Human genetic evidence supports MAP3K15 inhibition as a therapeutic strategy for diabetes

Diabetes mellitus is a chronic health condition that can result in significant end-organ complications and is estimated to impact at least 8.5% of the global adult population. Here, we performed gene-level collapsing analysis on exome sequences from 454,796 multi-ancestry UK Biobank participants to detect genetic associations with diabetes. Rare non-synonymous variants in GCK, GIGYF1, HNF1A, and HNF4A were significantly associated (P<1x10-8) with increased risk of diabetes, whereas rare non-synonymous variants in MAP3K15 were significantly associated with reduced risk of diabetes. Recessive carriers of rare non-synonymous variants in the X chromosome gene MAP3K15 had a 30% reduced risk of diabetes (OR=0.70, 95% CI: [0.62,0.79], P=5.7x10-10), along with reduced blood glucose (beta=-0.13, 95% CI: [-0.15,-0.10], P=5.5x10-18) and reduced glycosylated haemoglobin levels (beta=-0.14, 95% CI: [-0.16,-0.11], P=1.1x10-24). Hemizygous males carrying protein-truncating variants (PTVs) in MAP3K15 demonstrated a 40% reduced risk of diabetes (OR=0.60, 95% CI: [0.45,0.81], P=0.0007). These findings were independently replicated in FinnGen, with a MAP3K15 PTV associating with decreased risk of both type 1 diabetes (T1DM) and type 2 diabetes (T2DM) (p<0.05). The effect of MAP3K15 loss on diabetes was independent of body mass index, suggesting its protective effect is unlikely to be mediated via the insulin resistance pathway. Tissue expression profile of MAP3K15 indicates a possible involvement of pancreatic islet cell or stress response pathways. No safety concerns were identified among heterozygous or recessive MAP3K15 PTV carriers across over 15,719 studied endpoints in the UK Biobank. Human population genetic evidence supports MAP3K15 inhibition as a novel therapeutic target for diabetes.