Abstract
Purpose :
Genome-wide association studies (GWAS) have identified nearly 200 genetic variants linked to age-related macular degeneration (AMD). These discoveries have motivated clinical trials for AMD such as the use of complement inhibitors. However, the function and relevance of many AMD risk variants are poorly understood, especially if they reside in noncoding regions of the genome. To better interpret AMD risk variants, we report a combined single-cell RNA sequencing (scRNA-seq) and single-cell assay for transposase-accessible chromatin sequencing (scATAC-seq) atlas of adult human retinal pigment epithelium (RPE) and choroid.
Methods :
RPE and choroid tissue were dissected from eyes of postmortem donors with no known ocular disease (9 eyes from 5 donors) or with a history of non-exudative AMD (3 eyes from 2 donors). Tissues were flash-frozen in liquid nitrogen, and frozen tissues were homogenized in a detergent buffer and isolated from density gradients to obtain suspensions of individual nuclei. Nuclei were processed using the 10x Genomics Single Cell Multiome ATAC + Gene Expression kit. Libraries were sequenced on an Illumina NextSeq or NovaSeq. scRNA-seq and scATAC-seq data were analyzed in R using Seurat and ArchR packages.
Results :
>25,000 cells from human RPE and choroid passed quality-control filtering and were assigned based on marker genes to 8 cell types: RPE, fibroblasts, melanocytes, macrophages, vascular endothelium, T cells, Schwann cells, and B cells. scATAC-seq of these cell types identified >250,000 unique chromatin accessibility peaks. Stratification of scRNA-seq data by donor history revealed up- and down-regulated transcripts in each of the above 8 cell types during AMD. Intersection of chromatin accessibility peaks with the locations of variants from AMD risk loci nominated target cell types for specific variants.
Conclusions :
Single-cell multiomics of human RPE and choroid enable cell-type specific comparisons of gene expression and chromatin accessibility in health or during AMD. Integration of this data with GWAS may nominate cell and gene targets for specific AMD risk variants and help prioritize functional investigations.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.