Abstract
Background Non-coding genomic variation can influence tissue- and time-specific gene expression. With the widespread integration of genomic diagnostic strategies, there is increasing opportunity to understand the role of non-coding variants in common and rare ophthalmic disorders. We present a novel resource of paired whole genome (WGS) and transcriptome sequencing of the human adult retina and retinal pigment epithelium (RPE), which uncovers new understanding of the regulatory landscape in these tissues.
Methods WGS and bulk RNA-sequencing of retina and RPE was performed for 201 post-mortem eyes. A linear regression model was subsequently used to detect genomic variants driving differences in gene expression (cis-eQTLs, expression quantitative trait loci) in these tissues.
Results We identified 1,575,451 significant eQTL associations (FDR < 0.05) in the retina and 540,830 in the RPE, of which 479,229 were found in both eye tissues, and 47% are unique to this cohort. Genomic variants associated with observed differences in gene expression (eVariants) were enriched in annotated candidate cis-regulatory elements (cCREs) from different tissues, most significantly in the retina (p-value = 3.6×10-23) and RPE (p-value = 3.9×10-7). However, 84% eVariants did not overlap with any known cCREs in any tissue. Notably, we identified 8,248 novel eQTLs that impact genes with known association to eye disease and observed high variability in allelic frequency of eQTLs across diverse populations with diverse genetic ancestries. eVariants that impact known disease-genes have differences in properties compared to eVariants that impact non-disease genes, including allele frequency and overall impact on gene expression.
Conclusion This unique resource facilitates the discovery of novel retina-specific regulatory mechanisms and provides insight into non-coding variants that regulate known eye disease genes. Additionally, our findings suggest differences in the cis-regulation of genes associated with eye disorders compared to non-disease genes.
