Definitive surface markers for retinal progenitor cells (RPCs) are still lacking. Therefore, we sorted c-Kit+ and stage-specific embryonic antigen-4− (SSEA4−) retinal cells for further biological characterization. RPCs were isolated from human fetal retinas (gestational age of 12–14 weeks). c-Kit+/SSEA4− RPCs were sorted by fluorescence-activated cell sorting, and their proliferation and differentiation capabilities were evaluated by using immunocytochemistry and flow cytometry. The effectiveness and safety were assessed following injection of c-Kit+/SSEA4− cells into the subretina of Royal College of Surgeons (RCS) rats. c-Kit+ cells were found in the inner part of the fetal retina. Sorted c-Kit+/SSEA4− cells expressed retinal stem cell markers. Our results clearly demonstrate the proliferative potential of these cells. Moreover, c-Kit+/SSEA4− cells differentiated into retinal cells that expressed markers of photoreceptor cells, ganglion cells and glial cells. These cells survived for at least 3 months after transplantation into the host subretinal space. Teratomas were not observed in the c-Kit+/SSEA4−-cell group. Thus, c-Kit can be used as a surface marker for RPCs, and c-Kit+/SSEA4− RPCs exhibit the ability to self-renew and differentiate into retinal cells.
Photoreceptor degeneration occurs as a result of disorders affecting either the photoreceptors themselves or the associated retinal pigment epithelium (RPE) cells. This disease is a common cause of blindness and severely affects an individual’s quality of life (Pinilla et al., 2004). Photoreceptor degeneration is difficult to treat because the pathological process involves the apoptosis of photoreceptors or RPE cells, and photoreceptor cells cannot regenerate or self-repair. Although RPE cells have the capacity to proliferate in vivo and in vitro (Chiba, 2014; Stanzel et al., 2014), it is difficult for damaged RPE cells to repair themselves (Chiba, 2014). Currently, there are several treatment methods available, including gene therapy, transplantation therapy, drug therapy and artificial vision prostheses. Rescuing or regenerating photoreceptor cells or RPE in individuals with retinal degeneration is the key to treatment. Recently, stem-cell-based cell therapy has become a hot topic. Schwartz and colleagues have reported the safety and tolerability of human embryonic stem cell (ESC)-derived RPE cells for the treatment of dry age-related macular degeneration (AMD) and Stargardt's disease, and that report is the first description of the transplantation of human-ESC-derived cells into humans for the treatment of such diseases (Schwartz et al., 2012). Moreover, the authors observed no adverse proliferation or systemic safety issues related to the transplanted cells, and the best-corrected visual acuity was improved in some individuals (Schwartz et al., 2015). Thus, this human-ESC-based cell therapy in the treatment of blindness-causing diseases is of great significance because it provides a promising foundation for the use of cell-therapy intervention in numerous diseases (Sowden, 2014).
Tissue-specific retinal progenitor cells (RPCs) are an ideal source of differentiated cells with low tumor risk (Kajstura et al., 2011). Currently, most RPCs that are used for transplantation are derived from ESCs or are isolated from fetal tissues, but ESCs are difficult to differentiate in vitro, are not easily purified and might contain a variety of cell types or cells at different development stages. The continuous development of flow cytometry techniques to assess cell surface antigens has provided biomarkers that can be used to obtain highly purified tissue-specific RPCs. To date, several cell markers have proven to be suitable for the specific identification, isolation and enrichment of RPCs (Carter et al., 2009; Koso et al., 2009); these markers are invaluable for RPC research.
Stem cell factor receptor c-Kit (CD117), a progenitor cell marker, is a recognized antigen that is located on the cell surface which plays an important role in the survival and proliferation, as well as the prevention of apoptosis, of hematopoietic stem cells (HSCs) and lung cancer cells (Lennartsson and Ronnstrand, 2012). Koso and colleagues have identified c-Kit as an RPC marker in the mouse retina and have demonstrated a dramatic change in the expression profiles of the cell surface antigens c-Kit and stage-specific embryonic antigens (SSEAs) on RPCs during development (Koso et al., 2007). Hasegawa et al. have demonstrated that the human embryonic retina has a pool of c-Kit+ cells; however, the authors did not further culture or characterize them (Hasegawa et al., 2008). Stage-specific embryonic antigen-4 (SSEA-4), a human-ESC-associated antigen (Wright and Andrews, 2009), has previously been used as a marker to distinguish primitive ESCs (Kawanabe et al., 2012). A subpopulation of c-Kit+ cells expressing SSEA1 shows higher proliferative potential than c-Kit+/SSEA1− cells (Koso et al., 2007). However, in contrast to mouse ESCs, human ESCs lack SSEA1 and express SSEA4 (Wright and Andrews, 2009). Therefore, to reduce the risk of tumorigenicity, we used SSEA4 as a surface marker in fluorescence-activated cell sorting (FACS) analyses to exclude cells with high proliferative potential that had been isolated from the retina of human fetuses.
