Abstract
Purpose.:
Previous studies indicate that the expression of platelet-derived growth factor (PDGF) receptor α (PDGFRα) dramatically increases the ability of fibroblasts to induce experimental proliferative vitreoretinopathy (PVR). The purpose of this study was to determine whether PDGFRα contributed to the PVR potential of retinal pigment epithelial (RPE) cells, one of the most abundant cell types in PVR membranes.
Methods.:
PDGFRα expression in human ARPE19 cells was increased or decreased by stably expressing the PDGFRα cDNA or short hairpin (sh) RNA directed at PDGFRα, respectively. The level of PDGFRα expression in the resulting panel of cell lines was either barely detectable (KD), standard (similar to the level of primary RPE cells), or overexpressed approximately 80-fold. Western blot analysis was used to assess the level of p53 and the activation state of PDGFRα and Akt. The following cellular responses were monitored: proliferation, apoptosis, and contraction. The PVR potential of cells was tested in a rabbit model of PVR in which cells were coinjected with platelet-rich plasma into the vitreous.
Results.:
Comparison of KD and overexpressing cells indicated that high-level expression of PDGFRα dramatically augmented signaling events, cellular responses, and the PVR potential of ARPE19 cells. However, all these outcomes were also significantly increased, albeit not as robustly, by PDGFRα expression to the level typically present in RPE cells.
Conclusions.:
Even though RPE cells express substantially less PDGFRα than fibroblasts, it significantly boosts PVR-related signaling events, cellular responses, and the PVR potential of ARPE19 cells. These studies suggest that inhibiting activation, signaling, or both by PDGFRα has the potential to prevent the development of PVR.
Proliferative vitreoretinopathy (PVR) is a disorder characterized by the formation of membranes on both surfaces of the retina that contract and thereby induce retinal detachment. PVR results from various causes, but it is most often encountered after retinal tears and retinal detachment and after therapeutic interventions for these conditions. It occurs in approximately 8% to 10% of patients undergoing primary retinal detachment surgery and is the principal cause of failure of this procedure.
1 –5
The PVR membrane consists of extracellular matrix proteins (collagen I and II and fibronectin), and cells (retinal pigment epithelial cells [RPE], retinal glial cells, fibroblasts, and macrophages).
6 –9 RPE cells are among the most abundant cell type in PVR membranes, and this probably relates to the retinal break and dispersion of viable RPE cells into the vitreous during cryopexy treatment of retinal tears.
10 –12 Although there has been increased success in reattaching the retina, understanding the molecular mechanism of PVR is likely to enable the development of effective pharmacologic approaches to protect patients undergoing surgery to correct a retinal detachment from succumbing to PVR.
Results of recent studies support the growth factor hypothesis regarding the pathogenesis of PVR.
13 For instance, growth factors are present in the vitreous and promote many of the cellular events that are intrinsic to PVR. Furthermore, the expression of platelet-derived growth factor (PDGF) receptor α (PDGFRα) increases the ability of fibroblasts to induce experimental PVR.
14 Moreover, blocking growth factor-dependent activation of receptors or the downstream signaling events protects animals from developing experimental PVR.
15 –18 These studies begin to elucidate key events in the development of PVR and thereby identify potential therapeutic targets.
A potential shortcoming of the information obtained using a fibroblast-driven model of PVR is that fibroblasts are not the major cell type within clinical PVR membranes. Consequently, the goal of this study was to identify an Achilles heel in RPE cells, one of the most abundant cell types within clinical PVR membranes. In light of the fact that cultured RPE cell lines express PDGFRα
19 and that this receptor is expressed and activated in PVR membranes isolated from patients,
20,21 we focused on evaluating the importance of PDGFRα for RPE cells to induce experimental PVR.
An oligo (CCTGGAGAAGTGAGAAACAAA) corresponding to NM_011058.1–937 in a hairpin-pLKO.1 retroviral vector, the packaging plasmid (pCMV-dR8.91), the envelope plasmid (VSV-G/pMD2.G), and 293T packaging cells were from Dana Farber Cancer Institute/Harvard Medical School (Boston, MA).
To prepare PDGFRα shRNA lentivirus, a mixture of packaging plasmid (0.9 μg), envelope plasmid (0.1 μg), hairpin-pLKO.1 vector (1 μg) (or a hairpin-pLKO.1 containing the PDGFRα shRNA oligo), and TransIT-LT1 were mixed and incubated at room temperature for 30 minutes. The transfection mix was transferred to 293T cells that were approximately 70% confluent. After 18 hours, the medium was replaced with growth medium modified to contain 30% serum, and virus was harvested 40 hours after transfection. The viral harvest was repeated three times at 24-hour intervals. Virus-containing media were pooled and centrifuged at 800g for 5 minutes, and the supernatant was used to infect ARPE19 cells. Successfully infected cells were selected on the basis of their ability to proliferate in media containing puromycin (1 μg/mL). Resultant cells were characterized by Western blot analysis using an antibody against PDGFRα.
