Otx2 enhances transdifferentiation of Müller cells‐derived retinal stem cells into photoreceptor‐like cells

Abstract Retinal Müller glial cells have the potential of neurogenic retinal progenitor cells, and could reprogram into retinal‐specific cell types such as photoreceptor cells. How to promote the differentiation of Müller cells into photoreceptor cells represents a promising therapy strategy for retinal degeneration diseases. This study aimed to enhance the transdifferentiation of rat Müller cells‐derived retinal stem cells (MC‐RSCs) into photoreceptor‐like cells and explore the signalling mechanism. We dedifferentiated rat Müller cells into MC‐RSCs which were infected with Otx2 overexpression lentivirus or control. The positive rate of photoreceptor‐like cells among MC‐RSCs treated with Otx2 overexpression lentivirus was significantly higher compared to control. Furthermore, pre‐treatment with Crx siRNA, Nrl siRNA, or GSK‐3 inhibitor SB‐216763 reduced the positive rate of photoreceptor‐like cells among MC‐RSCs treated with Otx2 overexpression lentivirus. Finally, Otx2 induced photoreceptor precursor cells were injected into subretinal space of N‐methyl‐N‐nitrosourea induced rat model of retinal degeneration and partially recovered retinal degeneration in the rats. In conclusion, Otx2 enhances transdifferentiation of MC‐RSCs into photoreceptor‐like cells and this is associated with the inhibition of Wnt signalling. Otx2 is a potential target for gene therapy of retinal degenerative diseases.

ESCs remain highly controversial of ethical issues. Therefore, autologous retinal stem cell therapy has a significant advantage and potential. However, retinal stem cells only exist in the ciliary margin zone (CMZ) and retinal pigmented epithelium (RPE), and are far short of clinical need. 9 Accordingly, it is urgent to develop novel methods to generate substantial photoreceptor cells that can be integrated into the injured retina.
Müller glia cells are the major cell type in the mammalian retina and contribute to the maintenance of retinal homeostasis and trophic support for retinal neurons. 10 In acutely injured retinas, Müller glial cells can proliferate and reprogram into a neural progenitor state, which then differentiate into retinal neurons, but the efficiency is low. The reprogrammed progenitors express transcription factors similar to embryonic retinal progenitors such as Chx10, Pax6, Sox2, Sox9, Six3 and Ascl1a. 11 Previous studies have demonstrated that Müller glia cells proliferate in vitro to form neurosphere which can be differentiated into retinal-specific cell types, including retinal ganglion cells, bipolar neurons and rod photoreceptors. 12,13 However, the induce photoreceptor cells are too few to replace the lost or injured cells. Thus, how to promote the differentiation of Müller cells into photoreceptor cells represents a promising therapy strategy for RDDs.
The differentiation of retinal stem cells involves several transcription factors. Among them, orthodenticle homeobox 2 (Otx2) is a crucial transcription factor for photoreceptor cell specification, 14 it regulates the expression of cone-rod homeobox (Crx) by binding to Crx promoter, 15 and then both of them act on the downstream target gene Nrl and promote the generation of rhodopsin. [16][17][18] Wnt signalling is known to be crucial for the renewal of retinal progenitor cells and the maintenance of steady-state of stem cells. 19 Wnt signalling is negatively regulated by Otx2 in the midbrain dopaminergic neural stem cells. 20 In this study we constructed lentivirus pGC-FU-Otx2-EGFP and

| Immunofluorescence analysis
Rat retinal tissue sections or retinal cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature and permeabilized with PBS containing 3% bovine serum albumin (BSA), 5% goat serum and 0.3% Triton X-100 at 37°C for 1 hour. The samples were incubated overnight at 4°C with the primary antibodies (listed in Table 1). After washing, the samples were incubated in the dark at room temperature for 2 hours with Alexa Fluor-conjugated secondary antibodies (Invitrogen or Molecular Probes). For negative controls PBS was used instead of primary antibody. After immunostaining, the samples were counterstained with DAPI. Six randomly selected fields from each sample were observed and fluorescent signals were captured under fluorescent microscopy (Leica DMI4000B, Solms, Germany).

