The retina converts visual information into neural signals that are processed and transmitted to the brain. During retinal development, seven major cell types, six neuronal and one glial, are generated from a common neuroepithelium during discrete but overlapping time periods. Here, we study the retinal expression and function of three basic helix-loop-helix (bHLH) transcription factors: Atoh7/Math5 (atonal homologue 7), Neurog2/Neurogenin2, and Ascl1/Mash1 (achaete-scute complex like 1). Proneural bHLH transcription factors are critical for neuronal differentiation and cell type specification in the retina. Atoh7 and Neurog2 are expressed at the initiation of retinal development, and Atoh7 is critical for the generation of the first-born cell type, retinal ganglion cells (RGCs), which transmit visual information to the brain via the optic nerve. Ascl1 is expressed later in retinogenesis, and is required for normal bipolar interneuron and Müller glial genesis, two later-born cell types.
First, I explored the regulation of Atoh7 expression, using GFP-expressing transgenes under control of Atoh7 regulatory DNA, which expressed GFP in Atoh7-expressing progenitor cells and nascent RGC axons as they sent projections into the optic nerve and established connections with the brain. In addition to the visual system, Math5-GFP transgenic expression was observed ectopically in developing auditory and proprioceptive systems in the developing brain, spinal cord, and inner ear that normally express Atoh1/Math1, the other atonal semi-orthologue. I found similarities in the genetic regulation of the proximal 2.1 Kb of 5’ Atoh7 DNA and the Atoh1 3’ enhancers, and concluded that these highly-related bHLHs share common regulatory features that, during evolution from a common precursor, were restricted to nonoverlapping expression domains by as of yet unknown DNA repressor elements.
Second, I examined the function of Neurog2 at the initiation of retinal neurogenesis. Neurog2 and Atoh7 expression was observed sequentially in progenitor cells that give rise to the first neurons in the central retina. I determined that Neurog2, but not Atoh7, is essential for the peripheral expansion of neurogenesis and RGC genesis. In Neurog2 mutant mice, neurogenesis was delayed until the onset of retinal Ascl1/Mash1 expression, but by birth the proportions of early-born cell types are returned to normal. Ascl1 replacement of Neurog2 rescued the delay in both neural differentiation and RGC genesis, signifying that retinal development proceeds as overlapping waves of neurogenesis regulated by these bHLH factors.
Finally, I further explored the interchangeability of bHLH transcription factors. To test the hypothesis that Ascl1 and Atoh7 have distinct functions in cell cycle exit and fate specification in retinal progenitor cells, I used a previously constructed mouse model, the Atoh7Ascl1KI allele, which misexpresses Ascl1 in Atoh7-lineage cells. Ascl1 replacement of Atoh7 did not rescue RGC development but increased bipolar interneuron and decreased Müller glia number in adult eyes. During the initiation of neurogenesis, ectopic Ascl1 prolonged proliferation of Atoh7-expressing cells that normally exit the cell cycle, dominant to endogenous Atoh7 function.
In sum, this thesis provokes new mechanisms for the divergence of bHLH regulation and function in mouse retinal development. Neurog2 and Atoh7 have separate roles in early retinal progenitor cells during the initiation of neurogenesis. While Ascl1 can compensate for neural differentiation defects in Neurog2 mutant mice, it does not promote cell cycle exit or rescue RGC specification in terminally mitotic Atoh7-lineage cells. Together, bHLH factors have overlapping and distinct functions in the mammalian retina, defined by a combination of evolutionary homology, phase of cell cycle expression, and developmental timing.