ASN's Mission

ASN leads the fight to prevent, treat, and cure kidney diseases throughout the world by educating health professionals and scientists, advancing research and innovation, communicating new knowledge, and advocating for the highest quality care for patients.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: FR-OR11

Deep Learning Uncovers Clinical Subphenotypes of Diabetic Kidney Disease Driven by Genetic Variation in Rac1 Pathway

Session Information

Category: Diabetic Kidney Disease

  • 601 Diabetic Kidney Disease: Basic

Authors

  • Paranjpe, Ishan, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Wang, Xuan, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Anandakrishnan, Nanditha, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Zhou, Weibin, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Azeloglu, Evren U., Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • He, John Cijiang, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Nadkarni, Girish N., Icahn School of Medicine at Mount Sinai, New York, New York, United States
Background

Although diabetic kidney disease (DKD) is a leading cause of end stage renal disease, therapeutic development targeting causal pathways has been limited by disease heterogeneity. Integration of clinical data and genomics may uncover hidden DKD subphenotypes.

Methods

DKD patients from the Mount Sinai BioMe Biobank were included. Using laboratory measurements, vitals, and clinical notes in a deep leaning framework (Fig 1 A,B) we performed unsupervised clustering, accounting for population structure. We then performed a genome wide association study comparing patients in each cluster with healthy controls.

Results

We identified two clusters (Fig 1C), M (mild, N = 972) and S (severe, N = 390). Cluster M had greater ESKD prevalence (16% vs 5%; p <0.001) and higher baseline serum creatinine (1.2 vs 1.1; p<0.001). Using exome sequencing, a missense variant in ARHGEF18, rs117824875, was significantly associated with DKD in cluster S, but not cluster M (OR = 7.7; p = 9.56x10-8). This variant was also associated with DKD in an external cohort, UK Biobank (OR = 2.4, p =0.044).

ARHGEF18 knockdown in a diabetic zebrafish model induced whole body edema (Fig 1D). Stable overexpression of the rs117824875 mutant ARHGEF18 transcript in a human podocyte cell line led to decreased cell viability (Fig 1E), actin cytoskeleton reorganization (Fig 1F) and induced RhoA and Rac1 activation (Fig 1G). Mutant ARHGEF18 transcripts exhibited slower ubiquitin mediated degradation (Fig 1H).

Conclusion

Integration of electronic health records with exome sequencing using deep learning uncovered DKD heterogeneity driven by a gain of function variant in ARHGEF18. ARHGEF18 knockdown caused kidney failure in a zebrafish model. Mutant ARHGEF18 was resistant to degradation and activated the Rac1 pathway, suggesting pharmacological inhibition of ARHGEF18 may be a therapeutic target by preventing Rac1mediated podocyte damage (Fig 1I).

Figure 1