The retinal stem cells (RSCs) are defined as those that can self-renew and at the same time generate diverse retinal cell types. RSCs exist both in embryos and adult. During embryogenesis, RSCs produce6types of neurons and1type of glial cells sequentially. In adult vertebrates, especially lower vertebrates like fish and frogs, cells that located in the ciliary marginal zone (CMZ) can produce new neurons in the retina, thereby regarded as RSCs. In higher vertebrates like quail and chick, the CMZ cells produce mainly bipolar and amacrine cells after injury. In primates, RSC-like spheres were grown from the CMZ regions of post-mortem eyes and were shown to generate neuronal cells of the retina.At the present, the identity of RSCs is not clear. Whether RSCs in embryogenesis vs. adults are the same is unknown. A number of eye-field transcription factors, including Pax6, Soxl, Sox2, Otx2, MITF, Chx10, Six3and Rx, have been used to mark RSCs as these factors are present in the earliest stage of eye development in some animals. However, many factors (MITF, CHX10and Rx) are expressed by differentiated retinal cells whereas others (PAX6, SOX1, SOX2, SIX3, and OTX2) are similarly expressed by neuroepithelia. Therefore, it is difficult to distinguish RSCs from neural stem cells, especially in vitro. Even the most specific RSC transcription factor PAX6is universally expressed by neuroepithelia in primates. There is thus a need to identify markers for RSCs, especially in humans. This will in turn allow identification, isolation, and maintenance of RSCs, which will be critical for development of therapeutics for blindness.I first examined the temporal expression pattern of eye-field transcription factors during early human embryogenesis (stage12to16) by immunocytochemistry. I found that while Pax6, Otx2, Six3and Rx are broadly expressed by the neuroectoderm, Sox1/2is absent in the optic vesicle. In the optic vesicle epithelia, but not other neuroectoderm areas, MitF is expressed. In addition, the fluorescent intensity (expression level) of Pax6is higher in the optic vesicle than in other neuroepithelia. Therefore, expression of a high level of Pax6and MITF but not Soxl/2may be used to reliably identify RSCs in the human optic vesicle.I then used this combination of transcription factors (Pax6+/MitF+/Soxl/2-) to identify putative RSCs in the cultures that were differentiated from human embryonic stem cells (hESCs) using our conventional neural differentiation. I found a small number (4%) of Pax6+/MitF+/Sox1/2-cells that were present mainly at day12. These cells disappeared after day15. This result suggests that the appearance of RSCs corresponds to eye field development. It also suggests that the putative RSCs differentiate to retinal cells over time.To increase the differentiation of RSCs, I used a retinal differentiation medium that contains knockout serum replacement and nicotinomide. Under this culture condition, about13%of the differentiated cells were Pax6+/MitF+/Soxl/2-at day12. To further identify molecules that promote RSC generation, I applied sonic hedgehog (SHH), bone morphogenetic protein4(BMP4), FGF8, Wnt3a, DKK, and IGF1which are shown to increase the generation of retina stem/progenitor cells in animals. Unfortunately, I found that none of these factors further promote the generation of RSCs on top of the retinal differentiation medium.Since the putative RSCs are present only for a short period, I asked if these RSCs may be maintained. In DMEM/F12,I added LIF and a small molecular inhibitor of GSK3β, CHIR99021. I found that the putative RSCs were present as late as day24-27.Finally, I intended to confirm the presence of RSCs by treating the day-12cultures with FGF2or activinA to assess if the RSCs may be biased toward neural retina or pigment epithelium. Indeed, in the presence of FGF2, a large population of the cells expressed Chx10whereas in the presence of ActivinA, most of the spheres became pigmented. Therefore, there were RSCs in my cultures and these RSCs may be regulated to generate neural retinal cells or pigment epithelia. |