An optimized protocol for generating labeled and transplantable photoreceptor precursors from human embryonic stem cells

Cell replacement therapy is a promising approach for treatment of retinal degenerative diseases. Several protocols for the generation of photoreceptor precursors (PRP) from human embryonic stem cells (hESC) have been reported with variable efficiency. Herein, we show the advantages of use of size-controlled embryoid bodies in the ESC differentiation process using two differentiation protocols. We further explored cell-labeling methods for following the survival of PRP transplanted subretinally in rat eyes. Size-controlled embryoid bodies (EBs) generated using microwell dishes and non-size-controlled EBs generated using V-shaped 96-well plates were differentiated into PRP using two differentiation protocols. The differentiation protocols utilized two different combinations of growth factors. The first, Dkk1, Noggin, and IGF1, and the second protocol used IWR1e, SAG, and CHIR99021. Differentiation efficiency to PRP was analyzed by qPCR, immunocytochemistry, and fluorescence-assisted cell sorting (FACS). Size-controlled IWR1e yielded a significantly higher percent (86.4%) of PRP cells expressing CRX, compared with non-size-controlled IWR1e (51.4%, P = 0.026) or the size-controlled DKK1 protocol (70.5%, p = 0.007). In addition, the IWR1e differentiated cells exhibited a significantly higher fluorescence intensity of CRX immunostaining, compared with the DKK1 protocol, consistent with higher protein expression levels. The IWR1e cells exhibited higher maturation levels, as manifested by lower early neuronal marker PAX6 and pluripotency marker OCT4 levels compared with the DKK1 protocol. The expression of other late photoreceptor markers (NRL, recoverin) were similar among the differentiation groups. PRP cells were labeled by using hESC constitutively expressing EGFP or by AAV-GFP transduction. Finally, we transplanted the cells in the subretinal space of wild-type rats and monitored their survival over several weeks. The AAV2 serotype efficiently transduced the PRP cells, whereas other serotypes yielded low or no transduction. Following subretinal transplantation of GFP-labeled PRP, 63% of the cells were detected at 4 weeks post-transplantation. In conclusion, we show here that the IWR1e protocol using size-controlled EBs efficiently generated of PRP that could be labeled and followed in-vivo for weeks. The data from this study is an advance toward the goal of PRP transplantation therapy for retinal degenerative diseases.