Vascular Endothelial Growth Factor-C (VEGF-C/VEGF-2) Promotes Angiogenesis in the Setting of Tissue Ischemia

Recombinant Protein

Heterodimeric recombinant human (rh) VEGF-A (VEGF-1) protein, purified from Escherichia coli, was a generous gift of Drs. N. Ferrara, B. Keyt, and S. Bunting at Genentech Inc. (South San Francisco, CA). E. coli-derived rhVEGF-A appeared as a single protein band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis with a molecular weight of 39.8 kd under nonreducing conditions. rhVEGF-C (VEGF-2) was expressed in the BaculoGold Virus Expression Vector System (Pharmingen) and purified by ion exchange and hydrophobic interaction chromatography as described before. 41 Under nonreducing conditions, sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis yielded bands of 52 kd and 90 kd, representing multimeric, disulfide-linked rhVEGF-C complexes. 20 For all recombinant proteins, endotoxin levels were nondetectable using the limulus lysate assay.

Plasmids

Complementary DNA clones for rhVEGF-C, isolated from a human osteoclast cDNA library, were assembled into the pcDNAI/Amp mammalian expression vector that includes a cytomegalovirus promoter/enhancer to drive gene expression. This plasmid construct was named pcVEGF-C. The plasmid pGSVLacZ (courtesy of Dr. C. Bonnerot), containing a nuclear targeted β-galactosidase sequence coupled to the SV40 early promoter, was used for control transfection experiments. 42

Cell Culture

Human umbilical vein ECs (HUVECs) were isolated from umbilical cord vein by collagenase treatment as previously described 43 and grown in medium 199 (Life Technologies, Inc., Grand Island, NY) supplemented with 20% fetal bovine serum (FBS) (Life Technologies), 100 μg/ml EC growth supplement, and 50 U/ml heparin (Sigma Chemical Co., St. Louis, MO). HUVECs were used between passages 1 and 5.

Human microvascular (HM) ECs of dermal origin were purchased from Cascade Biologics, Inc. Immunostaining of representative plates with the antibody to PAL-E 44 was used to confirm the absence of lymphatic ECs. The cells were grown in EBM medium containing human epidermal growth factor (10 ng/ml), hydrocortisone (10 ng/ml), 5% FBS, and 0.4% bovine brain extract. HMECs were used between passages 4 and 6.

EC Proliferation Assay

Mitogenic activity was assayed using a previously validated 45 colorimetric MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) assay with the electron coupling reagent phenazine methosulfate (CellTiter 96 AQ; Promega, Madison, WI). Cells were seeded in a 96-well plate (5000 cells/well) in 0.1 ml serum-supplemented medium and allowed to attach overnight. VEGF (rhVEGF-A or rhVEGF-C) in 5% FBS with heparin (10 U/ml) was added to the wells for 48 hours. MTS/phenazine methosulfate mixture (20 μl; 1:0.05) was added per well and allowed to incubate for 1 hour at 37°C before absorbance was measured at 490 nm in an enzyme-linked immunosorbent assay plate reader. Background absorbance from control wells was subtracted, and seven wells were performed in parallel and averaged for each condition. Proliferation experiments were each repeated three times.

EC Migration Assay

EC migration assays were performed using a 48-well microchemotaxis chamber (Neuroprobe). 46 Polyvinylpyrrolidone-free polycarbonate filters with a pore size of 8 μm (Nuclepore) were coated with 0.1% gelatin for at least 6 hours at room temperature and dried under sterile air. Test substances were diluted to appropriate concentrations in medium 199 supplemented with 1% FBS, and 25 μl of the final dilution was placed in the lower chamber of a modified Boyden chamber. Subconfluent HUVEC or HMEC cultures were washed and trypsinized for the minimum time required to achieve cell detachment. After placing the filter between lower and upper chambers, 2.5 × 10⁵ cells suspended in 50 μl medium 199 containing 1% FBS were seeded in the upper compartment. The apparatus was then incubated for 5 hours at 37°C in a humidified chamber with 5% CO₂ to allow cell migration. After the incubation period, the filter was removed, and the upper side of the filter containing the nonmigrated cells was scraped with a rubber policeman. The filters were fixed with methanol and stained with Giemsa solution (Diff-Quick, Baxter Diagnostics, McGaw Park, IL). Migration was quantified by counting cells in three random high-power fields (×100) in each well. All groups were studied in quadruplicate.

