Comparison of human retinal progenitor cells cultured in media with or without serum: in vitro and in vivo characteristics and retinal transplantation

Abstract

Retinal progenitor cells (RPCs) have a potential role in the treatment of retinal degenerative diseases. This study is to investigate in vitro and in vivo characteristics and retinal transplantation of RPCs cultured in media with or without serum. Progenitor cells obtained from the neural retina of human eyes at 6-16 weeks gestation were cultured in serum-free media (SF-hRPCs) or in media containing 10% fetal bovine serum (FBS) (S-hRPCs). The differences were characterized between the cells cultured in vitro and transplanted (retinal transplantation) into Royal College of Surgeons (RCS) rats. The functional status of the rats was examined by flash-electroretinogram recordings. The result was that S-hRPCs exhibited higher proliferative dynamics in vitro. On the basis of outer nuclear layer thickness and flash-electroretinograms, S-hRPCs were more efficacious in slowing the progression of retinal degeneration following transplantation compared with SF-hRPCs. Moreover, retinal mesenchymal-like stem cells were isolated and identified from the S-hRPCs cultures. Our study demonstrated the potential of retinal MSCs for the treatment of retinal degeneration.

Introduction

There are limited therapeutic options for retinal degenerative diseases, and currently the possibility of stem cell-mediated regenerative treatment is being actively explored [1]. The proposed mechanisms of retinal restoration are based on cell replacement and bystander effects (i.e. trophic support, immunomodulation, and enhanced neuronal plasticity of stem or progenitor cells) [2]. Several populations of stem cells or progenitor cells, including retinal stem/progenitor cells (RSCs/RPCs), bone marrow mesenchymal stem cells (BMSCs), neural stem cells, and embryonic stem cells have been used for retinal transplantation [3-6]. In particular, RPCs of rodents, pigs, and humans (hRPCs) have been identified and can be cultured in serum-free media, which allows the survival of neuronal cultures with very few glia [6-10]. These cells can express mature retinal neuron markers following transplantation into the subretinal space of allo recipients [7,9].

To date, mesenchymal stem (progenitor) cells (MSCs) have been isolated from a variety of human tissues, such as bone marrow, adipose tissue, deciduous teeth, and perivascular cells of human umbilical cord derived from Wharton’s jelly [11-15]. Multipotent perivascular MSCs have also been extracted from human brain, and MSCs have been identified in cornea, conjunctiva and trabecular meshwork of the eye [16-19]. In this study, we aimed to determine whether a population of MSCs could be isolated from human fetal neural retina.

Materials and methods

Isolation and culture of hRPCs cells

Human fetal eyes (6-16 weeks gestational age, n = 16) were obtained from legal routine therapeutic abortions. Under sterile conditions, both eyes were dissected free from other tissues and the neuroretina was separated as previously described [8,9]. Retinas were minced and digested for 20-30 min with 1 U/ml papain (Worthington Biochemical Corp., Lakewood, NJ) at 37°C and 5% CO₂, and then mechanically dissociated into single cells with a small-bore pipette. Dissociated cell suspension was centrifuged at 1000 rpm for 5 min and the remaining undissociated tissue processed through another cycle in fresh papain. The cell pellets were re-suspended in culture medium (Ultraculture; Lonza, Basel, Switzerland) supplemented with 10 ng/mL epidermal growth factor (EGF; Sigma-Aldrich, Inc., St. Louis, MO), 20 ng/mL basic fibroblastic growth factor (bFGF; Invitrogen, Carlsbad, CA), 2 mM L-glutamine (Invitrogen), 1% Penicillin/Streptomycin (P/S; Invitrogen), and 5% fetal bovine serum (FBS, Invitrogen). After counting the number of live and dead cells using a trypan blue assay (Sigma-Aldrich), isolated cells were plated onto fibronectin-coated (100 μg/mL) tissue culture flasks at a density of 9-13×10³ cells/cm², replaced with FBS-free complete medium 24 hours after electroplating, and used for the remainder of the entire experiment (SF-hRPCs).

S-hRPCs were isolated in the same manner as SF-hRPCs described above. The cell pellets were re-suspended in DMEM/F12 with Glutamax (Gibco; Grand Island, NY) containing 10% FBS, 10 ng/mL EGF, 20 ng/mL bFGF, and 1% P/S, transferred to an uncoated six-well plate, and cultured at 37°C with 5% CO₂.

