Retinal Pigment Epithelium and Müller Progenitor Cell Interaction Increase Müller Progenitor Cell Expression of PDGFRα and Ability to Induce Proliferative Vitreoretinopathy in a Rabbit Model

Purpose. Proliferative vitreoretinopathy (PVR) is a complication of retinal detachment characterized by redetachment of the retina as a result of membrane formation and contraction. A variety of retinal cells, including retinal pigment epithelial (RPE) and Müller glia, and growth factors may be responsible. Platelet-derived growth factor receptor alpha (PDGFRα) is found in large quantities in PVR membranes, and is intrinsic to the development of PVR in rabbit models. This study explores the expression of PDGFR in cocultures of RPE and Müller cells over time to examine how these two cell types may collaborate in the development of PVR. We also examine how changes in PDGFRα expression alter Müller cell pathogenicity. Methods. Human MIO-M1 Müller progenitor (MPC) and ARPE19 cells were studied in a transmembrane coculture system. Immunocytochemistry and Western blot were used to look at PDGFRα, PDGFRβ, and GFAP expression. A transfected MPC line cell line expressing the PDGFRα (MIO-M1α) was generated, and tested in a rabbit model for its ability to induce PVR. Results. The expression of PDGFRα and PDGFRβ was upregulated in MIO-M1 MPCs cocultured with ARPE19 cells; GFAP was slightly decreased. Increased expression of PDGFRα in the MIO-M1 cell line resulted in increased pathogenicity and enhanced ability to induce PVR in a rabbit model. Conclusions. Müller and RPE cell interaction can lead to upregulation of PDGFRα and increased Müller cell pathogenicity. Müller cells may play a more active role than previously thought in the development of PVR membranes, particularly when stimulated by an RPE-cell-rich environment. Additional studies of human samples and in animal models are warranted.


Introduction
Proliferative vitreoretinopathy (PVR) occurs in 5-10% of rhegmatogenous retinal detachments [1]. It is a complex cellular process characterized by preretinal and subretinal membrane formation, intraretinal degeneration, gliosis, and contraction. The disease is characterized by (1) migration and proliferation of retinal pigment epithelial cells (RPE) and glial cells along with synthesis of extracellular matrix (ECM) proteins, such as collagen or fibronectin, which organize into retinal and vitreous membranes; (2) intraretinal glial cell proliferation, photoreceptor degeneration, and disorganization of retinal cell layers [2,3]. In a way, PVR can be viewed as maladaptive and/or aberrant wound healing [4], the severity of which is often determined by the circumstances in which it occurs.
Certain clinical characteristics are associated with an increased risk of PVR development [5,6]. These can be classified into two categories. In the first group are risk factors which increase RPE cell dispersion into the subretinal and preretinal space, and cellular proliferation. These include large retinal tears and detachments, cryotherapy and sclera 2 Stem Cells International indentation, and retinal detachments of long duration. In the second group are characteristics which increase the presence of growth factors and inflammatory cytokines in the environment, with or without breakdown of the blood-ocular barrier [7]. This includes vitreous hemorrhage, choroidal hemorrhage, and cryotherapy. These risk factors for PVR are additive-the more characteristics, the higher the risk of PVR development. Patients with traits in both categories are faced with a perfect storm, in which cell migration and proliferation occur in an environment primed for cellular misbehavior.
There has been controversy in the literature regarding the extent of involvement of cells other than RPE, such as Müller glia, in the pathogenesis of PVR. It is a fact that Müller cells are active participants. Recent work demonstrating the reactivity of Müller glia during retinal detachment and other forms of retinal injury suggests that these cells play a significant role in diseases involving retinal injury and degeneration, such as PVR. Although RPE cells have long been considered the principal mediators of this disease, Müller cell activation, migration, proliferation and transformation in retinal detachment, and retinal injury have all been documented [8][9][10]. Increased expression of GFAP and vimentin, indicative of increased reactivity, has been demonstrated in Müller glia in detached human retinas and experimental models of retinal detachment [11,12]. Experimental detachment models have also shown Müller cell proliferation which peaks at 3-4 days retinal detachment and continues at a slower rate for weeks to months [13], as well as migration of Müller cell processes and nuclei throughout the retinal layers and into the subretinal space [8]. Certainly, the data supports the need to explore more closely how these cells participate in PVR pathogenesis, and what drives them to do so.
The question then arises of whether the behavior of Müller cells, already primed and activated in the context of retinal detachment, can be altered by the presence of RPE cells and growth factors in the vitreous environment. Our experiments were designed to help us understand how RPE and Müller cells might affect each other when forced to interact in the context of retinal detachment, and how Müller glia altered by this environment participate in PVR. Because our goal is to better understand this process in human disease, we have chosen to work with the ARPE19 and MIO-M1 Müller progenitor cell lines.  patent application PCT/GB2004/005101). Primary rabbit conjunctival fibroblasts (RCFs) were obtained as previously described [14].

