Abstract: FR-PO368
Heritability Enrichment Analyses in Kidney Function Genome-Wide Association Study Identifies Trait-Specific Kidney Cell Types
Session Information
- CKD: Mechanisms - II
November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: CKD (Non-Dialysis)
- 2103 CKD (Non-Dialysis): Mechanisms
Authors
- Wuttke, Matthias, Institute of Genetic Epidemiology, Freiburg, Germany
- Li, Yong, Institute of Genetic Epidemiology, Freiburg, Germany
- Kottgen, Anna, Institute of Genetic Epidemiology, Freiburg, Germany
Group or Team Name
- CKDGen Consortium
Background
Identifying relevant tissues and cell types underlying kidney function and disease informs experimental follow-up studies to understand disease biology. Novel statistical methods allow for unbiased identification of trait-relevant cell types by incorporating RNA-seq data with GWAS summary statistics.
Methods
We used LD score regression for specifically expressed genes (LDSC-seg) to partition heritability in GWAS summary statistics of European ancestry participants from the CKDGen Consortium of estimated glomerular filtration rate (eGFR, n=567,460), urinary albumin-to-creatinine ratio (UACR, n=547,361), blood urea nitrogen (BUN, n=243,031) and serum urate (n=288,666). GWAS of asthma (UK Biobank, n=452,264), and schizophrenia (CLOZUK+PGC Consortia, n=105,318) were used as negative controls. Publicly available kidney single-cell RNA-seq datasets (human, 24 cell types; mouse, 16 cell types) were used to construct the top 10% specifically expressed genes per cell type followed by testing heritability enrichment in each trait. For examination at tissue level, the same procedure was applied using GTEx V7 data.
Results
Across tissues, we found significant enrichment of heritability in trait-associated loci containing genes that are highly expressed in kidney (eGFR: 2.2-fold enrichment, p=9.1e-8; urate: 2.1-fold enrichment, p=1.2e-5); liver was also enriched. Within the kidney, enrichment was observed in regions containing genes specifically expressed in proximal tubule cells in human (eGFR 2.3-fold, p=8.5e-5; BUN 1.7-fold, p=0.005; urate 2.3-fold, p=7.8e-6) and mice (eGFR 2.3 fold, p=0.0003; BUN 1.8-fold, p=0.02; urate 2.3-fold, p=0.0002), as well as in human podocytes (UACR 1.7-fold, p=0.009). Both asthma and schizophrenia did not show significant enrichment of heritability in regions with genes that are highly expressed in kidney cell types, but instead in brain tissues (schizophrenia, smallest p=9.8e-16).
Conclusion
GWAS signals of kidney function traits are enriched for genes that are highly expressed in relevant tissues and cell types such as proximal tubular cells for eGFR, BUN and urate, and glomerular cells for UACR. These results allow for identifying relevant cell types for experimental research to translate GWAS loci into a mechanistic understanding.