Platelet-derived growth factor D, tissue-specific expression in the eye, and a key role in control of lens epithelial cell proliferation.

Platelet-derived growth factor D (PDGF-D), also known as Iris-expressed growth factor, is a member of the PDGF/vascular endothelial growth factor family. The expression of PDGF-D in the eye is tissue-specific. In the anterior segment, it is localized to iris and ciliary body, whereas in the retina, PDGF-D is restricted to the outer plexiform layer. PDGF-D is present in aqueous humor but is not detectable in mature lens or in mouse lens-derived alphaTN4-1 cells. However, it is expressed in rabbit lens-derived N/N1003A cells. N/N1003A cell-conditioned medium stimulates proliferation in rat lens explants, and this is blocked by immunodepletion of PDGF-D. Immunopurified PDGF-D also stimulates cell proliferation in rat lens explants and in NIH 3T3 cells. In organ culture of rat eye anterior segments, anti-PDGF-D strongly inhibits lens epithelial cell proliferation. This finding suggests a major in vivo role for PDGF-D in the mechanisms of coordinated growth of eye tissues. Intervention in the PDGF-D pathway in the eye, perhaps by antibody or blocking peptide, could be useful in the treatment of certain cataracts, including post-operative secondary cataract.

Coordinated growth of different tissues of the eye is essential for normal eye development and for normal vision (1,2). In the anterior segment of the eye, the lens has important effects on the normal development of surrounding tissues (3), whereas the optical properties of the eye at all ages depend on maintaining the correct size and shape of the lens for the growing eye. Continuing growth of the lens past maturity has deleterious consequences for visual acuity with age (4).
Because of its highly organized structure and regulated cell growth, the lens is also one of the classic systems for studying growth and differentiation (5)(6)(7)(8)(9)(10). Lens cells in different states of proliferation and differentiation are spatially segregated in specific zones (7,11). An anterior monolayer of relatively undifferentiated epithelium contains a central region of quiescent cells surrounded by zones of cells that proliferate and migrate toward the lens equator. The proliferative zone is demarcated by the posterior chamber of the aqueous humor bounded by the lens, the lens zonules, the ciliary body, and the iris. As migrating epithelial cells reach the equator of the lens and leave the posterior chamber, they come under the influence of factors secreted into the vitreous (12). In this differentiation zone, migration and proliferation cease and the epithelial cells reorganize and elongate to form layers of fiber cells that reach half-way around the lens, express high levels of crystallins, and make up most of the refractive structure of the lens. FGF 1 family growth factors play important roles in the control of the differentiation process (12)(13)(14). The control of proliferation is not so well understood. Early work implicated insulin and epidermal growth factor (15,16). More recently, it has been shown that both PDGF-A and PDGF-B are expressed in the iris in the newborn mouse eye and that, in combination with FGF-2, PDGF-A affects proliferation and differentiation in lens epithelial explants (17,18). VEGFA has also been detected in lens epithelium, and the level of this factor, usually associated with angiogenesis, actually increases with age in the avascular lens (19).
In the course of the NEIBank project for ocular genomics (20), a new member of the PDGF/VEGF family of growth factors was discovered in the human iris and named iris-expressed growth factor (IEGF) (21). The same protein was independently identified by other groups and named spinal cordderived growth factor B or PDGF-D (22)(23)(24)(25)(26). The latter has become the accepted designation, although it may overemphasize a fairly distant superfamily relationship with the familiar PDGF-A and PDGF-B (see family tree in Ref. 21). The closest sequence relative of PDGF-D is another protein variously known as fallotein, spinal cord-derived growth factor A, or PDGF-C (27)(28)(29). Both of these newly identified proteins have an N-terminal CUB domain, an immunoglobulin-like protein interaction module (30), and a C-terminal PDGF-like growth factor domain (31). Recombinant PDGF-D has been studied in cell culture systems, and both proteolytically cleaved and fulllength proteins have been shown to have activity, including proliferative, transforming, and angiogenic activities, mediated through PDGFR␤ and PDGFR␣/␤ receptors (24 -26). Here we describe the localization of PDGF-D in the eye and its effects on lens epithelial cell proliferation.

