Mishra, K.; Sakban, R. B.; Shankar, S.; Wong, J.; Farah, B. L.; Guo, J.; Ching, J.; Tham, M. S.; Kovalik, J.-P.; Gurley, S. B.; Petretto, E.; Tolwinski, N. S.; Coffman, T. M.; Behmoaras, J.

Diabetic nephropathy (DN) is the leading cause of kidney failure in the developed world, but the genetic architecture of DN susceptibility is not well characterised. Here we apply a systems genetics approach in a mouse model of DN to discover novel QTLs for clinically relevant phenotypes including albuminuria, glomerulosclerosis, and macrophage infiltration. For context and prioritisation, we combined single-cell-transcriptomics-guided pQTL and eQTL mapping with cell-type-specific co-expression networks, identifying 192 candidate pGenes for albuminuria. While many were novel, 27% had prior genetic associations, and 40% were validated in a human diabetic cohort. Twelve genes belong to a podocyte network enriched for human GWAS signals. Among those, functional significance of the E3 ubiquitin ligases DCAF6 and ZNRF2 was confirmed by knockdown in Drosophila nephrocytes. Our systems genetics approach identified DN susceptibility genes previously validated in human GWAS while uncovering potential new genes and pathways that could be exploited for risk stratification and therapeutics development.