Effects of Adult Müller Cells and Their Conditioned Media on the Survival of Stem Cell-Derived Retinal Ganglion Cells

Retinal neurons, particularly retinal ganglion cells (RGCs), are susceptible to the degenerative damage caused by different inherited conditions and environmental insults, leading to irreversible vision loss and, ultimately, blindness. Numerous strategies are being tested in different models of degeneration to restore vision and, in recent years, stem cell technologies have offered novel avenues to obtain donor cells for replacement therapies. To date, stem cell–based transplantation in the retina has been attempted as treatment for photoreceptor degeneration, but the same tools could potentially be applied to other retinal cell types, including RGCs. However, RGC-like cells are not an abundant cell type in stem cell–derived cultures and, often, these cells degenerate over time in vitro. To overcome this limitation, we have taken advantage of the neuroprotective properties of Müller glia (one of the main glial cell types in the retina) and we have examined whether Müller glia and the factors they secrete could promote RGC-like cell survival in organoid cultures. Accordingly, stem cell-derived RGC-like cells were co-cultured with adult Müller cells or Müller cell-conditioned media was added to the cultures. Remarkably, RGC-like cell survival was substantially enhanced in both culture conditions, and we also observed a significant increase in their neurite length. Interestingly, Atoh7, a transcription factor required for RGC development, was up-regulated in stem cell-derived organoids exposed to conditioned media, suggesting that Müller cells may also enhance the survival of retinal progenitors and/or postmitotic precursor cells. In conclusion, Müller cells and the factors they release promote organoid-derived RGC-like cell survival, neuritogenesis, and possibly neuronal maturation.


Introduction
Degenerative diseases of the retina are one of the main causes of irreversible vision loss. Notably, retinal ganglion cell (RGC) death is the common hallmark of several ocular conditions, including glaucoma, Leber s hereditary optic neuropathy (LHON), and other optic neuropathies [1][2][3]. RGCs are the only output neurons of the retina as their axons form the optic nerve and connect the retina with the brain. As a result, RGC degeneration can lead to vision loss and, in fact, glaucoma is one of the leading causes of blindness worldwide [4]. Although RGC loss is the cornerstone in the management of glaucomatous degenerations, the only FDA-approved treatments are designed to lower intraocular Fisher, Carlsbad, CA, USA), non-essential amino acids (NEAAs) (Thermo Fisher, Waltham, MA, USA), sodium pyruvate (Thermo Fisher, Waltham, MA, USA), 0.1 mM β-mercaptoethanol (Sigma Aldrich, Saint Louis, MO, USA), 100 µL of LIF (10 million units/mL: ESGRO, Millipore, Billerica, MA, USA), 3 µM of a GSK3β inhibitor (Stemgent, Cambridge, MA, USA), and 0.4 µM of a MEK inhibitor (Stemgent). All undifferentiated cells were maintained in feeder-free conditions using growth factor-reduced Matrigel-coated plates (GFR-Matrigel was from Corning, New York, NY, USA). Only low-passage (<passage 30) cultures were used for these experiments.

Retinal Differentiation of mESCs
Retinal differentiation of mESCs was performed following a previously defined protocol, with minor variations [9].

Adult Müller Cell Cultures
Mice were sacrificed by cervical dislocation and their eyes were enucleated. The corneas, crystalline, lens, and the vitreous bodies were removed, and the retinas were carefully extracted in fresh DMEM/-CO 2 medium. The retinas were then cut into small fragments and they were incubated at 37 • C for 30 min in a Sterile Earle's Balanced Salt Solution (EBSS) containing Papain (20 U/mL) and DNase (2000 U/mL: Worthington, Lakewood, NJ, USA). Enzyme digestion was stopped by adding Ovomucoid, according to manufacturer's instructions, and the tissue was then mechanically dissociated. Isolated cells were recovered by centrifugation at 1200 rpm for 5 min. The pelleted cells were resuspended in DMEM + 10% FBS, seeded onto sterile 13 mm glass coverslips in 24 well plates, coated with poly-l-lysine (100 µg/mL: Sigma-Aldrich) and laminin (10 µg/mL: Sigma-Aldrich), and further cultured in DMEM + 10% FBS. All the cells recovered from one mouse retina were seeded in one well. The cultures were maintained in a humidified incubator at 37 • C in an atmosphere of 5% CO 2 . The media was replaced with fresh media on day 1 of culture and subsequently, every 3 days. These cultures reached confluence after 7 days in vitro (DIV).

