Abstract
Purpose :
The currently most comprehensive genome-wide association study (GWAS) for age-related macular degeneration (AMD) identified disease associations for 52 independent genetic variants at 34 genomic loci. Collectively, the AMD-associated loci are enriched for genes contributing to a number of shared cellular processes. Of note, most current approaches focus on single loci and their influence on AMD pathogenesis without considering epistatic interactions. Here, we present a novel approach to investigate combined effects of apparently independent GWAS variants associated with AMD.
Methods :
The influence of genetic variants on genome-wide gene expression was analyzed in harmonized post-mortem liver tissue (n = 588; Strunz et al. Sci Rep. 8:5865, 2018). Multiple combinations of AMD-associated genetic variants were used to generate genetic risk scores (GRS) and gene expression in liver was compared between the respective low and high-risk groups.
Results :
We determined 26 genes (eGenes) with a significantly altered expression in high-risk vs. low-risk combinations. Seven of these, namely LIPC, CFHR1, CFHR4, CFHR3, PILRA, PILRB, and TSPAN10 were previously reported to be regulated by single AMD-associated variants in expression quantitative trait locus (eQTL) studies. Nineteen eGenes represent exciting new candidates for AMD etiology. Among the novel candidates, BRCA1 and ASNS show the highest effect sizes (ESs), whereby a high genetic risk for AMD is correlated with a downregulation of BRCA1 (ES = -1.18, SE = 0.21, Q-Value = 1.55 x 10-7), and an upregulation of ASNS (ES = 1.17, SE = 0.20, Q-Value = 9.36 x 10-8).
Conclusions :
Here we present a novel method combining GRS and eQTL mapping to investigate joint effects of seemingly independent genetic variants on gene expression. We replicated earlier eQTL findings, and report 19 novel genes to correlate with the genetic risk to develop AMD. All genes were exclusively identified by jointly analyzing several AMD-associated variants, which is in-line with the theory that the signals underlying GWAS associations contribute to shared biological mechanisms.
This is a 2021 ARVO Annual Meeting abstract.