We hypothesized that c-Kit+ cells isolated from fetal retina tissues represent a population of stem cells. The RPCs were evaluated for cell characteristics, including self-renewal capacity, clonogenicity, the ability to differentiate into three types of retinal cells in vitro and the ability to differentiate into photoreceptors in vivo. Additionally, we injected the cells into the subretinal space of Royal College of Surgeon (RCS) rats, an animal model in which vision deteriorates due to RPE dysfunction; these rats serve as a model for a recessive mutation in a receptor tyrosine kinase gene (Mertk) that results in the progressive death of the photoreceptors (Pinilla et al., 2004). Then, we observed the survival, migration and differentiation of the cells, and determined their role in photoreceptor rescue and its impact on visual function. Finally, we assessed the safety of transplantation of c-Kit+/SSEA4− retinal progenitor cells into severe combined immune deficiency (SCID) mice. In our study, we found that the transplantation of c-Kit+/SSEA4− cells derived from human fetal retina tissues could protect the neural retina and preserve the retinal outer nuclear layer in RCS rats. Therefore, this study describes a method that can be used to obtain tissue-specific RPCs that could be used to delay the photoreceptor degeneration and to preserve the retinal outer nuclear layer in an animal model of retinal degeneration.
Distribution of c-Kit+ cells in human fetal eyes
The fetal neural retina at 12–13 weeks mainly comprises two layers: the outer neuroblastic layer (ONbL) and the inner neuroblastic layer (INbL). The cells in the ONbL are densely packed, whereas the cells in the INbL are more loosely packed. The ganglion cell layer can be seen around the optic disc, but not the peripheral retina.
c-Kit+ cells were not only distributed in the retina, but also in the cornea and choroid of the eye from a 13-week-old human fetus. c-Kit+ cells were localized in the inner retina from the optic nerve to the ora serrata (supplementary material Fig. S1), although more cells were located in the peripheral portion than at the posterior pole of the retina; c-Kit+ cells were also scattered in the corneoscleral limbus of the cornea (supplementary material Fig. S2) and choroid (Fig. 1A–C).
c-Kit+/SSEA4− cell isolation and culture
We found c-Kit+ cells in the retinas of eyes from human fetuses. c-Kit+/SSEA4− cells were sorted directly from the fetal retina; however, these cells were difficult to culture. Therefore, we first cultured retinal cells for 2–3 passages. Then, we collected retinal cells (a minimum of 5×106–10×106 cells) and sorted c-Kit+/SSEA4− cells by using FACS. These cells were collected and plated onto 24-well plates (5×103/cm2) (Fig. 1G). The morphology within the adherent population was spindle-shaped (Fig. 1H). By contrast, spheres were formed when the cells were grown in serum-free proliferation medium (Fig. 1J). The c-Kit epitope remained detectable by using immunofluorescence and FACS analyses after passaging (Fig. 1I,K).
Characteristics of c-Kit+/SSEA4− cells
Cells did not express the embryonic stem cell or tumor cell markers SSEA4 and the multi-drug resistance protein (MDR, also known as ABCB1) (Fig. 2C,F). By contrast, they expressed the RPC markers – including Pax6, Sox2, Rax and nestin – as analysed by using flow cytometry and immunofluorescence. c-Kit+/SSEA4− cells were found to express Pax6 (91.6±2.7%), Sox2 (95.2±2.0%), Rax (94.1±2.5%) and nestin (98.9±2.8%) (Fig. 3). Because c-Kit+ has also been described as a stem cell marker in other organs [i.e. mesenchymal stem cells (MSCs) and HSCs], we stained the sorted cells with MSC and HSC markers. The cells were negative for both the HSC markers (CD11b and CD45, also known as ITGAM and PTPRC, respectively) (Fig. 2B,H) and MSC markers (CD29 and CD140b, also known as ITGB1 and PDGFRB, respectively) (Fig. 2E,I).
