In Vivo Expression of Recombinant Vascular Endothelial Growth Factor in Rabbit Carotid Artery Increases Production of Superoxide Anion
http://www.100md.com
动脉硬化血栓血管生物学 2005年第3期
From the Departments of Anesthesiology, Molecular Pharmacology and Experimental Therapeutics, Mayo Clinic College of Medicine, Rochester, Minn.
Correspondence to Zvonimir S. Katusic, MD, PhD, Departments of Anesthesiology, Molecular Pharmacology, and Experimental Therapeutics, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905. E-mail katusic.zvonimir@mayo.edu
Abstract
Objective— Vascular endothelial growth factor (VEGF) is one of the most important pro-angiogenic cytokines. Ability of VEGF to stimulate formation of superoxide anion in vivo has not been studied. We hypothesized that in vivo expression of recombinant VEGF in the rabbit carotid artery increases production of superoxide anion.
Methods and Results— Plaque-forming units (109) of adenovirus-encoding human VEGF165 (AdVEGF) or ?-galactosidase (AdLacZ) were delivered into the lumen of rabbit carotid arteries. Three days after gene delivery, expression of recombinant proteins was detected in endothelium and smooth muscle cells. Endothelium-dependent relaxations to acetylcholine were impaired in AdVEGF-transduced arteries (P<0.01; n=5). Treatment with superoxide dismutase mimetic, Mn(III) tetra(4-benzoic acid) porphyrin chloride (10–5 mol/L), improved relaxations to acetylcholine (P<0.01; n=5). Western blot analysis demonstrated increased expression of p47phox in AdVEGF-transduced arteries (P<0.05; n=8). Lucigenin chemiluminescence showed significantly higher production of superoxide anion in AdVEGF-transduced arteries (P<0.05; n=5 to 10).
Conclusions— Our results suggest that in vivo expression of recombinant VEGF in the vascular endothelium increases local production of superoxide anion. Superoxide anion appears to be an important mediator of vascular effects of VEGF in vivo.
In vivo adenovirus-mediated expression of recombinant vascular endothelial growth factor (VEGF) in endothelial cells of the rabbit carotid artery increases production of superoxide anion. This finding suggests that superoxide anion is an important mediator of vascular effects of VEGF. Therapeutic effect of VEGF may depend on redox state of the vascular wall.
Key Words: endothelial function ? endothelium ? nitric oxide ? nitric oxide synthase ? reactive oxygen species
Introduction
Vascular endothelial growth factor (VEGF) stimulates endothelial cell migration, differentiation, and regeneration,1,2 which are essential for its pro-angiogenic activity.1,3 VEGF has been used extensively to stimulate therapeutic angiogenesis.4 Previous in vivo studies have suggested that VEGF not only augments the formation of collateral vessels but also modifies their vasomotor responses.5 Bauters et al demonstrated that administration of VEGF resulted in improved endothelium-dependent responses of large collateral vessels.6 Despite the potential beneficial therapeutic effect of VEGF, experimental studies have raised concerns about the safety of VEGF-mediated angiogenesis. Recent reports indicated that transgenic7 or adenoviral8 overexpression of VEGF resulted in formation of leaky vessels in laboratory animals, and that plasmid-based intramuscular VEGF gene transfer led to transient edema in human subjects.9,10 Celletti et al showed that recombinant human VEGF was capable of mobilizing macrophages/monocytes while simultaneously increasing atherosclerotic plaque formation and progression.11,12 It has been known that vascular and cardiac tissues are rich sources of reactive oxygen species (ROS), including superoxide anion and nitric oxide (NO). Virtually every cell type in the vascular wall has been shown to produce and to be regulated by ROS.13,14 ROS play an important role in vascular homeostasis by affecting signal transduction, proliferation, apoptosis, aging, gene expression, and the biologically effective concentration of NO.15–18
A number of studies have demonstrated that the inhibitory effect of superoxide anion on endothelium-dependent relaxation is caused by chemical antagonism between superoxide anion and NO.19–24 Recently, 2 reports have suggested that VEGF stimulates production of superoxide anion in cultured endothelial cells by an NAD(P)H oxidase-dependent mechanism.25,26 However, there is no in vivo evidence that VEGF165 stimulates production of superoxide anion. In this study, we provide evidence that VEGF165 increases production of superoxide anion in vivo, and that this effect may be primarily caused by activation of NAD(P)H oxidase.
Methods
Recombinant Adenoviral Vectors
We used a replication-deficient recombinant adenoviral vector encoding the human VEGF165 gene (AdVEGF) with E1a, partial E1b, and partial E3 deletion and expression cassette in the E1 position containing the cytomegalovirus immediate early promoter/enhancer.27 This vector was provided by the vector core facility, Molecular Medicine Institute, at University of Pittsburgh (Pittsburgh, Pa). A replication-deficient recombinant adenoviral vector encoding the Escherichia coli ?-galactosidase gene (AdLacZ) driven by the cytomegalovirus promoter was used as a control. This vector was provided by the Gene Transfer Vector Core at University of Iowa.
In Vivo Carotid Artery Gene Transfer
Male New Zealand White rabbits weighing 2.0 to 3.0 kg were used in these experiments. Housing facilities and all procedures were in accordance with the Institutional Animal Care and Use Committee guidelines of Mayo Clinic. The method of transduction of the vessel segment was the same as previously described.28 Briefly, anesthesia was induced by an intramuscular injection of ketamine (35 mg/kg), xylazine (5 mg/kg), and acepromazine (2.3 mg/kg). The common carotid arteries were exposed, and heparin sodium (300 U/kg) was administered intravenously. Proximal and distal vascular clamps were then applied. After insertion of a 24-gauge vascular catheter into the proximal part of the isolated segment, an adenoviral (AdLacZ or AdVEGF) load of 1x109 plaque-forming units in 100 μL of sterile phosphate-buffered saline (PBS) was instilled intraluminally. After 20 minutes, vascular clamps were removed, flow was restored, and the rabbit was allowed to recover. After 3 days, the rabbit was euthanized, and carotid arteries were harvested. We used untreated arteries or those treated with the sterile PBS alone as a control.
