Endogenic Regulation of Proliferation and Zinc Transporters by Pigment Epithelial Cells Nonvirally Transfected with PEDF

Abstract

Purpose.: Genetic modification of cells before transplantation may allow the delivery of neuroprotective and other functional molecules to patients with neurodegenerative diseases. To avoid complications associated with virally transfected cells, we have explored the use of nonviral methods to insert genetic material into RPE cells.

Methods.: After transfection with plasmids encoding different pigment epithelium-derived factor (PEDF) fusion proteins, transfected cells were established and passaged up to 100 times. Gene expression of PEDF, ZnT3, ZIP2, CRALBP, CATD, and ZO-1 was determined by RT-PCR. Secretion dynamics were analyzed using ELISA and a spheroid-based assay was used to confirm the anti-angiogenic activity of the recombinant PEDF.

Results.: Transfection efficiency reached up to 98.7% with a plasmid encoding PEDF and enhanced green fluorescent protein (EGFP) separately and 87.2% with a plasmid encoding an EGFP–PEDF fusion. Immunoblotting revealed that transfected RPE cells express the appropriate PEDF or EGFP–PEDF. Expression of recombinant PEDF is stable, as shown by its secretion for the 2 years and the 100 passages the cells have been followed. PEDF expression was overexpressed and the transfected cells exhibited increased proliferation, up-regulation of ZnT3 and ZIP2, and inhibited sprouting in human umbilical vein endothelial cell spheroids.

Conclusions.: Genetic in vitro modification of pigment epithelial cells using nonviral transfection protocols should improve the potential therapeutic treatment of neurodegenerative diseases by transplantation of genetically modified cells without the disadvantages of virally mediated transfection. Here we have shown that genetically modified RPE cells overexpress a functional human recombinant PEDF, as evidenced by the autogenic regulation of proliferation, up-regulation of two distinct zinc transporters, and in vitro inhibition of endothelial cell sprouting.

AMD is believed to be the result of alterations to the RPE and Bruch's membrane caused by the long-term oxidative stress to RPE cells. ¹ In susceptible individuals, the long-term oxidative stress may lead to the degradation of RPE functions, such as a decrease in outer segment degradation, accumulation of undigested material on the basal lamina, formation of advanced glycation end products (AGEs) ² that damage the basal lamina, increased matrix metalloproteinase-2, ³ and the production of VEGF. ⁴ In addition, AGEs may decrease the production of the neurogenic and anti-angiogenic pigment epithelium-derived factor (PEDF) as has been shown in retinal microvascular endothelial cells. ⁵ The increased production of VEGF, the decreased production of PEDF, and the alterations to the basal lamina may be the impetus for the development of exudative AMD. 

Treatment options for exudative AMD have included laser treatment and membrane removal alone or coupled with macular translocation or pigment cell transplantation; however, none of these treatment modalities have resulted in significant visual improvement. VEGF plays a major role in the development of neovascularization in exudative AMD. This has been amply documented by the significant improvement in visual function in exudative AMD patients treated with a number of different VEGF inhibitors. ^{6 –10} Recently, treatment with a humanized Fab fragment of a monoclonal anti-VEGF antibody (ranibizumab) ^7,8 and with the monoclonal antibody itself (bevacizumab) ^9,10 has become widespread for the treatment of exudative AMD and other ocular neovascular diseases, such as diabetic retinopathy, ¹¹ diabetic macular edema, ¹² choroidal neovascularization associated with pathologic myopia, ¹³ and neovascular glaucoma. ^14,15 Even though the development of VEGF inhibitors has transformed the treatment of exudative AMD and other ocular neovascular diseases, the treatment is costly, demanding in logistics of the mostly elderly patients, and certainly not without side effects, such as endophthalmitis, ocular hypertension, and submacular hemorrhage. ^{16 –18} These side effects appear related to the short half-life of the anti-VEGFs and the repeated intravitreal injections necessary to prevent the recurrence of neovascularization. Even though data on long-term side effects are not known, the inhibitory effects of bevacizumab and ranibizumab on RPE cell proliferation and the inhibition of outer segment phagocytosis by bevacizumab suggest that long-term treatment with anti-VEGFs may lead to undesirable side effects. ¹⁹ Ideally, and to avoid possible adverse side effects, inhibition of neovascularization should be achieved by the continuous delivery of a naturally occurring ocular inhibitor of vascularization. Such therapeutic modality can be achieved by having cells being a continuous delivery system. For this purpose, we have transfected human RPE cells with a plasmid encoding the gene for the antivascular and neurogenic factor PEDF. Using nonviral protocols, RPE cells can be transfected with high efficiency, the stably transfected cells secrete a continuous and constant level of recombinant functional PEDF, and the overexpression of PEDF results in the autoregulation of some cells' intrinsic functionalities, as shown by the increased proliferation of the transfected cells and the up-regulation of two zinc transporters, ZnT3 and ZIP2. 

