Gene expression changes during retinal development and rod specification

PURPOSE: Retinitis pigmentosa (RP) typically results from individual mutations in any one of >70 genes that cause rod photoreceptor cells to degenerate prematurely, eventually resulting in blindness. Gene therapies targeting individual RP genes have shown efficacy at clinical trial; however, these therapies require the surviving photoreceptor cells to be viable and functional, and may be economically feasible for only the more commonly mutated genes. An alternative potential treatment strategy, particularly for late stage disease, may involve stem cell transplants into the photoreceptor layer of the retina. Rod progenitors from postnatal mouse retinas can be transplanted and can form photoreceptors in recipient adult retinas; optimal numbers of transplantable cells are obtained from postnatal day 3–5 (P3–5) retinas. These cells can also be expanded in culture; however, this results in the loss of photoreceptor potential. Gene expression differences between postnatal retinas, cultured retinal progenitor cells (RPCs), and rod photoreceptor precursors were investigated to identify gene expression patterns involved in the specification of rod photoreceptors. METHODS: Microarrays were used to investigate differences in gene expression between cultured RPCs that have lost photoreceptor potential, P1 retinas, and fresh P5 retinas that contain significant numbers of transplantable photoreceptors. Additionally, fluorescence-activated cell sorting (FACS) sorted Rho-eGFP-expressing rod photoreceptor precursors were compared with Rho-eGFP-negative cells from the same P5 retinas. Differential expression was confirmed with quantitative polymerase chain reaction (q-PCR). RESULTS: Analysis of the microarray data sets, including the use of t-distributed stochastic neighbor embedding (t-SNE) to identify expression pattern neighbors of key photoreceptor specific genes, resulted in the identification of 636 genes differentially regulated during rod specification. Forty-four of these genes when mutated have previously been found to cause retinal disease. Although gene function in other tissues may be known, the retinal function of approximately 61% of the gene list is as yet undetermined. Many of these genes’ promoters contain binding sites for the key photoreceptor transcription factors Crx and Nr2e3; moreover, the genomic clustering of differentially regulated genes appears to be non-random. CONCLUSIONS: This study aids in understanding gene expression differences between rod photoreceptor progenitors versus cultured RPCs that have lost photoreceptor potential. The results provide insights into rod photoreceptor development and should expedite the development of cell-based treatments for RP. Furthermore, the data set includes a large number of retinopathy genes; less-well-characterized genes within this data set are a resource for those seeking to identify novel retinopathy genes in patients with RP (GEO accession: GSE59201).