Abstract
Purpose :
As a complex neuronal tissue, the mouse retina is composed of a large number of cell types that can be distinguished based on morphology, function, location, and transcriptomic profile. In addition, distinct mRNA transcript isoforms have been observed among different cell types. However, the extent of mRNA alternative splicing in the mouse retina has not been systematically characterized in a cell type specific context.
Methods :
Single cell RNA sequencing was performed by combining short read and long read high throughput sequencing with droplet based single cell technology. A customized data analysis pipeline is developed to allow accurate cell barcode/unique molecular identifier assignment, high-confidence transcript isoforms identification and characterization in each cell type in the retina.
Results :
The transcriptome from a total of 16,323 single cells from the mouse retina has been profiled with 726 million Illumina short reads and 275 million Oxford Nanopore long reads. Our dataset includes 8,497 bipolar cells (BCs), 5,789 rods, 1,022 cones, 869 Müller glial (MGs) and 146 amacrine cells (ACs). In addition to the cell class level, sufficient cells are captured for each BC type. Based on our preliminary analysis, in total, 235,861 transcript isoforms are identified, and 69.43% of those are novel. Furthermore, over 13% of isoforms are cell class specific, including 19,208 unique isoforms in BC, 9,136 in rod, among others. Finally, isoforms that are differentially expressed among the five major retinal cell classes are identified.
Conclusions :
Our study represents the first comprehensive characterization of the full-length transcription isoforms in single cells in mouse retina with over 160,000 novel isoforms in five retina cell classes identified. These data correspond to a substantial addition to the annotated/known mouse transcriptome from previous studies. We also identified 31,807 cell class specific isoforms. Based on our analysis, the long read scRNA-seq approach is highly conducive to novel isoform identification especially the cell-class specific isoforms and further improves cell sub clustering using splicing information. Further insights gained from deep analysis of this dataset will be reported.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.