Written informed consents of embryo donation were obtained from the patients. This study was approved by the Medical Ethics Committee in General Hospital of Chinese People’s Liberation Army (WHO Trial Registration, ChiCTR-TNRC-08000193). All experiments adhered to the Declaration of Helsinki.

Growth curve of two subpopulation cells

The proliferative potential of SF- and S-hRPCs was assessed according to the growth curve. Cells were fed by replacing half the volume of the medium with fresh medium every 48 h and passaged when they reached 70-85% confluence. SF-hRPCs were plated onto new fibronectin-coated flasks, while S-hRPCs were passaged and re-plated onto uncoated flasks. SF- and S-hRPCs were plated at a constant seeding density of 9-13×10³ cells/cm². The cultured cells were digested with HyQtase (Hyclone, Logan, UT) for 3-4 minutes at 37°C, centrifuged at 1000 rpm for 5 min, and re-suspended in medium. The number of live and dead cells was counted by trypan blue method when the cells were passaged. Hemocytometer (Corning, Steuben County, NY) was also used to count live cell numbers.

Quantitative RT-PCR

Total RNA was isolated from 11-13 week fetal retinal cells of passage 2 (P2) by TRIzol reagent (Invitrogen). The following experiments were generally done with materials obtained from 11-13 week old fetuses. The hRPCs isolated from different gestational age might show different characteristics in vitro, especially when the gestational age varies over a wide range and the experimental variability were eliminated by using restricted time window.

After reverse transcription reaction, real-time PCR of the ABI 7900HT system was performed using SYBR® GreenRealtime PCR Master Mix (TOYOBO, Osaka, Japan) with the following conditions: 94°C for 10 seconds, 58°C for 20 seconds, and 72°C for 30 seconds for 40 cycles. A melting curve of melting stage at the end of the amplification procedure showed no non-specific amplification. Retinal progenitor markers (nestin, PAX6, SOX2 and OTX2), photoreceptor precursor marker (CRX), and cell proliferation marker (KI67) were analyzed. GAPDH was used as the internal control. Pfaffl method was used to analyze the real-time PCR data. After normalizing to GAPDH expression, the relative expression levels of the genes in S-hRPCs were expressed as a fold change relative to expression levels in the SF-hRPCs.

Immunocytochemistry

P2 of SF- and S-hRPCs were used for immunocytochemistry as previously described [8]. In brief, the cells were fixed with 0.1 M phosphate balanced solution (PBS) containing 4% paraformaldehyde for 30 min, then blocked with 10% goat serum and dealt with 0.1% Triton-x (Sigma-Aldrich) for 1 h after rinsed in PBS. The cells were then immunolabeled with antibodies that anti-KI67 (1:400; Abcam, Cambridge, UK), anti-nestin (1:200; Sigma-Aldrich), anti- PAX6 (1:200; Abcam), anti-SOX2 (1:200; Abcam), and anti-GFAP (1:200; Abcam) at 4°C overnight.The cells were subsequently incubated in Cy3-conjugated secondary antibodies (1:100; Santa Cruz Biotechnology; Santa Cruz, CA) for 1 h. Nucleus was stained with DAPI (Beyotime Institute of Biotechnology, Jiangsu, China) and observed under a fluorescence microscope (Olympus BX51, Tokyo, Japan).

Animals

Animal experiments were performed according to the NIH guidelines for the care and use of laboratory animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The study was approved by the Institutional Animal Care and Use Committee of General Hospitalof Chinese People’s Liberation Army.

A total of 20 three-week-old pigmented dystrophic RCS (RCS-p⁺) rats carrying a deletion mutation in MerTK encoding gene were used in this study. Rats were fed with water containing 210 mg/l cyclosporine A (Novartis, Basel Switzerland) from the day before transplantation until they were sacrificed. The eyes were divided into four groups with 10 eyes each: group one received sub-retinal injections of S-hRPCs; group two (contralateral eyes in group one) received sub-retinal injections of SF-hRPCs; group three was sham operated animals and received injections of the carrier medium (PBS) only; and the remaining group (contralateral eyes in group three) received no treatment.