Material and Methods

Cell Culture

ARPE-19 cells (ATCC No.: CRL-2302) were maintained and passaged weekly (1:10) according to the depositors recommendations. 

Electroporation and Establishment of RPE Transfectants

Construction of the plasmids pMS-N-PEDF-H-IV and pMS-L-EGFP-PEDF-H-IV has been described previously. ²⁰ For electroporation, either 4 μg of PvuI linearized plasmid encoding PEDF and EGFP (Fig. 1A) or 4 μg of PvuI linearized plasmid encoding EGFP-PEDF (Fig. 1B) were added to 10⁶ RPE cells. A total of 100 μL of celltype specific nucleofector solution (Amaxa, Cologne, Germany) was added to each cell sample and an electric pulse was applied using the Nucleofector I (Amaxa; program X-01). Cells exposed to the same electric field without the addition of plasmid DNA and to plasmid DNA without exposure to an electric field served as controls. In both plasmids, gene expression was under the control of a constitutive cytomegalovirus promoter, and the inclusion of a Zeocin resistance gene facilitated the selection of the transfected cells. 

Figure 1.

 

Schematic drawing of the recombinant PEDF and EGFP–PEDF encoding plasmids pMS-N-PEDF-H-IV (A) and pMS-L-EGFP-PEDF-H-IV (B) used to transfect ARPE-19 cells.

After electroporation, cells were transferred into 12-well tissue culture plates in 2 mL complete medium (Dulbecco's modified Eagle's medium [DMEM]/Ham's F-12; Biochrom AG, Berlin, Germany) supplemented with 10% fetal bovine serum (FBS; PAA Laboratories, Pasching, Austria), 80 U/mL penicillin, and 80 U/mL streptomycin (Lonza, Basel, Switzerland) and maintained at 37°C in a humidified atmosphere of 95% air and 5% CO₂. After 3 days, cell culture supernatants were analyzed for secretion of recombinant PEDF (rPEDF) and the transfected RPE cells were transferred into 75 cm² cell culture flasks and grown in complete medium supplemented with 10% FBS, 80 U/mL penicillin, 80 U/mL streptomycin, and 100 μg/mL Zeocin (Invitrogen, Carlsbad, CA) in a humidified atmosphere of 95% air and 5% CO₂. After 2 weeks, green fluorescent (transfected) colonies were identified by fluorescence microscopy, isolated and subcultured to establish transfected cell populations. Established ARPE-19 transfectants were passaged weekly. 

Flow Cytometric Analysis

Transfected cells express EGFP, either alone or fused to PEDF, facilitating the identification of transfected cells and determination of transfection rates by their fluorescence using flow cytometry. For this analysis, RPE cell monolayers were trypsinized using 0.05% trypsin-0.02% EDTA (PAA Laboratories), washed 3 times in PBS, resuspended in 500 μL PBS, and analyzed using a FACSCalibur (Becton Dickinson, Heidelberg, Germany). 

Protein Purification and Quantification

Purification of the His-tagged PEDF and EGFP–PEDF was accomplished by nickel nitrilotriacetic acid (Ni-NTA) metal affinity chromatography (Qiagen, Hilden, Germany) as described previously. ²⁰ Cell culture media were collected and clarified by centrifugation for 10 minutes at 1100 rpm at 4°C. For Western blot analysis, 900 μL of the supernatant was mixed with 30 μL of 50% Ni-NTA slurry in 300 μL of 4× incubation buffer (200 mM NaH₂PO₄, pH 8.0, 1.2 M NaCl, and 40 mM imidazole). After 1 hour incubation at room temperature, the Ni-NTA resin was pelleted by centrifugation for 1 minute at 5000 rpm and washed twice with 175 μL incubation buffer. Bound protein was eluted by incubation with 30 μL of elution buffer (50 mM NaH₂PO₄, pH 8.0, 300 mM NaCl, and 250 mM imidazole) for 20 minutes at room temperature. For functional assays, the His-tagged PEDF proteins were purified using Ni-NTA spin columns (Qiagen). Clarified cell culture media (3600 μL) were mixed with 1200 μL of 4× incubation buffer and applied to the columns. Bound protein was eluted with 100 μL of elution buffer. 

For PEDF and VEGF quantification, 350,000 cells from the stably PEDF-transfected RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} and from nontransfected ARPE-19 were plated in 2 mL of complete medium (DMEM/Ham's F-12 supplemented with 10% FBS, 80 U/mL penicillin, and 80 U/mL streptomycin) and maintained at 37°C in a humidified atmosphere of 95% air and 5% CO₂. After 20 hours, the cells were lysed using radioimmunoprecipitation assay (RIPA) buffer (Pierce Biotechnology, Rockford, IL) according to the manufacturer's protocol and the culture media were clarified by centrifugation at 1100 rpm for 15 minutes. Both cell lysates and clarified culture media were analyzed by ELISA using the ELISAquant kit for human PEDF (BioProducts MD, Middletown, MD) and the immunoassay kit for human VEGF (Invitrogen) according to the manufacturer's protocols. 

