Molecular Vision 2013; 19:1925-1936
Received 02 October 2012 | Accepted 10 September 2013 | Published 13 September 2013
1Division of Ocular Biology & Therapeutics, Institute of Ophthalmology, University College London, London, UK; 2NIHR Biomedical Research Centre for Ophthalmology, UCL Institute of Ophthalmology and Moorfields Eye Hospital, London, UK
Correspondence to: Silke Becker, Division of Ocular Biology and Therapeutics, UCL Institute of Ophthalmology, 11-43 Bath Street, London EC1V 9EL; Phone: 020 7608-6927; FAX: 020 7608-6887; email: email@example.com
Purpose: Human Müller glia with stem cell characteristics (hMGSCs) can be induced to express genes and proteins of retinal ganglion cells (RGCs) upon in vitro inhibition of Notch-1 activity. However, it is not known whether expression of these markers is accompanied by acquisition of RGC function. This study investigated whether hMGSCs that express RGC markers also display neural functionality, as measured by their intracellular calcium concentration ([Ca2+]i) responsiveness following neurotransmitter stimulation in vitro.
Methods: Changes in mRNA expression of RGC markers and neurotransmitter receptors were assessed either by conventional or quantitative reverse transcription PCR (RT-PCR), while changes in protein levels were confirmed by immunocytochemistry. The [Ca2+]i levels were estimated by fluorescence microscopy.
Results: We showed that while undifferentiated hMGSCs displayed a profound elevation of [Ca2+]i after stimulation with N-methyl-D-aspartate (NMDA), this was lost following Notch-1 inhibition. Conversely, untreated hMGSCs did not respond to muscarinic receptor stimulation, whereas [Ca2+]i was increased in differentiated hMGSCs that expressed RGC precursor markers. Differentiated hMGSC-derived RGCs, but not undifferentiated hMGSCs, responded to stimulation by nicotine with a substantial rise in [Ca2+]i, which was inhibited by the α4β2 and α6β2 nicotinic receptor antagonist methyllycaconitine. Notch-1 attenuation not only caused a decrease in the gene expression of the Notch effector HES1 and increased expression of RGC markers, but also an increase in the gene and protein expression of α4 and α6 nicotinic receptor subunits.
Conclusions: These observations suggest that in response to Notch-1 inhibition, hMGSCs differentiate into a population of RGCs that exhibit some of the functionality observed in differentiated RGCs.
The presence of Müller glia with stem cell characteristics has been reported in a range of mammalian species, including the adult human retina [1-5]. Evidence for the regenerative ability of Müller glia has been demonstrated in postnatal chick  and rat retinas , and compelling proof for the neurogenic properties of these cells was shown in the zebrafish, in which Müller glia can dedifferentiate, reenter the cell cycle, and differentiate into retinal neurons [8-10]. Although Müller glia with stem cell characteristics (hMGSCs) are present in the adult human retina , they appear to have lost their regenerative ability in situ. However, they can be isolated and grown indefinitely in culture  and retain their capacity to differentiate into retinal neurons in vitro in response to factors identified during development . This makes hMGSCs strong candidates for cell replacement strategies in retinal degenerative conditions characterized by neural cell damage.
Previous investigations have demonstrated that hMGSCs concomitantly display various features of neural stem cells and retinal progenitors, such as the expression of SOX2, PAX6, and CHX10. The presence of markers of differentiated retinal neurons strongly indicates that hMGSCs represent retinal progenitors with the capacity to form retinal neuronal cells , including retinal ganglion cells (RGCs) , the cell type lost in glaucoma. Notch-1, which controls the development of RGCs in the embryonic retina  and which is present in low quantities in Müller glia of healthy retina , has been shown to be highly expressed in hMGSCs . We have previously demonstrated that Notch-1 inhibition in hMGSCs, using the γ−secretase inhibitor N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT), induces these cells to acquire neural morphology and to express gene and proteins characteristic of RGC precursors, including BRN3B, ISL-1, and HUD .
Neural progenitors of the central nervous system express receptors for various neurotransmitters [16,17], which upon binding to ligands induce changes in the membrane potential . However, expression of these receptors is not confined to neurons, and changes in membrane potential do not necessarily lead to a rise in cytosolic calcium ([Ca2+]i), which has been progressively accepted as indicative of neuronal cell function [18,19]. While some neurotransmitter receptors have been identified in neural progenitors, others are exclusively expressed in differentiated neurons , providing a tool for the identification of the maturation stages of neural cells. At the time of optic cup formation, neural retinal progenitors in the ventricular zone express receptors for muscarinic, purinergic, γ-aminobutyric acid (GABA), and glutamatergic systems . These are thought to play a role in the differentiation of retinal progenitors  and their differentiation and function can be used as indicators of retinal neural differentiation.