| qRT-PCR
Total RNA was extracted from retinal cells or retinal tissue performed with Trizol reagent (Ambion, Austin, TX, USA) and treated with RNase-free DNase I. Reverse transcription was performed with the PrimeScript RT kit with gDNA Eraser (Takara, Tokyo, Japan).
PCR was performed with the SYBR Premix Ex Taq (Takara) and the primers listed in Table 2. Glyceraldehyde phosphate dehydrogenase (GAPDH) was an internal standard. Experiments were performed at least in triplicate. Target mRNA levels were quantified using the comparative threshold cycle (Ct) method, fold change = 2 ΔΔCt ,    Table 1. After several washes, the membranes were incubated with HRP-conjugated secondary antibodies (Gibco) at room temperature for 1 hour. The signal was visualized using Pierce ECL kit (Thermo Scientific). The blots were visualized on X-Ray Film and quantified using Image J software (NIH). GAPDH was used as loading control.

| Electroretinograms (ERG)
The rats were dark adapted for 2 hours, anaesthetized with ip, injection of 10% chloral hydrate and the pupils were dilated with 1% tropicamide and treated with a topical anaesthetic (0.5% proparacaine). A silver-impregnated wire loop record electrodes were mounted on the surface of the cornea. A reference electrode was placed behind the ears and a ground electrode was placed under the cheek to detect the electroretinogram of the rats. Hydroxypropyl methylcellulose was applied to maintain corneal hydration. The measurements followed the International Society for Clinical Electrophysiology of Vision (ISCEV) standardization protocol.

| Statistical analysis
All data were presented as mean ± SD. Comparison between groups was assessed by unpaired Student's t-test, and the differences among multiple groups were analysed by analysis of variance

RSCs into photoreceptor-like cells
We obtained and characterized MC-RSCs using our previously established protocols. 21  Fluorescence microscopy showed that photoreceptor-like cells were stained as red based on photoreceptor cell-specific marker Rhodopsin. The photoreceptor-like cells in pGC-FU-Otx2-EGFP group appeared on day 4, reached the peak on day 7 and then gradually reduced. They were elongated, the cytoplasm was rich, the axons were long and the nuclei were located in the centre of the cell body ( Figure 1D). At 7 days after differentiation, the positive rate of Rhodopsin in pGC-FU-Otx2-EGFP group (42.33 ± 8.96%) was significantly higher than that of pGC-FU-EGFP group (7.32 ± 1.67%) and blank group (7.01 ± 2.09%) and there was no significant difference between blank group and pGC-FU-EGFP group (P > 0.05) (Figure 1E).
qRT-PCR and Western blot analysis showed that in pGC-FU-Otx2-EGFP group Otx2, Crx and Dkk-1 mRNA and protein levels increased and reached the peak on day 7, while Nrl mRNA and protein levels gradually increased and reached the peak on day 14.
In contrast, nuclear β-catenin/total β-catenin ratio (the indicator of Wnt/β-catenin pathway activation) gradually decreased (Figure 2A  P < 0.05. B and C, PCR and Western blot analysis showed that si-Crx significantly decreased Crx and Nrl mRNA and protein levels, si-Nrl significantly decreased Nrl mRNA and protein levels, while SB-216763 increased nuclear β-catenin/total βcatenin ratio. Data were presented as mean ± SD (n = 3).*P < 0.05, **P < 0.01 vs uninfected group; # P < 0.05, ## P < 0.01 vs VC group  The total thickness of retina in Model group and GFP group had no obvious change. B, The thickness of the retinal outer nuclear layer (ONL) at 14 days and 28 days after subretinal injection was 26.44 ± 8.21 μm and 32.81 ± 9.06 μm, respectively, significantly thicker than that of P0. Data were presented as mean ± SD (n = 6).*P < 0.05, **P < 0.01 vs group A; # P < 0.05, ## P < 0.01 vs group B Several studies showed that Otx2 promoted the formation of photoreceptor cells by regulating the expression of downstream factors Crx and Nrl. 26,27,29 Crx was the first identified specific transcription factor associated with photoreceptor cell development, which is highly expressed in mature cones and rods and controls the development and function of photoreceptor cells. 29 Otx2 regulates the expression of Crx by binding to Crx promoter, and both of them act on the downstream target Nrl to promote the formation of rhodopsin. 24,30 In this study, in MC-RSCs infected by pGC-FU-Otx2-EGFP lentivirus, mRNA and protein expression levels of Otx2 and its downstream targets Crx and Nrl increase gradually. These data confirm that Otx2 up-regulates the expression of Crx and Nrl in MC-