Sub-retinal transplantation

Sub-retinal transplantation was performed according to that described previously [10]. Briefly, 1 h before the transplant surgery, SF- and S-hRPCs (P2) were gently digested and dissociated to cell suspension, and then over 96% of cells were successfully labeled with the red fluorescent marker (2 μg/ml lipophilic membrane stain CM-Dil; Invitrogen).

All recipient rats were fixed in a head holder and anaesthetized with a single intraperitoneal injection of medetomidine hydrochloride (0.01 mg/10 g body weight; Dormitor, Pfizer,Karlsruhe). The pupils were dilated with 1% tropicamide (Santen Pharmaceutical Co., Ltd. Osaka, Japan). A 10 μl Hamilton syringe (30 gauge; Hamilton, Nevada, USA) containing the cell suspension was tangentially inserted through the conjunctiva and sclera into the sub-retinal space causing a self-sealing wound tunnel. Cell suspensions (5 μl with total 8×10⁵ cells/eye) were slowly injected to produce a retinal detachment in the temporal retina. The control group was injected with 5 μl PBS. All surgical procedures were performed under operating ophthalmic microscope. The cornea was punctured to reduce intraocular pressure and limit the efflux of cells at the injection site. Fundus examination was performed with direct ophthalmoscope viewing following sub-retinal transplantation.

Flash-electroretinogram recordings

Flash-electroretinogram recording was undertaken 6 and 9 weeks postnatal (3 and 6 weeks postoperative; n = 5 eyes/group/time) according to methods described previously [6]. ERG b waves were generated by white-light flashes at intensities ranging from -6.3 log cd-s. m^-2 to 0.6 log cd-s. m^-2.

Histology and immunohistochemistry

Rats 6 and 9 weeks old were killed by an overdose of anesthetic at postnatal (n = 5 eyes/group/time), the eyes were enucleated and the eyecups were fixed in PBS (0.01 M, pH 7.4) containing 4% paraformaldehyde. Eyecups were immersed in a graded series of sucrose solutions (10%, 20%, and 30% in 0.01 M PBS) for 2 d, embedded in OCT, and sliced on a cryostat (10 μm; Leica CM1900; Leica, Inc., Nussloch, Germany). We then attached sections to poly-L-lysine-coated slides, observed the thickness of outer nuclear layer (ONL) by differential interference contrast fluorescence microscopy (Leica), and measured layer thickness with Image Pro-plus 6.0 system.

Retinal sections were blocked in PBS containing 10% normal goat serum, 3% bovine serum albumin, and 0.1% Triton X-100 for 1 h at room temperature. Then the slices were incubated in human specific mouse anti-mitochondrial antibody (1:200, Abcam) and rabbit anti-mitochondrial antibody (1:200, Millipore, Billerica, MA) at 4°C overnight to identify the DiI labeled transplanted cells. Rabbit anti-recoverin polyclonal (1:1000, Millipore) and mouse anti-rhodopsin monoclonal antibody (1:8000; Sigma-Aldrich) were used to assess the differentiation of the donor cells. The sections were incubated with Alexa Fluor 488 and 568 goat anti-rabbit IgG (H+L) (1:500; Invitrogen) or Alexa Fluor 488 and 568 goat anti-mouse IgG (H+L) (1:500; Invitrogen) for 1 h at room temperature and then counter-stained with DAPI. The samples were washed with PBS at each step. As negative control, sections were processed as described above except without incubation with the primary antibodies. The retinal sections were examined using fluorescence (Leica, DM TRET) or confocal microscopy (Olympus, FV 1000; Tokyo, Japan). Sections were incubated in 70% ethanol containing 0.01% Sudan black B (Merck KGaA, Darmstadt, Germany) for 1 min to reduce autofluorescence [20]. The percentage of positive cells was determined by dividing the number of immunopositive cells by the number of DAPI stained nuclei. 500 to 1000 cells for each RPCs subgroup and each culture were counted in random fields.