Co-Cultures of Stem Cell-Derived RGC-Like Cells and Müller Cells.
Day 10-15, EBs were collected and mechanically dissociated with Accutase (Stem Cell Technologies), centrifuged at 1200 rpm for 2 min and resuspended in DMEM + 10% FBS. Dissociated EBs were seeded on a monolayer of Müller cells and maintained for 3 days in one of three different media: (1) Neurobasal supplemented with B27 and N2; (2) Neurobasal supplemented with B27, N2, and 10% FBS; or (3) DMEM supplemented with 10% FBS. Dissociated EBs seeded onto sterile poly-l-lysine and Laminin-coated 13 mm glass coverslips in 24 well plates were used as controls.

Müller Cell Conditioned Media Collection
Conditioned media (CM) was collected when Müller cell cultures had reached confluence (day 7), first washing the wells three times with DMEM medium supplemented with 1% l-glutamine and 0.1% gentamicin (Thermo-Fisher Waltham, MA, USA). DMEM was then added to each well and left for 3 h before the medium was changed to eliminate the rest of the FBS. Fresh DMEM was then added for 2 days before it was collected and sterilized by passing through a 0.22 µm filter. The CM was frozen in aliquots at −20 • C, and the Müller cells that produced it were fixed for 10 min with methanol at −20 • C.

Application of Conditioned Media
EBs dissociated on day 10 were plated (10 EBS/well) on 13-mm poly-l-lysine and Laminin coated glass coverslips in 24-well plates to test the activity of the CM. The cultures were maintained in a mixture of 50% Neurobasal medium supplemented with 2% B27 and N2, 1% l-glutamine (Thermo-Fisher) and 0.1% gentamicin, and 50% CM. The cells were cultured for 3 days at 37 • C in a humidified atmosphere containing 5% CO 2 , and the medium was changed every 2 days. Cells maintained in Neurobasal medium supplemented with 2% B27 and 1% l-glutamine were used as controls. The cells were fixed for 10 min with ice-cold methanol at −20 • C on day 3.
Undissociated day 10 EBs were also maintained in a mixture of 50% Tom's and 50% CM for 3 days, using EBs maintained in Tom´s medium alone as controls. The EBs were then collected for RNA extraction and at least three replicates of each culture were made, performing the procedure in triplicate.

Immunocytochemistry
After 3 days in co-culture or in the presence of CM, the cells were fixed in cold methanol (−20 • C) and washed with PBS (pH 7.0). After blocking non-specific antigens with blocking buffer (10% NGS and 0.1% Triton X-100 in PBS), the cells were incubated with primary antibodies at a dilution indicated in Table 1 at 4 • C overnight. The next day, the cells were extensively washed with PBS and incubated with Alexa Fluor secondary antibodies as indicated in Table 1. Subsequently, the cells were washed with PBS, the cell nuclei were labeled with DAPI at a dilution of 1:10,000, and the coverslips were then mounted with Fluoromount-G (Southern Biotech, Birmingham, AL, USA).

qPCR
Total RNA was extracted from the whole EBs maintained in CM using Trizol (Invitrogen) and chloroform extraction, according to the manufacturer s instructions. The RNA was digested with DNase1 (Qiagen, Hilden, Germany), cleaned using the Qiagen RNA mini clean-up kit and reverse transcribed into cDNA using the Superscript III RT kit (Invitrogen) following the manufacturer s instructions. qPCR was performed using the primers indicated in Table 2. Table 2. List of qPCR primers (5 to 3 ).

Quantification and Statistical Analysis of RGC-Like Cells
Fluorescent images of RGC-like cells and Müller cells were taken on a Leica DM 5000 M fluorescence microscope with a Leica DFC 500 camera using the same exposure times to compare control and treatment conditions. The total number of RGC-like cells per coverslip was quantified using the specific markers indicated for each experiment. The RGC-like cells and Müller cells in co-cultures were counted, and statistical analyses were carried out using the IBM SPSS Statistics software v.21-0 and the homogeneity of the variances was assayed with the Levene s test. A Mann-Whitney U test or ANOVA were used to assess whether there were significant differences between the groups. The minimum value of significance for both tests was defined as p < 0.05. At least four complete coverslips and three independent experiments were analyzed for each experimental condition.