Western Blot Analysis

For SDS-polyacrylamide gel electrophoresis, 15 μL of Ni-NTA purified protein were mixed with 2× SDS sample buffer, ²¹ heated for 5 minutes at 95°C, and separated on 10% SDS-polyacrylamide gels. After electrophoresis, the proteins were transferred onto a 0.45-μm pore size nitrocellulose membrane (Whatman, Maidstone, Kent, UK) using the semidry blotting technique (Bio-Rad Laboratories, Hercules, CA), followed by Ponceau S staining to confirm transfer. Blots blocked with 3% BSA/tris-buffered saline (TBS) overnight at 4°C were incubated for 2 hours at room temperature with anti-Penta-His antibodies (mouse monoclonal, 1:500; Qiagen) or (for confirmation) anti-PEDF (I-15; goat polyclonal, 1:200; Santa Cruz Biotechnology, Santa Cruz, CA) diluted in 3% BSA/TBS. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was detected with anti-GAPDH antibodies (mouse monoclonal, 1:5000; Novus Biologicals, Littleton, CO). Afterward, the blots were incubated for 1 hour at room temperature with horseradish peroxidase-conjugated anti-mouse antibodies (rabbit polyclonal, 1:1000; Dako, Glostrup, Denmark) for the anti-Penta-His or with horseradish peroxidase-conjugated anti-goat antibodies (donkey, 1:1000; Santa Cruz Biotechnology) for the PEDF specific detection, both diluted in 10% milk powder/TBS. Protein bands were visualized by chemiluminescence using the LAS-3000 imaging system (FujiFilm, Tokyo, Japan). 

Autogenic Regulation of Cell Proliferation

To define the autogenic effect of the PEDF overexpression, nontransfected ARPE-19 cells and the stably PEDF-transfected ARPE-19 cell lines RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} were seeded in 96-well plates at an initial density of 2500 cells/cm². Medium was changed on days 3, 7, and 10. Cell proliferation was monitored with crystal violet every 1 or 2 days beginning at day 1 until day 14 of culture. Briefly, the medium was removed and 100 μL crystal violet solution (0.5% in 20% methanol) were added. After 10 minutes, the cells were rinsed three times with water. After overnight drying, 50 μL sodium citrate solution (10 mM in 50% ethanol) were added to each plate and the absorbance was measured at 595 nm using an ELISA reader. Each cultivation time was tested in six individual wells in triplicate experiments. 

Autogenic Regulation of Zinc Transporters

Nontransfected ARPE-19 cells and the two established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} were grown to confluence, washed twice with zinc-free and serum-free MEM, and cultured in zinc-free and serum-free MEM. After 48 hours, total RNA was isolated from all cultures and used to analyze the expression of the zinc transporters ZnT3 and ZIP2. ²²  

RNA Isolation, RT-PCR, and Quantitative Real-Time PCR

Total RNA was isolated using the RNeasy Mini Kit (Qiagen) together with the RNase-free DNase Set (Qiagen) according to the manufacturer's protocol. Reverse transcription was carried out on 1 μg of total RNA using the Reverse Transcription System (Promega, Madison, WI) according to the manufacturer's instructions. RT-PCR (GoTaq Hot Start Polymerase, Promega) was performed at an annealing temperature of 58°C for 40 cycles for PEDF (F: 5′-TAG ACC GAG AAC TGA AGA CCG-3′; R: 5′-TGT GTC TGT GTC CCT CAG TAC-3′), ZnT3 ²² (F: 5′-ATG GTC ACT GGC ATC CTC CTG TA-3′; R: 5′-AGA TGG AGA AGA GGA AGG TGC TGA-3′), ZIP2 ²² (F: 5′-CCC TTG TCC TCT TGC TGT CAC TCT-3′; R: 5′-AGC TCC CGT GGA AGA ATT TCT AGG-3′), CRALBP (F: 5′-TCA CCA CGA CCT ACA ATG TGG-3′; R: 5′-AAC TAC AGT TCA GCT GGC AGG-3′), CATD (F: 5′-AGG CAA AGG CTA CAA GCT GTC-3′; R: 5′-TGT GCT CTG GAT CAG CTC TAC-3′), and ZO-1 (F: 5′-ACA CTG CTG AGT CCT TTG GTG-3′; R: 5′-CTA GCC AAT ACC AAC AGT CCC-3′) primers. The housekeeping gene GAPDH (F: 5′-ATC CCA TCA CCA TCT TCC AG-3′; R: 5′-ATG AGT CCT TCC ACG ATA CC-3′) was used as the internal RNA loading control. Specificity of the selected primers was ascertained by DNA sequence analysis (Eurofins MWG Operon, Ebersberg, Germany). PCR products were resolved by gel electrophoresis on 1.5% agarose gels. 

