Introduction

Pathological neovascularization in most tissues of the body, including the eye, appears to result from an imbalance between stimulators and inhibitors of angiogenesis. The majority of therapeutic efforts have focused on neutralizing a single pro-angiogenic factor (primarily vascular endothelial growth factor, VEGF-A) with notable success in the treatment of choroidal neovascularization (CNV).^1,2 However, boosting the activity of a natural anti-angiogenic factor offers the advantage of simultaneously antagonizing the activities of multiple pro-angiogenic stimuli, with the potential for a more robust therapeutic effect.

Pigment epithelium-derived factor (PEDF) is a natural anti-angiogenic protein³ that is produced by retinal pigment epithelial (RPE) cells and other cells in the eye.^4,5 PEDF induces apoptosis of dividing endothelial cells in new blood vessels, without effect on quiescent endothelial cells,⁶ and it has been postulated to be part of the endogenous system for the control of new vessel growth within the eye.^7,8,9,10 In support of this hypothesis, an increased vitreous VEGF/PEDF ratio has been reported in patients with neovascular age-related macular degeneration (AMD) or proliferative diabetic retinopathy. Injection of recombinant PEDF or increased expression of PEDF by gene transfer suppresses the development of retinal or choroidal neovascularization in animal models^{6,11,12,13,14,15,16,17,18} thereby suggesting that altering the balance between VEGF and PEDF may represent an attractive therapeutic strategy. Furthermore, any increase in PEDF levels after neovascularization has developed can cause regression of the neovascularization,⁶ thereby suggesting that it may be useful for treatment as well as suppression of disease.

Given the extensive evidence demonstrating anti-angiogenic activity of PEDF in the eye, we chose to investigate an alternative strategy for increasing its intraocular levels, namely activation of the endogenous Pedf gene through the action of an engineered zinc finger protein transcription factor (ZFP TF). ZFP TFs can be designed to specifically control the expression of virtually any gene, and their use has been shown to stimulate biological effects of target genes.^19,20 Because the ZFP activator functions in the context of the natural promoter, and only two copies of the PEDF gene are present in each cell, the level of PEDF expression driven by ZFP is predicted to be more physiological than that produced by complementary DNA (cDNA). This is potentially critical for long-term in vivo efficacy, as many natural anti-angiogenic factors (including PEDF) exhibit a biphasic dose response and are ineffective at high doses.^21,22 In this study, we investigated whether gene transfer of a PEDF-activating engineered ZFP TF would result in sufficient up-regulation of the endogenous Pedf gene and provide an anti-angiogenic environment in mouse eyes capable of inhibiting laser-induced CNV.

Results

Identification of engineered ZFP TF activators of Pedf expression in cultured cells

In order to identify ZFP TF activators of the human Pedf gene, a panel of 50 ZFP DNA binding domains targeting distinct sequences located within the -400 to +100-base pair (bp) region relative to the transcription start site of the Pedf gene were designed from an archive of zinc finger DNA binding modules. The resulting ZFP DNA binding domains, each recognizing a distinct 18-bp target sequence within the Pedf promoter region, were then linked to the nuclear factor-B p65 activation domain to generate candidate ZFP activators of hPedf. The binding affinity of each assembled ZFP to its intended target site was determined in vitro and was found to be similar to those of a small set of reference ZFPs that are robust regulators of other endogenous target genes (data not shown). This panel of ZFP TFs was then screened for the ability to drive an increase in Pedf messenger RNA (mRNA) levels by transient transfection into HEK293 cells. Two ZFPs that have overlapping binding sites were found to activate Pedf (data not shown); the more potent activator, ZFP-6961-p65, was selected for further studies. ZFP-6961-p65 (ZFP-hPEDF) recognizes the sequence 5'-GGATGGtGGTGCAGCAGTG-3' (where lower case t denotes a skipped base not directly contacted by the zinc finger modules) located -92 bp relative to the transcription initiation site of the human Pedf gene. In HEK293 cells, an approximately 14-fold activation in Pedf mRNA levels was observed (Figure 1a), which was dependent upon the presence of the nuclear factor-B p65 activation domain (data not shown). The activity of ZFP-hPEDF was further tested in ARPE-19 cells, a line derived from human RPE cells,²³ the major source of endogenous PEDF in the eye. Although the basal level of Pedf mRNA in ARPE-19 cells was approximately tenfold higher than that in HEK293 cells (data not shown), ZFP-hPEDF was able to further increase the mRNA level approximately twofold (Figure 1b). Western blot of the cultured media showed that the increase in Pedf mRNA was accompanied by a significant increase in the levels of secreted PEDF protein in both HEK293 cells and ARPE-19 cells 3 days after transfection (Figure 1c). Transfection of ZFP-hPEDF into a third human cell line, U87MG glioblastoma cells, with undetectable basal Pedf mRNA levels, also drove ZFP-dependent up-regulation of PEDF (Figure 1d). Genome-wide microarray analysis revealed that only one other gene (SERPINA1, NCBI Accession AF119873) was up-regulated by ZFP-hPEDF in U87MG cells when compared with an empty vector control (see Materials and Methods); both SERPINA1 and PEDF (SERPINF1) belong to the Serpin superfamily of protease inhibitors. Using real-time reverse transcriptase polymerase chain reaction (RT-PCR), we confirmed that SERPINA1 was indeed activated by ZFP-hPEDF, but not by over-expression of PEDF cDNA in U87MG cells (data not shown), thereby suggesting that the regulation of SERPINA1 is mostly likely an off-target effect of ZFP-hPEDF rather than a secondary response to increased Pedf expression. No genes other than PEDF and SERPINA1 were scored as up- or down-regulated by the microarray analysis. Such an exceptional degree of specificity is in full agreement with two previous genome-wide studies of engineered ZFP TF action in mammalian²⁴ and plant genomes.²⁵ Together, these results demonstrate that ZFP-hPEDF is a highly specific, potent activator of Pedf, and can function in multiple cell types having a wide range of basal levels of Pedf expression.