Analyses of Vasomotor Function
Rings (4-mm-long) from each carotid artery were connected to a force transducer for recording of isometric force and placed in organ baths filled with 25 mL of the modified Krebs solution (118.3 mmol/L NaCl, 25 mmol/L NaHCO3, 4.7 mmol/L KCl, 2.5 mmol/L CaCl2, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 0.026 mmol/L EDTA, and 11.1 mmol/L glucose; pH 7.4) at 37°C (94% O2, 6% CO2). Isometric force was recorded continuously, and the rings were allowed to stabilize at 0.2 grams for 1 hour. Each ring was then gradually stretched to 3.0 grams. The rings were incubated with indomethacin (10–5 mol/L) for 30 minutes before exposure to other drugs. The contraction responses to phenylephrine (10–9 to 10–5 mol/L) were examined before exposure to acetylcholine or diethylammonium(Z)-1-(N,N-diethylamino)diazen-1-IM1,2-diolate (DEA-NONOate). These contractions were expressed as a percentage of the maximal contraction induced by phenylephrine. The relaxation responses to acetylcholine (10–9 to 10–5 mol/L) or DEA-NONOate (10–9 to 10–5 mol/L) were studied during submaximal contractions to phenylephrine. These relaxations were determined as a percentage of relaxation to a high concentration of papaverine (3x10–4 mol/L). In a separate protocol, a novel cell-permeable superoxide dismutase (SOD) mimetic, Mn(III) tetra(4-benzoic acid) porphyrin chloride (MnTBAP) (10–5 mol/L), was incubated for 15 minutes before the contraction or relaxation responses.
Detection of ?-Galactosidase Expression
Segments of rabbit carotid arteries transduced with AdLacZ or AdVEGF were washed in PBS and fixed for 30 minutes at room temperature in 2% formaldehyde and 0.2% glutaraldehyde in PBS (pH 7.4). One milliliter of a solution of 500 μg/mL 5-bromo-4-chloro-3-indolyl-?-D-galactoside (X-gal) was added to the rings, and these were incubated for 1 hour at 37°C. After a rinsing with PBS, vessel segments were cut into 3-mm rings and embedded in paraffin. Five-μm-thick cross-sections spaced at least 100 μm apart were cut from each segment and counterstained with nuclear fast red (ICN Biomedicals Inc). The sections were examined under a light microscope. Efficiency of gene transfer to the endothelium was determined by counting and calculating percentage of stained cells.
Immunohistochemical Analysis
Staining was performed on deparaffinized sections. Endogenous peroxidases were blocked with 20-minute incubation in 0.3% H2O2 in methanol. Nonspecific binding was blocked by incubation of tissue with diluted goat serum for 20 minutes. Sections were then incubated with monoclonal anti-VEGF antibody (1:100; Santa Cruz Biotechnology) for 3 hours at room temperature. Secondary antibody from Vectastain Elite ABC Kit (Vector Laboratories) was applied to sections for 1 hour at room temperature. For visualization, sections were incubated with DAB substrate (Vector Laboratories) for 10 seconds and counterstained with hematoxylin. Negative controls were performed by omitting the primary antibody.
Western Blot Analyses
After collection and removal of connective tissue, carotid arteries were homogenized on ice in lysis buffer (pH 7.5) containing 50 mmol/L Tris-HCl, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 1% IGEPAL, 0.1% SDS, 0.1% deoxycholate, and a 100-fold dilution of a mammalian protease inhibitor cocktail (Sigma Chemical Co). Equal amounts of protein (100 μg) were separated by SDS-PAGE and transferred to nitrocellulose membrane (Amersham Biosciences). Monoclonal anti-VEGF (1:200; Santa Cruz Biotechnology), anti-p47phox (1:500; BD Transduction Laboratories), anti-phosphorylated endothelial nitric oxide synthase (phosphorylated eNOS) (1:500; BD Transduction Laboratories), anti-inducible nitric oxide synthase (iNOS) (1:500; BD Transduction Laboratories), and polyclonal anti-manganese superoxide dismutase (MnSOD) (1:500; Stressgen Biotechnologies) antibodies were used. Monoclonal anti-actin (1:50 000; Sigma-Aldrich Chemical Co) antibody was used as loading control. Bands were visualized by enhanced chemiluminescence using an ECL system kit (Amersham Biosciences).
Densitometry was performed using National Institutes of Health Image (Scion-Image; Scion Corporation), and the results were expressed in relative densitometry as compared with actin.
Oxidative Fluorescence Microscopy
The oxidative fluorescent dye dihydroethidium was used to detect superoxide anion.29 Unfixed frozen segments of carotid arteries were cut into 15-μm-thick sections and placed on a glass slide. Samples were incubated with dihydroethidium (1x10–6 mol/L) in a light-protected humidified chamber for 30 minutes at 37°C. Tissue sections were imaged of an Olympus Fluoview laser scanning confocal microscope.
Quantification of Vascular Superoxide Anion Production
Production of superoxide anion was measured by lucigenin-enhanced chemiluminescence as previously described.30 Briefly, rings from each carotid artery were opened length-wise and equilibrated for 30 minutes at 37°C (5% CO2 incubator) in the modified Krebs-HEPES buffer (135.3 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaCl2, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 0.026 mmol/L EDTA, 10.1 mmol/L glucose, and 20 mmol/L HEPES; pH 7.4). Scintillation vials containing 2 mL modified Krebs–HEPES buffer with 5 μmol/L lucigenin were placed into a scintillation counter (LS 5000; Beckman Instruments Inc) switched to the out-of-coincidence mode. Background signals were recorded, and vascular segments were then added to each vial. In some experiments, vessels were preincubated with a NAD(P)H oxidase inhibitor, apocynin31 (100 μmol/L), or a phosphatidylinositol-3 kinase inhibitor, wortmannin32 (30 μmol/L), for 30 minutes at 37°C (5% CO2 incubator) before addition of lucigenin. The results were expressed as counts per minute per milligram dry weight.
Measurement of Cyclic Guanosine 3', 5'-Monophosphate
From each harvested carotid artery, rings were immersed immediately in a solution of 3-isobutyl-L-methylxanthine (IBMX; 1 mmol/L) and incubated for 30 minutes at 37°C (5% CO2 incubator). The tissues were frozen in liquid nitrogen and stored at –80°C until the time of assay. After homogenization of arteries, a radioimmunoassay kit (Amersham Biosciences) was used to perform the measurement as previously described.33 Tissues were analyzed in parallel, and the results were expressed as pmol/mg protein.
Measurements of Biopterin and GTPCH I Activity
Tetrahydrobiopterin (BH4) and oxidized biopterin (BH2) were determined after differential oxidation in acid and base conditions by reverse-phase high-performance liquid chromatography as previously described.28 Activity of GTP cyclohydrolase I (GTPCH I), the rate-limiting enzyme in BH4 synthesis, was assessed as a function of neopterin production under standard conditions with GTP as a substrate.28 All results in carotid arteries were normalized against tissue protein levels.