The nicotinic, glutamatergic, and muscarinic receptor-ligand systems play a significant role in RGC development [22,23]. Since their expression changes throughout various stages of RGC differentiation, they can be examined to identify whether acquisition of markers of RGC-committed precursors by differentiated Müller stem cells is accompanied by expression of RGC functionality. In particular, the expression of nicotinic acetylcholine receptors (nAChR), which are present in retinal stem cells and early retinal progenitors, is greatly upregulated in late retinal progenitors . The expression of different nAChR subunits is likely to be differentially regulated throughout development . Conversely, functional expression of N-methyl-D-aspartate (NMDA) receptors is highest in late retinal precursors [19,22,23] and in mature RGCs  , as well as in Müller glia cells , but not early retinal precursors . Muscarinic receptors, which are only sparsely expressed in early retinal progenitors and Müller glia cells, have been shown to be abundantly expressed in late retinal progenitors [20,26] (Figure 1).
Although these neurotransmitter receptors are also expressed by Müller glia [20,25,27], changes in levels of expression of these molecules by hMGSCs may indicate acquisition of neural function and can be used to estimate the ontogenetic stage of the retinal precursors generated. On this basis, we investigated whether downregulation of Notch-1 in hMGSCs, in addition to inducing phenotypic changes characteristic of RGCs, also leads to neural functionality as judged by an increase in [Ca2+]i in response to selective neurotransmitter stimulation.
An hMGSC line derived in our laboratory and known as MIO-M1 was maintained for up to 40 passages in Dulbecco’s Modified Eagle Medium (DMEM, 1× with GlutaMAX™, without sodium pyruvate; Gibco, Life Technologies, Carlsbad, CA or DMEM high glucose™, PAA laboratories, Pasching, Austria), supplemented with 10% fetal calf serum (FCS, PAA laboratories) as well as 20 U/ml penicillin and 20 μg/ml streptomycin (Gibco, Life Technologies). To passage cells, confluent monolayers were usually detached once a week using TrypLE™Express (Gibco, Life Technologies) and subcultured at a dilution of 1:5 to 1:6.
Differentiation of MIO-M1 cells into RGC precursors was induced as previously described  by culturing cells for 7 days on surfaces coated with 0.5 µg/ml basement membrane protein (BMP, ECM gel from Engelbreth-Holm-Swarm murine sarcoma, Sigma-Aldrich, St. Louis, MO) with 20 ng/ml basic fibroblast growth factor-2 (FGF2, Sigma-Aldrich) in the absence or presence of 50 µM DAPT (Sigma-Aldrich, St. Louis, MO). MIO-M1 cells cultured in the absence of these factors were used as controls.
MIO-M1 cells were grown for 7 days on LAB-TEK™ 8-well chambered coverglasses (Nalge Nunc™, Rochester, NY) to approximately 60% confluency. Cells were divided into three treatment groups, which received either no treatment (control) or which were cultured on BMP-coated surfaces with 20 ng/ml FGF2 in the absence (BMP/FGF2) or presence of 50 µM DAPT (BMP/FGF2/DAPT). Cells were loaded with Fura Red-AM (2 µg/ml, Invitrogen, Life Technologies) in serum-free DMEM for 30 min at 37 °C. Before stimulation with neurotransmitters, cells were serum-recovered in DMEM supplemented with 10% FCS for at least 30 min at 37 °C, to allow for deesterification of the dye, and were subsequently transferred into 200 µl phenol-red free Leibovitz’s medium L-15 (Gibco, Life Technologies). Inhibitors of nicotinic acetylcholine receptors (100 nmoles/l methyllycaconitine citrate hydrate (MLA), Sigma-Aldrich, or 100 nmoles/l α-conotoxin MII (α-CT), Tocris, Bristol, UK) were added in Leibovitz’s medium L-15 (Gibco, Life Technologies) at least 15 min before the addition of nicotine.