Figure 1.

Engineered zinc finger proteins (ZFPs) activate Pedf expression. (a) The expression plasmid for ZFP-hPEDF or empty vector control was transfected into HEK293 cells in duplicate; 72 hours after transfection, total RNA was isolated, and the level of Pedf messenger RNA (mRNA) and 18S RNA were measured by real time reverse transcriptase polymerase chain reaction (RT-PCR) (Taqman), the Pedf/18S ratio was used for normalizing Pedf expression. The Pedf/18S ratio of cells transfected with empty vector control was set as 1. The bars represent the mean (SD) of PEDF/18S ratio of duplicate transfections, each measured by duplicate RT-PCRs. (b) The same as a, except that ARPE-19 cells were transfected. (c) Seventy-two hours after transfection into HEK293 or ARPE-19 cells, culture media were collected and analyzed by Western blotting with an anti-PEDF polyclonal antibody. The cell numbers at the time of harvesting were not significantly affected by the transfection of either the ZFP-hPEDF or the empty vector, as verified by cell counting and the levels of 18S RNA (RT-PCR). Recombinant PEDF was used as a positive control. (d) The same as a, except that U87MG cells were transfected. Because the basal level of PEDF mRNA was near the lower detection limit of RT-PCR, the ratio of PEDF/18S ratio of cells transfected with ZFP-hPEDF was set as 1. (e) The expression plasmid for ZFP-mPEDF and empty vector control were transfected into Neuro2A cells in duplicate; RNA analysis was done as in a. (f) Neuro2A cells were transfected as described earlier, and after 48 hours the media were replaced. Replicates of transfected cells were maintained under normoxic (20% O₂) or hypoxic (0.5% O₂) conditions, and after 24 hours media were collected for Western blots analysis. PEDF, pigment epithelium-derived factor.

Full figure and legend (22K)

In order to determine the efficacy of the ZFP TF activator of Pedf in vivo, a mouse-specific version of the ZFP-PEDF activator was generated (the recognition sequence for ZFP-hPEDF is not conserved in mouse). We designed a panel of ZFP TFs that target the mouse Pedf promoter and tested their activity in Neuro2A cells by transient transfection. The ZFP TF that supported the greatest increase in Pedf mRNA levels, ZFP-6078-p65 (ZFP-mPEDF), was selected for further studies. ZFP-mPEDF recognizes the sequence 5'-GTGGTGgGAGAGGAGGGTG-3' (where the lowercase g denotes a skipped base not directly contacted by the zinc finger modules) located -227 bp relative to the transcription initiation site of the mouse Pedf gene. Transfection of Neuro2A cells with a plasmid expressing the ZFP-mPEDF activator resulted in an approximately tenfold increase in Pedf mRNA levels relative to an empty vector control (Figure 1e). ZFP-mPEDF also led to a significant increase in secreted PEDF protein, whereas this was below the level of detection in Neuro2A cells transfected with the empty vector control (Figure 1f, two left lanes). Since endogenous PEDF is down-regulated at the protein level by hypoxia,³ and low oxygen conditions may also be present in ocular diseases associated with neovascularization, we wished to confirm that the transcriptional activation of Pedf by ZFP-mPEDF would support increased PEDF protein levels under hypoxic conditions also. Equivalent increases in the secreted PEDF levels were obtained in cells transfected with the ZFP TF-expressing plasmid, whether they were maintained under normoxic or hypoxic conditions (Figure 1f). This indicates that ZFP-mPEDF (referred to below as ZFP-PEDF) is a potent activator of Pedf.

Subretinal or intravitreous injection of AAV.ZFP-PEDF increases Pedf mRNA and suppresses CNV

In order to determine whether ZFP-PEDF would increase Pedf mRNA levels in vivo, thereby providing an anti-angiogenic environment in the eye, we constructed adeno-associated virus type 2 (AAV-2) vectors expressing ZFP-PEDF (AAV.ZFP-PEDF), as well as a control vector expressing green fluorescent protein (AAV.GFP). Consistent with published studies,^26,27 6 weeks after subretinal injection of AAV.GFP, we observed strong fluorescence in RPE cells, with sporadic and low intensity fluorescence in other cells in the overlying retina (Figure 2a). By contrast, 6 weeks after intravitreous injection of AAV.GFP, we observed strong fluorescence in ganglion cells along the inner surface of the retina, and low intensity fluorescence in some cells in the inner nuclear layer (Figure 2b). Total RNA was isolated from posterior eyecups or from retina 6 weeks after subretinal injection or intravitreous injection, respectively. Real time RT-PCR analysis (using a primer/probe set that amplifies a sequence within the 5' untranslated region that is common to AAV.GFP and AAV.ZFP-PEDF) showed no statistical difference between the levels of mRNA transcript produced by AAV.GFP and AAV.ZFP-PEDF vectors (Supplementary Figure S2a), thereby suggesting that the transduction efficiency and the expression kinetics of AAV.ZFP-PEDF is similar to that of AAV.GFP. A significant increase in Pedf mRNA levels was observed in eyes that received subretinal injection of AAV.ZFP-PEDF compared to eyes injected with the AAV.GFP control (Figure 2c). This increase in Pedf mRNA level was independent of the route of vector delivery, as seen from the fact that the retinas of eyes that received an intravitreous injection of AAV.ZFP-PEDF also showed a significant increase in Pedf mRNA when compared with the AAV.GFP-treated controls (Figure 2d and Supplementary Figure S2b). It is therefore clear that the Pedf-activating ZFP TF identified in Figure 1 promotes Pedf transcription in vivo after either subretinal or intravitreous delivery with an AAV vector.

