Poster
Overview
The complexity of the cell diversity, classification and characterization of function in the mammalian visual system cannot be overstated. The process of translating light information through photoreceptors of rods and cone cells and transducing the signal through a complex network of bipolar cells, horizontal cells, amacrine cells and retinal ganglion is both intricate and subtle. There are two challenges that prevent proper characterization and study of the visual system: 1) Cell populations of subtypes in the system are relatively small. 2) Subtype characterization often relies on a few marker genes and can be missed due to poor gene detection. The mammalian visual system represents a pinnacle of challenges for single cell RNA-seq. As a proof-of-concept, 1 million cells isolated from murine eyeballs were processed from two animals (male and female), put through the barcoding workflow, and sequenced. Clustering of the subsequent data reveals a diversity of cell types including (but not limited to) retinal ganglion cells (RGC), amacrine cells, horizontal cells, bipolar cells, astrocytes, and Muller cells. Additional supporting cells such as keratocytes, lens fiber cells, melanocytes, microglia, and retinal endothelial cells are also detected. Further sub-clustering of amacrine cells and RGC reveals even more distinct subtypes (46 RGC subtypes and >35 amacrine subtypes) whose markers are consistent with putative subtypes. With >100 profiled cell types, the number of profiled cells per type represents a very small percentage. Scaling to at least 1 million cells is necessary to detect as many types as possible since some populations represent <0.1% of the eye. Furthermore, high gene
sensitivity allows for detection and sub-clustering of distinctive subtypes. This combination of scaling and sensitivity is needed to achieve a comprehensive and robust atlas. This murine eye atlas is just the beginning of using combinatorial barcoding technology to study the mechanisms of ophthalmologic disease models.