Quantitative real-time PCR was performed on a LightCycler 1.2 using the LightCycler FastStart DNA Master SYBR Green I kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol. A single run included the analysis of the PEDF gene together with the internal control gene GAPDH. Reactions were performed with diluted cDNA, corresponding to 20 ng of initially used total RNA, 4 mM MgCl₂, and a primer concentration of 0.2 μM. The following LightCycler conditions were used: initial denaturation at 95°C for 10 minutes, followed by 45 cycles denaturation at 95°C for 10 seconds, annealing at 62°C for 7 seconds, and elongation at 72°C for 15 seconds. Subsequently, a melting curve analysis was performed to confirm amplification specificity from each primer pair. To analyze the expression levels of the PEDF gene the 2^-ΔΔCT method, which expresses the relative gene expression ratio, ^23,24 was used. The data for the established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} were expressed as change in gene expression normalized to the internal control gene (ΔCP) and relative to the values of the nontransfected ARPE-19 cells (ΔΔCP). By definition, the ΔΔCP for nontransfected ARPE-19 cells equaled 0, resulting in a relative gene expression rate of 1. For the established ARPE-19 transfectants, a ratio of 1 would indicate a gene expression similar to the nontransfected cells, whereas a ratio higher than 1 would indicate an up-regulation of gene expression and a ratio lower than 1 would indicate gene downregulation. 

Spheroid-Based Anti-Angiogenesis Assay

The spheroid-based anti-angiogenic assay was carried out according to the procedure of Korff and Augustin. ²⁵ Human umbilical vein endothelial cells (HUVECs) were isolated and cultured in endothelial cell growth medium (PromoCell, Heidelberg, Germany). Spheroids of defined cell numbers were generated by suspending HUVECs in endothelial cell growth medium containing 20% (w/v) carboxymethylcellulose (Sigma-Aldrich, Munich, Germany) and seeding 25-μL drops containing 400 cells on nonadherent plastic Petri dishes (Greiner Bio-One, Frickenhausen, Germany). After overnight incubation at 37°C in a humidified atmosphere of 95% air and 5% CO₂, the spheroids were embedded into collagen gels and rapidly transferred to 24-well tissue culture plates. After 30 minutes at 37°C, to allow the collagen to polymerize, 100 μL of conditioned medium from 2-day-old cultures of stably transfected RPE^{[N-PEDF-H-IV]} or nontransfected ARPE-19 cells in the presence or absence of endothelial cell growth medium supplement mix (0.02 mL/mL fetal calf serum, 0.004 mL/mL endothelial cell growth supplement, 0.1 ng/mL epidermal growth factor, 1.0 ng/mL recombinant human basic fibroblast growth factor, 90 μg/mL recombinant human heparin, and 1.0 μg/mL hydrocortisone; PromoCell) used as stimulus were layered on top of the gels in triplicate experiments. To ascertain the effect of PEDF secreted by the transfected cells, spheroids were incubated with conditioned medium that was depleted of rPEDF by treatment with Ni-NTA resin for 2 hours, followed by centrifugation at 5000 rpm for 5 minutes. 

After 24 hours, capillary sprouting was quantified by measuring the cumulative length using digital imaging software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html). ²⁶ The mean cumulative sprout length per spheroid was calculated by measuring and averaging the length of all sprouts for 10 randomly selected spheroids. Data were expressed as mean ± SEM and analyzed for statistical significance using a two-tailed Student's t-test. P < 0.05 was considered statistically significant. 

Secretion of rPEDF by Subconjunctivally Implanted Stably Transfected Cells

Animal experiments were conducted in accordance with the principles for the care and use of research animals and were carried out with the permission and supervision of the District Government of Cologne, Germany. Stably transfected RPE^{[N-PEDF-H-IV]} cells were cultured in complete medium supplemented with 10% FBS, 80 U/mL penicillin, and 80 U/mL streptomycin, trypsinized using 0.05% trypsin-0.02% EDTA, washed three times with 0.9% NaCl, and suspended in 0.9% NaCl at a concentration of 300,000 cells/μL; 10 μL containing 3 million cells were injected into the rabbit conjunctiva using a 27-gauge needle. Rabbits were killed at 3 and 7 days after injection. The eyes were enucleated and dissected into conjunctiva, sclera, RPE/choroid, neural retina, and vitreous. The tissues were lysed using 5 μL RIPA buffer/mg tissue and subjected to Western blot analysis for the detection of PEDF fusion protein and GAPDH. rPEDF was detected using anti-Penta-His antibodies, and GAPDH was detected using mouse monoclonal antibodies. 