Figure 2.

Increased expression of pigment epithelium-derived factor (PEDF) and suppression of choroidal neovascularization (CNV) after intravitreous or subretinal injection of AAV.ZFP-PEDF. (a) Six weeks after subretinal injection of AAV.GFP, fluorescence microscopy of ocular sections showed strong green fluorescent protein (GFP) fluorescence in retinal pigment epithelial cells. (b) Six weeks after intravitreous injection of AAV.GFP, fluorescence microscopy of ocular sections showed strong GFP fluorescence in ganglion cells. (c) Six weeks after subretinal injection of AAV.GFP or AAV.ZFP-PEDF, total RNA was isolated from posterior eyecups and assayed by real time reverse transcriptase polymerase chain reaction as described in Figure 1. The bars represent the mean (SEM) of the Pedf mRNA/18S RNA ratio. The Pedf mRNA/18S ratio for eyes injected with AAV.GFP was set to 1. P = 0.0024 for difference between eyes injected with AAV.ZFP-PEDF and those injected with AAV.GFP, as arrived at by a two-tailed t-test. (d) Six weeks after intravitreous injection of AAV.GFP or AAV.ZFP-PEDF, total RNA was isolated from retina and assayed as in c. P = 0.0114 for difference between eyes injected with AAV.ZFP-PEDF and those injected with AAV.GFP, as arrived at by a two-tailed t-test. (e) Six weeks after intravitreous (top panels) or subretinal (bottom panels) injection of AAV.GFP (left panels) or AAV.ZFP-PEDF (right panels), mice had rupture of Bruch's membrane in three locations in each eye, and after 2 weeks they were perfused with fluorescein-labeled dextran and choroidal flat mounts were examined by fluorescence microscopy. The area of CNV at Bruch's membrane rupture sites appeared smaller in eyes that had been injected with AAV.ZFP-PEDF when compared with those injected with AAV.GFP. Measurements by image analysis confirmed that there were significant reductions in the area of CNV after (f) subretinal (P = 0.038 by linear mixed model) or (g) intravitreous (P = 0.0318) injection of AAV.ZFP-PEDF as compared to AAV.GFP. AAV, adeno-associated viral; mRNA, messenger RNA; ZFP, zinc finger protein.

Full figure and legend (45K)

In order to determine whether the increased expression of endogenous Pedf mediated by the ZFP TF would provide any reduction in neovascularization, we compared the AAV.ZFP-PEDF- and AAV.GFP-treated eyes in an experimental model of CNV. Eight weeks after vector injection and two weeks after rupture of Bruch's membrane with laser photocoagulation, there were large areas of CNV at rupture sites in eyes that had been given intravitreous or subretinal injections of AAV.GFP (Figure 2e, left panels). By contrast, CNV was reduced in eyes that had been given an intravitreous or subretinal injection of AAV.ZFP-PEDF (Figure 2e, right panels). Measurement of the area of CNV at rupture sites by image analysis, with the investigator masked with respect to treatment group, confirmed that the mean area of CNV was significantly less in eyes given either subretinal (Figure 2f) or intravitreous (Figure 2g) AAV.ZFP-PEDF when compared with eyes given AAV.GFP by the corresponding route of administration.

Subretinal injection of AAV.ZFP-PEDF reduces CNV area, but not hyperpermeability

In order to confirm the results obtained thus far and to assess the potential impact of increased PEDF levels on vasopermeability, fluorescein angiograms were performed 8 weeks after subretinal injection of AAV vectors and 2 weeks after rupture of Bruch's membrane by laser photocoagulation. Photographs taken 90 seconds after dye injection showed smaller areas of CNV in AAV.ZFP-PEDF-injected eyes in comparison with photographs of eyes injected with AAV.GFP as controls (Figure 3a, left panels). Photographs obtained 7 minutes after dye injection showed substantial leakage of dye into surrounding tissue in both experimental and control groups (Figure 3a, right panels). Measurement by image analysis confirmed a significant reduction in the area of CNV in eyes injected with AAV.ZFP-PEDF as compared to eyes injected with AAV.GFP (Figure 3b). However, there was no significant difference between the groups with regard to the amount of dye leakage from the CNV (Figure 3c). Ocular sections from eyes injected with AAV.ZFP-PEDF activator also showed less CNV at rupture sites (Figure 3d) compared to that seen in eyes injected with AAV.GFP (Figure 3e).

Figure 3.

Suppression of size of choroidal neovascularization (CNV), but not excessive permeability after subretinal injection of AAV.ZFP-PEDF. Adult C57BL/6 mice had subretinal injection of AAV.ZFP-PEDF or AAV.GFP, and after 6 weeks had laser-induced rupture of Bruch's membrane in each eye. Representative early- and late-phase fluorescein angiograms show a relatively small CNV lesion that leaks fluorescein in an eye injected with AAV.ZFP-PEDF (a, top row), and a relatively large CNV lesion that leaks fluorescein in an eye injected with AAV.GFP (a, bottom row). The bars show (b) the mean (SEM) area of CNV and (c) the mean (SEM) area of leakage. There was a 51.4% reduction in area of CNV in eyes injected with AAV.ZFP-PEDF as compared to eyes injected with AAV.GFP (P = 0.0025), but no significant difference in the area of leakage. Representative histology shows (d) a small CNV lesion in an eye injected with AAV.ZFP-PEDF and (e) a large CNV lesion in an eye injected with AAV.GFP. Arrowheads delineate the borders of CNV complexes. AAV, adeno-associated viral; GFP, green fluorescent protein; mRNA, messenger RNA; PEDF, pigment epithelium-derived factor; ZFP, zinc finger protein.