Flow cytometry analysis of S-hRPCs and SF-hRPCs

To identify the nature of S-hRPCs and SF-hRPCs, P2 and P7 of confluent hRPCs were digested and labeled with the following commercial antibodies: Negative marker cocktail (CD45/CD34/CD11b/CD19/HLA-DR PE), anti-CD105-PerCP-Cy5.5, anti-CD73-APC, anti-CD90-FITC, anti-CD44-PE antibodies respectively, all antibodies were bought from BD Biosciences, San Diego, CA; Cat. No. 562245). After fixation, flow cytometry analysis was performed on a BD FACS Calibur Flow Cytometer (BD Biosciences). Negative control immunofluorescence experiments were performed in parallel with unrelated antibodies.

The properties of differentiation into a mesenchymal lineage

To find out if cultures contained retinal MSCs and these retinal MSCs had the properties of stromal and mesenchymal potential, 6^th passage S-hRPCs and SF-hRPCs were induced to undergo osteogenic, adipogenic, or chondrogenic differentiation as previously described [21]. Cells were treated with osteogenic medium which consisted of DMEM-LG supplemented with 10% FBS, 50 μg/ml ascorbate-2 phosphate, 10^-7 M dexamethasone, and 10 mM β-glycerophosphate (all from Sigma) for three weeks to stimulate osteogenic differentiation. After culture for three weeks, cells were subsequently fixed with 4% formaldehyde and stained with oil-red O (Sigma). Differentiated cells were treated with 2% silver nitrate (Sigma) under a UV lamp for 1 h, fixed with 2.5% sodium thiosulfate (Sigma) for 5 mins, washed with dH₂O, and then counterstained with 1% alizarin red (Sigma) to evaluate their mineralized matrix. In order to induce chondrogenesis, cells were cultured as high-density pellet (2.5×10⁵ cells/pellet) in DMEM-LG serum-free medium supplemented with 10^-7 M dexamethasone, 40 μg/ml L-proline, 100 μg/ml sodium pyruvate, 1% ITS-premix, and 10 ng/ml transforming growth factor-β3 (TGF-β3) (R&D Systems, Minneapolis, MN) for 4 weeks. Then the extracellular matrix in chondrocyte-like beads and pellets were embedded in paraffin, cut into histological sections, and stained with alcian blue (Sigma).

Statistical analysis

Experimental data are expressed as the mean ± standard deviation (SD). Each experiment was repeated at least three times unless otherwise specified. Statistical comparisons were done by Student’s two-tailed (t-test). Differences were considered to be significant when P ≤ 0.05.

Results

Morphology and expansion potentials of SF- and S-hRPCs

To unravel the effects of different culture conditions on hRPCs, the morphological changes were observed. After dissociation into single cells (Figure 1A), SF-hRPCs adhered to the bottom of the flask within 34.3 ± 16.8 h (n = 8) (Figure 1B). Once adherent, SF-hRPCs flattened and began to spread out as a monolayer over the next 1-4 days, attaining a cobblestone-like appearance. During this time, some cells presented retinal ganglion-like cell growth with long axons (1-2 days, Figure 1C, white arrow) or showed bipolar-like shapes (Figure 1E, red arrow). The morphology of SF-hRPCs remained relatively consistent within six passages (Figure 1F), and beyond 6^th passage, growth ceased and cells either started undergoing cell death or differentiated into a more fibroblastic morphology (Figure 1G). S-hRPCs took a longer time to adhere to the flask (49.0 ± 14.5 h; n = 4, P < 0.05) (Figure 1D) and its morphology was maintained for at least 15 passages (Figure 1H).

Figure 1.

Morphology and expansion potential of SF-hRPCs and S-hRPCs. Fresh neuroretinas were dissociated into cell cultures (A). Cells cultured in media without or with serum on day 1 and 5 in vitro respectively (B, D). A few SF-hRPCs showed ganglion cell-like (C, white arrow) and bipolar-like (E, red arrow) growth in primary cultures. Most SF-hRPCs showed cobblestone-shaped morphology that was consistent throughout 6 passages (F), then the cells took on a fibroblastic phenotype (G) or died beyond 6 passages. S-hRPCs assumed a spindle- shaped morphology and maintained in beyond 15 passages (H). SF-hRPCs isolated from older donor tissue (15-16 weeks gestation) exhibited stronger expansion potential than cells isolated from younger donor tissue (6-7 weeks gestation); smaller numbers of hRPCs resulted from younger donor tissues (I). S-hRPCs had stronger expansion potential than SF-hRPCs (J).