Results

Expression Dynamics of the Established ARPE-19 Transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}

The simultaneous expression of PEDF and EGFP by the RPE^{[N-PEDF-H-IV]} transfectants and the expression of the fusion protein EGFP–PEDF by the RPE^{[L-EGFP-PEDF-H-IV]} transfectants were used to determine the percentage of transfected cells using flow cytometric analysis. The results showed that 98.7% of the RPE^{[N-PEDF-H-IV]} cells (Fig. 2A) and 87.2% of the RPE^{[L-EGFP-PEDF-H-IV]} cells (Fig. 2B) were transfected. 

Figure 2.

 

Expression of EGFP in the established transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}. Flow cytometric analysis of the Zeocin-selected ARPE-19 transfectants showed that 98.7% of the cells were transfected with pMS-N-PEDF-H-IV (A) and 87.2% of the cells were transfected with pMS-L-EGFP-PEDF-H-IV (B). Nontransfected ARPE-19 cells were used as control.

Western blot analysis of Ni-NTA–purified rPEDF showed that after transfection there is an increase in rPEDF and rEGFP–PEDF amount from day 3 to day 14, indicating proliferation of the transfected cells and continuous secretion of the recombinant protein (Fig. 3A). The secretion of rPEDF and rEGFP–PEDF into the media remained constant at high levels for the 2 years that the cells have been followed (Fig. 3B). The stable incorporation of the human PEDF gene into the human pigment epithelial cell line using nonviral protocols was evident from the secretion of rPEDF into the culture media for 2 years from the date of electroporation and after 100 passages, both in the presence or absence of Zeocin selection pressure (Fig. 3C). Comparison of the rPEDF levels in the culture media and the cell lysates from RPE^{[N-PEDF-H-IV]} (N-P) and RPE^{[L-EGFP-PEDF-H-IV]} (L-P) transfected cells showed that the majority of the rPEDF was secreted into the media (Fig. 3D). Endogenous PEDF levels in the culture medium and cell lysate of non-transfected ARPE-19 cells were below the detection limit (Fig. 3D, Co). Note that rPEDF secreted by RPE^{[N-PEDF-H-IV]} and EGFP-PEDF secreted by RPE^{[L-EGFP-PEDF-H-IV]} cells are of the expected molecular weight (∼48 and ∼78 kDa, respectively). The rather constant levels of GAPDH protein observed in the cell lysates for the 1- and 2-year cultures and the cell lysates from transfected cells grown in the presence or absence of Zeocin (Figs. 3B–D) confirmed that equal amounts of protein were loaded on the gels. 

Figure 3.

 

Western blot analysis of purified PEDF and EGFP–PEDF fusion protein from the established transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}. (A) After transfection, cells in culture showed increased rPEDF secretion into the media over a period of 14 days, indicating cell proliferation and continuous protein secretion of the transfected cells. (B) Secretion continues for the 2 years the cells have been analyzed. The use of the same cell numbers from which the medium was obtained and the similar gel loading, as shown by the similar band intensities of the internal GAPDH control at 1 and 2 years, indicated not only the continuous secretion of rPEDF, but also an increase in rPEDF secreted into the media especially by RPE^{[L-EGFP-PEDF-H-IV]} transfected cells. (C) Analysis of Ni-NTA–purified proteins from cell culture supernatants 2 years after electroporation showed nearly similar expression levels of rPEDF in the presence or absence of Zeocin. Similar loading was assured by the similar band density of GAPDH. (D) Analysis of cell culture media and the respective cell lysates indicated that most of the PEDF fusion protein was secreted. The rPEDF from RPE^{[N-PEDF-H-IV]} transfected cells (N-P) and the EGFP–PEDF fusion protein from RPE^{[L-EGFP-PEDF-H-IV]} transfected cells (L-P) were detected using anti-Penta-His antibodies. Note the expected molecular weights of rPEDF and recombinant EGFP–PEDF (∼48 and ∼76 kDa, respectively). Loading of equal protein amounts was assured by the similar density of GAPDH protein bands, which migrated with the expected molecular weight of ∼36 kDa.

Quantification of rPEDF Secretion by the Established ARPE-19 Transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}

Using ELISA, total PEDF levels were analyzed in the culture media and cell lysates of 350,000 cells from nontransfected ARPE-19 and from the established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}. After 20 hours in culture, nontransfected ARPE-19 cells secreted 0.13 ng PEDF per hour per 100,000 cells, whereas RPE^{[N-PEDF-H-IV]} cells secreted 32 ng and RPE^{[L-EGFP-PEDF-H-IV]} cells secreted 8.26 ng PEDF per hour per 100,000 cells (Table 1). The cell lysates contained 0.02 ng (for nontransfected ARPE-19), 0.46 ng (for RPE^{[N-PEDF-H-IV]}), and 0.58 ng PEDF per hour per 100,000 cells (for RPE^{[L-EGFP-PEDF-H-IV]}; Table 1). The levels of secreted VEGF were 0.002 ng for nontransfected ARPE-19, 0.001 ng for RPE^{[N-PEDF-H-IV]}, and 0.006 ng VEGF per hour per 100,000 cells for RPE^{[L-EGFP-PEDF-H-IV]} (Table 1). 

