Identification and functional analysis of glycemic trait loci in the China Health and Nutrition Survey
PLoS Genet. 2018-04-05;14(4):e1007275.
- Abstract
- To identify genetic contributions to type 2 diabetes (T2D) and related glycemic traits (fasting glucose, fasting insulin, and HbA1c), we conducted genome-wide association analyses (GWAS) in up to 7,178 Chinese subjects from nine provinces in the China Health and Nutrition Survey (CHNS). We examined patterns of population structure within CHNS and found that allele frequencies differed across provinces, consistent with genetic drift and population substructure. We further validated 32 previously described T2D- and glycemic trait-loci, including G6PC2 and SIX3-SIX2 associated with fasting glucose. At G6PC2, we replicated a known fasting glucose-associated variant (rs34177044) and identified a second signal (rs2232326), a low-frequency (4%), probably damaging missense variant (S324P). A variant within the lead fasting glucose-associated signal at SIX3-SIX2 co-localized with pancreatic islet expression quantitative trait loci (eQTL) for SIX3, SIX2, and three noncoding transcripts. To identify variants functionally responsible for the fasting glucose association at SIX3-SIX2, we tested five candidate variants for allelic differences in regulatory function. The rs12712928-C allele, associated with higher fasting glucose and lower transcript expression level, showed lower transcriptional activity in reporter assays and increased binding to GABP compared to the rs12712928-G, suggesting that rs12712928-C contributes to elevated fasting glucose levels by disrupting an islet enhancer, resulting in reduced gene expression. Taken together, these analyses identified multiple loci associated with glycemic traits across China, and suggest a regulatory mechanism at the SIX3-SIX2 fasting glucose GWAS locus.
- Consortium data used in this publication
- Summary genetic association results data can be found at http://mohlke.web.unc.edu/data/. All other relevant data are in the paper and Supporting Information files. We searched the following publicly available eQTL databases to identify cis-eQTLs at the observed loci: GTEx v7 [85], the University of Chicago eQTL browser [86], the Islet eQTL Explorer (http://theparkerlab.org/tools/isleteqtl/) [47], and the Blood eQTL Browser [45]. We also searched for cis-eQTLs in subcutaneous adipose tissue from the METSIM study [87]. All eQTL data sources used a false discovery rate (FDR) <5% for identifying cis-eQTLs, with the exception of the METSIM study, which used an FDR <1%.