S- and SF-hRPCs could proliferate in vitro and be passaged several times. SF-hRPCs isolated from older donor tissue (15-16 weeks; n = 3) exhibited stronger proliferative capacity than those isolated from younger donor tissue (6-7 weeks) (Figure 1I; n = 3, respectively; P < 0.05). A smaller number of hRPCs were yielded from younger donor tissue at the time of tissue collection. However, SF-hRPCs from younger donor tissue could also continue proliferating to 5-6 passages, which is different from a previous study arguing that proliferation of SF-hRPCs isolated from younger donor tissue slowed after 1-2 passages [9]. S-hRPCs had a greater potential in average expansion by the 6^th passage compared with SF-hRPCs (Figure 1J; n = 4, respectively; P < 0.05) and their continued proliferation over 15 passages had been assessed (data not shown). S- and SF-hRPCs could be frozen and thawed without loss of proliferative capacity.

Expression of progenitor and proliferative markers

In order to study the self-renewal and expansion potential of SF-hRPCs vs. S-hRPCs, expression of Nestin, PAX6, SOX2, OTX2, CRX, and Ki67 was detected. The results showed that the expression of Nestin, PAX6, SOX2, OTX2 and CRX was significantly higher, while that of Ki67 was far lower in the SF-hRPCs comparing with S-hRPCs (P < 0.05) (Figure 2A).

Figure 2.

Expression of progenitor markers and proliferation markers in SF-hRPCs and S-hRPCs after the 2^nd passage (11-13 gestation material). qPCR results showed: the expression of nestin, PAX6, SOX2, OTX2 and CRX was significantly higher and that of Ki67 was far lower in SF-hRPCs compared with S-hRPCs (n = 3, P < 0.05) (A). SF-hRPCs and S-hRPCs cultures were successfully immunolabeled with antibodies against KI67 (B, G), PAX6 (C, H), SOX2 (D, I), nestin (E, J), and GFAP (F, K). The percentage of SF-hRPCs expressed PAX6, SOX2 and Nestin was much higher, while percentage of cells that expressed Ki67 and GFAP were lower compared with S-hRPCs cultures (L) (n = 3, P < 0.05).

Consisted with the qPCR results, most of the SF-hRPCs were immunopositive to PAX6 (89.5% ± 3.7), SOX2 (94.8% ± 1.0), and Nestin (100% ± 0.0) (Figure 2C-E), which remained fairly constant through all passages tested (P5-P6) (data not shown). However, the proportion of cells expressing PAX6 (Figure 2H), SOX2 (Figure 2I), and Nestin (Figure 2J) was lower in S-hRPCs cultures (77.2% ± 6.5, 71.6% ± 2.2 and 74.7% ± 1.2, respectively; P < 0.05), which decreased sharply within six passages. A far higher proportion of S-hRPCs cells expressing Ki67 was found (Figure 2G) compared with SF-hRPCs (Figure 2B) (80.3% ± 7.3, 71.2% ± 3.1, respectively; P < 0.05). GFAP was not detected in SF-hRPCs cultures (Figure 2F), but low expression (7.3% ± 2.2) was detected in the S-hRPCs cultures (Figure 2K).

In-vivo retinal transplantation

S- or SF-hRPCs from 11-13 week fetal material and P2 were successfully injected into the sub-retinal space of postnatal 3 week old RCS-p⁺ rats. S- and SF-hRPCs (prelabeled red with DiI and human specific anti-mitochondrial antibody) were present in the host sub-retinal space 3 and 6 weeks post-transplantation in 100% of cases. At 6 weeks postnatal, small number of S- and SF-hRPCs were found migrated towards the host outer nuclear layer (ONL) (Figure 3A, 3C), while very few cells expressed recoverin (expressed by mature photoreceptor and midget cone bipolar cells) or rhodopsin (a rod photoreceptor marker). The thickness of the ONL appeared to be well-maintained following S-hRPCS transplants (Figure 3B; see below).

Figure 3.