MATERIALS AND METHODS
Antibody Production-A peptide from human PDGF-D (IEGFp1, KSKVDLDRLNDDAKRYSC) was synthesized at BIOSOURCE/QCB (Hopkinton, MA), conjugated to carrier, and used to immunize rabbits. Antiserum was tested on Western blots. The most potent bleeds were affinity-purified at BIOSOURCE.
Western Blots-Proteins were extracted from rat eye tissues in TE buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA). For cell-conditioned medium, cells were grown overnight at 37°C, 5% CO 2 , washed three * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  times with serum-free medium, and grown overnight in serum-free medium. Conditioned medium was collected at 0 and 21 h and concentrated using Amicon YM10 filters (Millipore, Billerica, MA). Bovine aqueous humor was extracted from 12-18-month-old bovine eyes using a Hamilton syringe. The eyes were obtained through J. W. Treuth & Sons (Baltimore, MD).
For Western blots, 10-g protein samples (or 30 g for aqueous humor) were run in 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes. Western blots were processed following the protocols of the Western Breeze chemiluminescent Western blot kit (Invitrogen). Blots were incubated with anti-PDGF-D primary antibody (IEGFp1) diluted 1:1000 for 1 h at room temperature. Primary antibody was also biotinylated using EZ-Link sulfo-NHS-LC-biotin (Pierce). The biotinylated antibody was used at 1:5000 dilution in Western blots and visualized with streptavidin-horseradish peroxidase conjugate (Pierce) and Supersignal West Pico substrate (Pierce) following the manufacturer's protocols.
Conditioned Media and Immunodepletion-N/N1003A cells (32) were cultured in minimum Eagle's medium with L-glutamine (Sigma) containing 8% rabbit serum (Invitrogen) and gentamycin. ␣TN4-1 cells (33) were cultured in DMEM with L-glutamine containing 10% fetal bovine serum and penicillin/streptomycin. Conditioned media were collected after 21-h serum starvation. 1 ml of conditioned media was incubated with 1 g of anti-PDGF-D antibody or normal IgG (Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 1 h. 50 l of 50% protein G-agarose (Amersham Biosciences) were incubated with the antigen-antibody complex at 4°C for 1 h. Complexes were precipitated by centrifugation. The supernatant (depleted medium) was removed.
Immunopurification of PDGF-D-Anti-PDGF-D antibody was crosslinked to protein A beads using the Seize-X immunoprecipitation kit (Pierce). 15 ml of N/N1003A cell-conditioned medium was concentrated to 1 ml by YM10 filtration (Millipore) and immunoprecipitated using the cross-linked antibody, and bound PDGF-D was eluted.
Cell Survival and Proliferation in Lens Explants-Lens epithelial explants were dissected from newborn rat eyes (14). Five explants were placed in individual wells of a 96-well tissue culture plate and treated with 200 l of conditioned medium from N/N1003A cells, conditioned medium from ␣TN4-1 cells, N/N1003A medium immunodepleted for PDGF-D, or N/N1003A medium mock-immunodepleted by normal IgG. Each treatment was performed in triplicate. Explants were incubated at 37°C in 5% CO 2 .
Cell survival was measured using trypan blue (Invitrogen). Explants in a 96-well tissue culture plate were cultured as above. Plates were centrifuged for 5 min at 300 ϫ g at 4°C. Cells were trypsinized, centrifuged in minimum Eagle's medium, and incubated for 5 min with trypan blue at room temperature. Live and dead cells, as judged by staining, were counted using a hemocytometer.
For the proliferation assay, explants were treated in the same way. Proliferation assays were performed using the BrdUrd labeling kit III (Roche Applied Science). 10 M BrdUrd was added at 21 h and incu-bated at 37°C for 2 h. The cells were fixed with 0.5 M ethanol/HCl and incubated with nucleases. Incorporated BrdUrd was detected with monoclonal anti-BrdUrd-POD Fab fragments using the soluble chromogenic substrate (2,2Ј-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid)) measuring absorbance 405 nm.
For testing immunoprecipitation (IP)-purified PDGF-D, explants were taken as before, incubated with 200 l of medium (minimum Eagle's medium ϩ penicillin/streptomycin), and treated with two different concentrations of IP-purified PDGF-D. Cell survival and proliferation were assayed in triplicate as before.
