Sample multiplexing for retinal single-cell RNA-sequencing

Abstract

Single-cell RNA sequencing (scRNA-seq) is a powerful tool for evaluating cell-type specific transcription but can be challenging to apply to rare cell populations. Efficient collection of low abundance cell types can be achieved by pooling biological samples, but this precludes resolution of sample origin together with phenotypic data. This limitation is particularly problematic in experiments in which biological or technical variation is expected to be high (e.g., disease models which vary in phenotypic severity, high throughput genetic perturbation screens, or human patient samples). One solution is sample multiplexing where each sample is tagged with a unique sequence barcode that is resolved bioinformatically. We have established a scRNA-seq sample multiplexing pipeline for mouse retinal ganglion cells (RGCs) using cholesterol-modified-oligos (CMOs). RGCs are a scarce and highly heterogenous cell type (~47 types) that represent less than 1% of all retinal cells. We found that CMOs labeled RGCs efficiently and did not have a significant impact on cell viability, gene expression, or cluster type representation. Through these assignments, we evaluated differential gene expression and identified transcriptional variance between RGCs of individual retinas both as a whole and within specific RGC types. Overall RGC type distributions and transcriptomic correlations between retinas were remarkably similar. Additionally, sample multiplexing enabled the identification of multiplets and resolution of key biological features such as sex-specific gene expression differences. In conclusion, we found that the addition of CMOs is not detrimental to scRNA-seq analysis and enabled the evaluation of differential gene expression among individual biological samples for a rare cell population. As single-cell transcriptomics are becoming a more widely used approach to research development and disease, sample multiplexing represents a useful method to enhance the precision of scRNA-seq analysis.

Comprehensive single-cell atlas of the mouse retina

“…, 13 and Ma et al. 15 studies. Examination of the marker genes Calb1 and Calb2 demonstrated that C39 is the ventral selective type with high expression of Calb2 , and C31 is the dorsal selective type with Calb1 expression.…”

“…study. 15 Examination of 18_Novel markers Pcdh20 and 4833424E24Rik revealed increased expression of both markers in 18_Novel_b, yet Pcdh20 expression was absent in 18_Novel_a ( Figure S9 E). The DEG analysis further demonstrated many genes selectively expressed in these two clusters, suggesting they are transcriptionally similar but independent types ( Figure S9 D).…”

“…, 13 and Ma et al . 15 The integration of the collected data identified all 45 previously identified RGC types ( Figure 5 A and STAR Methods ). Examination of known cell type markers in the integrated data with the final annotation showed proper expression profiles in corresponding types 9 , 18 , 30 ( Figure 5 B).…”

“…The mouse retina provides an important model for the study of neurobiology, with more than 130 distinct cell types characterized through previous scRNA-seq studies. 7 , 9 , 11 , 12 , 13 , 14 , 15 However, the scRNA-seq datasets have been generated separately for BC 11 AC, 12 and RGC, 9 , 13 , 15 with the largest dataset containing just under 36,000 cells, making it difficult to use in aggregate. Though most of these datasets are independently browsable on the Broad Single Cell Portal 16 and accessible through separate databases such as the Gene Expression Omnibus (GEO) repository, it can be challenging to assess gene expression patterns across all retinal cell types.…”

“…Ensuring these atlases define a complete set of retinal cell types remains a major challenge that can only be addressed by powering studies to sufficiently profile the rarest retinal cell types. Here, we generated scRNA-seq data of over 189,000 cells in the mouse retina to complement 141,000 cells from six publicly available scRNA-seq datasets, 9 , 11 , 12 , 13 , 14 , 15 creating a unified cell atlas of the wild-type mouse retina containing over 330,000 cells. Our integrated analysis presents a comprehensive characterization of all major cell classes in the retina, including non-neuronal types, as well as a consensus cell type annotation of BCs, ACs, and RGCs.…”