In-vivo retinal transplantation. A small number of S-hRPCs (A) and SF-hRPCs (C) (red) migrating towards the host ONL (counterstained with green recoverin and blue DAPI) at six weeks postnatal; S-hRPCs after transplantation maintained the thickness of the ONL in RCS-p⁺ rats (B). Abbreviations: INL, inner nuclear layer; ONL, outer nuclear layer; SRS, sub-retinal space; PN, postnatal.

The thickness of the ONL

In order to identify the influence of S-hRPCs and SF-hRPCs transplants on ONL, ONL thickness of five rats from each group (two sections per rat) was measured.

Both S-hRPCs and SF-hRPCs transplants had reduced loss of ONL cells, and the distribution of DiI-labeled transplanted cells covered over half of the retina, although ONL thickness still diminished over the 6-week transplant period (Figure 4). The protective effect was accompanied by the presence of the DiI-prelabeled transplanted cells. Moreover, surviving ONL cells extended beyond the boundaries of donor cell distribution.

Figure 4.

S-hRPCs and SF-hRPCs transplantation helped to maintain the thickness of the ONL. Comparison of ONL thickness between S-hRPCs-transplanted (A, E), SF-hRPCs transplanted (B, F), PBS-treated (C, G), and untreated groups (D, H) 6 and 9 weekspostnatal. (I) The ONL were significantly thicker in S- and SF-hRPC transplanted groups at both time points compared with PBS and untreated animals (*P < 0.05; n = 5; ANOVA). S-hRPCs transplantation resulted in far thicker ONLs compared with SF-hRPCs (n = 5; P < 0.05). GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; PN, postnatal. Scale bar: 50 μm.

Compared with SF-hRPCs transplants, ONL thickness was significantly thicker after S-hRPCs transplants (60.1 ± 6.4 vs. 45.8 ± 5.4 μm respectively; n = 5 rats, P < 0.05) at postnatal 6 weeks. Although the ONL was still thicker following S-hRPCs transplantation compared with SF-hRPCs (41.6 ± 3.1 vs. 30.3 ± 4.3 μm, respectively; n = 5 rats, P < 0.05) 9 weeks post-transplantation, the ONL was far thinner at 9 weeks compared with 6 weeks (n = 5, P < 0.05) post-transplantation. Both transplants resulted in obvious thicker ONL compared with PBS-injection (27.5 ± 2. 8 and 17.9 ± 2.0 at 6 weeks and 9 weeks postnatal, respectively) or untreated eyes (postnatal, 29.4 ± 2.3 and. 17.2 ± 2.2 μm at 6 weeks and 9 weeks postnatal, respectively) (n = 5, P < 0.05). No significant difference was found between PBS and untreated groups.

ERG amplitude measurements

ERG amplitudes indicated that both S- and SF-hRPCs transplants prolonged the functional integrity of the RCS retina, which was consistent with ONL thickness measurement. S- and SF-hRPCs injected animals got significantly higher amplitude rod-ERG b waves amplitude compared with PBS injected group (149.7 ± 27.5, 118.5 ± 20.9 vs. 82.8 ± 15.4 μv, respectively; n = 5, P < 0.05) at 6 weeks. Max-ERG b wave amplitude was also far higher compared with that of PBS treated rats (175.3 ± 18.0, 145.5 ± 25.1 vs. 99.7 ± 22.9 μv, respectively; n = 5, P < 0.05). These differences lasted until postnatal week 9 (rod-ERG-b 110.0 ± 24.7, 81.3 ± 18.5 vs. 35.8 ± 11.1 μv; Max-ERG-b 144.0 ± 25.5, 109.5 ± 21.4 vs. 51.0 ± 13.8 μv; n = 5, P < 0.05) albeit with reduced amplitude compared with that observed at postnatal week 6 (n = 5, P < 0.05) (Figure 5).

Figure 5.

ERG measurements after transplantation. Max-ERG were recorded at 6 weeks (A) and 9 weeks postnatal (B). Recorded Rod-ERG at the same time points (C) and (D) respectively. (E, F) showed amplitudes of ERG Max-b wave and Rod-b wave, respectively. Two wave amplitudes were significantly higher after S- and SF-hRPC transplantation than those PBS-treated and untreated groups at 6 and 9 weeks postnatal (*P < 0.05; n = 5; ANOVA). Also, the amplitudes of 2 waves were obviously higher in S-hRPC transplanted animals compared with SF-hRPC transplanted animals at 2 time points (P < 0.05).