Histology
Common carotid arteries were fixed in buffered formalin (4%) and kept for 48 hours at room temperature. Tissues were embedded in paraffin and processed for light microscopy using standard hematoxylin and eosin staining.
Drugs
The following pharmacological agents were used: indomethacin, phenylephrine, acetylcholine, papaverine, apocynin, wortmannin, and IBMX (all from Sigma Chemical Co), DEA-NONOate (Cayman Chemical Co), MnTBAP (BIOMOL Laboratories), Xgal (Invitrogen Corporation), and dihydroethidium and lucigenin (Molecular Probes Inc). Protein levels throughout the study were determined with the use of a DC protein assay kit (Bio-Rad Laboratories).
Statistical Analyses
Results are expressed as mean±SEM, for n rabbits used in each experimental protocol. The concentration-response curves were analyzed by 2-way ANOVA followed by Bonferroni method for multiple comparisons. Half-maximal effective concentrations (EC50) were calculated by nonlinear regression and expressed as –log M. The superoxide anion production was analyzed by 1-way ANOVA followed by Dunnett method for multiple comparisons. The optical intensity values, cyclic guanosine 3', 5'-monophosphate (cGMP), biopterin levels, and GTPCH I activity were analyzed by an unpaired Student t test for simple comparisons between 2 groups. A value of P<0.05 was considered statistically significant.
Results
Confirmation of Gene Transfer
Arteries transduced with AdLacZ showed transgene expression in the endothelium as confirmed with X-gal staining (Figure 1A). In contrast, there was no X-gal staining in arteries transduced with AdVEGF (Figure 1A). The efficiency of gene transfer was quantified at 31±7% (n=4) of endothelial cells.
Figure 1. A, Histological localization of ?-galactosidase reporter gene expression in rabbit carotid arteries after exposure to AdLacZ (adenovirus vector encoding the Escherichia coli ?-galactosidase gene) or AdVEGF (n=4). Blue (X-gal) staining demonstrates endothelium-specific gene transfer of ?-galactosidase. Original magnification, x40. B, Immunohistochemical localization of VEGF in carotid arteries with AdLacZ or AdVEGF. Positive staining was seen in the endothelium and media in AdVEGF arteries, but not in AdLacZ arteries (n=3). Original magnification, x40. C, Representative Western blot analysis demonstrated high expression of human VEGF165 protein in AdVEGF arteries compared with AdLacZ arteries (n=3).
Detection and Localization of VEGF Expression
Immunohistochemical staining of sections of AdVEGF arteries showed the expression of VEGF in the endothelium and media (Figure 1B). In contrast, there was not detectable expression of VEGF in AdLacZ arteries (Figure 1B). High expression of recombinant human VEGF165 protein was confirmed in carotid arteries after 72 hours of transduction with AdVEGF by Western blot analysis (Figure 1C). There was not detectable expression of VEGF in untreated arteries or those treated with PBS alone (n=3; data not shown).
Histology and Vascular Structure
No morphological or inflammatory changes were observed in rabbit carotid arteries transduced with AdLacZ or AdVEGF, 3 days after gene delivery (n=3; data not shown).
Vasomotor Function
In AdVEGF arteries, contractions to phenylephrine were significantly reduced as compared with AdLacZ arteries (Figure 2A). Endothelium-dependent relaxations to acetylcholine were significantly impaired in AdVEGF arteries (Figure 2B), whereas endothelium-independent relaxations to DEA-NONOate did not differ between AdLacZ and AdVEGF arteries (P=0.98; n=5). Treatment with MnTBAP significantly improved relaxations to acetylcholine in AdVEGF arteries (Figure 2C). In contrast, contractions to phenylephrine and relaxations to DEA-NONOate were not affect by MnTBAP in AdLacZ and AdVEGF arteries (n=5; data not shown). In addition, there was no difference of vasomotor function among untreated arteries, those treated with PBS alone and AdLacZ arteries (EC50 to acetylcholine=7.2±0.2 –log M, 6.9±0.1 –log M, and 6.9±0.1 –log M; maximum relaxation=97±3%, 96±4%, and 90±4%, respectively; n=5).
Figure 2. Concentration response curves to phenylephrine (A), acetylcholine (B) in AdLacZ or AdVEGF arteries, and acetylcholine with or without MnTBAP (C) in AdVEGF arteries. A, Contractions to phenylephedrine in AdVEGF arteries were significantly reduced as compared with AdLacZ arteries (*P<0.05). B, Relaxations to acetylcholine in AdVEGF arteries were significantly impaired as compared with AdLacZ arteries (*P<0.01). C, Treatment with MnTBAP significantly improved relaxations to acetylcholine in AdVEGF arteries (*P<0.01). All results are shown as mean±SEM (n=5) and were analyzed by ANOVA, followed by Bonferroni method.
Vascular Superoxide Anion Production
After loading with the oxidation-sensitive dye dihydroethidium, a marked increase in ethidium bromide fluorescence was found throughout the vascular wall of AdVEGF arteries, which reflected an increase in superoxide anion compared with AdLacZ arteries (Figure 3A) or those treated with PBS alone (data not shown). The increase in ethidium bromide fluorescence was observed mainly in endothelial cells, and to a lesser degree in vascular smooth muscle cells (Figure 3A). Lucigenin-enhanced chemiluminescence showed significantly higher superoxide anion levels in AdVEGF arteries compared with AdLacZ arteries (Figure 3B). Increased production of superoxide anion was significantly reduced in the presence of 100 μmol/L apocynin (Figure 3B). In contrast, 30 μmol/L wortmannin had no effect (Figure 3B).
Figure 3. A, Fluorescent photomicrographs showing in situ detection of superoxide anion with use of confocal microscopic sections of AdLacZ or AdVEGF arteries. High ethidium bromide fluorescence was found in the vascular wall of AdVEGF arteries compared with AdLacZ arteries and was localized mainly in endothelial cells, and to a lesser degree in vascular smooth muscle cells. Original magnification, x10 (upper panels); x63 (lower panels). B, Bar graphs showing superoxide anion levels in AdLacZ arteries, AdVEGF arteries, and AdVEGF arteries treated with apocynin or wortmannin. (*P<0.05 versus AdLacZ arteries or AdVEGF arteries with apocynin). Superoxide anion levels in AdVEGF arteries were not affected by wortmannin. All results are shown as mean±SEM (n=5 to 10) and were analyzed by ANOVA, followed by Dunnett method.
Expressions of Protein by Western Blot Analysis
Expressions of p47phox (Figure 4) and phosphorylated eNOS (Figure 5A) protein were increased in AdVEGF arteries compared with AdLacZ arteries. In contrast, expressions of iNOS and MnSOD protein were not different between AdLacZ and AdVEGF arteries (n=4, data not shown).
