with increased serum urate was also associated with gout. Two of the loci were associated in the Polynesian samples (SFMBT1, VEGFA) with the same allele associated with gout as associated with increased serum urate in Europeans. Analysis of participants with higher Polynesian ancestry revealed association at SFMBT1, VEGFA, PRKAG2 and HLF. For PRKAG2 and HLF, the allele associated with decreased serum urate in European was associated with increased risk of gout in Polynesian. Meta-analysing the NZ European and Polynesian data sets revealed association at MAF. Conclusions: Nine of the 18 loci had previously been associated with gout by Kottgen et al. (TRIM46, SFMBT1, TMEM171, VEGFA, BAZ1B, PRKAG2, STC1, A1CF, NFAT5). We replicated four of these loci (TRIM46, SFMBT1, PRKAG2, A1CF) and associated a further four loci with gout (INHBB, IGF1R, MAF, HLF). Our results also suggest alleles with opposing effects between European and Polynesian at PRKAG2 and HLF. This may indicate recombinant haplotypes, which can be exploited in fine-mapping the etiological variants. References: 1. Köttgen et al. Nat Genet. 2013;45:145 Disclosure of Interest: None declared DOI: 10.1136/annrheumdis-2014-eular.4679Citation: Annals of the Rheumatic Diseases, volume 73, supplement 2, year 2014, page 353Session: Genomics, genetics and epigenetics of rheumatic diseases (Poster Presentations )