Mouse Embryonic Stem Cells can be Differentiated into RGC-Like Cell Fates
R1 undifferentiated mouse embryonic stem cells (mESCs) were directed towards retinal cell fates following a protocol originally described by Yoshiki Sasai's laboratory [12], with some modifications from our previously described method [39]. Upon differentiation, free-floating mESCs formed EBs that developed a neural epithelium that was apparent by day 3 as a well-defined clear layer in the outer part of each EB ( Figure 1A). By day 5, the neuroepithelial layer evaginated and formed optic vesicle-like structures. These optic regions increased in size and thickness in the subsequent culturing days ( Figure 1A). To further confirm that the mESCs were directed towards retinal cell fates, we performed RT-qPCR analyses. At day 6, several retinal progenitor genes (Pax6, Lhx2 and Atoh7) were highly upregulated compared to the undifferentiated samples ( Figure 1B, upper panels). Atoh7/Math5 is normally expressed in a subpopulation of retinal progenitor cells in the developing retina and is required for RGC and optic nerve development in mouse and humans [40][41][42]. Thus, the observed expression in the organoid cultures suggests that the retinal progenitors in the EBs acquired the competence to generate RGC-like cells at these early differentiation stages. By day 8, a significant increase in the expression of the RGC-specific genes Brn3a (Pou4f1), Brn3b (Pou4f2) and Brn3c (Pou4f3) was detected ( Figure 1B, lower panels). Stem cell-derived RGC-like cells were dissociated from EBs cultured for 10−15 days and, as shown in Figure 1C, a pan-Brn3 antibody abundantly co-labeled β-III-tubulin+ neurons resembling RGCs (white arrows in Figure 1C).

Adult Mouse Müller Glia Cells Cultured for 7 Days Express Normal Müller Glia Markers
We have recently established a method to purify and culture adult Müller glia cells from mouse, rat and pig retinas, and we have previously shown that cultured murine Müller glia cells express GFAP, Glutamine Synthetase (GS), p75 NTR , and Cralbp. Flow cytometry indicated that >94% of all the cells in these cultures express p75 NTR [43]. In order to further characterize our Müller glia culture system, we used a combination of markers normally expressed in Müller glia cells, including the cytoskeletal protein Vimentin (Figure 2A,A', green) and the transcription factors Lhx2, Pax6, and Sox2 ( Figure 2B-D, red). Negative controls for each secondary antibody employed were also performed in order to verify the specificity of the primary antibodies. As expected, cultured Müller glia cells expressed high levels of Vimentin, and variable but detectable levels of Lhx2, Pax6 and Sox2 were observed in Müller glia nuclei.

Co-Culturing Stem Cell-Derived Cells with Adult Müller Glia Highly Increased RGC-Like Cell Survival
In order to assess the possible neuroprotective effects of Müller glia on stem cell-derived RGC-like cells, dissociated EBs were co-cultured on a monolayer of adult mouse Müller cells for 3 days in three different media: Neurobasal medium supplemented with B27 and N2 with or without 10% FBS (more favorable to neurons), and DMEM + 10% FBS (more favorable to glia, Figure 3). Neurobasal is a medium optimized for neuronal cell culture [44], and the use of Neurobasal instead of DMEM has been shown to be essential for good survival of purified RGCs [45]. Similarly, our previous data indicated that Müller glia cells grow better when cultured in DMEM supplemented with 10% FBS. Given that the ideal culturing conditions for RGCs and Müller glia are different, we tested different media compositions to address what is the best culturing condition to support RGC survival when neurons and glia are combined in a co-culture system. Interestingly, when stem cell-derived cells were grown on Müller cells, RGC-like cell survival was enhanced in all the different media tested in comparison to the control conditions (the absence of Müller cells). On average, RGC-like cell survival increased 8.11 fold ± 0.25 (mean ± std. deviation) when the cells were cultured in DMEM + 10% FBS (p = 0.00086), 7.5 fold ± 0.26 (mean ± std. deviation) when cultured in Neurobasal + B27 + N2 + 10% FBS (p = 0.000109), and 10.97 fold ± 0.68 (mean ± std. deviation) when cultured in Neurobasal with B27 and N2 (p = 0.00001, Figure 3E). While we detected clear differences between the control cells and the co-cultures, there were no statistical differences between the different culture conditions assayed (DMEM vs. Neurobasal + 10% FBS p = 0.3482, DMEM vs. Neurobasal p = 0.2392, Neurobasal vs. Neurobasal + 10% FBS p = 0.86).