Full figure and legend (79K)

Intravitreous delivery of AAV.ZFP-PEDF maintains anti-angiogenic activity 3 months after injection

Intravitreous injection, an outpatient procedure, is preferable to subretinal injection, which is more invasive and is done in the operating room. We therefore focused on investigating long-term efficacy after intravitreous injection of ZFP-PEDF. Consistent with the ability of AAV vectors to provide sustained transgene expression, we observed no significant difference between the levels of mRNA transcript produced by AAV.ZFP-PEDF at 6 weeks and at 3 months after injection (Supplementary Figure S2c). At 3 months after intravitreous injection, there was a significant increase in Pedf mRNA levels in the AAV.ZFP-PEDF group relative to the AAV.GFP-injected controls (Figure 4a). Fourteen weeks (three months) after vector injection and two weeks after rupture of Bruch's membrane with laser photocoagulation, choroidal flat mounts were performed to access the area of CNV. There was a significant reduction in the area of CNV in eyes injected with AAV.ZFP-PEDF as compared to those injected with AAV.GFP (Figure 4b). This indicates that AAV delivery of a ZFP TF activator of Pedf provides a long-term anti-angiogenic environment that is sufficient to suppress CNV.

Figure 4.

Intravitreous administration of AAV.ZFP-PEDF maintains anti-angiogenic activity 3 months after injection. (a) Three months after intravitreous injection of AAV.GFP or AAV.ZFP-PEDF, total RNA was isolated from the retina and assayed as in Figure 2c (P = 0.0203). (b) Three months after intravitreous injection of AAV.ZFP-PEDF or AAV.GFP, Bruch's membrane was ruptured, and the area of choroidal neovascularization (CNV) was measured as in Figure 2. (P = 0.0327 by linear mixed model). AAV, adeno-associated viral; GFP, green fluorescent protein; PEDF, pigment epithelium-derived factor; ZFP, zinc finger protein.

Full figure and legend (5K)

Discussion

CNV is a highly prevalent cause of severe vision loss, and occurs as a complication of AMD and other diseases that compromise Bruch's membrane and the RPE. These diseases are characterized by growth of new vessels from the choroid and/or the deep capillary bed of the retina into the subretinal space, where they leak fluid causing reversible vision loss, and gradually stimulate scarring, resulting in permanent vision loss.

The pathogenesis of CNV is not completely understood, but it seems to be caused by an imbalance between stimulators and inhibitors of neovascularization. The major stimulator involved is VEGF-A, and antagonists of VEGF suppress CNV in animal models.^28,29 Ranibizumab, a humanized antibody of VEGF-A, is the first treatment that has been able to produce significant improvement in vision in a substantial percentage of patients with CNV caused by AMD.^1,2 These beneficial effects were obtained with monthly injections for 1 year, and the long-term requirements and results are unknown. There appears to be chronic deregulation of VEGF-A expression, and while some patients may not require monthly injections to maintain vision, most patients are likely to require at least six injections per year for the remainder of their lives. Identification of new treatments that can reduce the frequency of injection of VEGF antibody would provide important benefits.

In this study, we demonstrated that expression of the endogenous Pedf gene, which encodes a potent anti-angiogenic factor, can be activated using engineered ZFP-TFs that are designed to target the Pedf promoter. We provided evidence that gene transfer of a construct for a ZFP-PEDF activator successfully rebalanced the intraocular environment to one that is more anti-angiogenic and significantly inhibited CNV in a mouse model. Moreover, the anti-angiogenic effect of the ZFP activator was observed 3 months after a single intravitreous injection. Because AAV vectors have been shown to support sustained transgene expression in the eye, in some cases up to several years,³⁰ the combination of a ZFP-PEDF activator and an AAV vector could potentially be used in patients with CNV to reduce the frequency of, or perhaps even eliminate the need for, repeated intraocular injections of VEGF antagonists such as Ranibizumab. Our data in cell culture suggests that ZFP-driven PEDF activation can be achieved in multiple cell types, this is supported by the in vivo results that the anti-angiogenic effects are similar whether AAV.ZFP-Pedf activator is injected into the vitreous cavity or subretinal space, despite the fact that cell types transduced by intravitreous injection of AAV have low basal levels of PEDF expression. This lowers the barrier to potential application of PEDF activator to humans, as intravitreous injections can be done in an outpatient setting, and is preferred to subretinal delivery. Moreover, a single intravitreous injection is generally quite safe and well-tolerated. Injection of AAV.ZFP-Pedf activator could therefore be considered for prophylaxis in patients at high risk for the development of CNV. Patients who are at high risk of developing neovascular AMD have been defined; patients with large drusen and pigmentary changes in both eyes have about a 50% chance of developing CNV over 5 years compared to 0.5% in patients lacking these features.³¹ Increasing the anti-angiogenic activity in the eye may reduce the incidence and/or delay the onset of CNV in high risk patients, which would delay the initiation and/or lower the frequency of Ranibizumab injections. This could result in substantial savings in health care expenditures. Notably, although Pedf mRNA and protein levels induced by the ZFP activator were approximately 100-fold less than that produced by the Pedf cDNA in transient transfections (data not shown), the extent of inhibition of CNV driven by the ZFP activator of Pedf in vivo is comparable to the delivery of Pedf cDNA,³² suggesting that a modest increase in PEDF levels (i.e., via activating the endogenous gene) is sufficient to create an anti-angiogenic environment in the eye. Similar to several other endogenous anti-angiogenic factors that do not function at high concentrations,²² very high levels of recombinant PEDF promote angiogenesis in the mouse eye.²¹ Therefore, ZFP-driven PEDF expression may ultimately prove to be the ideal mode for obtaining the intended efficacy with tolerability. Furthermore, ZFPs can be engineered so their activities are pharmacologically regulated. (e.g., Fusion protein consisting of ZFP and the ligand-binding domain of progesterone receptor whose activity can be regulated by antiprogestin.^33,34) This provides additional safety for potential therapeutic applications with a simplified vector design for in vivo implementation compared to two-component protocols (e.g., the Tet-inducible system), which are required for most cDNA-based approaches. Critical to the long-term therapeutic application of the ZFP PEDF activator is its exceptional genome-wide specificity. The PEDF activator affected the expression of remarkably few genes in addition to Pedf itself. In agreement with such high specificity, we observed no overt toxicities or adverse events in animals following 3 months of ZFP expression.