NO Measurement

An NO-specific polarographic electrode connected to an NO meter (Iso-NO, World Precision Instruments, Sarasota, FL) 47 was used to measure NO released from a cultured HUVEC monolayer. Calibration of the NO electrode was performed according to the following equation:

The standard calibration curve was obtained by adding graded concentrations of KNO₂ (0, 5, 10, 25, 50, 100, 250, and 500 nmol/L) to the calibration solution containing KI and H₂SO₄. Specificity of the iso-NO electrode for NO was previously determined by measurement of NO from authentic NO gas. 48 The culture medium was removed, and HUVECs were washed twice with Dulbecco’s phosphate-buffered saline. Cells were then bathed in 5 ml of filtered Krebs-Henseleit solution in six-well plates, and the cell plates were placed on a slide warmer (Lab Line Instruments) to maintain the temperature at 37°C. The NO sensor probe was inserted vertically into the wells, with the tip of the electrode placed 2 mm under the surface of the solution before reagents were added. S-Nitroso-N-acetyl-penicillamine (Sigma Chemical Co.) was used as a positive control. The amount of released NO was expressed as pmol per 1 × 10⁶ ECs. All values reported represent the mean of four to six measurements performed in each group (number of cell culture wells).

Miles Assay

Hairless male albino guinea pigs (200 to 600 g; Charles River Breeding Laboratories, Wilmington, MA) were lightly anesthetized with ether (Fisher Scientific, Fair Lawn, NJ). Evans blue dye (Sigma Chemical Co.) was filtered through 0.2-μm micropore filters (Corning Glass, Corning, NY) and then infused as a 0.5- to 1.0-ml bolus of 0.5% solution in saline via the left femoral vein. Evans blue dye binds to circulating plasma proteins and extravasates in response to certain reagents, rendering hyperpermeable dermal sites blue. 49 After the animals regained consciousness, reagents were injected intradermally in a volume of 0.1 ml. Intradermal injections were made into the trunk, posterior to the shoulder, 20 minutes after intravenous injection of Evans blue dye with a 30-gauge needle (Becton-Dickinson) causing a bleb of 9 to 11 mm in diameter. Increase in vascular permeability was assessed by leakage of blue dye into the bleb. 49 As originally described by Miles and Miles, 49 a small area of traumatic bluing 1 to 3 mm in diameter may be seen at the center of the bleb after intradermal injection of saline control. Intensity and area of the blue color changes within the blebs were identified and assessed, and the site of intradermal injection was photographed. The following reagents were injected intradermally: 1) saline alone (as vehicle control), 2) increasing concentrations of rhVEGF-A (4, 8, 16, 32, 64, and 128 ng), and 3) increasing concentrations of rhVEGF-C (4, 8, 16, 32, 64, and 128 ng).

To investigate the role of NO on VEGF-induced vascular permeability, N_ω-nitro-l-arginine methyl ester (l-NAME; 20 mg/kg; Sigma Chemical Co.) was injected via the femoral vein immediately before administration of Evans blue dye. Twenty minutes later, the following were injected intradermally: 1) saline alone (negative control), 2) increasing concentrations of rhVEGF-A (8, 16, 32, 64, and 128 ng), and 3) increasing concentrations of rhVEGF-C (8, 16, 32, 64, and 128 ng).

Analysis of VEGF Receptor mRNA in Explanted Arterial and Venous Specimens

Total RNA from unused segments of human internal mammary artery and saphenous vein was isolated by phenol/chloroform extraction, 50 and RNA concentration was calculated from absorbance at 260 nm.