Characterization of proliferation
The cells exhibited the ability to proliferate and grow in a monolayer on plastic plates in proliferation medium (Fig. 1H), and to form spheres in serum-free proliferation medium (Fig. 1J). Proliferating cells were identified by the proliferation marker Ki67 (also known as MKI67); the percentage of cells that expressed Ki67 was determined to be 82.0±3.1% by FACS analysis, which was consistent with the immunofluorescence staining results (Fig. 4A,B). c-Kit+/SSEA4− cells were seeded onto 24-well plates at a density of 10,000 cells/well, and the cell numbers were counted 1, 3, 5, 7 and 9 days after seeding (n=3). Cell proliferation reached a maximum and then plateaued after the cells were plated; by day 7, the cell number had increased by more than 20-fold (Fig. 4D). Additionally, we examined the cell cycle distribution of the c-Kit+ cells and found that 41.13±2.99% of cells were in the G2 and S phases (Fig. 4C).
c-Kit+ cells were cultured under two different conditions – adherent conditions with proliferation medium and non-adherent conditions with serum-free medium. When a single c-Kit+ cell (sorted using flow cytometry) was seeded onto a 96-well plate, clones formed approximately 20 days later (Fig. 1H). The clones were digested, 500 cells were seeded onto 10-cm dishes and more colonies formed. When these cells were cultured in serum-free medium, neurospheres formed after 20–30 days (Fig. 1J). The spheres were defined as free-floating with a diameter >40 μm. The percentage of cells that formed colonies after 30 days was 0.06±0.01%, but the efficiency of colony formation was less than that observed under adherent conditions (3.33%±1.53%) (P<0.01). This finding is consistent with the results of a previous report (Rangel et al., 2013).
Characterization of differentiation
Multipotency is another property of stem cells. Therefore, we used two differentiation media – photoreceptor differentiation medium and glia differentiation medium. The in vitro differentiation was assessed by culturing c-Kit+ cells in differentiation medium for 3 weeks using the two types of differentiation medium. Few differentiated cells expressed rhodopsin in the differentiation medium. By contrast, many cells differentiated into glial cells (66.7±5.8%). Because previous studies have shown that retinoic acid can promote the differentiation of photoreceptor cells in vitro, we changed the medium in order to induce photoreceptor differentiation, as previously described (Nakano et al., 2012; Li et al., 2013).
Photoreceptor differentiation medium containing retinoic acid was used to culture c-Kit+/SSEA4− cells for 3 weeks. Then, we stained for photoreceptor cell markers (Otx2, Crx, rhodopsin and recoverin), a ganglion cell marker (Thy1) and a glial cell marker (GFAP) (Fig. 5G–L). Otx2 (49.9±4.1%), Crx (59.9±4.0%), recoverin (68.1±5.1%) and rhodopsin (4.0±0.3%) were expressed in the differentiated cells, and 29.1±5.4% of the cells expressed Thy1 (Fig. 5A–F). By contrast, the cells that had been cultured in the initial differentiation medium only expressed GFAP and Thy1.
Differentiation of grafted cells
The results of our immunostaining showed that some cells migrated into the inner retina at 8 and 12 weeks after transplantation, and that a small number of cells expressed recoverin (Fig. 6H). However, the majority of the cells were located in the subretinal space. The percentage of transplanted cells that expressed photoreceptor cell markers at 4, 8 and 12 weeks were 1.01±0.11%, 2.36±0.25% and 5.22±0.14%, respectively.
Outer nuclear layer thickness
To verify the protective effect of transplanted cells on retinal degeneration, we measured retinal outer nuclear layer (ONL) thickness. The ONL of the cell-grafted retina was significantly thicker compared to the control and untreated groups at 4 weeks (28.43±1.95 μm versus 7.67±1.08 μm and 8.50±1.47 μm, n=3, P<0.01), 8 weeks (23.27±0.85 μm versus 6.61±0.65 μm, 6.83±1.08 μm, n=3, P<0.01) and 12 weeks (19.43±0.84 μm versus 4.17±0.75 μm, 4.80±1.08 μm, n=3, P<0.01). There was no significant difference in ONL thickness between the control and untreated groups (Fig. 6M).
