Abstract
Purpose :
Genome-wide association studies (GWASs) have identified many genetic loci for glaucoma, which provide a starting point for functional follow-up studies. However, prioritization of the underlying causal genes and identification of their functional characteristics is largely lacking. The study aims to: (i) prioritize the most likely ‘causal’ genes, and (ii) identify underlying biological pathways and functional characteristics of candidate primary open-angle glaucoma (POAG) genes.
Methods :
We used the results of the most recent and largest GWAS based on 12,315 cases and 227,987 controls of UKB and GERA cohorts. First, we selected independent single nucleotide polymorphisms (SNPs) associated with POAG and performed gene-prioritization analyses based on: i) in-silico sequencing; ii) Data-driven expression prioritized integration for complex traits (DEPICT); iii) Summary statistics-based Mendelian Randomization (SMR); and iv) MetaXcan. Next, based on the resulting prioritized genes, we performed functional and tissue enrichment analysis. We also estimated the gene expression value of the prioritized genes in 10 ocular tissues.
Results :
In-silico sequencing of 50 independent variants yielded 50 nearest genes and nine non-synonymous linked SNPs (nsSNPs) (r2>0.5) mapping to seven genes. Three of these nsSNPs were found to have deleterious effects. DEPICT prioritized 72 additional genes at a false discovery-corrected (FDR) p-value <0.05. SMR analysis of blood and brain eQTL data, identified 18 genes as the most likely causal genes for POAG (HEIDI test p-value ≥ 2.78x10-3), of which TXNRD2, MVB12B, NR1H3 and LTBP3 were at top hits (p-value ≤7.79x10-8). Furthermore, MetaXcan analysis identified 12 significant genes of which eight showed acceptable evidence of colocalized signals with glaucoma.
With 140 prioritized genes, Gene ontology and Ingenuity pathway enrichment analysis revealed 200 and 64 pathways associated with POAG, respectively. Cardiovascular related pathways, including cardiac muscle cell proliferation (GO:0060038), and blood vessel development (GO:0001568), were among the top hits (p-value ≤ 2.16x10-11, FDR ≤1.14x10-7). Glaucoma genes were overrepresented in optic nerve head, trabecular meshwork, genitalia, heart and blood vessel tissues.
Conclusions :
Our findings may help in understanding the molecular mechanisms underlying glaucoma pathogenesis.
This is a 2020 ARVO Annual Meeting abstract.