Table 1.

 

PEDF and VEGF Levels Secreted into the Culture Media and in Cell Lysates of Nontransfected ARPE-19 and Stably Transfected RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} Cells

Culture	PEDF		VEGF
	Lysate (ng/h/105 cells)	Culture Medium (ng/h/105 cells)	Culture Medium (ng/h/105 cells)
ARPE-19	0.02	0.13	0.002
RPE[N-PEDF-H-IV]	0.46	32.00	0.001
RPE[L-EGFP-PEDF-H-IV]	0.58	8.26	0.006

 PEDF, pigment epithelium-derived factor; RPE, retinal pigment epithelium; VEGF, vascular endothelial growth factor.

Overexpression of Human PEDF in the Established ARPE-19 Transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}

Semiquantitative RT-PCR clearly showed the overexpression of the PEDF gene in both established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}. In nontransfected ARPE-19 cells, the PEDF PCR fragment was only apparent up to a concentration of 1 ng of initial total RNA (Fig. 4A, column 1, lanes 1–3), whereas in the established transfected cell lines, the PEDF PCR fragment was evident up to a concentration of 0.008 ng of initial total RNA (Fig. 4A, columns 2 and 3, lanes 1–6). In contrast, only minor changes were observed for the GAPDH PCR fragment between nontransfected and stably PEDF-transfected cells (Fig. 4A). The up-regulation of PEDF gene expression was also clearly evident from the results of the quantitative RT-PCR (Fig. 4B), where the relative PEDF gene expression ratio in nontransfected ARPE-19 cells was set to 1. 

Figure 4.

 

Gene expression profile in the established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]}. (A) Composite photographs show RT-PCR products from total RNA for the housekeeping gene GAPDH and for the genes PEDF, ZnT3, ZIP2, CRALBP, CATD, and ZO-1 prepared from nontransfected ARPE-19 (left panel), RPE^{[N-PEDF-H-IV]} (center panel), and RPE^{[L-EGFP-PEDF-H-IV]} (right panel) cells. Lanes 1 to 6 showed a series of 1:5 cDNA dilutions from 25 ng (lane 1) to 0.008 ng (lane 6) initially used total RNA. In RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} cells, PEDF gene expression was clearly overexpressed and gene expression of ZnT3 and ZIP2 was up-regulated, whereas expression of the RPE differentiation marker genes CRALBP, CATD, and ZO-1 remained unchanged compared to nontransfected ARPE-19 cells. (B) Quantitative real-time PCR illustrates the increased gene expression ratio of PEDF relative to the housekeeping gene GAPDH. The relative gene expression ratio of PEDF in nontransfected ARPE-19 cells was set to 1.

Autogenic Regulation of ZnT3 and ZIP2 Zinc Transporter

The functionality of the rPEDF produced by the stably transfected ARPE-19 cells was evident by the up-regulation of gene expression for the two zinc transporters ZnT3 and ZIP2. Compared to the expression of nontransfected ARPE-19 cells, semiquantitative gene expression analysis of ZnT3 and ZIP2 clearly showed that these two genes were up-regulated in both of the established transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} (Fig. 4A). In contrast, gene expression of CRALBP (retinol recycling), CATD (phagocytosis), and ZO-1 (tight junctions) was unchanged (Fig. 4A). 

Autogenic Regulation of Proliferation

During the first 5 days in culture, no differences in proliferation rates were evident between transfected and nontransfected ARPE-19 cells (Fig. 5). However, after day 5, the proliferation rates of the transfected cells increased exponentially and the cells reached confluence and became contact inhibited by day 8, whereas the nontransfected cells proliferated at a much lower rate, had not reached confluence by day 14, and the cell number was 30% below the number of the transfected cells at day 8. 

Figure 5.

 

Proliferation rates of the established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} compared to nontransfected ARPE-19 cells. Beginning at day 5 of culture, the proliferation rates of RPE^{[N-PEDF-H-IV]} and RPE^{[L-EGFP-PEDF-H-IV]} increased markedly compared to that of the nontransfected cells. The transfected cells reached confluence by day 8, whereas the nontransfected cells were not confluent by day 14. The cell number in nontransfected ARPE-19 cells at day 14 was 30% less than in the cultures of the transfected cells at day 8.