Electroretinogram (ERG) analysis was performed, and b-waves were measured at 4 weeks, 8 weeks and 12 weeks. Significant high-amplitude b-waves were detected in the transplanted group at 4 and 8 weeks compared to the control and untreated groups (P<0.01). Recordings at 12 weeks were not significantly different between the sham-surgery and untreated groups (P>0.05) (Fig. 6N–Q).
The safety of c-Kit+/SSEA4− cells was tested through subcutaneous injection into the groin of six SCID mice; human ESCs were injected into another six SCID mice as a positive control. After 8 weeks, no gross inflammatory reaction was observed in any of the animals, and teratomas were not observed in the c-Kit+/SSEA4−-cell group. By contrast, teratomas were seen in the human-ESC group at 8 weeks post injection (Fig. 7).
In 2011, the FDA approved PhaseI/II clinical trials of cell transplantation therapy for AMD and Stargardt's disease, which led to immense advances in transplantation treatment of retinal degeneration (Schwartz et al., 2012). RPC transplantation for the treatment of retinal degenerative diseases has become the most promising therapeutic strategy. c-Kit+ RPCs are derived from fetal retinas and have advantages that include the potential to differentiate into retinal cells and a low risk of tumor formation. These factors make c-Kit a good marker for the selection of candidate cells for transplantation in order to treat retinal degeneration.
Our results showed that cells expressing the c-Kit epitope on their cell surface were distributed in the eye. c-Kit+/SSEA4− cells possessed characteristics of self-renewal and the ability to differentiate into three types of retinal cells. Owing to the development of flow cytometry technology and the discovery of new cell surface antigens, researchers have been able to obtain a higher purity of stem cells by using FACS isolation. c-Kit defines a regionally and temporally restricted immature subset of RPCs, the expression of which starts centrally and progresses centrifugally (Koso et al., 2007).
At 9–10 fetal weeks, the retina is divided into the ONbL and the INbL. At 11–13 fetal weeks, the ganglion cell layer (GCL) is several cell layers thick. At this stage, the outer boundary of the ONbL adjacent to the RPE is bordered by a row of cones, and the ONbL is separated from the differentiating GCL by a thin inner plexiform layer. The nerve fiber layer also becomes apparent on the inner boundary of the GCL close to the optic disc. The expression of recoverin and of the cone marker S-opsin does not begin until 11–12 weeks, and rod markers are expressed at 15–16 weeks (O'Brien et al., 2003; Hendrickson et al., 2008). In our study, fetal neural retina at 12–13 weeks mainly comprised two layers – the ONbL and the INbL, which is consistent with previous reports on the development of the fetal retina (O'Brien et al., 2003; Hendrickson et al., 2008). We found that c-Kit+ cells not only existed in the retina of the fetal eye but that they were also located in the cornea and choroid. Our results showed that c-Kit cells were located in the inner layer of the retina, and this finding is in agreement with the demonstration that the c-Kit ligand SCF is also expressed in the inner retina (Hasegawa et al., 2008). We successfully isolated and cultured c-Kit+ cells that had been derived from the human fetal retina, and identified their self-renewal, proliferation and differentiation characteristics. We also successfully used surface markers to isolate tissue-specific RPCs that did not express SSEA-4.
It is important to obtain pure RPCs and to exclude other types of ESCs after isolation of eye tissue from the human fetus. The sorting of both the c-Kit+ and SSEA4− surface markers allowed us to obtain purified RPCs. Additionally, the sorted cells were stained for stem cell markers (MDR) and then isolated by using flow cytometry to ensure that there was no contamination with ESCs; a total of 99% of the double-marker-sorted cells did not express the MDR marker.
c-Kit+ cells have been previously reported to be able to survive under suspension or adherent growth conditions (Koso et al., 2007; Kajstura et al., 2011; Rangel et al., 2013). The c-Kit cells isolated in this study could also survive under both conditions. A higher proliferation rate was observed under adherent conditions [in medium supplemented with fetal bovine serum (FBS)], which is consistent with the findings of Rangel et al. (Rang et al., 2013). One possible reason for this finding is that the cell–cell interactions and the cell adhesions that are present under adherent conditions are important for c-Kit+-cell growth (Rangel et al., 2013). Retinal progenitor cells grown under adherent conditions have also been reported to exhibit a greater proliferative potential than cells grown in suspension (Xia et al., 2012).