Müller Glia Conditioned Media (CM) Enhances RGC-Like Cell Survival
To analyze if the factors secreted by adult Müller cells have the same effect on the survival of stem cell-derived RGC-like cells similar to the co-cultures, dissociated EBs were cultured with adult mouse Müller cell CM. RGC-like cells cultured in DMEM + 10% FBS (control) were compared to those maintained in Neurobasal medium supplemented with B27 and N2 (Neurobasal control) or to cells maintained in 1:1 Neurobasal medium B27:Müller cell CM. Notably, we observed significantly more RGC-like cells when the cells were maintained in the presence of Müller cell CM. Exposing the cultured stem cell-derived cells to the Müller cell CM enhanced the survival of RGC-like cells 1.85 ± 0.33 folds (mean ± std. deviation, p-value: 0.00174), while there were no significant differences between DMEM and Neurobasal culturing conditions (p-value: 0.715, Figure 4).

Müller Glia Conditioned Media (CM) Increases Atoh7 and Brn3b Expression
The effect of the Müller cell CM on the survival of stem cell-derived RGC-like cells was also assessed in non-dissociated EBs by analyzing the expression of Brn3b and Atoh7 by RT-qPCR. Brn3b (Pou4f2) is a protein expressed in most RGCs [46][47][48][49][50], while Atoh7 is a transcription factor that is transiently expressed during early retinal histogenesis and is necessary for RGC development [40,51,52]. Both Brn3b and Atoh7 expression were enhanced when EBs were cultured with adult Müller cell CM; Atoh7 was upregulated 6.14-fold (p = 0.0309) and Brn3b 4.12-fold (p = 0.021, Figure 5).

Co-Culturing RGC-Like Cells with Müller Glia Cells as Well as Müller Glia Conditioned Media (CM) Increases Neuritogenesis
Previous data have shown that Müller glia can enhance neuritogenesis of cultured adult RGCs [32,37]. To assess whether a similar effect was also elicited on stem cell-derived RGC-like cells, we quantified the neurite length of the longest neurite of cells cultured in control conditions (DMEM + 10% FBS), cells exposed to CM (1:1, DMEM + 10% FBS: CM), and RGC-like cells co-cultured with Müller glia. To avoid including dying cells in the quantifications, all experiments were co-stained with DAPI and we excluded from the counts any cells exhibiting pyknotic nuclei or axonal fragmentation. Remarkably, both CM and co-culturing conditions significantly increased neurite length (Figure 6). At least, 70 individual neurites were measured from three different biological replicates for each condition. In control conditions, the longest neurite grew 129.7 ± 112 µm (mean ± std. deviation); the cells exposed to Müller glia-conditioned media extended neurites that measured 270.6 ± 155 µm (mean ± std. deviation), while the RGCs co-cultured with Müller glia exhibited significantly longer neurites that grew 434.3 ± 262 µm (mean ± std. deviation), with the longest neurites measuring over 1 mm.