Cells were transferred onto the stage of a Leica TCS-SP2 inverted microscope (Leica, Wetzlar, Germany) and kept at 37 °C inside a heated stage chamber. Cells were exposed to excitation light at a wavelength of 488 nm, and emission light of 550–600 nm was recorded. Images were taken at 20× magnification at a frequency of 2 Hz for 75 s and were stored on a PC for off-line analysis. All agonists were dissolved in water (ddH2O) at a concentration of 10 mM, and 50 µl of this solution were added to the bath to achieve a final concentration of 2 mM. Fluorescence intensity during the recording period was measured with LCS Lite software individually for each cell with background subtraction, with the inverted fluorescence intensity used as an estimate of [Ca2+]i. Changes in [Ca2+]i were determined as the percentage of the [Ca2+]i at rest and were averaged for all cells recorded in the experiment. Data are displayed as average [Ca2+]i ± SEM in n separate experiments. The rise in [Ca2+]i was validated by exposure of untreated and differentiated MIO-M1 cells to 2 mM histamine (n=3), which caused a profound reduction in the fluorescence intensity, corresponding to a substantial sustained rise in [Ca2+]i (data not shown). The fraction of cells responding to agonist stimulation with a rise in [Ca2+]i (“responders”) was determined as a percentage of the total number of cells in the field, and is displayed as mean ± SEM.
Cells were grown to approximately 90% confluency in DMEM or DMEM supplemented with growth and differentiation factors (see above), detached with TrypLE™Express, washed in phosphate buffered saline (PBS), and pelleted at 4 °C. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany), including DNase treatment according to the manufacturer’s instructions. The amount of 2 µg RNA was reverse transcribed using the First Strand cDNA Synthesis kit (0.5 µg RNA/10 µl reaction, Roche Applied Science, Penzberg, Germany). The reaction mixture was incubated for 10 min at 25 °C, 60 min at 42 °C, 5 min at 99 °C, and 5 min at 4 °C in a Mastercycler gradient thermal cycler (Eppendorf, Hamburg, Germany).
PCR was subsequently performed using the Expand High Fidelity PLUS PCR System (Roche Applied Science) with specific primers for BRN3B, HES1, nAChR α subunits 1–4, 6, and 7, β-actin and GAPDH (Invitrogen, Life Technologies, see Table 1). PCR mixtures were incubated at 94 °C for 2 min, followed by 24 to 40 cycles of 94 °C for 30 s, 58 °C for 30 s, 72 °C for 1 min and finally held at 72 °C for 5 min. PCR products were analyzed by electrophoresis on 1% agarose gel. Band density was quantified by densitometry using ImageJ software, and results were expressed as the ratio of the band density of the target gene over that of β-actin or GAPDH as the housekeeping gene.
Prior to running quantitative RT-PCR reactions, primer concentrations and amplification conditions were optimized and the presence of a single PCR product was confirmed by agarose gel electrophoresis. The efficiency of PCR product amplification was assumed to be 100%. Quantitative RT-PCR reactions were subsequently performed using SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich) according to the manufacturer’s instructions with 0.4 pmoles/l primers and 2 μl cDNA in a total volume of 25 μl. Twelve repeats were prepared for each experimental condition and transferred to an AB 7900HT Real-Time PCR System (Applied Biosystems, Life Technologies). Quantitative RT-PCR mixtures were incubated at 94 °C for 2 min, followed by 40 cycles of 94 °C for 15 s, 60 °C for 30 s, and 72 °C for 1 min as well as a dissociation stage with a SYBR green detector for quality control of the PCR product. Data analysis was performed using DART-PCR version 1.0 software. The relative expression values (R0) of the test genes were normalized to the R0 values of the housekeeping gene, and the ratios obtained were compared as the fold change in mRNA expression.
MIO-M1 cells were grown for 7 days in DMEM supplemented with 2% FCS on BMP-coated glass coverslips in 24-well plates and divided into two treatment groups, which received either no treatment (control) or treatment with 20 ng/ml FGF2 and 50 µM DAPT (BMP/FGF2/DAPT). After 7 days cells were fixed in 4% paraformaldehyde for 10 min, cryoprotected using 30% sucrose and stored at −20 °C until required. After defrosting, slides were blocked for 1 h at room temperature using either Roche blocker (0.5% Blocking Solution, Roche Applied Science, with 5% donkey serum) or Tris-buffered saline (TBS) with 0.3% triton and 5% donkey serum. Primary antibodies (see Table 2) were diluted in blocking serum and incubated overnight at 4 °C. Primary antibody labeling was detected using donkey antigoat or antimouse antibodies labeled with Alexa Fluor 488 (1:500, Molecular Probes, Invitrogen) for 3 h at room temperature. 4',6-diamidino-2-phenylindole (DAPI; 1 μg/ml, Sigma-Aldrich) was used to counterstain the cell nuclei, and slides were mounted with Vectashield (Vectashield, Vector Laboratories, Burlingame, CA). Fluorescent images were captured with identical exposure times using a Zeiss LSM710 confocal microscope and identically processed using Carl Zeiss Zen imaging software (Carl Zeiss Microscopy GmbH, Jena, Germany).