Figure 4. Representative Western blot analysis demonstrated high expression of p47phox protein expression in AdVEGF arteries compared with AdLacZ arteries. Bar graphs indicate the results of the relative densitometry compared with actin. Results are shown as mean±SEM (n=8) and were analyzed by an unpaired Student t test (*P<0.05).
Figure 5. A, Representative Western blot analysis demonstrated high expression of phosphorylated eNOS (phospho-eNOS) protein in AdVEGF arteries compared with AdLacZ arteries (n=5). B, Bar graphs showing basal levels of cGMP in AdLacZ or AdVEGF arteries. Results are shown as mean±SEM (n=5) and were analyzed by an unpaired Student t test (*P<0.05).
cGMP Levels
Production of cGMP was significantly increased in AdVEGF arteries as compared with AdLacZ arteries (Figure 5B).
Biopterin Levels and GTPCH I Activity
BH4 (Figure 6A), BH2 (Figure 6B) levels, and the BH4 as a percentage of total biopterin (Figure 6C) did not differ between AdLacZ and AdVEGF arteries. In addition, GTPCH I activity was not significantly affected by AdVEGF delivery to the carotid arteries (n=7; data not shown).
Figure 6. BH4 (A), BH2 (B) levels, and the BH4 as a percentage of total biopterin (C) were not different between AdLacZ and AdVEGF arteries. All results are shown as mean±SEM (n=7) and were analyzed by an unpaired Student t test (P=0.78, 0.46, and 0.47, respectively).
Discussion
This is the first in vivo study to demonstrate stimulatory effect of recombinant VEGF on generation of superoxide anion in arterial wall. Pharmacological analysis of vasomotor function in arteries expressing recombinant VEGF demonstrates that impairment of endothelium-dependent relaxations to acetylcholine is mediated by increased formation of superoxide anion. Consistent with this observation, we provide evidence that recombinant VEGF upregulates expression of p47phox protein, essential subunit of NAD(P)H oxidase, and that pharmacological inhibition of NAD(P)H oxidase normalizes production of superoxide anion. These findings strongly suggest that NAD(P)H oxidase may be the major source of superoxide anion in arteries exposed to high local concentration of VEGF. Our findings are also the first to our knowledge to demonstrate that recombinant VEGF does not affect vascular metabolism of tetrahydrobiopterin, an essential co-factor required for enzymatic activity of eNOS, suggesting that uncoupled eNOS is an unlikely source of superoxide anion in the VEGF-transduced rabbit carotid artery.
Consistent with our previous studies,28,34 adenovirus-mediated gene delivery into endothelial cells of carotid artery resulted in significant increase of recombinant VEGF expression in arterial wall. Increase in VEGF expression was associated with high expression of phosphorylated eNOS protein, as well as high levels of cGMP. Previous studies demonstrated that in vascular endothelium VEGF stimulates production of NO by phosphorylating eNOS via activation of phosphatidylinositol-3-OH-kinase (PI3K) and protein kinase B (Akt).35,36 Although we did not directly measure production of NO, high levels of cGMP most likely reflect increased production of NO in VEGF transduced arteries. We did not detect expression of iNOS, further suggesting that high local production of NO in carotid artery was caused by stimulatory effect of VEGF on eNOS activity. This effect of VEGF may explain reduced vasoconstriction in response to phenylephrine. In our previous studies, we demonstrated that elevation of NO production and its second messenger cGMP had inhibitory effect on vasomotor function, including vasoconstrictor effects to endothelin-1 and uridine 5'-triphosphate.37,38 Overexpression of VEGF reduced endothelium-dependent relaxations to acetylcholine. The inhibitory effect of VEGF was prevented by the presence of superoxide anion scavenger, MnTBAP, strongly suggesting that high production of superoxide anion is responsible for impairment of endothelial function. MnTBAP normalized endothelium-dependent relaxations in VEGF-transduced arteries but did not affect relaxations to acetylcholine in LacZ-transduced blood vessels, demonstrating selectivity of superoxide anion scavenger.
To characterize the source of superoxide anion in arterial wall, we examined expression of p47phox protein, an important component of NAD(P)H oxidase.39 In vivo transduction of carotid artery with recombinant VEGF caused significant increase of p47phox protein expression, suggesting that increased activity of NAD(P)H oxidase is the most likely source of superoxide anion. Our conclusion was reinforced by the fact that an inhibitor of NAD(P)H oxidase, apocynin, normalized production of superoxide anion. These findings are in agreement with reported ability of VEGF to increase formation of ROS and activity of NAD(P)H oxidase in cultured endothelial cells.25 Interestingly, Abid et al proposed that NAD(P)H oxidase mediated increase in concentration of ROS is an important signaling mechanism required for upregulation of MnSOD expression. In vivo, we could not detect higher expression of MnSOD in rabbit carotid artery transduced with recombinant VEGF. The reasons for discrepancy between in vitro and in vivo findings are unclear, but could be caused by inherent difference in superoxide anion production of cultured endothelium versus endothelium of intact carotid artery. Expression of VEGF did not affect metabolism of BH4. This conclusion was supported by the fact that VEGF did not affect BH4 levels or GTPCH I enzymatic activity. Most importantly, despite detected increase in superoxide anion formation in VEGF-transduced arteries, ratio between reduced (BH4) and oxidized (BH2) biopterin remained normal. This is an important observation demonstrating that level of oxidative stress induced by expression of recombinant VEGF does not cross the threshold needed for oxidation of BH4 reported in cardiovascular diseases associated with high production of ROS and subsequent uncoupling of eNOS.40 Consistent with this conclusion, wortmannin, a selective inhibitor of PI3K, did not affect production of superoxide anion, ruling out the possibility that uncoupled eNOS activated by VEGF (via PI3K/Akt pathway) contributes to production of superoxide anion. Whether recombinant VEGF expression in diseased arteries chronically exposed to high oxidative stress may contribute to vascular dysfunction by stimulating superoxide anion production is unknown. However, our observations regarding ability of VEGF to activate NAD(P)H oxidase and increase production of superoxide anion may help to explain controversial findings reported in patients treated with VEGF.9,10,41 It is conceivable that in patients with cardiovascular disease, differences in redox background may have decisive impact on therapeutic effect of VEGF.
The results of the present study demonstrate that in the rabbit carotid artery, adenovirus-mediated delivery of recombinant VEGF stimulates production of superoxide anion most likely caused by upregulation of NAD(P)H oxidase expression and enzymatic activity. Our findings suggest that redox state of the vascular wall may be an important determinant of the beneficial effect of VEGF in treatment of cardiovascular disease.