ZFP TF-induced expression of PEDF in the eye resulted in smaller CNV lesions at Bruch's membrane rupture sites, but the smaller lesions still leaked fluorescein. This is consistent with our knowledge regarding the mechanism of the anti-angiogenic effects of PEDF, i.e., the induction of apoptosis in proliferating endothelial cells. While it has been suggested that PEDF has anti-permeability effects, and while we cannot rule out an anti-permeability effect on well-established vessels, our data indicate that PEDF does not prevent leakage from new vessels. On the other hand, Ranibizumab and other VEGF antagonists clearly reduce excessive permeability from CNV, providing rationale for combination therapy. In addition, PEDF may provide additional useful activities because its induction of apoptosis of proliferating endothelial cells is independent of the nature of the pro-angiogenic stimulus and can therefore inhibit the effects of stimuli other than VEGF.^35,36 PEDF also has neurotrophic activity^37,38 which may help protect the photoreceptors in degenerative diseases such as AMD.

In the present study, we engineered and characterized a ZFP activator of PEDF, because it is the most extensively studied anti-angiogenic protein in the eye.^{6,11,12,13,14,15,16,17,18} Intravitreous injection of an adenoviral vector expressing PEDF from a cDNA has been evaluated in a Phase I trial in patients with advanced neovascular AMD and was found to be safe with some possible hints of biological activity.³⁹ However, the versatility and specificity of the ZFP technology means it can be used to increase the expression of other anti-angiogenic proteins instead of or in addition to PEDF, or to decrease the expression of pro-angiogenic proteins. Since the eye is a relatively isolated tissue compartment allowing transduction of ocular cells with little or no effect on extraocular tissues, the possibilities for specifically enhancing the anti-angiogenic environment within the eye are outstanding. The present study provides proof-of-concept for such an approach.

Materials and Methods

Generation of ZFP TFs for Pedf activation. The -400 to +100 bp regions relative to the transcription start sites of the human and mouse Pedf genes were used for ZFP targeting. Six-finger ZFP TFs were assembled by PCR linking three two-finger modules as described.^40,41 Requests for reagents should be addressed to: szhang@sangamo.com

Cell culture and transfections. HEK293 cells, U87MG cells, and mouse Neuro2A cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. ARPE-19 cells were cultured in 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F12 medium with 10% fetal bovine serum. Transient transfections were performed as previously described.⁴² Greater than 80% transfection efficiency was achieved in all experiments, as monitored by a GFP expression plasmid. Transfection media were removed after 8 hours and fresh media were added. Cells were harvested 72 hours after transfection for RNA isolation and Taqman analysis. Culture media were collected for analyzing secreted PEDF by Western blots.

Real-time RT-PCR analysis. For cultured cells, RNA was isolated using High Pure RNA isolation kit (Roche, Indianapolis, IN). For retina and posterior eyecups (retina, RPE, choroid, and sclera), RNA was isolated using TRIzol Plus RNA purification system (Invitrogen, Carlsbad, CA). Real-time RT-PCR (Taqman) assays were performed as previously described.⁴² The levels of Pedf mRNA and 18S RNA were quantified using standard curves spanning a 125-fold concentration range. Each RNA sample was assayed in duplicate Taqman reactions. The ratio of Pedf/18S was used for determining the relative levels of Pedf mRNA. The sequences for the Taqman primer/probes are included as Supplementary Table S1.

Western blot analysis. Seventy-two hours after the HEK293, ARPE-19, and Neuro2A cells were transfected, culture media were collected for assaying PEDF levels by Western blot. The cell numbers at the time of harvesting were not significantly affected by the transfection of either the ZFP-encoding plasmid or the empty vector, as verified by cell counting and the levels of 18S RNA (RT-PCR). Thirty microliter of media from each transfected sample were analyzed by Western blot as previously described²⁴ using a polyclonal anti-PEDF antibody (Bioproducts MD, Middletown, MD).

Microarray analysis. The expression plasmid of ZFP-hPEDF (ZFP-6961-p65) and the empty vector were transfected into U87MG cells in triplicate. Seventy-two hours after transfection, RNA samples were prepared as per the manufacturer's recommendations. Global changes in gene expression were analyzed using a Human Genome Focus GeneChip array and Affymetrix Scanner. Data analysis to determine differentially expressed genes was carried out using Array Assist's (Stratagene, La Jolla, CA) RMA algorithm. The "Change Call" was made using criteria of (i) a twofold difference in expression level between experiment and control, and (ii) P-value < 0.005.