In Vitro Inhibition of Endothelial Cell Sprouting

Anti-angiogenic activity of rPEDF was examined by analyzing in vitro microvessel formation by culturing HUVEC spheroids in medium conditioned by ARPE-19 cells, by stably transfected RPE^{[N-PEDF-H-IV]} cells, and by PEDF-depleted conditioned medium from stably transfected RPE^{[N-PEDF-H-IV]} cells. The average cumulative sprout length in endothelial cell spheroids was 692 ± 84 μm (Fig. 6). The addition of endothelial cell growth medium supplement mix to conditioned medium from ARPE-19 cells stimulated sprouting and cumulative sprout length reached 4130 ± 516 μm (P < 0.005), whereas in the presence of conditioned medium from the established ARPE-19 transfectants RPE^{[N-PEDF-H-IV]} and endothelial cell growth medium supplement mix, cumulative sprout length was only 1625 ± 445 μm (Fig. 6). To show that the inhibition of sprouting was caused by the rPEDF, conditioned medium from RPE^{[N-PEDF-H-IV]} cells was depleted of rPEDF by treatment with Ni-NTA resin. The depleted conditioned medium showed a significant increase in cumulative sprout length reaching 3648 ± 234 μm, or 55% above not-depleted conditioned medium (P < 0.005; Fig. 6). 

Figure 6.

 

The spheroid-based anti-angiogenesis assay confirmed the anti-angiogenic activity of rPEDF. Cumulative sprout length was determined from all sprouts generated from five randomly selected spheroids per well, three wells for each condition. Composite photographs show representative spheroids. Supplementing the conditioned medium of ARPE-19 cells with markedly increased sprouting from 692 ± 82 μm to 4130 ± 516 μm can be seen by the graph (compare first and second bar) and from the spheroid micrographs. Conditioned medium from stably PEDF-transfected RPE^{[N-PEDF-H-IV]} (RPE^[PEDF]) cells inhibited significantly the stimulation of sprouting by the endothelial cell growth medium supplement mix; cumulative sprouting was only 1625 ± 445 μm (compare second and third bar and the spheroid micrographs). Depletion of rPEDF from the conditioned medium of stably PEDF-transfected RPE^[PEDF] cells almost entirely abolished the inhibition of sprouting (3648 ± 234 μm, compare third and fourth bar and the spheroid micrographs).

In Vivo Secretion of rPEDF by Stably Transfected Cells

Secretion of rPEDF in vivo by stably transfected cells was analyzed after subconjunctival cell implantation. Three days after implantation, Western blot analysis of ocular tissues showed evidence of rPEDF secretion by RPE^{[N-PEDF-H-IV]} cells in the conjunctiva, and diffusion into the sclera and the RPE/choroid complex, but not into the neural retina or vitreous (Fig. 7A). At 7 days postimplantation, rPEDF increased in the conjunctiva, sclera, and RPE/choroid complex, but again there was no evidence that the secreted rPEDF diffused into the neural retina or vitreous (Fig. 7B). 

Figure 7.

 

Western blot analysis of rPEDF at days 3 (A) and 7 (B) after subconjunctival injection of stably PEDF-transfected RPE^{[N-PEDF-H-IV]} cells in a rabbit model. Note the presence of rPEDF in the conjunctiva (Co), sclera (Sc) and RPE/choroid complex (R/C), but not in the neural retina (Re) and the vitreous (Vi). rPEDF detected using anti-Penta-His antibodies was of the expected molecular weight (∼48 kDa). Protein loading was assessed by the use of anti-GAPDH antibodies (∼36 kDa). Note that in the retina, rPEDF was not present, even though protein loading was greater than for the other tissues.

Discussion

The proper balance between angiogenic and anti-angiogenic activities in tissues maintains the microvascular environment necessary to deliver nutrients, signaling mediators, and to remove waste. In the eye, VEGF is the main angiogenic factor responsible for the maintenance of a normal vasculature, while PEDF is the main anti-angiogenic agent that prevents pathologic neovascularization and maintains the proper neurogenic environment for visual function. When the balance between angiogenic and anti-angiogenic activities is altered in favor of angiogenic activities, neovascularization develops. In the eye, neovascularization is responsible for a number of blinding diseases, such as exudative AMD, diabetic retinopathy, retinopathy of prematurity, and neovascular glaucoma. ²⁷ The demonstration by Ohno-Matsui et al. ²⁸ and Ohno-Matsui ²⁹ that overexpression of VEGF in the retina is sufficient to elicit retinal and choroidal neovascularization has led to the development of anti-angiogenic therapies for exudative AMD. 

Even though treatment with anti-VEGF is successful, treatment requires sustained regimens and frequent intravitreal injections, with all its accompanying risks, including the possibility of systemic penetration resulting in long-term thromboembolic events. ^{30 –32}  

A more promising avenue of anti-angiogenic therapy would be the delivery of the natural anti-angiogenic and neurogenic factor PEDF. Even though it is possible to administer PEDF locally at the affected site by intravitreal injection, by slow release systems or mini-pumps, these systems are fraught with complications and deliver only for a short duration. The optimal means of delivering biologic factors locally would be the implantation of compatible cells that would constitutively express the desired factor. For retinal degeneration, the transplantation of pigment epithelial cells that would continuously secrete PEDF would not only inhibit the neovascularization associated with AMD but would also maintain a healthy neurogenic microenvironment. A number of investigators have attempted to transfer the PEDF gene using viral vectors, ^{33 –35} microbubbles, ³⁶ and nanoparticles ³⁷ either directly to the retina or by transplanting cells that had been transfected with the PEDF gene using viral vectors, and have shown functional rehabilitation in a number of animal models. Even though very efficient, the problems associated with gene delivery to ocular tissues using viral vectors is limited by possible undesirable side effects including toxicity, immunogenicity, and the limited size of inserted DNA. ^38,39  

