Abstract: SA-PO0300
Systems Genetics Identifies a Podocyte Activity Network in Diabetic Nephropathy
Session Information
- Diabetic Kidney Disease: Basic and Translational Science Advances - 2
November 08, 2025 | Location: Exhibit Hall, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- 701 Diabetic Kidney Disease: Basic
Authors
- Mishra, Kunal, Duke-NUS Medical School, Singapore, Singapore
- Sakban, Rashidah, Duke-NUS Medical School, Singapore, Singapore
- Guo, Jing, Duke-NUS Medical School, Singapore, Singapore
- Petretto, Enrico, Duke-NUS Medical School, Singapore, Singapore
- Gurley, Susan B., University of Southern California Keck School of Medicine, Los Angeles, California, United States
- Coffman, Thomas M., Duke-NUS Medical School, Singapore, Singapore
- Behmoaras, Jacques, Duke-NUS Medical School, Singapore, Singapore
Group or Team Name
- Diabetes studY in Nephropathy And other Microvascular cOmplications (DYNAMO).
Background
Diabetic nephropathy (DN) is a common cause of kidney failure affecting millions. Only 30-50% of people with diabetes eventually develop DN, and familial clustering and heritability of DN risk have been long recognised. However, genetic mechanisms determining risk for DN have been difficult to identify. To address this problem, we applied systems genetics in a mouse model closely mimicking human DN.
Methods
We previously showed strain-dependent differences in DN susceptibility of Akita (A) mice with T1DM bearing a renin transgene (R). To study genetic susceptibility to DN, we carried out an F2 cross between (AR) mice from DN-susceptible (129/SvEv) and resistant (C57BL/6) strains. 285 male F2 mice were bred and phenotyped by determining (i) albuminuria, (ii) kidney pathology scores and (iii) renal macrophage infiltration. We also conducted (iv) whole-genome sequencing and (v) bulk-RNA sequencing on whole kidneys to map phenotypic quantitative trait loci (pQTL) using linear mixed modelling, and expression QTLs (eQTL) using a Bayesian approach. To build functionally annotated gene co-expression networks specific to kidney cell types, our results were integrated with single-cell RNA-seq data obtained from AR and WT kidneys from parental strains (n=12; 193,728 cells; 43% glomerular).
Results
We found 28 significant pQTLs for albuminuria (p.adj<0.05), 26 of which were novel, spanning 427 genes. 6 pQTLs were also found to be trans-eQTLs controlling 26 genes. Across the albuminuria pQTLs, one locus and 80 genes were also associated with macrophage infiltration, suggesting a mechanistic link between albuminuria and macrophage activity. In the single-cell data from the parental lines, 28% of the 427 genes within QTLs are differentially expressed in podocytes. In parallel, we identify a disease-associated podocyte activity network (PAN) consisting of 331 genes enriched for loci from GWAS of microalbuminuria in DN (hypergeometric P=0.0014). 49% of these genes are downregulated in podocytes from the susceptible 129 strain. 11 PAN genes mapped to our albuminuria QTLs and are prioritized for functional testing.
Conclusion
In a mouse model recapitulating features of human DN, we identify novel QTLs linked to DN susceptibility. Our results suggest that genes regulating podocyte function may be critical determinants of DN pathogenesis.
Funding
- Government Support – Non-U.S.