Packaging of Pedf-ZFP TF in AAV vectors. The coding sequence of 6078-p65 was cloned into the AAV-TetO2-MCS vector (Supplementary Figure S1), which was constructed by inserting two copies of the Tet operator sequence 3' of the cytomegalovirus promoter of AAV-MCS vector (Stratagene, La Jolla, CA). The AAV production, purification, and virus genome quantification were performed as described.^43,44

Injections of AAV vectors in mice. Adult C57BL/6 mice were treated humanely in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Experiments were performed independently in the Campochiaro and Ali laboratories using slightly different techniques. In the Campochiaro laboratory, intravitreous or subretinal injections were administered with a Harvard pump microinjection apparatus and pulled glass micropipets as previously described.¹¹ Each micropipet was calibrated to deliver 1 l of vehicle containing 5 10⁸ or 1 10⁹ particles of AAVPedf-ZFP TF or AAVGFP upon depression of a foot switch. In the Ali laboratory, injections of 2 l of vehicle containing 1 10⁹ or 2 10⁹ particles of AAVPedf-ZFP TF or AAVGFP were administered using an operating microscope (for details, see Supplementary Materials and Methods).

Fluorescence microscopy after intravitreous or subretinal injection of AAV.GFP. Six weeks after vector injection, mice were killed and eyes were removed. Eyes were immersed in 4% paraformaldehyde/0.1 mol/l phosphate buffer (pH 7.2) for 2 hours at 4 °C. Following removal of the cornea and lens, posterior eyecups were cryoprotected in 30% sucrose in phosphate-buffered saline for 6–12 hours and then frozen. Ten micrometer frozen sections were examined by fluorescence microscopy using an Axioskop microscope (Zeiss, Thornwood, NY).

Laser-induced CNV. Six weeks after vector injection, Bruch's membrane was ruptured by laser photocoagulation at three locations in each eye as previously described.⁴⁵ The area of CNV at Bruch's membrane rupture sites was measured 2 weeks after laser treatment in choroidal flat mounts from fluorescein-labeled dextran-perfused mice.³² Statistical comparisons were made between groups by linear mixed model⁴⁶ using SAS software (SAS Institute, Cary, NC).

Histology. At 12 weeks after vector injection and 6 weeks after rupture of Bruch's membrane, mice from each group were terminally perfused with 10% formalin. Eyes (n = 6–8 per group) were removed, post-fixed in 10% formalin overnight, processed and embedded in paraffin wax. Eyes were serially sectioned at 4 m thickness, mounted onto polylysine-coated slides, stained with hematoxylin and eosin, and examined by light microscopy.

