Genome-wide association studies (GWAS) have previously shown that variants in introns of the FTO gene are associated with a predisposition to obesity and type 2 diabetes in humans. Now a study (Nature doi:10.1038/nature13138, 12 March 2014) in human tissue and mouse models show that these variants alter the gene expression of transcription factor IRX3, and not FTO, which was formerly thought to be the key player in obesity in this region.

Previous studies have failed to link FTO gene variants with alterations in its expression in humans, which prompted Smemo et al. to use techniques that allow the identification of long-range chromatin interactions of the variants. In the brains of embryonic and adult mice, these obesity-associated variants physically interacted with the Irx3 promoter. Furthermore, in tissues from human brains, IRX3 expression correlated with the presence of single nucleotide polymorphisms associated with increased body-mass index. The authors found that mice lacking Irx3 were resistant to high-fat diet–induced obesity and insulin resistance and had 'browning' of white adipose tissue, suggesting that Irx3 may regulate energy homeostasis.

The authors' tale is a cautionary one, as it underscores the need to rethink how we follow up GWAS hits in noncoding intervals to confirm candidate target genes in human diseases.