Rod-ERG b waves and Max-ERG b wave of S-hRPCs transplanted rats were significantly higher than SF-hRPCs transplanted rats at both postnatal 6 and 9 weeks (n = 5, P < 0.05). No significant differences between PBS-treated and untreated groups were identified (Figure 5).

Mesenchymal stem cell characterization of S-hRPCs and its potential for osteogenic, chondrogenic, and adipogenic differentiation

S-hRPCs derived from eye tissues of 11-13 week animals displayed MSCs characteristics after the 6^th passage. According to the results of flow cytometry analyses, the CD73, CD90, CD44, and CD105 detection in S-hRPCs after the 6^th passage were strongly positive (Figure 6); almost half of S-hRPCs from P2 expressed CD105 antigen (43.9% ± 7.6%) (Figure 7), however, CD14, CD45 and CD34, B-lymphocyte antigen CD19, and microglial antigen CD11b were negative. The expression of surface human leukocyte antigen DR (HLA-DR) could not be detected (Figure 6). In contrast, SF-hRPCs showed no MSCs characteristics after the 6^th passage, they were generally negative for CD105 (1.4% ± 0.3%, n = 3) (Supplementary Figure 1), which was the same with SF-hRPCs from P2 (Data not shown, n = 3). These results indicated that S-hRPCs from P2 contain a group of retinal MSCs.

Figure 6.

The MSCs characteristics and mesodermal potential of S-hRPCs. (A-E) S-hRPCs were cultured until the 6^th passage and labeled with monoclonal antibodies anti- CD73, CD90, CD105, CD44, CD34, CD45, CD11b, CD19 and HLA-DR, and then analyzed by FACS (n = 3). Isotype-matched IgG controls (non-shaded green curves) and S-hRPCs curves (blue) are shown. (F) S-hRPCs exhibited matrix mineralization formation three weeks after the addition of osteogenic induction medium (red). Adipose (red, G) and chondrogenesis (blue, H) formation were observed 4 weeks after the addition of adipogenic induction medium and chondrogenic medium TGF-β3.

Figure 7.

Flow cytometry analysis of S-hRPCs (P2). S-hRPCs (P2, n = 3) expressed MSC markers including CD73 (99.3% ± 1.7%), CD90 (90.4% ± 4.0%), CD44 (97.4% ± 2.3%). Almost half of the cells expressed the CD105 antigen (43.9% ± 7.6%), but the CD14, CD45, CD34, B-lymphocyte antigen CD19, microglial antigen CD11b and HLA-DR could not be detected. The results indicated that S-hRPCs (P2) contained a group of retinal MSCs.

To investigate mesenchymal nature of S-hRPCs after 6 passages, cells were treated with appropriate osteo-, chondro-, adipo media, and their differentiation was confirmed via appropriate staining. As shown in Figure 6F, S-hRPCs stained by alizarin red exhibited matrix mineralization formation after 21 days in culture. S-hRPCs were also able to differentiate towards adipose phenotype after culture for 21 days with adipogenic medium (Figure 6G; Oil-red O staining). In addition, after the formation of high-density pellet and cultured in a serum-free chondrogenic medium with TGF-β3, cells could also be induced towards chondrogenic phenotype after 28 days later (Figure 6H, positive alcian blue staining). These results indicated that the S-hRPCs after the 6^th passage had the potential to differentiate along stromal and mesenchymal lines, while SF-hRPCs grown under the same conditions did not exhibit potential to differentiate into other cell lineages (osteo/chondro/adipo).

Discussion

RPCs represent a subtype of tissue-specific multipotent cells and are often referred to as retinal neural progenitor cells [22]. MSCs are the conceptual progenitors of most derivatives from mesoderm and have been identified from several different tissues. Both RPCs and MSCs show great potential in the treatment of retinal degenerative disease [5,23]. In this study, we investigated the growth kinetics and in vitro and in vivo characteristics of SF and S-hRPCs. The result suggested that SF-hRPCs were characterized as retinal neural progenitor cells and sustained their immature, undifferentiated identity in vitro. However, mesenchymal-like stem cells were successfully isolated and identified in S-hRPCs cultures, which had a greater capacity of self-renewal in vitro. Moreover, S-hRPCs had more maintenance of ONL thickness and prolonged visual function after transplantation compared with SF-hRPCs.