Discussion
In recent years, stem cell therapies have become a promising strategy to develop novel treatments for retinal degenerations. As stem cell protocols have advanced, derivation methods and utilization of specific culturing conditions have been established to set the proper conditions for controlled and directed differentiation towards retinal cell fates, including RGC-like cells [11,15,20,39,53,54]. Our data indicates that, in culture, organoids acquire the competence to generate RGC-like cells at early stages of differentiation as retinal progenitor genes are upregulated from day 6 ( Figure 1B) and RGC-associated genes (Brn3a, b and c) are increased from day 8, mimicking the normal developmental progression of the embryonic retina [55][56][57]. Surprisingly, Brn3b initially exhibits a decline in expression (day 6 in vitro, Figure 1B). While most of the commonly used retinal markers are not expressed in undifferentiated stem cells, it has been previously shown that Brn3b is expressed in germ cells and stem cells at the earliest stages of embryonic development [58], and thus, the initial decline probably reflects the ongoing differentiation from Brn3b+ undifferentiated cells.
Stem cell-derived RGC-like cells have been previously characterized by their molecular signatures and physiological properties [23,24,59,60]. These experiments validate that the cells made from organoids in vitro are remarkably similar to endogenous RGCs even though differences in their maturation status have been reported [61]. Correspondingly, retinal organoids yield different RGCs subtypes, can fire action potentials, and respond to normal chemoattractant and chemorepellent cues [23,62,63]. However, within 3D retinal organoid cultures, RGC-like cells die over time [29] and, in fact, these cells are not observed in mouse retinal organoids beyond differentiation day 25 [64]. This currently unavoidable cell loss is likely the result of two main factors: RGC-like cell number in culture is probably governed by the same programmed cell death processes that occur during normal development [65,66], and a failure to locate brain targets and establish normal synaptic connections in vitro. Another factor driving RGC-like cell degeneration in culture may be a technical limitation: Organoids grown in free-floating 3D cultures have limited access to oxygen and nutrients. This is in stark contrast to a living animal with constant blood flow to replenish all tissues and cells with nutrients and oxygen. Regardless of the molecular causes of the degeneration, novel strategies must be devised to increase organoid-derived RGC survival before we can translate these technologies to clinical applications.
Glia are typically considered the support cells of the nervous system due to their critical functions in providing structural and metabolic support, regulating the blood-brain barrier, and controlling synaptic modulation and tissue homeostasis [67]. One of the most abundant glial types in the retina, the Müller glia, expands across all the layers of the retina and it has been well-characterized that these cells can enhance the survival of different retinal neurons [68]. Importantly, Müller glia functions have been shown to increase survival of RGCs [32,37,69]. For instance, Müller glia secrete neurotrophic factors like BDNF [70], CTNF [71], basic fibroblast growth factor (bFGF) [72], pigment epithelium-derived growth factor (PEDGF) [73], and glial-derived neurotrophic factor (GDNF) [74]. Classic experiments have shown that these trophic factors can promote RGC survival in vitro [36,75,76] and in experimental models of RGC damage [77][78][79]. In addition to these roles, Müller glia are also known to phagocytose cell debris in physiological and pathological conditions [80][81][82]. Given the strong neuroprotective effect of Müller glia on neurons, we have developed a strategy to use adult Müller cells to enhance the survival of stem cell-derived RGC-like cells during retinal differentiation.
Notably, we have shown that co-culturing RGC-like cells derived from stem cells together with Müller glia greatly enhances survival and neuritogenesis. Similarly, dissociated EBs cultures were maintained with medium conditioned by Müller cells to assess whether the increase of organoid-derived RGC-like cell survival was due to factors released by the Müller cells. The conditioned medium also significantly increased the survival and neuritogenesis (Figures 4 and 6), indicating that the effect of Müller cells is not mediated by substrate alone as we have previously observed [32,38]. However, as the effects of co-culturing cells on the survival of stem cell-derived RGC-like cells were greater than that observed with CM alone, there appears to be a synergistic effect of membrane bound and diffusible factors. It is also possible that the half-life of factors necessary to significantly improve the health of stem cell-derived RGCs is short, and the presence of Müller glia that constantly replenish these factors increases the beneficial effect these factors have on RGCs in vitro. Additionally, Müller glia cells can elicit phagocytic functions in coordination with the microglia, often considered the "professional" phagocytic cells of the retina [83][84][85]. During development, the microglia originates in the primitive yolk sac [86] and infiltrate in the retina during early development. Since they have a different embryonic origin, microglia cells are not present in stem cell-derived retinal organoids as the stem cells are first directed towards ectodermal fates. Importantly, it has been suggested that the phagocytic activity of Müller cells becomes more relevant in degenerative situations where the microglia are absent [87]. By clearing debris and removing cellular corpses, Müller glia may prevent subsequent damage to other neurons in stem cell-derived cultures.
In a series of elegant experiments, the Meyer lab has recently shown that co-culturing hiPCS-derived RGC-like cells with astrocytes, another important glial type of the retina, increased RGC neurite length and complexity, and improved RGC maturation, including a significant enhancement of the RGC electrophysiological properties [61]. Overall, the neurite length in control conditions measured in this study as well as the effect elicited by co-culturing with astrocytes are comparable to our results even though we are using mouse cells. In contrast, astrocyte-conditioned media showed no effects in RGC neuritogenesis. In our experiments, Müller glia CM not only increased neuritogenesis but also enhanced the expression of Brn3b in non-dissociated EBs, suggesting an enhanced survival of stem cell-derived RGCs in non-dissociated EBs ( Figure 5), similar to the survival detected in isolated cells. We also assessed the expression of Atoh7, a transcription factor that is transiently expressed during early retinal histogenesis and that is necessary for RGC differentiation [40,52] or survival [88]. Our results show that Atoh7 expression is enhanced in EBs grown in CM, suggesting that the factors released by Müller cells not only enhance the survival of RGCs but also they increase the survival of the progenitor cells that can differentiate into RGCs.
In conclusion, Müller cells release factors that enhance the survival of stem cell-derived RGC-like cells and also contribute to neuritogenesis and maturation. These findings could improve stem cell-based transplant strategies currently under development. In future studies, the role of Müller glia should be considered when developing in vitro models using RGCs. Moreover, given the important role of Müller glia in RGC survival, future studies should also shed light on the molecular relationships between the Müller glia and the RGCs in pathological conditions, perhaps opening novel targets for therapeutic strategies for the treatment of glaucoma and other RGC degenerations.