To ensure that the c-Kit+/SSEA4− cells that had been sorted using FACS were highly tissue-specific, these cells were further evaluated for the human RPC tissue-specific markers nestin, Rax, Pax6 and Sox2 by using immunohistochemistry and flow cytometry analyses. More than 90% of the cells expressed the RPC markers Pax6, Sox2, Rax and nestin, indicating that the cells were in an immature retinal cell state (Schmitt et al., 2009). Additionally, cell proliferation potential was assessed using the Ki67 marker. More than 80% of the c-Kit+/SSEA4− cells expressed this marker, demonstrating that the c-Kit+/SSEA4− cells possessed a high proliferation ability in vitro and represented a highly pure and tissue-specific RPC with a certain proliferative capacity.
Furthermore, we found that c-Kit+/SSEA4− cells could form colonies under adherent and suspension conditions. This finding is consistent with the other properties of c-Kit cells that have been reported by Kajstura et al. and Rangel et al. (Kajstura et al., 2011; Rangel et al., 2013).
In addition to proliferation capability, another characteristic of progenitor cells is the capacity to differentiate. In differentiation medium, c-Kit+/SSEA4− cells could be induced into photoreceptor cells, ganglion cells and glial cells that expressed the corresponding cell-specific markers (Otx2, Crx, recoverin, rhodopsin, Thy1 and GFAP). However, in our study the proportion of cells expressing rhodopsin (4.0±0.3%) was lower than the proportion reported in the study by Coles et al. (34.5±9.1%); furthermore, we found that more cells differentiated into glial cells (66.7±5.8%) than has been previously reported (19.7±10.6%) (Coles et al., 2004).
The transplantation of stem cells into the subretina of rats or mice that are afflicted by retinal degeneration has been demonstrated to produce a neuroprotective effect (Tian et al., 2011; Tzameret et al., 2014). In our study, we found that transplantation of c-Kit+/SSEA4−-cells that had been derived from human fetal retina tissue could protect neural retinas in RCS rats. The RCS rat is characterized by a recessive mutation in the Mertk gene, which encodes a receptor tyrosine kinase. This mutation precludes RPE cells from phagocytosing rod outer segments that have been shed, leading to the progressive death of photoreceptor cells (Pinilla et al., 2004). At (P)18 in these rats, numerous morphological changes are already apparent, including the disturbance of outer segments (Davidorf et al., 1991). After P21, the rod contribution to the mixed b-wave starts declining (Pinilla et al., 2004) and, at P22, changes in photoreceptor nuclei are found (Cuenca et al., 2005). We transplanted cells into the subretina at P21, which is when photoreceptor degeneration begins (Luo et al., 2014). Our experiment showed that grafted c-Kit+/SSEA4− cells improved ERG b-wave amplitude, which represents the electrical function of the retina (Tian et al., 2011), in RCS rats for 2 months after transplantation; this finding is consistent with the results of a previous study by Tian and colleagues (Tian et al., 2011).
It has been noted by Gonzalez-Cordero and colleagues (Gonzalez-Cordero et al., 2013) that reliable electroretinographic responses are only achieved in mice following the rescue of 150,000 functioning rods. Our study showed that ONL thickness was maintained for at least 3 months after cell transplantation; however, the thickness decreased over time. This decrease is likely to be due to the fact that there were not enough functional photoreceptors in the third month after transplantation. The grafted cells expressed the photoreceptor marker recoverin. However, the migrated human RPCs failed to express the ganglion cell marker Thy1. It has been previously reported (Luo et al., 2014) that stem cells grafted into the degenerative retina have difficulty differentiating into ganglion and photoreceptor cells in vivo; this finding was consistent with the results from our study.
Finally, the major concern regarding the use of progenitor cells for transplantation is tumorigenesis. The results from the teratoma assay showed that the transplantation of human RPCs is safe. There was no evidence of tumor formation 8 weeks after human RPC transplantation into the subretinal space of RCS rats.
