Single-cell transcriptome comparison of myeloid cells' in the retina in response to transplantation of human stem-cell derived neurons and to ocular damage reveals the reversibility of the microglia activation.

Retinal ganglion cells (RGCs) play a critical role in the transmission of visual signals from the retina to the brain required for proper vision. The optic neuropathies, including glaucoma, NAION, LHON, traumatic optic neuropathy, optic pathway glioma result in the reversible and irreversible changes in RGCs, ultimately leading to their death. RGC loss is permanent given that these cells cannot regenerate in mammals. Neuroregenerative strategies for managing optic neuropathies have mainly focused on restoring the lost RGC cell population in a glaucomatous retina. This treatment approach not only aims to halt disease progression, but also to restore the lost vision by replacing the damaged/lost RGCs through RGC transplantation. Significant strides have been made in this field of RGC transplantation and regeneration. As a first step of RGC transplantation, several protocols now exist to differentiate cells into primary RGCs, and several sources of donor RGC have been established including human ES and iPSC-derived RGCs. To achieve functional regeneration, the transplanted RGCs must survive, integrate, and extend their axons and establish connections within the retina and the brain. Of the many factors influencing initial donor RGCs survival, the innate immunity [ref] and primarily microglia reactivity is of particular interest. In ocular models of glaucoma, microglia have been heavily implicated in driving disease progression, promoting a pro-inflammatory environment. In some instances, activated microglia have been directly implicated in RGC loss by mediating phagocytosis of the damaged or injured RGCs. In line with this, our recent work, shows that activated host microglia are detrimental to donor RGCs, presumably via phagocytosis of the donor RGCs. We observed that pre-treatment of donor RGCs with FasL and annexin V to block the exposed phosphatidylserine residues of the stressed donor RGCs prior to transplantation reduced activation of host microglia, and better RGC survival rates post transplantation. Studies like these and others underscore the importance of modulating host microglia reactivity for better transplantation outcomes. To further improve donor RGC survival post transplantation, a better understanding of the host myeloid cell reactivity following donor RGC transplantation is paramount. We and others have previously shown the importance of modulating the retinal microenvironment to promote RGC survival. In the current study, we profiled the transcriptome of the host myeloid cell population (CX3CR1-GFP) using a single cell RNA sequencing approach, to delineate the myeloid cell population most relevant for donor RGC survival post transplantation. A focused analysis on the microglia cell population was performed, given their established role in glaucoma disease progression. In addition to profiling the expression pattern of the known classic microglia disease associated signatures such as ApoE, Spp1, and Lgals, we identified several other genes associated with, and or driving microglia activation, including Cybb, Atf3, Aurka and Kif22. Given the known function of these genes in other disease contexts, we believe that these genes are important for modulating donor RGC survival post transplantation, thus warrant further investigation in this context. Taken together, with reference to our established adult mouse retina cell atlas, we present a comprehensive analysis of the host myeloid cell population, specifically focusing on microglia reactivity status, at a single cell resolution level, providing insights into the regulatory mechanisms that may be important for donor RGC survival and integration into glaucomatous retina post transplantation. Overall design: Three days following RGC transplantation, mice were euthanized, eyes enucleated and retinal tissue isolated and transferred into a tube containing PBS with 0.04% BSA on ice. Eight mouse retinas were pooled for each sample. The tissue was dissociated by mechanical and enzymatic methods using Gentle MACS protocols. Briefly, retinas were allowed to settle to the bottom of the tube and excess PBS carefully removed. Tissues were then rinsed with HBSS and transferred to a MACS M-tube and let to settle to the bottom of the tube before again removing excess HBSS. Tissues were then resuspended in 2ml of dissociation solution containing DNAse 1 (1mg/ml). The mixture was incubated at 37°C for 5min in a heating bath. The M-tube was then inverted ensuring that all retina tissues remained in solution, and the tube placed into the MACS tissue dissociator and the brain 1 program run. At the end of brain 1 program, the M-tube was removed from the MACS dissociator and incubated at 37°C for 5mins in a heating bath. The tube was then inverted again and placed back into the MACS dissociator and program brain 2 initiated, after which 10mL of isolation buffer was added to the dissociate and the mixture triturated 7-10 times with a 1000µl pipette tip. Then 12ml of the homogenate was transferred into a 15ml tube and cells pelleted at 200g for 5 mins at 4°C. The cell pellet was resuspended in 500µl of isolation buffer and filtered using a 70um cell strainer into a FACS tube for sortingAll tubes for this procedure were pre-blocked with 2% BSA in HBSS. Separate single stain samples for TdTomato+ and GFP+ were prepared and used to set the gates. The harvested cells were sorted separately for TdTomato +ve, and GFP +ve, and collected into separate collection tubes each containing 200µl of PBS with 0.04% BSA. The sorted cells were taken for single cell RNA sequencing library preparation. Of the sorted cells, up to 10,000 viable cells per sample were used for single cell RNA library preparation following the 10x genomics Chromium Next GEM Single Cell 3' Reagent Kits v3.1 protocol. Briefly, the droplet-based encapsulation method was used to encapsulate single cells with gel beads, partitioning oil, and a reverse transcription master mix generating GEMs. The generated GEMs were cleaned, gene expression libraries constructed and sequenced at a sequencing depth of 20,000 read pairs per cell, paired-end dual indexing (See manufactures instructions for detail).