Expression of PDGFRα in ARPE19 Cells Enhanced Vitreous-Dependent Signaling Events Intrinsic to PVR
Expression of PDGFRα to the Level Typical of RPE Cells Significantly Enhanced RV-Induced Cellular Responses Intrinsic to PVR
Exposure of ARPE19KD cells to vitreous promoted a small increase in proliferation (
Fig. 2). The proliferative response to vitreous was significantly greater in ARPE19 cells, indicating that expressing PDGFRα to a physiologically relevant level increased the responsiveness of RPE cells to vitreous. Overexpressing PDGFRα did not further enhance the ability of cells to proliferate.
A similar phenomenon was observed when we monitored the ability of vitreous to protect ARPE19 cells from apoptosis. Survival was significantly improved by expression of PDGFRα to the level typically observed in RPE cells, and overexpression was not further advantageous (
Fig. 3 and
Supplementary Fig. S2).
Given that PDGFRα expression resulted in both a decline in apoptosis and an increase in proliferation, it is possible that the increased percentage of surviving cells led to an overestimation of the cell proliferation response. In the worst case scenario, the error would be minor because the magnitude of the change in percentage of surviving cells was much smaller than the percentage increase in proliferation. Vitreous induced 1.9%, 2.4%, and 1.5% increases in surviving cells, whereas the increases in cell proliferation for ARPE19α, ARPE19, and ARPE19KD cells were 169%, 157%, and 133%, respectively.
As with proliferation and survival, vitreous-induced contraction of ARPE19 cells was enhanced by the expression of PDGFRα to the standard level (
Fig. 4 18,26 ). However, in contrast to the proliferative and survival responses, overexpressing PDGFRα improved the ability of cells to contract (
Fig. 4). These studies show that the expression of PDGFRα in RPE cells to the standard level enhanced three PVR-related cellular responses when exposed to vitreous. Furthermore, contraction was the only vitreous-dependent response that was further enhanced by the overexpression of PDGFRα.
By characterizing a panel of ARPE19 cells that harbor various levels of PDGFRα, we learned that the level of PDGFRα that RPE cells normally express was sufficient to boost their responsiveness to vitreous and enhance their ability to promote PVR. Because RPE cells are among the most abundant cell types present in an epiretinal membrane,
7 –9 our studies suggest that approaches to block vitreous-driven activation of PDGFRα have the potential to provide protection from PVR.
Although the PVR potential of RPEs and fibroblasts is dependent on the expression of PDGFRα, the level of expression in fibroblasts is typically much higher than in ARPE19 cells (
Fig. 1). This raises the possibility that fibroblasts need more PDGFRα to achieve their full PVR potential. Although we have not evaluated the impact of PDGFRα expression level on PVR potential with a panel of cell lines expressing different levels of PDGFRα, our work with a dominant negative PDGFRα suggests that modest expression of PDGFRα is sufficient to boost the PVR potential of fibroblasts, just as in RPEs. The dominant negative PDGFRα mutant is a truncation that ends at the beginning of the kinase domain, and it effectively prevents PVR, provided that it is expressed roughly 50-fold above the level of the full-length receptor.
16 At lower levels of expression, the dominant negative is no longer capable of blocking PVR (H.L. and A.K., unpublished observations, 2009). These studies suggest that not all PDGFRα expressed in fibroblasts is necessary to enhance the PVR potential of this cell type.
Overexpressing PDGFRα did not increase all vitreous-driven cellular responses above the level observed in cells expressing the standard level of PDGFRα. This observation suggests that different thresholds of signaling are required for the three cellular responses we monitored. Since the greater decline in p53 (achieved in overexpressing cells) was associated with enhanced contraction, we speculate that p53 controls the expression of genes that prevent contraction. Identification of such genes may lead to approaches to prevent this most sight-threatening component of PVR.
It is important to point out that this model of PVR may underestimate the pathogenic potential of PDGFRα overexpression. Overexpressing cells may be substantially more responsive in traumatic settings that involve the release of large amounts of PDGF from platelets. Additional studies to evaluate the level of PDGFRα expression and how it correlates with the severity of PVR will provide the information needed to address this question.
Although our studies indicate that PDGFRα enhances the PVR potential of RPE cells, it is important to note that even when PDGFRα expression is very low (or null, as can be achieved using knockout fibroblasts
14 ), the cells are still capable of inducing PVR. These observations indicate that PDGFRα is not the only driver of pathogenesis. Identification of other contributors will further increase our appreciation of how PVR develops and will enable us to prevent it.
In summary, our findings identify PDGFRα as an essential mediator of the PVR potential of RPE cells, one of the major cell types in PVR membranes. As a result, efforts directed at inhibiting vitreous-dependent activation of PDGFRα, signaling events, or cellular responses intrinsic to PVR have the potential to lead to therapies that effectively protect patients from PVR.