To avoid the limitations of viral gene transfer, we have shown that it is feasible to efficiently transfect pigment epithelial cells with the human PEDF gene using electroporation. ²⁰ However, transfection does not ensure that the transfected cells efficiently and continuously produce the appropriate gene product. Here we show that transfected RPE cells secrete a PEDF or an EGFP–PEDF fusion protein of the expected molecular weight (∼48 and ∼76 kDa, respectively), even after 2 years and 100 passages in the presence or absence of Zeocin selection pressure, indicating the stable incorporation of the gene into the cell's genome. The transfected RPE cells produce and secrete rPEDF into the culture media at concentrations from 8 to 32 ng per hour per 100,000 cells, whereas nontransfected RPE cells secrete only 0.13 ng per hour per 100,000 cells. Transfection with PEDF did not alter the expression of genes critical to RPE cell function—specifically, CRALBP, CATD, and ZO-1—or the secretion of VEGF. Both nontransfected and stably transfected ARPE-19 cells produced VEGF at low levels that were not altered significantly after PEDF transfection. 

The goal of transfecting pigment epithelial cells with the gene for human PEDF is to transplant PEDF-transfected cells in the subretinal space as a treatment modality for retinal degenerative diseases. Because PEDF is a multifunctional molecule that has been shown to have neuroprotective ^40,41 and anti-angiogenic ^42,43 effects, protect against AGE-mediated effects, ^44,45 and suppress metalloproteinase-9, ⁴⁶ it is critical to verify that the PEDF-transfected cells are able to secrete a functional PEDF protein. To define the biologic activity of the rPEDF secreted into the media by the transfected ARPE-19 cells, we have examined their ability to autoregulate their proliferation and zinc transporter gene expression and the anti-angiogenic activity of rPEDF using an in vitro HUVEC sprouting assay. Here we show that ARPE-19 cells stably transfected with PEDF are capable of self-regulating cell proliferation and zinc transporter gene expression. Cell proliferation rates were increased dramatically beginning at day 5 of culture in transfected cells and reached confluence by day 8. However, nontransfected cells did not reach confluence by day 14, and in fact the cell number of nontransfected cells at day 14 was 30% lower than the cell number of transfected cells at day 8. On reaching confluence, the transfected cells became contact inhibited, a characteristic that has important implications regarding the application of these cells for regenerative transplantation. The higher proliferation and viability may accelerate and enhance engraftment of the transfected cells after implantation whereas the potential malignant transformation by the enhanced PEDF expression is precluded by the efficient contact inhibition. Nevertheless, mutagenic effects of random genomic integration are still a concern and must be considered. 

We have previously shown that exogenous rPEDF up-regulates the gene expression of the zinc transporters ZnT2 and ZIP2 in cultivated nontransfected ARPE-19 cells. ²⁰ Here we show that stably PEDF-transfected ARPE-19 cells are capable to autoregulate the expression of the zinc transporters ZnT3 and ZIP2, suggesting that these cells are able to regulate endogenous properties to become functionally more efficient. In addition, anti-angiogenic activity of the rPEDF was shown by the significant sprouting inhibition by HUVEC spheroids in vitro. 

The studies reported here show that pigment epithelial cells can be transfected with a linearized plasmid encoding the human PEDF gene with very high efficiency, that the PEDF gene is integrated within the cell's genome, as shown by the continuous secretion of rPEDF for 2 years and for over 100 passages, that transfection with PEDF does not affect the expression of genes essential to RPE cell function, and that in vivo, subconjunctivally implanted transfected cells survive and secrete rPEDF. We have also shown that the secreted rPEDF is functionally active and that the transfected cells acquire the functional attributes conferred by the higher levels of PEDF expression—namely, increased proliferation and autoregulation of zinc transporter gene expression. 

The studies detailed here are a crucial step in the development of cell-based, nonviral gene therapy for retinal degenerative diseases with the ultimate objective of isolating autologous cells from a patient, transfecting the cells with the appropriate transgene, and implanting the transfected cells during a single surgical session. 

Footnotes

 Supported by a grant from the Interdisciplinary Center for Clinical Research Aachen within the faculty of medicine at the Rheinisch-Westfälische Technische Hochschule Aachen University.

Footnotes

 Disclosures: S. Johnen, None; O. Kazanskaya, None; N. Armogan, None; C. Stickelmann, None; M. Stöcker, None; P. Walter, None; G. Thumann, None.

The authors thank Anna Dobias (Department of Ophthalmology, RWTH Aachen University, Aachen, Germany) for excellent technical assistance.