Proliferation Assay on NIH 3T3 Cells-NIH 3T3 cells were grown in 96-well plates in DMEM, 10% fetal bovine serum/penicillin/streptomycin at 37°C, 5% CO 2 overnight. They were changed to medium with 1% fetal bovine serum and grown overnight again. Cells were treated for 21 h with 100 ng/ml PDGF-A, PDGF-B (Roche Applied Science), and IP-purified rabbit PDGF-D with or without the addition of the appropriate antibody (PDGF-A and PDGF-B antibodies from R&D, Benicia, CA) at dilutions of 1:200, 1:100, and 1:50. Proliferation was assayed using the BrdUrd labeling kit III.
Anterior Segment Organ Culture-Whole anterior segments were dissected from newborn rat eyes. Five anterior segments were placed in each well of a 96-well tissue culture plate and incubated with DMEM/ F-12 medium containing 2 mM L-glutamine/1 mM ascorbic acid/penicillin/streptomycin in the presence of antibodies to PDGF-A, PDGF-B (R&D), FGF-2 (Santa Cruz Biotechnology), and PDGF-D at a 1:100 dilution for 21 h at 37°C, 5% CO 2 . Lens epithelia were dissected, and proliferation was assayed by BrdUrd incorporation as before.

Expression of PDGF-D in Eye Tissues and Cultured Cells-A
peptide antibody (designated IEGFp1) was designed to target a region of the growth factor-like domain of the human PDGF-D sequence that was predicted to be both antigenic and specific and to represent a core functional region in case of a loss of other regions by post-translational processing (Fig. 1). In Western blots of adult rat eye tissues, the anti-PDGF-D antibody detected protein in the iris, cornea, and retina ( Fig. 2A). In contrast, the expression in lens was very low or absent. Although there was some minor size heterogeneity among different eye tissues, the immunoreactive protein was consistent in size with full-sized glycosylated PDGF-D (ϳ50 kDa), rather than with proteolytically cleaved fragments. To test for the presence of PDGF-D in the aqueous humor, bovine eyes were used to overcome the difficulty in obtaining this fluid cleanly from rodent eyes. The antibody retained its activity across species, and apparently full-sized PDGF-D was also detected in bovine aqueous humor (Fig. 2B). As a control, the antibody was also used to test the aqueous sample for a known aqueous marker, macrophage migration inhibitory factor (34).
Conditioned media from two lens-derived cell lines were also tested for PDGF-D (Fig. 2C). Mouse lens-derived ␣TN4-1 cells, which are transformed with SV40 T-antigen (33), gave no reaction, but non-transformed rabbit lens-derived N/N1003A cells (32) were positive. Because the rabbit cells are grown in the presence of rabbit serum, the cells were washed three times, transferred to serum-free medium, and tested by Western blot at 0 and 24 h after transfer to serum-free conditions using biotin-labeled anti-PDGF-D to avoid secondary antibody detection of any contaminating rabbit IgG heavy chains. No PDGF-D was detected at 0 h, but full-sized PDGF-D was detected in the medium after 24 h. To confirm the expression of rabbit PDGF-D in N/N1003A cells, intron-crossing PCR primers were designed from the conserved regions of human and mouse sequences and used to amplify PDGF-D transcripts by RT-PCR from mouse lens and from the lens-derived cells. Consistent with previous results, no product was obtained from mouse lens or ␣TN4-1 cells but a 900-bp cDNA fragment for rabbit PDGF-D was obtained from N/N1003A cells (data not shown). This was cloned and sequenced to confirm the identity and to provide partial sequence for the rabbit protein (Gen-Bank TM accession AY347260).
The anti-PDGF-D antibody was used in immunofluorescence analysis of sections of rodent and monkey eye tissues (Figs. [3][4][5]. In newborn rat anterior segment (Fig. 3), specific staining was seen in the iris and ciliary body but not in the lens, whereas limbal cells of the corneal epithelium also gave a positive reaction. Staining was intense in the stroma of both ciliary body and iris but was also present in the epithelial cell layers. To see whether similar expression is present in primate eyes, adult monkey anterior segment was also examined by immunofluorescence. In ciliary body (Fig. 4A), there was again intense stain in the muscle layer and significant staining throughout the epithelium. Intense staining was also seen in the sphincter muscle of the iris and in an apparently connected layer basal to the pigmented epithelium (Fig. 4B). There was no obvious stain in the intensely pigmented cells themselves. This could either reflect the absence from these cells in the adult monkey iris or possibly the quenching of a weaker signal by melanin. Overall, these results show that PDGF-D is differentially expressed in the anterior segment in both rodent and primate eye in a pattern that is consistent with a role in lens growth through secretion into the posterior chamber.
