Abstract: PO1628
Identification of Genetic Drivers of Age-Related Renal Histopathology
Session Information
- Genetic Diseases of the Kidneys: Non-Cystic - 1
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Genetic Diseases of the Kidneys
- 1002 Genetic Diseases of the Kidneys: Non-Cystic
Authors
- Sheehan, Susan Marie, The Jackson Laboratory, Bar Harbor, Maine, United States
- Korstanje, Ron, The Jackson Laboratory, Bar Harbor, Maine, United States
Background
Studies to understand age-related changes in the human kidney have been performed by measuring kidney function and damage markers in the urine. These studies have provided valuable information, including clear genetic components underlying kidney disease. However, due to the highly invasive nature of kidney biopsies, it is not possible to identify early causal changes in humans by histological analyses that are hypothesized to precede changes in function and renal damage. However, mouse models provide access to kidneys at specific time points enabling us to conduct histological analyses across lifespan. We established the Aged Mouse Kidney Resource, which consists of kidneys from 600 genetically diverse mice (males and females) at three ages (6, 12, and 18 months). Scanned PAS slides for all mice are publically available at korstanjelab.jax.org, as well as gene expression, protein expression, and DNA methylation data for a subset of kidneys
Methods
Renal histology has been mostly a qualitative or semi-quantitative discipline. We leverage new approaches in image analysis and machine learning and demonstrate the feasibility of quantification on entire sections of mouse kidneys (pathomics) from a large number of animals. We have developed a pipeline that uses machine learning on scanned slides, which allows us to automatically segment glomeruli and quantify mesangial matrix expansion (MME) in a high-throughput fashion.
Results
Applying our pipeline on the 12-month kidneys from our Resource shows an estimated heritability (h2) of 0.76 for MME and genetic analysis identifies three significant loci with Abca13 and Cfl2 as strong candidate genes for two of these loci. On the other hand, we find that the heritability drops to 0.61 and no significant loci were found in the 18-month old kidneys. We hypothesize that this is caused by the increasing effect of environmental variation with age and death before 18 months of animals with fast age-related renal functional decline that reduces the genetic variability and mapping power in the population.
Conclusion
Our results demonstrate the importance of genetic factors contributing to histological phenotypes and the power of combining pathomics and genetics to identify genes involved in age-related histological changes.
Funding
- Other NIH Support