Dicky Struik¹, Janine K. Kruit, Helen Westra¹, A.J.A. van de Sluis¹, and Johan W. Jonker¹

¹Section of Molecular Metabolism and Nutrition, Department of Pediatrics, University of Groningen, University Medical Center Groningen, The Netherlands.

Background: Recent large-scale human genetic sequencing studies have identified the MAP3K15 gene, a kinase with unknown function, as a new and universal risk locus for type 1, type 2, and gestational diabetes. Recessive carriers of rare nonsynonymous variants in the MAP3K15 exhibit lower blood glucose and HbA1c levels and a 30% reduction in diabetes risk. However, the functional consequences of these variants on MAP3K15 protein activity and their role in modulating diabetes risk remain unclear.

Methods: We overexpressed the full-length human MAP3K15 protein and conducted unbiased phospho-proteome profiling to identify protein substrates indicative of MAP3K15 protein activity. Site-directed mutagenesis was used to generate diabetes-associated nonsynonymous MAP3K15 variants (R383H, S796L, P857L, A1127T, R1122*, and R1136*), and their activity was compared to wild-type MAP3K15. Additionally, we investigated streptozotocin-induced diabetes development in wildtype and MAP3K15 knockout mice.

Results: Overexpression of the wild-type human MAP3K15 protein led to dowstream phosphorylation of MAPK8, MAPK14, AKT1, and HSBP1. Phosphorylation of MAPK8 by nonsynonymous MAP3K15 variants was reduced by more that 90% compared to wild-type MAP3K15, while phosphorylation of MAPK14, AKT1, and HSBP1 remained largely unaffected. Five days post-streptozotocin treatment, blood glucose levels were significantly lower in MAP3K15 knockout mice compared to wildetype mice (P=0.0397; KO: 10.96±0.86 versus WT: 13.10±1.18).

Conclusion: We identified MAPK8 as a specific substrate of MAP3K15 and demonstrated that diabetes-associated nonsynonymous MAP3K15 variants act as loss-of-function mutations. Furthermore, MAP3K15 deficiency in mice partly protected against streptozotocin-induced diabetes development, providing the first evidence for a causal relationship between MAP3K15 and diabetes.