Results are expressed as means ± standard error of the mean (SEM), and statistical significance was determined by one-way repeated measures ANOVA with differences considered to be significant for p<0.05.
Acquisition of RGC precursor markers in hMGSCs triggered by the attenuation of Notch-1 activity was confirmed using RT-PCR and quantitative RT-PCR. Figure 2A illustrates that treatment of hMGSCs with FGF2 and DAPT on BMP-coated surfaces induced strong mRNA downregulation of the Notch-1 effector HES1 (***p<0.001) and upregulation of the RGC marker BRN3B (**p<0.01), as demonstrated by quantitative and conventional RT-PCR, respectively. Ct values for quantitative RT-PCR of HES1 are shown in Table 3.
As shown in Figure 2B immunocytochemical staining demonstrated high intracellular expression of the Müller cell markers vimentin and CRALBP in undifferentiated hMGSCs, while the RGC and neuronal markers THY-1, βIII-tubulin, and ISL-1 were not detectable or expressed at very low levels (upper panel). Conversely, after differentiation of hMGSCs by Notch-1 inhibition using BMP, FGF2, and DAPT, protein expression of vimentin and CRALBP was greatly attenuated, while THY1, βIII-tubulin, and ISL-1 were shown to be strongly upregulated (lower panel), indicative of a change from a glial to a retinal neuronal phenotype.
Figure 3A shows that following Notch-1 inhibition in hMGSCs, the average rise in [Ca2+]i in response to NMDA receptor stimulation was largely decreased, compared with untreated cells (***p<0.001). There was no significant difference in the average [Ca2+]i response for cells treated with BMP and FGF2 alone or in the presence of DAPT. As demonstrated in Figure 3B, the proportion of cells responsive to NMDA stimulation was greatly reduced after Notch-1 inhibition (*p<0.05). There was no significant difference in the percentage of cells responsive to NMDA for cells treated with BMP and FGF2 alone or in the presence of DAPT.
Similarly, we investigated the effect of the muscarinic receptor agonist McN-A343 (2 mmoles/l) on [Ca2+]i in undifferentiated and differentiated hMGSCs. Figure 4A shows that undifferentiated MIO-M1 cells responded to muscarinic receptor activation with a small transient rise in [Ca2+]i. This response was substantially augmented following treatment with BMP and FGF2 alone (*p<0.05), but in comparison to control cells, remained unaltered in the in the presence of BMP, FGF2 and DAPT (p>0.05). Conversely, Figure 4B illustrates that the proportion of undifferentiated or differentiated cells, which responded to treatment with McN-A343 with an augmentation of [Ca2+]i, was not statistically affected (p>0.05).
The [Ca2+]i responsiveness to nicotine has previously been shown to be greatly augmented in hMGSCs after differentiating treatment with BMP, FGF2, and DAPT . As demonstrated in Figure 5A, in undifferentiated hMGSCs mRNA for the nAChR subunits α1–4, 6, and 7 was detected in low quantities. After blocking Notch-1 activity with BMP, FGF2, and DAPT treatment, mRNA for the α4 and α6 nAChR subunits was demonstrated by conventional RT-PCR to be significantly upregulated in comparison to untreated hMGSCs (*p<0.05, n=6). Consistently, Figure 5B shows increased immunocytochemical staining for the α4 and α6 nAChR subunits, concomitantly with upregulation of the RGC marker THY1, in RGC-committed precursors.