Localized staining was also seen in the retina. In the mature adult rat retina (Fig. 5), this expression localizes intensely to the outer plexiform layer, which contains photoreceptor axons and the synaptic layer between photoreceptors and second order neurons.
Expression of Receptors for PDGF/VEGF Family Members in Lens-Previous results have shown that PDGF-D, in contrast to PDGF-A, acts mainly through the PDGFR␤ receptor (24 -26). Rat eye tissues were examined for the expression of PDGFR␤ and for several other receptors for members of the VEGF/PDGF family using RT-PCR (Fig. 6). Primers were designed for rat sequences taken from GenBank TM and dbEST for PDGFR␣, PDGFR␤, FLT1 (VEGFR1), FLT3, and FLK1 (VEGFR2) receptors. The products of the expected size for all five receptors were detected in total RNA extracted from newborn rat lens epithelium, lens fibers, and iris, whereas no RT controls were blank. Bands were sequenced to confirm their identity. The same primers were also used on RNA from the rabbit lens-derived cell line. Positive results were obtained for PDGFR␤ and for FLT3 (Fig. 6). The failure of the other primers to work may reflect differences in the (unknown) rabbit sequences for these receptors. the presence of appropriate receptor expression in lens epithelium, the effects of PDGF-D on lens epithelial cell proliferation were examined. N/N1003A cell-conditioned medium was used as a source of native PDGF-D. Newborn rat lens epithelial explants were dissected and cultured. Cell survival was examined by trypan blue staining in explants at 0 and 21 h after incubation in N/N1003A-conditioned medium, conditioned medium immunodepleted using the anti-PDGF-D antibody, mockdepleted medium, and conditioned medium from ␣TN4-1 cells (Fig. 7A). Lens epithelial cells in culture have previously been shown to produce an unidentified cell survival factor and to be able to survive in minimal medium (35). After 21 h, both conditioned and depleted media retained a 75-80% number of live cells seen at 0 h with an increased number of dead cells. However, there is a distinct enhancement of cell survival in N/N1003A-conditioned medium compared with depleted medium (essentially serum-free medium) or ␣TN4-1 medium, suggesting that PDGF-D has survival activity in this system.
Proliferative activity in the same system was examined by BrdUrd incorporation, normalized by the number of live cells (Fig. 7B). Significant cell proliferation was detected for explants in N/N1003A-conditioned medium. This was substantially reduced by immunodepletion of PDGF-D, whereas mock depletion with total rabbit IgG had no effect. In addition, ␣TN4-1-conditioned medium, which lacks significant amounts of PDGF-D, was much less effective in promoting cell proliferation. These results show that PDGF-D is strongly mitogenic in the lens explant system. The proliferative effect of PDGF-D seems to be greater than its survival effects. Although immu-nodepletion of PDGF-D reduced cell survival by ϳ20% in this system, proliferation was reduced Ͼ5-fold.
In addition to using N/N1003A-conditioned medium, PDGF-D protein was purified by immunoprecipitation from the medium. The IP PDGF-D was tested on isolated rat lens explants for its effect on cell viability (Fig. 8A) and BrdUrd incorporation (Fig. 8B). The IP material promoted cell proliferation, confirming the activity of the protein recognized by the antibody. Quantification of PDGF-D in this preparation is uncertain because of contaminating antibody heavy chain that is present despite cross-linking the antibody to beads. However, two concentrations of the IP material (simply 2-and 4-l samples of the preparation) were tested on explants and both induced a 2-fold increase in cell proliferation.