Reverse transcription 51 was performed by heating a 10-μl reaction mixture containing 1 μg total RNA and 20 μg/ml oligodeoxy-thymidine (Life Technologies) at 70°C for 10 minutes. After cooling, 60 units of human placenta ribonuclease inhibitor (Promega) and 200 U Moloney’s murine leukemia virus RNase H reverse transcriptase (Life Technologies) were added in a final 20-μl reaction mixture containing (mmol/L) deoxynucleotide triphosphate (Pharmacia) 1 each, dithiothreitol 10, Tris-HCl (pH 8.3) 25, KCl 75, and MgCl₂ 3; incubated for 1 hour at 42°C; heated 5 minutes at 95°C; and diluted to 50 μl with double-distilled water. For polymerase chain reaction (PCR) with a thermostable DNA polymerase, 52 a 50-μl reaction containing 5 μl of cDNA-RNA hybrids, 200 μmol/L deoxynucleotide triphosphate (Pharmacia), 20 mmol/L Tris-HCl (pH 8.55), 2.5 mmol/L MgCl₂, 16 mmol/L (NH₄)₂SO₄, 150 μg/ml bovine serum albumin, 1 μmol/L each oligonucleotide primer, and 1.25 U of Taq polymerase (Perkin-Elmer Corp., Norwalk, CT) was subjected to the following temperature cycles: 30 seconds of denaturing at 95°C, 5 seconds of annealing at 56°C, and 1 minute of extension at 72°C. At the end of the last cycle, a prolonged extension step was carried out for 10 minutes. PCR products were analyzed by electrophoresing 10 μl of each PCR reaction mixture in a 1.5% agarose gel, and bands were visualized by ethidium bromide staining. The results represent three independent amplifications from two separate studies.

The primers chosen for human VEGFR-3 (GenBank accession number X68203) were: for sense (5′ to 3′), CAG GAT GAA GAC ATT TGA, and for antisense (5′ to 3′), AAG AAA ATG CTG ACG TAT GC 53 (190 bp PCR-product); for human VEGFR-2 (GenBank accession number X61656), CAA CAA AGT CGG GAG AGG AG and ATG ACG ATG GAC AAG TAG CC (819 bp); for human VEGFR-1 (GenBank accession number X51602), CAG CGG CTT TTG TGG AAG ACT CAC and ACA TCT CGG TGT CAC TTC TTG GAC (735 bp); and for human glyceraldehyde-3-phosphate dehydrogenase, TGA AGG TCG GAG TCA ACG GAT TTG and CAT GTG GGC CAT GAG GTC CAC CAC (983 bp). The number of cycles was 40 for VEGFR-1, VEGFR-2, and VEGFR-3, and 30 for glyceraldehyde-3-phosphate dehydrogenase.

Animal Model

The potential for VEGF-C to promote angiogenesis in vivo was investigated in a previously described rabbit model of hindlimb ischemia. 54,55 All protocols were approved by the Institutional Animal Care and Use Committee of St. Elizabeth’s Medical Center. A total of 28 (seven per group) male New Zealand White rabbits (3.5 to 4.0 kg) (Pine Acre Rabbitry) were anesthetized with a mixture of ketamine (50 mg/kg) and acepromazine (0.8 mg/kg) after premedication with xylazine (2 mg/kg). The femoral artery of one hindlimb was completely excised from its proximal origin as a branch of the external iliac artery to the point distally where it bifurcates into the saphenous and popliteal arteries. Consequently, blood flow to the ischemic limb is dependent on collateral vessels originating from the internal iliac artery. 54,55 An interval of 10 days was allowed for postoperative recovery and development of endogenous collateral vessels. Beyond this time point, studies performed up to 90 days postoperatively 55 have demonstrated no significant collateral vessel augmentation.

Percutaneous Arterial Gene Transfer of pcVEGF-C

At 10 days postoperatively (day 0), after performing a baseline angiogram (see below), the internal iliac artery of the ischemic limb of eight animals was transfected with pcVEGF-C as naked DNA using a 2.0-mm hydrogel-coated balloon catheter (Slider with Hydroplus, Boston Scientific, Natick, MA). The angioplasty balloon was prepared ex vivo by first advancing the deflated balloon through a 5-F Teflon sheath (Boston Scientific), applying 500 μg of pcVEGF-C solution to the 20 μm-thick layer of hydrogel coating the external surface of the inflated balloon, and then retracting the inflated balloon back into the protective sheath. The sheath and angioplasty catheter were then introduced via the right carotid artery and advanced to the lower abdominal aorta using a 0.014-inch guide wire (Hi-Torque Floppy II, Advanced Cardiovascular Systems) under fluoroscopic guidance. The balloon catheter was then advanced into the internal iliac artery of the ischemic limb, inflated for 1 minute at 6 atmospheres, deflated, and withdrawn. An identical protocol was used to transfect the internal iliac artery of control animals (n = 8) with the promoter-matched plasmid pGSVLacZ containing a nuclear targeted β-galactosidase sequence.

On day 30, all measurements were repeated (see below), the animals were sacrificed, and tissue samples were retrieved for histological examination.