References

REFERENCES

1. Rosenfeld, PJ, Brown, DM, Heier, JS, Boyer, DS, Kaiser, PK, Chung, CY et al. (2006). Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 355: 1419–1431. | Article | PubMed | ISI | ChemPort |
2. Brown, DM, Kaiser, PK, Michels, M, Soubrane, G, Heier, JS, Kim, RY et al. (2006). Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 355: 1432–1444. | Article | PubMed | ISI | ChemPort |
3. Dawson, DW, Volpert, OV, Gillis, P, Crawford, SE, Xu, H, Benedict, W et al. (1999). Pigment epithelium-derived factor: a potent inhibitor of angiogenesis. Science 285: 245–248. | Article | PubMed | ISI | ChemPort |
4. Tombran-Tink, J, Chader, GG and Johnson, SV (1991). PEDF: a pigment epithelium-derived factor with potent neuronal differentiating activity. Exp Eye Res 53: 411–414. | Article | PubMed | ChemPort |
5. Steele, FR, Chader, GJ, Johnson, LV and Tombran-Tink, J (1993). Pigment epithelium-derived factor: neurotrophic activity and identification as a member or the serine protease inhibitor gene family. Proc Natl Acad Sci USA 90: 1526–1530. | Article | PubMed | ChemPort |
6. Mori, K, Gehlbach, P, Ando, A, McVey, D, Wei, L and Campochiaro, PA (2001). Regression of ocular neovascularization by increased expression of pigment epithelium-derived factor. Invest Ophthalmol Vis Sci 43: 2428–2434. | ISI |
7. Spranger, J, Osterhoff, M, Reimann, M, Mohlig, M, Ristow, M, Francis, MK et al. (2001). Loss of the anti-angiogenic pigment epithelium-derived factor in patients with angiogenic eye disease. Diabetes 50: 2641–2645. | Article | PubMed | ISI | ChemPort |
8. Gao, G, Li, Y, Zhang, D, Gee, S, Crosson, C and Ma, J (2001). Unbalanced expression of VEGF and PEDF in ischemia-induced retinal neovascularization. FEBS Lett 489: 270–276. | Article | PubMed | ISI | ChemPort |
9. Holekamp, NM, Bouck, N and Volpert, OV (2002). Pigment epithelium-derived factor is deficient in the vitreous of patients with choroidal neovascularization due to age-related macular degeneration. Am J Ophthalmol 134: 220–227. | Article | PubMed | ISI | ChemPort |
10. Ogata, N, Nishikawa, M, Nishimura, T, Mitsuma, Y and Matsumura, M (2002). Unbalanced vitreous levels of pigment epithelium-derived factor and vascular endothelial growth factor in diabetic retinopathy. Am J Ophthalmol 134: 348–353. | Article | PubMed | ISI | ChemPort |
11. Mori, K, Duh, E, Gehlbach, P, Ando, A, Takahashi, K, Pearlman, J et al. (2001). Pigment epithelium-derived factor inhibits retinal and choroidal neovascularization. J Cell Physiol 188: 253–263. | Article | PubMed | ISI | ChemPort |
12. Stellmach, V, Crawford, SE, Zhou, W and Bouck, N (2001). Prevention of ischemia-induced retinopathy by the natural ocular anti-angiogenic agent pigment epithelium-derived factor. Proc Natl Acad Sci USA 98: 2593–2597. | Article | PubMed | ChemPort |
13. Mori, K, Gehlbach, P, Yamamoto, S, Duh, E, Zack, DJ, Li, Q et al. (2002). AAV-mediated gene transfer of pigment epithelium-derived factor inhibits choroidal neovascularization. Invest Ophthalmol Vis Sci 43: 1994–2000. | PubMed | ISI |
14. Raisler, BJ, Berns, KI, Grant, MB, Beliaev, D and Hauswirth, WW (2002). Adeno-associated virus type 2 expression of pigmented epithelium-derived factor or Kringles 1-3 of angiostatin reduce retinal neovascularization. Proc Natl Acad Sci USA 99: 8909–8914. | Article | PubMed | ChemPort |
15. Duh, EJ, Yang, HS, Suzuma, I, Miyagi, M, Youngman, E, Mori, K et al. (2002). Pigment epithelium-derived factor suppresses ischemia-induced retinal neovascularization and VEGF-induced migration and growth. Invest Ophthalmol Vis Sci 43: 821–829. | PubMed | ISI |
16. Auricchio, A, Behling, KC, Maguire, AM, O'Conner, EM, Bennett, J, Wilson, JM et al. (2002). Inhibition of retinal neovascularization by intraocular viral-mediated delivery of anti-angiogenic agents. Mol Ther 6: 490–494. | Article | PubMed | ISI | ChemPort |
17. Gehlbach, P, Demetriades, AM, Yamamoto, S, Deering, T, Duh, EJ, Yang, HS et al. (2003). Periocular injection of an adenoviral vector encoding pigment epithelium-derived factor inhibits choroidal neovascularization. Gene Ther 10: 637–646. | Article | PubMed | ISI | ChemPort |
18. Saishin, Y, Silva, RL, Saishin, Y, Kachi, S, Aslam, S, Gong, YY et al. (2005). Periocular gene transfer of pigment epithelium-derived factor inhibits choroidal neovascularization in a human-sized eye. Hum Gene Ther 16: 473–478. | Article | PubMed | ISI | ChemPort |
19. Rebar, EJ, Huang, Y, Hickey, R, Nath, AK, Meoli, D, Nath, S et al. (2002). Induction of angiogenesis in a mouse model using engineered transcription factors. Nat Med 8: 1427–1432. | Article | PubMed | ISI | ChemPort |
20. Dai, Q, Huang, JS, Klitzman, B, Dong, C, Goldschmidt-Clermont, PJ, March, KL et al. (2004). Engineered zinc finger-activating vascular endothelial growth factor transcription factor plasmid DNA induces therapeutic angiogenesis in rabbits with hindlimb ischemia. Circulation 110: 2467–2475. | Article | PubMed | ISI | ChemPort |
21. Apte, RS, Barreiro, RA, Duh, E, Volpert, O and Ferguson, TA (2004). Stimulation of neovascularization by the anti-angiogenic PEDF. Invest Ophthalmol Vis Sci 45: 4491–4497. | Article | PubMed | ISI |
22. Folkman, J (2006). Angiogenesis. Annu Rev Med 57: 1–18. | Article | PubMed | ISI | ChemPort |
23. Dunn, KC, Aotaki-Keen, AE, Putkey, FR and Hjelmeland, LM (1996). ARPE-19, a human retinal pigment epithelial cell line with differentiated properties. Exp Eye Res 62: 155–169. | Article | PubMed | ISI | ChemPort |
24. Tan, S, Guschin, D, Davalos, A, Lee, YL, Snowden, AW, Jouvenot, Y et al. (2003). Zinc-finger protein-targeted gene regulation: genomewide single-gene specificity. Proc Natl Acad Sci USA 100: 11997–12002. | Article | PubMed | ChemPort |
