Abstract: SA-PO129

Phenotype Agglomeration Analysis of Clinical and Histological End Points of Diabetic Kidney Disease Defines Modules with Improved Transcriptional Associations

Session Information

Category: Diabetes

  • 503 Diabetes Mellitus and Obesity: Translational

Authors

  • Guarnieri, Paolo, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
  • Najafian, Behzad, University of Washington, Seattle, Washington, United States
  • Mauer, Michael, University of Minnesota, Minneapolis, Minnesota, United States
  • Kretzler, Matthias, University of Michigan, Ann Arbor, Michigan, United States
  • Hill, Jonathan, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
  • Nair, Viji, University of Michigan, Ann Arbor, Michigan, United States
  • Harder, Jennifer L., University of Michigan, Ann Arbor, Michigan, United States
  • Wang, Junke, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
  • Hawkins, Julie, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
  • Pullen, Steven S., Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
  • Nelson, Robert G., National Institutes of Health, Phoenix, Arizona, United States
  • Boustany, Carine, Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut, United States
Background

Diabetic kidney disease (DKD) is the primary cause of end stage renal disease worldwide. While the structural and functional determinants of DKD have been studied extensively in various populations, comprehensive aggregation of longitudinal phenotypic measures with molecular mechanisms is lacking.

Methods

We collected an extensive set of clinical and histological endpoints from a cohort of 70 Pima Indians with type 2 diabetes and DKD. Patients underwent annual research examinations that included measurement of glomerular filtration rate (GFR, iothalamate) and had 2 research kidney biopsies performed 10 years apart. We developed a method in which missing values were interpolated with the mean, normalized and log transformed, if applicable. For inter-related traits only the most representative measurement was selected. The remaining traits were then clustered across samples balancing silhouette coefficients with minimum cluster size. A single composite trait – eigentrait – was then generated by averaging its components. Weighted gene co-expression network analysis was used to generate modules of genes with highly correlated expression, associated with either measured traits or eigentraits. Enriched biological functions were determined using Ingenuity Pathway Analysis on composite genes of significantly associated modules.

Results

We identified 7 eigentraits defining modules of mixed clinical and histological endpoints. A highly robust module included the slope of GFR, glomerular basement membrane thickening and mesangial expansion, and was associated in the first biopsy with immune cell migration while in the second biopsy with extracellular matrix synthesis, thereby supporting a role for inflammation in the initiation of DKD.

Conclusion

Phenotype agglomeration analysis provided a means of distilling biologically meaningful signals from a complex collection of traits in Pima Indians with DKD.

Funding

  • Commercial Support