In summary, our study demonstrates that c-Kit+/SSEA4− cells exist in human fetal eyes and that these cells possess the stem cell properties of self-renewal, colony formation and pluripotent differentiation. Although only a small proportion of the engrafted c-Kit+/SSEA4− cells in a rat model of retinal degeneration could differentiate to express the photoreceptor marker, c-Kit+/SSEA4− cells delayed apoptosis-mediated photoreceptor death or rescued host photoreceptor cells for at least 3 months. Therefore, c-Kit+ and SSEA4− might serve as good markers for the selection of candidate cells for transplantation in order to delay retinal degeneration.
MATERIALS AND METHODS
Cell isolation and culture
Eyes from human fetuses with a gestational age ranging from 12 to 14 weeks were obtained from spontaneous abortions at the Southwest Hospital, Third Military Medical University (Chongqing, China). The Ethics Committee of Southwest Hospital specifically approved this study, and it is registered in the Chinese Clinical Trial Register (ChiCTR; registration number – ChiCTR-TNRC-08000193). The gestational age of each fetus was determined using the last menstruation date and fetal foot length (Merz et al., 2000; Drey et al., 2005). Postmortem times of less than 1 h were used because they do not alter the ability of progenitor cells to survive in culture (Carter et al., 2007).
Cells were isolated from the neuroretinas of human fetal eyes as previously described (Coles et al., 2004; Klassen et al., 2004; Aftab et al., 2009; Schmitt et al., 2009; Baranov et al., 2013). Briefly, the eyes were rinsed in cold Hank's buffered salt solution (HBSS; Hyclone, South Logan, UT). The neuroretina was dissociated into small pieces and enzymatically digested with 1 ml of papain (12 units/ml; Worthington, Lakewood, NJ). The digested retinal tissue supernatant was filtered through a 40 μm filter (BD Biosciences, Franklin Lakes, NJ) to obtain single cells. The cells were centrifuged and re-suspended in proliferation medium that had been supplemented with FBS. The isolated cells were seeded into 6-well plates at a density of 5×105 cells/well. The cells were cultured at 37°C under 5% CO2, and the medium was changed every 3 days.
The following antibodies were used – allophycocyanin (APC)-conjugated anti-human c-Kit antibody (Biolegend, San Diego, CA), fluorescein isothiocyanate (FITC)-conjugated anti-human SSEA-4 antibody (BD Biosciences), phycoerythrin (PE)-conjugated anti-human CD29 antibody (Biolegend), PE-conjugated anti-human multidrug resistance protein (MDR) antibody (number 348605, Biolegend), APC-CY7-conjugated anti-human CD45 antibody (Biolegend), FITC-conjugated anti-human CD11b antibody (BD Biosciences), and PE-conjugated anti-human CD140b antibody (Biolegend).
The proliferation medium included Dulbecco's modified Eagle medium with nutrient mixture F-12 (DMEM/F-12; Hyclone) that had been supplemented with 20 ng/ml fibroblast growth factor-basic (bFGF, PeproTech, Rocky Hill, NJ); 20 ng/ml epidermal growth factor (EGF, PeproTech); 1×insulin, transferrin and selenium (ITS; GIBCO, Carlsbad, CA); 1×penicillin-streptomycin (GIBCO); and 10% FBS (GIBCO).
Immunocytochemistry was performed as previously described (Tian et al., 2011; Pearson et al., 2012). Briefly, rats were euthanized with an overdose of anesthesia, and the eyes were enucleated and fixed in 4% paraformaldehyde (0.01 M, pH 7.4). For human fetuses, the eyes were enucleated and fixed in 2% paraformaldehyde (0.01 M, pH 7.4) (Hendrickson et al., 2008). The eye cups were immersed in a graded series of sucrose solutions overnight, embedded in an optimal cutting temperature compound and then sectioned on a cryostat (Leica CM190, Wetzlar, Germany). Frozen tissues sections were cut as 10-μm-thick transverse sections.
Immunocytochemistry was performed using our previously described methods (Duan et al., 2013). Briefly, the cells or slides were fixed and permeabilized, then blocked in 10% serum for 30 min. The cells were incubated with primary antibody at 37°C for 2 h, washed with PBS, incubated with secondary antibody, washed with PBS, incubated with 6-diamidino-2-phenylindole (DAPI) for 10 min at room temperature and then washed with PBS.
