The mammalian retina contains over sixty diverse cell types, which are grouped into seven major cell classes: rod photoreceptors, cone photoreceptors, bipolar interneurons, amacrine interneurons, horizontal interneurons, retinal ganglion cells (RGCs), and Muller glia. During development, all cell classes arise from a common pool of retinal progenitor cells (RPCs), in a highly conserved and partially overlapping sequence. RPCs simultaneously regulate the growth of the retina by undergoing continuous rounds of cell division. Therefore, a complex network of extrinsic signals and intrinsic transcription factors is required to maintain the balance between cell cycle progression and neuronal (or glial) differentiation.
Proneural basic helix-loop-helix (bHLH) factors coordinate multiple elements of retinal neurogenesis. Although Notch regulation of bHLH genes is an evolutionarily conserved module, the tissue-specific mechanisms are incompletely defined. In the developing mouse retina, Atoh7 regulates RGC competence and Neurog2 is required for the progression of neurogenesis. These two transcription factors are extensively coexpressed in RPCs, thereby suggesting a similar mode of regulation. In Chapter 2, we directly compared Atoh7 and Neurog2 expression at the earliest stages of retinal neurogenesis, in a broad spectrum of Notch pathway mutants. Here, a Notch1, Rbpj, and Hes1-mediated signal represses Atoh7. Yet, the combined activities of Notch1, Notch3, and Rbpj regulate Neurog2 patterning, independent of Hes1, Hes3, or Hes5. Finally, we tested Notch regulation of Jag1 and Pax6, to establish the proper context for distal Neurog2 patterning. We found that Rbpj blocks coexpression of Jag1 and Neurog2 most-distally, while stimulating Pax6 in an adjacent domain. Together, our data suggest that Notch signaling controls the overall tempo of retinal neurogenesis, by integrating cell fate specification with the developmental status of cells ahead of this wave.
Although Neurog2 is required for the overall temporal progression of neurogenesis, the molecular mechanism behind this regulation remains elusive. Additionally, it was previously shown that ectopic Ascl1, another bHLH, could rescue the delay of RGC differentiation seen in Neurog2 mutants. Combined, this suggests that Neurog2 and Ascl1 regulate similar processes, with respect to neuronal wave propagation. In Chapter 3, we directly compared the average RPC cell cycle length between Neurog2GFP/+, Neurog2GFP/GFP, and Neurog2GFP/Ascl1 retinas. Here, we found that Neurog2 is required for normal cell cycle exit, whereby ectopic Ascl1 only partially rescues the mutant defects. RNA-seq analysis for all three genotypes demonstrated that genes associated with RGC differentiation and cell cycle progression were significantly downregulated in Neurog2 mutants, compared to controls. Interestingly, ectopic Ascl1 only rescued one RGC gene, at the expense of all others. Combined, this data implies that although ectopic Ascl1 restores the wave of neurogenesis, it does not completely substitute for Neurog2 at the same molecular level. Therefore, each factor must target a unique subset of downstream genes, in addition to common genes involved in general neurogenesis.