RSCs.
Otx2 promoted the differentiation of retinal stem cells by acting on its downstream gene Dkk-1 (Dickkopf-1), which negatively regulated Wnt pathway. 31 Wnt pathway regulates the development of nervous system, neurogenesis, neuronal proliferation and maintenance of stable state through complex signal transduction. 32 Dkk-1 is an inhibitor of Wnt pathway, which can form a trimer with the receptor LRP5/ 6 and Kremen1/2 and block intracellular transmission of Wnt signal. 33 Ip et al demonstrated that Dkk-1 was the direct target of OTX2 and OTX2 bound to the H1 control element of Dkk-1 to activate its and Rhodopsin in each group were calculated among total cells stained by DAPI. Data were presented as mean ± SD (n = 6). D, 28 days after subretinal injection, immunofluorescence staining showed that synaptophysin (red fluorescence) and GFP (green fluorescence) were positive in OPL of Otx2 group but not in OPL of Model and GFP group (bar = 50 μm). GCL, ganglion cell layer; ONL, outer nuclear layer; INL, inner nuclear layer; IPL, inner plexiform layer; OPL, outer plexiform layer. *P < 0.05 vs Model group; # P < 0.05 vs GFP group expression. 31 In this study, after Otx2 overexpression the mRNA and protein levels of Dkk-1 gradually increased but total β-catenin and nuclear β-catenin levels decreased gradually, consistent with previous studies. 31 Taken together, these data suggest that Otx2 could up-regulate the expression of Dkk-1 to inhibit Wnt pathway.
To explore signal pathway by which Otx2 enhanced the differentiation of MC-RSCs into photoreceptor cells, we used siRNA interference and Wnt activator SB-216763. SB-216763 is an inhibitor of GSK-3β, which is a key kinase to inhibit Wnt/β-catenin pathway. 33 Our results showed that the expression of . Scotopic a-wave (B) and b-wave (C) in Otx2 group were higher than in Model group, but were lower than in Normal group. Data were presented as mean ± SD (n = 6). *P < 0.05 interference or SB-216763 treatment in MC-RSCs infected with Otx2 lentivirus. These data suggest that Otx2 enhanced differentiation of MC-RSCs into photoreceptor-like cells is associated with the inhibition of Wnt pathway.
In addition, matrix environment affects the differentiation of photoreceptor cells. Osakada et al reported that in medium supplemented with Dkk-1 (Wnt antagonist), Lefty-A (Nodal antagonist), RA and taurine, nearly 20% ESs differentiated into Rhodopsin-positive Rod cells. 7 Zhong et al reported that RA and taurine could promote hiPSC differentiation into rod-like cells in vivo and in vitro. 6 On the basis of these data, we added RA and taurine into the differentiation medium to improve the differentiation efficiency of photoreceptor-like cells. To evaluate the in vivo differentiation of Otx2 induced MC-RSCs, we employed MNU-induced photoreceptor degeneration rat model as established previously. 22 The subretinal cavity is recognized as immune privileged region, and cell transplantation has advantages such as the low immune rejection and less systemic adverse reactions. 34

CONFLI CT OF INTEREST
The authors declare no conflict of interest.