The [Ca2+]i responsiveness to nicotine was confirmed by inhibition of nicotinic receptors with pharmacological antagonists. To demonstrate that upregulated nAChR subunits form functional ion channels and that an increase in [Ca2+]i responsiveness to nicotine occurred following Notch-1 inhibition with BMP, FGF2, and DAPT treatment, we tested the effects of a pharmacological inhibitor of the nAChR α3β2, α-conotoxin MII (α-CT, 100 nmoles/l), and methyllycaconitine (MLA, 100 nmoles/l), which blocks α4β2, α6β2, and α7β2 nAChR activation. Figure 6 demonstrates that the [Ca2+]i responsiveness to nicotine after Notch-1 inhibition was greatly attenuated by methyllycaconitine to a level similar to untreated hMGSCs (***p<0.001 n=3), while α-conotoxin MII had no significant effect on [Ca2+]i at the concentration used (n=4).
The human retina is not known to regenerate or repair after disease or injury. However, it harbors a population of Müller glia with stem cell characteristics, similar to those observed in lower vertebrates, which have been demonstrated to repair and regenerate after retinal damage [5,12]. In vitro, hMGSCs have recently been shown to differentiate toward RGCs, the cell type predominantly damaged in glaucoma . We have previously reported that intravitreal transplantation of cells with RGC phenotype in rats, depleted of RGCs by NMDA injection, are capable of partially restoring RGC function . Although this effect may to some extent be ascribed to neurotrophic factors, it is possible that it can also be attributed to the neural function of the transplanted cells. Transplantation success may be affected by cell-fate determination and the ontogenetic stage of the graft . It is therefore vital that the phenotype of potential candidates for cell transplantation be thoroughly investigated. In the present study, we have investigated the functional characteristics, as well as the ontogenetic stage, of hMGSC-derived RGCs.
hMGSCs express high levels of Notch-1 , which has been shown in developmental studies to maintain the progenicity of neural stem cells . Since generation of RGCs requires the cessation of Notch-1 activity in the embryonic retina [14,30], we have used Notch-1 inhibition to differentiate hMGSCs toward RGCs, as previously reported . To confirm that we were examining an RGC population, we examined the expression of RGC markers following Notch-1 inhibition. As previously noted, the expression of its downstream effector HES1 was markedly downregulated , while the early RGC marker BRN3B was greatly increased. Following Notch-1 inhibition, the expression of the RGC markers ISL-1, THY-1, and βIII-tubulin were greatly augmented with concomitant downregulation of vimentin and CRALBP. These findings substantiate our previous reports that cells expressing neural morphology and RGC markers can be generated from hMGSCs by Notch-1 inhibition .
We subsequently studied the functional properties of hMGSC-derived retinal cells committed to an RGC fate by investigating their [Ca2+]i responsiveness to various key neurotransmitters, as well as changes in their expression of neurotransmitter receptors. Although it was initially thought that only neurons contained functional neurotransmitter receptors, it is now generally accepted that while some receptors are exclusively expressed by mature neurons , other cell types in the central nervous system, including glial cells, can also respond to certain neurotransmitters [31-33]. In addition, the expression of neurotransmitter receptors in neural precursors varies depending on their developmental stage [16,17]. Müller glial cells and retinal neurons respond to certain neurotransmitters, and these cell types can nevertheless be distinguished by selective agonists, which activate distinct neurotransmitter receptors. Functional expression of nicotinic, NMDA-sensitive glutamatergic, and muscarinic receptors were investigated in this study, since these are differentially expressed in Müller glial cells, RGC precursors, and differentiated RGCs.
While low expression of NMDA receptor subunits has been reported in early retinal development [20,34], this is greatly upregulated during the late stages of retinal development or in the early postnatal period. Concomitantly, a functional response of RGCs and their precursors to the stimulation of glutamatergic receptors with NMDA has been reported to emerge [19,35-37]. In addition, NMDA receptors are highly expressed by mature RGCs , as well as by Müller glia . In the present study, we have shown that the [Ca2+]i responsiveness to the glutamatergic neurotransmitter receptor agonist NMDA is greatly diminished in hMGSCs following Notch-1 inhibition. This change in sensitivity to NMDA receptor activation after treatment with BMP and FGF2 alone or in combination with the γ-secretase inhibitor DAPT is consistent with hMGSCs acquiring a functional phenotype akin to RGCs. Combined treatment of BMP and FGF2 with DAPT generated cells with RGC phenotype, as this has been shown to promote the expression of molecular markers of newly born RGCs, such as BRN3B.
