Single-cell Multiomics Analysis of Human Retinal Development

Abstract

Purpose : Development of the retina is under precise temporal and spatial regulation. Although it has been observed that the chromatin state changes dynamically during retinal development, how it changes and its relationship with transcriptomics are unclear. Therefore, we aim at applying the latest single-cell multi-omics and spatial transcriptomics tools to gain insights into this complex development process.

Methods : Samples from foveal and peripheral human fetal retina between 10- and 23-weeks gestational age were collected and subjected to single nuclei multiome profiling using the 10X Genomics platform to generate transcriptomic and open chromatin data from the same nuclei. Batch corrected data obtained from all samples were analyzed together to identify cell types and subtypes. Transcription factors (TFs) and gene regulatory networks (GRNs) that correlate with the developmental process were analyzed based on the multiome dataset. Finally, the single-cell spatial developmental maps will be generated using the massively multiplexed single-molecule RNA in situ hybridization (MERFISH) technology.

Results : In total, single nuclei multiome profiling of 280,000 nuclei has been generated from 24 samples of the peripheral and foveal retina. All major cell types and over 30 cell subtypes have been identified in our dataset. Consistent with the idea that cell differentiation starts from the central retina and progresses toward the periphery, delayed development onset of the peripheral retina was clearly observed. Furthermore, based on pseudotime analysis, the trajectory matched well with the established birth order of all cell types. By analyzing cells close to the branch points among trajectory, known and novel TFs, GRNs that correlate with the developmental process were identified. Both concordant and discordant regulations of transcripts and open chromatin were observed, providing insights on different modes of regulatory switch during development. Finally, the generation of a single-cell spatial developmental map of the retina is underway.

Conclusions : Our study produces a spatial-temporal multi-omics map of the human developing retina. Integrative analysis of this multi-omics dataset reveals candidate TFs and GRNs controlling the developmental process. Comparison between transcriptome and open chromatin profiles reveals the complex dynamics of gene regulation between genomic and epigenomics during retinal development.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.