The most commonly used and recommended culture method of RPCs expansion requires serum-free medium [8]. Our studies demonstrated that SF-hRPCs could self-renew and proliferate, and stayed in an undifferentiated state that expressed immature markers when cultured. qPCR and immunofluorescence demonstrated that the SF-hRPCs from P2 (cultured in serum free medium, except that the first day, medium contained 5% FBS to increase cell survival rate and plating efficiency) expressed retinal progenitor markers including Nestin and SOX2, and cell proliferation marker Ki67, indicating they stayed in an immature state. The expression of Nestin and PAX6 remained at a constant level through several cell passages (P5-P6), the same with previous reports [8]. Nestin is a marker of neural intermediate filaments and abundantly expressed in undifferentiated cells. PAX6 is a paired box transcription factor of region-specific neural progenitors, which plays an important role in ocular development [24]. Ki67, a proliferation marker, is a nuclear protein associates with cell cycle and expresses only in cells that are actively dividing. We found that SF-hRPCs isolated from younger donor tissues could continually proliferate over 5-6 passages, which is different from previous results that proliferation slowed down after 1-2 passages [9]. This might mainly be due to the short delay between collection and culture (within 30 mins) and the difference of culture medium supplier: our culturing medium was Ultra culture from Lonza while they used Rencell media from Chemicon.

The qPCR and immunofluorescence results showed that the majority of S-hRPCs (P2, cultured in medium containing 10% FBS) remained in an undifferentiated state. Flow cytometry analyses indicated S-hRPCs had characteristics of MSCs; though the expression of Nestin, PAX6, and SOX2 decreased sharply after 6 passages, they retained strong self-renewal and proliferative capacity up to the 15^th passage. After 6 weeks’ sub-retinal transplantation, small number of SF-hRPCs and S-hRPCs could be found in degenerated retinas and a small proportion of them expressed recoverin (a marker of cone and rod photoreceptors and cone bipolar cells) and mature rod marker rhodopsin. Though SF-hRPCs and S-hRPCs transplantation suggests that photoreceptor replacement can occur, there were a few transplanted cells that could migrate and integrate towards the host ONL. The visual function of RCS-p⁺ retina is kept by the maintenance of ONL thickness, suggesting that the survival of photoreceptors is achieved through slowing loss of ONL cell rather than cell replacement.

Our studies demonstrated that S-hRPCs (P2) are composed of retinal neural progenitor cells and retinal MSCs. When injected into the subretinal space of RCS-p⁺ rats, compared with SF-hRPCs, S-hRPCs showed a stronger capacity in maintaining the thickness of ONL and prolonging the functional integrity of the retina in a way similar to human BMSCs. Transplantion of S-hRPCs causing slowed photoreceptor cell loss may be attributed to the improved circulation and secretion of trophic factors essential for photoreceptor survival [4,5,25]. It is reported that organ-matched mesenchyme permits progenitor proliferation and self-renewal in vitro and in vivo [26]. It is commonly believed that MSCs can secrete various neurotrophic factors (e.g., nerve growth factor [NGF], brain-derived neurotrophic factor [BDNF], ciliary neurotrophic factor [CNTF], insulin-like growth factor-1 [IGF-1]), and cytokines to modulate inflammation, augment tissue repair, enhance regeneration and modulate many facets of the immune system [27-30]. Further research is needed to investigate whether there is a synergistic effect of retinal MSCs on retinal neural progenitor cells.

In this study, we demonstrated that SF-hRPCs could maintain their immature, undifferentiated neural retinal progenitor state during culture. S-hRPCs exhibited greater proliferative dynamics and greater capacity to inhibit morphological and functional degeneration when transplanted to the subretinal space of RCS rats compared with SF-hRPCs. We also successfully isolated retinal mesenchymal-like stem cells from the S-hRPCs population, which might be a new source of pluripotent MSCs. Transplantation of retinal MSCs or combination with retinal neural progenitor cells might be a novel therapeutic approach for retinal degeneration.

Disclosure of conflict of interest

None.