Acknowledgments
This work was supported in part by National Heart, Lung, and Blood Institute grants HL-53524, HL-58080, and HL-66958, the American Heart Association Bugher Foundation Award for the Investigation of Stroke, and by the Mayo Foundation. The authors thank Janet Beckman for editorial assistance.
Received August 3, 2004; accepted November 22, 2004.
References
Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med. 1995; 1: 27–31.
Ferrara N. Role of vascular endothelial growth factor in the regulation of angiogenesis. Kidney Int. 1999; 56: 794–814.
Risau W. Mechanisms of angiogenesis. Nature. 1997; 386: 671–674.
Isner JM, Losordo DW. Therapeutic angiogenesis for heart failure. Nat Med. 1999; 5: 491–492.
Asahara T, Chen D, Tsurumi Y, Kearney M, Rossow S, Passeri J, Symes JF, Isner JM. Accelerated restitution of endothelial integrity and endotheliumdependent function after phVEGF165 gene transfer. Circulation. 1996; 94: 3291–3302.
Bauters C, Asahara T, Zheng LP, Takeshita S, Bunting S, Ferrara N, Symes JF, Isner JM. Recovery of disturbed endothelium-dependent flow in the collateralperfused rabbit ischemic hindlimb after administration of vascular endothelial growth factor. Circulation. 1995; 91: 2802–2809.
Thurston G, Suri C, Smith K, McClain J, Sato TN, Yancopoulos GD, McDonald DM. Leakage-resistant blood vessels in mice transgenically overexpressing angiopoietin-1. Science. 1999; 286: 2511–2514.
Thurston G, Rudge JS, Ioffe E, Zhou H, Ross L, Croll SD, Glazer N, Holash J, McDonald DM, Yancopoulos GD. Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat Med. 2000; 6: 460–463.
Baumgartner I, Pieczek A, Manor O, Blair R, Kearney M, Walsh K, Isner JM. Constitutive expression of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation. 1998; 97: 1114–1123.
Isner JM, Baumgartner I, Rauh G, Schainfeld R, Blair R, Manor O, Razvi S, Symes JF. Treatment of thromboangiitis obliterans (Buerger’s disease) by intramuscular gene transfer of vascular endothelial growth factor: preliminary clinical results. J Vasc Surg. 1998; 28: 964–975.
Celletti FL, Waugh JM, Amabile PG, Brendolan A, Hilfiker PR, Dake MD. Vascular endothelial growth factor enhances atherosclerotic plaque progression. Nat Med. 2001; 7: 425–429.
Celletti FL, Hilfiker PR, Ghafouri P, Dake MD. Effect of human recombinant vascular endothelial growth factor165 on progression of atherosclerotic plaque. J Am Coll Cardiol. 2001; 37: 2126–2130.
Griendling KK, Ushio-Fukai M. NADH/NADPH oxidase and vascular function. Trends Cardiovasc Med. 1997; 7: 301–307.
Suzuki YJ, Ford GD. Redox regulation of signal transduction in cardiac and smooth muscle. J Mol Cell Cardiol. 1999; 31: 345–353.
Rao GN, Berk BC. Active oxygen species stimulate vascular smooth muscle cell growth and proto-oncogene expression. Circ Res. 1992; 70: 593–599.
Katusic ZS. Superoxide anion and endothelial regulation of arterial tone. Free Radic Biol Med. 1996; 20: 443–448.
Lander HM. An essential role for free radicals and derived species in signal transduction. Faseb J. 1997; 11: 118–124.
Abe J, Berk BC. Reactive oxygen species as mediators of signal transduction in cardiovascular disease. Trends Cardiovasc Med. 1998; 8: 59–64.
Hattori Y, Kawasaki H, Abe K, Kanno M. Superoxide dismutase recovers altered endothelium-dependent relaxation in diabetic rat aorta. Am J Physiol. 1991; 261: H1086–1094.
Nakazono K, Watanabe N, Matsuno K, Sasaki J, Sato T, Inoue M. Does superoxide underlie the pathogenesis of hypertension? Proc Natl Acad Sci U S A. 1991; 88: 10045–10048.
White CR, Brock TA, Chang LY, Crapo J, Briscoe P, Ku D, Bradley WA, Gianturco SH, Gore J, Freeman BA, et al. Superoxide and peroxynitrite in atherosclerosis. Proc Natl Acad Sci U S A. 1994; 91: 1044–1048.
Keaney JF, Jr., Xu A, Cunningham D, Jackson T, Frei B, Vita JA. Dietary probucol preserves endothelial function in cholesterol-fed rabbits by limiting vascular oxidative stress and superoxide generation. J Clin Invest. 1995; 95: 2520–2529.
Rajagopalan S, Kurz S, Munzel T, Tarpey M, Freeman BA, Griendling KK, Harrison DG. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation. Contribution to alterations of vasomotor tone. J Clin Invest. 1996; 97: 1916–1923.
Laursen JB, Rajagopalan S, Galis Z, Tarpey M, Freeman BA, Harrison DG. Role of superoxide in angiotensin II-induced but not catecholamine-induced hypertension. Circulation. 1997; 95: 588–593.
Abid MR, Tsai JC, Spokes KC, Deshpande SS, Irani K, Aird WC. Vascular endothelial growth factor induces manganese-superoxide dismutase expression in endothelial cells by a Rac1-regulated NADPH oxidase-dependent mechanism. FASEB J. 2001; 15: 2548–2550.
Ushio-Fukai M, Tang Y, Fukai T, Dikalov SI, Ma Y, Fujimoto M, Quinn MT, Pagano PJ, Johnson C, Alexander RW. Novel role of gp91(phox)-containing NAD(P)H oxidase in vascular endothelial growth factor-induced signaling and angiogenesis. Circ Res. 2002; 91: 1160–1167.
Muhlhauser J, Merrill MJ, Pili R, Maeda H, Bacic M, Bewig B, Passaniti A, Edwards NA, Crystal RG, Capogrossi MC. VEGF165 expressed by a replication-deficient recombinant adenovirus vector induces angiogenesis in vivo. Circ Res. 1995; 77: 1077–1086.
Hynes SO, Smith LA, Richardson DM, Kovesdi I, O’Brien T, Katusic ZS. In vivo expression and function of recombinant GTPCH I in the rabbit carotid artery. Am J Physiol Heart Circ Physiol. 2004; 286: H570–H574.
d’Uscio LV, Baker TA, Mantilla CB, Smith L, Weiler D, Sieck GC, Katusic ZS. Mechanism of endothelial dysfunction in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol. 2001; 21: 1017–1022.