The active IP material was then used to test the ability of the anti-PDGF-D antibody to act as a blocking activity. The IPpurified PDGF-D was tested on NIH 3T3 cells alongside recombinant PDGF-A and PDGF-B. The anti-PDGF-D antibody and commercial blocking antibodies for the other two growth factors were tested in blocking cell proliferation. In this qualitative system, all three antibodies exhibited blocking activity and had their maximum effect at 1:100 dilution (Fig. 8C). Higher BrdUrd incorporation in explants cultured as in part a. BrdUrd incorporation is measured by absorbance at 405 nm after subtraction of blank control value and normalized by live cell content from A. C, effect of PDGF family growth factors and their blocking antibodies on NIH 3T3 cells. BrdUrd incorporation was measured as above.
itself, complete anterior segments with cornea, lens, ciliary body, and iris intact were dissected from newborn rat eyes, maintaining the native environment for lens cell proliferation (Fig. 9A). The anterior segments were maintained in culture for 21 h in DMEM with no added growth factors, relying on endogenous factors from the anterior segment tissues. To test the involvement of specific growth factors, the anterior segments were also incubated in DMEM supplemented with anti-PDGF-D antibody or blocking antibodies for PDGF-A, PDGF-B, or FGF-2. After 21 h, lens epithelia were dissected and assayed for cell survival (Fig. 9B). Under all of the conditions tested, there was significant cell survival at a higher level (2-3-fold) than was seen for isolated explants cultured for 21 h. This was not surprising, reflecting the more native environment of the epithelia during culture. Blocking antibody for PDGF-D had the largest effect on cell survival, with a 15% reduction in the number of surviving cells compared with incubation in DMEM alone.
BrdUrd incorporation by lens epithelia was measured in the same system (Fig. 9C). As measured in this assay, epithelial cell proliferation, normalized for the number of live cells, was reduced by 75% by treatment with the antibody to PDGF-D. In contrast, blocking antibodies to PDGF-A and PDGF-B reduced lens epithelia cell proliferation by 15% or less. Anti-FGF-2 (FGF-2 is associated with lens cell differentiation) had no significant effect on proliferation. These results suggest that PDGF-D is not only a factor capable of inducing or maintaining lens cell proliferation but is a major component of this process. DISCUSSION The complex architecture of the vertebrate eye is often cited as one of the most remarkable products of evolution (36 -38). Normal vision depends on the tightly coordinated growth of several highly differentiated tissues of different embryological origins (1). A key mechanism for control of growth among the different tissues of the eye is the exchange of growth factors as each tissue influences the growth and development of its neighbors. This is illustrated particularly clearly in experiments that show that the lens plays a central role in the development of the rest of the anterior segment, whereas growth and differentiation of the lens itself depend on gradients of growth factors from the aqueous to the vitreous compartments (2,3,5,6,12,14). A model has emerged with FGF family members secreted into the vitreous having a major role in the control of lens cell differentiation, whereas growth factors in the posterior chamber of the aqueous humor control lens cell proliferation. Recently, it has been shown that PDGF-A can stimulate epithelial cell proliferation (17,18). Here we show that a new addition to the repertoire of growth factors, PDGF-D, has a major role in this part of the control mechanism for lens growth.
PDGF-D, which is related to both PDGF and VEGF growth factors, has mitogenic, angiogenic, and transforming activities in various cell types in culture and has been shown to act as a homodimer (22)(23)(24)(25)(26). Originally named "iris-expressed growth factor," ocular PDGF-D was discovered by expressed sequence tag analysis of the human iris (21). As shown here, PDGF-D is expressed with striking specificity in the eye. In both rodents and primates, the protein is found principally in the tissues bounding the anterior surface of the posterior chamber of the aqueous humor and full-sized secreted PDGF-D is present in bovine aqueous humor. In rodents and primates, there is in- tense immunoreactivity in the stroma and muscles of the iris and ciliary body and significant but lower levels in the ciliary epithelium. In the newborn rat, there is clear immunostaining throughout the epithelial layers of both ciliary body and iris. In the adult monkey, the pattern is very similar in ciliary body but it is not so clear whether there is any PDGF-D in the pigmented epithelium of the iris, although there is a layer of strong staining basal to these cells. The factors secreted from ciliary body/ iris into the aqueous humor can bathe the epithelial cells of the proliferative zone of the lens. The strong staining for PDGF-D in the muscles of the ciliary body and iris is intriguing. It remains to be seen whether this is a principal site of synthesis with possible transport to the epithelia through the blood/eye barrier or whether there is a specific role for this protein in these ocular muscles. However, PDGF-D is present in aqueous humor from bovine eye and, as such, can have effects on lens growth.
