Abstract
Purpose :
Single-cell RNA sequencing (scRNA-seq) is a powerful tool for studying the complexity of the transcriptome, but addressing sample variance can be a challenge. Examination of rare and highly heterogenous cell populations, such as retinal ganglion cells (RGCs), often requires pooling multiple samples for experiments. However, samples may have high technical or natural variance, as may be the case in gene perturbation studies or human tissues, respectively. Sample identities from pooled cells are not distinguishable by standard scRNA-seq and thus phenotypic differences may be masked by batch effects. While sample variance may not be avoidable, it is resolvable by sample-based multiplexing. To demonstrate this, we performed scRNA-seq on mouse RGCs with retina-multiplexing using cholesterol-modified oligos (CMOs).
Methods :
We generated congenic C57bl6/j Vglut2-cre;Ai9 reporter mice to label RGCs. Single cells were obtained by enzymatic dissociation of whole retinas. CMOs were used to tag cells from each retina and RGCs were purified by fluorescence-activated cell sorting (FACS) using the CD90.2 antibody. Demultiplexing was done using the R package HTOdemux.
Results :
We profiled 43,108 CMO-labeled RGCs from 35 retinas. The inclusion of CMOs did not affect transcriptomic quality, aligned with previously published RGC scRNA-seq data, and improved resolution of two transcriptionally-similar clusters. CMOs enabled attribution of individual cells to a specific retinas (~85% efficiency), allowing detection of multiplets, and identification of retinas with poor representation. RGC type frequencies were consistent across retinas. We identified differentially expressed genes between RGCs from male vs female but not between left vs right retinas. Sex-specific differences were restricted to genes on sex chromosomes.
Conclusions :
In the current scope of research, scRNA-seq is increasingly used to evaluate transcriptomic shifts in disease and development. As variability is often unavoidable, we conclude that multiplexing has the capability to resolve sample variance and may enhance the precision of scRNA-seq analyses. Our dataset demonstrates that, aside from select sex-specific differences, age-matched congenic adult mouse RGCs exhibit little natural variance in transcriptional profile or cell type-frequencies.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.