Zanetti M, d’Uscio LV, Kovesdi I, Katusic ZS, O’Brien T. In vivo gene transfer of inducible nitric oxide synthase to carotid arteries from hypercholesterolemic rabbits. Stroke. 2003; 34: 1293–1298.
Li L, Fink GD, Watts SW, Northcott CA, Galligan JJ, Pagano PJ, Chen AF. Endothelin-1 increases vascular superoxide via endothelin(A)-NADPH oxidase pathway in low-renin hypertension. Circulation. 2003; 107: 1053–1058.
Zubkov AY, Rollins KS, Parent AD, Zhang J, Bryan RM, Jr. Mechanism of endothelin-1-induced contraction in rabbit basilar artery. Stroke. 2000; 31: 526–533.
Matsumoto T, D’Uscio LV, Eguchi D, Akiyama M, Smith LA, Katusic ZS. Protective effect of chronic vitamin C treatment on endothelial function of apolipoprotein E-deficient mouse carotid artery. J Pharmacol Exp Ther. 2003; 306: 103–108.
Sato J, Mohacsi T, Noel A, Jost C, Gloviczki P, Mozes G, Katusic ZS, O’Brien T, Mayhan WG. In vivo gene transfer of endothelial nitric oxide synthase to carotid arteries from hypercholesterolemic rabbits enhances endotheliumdependent relaxations. Stroke. 2000; 31: 968–975.
Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K, Franke TF, Papapetropoulos A, Sessa WC. Regulation of endothelium-derived nitric oxide production by the protein kinase Akt. Nature. 1999; 399: 597–601.
Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM. Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature. 1999; 399: 601–605.
Akiyama M, Eguchi D, Weiler D, O’Brien T, Kovesdi I, Scotland RS, Sessa WC, Katusic ZS. Expression and function of recombinant S1179D endothelial nitric oxide synthase in canine cerebral arteries. Stroke. 2002; 33: 1071–1076.
Eguchi D, D’Uscio LV, Wambi C, Weiler D, Kovesdi I, O’Brien T, Katusic ZS. Inhibitory effect of recombinant iNOS gene expression on vasomotor function of canine basilar artery. Am J Physiol Heart Circ Physiol. 2002; 283: H2560–H2566.
Jones SA, O’Donnell VB, Wood JD, Broughton JP, Hughes EJ, Jones OT. Expression of phagocyte NADPH oxidase components in human endothelial cells. Am J Physiol. 1996; 271: H1626–H1634.
Katusic ZS. Vascular endothelial dysfunction: does tetrahydrobiopterin play a role? Am J Physiol Heart Circ Physiol. 2001; 281: H981–H986.
Rajagopalan S, Mohler ER, 3rd, Lederman RJ, Mendelsohn FO, Saucedo JF, Goldman CK, Blebea J, Macko J, Kessler PD, Rasmussen HS, Annex BH. Regional angiogenesis with vascular endothelial growth factor in peripheral arterial disease: a phase II randomized, double-blind, controlled study of adenoviral delivery of vascular endothelial growth factor 121 in patients with disabling intermittent claudication. Circulation. 2003; 108: 1933–1938.(Osamu Suda; Leslie A. Smi)
Correspondence to Zvonimir S. Katusic, MD, PhD, Departments of Anesthesiology, Molecular Pharmacology, and Experimental Therapeutics, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905. E-mail katusic.zvonimir@mayo.edu
Abstract
Objective— Vascular endothelial growth factor (VEGF) is one of the most important pro-angiogenic cytokines. Ability of VEGF to stimulate formation of superoxide anion in vivo has not been studied. We hypothesized that in vivo expression of recombinant VEGF in the rabbit carotid artery increases production of superoxide anion.
Methods and Results— Plaque-forming units (109) of adenovirus-encoding human VEGF165 (AdVEGF) or ?-galactosidase (AdLacZ) were delivered into the lumen of rabbit carotid arteries. Three days after gene delivery, expression of recombinant proteins was detected in endothelium and smooth muscle cells. Endothelium-dependent relaxations to acetylcholine were impaired in AdVEGF-transduced arteries (P<0.01; n=5). Treatment with superoxide dismutase mimetic, Mn(III) tetra(4-benzoic acid) porphyrin chloride (10–5 mol/L), improved relaxations to acetylcholine (P<0.01; n=5). Western blot analysis demonstrated increased expression of p47phox in AdVEGF-transduced arteries (P<0.05; n=8). Lucigenin chemiluminescence showed significantly higher production of superoxide anion in AdVEGF-transduced arteries (P<0.05; n=5 to 10).
Conclusions— Our results suggest that in vivo expression of recombinant VEGF in the vascular endothelium increases local production of superoxide anion. Superoxide anion appears to be an important mediator of vascular effects of VEGF in vivo.
In vivo adenovirus-mediated expression of recombinant vascular endothelial growth factor (VEGF) in endothelial cells of the rabbit carotid artery increases production of superoxide anion. This finding suggests that superoxide anion is an important mediator of vascular effects of VEGF. Therapeutic effect of VEGF may depend on redox state of the vascular wall.
Key Words: endothelial function ? endothelium ? nitric oxide ? nitric oxide synthase ? reactive oxygen species
Introduction
Vascular endothelial growth factor (VEGF) stimulates endothelial cell migration, differentiation, and regeneration,1,2 which are essential for its pro-angiogenic activity.1,3 VEGF has been used extensively to stimulate therapeutic angiogenesis.4 Previous in vivo studies have suggested that VEGF not only augments the formation of collateral vessels but also modifies their vasomotor responses.5 Bauters et al demonstrated that administration of VEGF resulted in improved endothelium-dependent responses of large collateral vessels.6 Despite the potential beneficial therapeutic effect of VEGF, experimental studies have raised concerns about the safety of VEGF-mediated angiogenesis. Recent reports indicated that transgenic7 or adenoviral8 overexpression of VEGF resulted in formation of leaky vessels in laboratory animals, and that plasmid-based intramuscular VEGF gene transfer led to transient edema in human subjects.9,10 Celletti et al showed that recombinant human VEGF was capable of mobilizing macrophages/monocytes while simultaneously increasing atherosclerotic plaque formation and progression.11,12 It has been known that vascular and cardiac tissues are rich sources of reactive oxygen species (ROS), including superoxide anion and nitric oxide (NO). Virtually every cell type in the vascular wall has been shown to produce and to be regulated by ROS.13,14 ROS play an important role in vascular homeostasis by affecting signal transduction, proliferation, apoptosis, aging, gene expression, and the biologically effective concentration of NO.15–18
A number of studies have demonstrated that the inhibitory effect of superoxide anion on endothelium-dependent relaxation is caused by chemical antagonism between superoxide anion and NO.19–24 Recently, 2 reports have suggested that VEGF stimulates production of superoxide anion in cultured endothelial cells by an NAD(P)H oxidase-dependent mechanism.25,26 However, there is no in vivo evidence that VEGF165 stimulates production of superoxide anion. In this study, we provide evidence that VEGF165 increases production of superoxide anion in vivo, and that this effect may be primarily caused by activation of NAD(P)H oxidase.