Mature lens does not express PDGF-D, although preliminary results (data not shown) suggest that the protein may be expressed in the lens early in development at stages when the growth rate is at its highest. This is under further investigation. Similar to mature lens, the ␣TN4-1 cell line derived from transgenic mouse lens transformed by SV40 T-antigen (33) does not express detectable PDGF-D. However, the untransformed rabbit lens N/N1003A line (32) does express this growth factor. The possibility that the expression of PDGF-D may play a role in the immortalization of these cells is also under examination. Interestingly, in these cells as well as in native eye tissues, PDGF-D seems to be full-sized with no evidence for the proteolytic cleavage that has been observed in some cultured cells and which seems to be required for activation in some experimental systems (26).
The principal receptor for PDGF-D is the PDGFR␤ receptor (26). Rat lens epithelium was examined for expression of PDGFR␣ and PDGFR␤ and also for the VEGF receptors FLT1 (VEGFR1) and FLK1 (KDR or VEGFR2) and for FLT3, another tyrosine kinase receptor associated with angiogenesis. All five were detected in the epithelial cells by RT-PCR. In addition, the PDGFR␤ and FLT3 receptors were identified in the rabbit lens N/N1003A cells. The expression of receptors usually associated with angiogenesis in the avascular lens and in lensderived cells suggests that these factors may have additional roles in this tissue. PDGFR␣ receptor has been identified in lens by other groups (17,39), and VEGFR2 (FLK1) receptor been shown to be expressed in mouse lens epithelium (19). The wide variety of receptors expressed in the lens suggests that many pathways are available for growth factor-mediated communication among ocular tissues. The presence of PDGFR␤ in the iris raises the likelihood that PDGF-D and related proteins expressed in this tissue may exert autocrine or paracrine effects that remain to be examined. N/N1003A conditioned medium is a convenient source for native PDGF-D, avoiding possible processing issues that may arise for recombinant protein in heterologous systems. The conditioned medium stimulates epithelial cell proliferation in rat lens explants, and this is efficiently abolished by immunodepletion of the growth factor. PDGF-D protein purified by immunoprecipitation from the conditioned medium is also able to stimulate proliferation in lens explants and in NIH 3T3 cells. Although there are also some effects on cell survival, the proliferative effects of the conditioned medium in the explant system are at least an order of magnitude greater.
Antibody to PDGF-D is able to block lens epithelial cell proliferation efficiently in intact rat eye anterior segments in organ culture. Complete anterior segments (cornea, iris, ciliary body, lens, and attached tissues) can be maintained in culture, and in this system, proliferation of lens epithelial cells continues for at least 21 h. Even in the presence of the surrounding tissues that provide the native sources of growth factors, incubation with anti-PDGF-D significantly reduces proliferation of lens epithelial cells. Blocking antibodies to PDGF-A, PDGF-B, and FGF-2 have a much less effect. The potent effect of the anti-PDGF-D antibody suggests that PDGF-D has a major role in the anterior segment of the eye and may indeed be the main factor controlling growth of the lens in vivo.
This observation could have direct clinical value. Post-operative secondary cataract is a major complication of standard cataract surgery (40). Residual epithelial cells attached to lens capsule can proliferate and cause a new opacity that must be addressed by laser treatment or other measures. Conceivably, intervention in the PDGF-D signaling pathway, perhaps by antibody fragments or blocking peptides, could be used at the time of intraocular lens implantation to reduce the occurrence of post-operative secondary cataract. Intervention in the PDGF-D-related proliferative pathways might also be therapeutically useful in addressing some forms of posterior subcapsular cataract where cell proliferation may be involved (41). Continued lens growth with age is also a general problem for the aging eye and contributes to age-related problems including presbyopia (4,8). Slowing the growth of the mature lens could help delay some of these problems.
In addition to this role in the anterior segment of the eye, PDGF-D also shows striking localization to the outer plexiform layer of the retina in the adult rat. This is a region that contains photoreceptor axons and the synaptic connections among photoreceptor, bipolar, and horizontal cells. The significance of the expression of this growth factor in the retina is not yet known and is under further investigation. One interesting possibility is that PDGF-D might be one of the factors involved in survival of photoreceptor cells (42) and, as such, might be useful in therapeutic approaches to retinal degeneration.