Intra-arterial Administration of rhVEGF-C Protein

Ten days postoperatively (day 0), baseline noninvasive and invasive hemodynamic parameters (see below) were recorded. A single intra-arterial bolus of rhVEGF-C (500 μg; n = 8) in 3 ml of phosphate-buffered saline or vehicle solution (3 ml of phosphate-buffered saline with 0.1% rabbit serum albumin (RSA), n = 8) was then administered over a period of 1 minute through a 3-F end-hole infusion catheter (Tracker-18; Target Therapeutics) positioned in the internal iliac artery of the ischemic limb. The catheter was then washed with 3 ml of phosphate-buffered saline containing 0.1% albumin. The dose of rhVEGF-C protein used in the present experiment (500 μg) has previously been shown for rhVEGF-1 to induce a maximal effect using a single bolus protocol. 12,56

Physiological Assessment

Anatomical Assessment

Statistical Analysis

Results were expressed as mean ± standard error of the mean (SEM). Statistical significance was evaluated using unpaired Student’s t-test for comparisons between two means and analysis of variance followed by Scheffé’s procedure for more than two means. A value of P < 0.05 was interpreted to denote statistical significance.

Mitogenic Effect of rhVEGF-C on ECs

The effects of rhVEGF-C on EC proliferation are shown in Figure 1 ▶ . At a concentration of 10 ng/ml, the mitogenic activity of VEGF-C protein on HUVECs was 35% of that seen with VEGF-A. This difference was statistically significant (P < 0.005) (Figure 1A) ▶ . The mitogenic effect of VEGF-C was more pronounced when applied to HMECs (Figure 1B) ▶ : at 10 ng/ml, the mitogenic response of VEGF-C was 72% as potent as VEGF-A (P < 0.01).

Effect of rhVEGF-C on EC Migration

Both VEGF-A and VEGF-C protein exhibited a pronounced chemotactic effect on ECs. The dose-response curves showed a typical Gaussian distribution, beginning with an effect on migration that was evident at a minimal concentration of 0.1 ng/ml (Figure 2) ▶ . For HUVECs, at a concentration of 10 ng/ml, the chemotactic response of VEGF-C was 83% of that seen with VEGF-A (P < 0.05) (Figure 2A) ▶ ; for HMECs, the chemotactic response to VEGF-A and VEGF-C was indistinguishable (Figure 2B) ▶ .

NO Release from Cultured HUVECs by VEGF-C

Treatment of HUVECs with VEGF-C (1, 10, and 100 ng/ml) resulted in a concentration-dependent increase in NO synthesis (Figure 3) ▶ . Peak NO synthesis was observed between 3 and 6 minutes after addition of VEGF-C. Treatment of HUVECs with VEGF-A (100 ng/ml) under the same experimental conditions induced NO synthesis as previously reported. 61 S-Nitroso-N-acetyl-penicillamine was used as a positive control, and 0.1 mmol/L resulted in a significant increase in NO concentration measured in the buffer medium. For comparison, peak NO measured in response to VEGF-C was 92.7% and 80.9% of that measured after the same dose of VEGF-A and 0.1 mmol/L S-nitroso-N-acetyl-penicillamine, respectively.

Effect of VEGF-C on Vascular Permeability

Repeated intradermal injection of the vehicle control saline (0.1 ml) did not increase vascular permeability. Intradermal injection of either VEGF-A or VEGF-C (16, 32, 64, or 128 ng) significantly increased vascular permeability in a dose-dependent manner (Figure 4A) ▶ . Development of a positive (blue) response began 155 ± 6 seconds (n = 3) after intradermal injection. Repeated experiments (n = 3) did not show significant differences in the induction of vascular permeability between VEGF-A and VEGF-C. As a positive control, we used bradykinine and histamine (100 nmol/L), both of which increased vascular permeability at the injected sites (data not shown).

To examine the role of NO in VEGF-C-mediated vascular permeability, the NO synthase inhibitor l-NAME (20 mg/kg) was injected systemically 20 minutes before the intradermal application of VEGF-A and VEGF-C. l-NAME markedly attenuated vascular permeability induced by both VEGF-A and VEGF-C (Figure 4B) ▶ . The inactive stereoisomer d-NAME (20 mg/kg), which does not inhibit endothelial NO synthesis, failed to inhibit the increase in vascular permeability observed with either VEGF-1 or VEGF-C (not shown). Thus, for both VEGF-A and VEGF-C, increased vascular permeability was dependent on local production of NO.