Methods
Recombinant Adenoviral Vectors
We used a replication-deficient recombinant adenoviral vector encoding the human VEGF165 gene (AdVEGF) with E1a, partial E1b, and partial E3 deletion and expression cassette in the E1 position containing the cytomegalovirus immediate early promoter/enhancer.27 This vector was provided by the vector core facility, Molecular Medicine Institute, at University of Pittsburgh (Pittsburgh, Pa). A replication-deficient recombinant adenoviral vector encoding the Escherichia coli ?-galactosidase gene (AdLacZ) driven by the cytomegalovirus promoter was used as a control. This vector was provided by the Gene Transfer Vector Core at University of Iowa.
In Vivo Carotid Artery Gene Transfer
Male New Zealand White rabbits weighing 2.0 to 3.0 kg were used in these experiments. Housing facilities and all procedures were in accordance with the Institutional Animal Care and Use Committee guidelines of Mayo Clinic. The method of transduction of the vessel segment was the same as previously described.28 Briefly, anesthesia was induced by an intramuscular injection of ketamine (35 mg/kg), xylazine (5 mg/kg), and acepromazine (2.3 mg/kg). The common carotid arteries were exposed, and heparin sodium (300 U/kg) was administered intravenously. Proximal and distal vascular clamps were then applied. After insertion of a 24-gauge vascular catheter into the proximal part of the isolated segment, an adenoviral (AdLacZ or AdVEGF) load of 1x109 plaque-forming units in 100 μL of sterile phosphate-buffered saline (PBS) was instilled intraluminally. After 20 minutes, vascular clamps were removed, flow was restored, and the rabbit was allowed to recover. After 3 days, the rabbit was euthanized, and carotid arteries were harvested. We used untreated arteries or those treated with the sterile PBS alone as a control.
Analyses of Vasomotor Function
Rings (4-mm-long) from each carotid artery were connected to a force transducer for recording of isometric force and placed in organ baths filled with 25 mL of the modified Krebs solution (118.3 mmol/L NaCl, 25 mmol/L NaHCO3, 4.7 mmol/L KCl, 2.5 mmol/L CaCl2, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 0.026 mmol/L EDTA, and 11.1 mmol/L glucose; pH 7.4) at 37°C (94% O2, 6% CO2). Isometric force was recorded continuously, and the rings were allowed to stabilize at 0.2 grams for 1 hour. Each ring was then gradually stretched to 3.0 grams. The rings were incubated with indomethacin (10–5 mol/L) for 30 minutes before exposure to other drugs. The contraction responses to phenylephrine (10–9 to 10–5 mol/L) were examined before exposure to acetylcholine or diethylammonium(Z)-1-(N,N-diethylamino)diazen-1-IM1,2-diolate (DEA-NONOate). These contractions were expressed as a percentage of the maximal contraction induced by phenylephrine. The relaxation responses to acetylcholine (10–9 to 10–5 mol/L) or DEA-NONOate (10–9 to 10–5 mol/L) were studied during submaximal contractions to phenylephrine. These relaxations were determined as a percentage of relaxation to a high concentration of papaverine (3x10–4 mol/L). In a separate protocol, a novel cell-permeable superoxide dismutase (SOD) mimetic, Mn(III) tetra(4-benzoic acid) porphyrin chloride (MnTBAP) (10–5 mol/L), was incubated for 15 minutes before the contraction or relaxation responses.
Detection of ?-Galactosidase Expression
Segments of rabbit carotid arteries transduced with AdLacZ or AdVEGF were washed in PBS and fixed for 30 minutes at room temperature in 2% formaldehyde and 0.2% glutaraldehyde in PBS (pH 7.4). One milliliter of a solution of 500 μg/mL 5-bromo-4-chloro-3-indolyl-?-D-galactoside (X-gal) was added to the rings, and these were incubated for 1 hour at 37°C. After a rinsing with PBS, vessel segments were cut into 3-mm rings and embedded in paraffin. Five-μm-thick cross-sections spaced at least 100 μm apart were cut from each segment and counterstained with nuclear fast red (ICN Biomedicals Inc). The sections were examined under a light microscope. Efficiency of gene transfer to the endothelium was determined by counting and calculating percentage of stained cells.
Immunohistochemical Analysis
Staining was performed on deparaffinized sections. Endogenous peroxidases were blocked with 20-minute incubation in 0.3% H2O2 in methanol. Nonspecific binding was blocked by incubation of tissue with diluted goat serum for 20 minutes. Sections were then incubated with monoclonal anti-VEGF antibody (1:100; Santa Cruz Biotechnology) for 3 hours at room temperature. Secondary antibody from Vectastain Elite ABC Kit (Vector Laboratories) was applied to sections for 1 hour at room temperature. For visualization, sections were incubated with DAB substrate (Vector Laboratories) for 10 seconds and counterstained with hematoxylin. Negative controls were performed by omitting the primary antibody.
Western Blot Analyses
After collection and removal of connective tissue, carotid arteries were homogenized on ice in lysis buffer (pH 7.5) containing 50 mmol/L Tris-HCl, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 1% IGEPAL, 0.1% SDS, 0.1% deoxycholate, and a 100-fold dilution of a mammalian protease inhibitor cocktail (Sigma Chemical Co). Equal amounts of protein (100 μg) were separated by SDS-PAGE and transferred to nitrocellulose membrane (Amersham Biosciences). Monoclonal anti-VEGF (1:200; Santa Cruz Biotechnology), anti-p47phox (1:500; BD Transduction Laboratories), anti-phosphorylated endothelial nitric oxide synthase (phosphorylated eNOS) (1:500; BD Transduction Laboratories), anti-inducible nitric oxide synthase (iNOS) (1:500; BD Transduction Laboratories), and polyclonal anti-manganese superoxide dismutase (MnSOD) (1:500; Stressgen Biotechnologies) antibodies were used. Monoclonal anti-actin (1:50 000; Sigma-Aldrich Chemical Co) antibody was used as loading control. Bands were visualized by enhanced chemiluminescence using an ECL system kit (Amersham Biosciences).