The following primary antibodies were used – anti-c-Kit antibody (1:100, R&D Systems, Minneapolis, MN), anti-Pax6 (1:200, Abcam, Cambridge, MA), anti-Sox2 (1:100, Abcam), anti-rhodopsin (1:200, Abcam), anti-Otx2(1:200, Abcam), anti-Rax (1:100, Abcam), anti-nestin (1:200, Sigma, St Louis, MO), anti-Ki67 (1:300, Sigma), anti-GFAP (1:300, Abcam), anti-Thy1 (1:100, BD Biosciences) and anti-recoverin (1:10,000, Millipore). The following secondary antibodies were used – Cy3-conjugated antibody (1:800, Beyotime, NanTong, JiangShu, China) and FITC-conjugated antibody (1:200, Beyotime). Additionally, the cells were stained with DAPI (1:10, Beyotime). The images were captured using an Olympus OP70 microscope (Olympus Microscopy, Japan) or Leica TCS SP50 confocal microscope (Leica Microsystems, Wetzlar, Germany).
Fluorescence-activated cell sorting
The cells were digested with HyQtase for 5 min, followed by the addition of 3 ml of wash buffer (Biolegend) and centrifugation at 400 g for 5 min at 4°C. Then, the cells were re-suspended in Stain Buffer (Biolegend), and 2 µl of Fc block (Biolegend) was added to each tube. The cells were incubated for 15 min at 4°C and then incubated with fluorochrome antibody conjugates for 20–30 min on a shaker at 4°C. The cells were washed with wash buffer and centrifuged at 200 g for 5 min at 4°C. The supernatant was removed, 300 μl of PBS was added to the cells, and the suspension was transferred to a standard flow cytometry tube for FACS analysis (BD Biosciences, AriaII, Franklin Lakes, NJ).
Expansion of c-Kit+ cells
Sorted cells were plated at a density of 3×103–5×103/cm2 and cultured in proliferation medium DMEM/F12 that had been supplemented with 20% FBS, 20 ng/ml bFGF, 20 ng/ml EGF and 1×ITS for 3 days. Then, the medium was changed, and the FBS concentration was decreased to 10%. The medium was changed every 3 days, and the cells were passaged until they reached 80% confluence (every 3–4 days) at a constant seeding density of 10×103–13×103 cells/cm2.
Analysis of cell differentiation was performed using previously described methods (Coles et al., 2004; Li et al., 2013). Briefly, 5×103 cells/well of RPCs that had been passaged three times were seeded into 24-well plates on glass coverslips pre-coated with 0.015 mg/ml poly-lysine (Sigma) in glia differentiation medium. For photoreceptor and other retinal cell differentiation, 5×103 cells were seeded into 24-well plates on coverslips pre-coated with poly-l-lysine in photoreceptor differentiation medium; the differentiation medium was changed every 4 days.
The glia differentiation medium comprised DMEM/F12 (Hyclone) that had been supplemented with 10 ng/ml bFGF (PeproTech), 1× penicillin-streptomycin (GIBCO), 1% FBS (GIBCO) and 2 μg/ml heparin (Sigma).
The photoreceptor differentiation medium comprised DMEM/F12 (Hyclone) that had been supplemented with 10 ng/ml bFGF (PeproTech), 1×penicillin-streptomycin (GIBCO), 500 nM retinoic acid (Sigma) and 2% B27 (GIBCO).
Cell proliferation curve
The cell proliferation assay was performed as previously described (Aftab et al., 2009). Briefly, RPCs that had been passaged three times were seeded into 15 wells of a 24-well plate at a density of 10,000 cells/well. The cell number was counted in three wells at 1, 3, 5, 7 and 9 days; the experiment was repeated three times. The average number of cells was used to generate the cell proliferation curve.
The animals were treated as described under a protocol approved by the Institutional Animal Care and Use Committee of the Third Military Medical University in accordance with the National Institutes of Health guidelines for the care and use of laboratory animals, and with the Use of Animals in Ophthalmic and Visual Research (ARVO) statement. Mice and rats were fed and housed under a 12 h light–dark cycle. The drinking water of the rats contained cyclosporine A (210 mg/l) from 1 day before transplantation until they were euthanized (Lu et al., 2010).
RCS rats (males and females, 2–3 weeks old) were used for cell transplantation. Rats with congenital microphthalmia, dysplasia of extremities or congenital cataracts were excluded from the study.
