Search

The genetic basis of type 2 diabetes (T2D) is poorly understood, with only three common gene variants confirmed. As part of the Wellcome Trust Case Control Consortium we have genotyped 500,568 single-nucleotide polymorphisms (SNPs) in 1954 type 2 diabetic patients and 2922 population controls. In a preliminary analysis, and after stringent quality control (including > 98% genotype success rate, HWE P > 0.01 in cases and controls), we analyzed ∼309,000 SNPs with minor allele frequencies of >5% that cover 62% of common variation in the human genome. Replication analyses were performed on 10 of the top signals using an additional 3073 UK cases and 3879 UK controls. SNPs showing the strongest associations included a cluster in the FTO gene on chromosome 16 (e.g. rs9939609; MAF=0.4; odds ratio (OR)=1.26 [95% CI 1.16-1.37], P = 3 x 10[sup -8]) that reached genome wide significance. This result was replicated (OR = 1.19 [95% CI 1.11 - 1.28], P = 2 x 10[sup -8]) in the additional cases and controls, giving a combined OR = 1.22 [95% CI 1.16-1.29], P = 5 x 10[sup -13]. Linkage disequilibrium around associated SNPs extended 47kb within FTO (r² > 0.2). Known type 2 diabetes risk alleles in the TCF7L2, PPARG and KCNJ11 genes occurred at ranks 1,264 and 790 respectively. The P 12A variant in PPARG and a SNP that tags the E23K variant (r² = 0.9) in KCNJ11, did not reach genome wide levels of significance (OR = 1.26 [95% CI 1.11 - 1.44], P=0.0004) and OR = 1.15 [95% CI 1.05 - 1.24], P = 0.0016) respectively, but effect sizes were consistent with previous large studies and meta-analyses. In conclusion, the first UK genome wide association study has identified FTO, which is widely expressed in early development as a gene contributing to T2D risk. Results in known T2D SNPs show that further large replication studies are needed to confirm potential novel genes. [ABSTRACT FROM AUTHOR]