25. Guan, X, Stege, J, Kim, M, Dahmani, Z, Fan, N, Heifetz, P et al. (2002). Heritable endogenous gene regulation in plants with designed polydactyl zinc finger transcription factors. Proc Natl Acad Sci USA 99: 13296–13301. | Article | PubMed | ChemPort |
26. Bennett, J, Duan, D, Engelhardt, JF and Maguire, AM (1997). Real-time, noninvasive in vivo assessment of adeno-associated virus-mediated retinal transduction. Invest Ophthalmol Vis Sci 38: 2857–2863. | PubMed | ISI | ChemPort |
27. Flannery, JG, Zolotukhin, S, Vaquero, MI, LaVail, MM, Muzyczka, N and Hauswirth, WW (1997). Efficient photoreceptor-targeted gene expression in vivo by recombinant adeno-associated virus. Proc Natl Acad Sci USA 94: 6916–6921. | Article | PubMed | ChemPort |
28. Kwak, N, Okamoto, N, Wood, JM and Campochiaro, PA (2000). VEGF is an important stimulator in a model of choroidal neovascularization. Invest Ophthalmol Vis Sci 41: 3158–3164. | PubMed | ISI | ChemPort |
29. Kryzstolik, MG, Afshari, MA, Adamis, AP, Gaudreault, J, Gragoudas, ES, Michaud, NM et al. (2002). Prevention of experimental choroidal neovascularization with intravitreal anti-vascular endothelial growth factor antibody fragment. Arch Ophthalmol 120: 338–346. | PubMed | ISI | ChemPort |
30. Acland, GM, Aguirre, GD, Bennett, J, Aleman, TS, Cideciyan, AV, Bennicelli, J et al. (2005). Long-term restoration of rod and cone vision by single dose rAAV-mediated gene transfer to the retina in a canine model of childhood blindness. Mol Ther 12: 1072–1082. | Article | PubMed | ISI | ChemPort |
31. Ferris, FL, Davis, MD, Clemons, TE, Lee, LY, Chew, EY, Lindblad, AS et al. (2005). A simplified severity scale for age-related macular degeneration: AREDS report No. 18. Arch Ophthalmol 123: 1570–1574. | PubMed |
32. Mori, K, Ando, A, Gehlbach, P, Nesbitt, D, Takahashi, K, Goldsteen, D et al. (2001). Inhibition of choroidal neovascularization by intravenous injection of adenoviral vectors expressing secretable endostatin. Am J Pathol 159: 313–320. | PubMed | ISI | ChemPort |
33. Beerli, RR, Schopfer, U, Dreier, B and Barbas, CF 3rd (2000). Chemically regulated zinc finger transcription factors. J Biol Chem 275: 32917–32326. | Article |
34. Dent, CL, Lau, G, Drake, EA, Yoon, A, Case, CC and Gregory, PD (2007). Regulation of endogenous gene expression using small molecule-controlled engineered zinc-finger protein transcription factors. Gene Ther (in press).
35. Volpert, OV, Zaichuk, T, Zhou, W, Reiher, F, Ferguson, TA, Stuart, PM et al. (2002). Inducer-stimulated Fas targets activated endothelium for destruction by anti-angiogenic thrombospondin-1 and pigment epithelium-derived factor. Nat Med 8: 349–357. | Article | PubMed | ISI | ChemPort |
36. Zaichuk, TA, Shroff, EH, Emmanuel, R, Filleur, S, Nelius, T and Volpert, OV (2004). Nuclear factor of activated T cells balances angiogenesis and inhibition. J Exp Med 199: 1513–1522. | Article | PubMed | ISI | ChemPort |
37. Araki, T, Taniwaki, T, Becerra, SP, Chader, GJ and Schwartz, JP (1998). Pigment epithelium-derived factor (PEDF) differentially protects immature but not mature cerebellar granule cells against apoptotic cell death. J Neurosci Res 53: 7–15. | Article | PubMed | ISI | ChemPort |
38. Bilak, MM, Corse, AM, Bilak, SR, Lehar, M, Tombran-Tink, J and Kuncl, RW (1999). Pigment epithelium-derived factor (PEDF) protects motor neurons from chronic glutamate-mediated neurodegeneration. J Neuropathol Exp Neurol 58: 719–728. | PubMed | ISI | ChemPort |
39. Campochiaro, PA, Nguyen, QD, Shah, SM, Klein, ML, Holz, E, Frank, RN et al. (2006). Adenoviral vector-delivered pigment epithelium-derived factor for neovascular age-related macular degeneration: results of a phase I clinical trial. Hum Gene Ther 17: 167–176. | Article | PubMed | ISI | ChemPort |
40. Moore, M, Choo, Y and Klug, A (2001). Design of polyzinc finger peptides with structured linkers. Proc Natl Acad Sci USA 98: 1432–1436. | Article | PubMed | ChemPort |
41. Moore, B, Klug, A and Choo, Y (2001). Improved DNA binding specificity from polyzinc finger peptides by using strings of two-finger units. Proc Natl Acad Sci USA 98: 1437–1441. | Article | PubMed | ChemPort |
42. Liu, PQ, Rebar, EJ, Zhang, L, Liu, Q, Jamieson, AC, Liang, Y et al. (2001). Regulation of an endogenous locus using a panel of designed zinc finger proteins targeted to accessible chromatin regions. Activation of vascular endothelial growth factor A. J Biol Chem 276: 11323–11334. | Article | PubMed | ISI | ChemPort |
43. Matsushita, T, Elliger, S, Elliger, C, Podsakoff, G, Villarreal, L, Kurtzman, GJ et al. (1998). Adeno-associated virus vectors can be efficiently produced without helper virus. Gene Ther 5: 938–945. | Article | PubMed | ISI | ChemPort |
44. Wright, JF, Qu, G, Tang, C and Sommer, JM (2003). Recombinant adeno-associated virus: formulation challenges and strategies for a gene therapy. Curr Opin Drug Discov Devel 6: 174–178. | PubMed | ISI | ChemPort |
45. Tobe, T, Ortega, S, Luna, JD, Ozaki, H, Okamoto, N, Derevjanik, NL et al. (1998). Targeted disruption of the FGF2 gene does not prevent choroidal neovascularization in a murine model. Am J Pathol 153: 1641–1646. | PubMed | ISI | ChemPort |
46. Verbeke, G and Molenberghs, G (2000). Linear Mixed Models for Longitudinal Data. Springer-Verlag, Inc.: New York.

Acknowledgments

Supported by a grant from Sangamo BioSciences, a senior scientist award from Research to Prevent Blindness, New York, NY, the Moorfield's Eye Hospital Special Trustees, William and Jean Lake and Richard and Charlotte Heffner. P.A.C. is the George S. and Dolores Dore Eccles Professor of Ophthalmology. We thank Tim Farries, Lutz-Peter Berg, and John Girdleston for ZFP designs for mouse Pedf, Jeffrey Miller, and Ed Rebar for ZFP designs for human Pedf. We thank Sarah Hinkley for assistance in ZFP construction, Gary Lee and Nhu Tran for AAV production. We also thank Sean Brennan for careful reading of the manuscript.