The RCS rats were randomly divided into two groups: the cell grafted group (n=9), in which the rats received a subretinal injection of 3 μl of a c-Kit+/SSEA4− cell suspension (cell concentration, 2×105 cells/μl) and the PBS group (n=9), in which the rats received a subretinal injection of 3 μl of HBSS. In both groups, the right eyes (oculus dexter, OD) received cell transplants (cell grafted group) or HBSS (group), and the left eyes (oculus sinister, OS) were untreated. The left eyes of the cell-grafted group served as the untreated group.
Transplantation methods were performed as previously described (Tian et al., 2011). All rats were anesthetized with a single intra-peritoneal injection of 4% chloral hydrate (0.8 ml/100 g of body weight). The pupils were dilated with 1% tropicamide. A Hamilton syringe (29 gauge; Hamilton, Reno, NV) containing the DiI-labeled cell suspension was injected into the subretinal space. DiI was obtained from Invitrogen (Grand Island, NY).
The ERG techniques used were performed as previously described (Tian et al., 2011). The animals were tested at 4 weeks, 8 weeks and 12 weeks. All rats were dark-adapted overnight. The anesthesia method used was as described in ‘Cell transplantation’. A Flash-ERG recording electrode, comprising a small silver ring, was positioned on the corneal surface with a drop of methyl cellulose and used to record responses (Roland system, Wiesbaden, Germany). Each ERG response represents the average of three flashes. For all Flash-ERG recordings, the b-wave amplitude was measured from the a-wave trough baseline to the peak of the b-wave, and b-wave latency was measured from the onset of the stimulus to the b-wave peak. ERG b-waves were generated with flashes of white light at intensities of −0.3 cds−1 m−2 and 3.0 cds−1 m−2.
c-Kit+/SSEA4− cells (1×107 cells/100μl) were injected into the groin in six SCID mice, and the animals were observed for 8 weeks to detect possible tumor formation; human ESCs were injected into six SCID mice as a positive control. The animals were anesthetized and examined by a pathologist to identify microscopic pathological changes and evidence of tumor formation. The human ESC line H-1 (WA-01; Ma et al., 2014) was kindly provided by Yue Huang (School of Basic Medicine, Peking Union Medical College, Beijing, China).
Analysis of the thickness of the outer nuclear layer
Three areas of retinal ONL thickness were examined in the transplanted area (but not in the area that contained the layers of transplanted cells) in the treated and untreated groups, and in the sham-surgery group. The thickness of the ONL was evaluated in DAPI-stained sections in three areas along the grafted half of the retina. The ONL thickness was measured using Image-Pro Express software.
Cell counts and analysis
The number of DiI and recoverin double-positive cells in each image was counted at three locations in three areas along the grafted half of the retina from the retinal margin to the posterior pole from three rats for statistical analysis. Every fifth section was counted (50 µm) to avoid counting the same cell in more than one section; the cells were counted at a ×400 magnification (Wan et al., 2007; Xu et al., 2013).
Statistical analyses were performed using SPSS for Windows Version 13.0. Data are described as mean±standard error. Statistical comparisons were made using either Student's two-tailed t-test or analysis of variance. Differences were considered to be statistically significant at P<0.05.
The authors thank Dr. Xiaoli Liu (Southwest Hospital/Southwest Eye Hospital, Third Military Medical University, Chongqing, China; Division of Pulmonary and Critical Care Medicine; Departments of Pediatric Newborn Medicine, Brigham and Women's Hospital and Harvard Medical School, USA) for technical support, and thank Yuxiao Zeng, Qiyou Li and Chuanhuang Weng for assistance with cell transplantation and ERG analyses.
Peng-Yi Zhou contributed to conception and design, data collection and analysis, as well as writing and revision of the manuscript. Guang-Hua Peng contributed to conception and design, data analysis and interpretation, and the provision of study material, as well as to writing and revision of the manuscript. Haiwei Xu contributed to experimental design, data analysis and interpretation, as well as revision of the manuscript. Zheng Qin Yin contributed to conception and design of the study, and provided study material.
This work was supported by the National Key Basic Research Program of China [grant 2013CB967001 to Guang-Hua Peng and grant 2013CB967002 to Zheng Qin Yin]; and the National Natural Science Foundation of China [grant 31271400 to Guang-Hua Peng].
The authors declare no competing or financial interests.