Transcriptomics analysis of Muller glia in response to different damage paradigms in the adult zebrafish retina

Abstract

Purpose : In the damaged adult zebrafish retina, Müller glia re-enter the cell cycle to produce neuronal progenitor cells that differentiate into the lost retinal neurons. Different damage models, such as constant light and ouabain, were used to kill different neuronal types. Additionally, coinjecting TNFa and RO4929097, a gamma-secretase inhibitor, stimulated Muller glia reprogramming. It is unknown if the different damage models induce similar or unique gene expression changes. Therefore, we examined transcriptional changes in Muller glia in response to these different models.

Methods : Albino; Tg[gfap:GFP] zebrafish were either light damaged (0, 10, 20, and 36 hours), treated with ouabain (36, 48, 72 hours), or coinjected with TNFa and RO4929097 (36, 48, 72 hours). Retinas were dissociated and FACS-sorted, with both the GFP-positive (Muller glia) and GFP-negative (retinal neurons) fractions being analyzed by RNAseq. Several of these timepoints in all three treatments were also analyzed by single-cell RNAseq. Bioinformatic analyses were performed to characterize transcriptional similarities and differences.

Results : RNAseq identified several genes that were known to be differentially expressed in Muller glia in the damaged retina, such as ascl1a (up-regulated) and notch signaling components (down-regulated). Single-cell RNAseq identified the reprogrammed Muller glia relative to the control Muller glia. This analysis also yielded several genes that were known to change in expression in reprogrammed Muller glia, as well as several new candidates that potentially regulate Muller glia reprogramming. Pseudotime analysis was performed on the Muller glia data from the single-cell RNAseq analysis to examine Muller glia progression from the control basal state to the reprogrammed state. Data from the pseudotime analysis will then be compared between the two different damage models and the TNFa and RO4929097 coinjected retinas to identify genes that are similarly regulated in all three models and those that are differentially regulated between the three models.

Conclusions : RNAseq analysis identified several genes that are differentially expressed between three different damage models. Combining single-cell RNAseq and pseudotime analysis revealed several candidate genes that are involved during Muller glia reprogramming in the different models.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.