VEGF-C Receptor Expression

To identify potential target vessels for VEGF-C activity in adults, we studied VEGFR-3 and VEGFR-2 mRNA expression in explanted segments of human internal mammary artery and saphenous vein by reverse transcription-PCR. For these experiments we used human specimens, because the corresponding nucleotide sequences for rabbit species have not yet been published. Figure 5 ▶ demonstrates that the VEGF-C receptors VEGFR-2 and VEGFR-3, as well as VEGFR-1, are expressed in both venous and arterial tissues. As a positive control, we used HUVECs, for which coexpression of VEGFR-1, VEGFR-2, and VEGFR-3 has been previously described. 53 RNA from human fibroblasts, from which amplification never resulted in positive bands for these receptors, was used as a negative control.

Effect of pcVEGF-C and rhVEGF-C on Blood Pressure and Flow

Calf blood pressure ratio was similar in all groups at day 0 (Figure 6A) ▶ . By day 30, blood pressure ratio in the pcVEGF-C and VEGF-C protein-treated groups had increased to 0.83 ± 0.03 and 0.76 ± 0.04, respectively. In both cases, this value exceeded that measured for the control groups (0.59 ± 0.04 and 0.58 ± 0.03; P = 0.002 and 0.009, respectively). Blood pressure ratio in rabbits undergoing gene transfer with phVEGF-C was not significantly different from that recorded in rabbits treated with VEGF-C recombinant protein.

Blood flow was measured in the internal iliac artery of the ischemic limb at days 0 and 30. At day 0, there were no differences among groups in resting or maximum blood flow (not shown). For pcVEGF-C plasmid, 30-day maximum iliac blood flow was clearly increased compared with pGSVLacZ controls (53.7 ± 5.9 versus 31.6 ± 1.2 ml/min; P = 0.0009). At day 30, maximum flow after nitroprusside infusion (39.1 ± 6.7 ml/min) was significantly higher in the VEGF-C protein-treated group compared with RSA control (26.1 ± 5.4 ml/min; P = 0.03). Comparison between the plasmid and protein treatment groups disclosed no statistically significant difference (P = 0.17).

Iliac flow reserve, the ratio between blood flow at rest and after maximal pharmacological stimulation, was 2.90 ± 0.31 and 2.65 ± 0.14 for the pcVEGF-C- and rhVEGF-C-treated rabbits, respectively. In both cases, this represented a statistically significant increase compared with controls (pGSVLacZ = 2.10 ± 0.17 and RSA = 2.04 ± 0.23; P = 0.04 and P = 0.04, respectively) (Figure 6C) ▶ .

Effect of pcVEGF-C and rhVEGF-C on Anatomical Evidence of Collateral Development

Quantitative analysis of collateral vessel development in the medial thigh is summarized in Figure 6B ▶ . Before treatment (day 0), there were no significant differences among groups in angiographic score. By day 30, the angiographic score in both VEGF-C gene transfer (0.85 ± 0.05) and protein (0.74 ± 0.08) groups exceeded that measured in the control groups (pGSVLacZ = 0.51 ± 0.02 (P = 0.00003) and RSA = 0.53 ± 0.03 (P = 0.023), respectively). The angiographic score in the pcVEGF-C gene transfer group was significantly higher than that measured for the protein-treated group (P = 0.02). Figure 7 ▶ shows representative internal iliac angiograms recorded at day 30 from controls, VEGF-C plasmid, and VEGF-C recombinant protein-treated animals. In both treatment groups, collateral artery development was more marked compared with corresponding control groups.

Tissue sections from the medial thigh muscles of the ischemic and nonischemic limbs were examined by light microscopy at day 30. Histological evaluation after alkaline phosphatase staining revealed that the capillary density in the pcVEGF-C-transfected group (252 ± 12/mm²) and that in the VEGF-C protein-treated group (229 ± 20/mm²) were both significantly greater than values observed for the corresponding control groups (pGSVLacZ = 183 ± 10/mm² (P = 0.001), and RSA = 164 ± 20/mm² (P = 0.043), respectively) (Figure 6D) ▶ . Analysis of capillary/muscle fiber ratio yielded similar results (data not shown). Analysis of capillary density and capillary/muscle fiber ratio in the medial thigh muscle of the nonischemic limb showed no differences among groups (data not shown). Representative examples of histological sections stained for alkaline phosphatase in the different experimental groups are shown in Figure 8 ▶ .