Cell Cultures.
Transmembrane cell cultures were set up using MIO-M1 and ARPE19 cells in DMEM/F12 media with 10%FBS (Gibco). MIO-M1 cells were plated in sixwell plates and allowed to reach confluency. ARPE19 cells were plated on transwell inserts with 0.4 μm pores and allowed to grow to confluency. Upon reaching confluency, the inserts containing ARPE19 cells were placed in MIIO-M1 containing six-well plates and cultured in a total volume of 3 mL of media. Control groups of each single-cell type were grown concurrently. Cells were fed with 1.5 mL of media on day 2 and 6, with complete media changes at day 4.

Western
Blotting. MIO-M1 and ARPE19 cells were harvested at days 1, 3, 5, and 7. Lysates were made for Western blot analysis. Media was removed and cells were collected from experimental and control plates and inserts using sterile cell scrapers in PBS. Lysates of these cells were created by incubation for 30 min at 4 • C in Ripa buffer 50 mM Tris HCl [pH 8.0], 0.1% SDS, 0.5% sodium deoxycholate, 1% NP-40, 150 mM NaCl, 2% protease inhibitor cocktail, 2% phosphatase inhibitor cocktail 1, and 2% phosphatase inhibitor cocktail 2 (Sigma), followed by 15 seconds of sonication and removal of cellular debris by centrifugation at 12,000 rpm for 12 min at 4 • C. The protein content of the lysates was determined with the BCA protein assay (Thermo Scientific). Expression of PDGFRα and PDGFRβ in the control and experimental samples was compared by Western blotting. 50 μg of protein was resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The protein bands were transferred onto nitrocellulose membranes (Bio-Rad Laboratories), and the membrane was subjected to Western blot analysis using anti-PDGFRα or anti-PDGFRβ primary antibodies. Band density was quantified using Image J and normalized to β-actin expression.

Immunocytochemistry.
MIO-M1 and ARPE19 cells were cocultured as described above on 16-well glass slides with ARPE19 cells in 8-well strips of 0.2 μm membrane inserts. Inserts were removed after 5 days and the plated cells were fixed for 20 minutes with 4% paraformaldehyde and permeabilized for one hour at room temperature with 10% goat serum and 0.1% Triton X-100 in PBS. MIO-M1 cells were stained with primary antibodies directed against PDGFRα, PDGFRβ, and incubated at 4 • C overnight. Primary and secondary antibodies were prepared in 10% GS in PBS solutions at a concentration of 1 : 100 for PDGFRα and β. The slides were then washed with PBS. Secondary antibodies were added to the slides for 1 hour. Slides were then washed with PBS, coated with mounting media containing DAPI, and covered. The cells were examined by fluorescent microscopy with Cy2 filter.  PVR was induced in the right eyes of pigmented rabbits purchased from Covance (Denver, PA, USA). Briefly, a gas vitrectomy was performed by injecting 0.1 mL of perfluoropropane (C 3 F 8 ) (Alcon, Fort Worth, TX, USA) into the vitreous cavity 4 mm posterior to the corneal limbus. One week later, all rabbits received two injections: (1) 0.1 mL of PRP (platelet-rich plasma) and (2) 0.1 mL DMEM containing 2 × 10 5 of rabbit conjunctival fibroblasts (RCFs), MIO-M1, and MIO-M1α cells. The extent of retinal detachment was evaluated by indirect ophthalmoscopy with a handheld +30 D fundus lens at days 2, 4, 7, and weekly thereafter for a total of 4 weeks. Extent of PVR was graded according to the Fastenberg classification from grade 0 through 5 [15]. On day 28 the animals were sacrificed and the eyes were enucleated. All surgeries were performed under aseptic conditions and pursuant to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The protocol for the use of animals was approved by the Schepens Animal Care and Use Committee. Mann-Whitney test for nonparametric data (P < 0.05) was used for statistical analysis.

