A key limitation of genome-wide association studies (GWAS) is that they do not specify either the underlying causal variant (s) or the effector gene (s) at a given locus. The widely reproduced obesity FAIM2 locus, named arbitrarily after the nearest gene, is relatively understudied given it is more strongly associated in children than in adults. Our previously reported trans-ancestral GWAS meta-analysis of childhood obesity enabled fine-mapping of the FAIM2 locus; in the ‘credible set’ of 6 SNPs, rs71329 had the highest probability of being causal. This non-coding variant resides in the 3’ untranslated region (UTR) of FAIM2. Given that obesity-related GWAS signals have been reported to be enriched in genes relevant to neuronal function, we generated ATAC-seq and high-resolution promoter-focused Capture C data in both primary astrocytes and ESC-derived hypothalamic neurons to carry out variant-to-gene mapping at this locus. We observed that rs71329 occupied a region of open chromatin, therefore a putatively functional element, and made physical contact with the FAIM2 promoter. These findings were also supported by both GTEx testes eQTL data and a transcriptome-wide association study in brain tissue. We sought to determine in vitro if the rs71329alleles influence enhancer activity in primary astrocytes. Using technical replicates repeated 8 times, the non-risk allele region increased luciferase reporter expression relative to the FAIM2 promoter alone by a fold change of 1.87, while the risk allele decreased expression by a fold change of 0.68 (non-risk allele+FAIM2 promoter vs. risk allele+FAIM2 promoter P<0.0001) . Our results show that rs71329resides in an enhancer element and suggest that the risk allele within the 3’ UTR of FAIM2 decreases the expression of this putative effector gene, which then confers risk of childhood obesity and in turn type 2 diabetes later in life.


S.H.Littleton: None. J.Bradfield: None. J.A.Pippin: None. C.Su: None. A.Chesi: None. A.D.Wells: Stock/Shareholder; Johnson & Johnson. R.I.Berkowitz: Research Support; Eisai Inc., Novo Nordisk. M.C.Pahl: None. S.F.Grant: None.


National Institutes of Health (HD056465)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.