Muscarinic acetylcholine receptors are expressed at a low level in undifferentiated retinal progenitor cells during early embryogenesis, although they are upregulated during later stages of retinal development . Electrophysiological studies have provided evidence that these receptors form functional ion channels in retinal progenitors and can facilitate an increase in [Ca2+]i in response to muscarinic receptor stimulation [17,38,39]. Conversely, Müller glia display low sensitivity to muscarinic receptor agonists . In the present study, partial inhibition of Notch-1 activity with BMP and FGF2 caused a significant increase in the [Ca2+]i response to muscarinic receptor stimulation, which suggests a change in the phenotype toward RGCs. On the other hand, differentiation of hMGSCs with a combination of BMP, FGF2, and DAPT, which leads to attenuation of Notch-1 activity , resulted in no significant augmentation in the [Ca2+]i response to McN-A343 in comparison to control cells. This finding is consistent with the hypothesis that complete Notch-1 inhibition with BMP, FGF2, and DAPT promotes the development of an earlier retinal progenitor phenotype, which does not express functional muscarinic receptors and may be capable of forming RGC progenitors.
Although mRNA expression of the downstream effector of Notch-1, HES1, was markedly attenuated by treatment with BMP and FGF2 alone or in combination with DAPT, this was not reflected by the different patterns of muscarinic receptor responsiveness in these cells. In fact, while treatment with BMP and FGF2 significantly increased the [Ca2+]i responsiveness to muscarinic receptor activation, this was attenuated by Notch-1 inhibition with BMP, FGF2, and DAPT. Although beyond the scope of the present study, it may be speculated that other downstream effectors of Notch-1, such as HES5 , may be differentially affected by Notch-1 inhibition with BMP, FGF2, and DAPT and contribute to the differentiation of Müller glia cells with stem cell characteristics toward RGCs.
One of the hallmarks of neurons is their responsiveness to nicotinic receptor activation , which is characteristic of differentiated RGCs , but not undifferentiated precursors committed to an RGC fate . Although Müller glia cells have been reported to express nAChR subunits  and respond to stimulation with increased [Ca2+]i, their sensitivity appears to be much lower than in retinal neurons . During development, RGC expression of nicotinic receptor subunits is first detected in early differentiating RGCs . In the present study, undifferentiated hMSCs and those treated with BMP and FGF2 alone remained largely unresponsive to the stimulation of nAChRs. However, following attenuation of Notch-1 activity with BMP, FGF2 and the γ-secretase inhibitor DAPT, stimulation of nAChRs with nicotine elicited a profound increase in [Ca2+]i, which may be ascribed to increased expression of the α4 and α6 nicotinic receptor subunits, as indicated by the increase in gene and protein expression observed in this study. Our findings were further substantiated by pharmacological intervention with the α4, α6 and α7 nicotinic receptor blocker methyllycaconitine, which greatly attenuated the [Ca2+]i responsiveness of differentiated MIO-M1 cells to nicotine. Since mRNA levels for the α7 nicotinic receptor subunit were not shown to be significantly altered, it is likely that the augmentation of [Ca2+]i by nicotine in differentiated hMSCs was largely mediated by the increased levels of α4 and α6 receptors. In addition, the α3β2 blocker α-conotoxin MII did not significant alter the rise in [Ca2+]i in response to nicotine, demonstrating the selectivity of its attenuation by methyllycaconitine.
Although Notch-1 inhibition with BMP, FGF2, and DAPT altered the [Ca2+]i responsiveness of hMGSCs toward a phenotype consistent with postmitotic RGCs, not all cells appeared to respond equally to neurotransmitter stimulation. This likely reflects the generation of a heterogenous cell population, which may also contain early and late retinal progenitors, although the predominant cell type obtained was functionally similar to RGCs.
Taken together, the results of our study strongly suggest that hMGSCs differentiated with BMP, FGF2, and DAPT generate functional newly born RGCs, a developmental stage that may be advantageous for successful engraftment . The present study therefore provides important evidence that hMGSCs can be differentiated in vitro into functional RGCs, with potential application in cell-based therapies for treating conditions in which RGCs are compromised.
This work was supported by the Medical Research Council (Grant No G0900002), Fight for Sight (Grant Ref 1946- donation by Mr Tony Bickford), the NIHR Biomedical Research Centre at Moorfields Eye Hospital and the UCL Institute of Ophthalmology (Grant Ref BRC-022), and The Wellcome Trust Institutional Strategic Support Fund (Grant Reference 097815/Z/11/Z).