As previously discussed, the goal of these experiments is to determine if RPE cells induce a change in Muller cells that would result in a PVR-inducing cell.
Immunocytochemical staining of MIO-M1 MPCs after 5 days of coculture with ARPE19 cells confirms MIO-M1 MPCs upregulate their expression of PDGFRα (Figures 1(a)-1(b)), with a moderate increase in expression of PDGFRβ (Figures 1(c)-1(d)). Consistent with these results, Western blot analysis showed upregulation of PDGFRα expression in experimental MIO-M1 MPCs by day 5 after-plating ( Figure 2). In contrast, expression of PDGFRβ remained low in MIO-M1 and ARPE 19 cells across the same time, with a measurable increase in MIO-M1 cells by day 7 only (Figure 2). A small but detectable decrease in GFAP

Discussion
PDGFRα has long been implicated in the pathogenesis of PVR. It is found extensively in preretinal membranes from PVR patients [16,17]. Experimental models using mouse embryonic fibroblasts as well as rabbit conjunctival fibroblasts have demonstrated the intrinsic role that PDGFRα, and not PDGFRβ, plays in the pathogenesis of the disease [14,18]. In fact, inhibition of the PDGFRα, either through inhibition of its tyrosine kinase or the ROS pathway, has been shown to be sufficient in these models to attenuate and/or inhibit the development of PVR [19,20].
Studies show that Müller cells are present in PVR membranes [21]. However, RPE rather than Müller cells, have dominated the literature and have been the focus of most studies and theories of PVR [22,23]. This is partly due to the fact that RPE cell markers are more abundant in PVR membranes. The assumption has been made that Müller cells play a less important role, yet this may not be the case.
Our in-vitro study observations and correlating rabbit model results suggest that one of the mechanisms by which Müller cells may play a role in PVR is by upregulating the expression of PDGFRα. When comparing with previous studies in the literature, this change has a bigger impact on Müller cell behavior than RPE cell behavior [24]. These experiments also suggest that Müller cell up-regulation of PDGFR might be the result of the changes in the RPE cell Stem Cells International rich environment which exists in retinal detachments with high-risk characteristics. High-risk retinal detachments for the development of PVR are more commonly characterized by increased RPE cell migration and presence in the vitreous cavity. In the presence of RPE cells in our study, Müller progenitor cells do two very important things-they change their expression of traditional Müller cell markers such as GFAP, and they increase their expression of PDGFR. Though these changes do not appear to be significant at first glance, studies in the rabbit model suggest that only a small increase in PDGFRα expression is necessary to dramatically alter the behavior of these cells to resemble that of fibroblasts. This is in stark contrast to studies performed using RPE cells, in which a more than 80-fold increase in receptor expression was necessary to significantly alter their behavior, with invivo results which were still inferior when compared to those using fibroblasts [24]. Müller cell plasticity and their capacity to transform their phenotype have already been demonstrated [25,26]. The ability of this cell type to alter its expression of GFAP in different environments helps to support our observations [9]. Dedifferentiation in particular is thought to play an important role in how Müller cells participate in PVR. Müller cells in the peripheral retina, where PVR most often occurs, have been shown to express stem cell markers, indicative of active proliferation, and dedifferentiation [27]. In-vitro, human-derived Müller cells, including the MIO-M1 Müller progenitor cell line, have been shown to exhibit neural stem cell traits [28,29].
We therefore propose the following theory to PVR development invivo. During retinal detachment, depending on the size and longevity of the detachment as well as the size and location of the retinal tear, there is an opportunity for RPE cells to abandon their natural monolayer and migrate onto the subretinal and preretinal space. The vitreous often acts as a scaffold onto which these cells can attach. Once allowed to migrate, RPE cells begin to produce cytokines and cofactors (yet to be fully identified) which can alter Müller cell phenotype and growth factor surface protein expression, leading to an increase in fibroblastic behavior and pathogenicity. Whether there is complete transformation of Müller cells invivo remains a question. It is possible that the abundance of RPE cells in PVR membranes is a "redherring", and that RPE cells play more of an effector role, with Müller cells doing most of the membrane formation and contraction.
The drawback of our study is that it is based on observations in an in-vitro setup and a limited rabbit animal model. These observations would be more difficult to make invivo, given the progressive and dynamic nature of the disease and the plasticity of Müller cells. Despite its limitations, our invitro model allows us to capture specific changes, and our rabbit model allows us to confirm their importance.
Certainly more work needs to be performed invivo. The next step would be a series of experiments in retinal detachment models looking at intraretinal GFAP and PDGFRα expression. Another set of experiments would be aimed at identifying those molecules which trigger Müller cell transformation and growth factor receptor expression. Once identified, therapeutic interventions can be designed to interfere with and redirect this process.