Densitometry was performed using National Institutes of Health Image (Scion-Image; Scion Corporation), and the results were expressed in relative densitometry as compared with actin.
Oxidative Fluorescence Microscopy
The oxidative fluorescent dye dihydroethidium was used to detect superoxide anion.29 Unfixed frozen segments of carotid arteries were cut into 15-μm-thick sections and placed on a glass slide. Samples were incubated with dihydroethidium (1x10–6 mol/L) in a light-protected humidified chamber for 30 minutes at 37°C. Tissue sections were imaged of an Olympus Fluoview laser scanning confocal microscope.
Quantification of Vascular Superoxide Anion Production
Production of superoxide anion was measured by lucigenin-enhanced chemiluminescence as previously described.30 Briefly, rings from each carotid artery were opened length-wise and equilibrated for 30 minutes at 37°C (5% CO2 incubator) in the modified Krebs-HEPES buffer (135.3 mmol/L NaCl, 4.7 mmol/L KCl, 2.5 mmol/L CaCl2, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 0.026 mmol/L EDTA, 10.1 mmol/L glucose, and 20 mmol/L HEPES; pH 7.4). Scintillation vials containing 2 mL modified Krebs–HEPES buffer with 5 μmol/L lucigenin were placed into a scintillation counter (LS 5000; Beckman Instruments Inc) switched to the out-of-coincidence mode. Background signals were recorded, and vascular segments were then added to each vial. In some experiments, vessels were preincubated with a NAD(P)H oxidase inhibitor, apocynin31 (100 μmol/L), or a phosphatidylinositol-3 kinase inhibitor, wortmannin32 (30 μmol/L), for 30 minutes at 37°C (5% CO2 incubator) before addition of lucigenin. The results were expressed as counts per minute per milligram dry weight.
Measurement of Cyclic Guanosine 3', 5'-Monophosphate
From each harvested carotid artery, rings were immersed immediately in a solution of 3-isobutyl-L-methylxanthine (IBMX; 1 mmol/L) and incubated for 30 minutes at 37°C (5% CO2 incubator). The tissues were frozen in liquid nitrogen and stored at –80°C until the time of assay. After homogenization of arteries, a radioimmunoassay kit (Amersham Biosciences) was used to perform the measurement as previously described.33 Tissues were analyzed in parallel, and the results were expressed as pmol/mg protein.
Measurements of Biopterin and GTPCH I Activity
Tetrahydrobiopterin (BH4) and oxidized biopterin (BH2) were determined after differential oxidation in acid and base conditions by reverse-phase high-performance liquid chromatography as previously described.28 Activity of GTP cyclohydrolase I (GTPCH I), the rate-limiting enzyme in BH4 synthesis, was assessed as a function of neopterin production under standard conditions with GTP as a substrate.28 All results in carotid arteries were normalized against tissue protein levels.
Histology
Common carotid arteries were fixed in buffered formalin (4%) and kept for 48 hours at room temperature. Tissues were embedded in paraffin and processed for light microscopy using standard hematoxylin and eosin staining.
Drugs
The following pharmacological agents were used: indomethacin, phenylephrine, acetylcholine, papaverine, apocynin, wortmannin, and IBMX (all from Sigma Chemical Co), DEA-NONOate (Cayman Chemical Co), MnTBAP (BIOMOL Laboratories), Xgal (Invitrogen Corporation), and dihydroethidium and lucigenin (Molecular Probes Inc). Protein levels throughout the study were determined with the use of a DC protein assay kit (Bio-Rad Laboratories).
Statistical Analyses
Results are expressed as mean±SEM, for n rabbits used in each experimental protocol. The concentration-response curves were analyzed by 2-way ANOVA followed by Bonferroni method for multiple comparisons. Half-maximal effective concentrations (EC50) were calculated by nonlinear regression and expressed as –log M. The superoxide anion production was analyzed by 1-way ANOVA followed by Dunnett method for multiple comparisons. The optical intensity values, cyclic guanosine 3', 5'-monophosphate (cGMP), biopterin levels, and GTPCH I activity were analyzed by an unpaired Student t test for simple comparisons between 2 groups. A value of P<0.05 was considered statistically significant.
Results
Confirmation of Gene Transfer
Arteries transduced with AdLacZ showed transgene expression in the endothelium as confirmed with X-gal staining (Figure 1A). In contrast, there was no X-gal staining in arteries transduced with AdVEGF (Figure 1A). The efficiency of gene transfer was quantified at 31±7% (n=4) of endothelial cells.
Figure 1. A, Histological localization of ?-galactosidase reporter gene expression in rabbit carotid arteries after exposure to AdLacZ (adenovirus vector encoding the Escherichia coli ?-galactosidase gene) or AdVEGF (n=4). Blue (X-gal) staining demonstrates endothelium-specific gene transfer of ?-galactosidase. Original magnification, x40. B, Immunohistochemical localization of VEGF in carotid arteries with AdLacZ or AdVEGF. Positive staining was seen in the endothelium and media in AdVEGF arteries, but not in AdLacZ arteries (n=3). Original magnification, x40. C, Representative Western blot analysis demonstrated high expression of human VEGF165 protein in AdVEGF arteries compared with AdLacZ arteries (n=3).
Detection and Localization of VEGF Expression
Immunohistochemical staining of sections of AdVEGF arteries showed the expression of VEGF in the endothelium and media (Figure 1B). In contrast, there was not detectable expression of VEGF in AdLacZ arteries (Figure 1B). High expression of recombinant human VEGF165 protein was confirmed in carotid arteries after 72 hours of transduction with AdVEGF by Western blot analysis (Figure 1C). There was not detectable expression of VEGF in untreated arteries or those treated with PBS alone (n=3; data not shown).
Histology and Vascular Structure
No morphological or inflammatory changes were observed in rabbit carotid arteries transduced with AdLacZ or AdVEGF, 3 days after gene delivery (n=3; data not shown).
Vasomotor Function
In AdVEGF arteries, contractions to phenylephrine were significantly reduced as compared with AdLacZ arteries (Figure 2A). Endothelium-dependent relaxations to acetylcholine were significantly impaired in AdVEGF arteries (Figure 2B), whereas endothelium-independent relaxations to DEA-NONOate did not differ between AdLacZ and AdVEGF arteries (P=0.98; n=5). Treatment with MnTBAP significantly improved relaxations to acetylcholine in AdVEGF arteries (Figure 2C). In contrast, contractions to phenylephrine and relaxations to DEA-NONOate were not affect by MnTBAP in AdLacZ and AdVEGF arteries (n=5; data not shown). In addition, there was no difference of vasomotor function among untreated arteries, those treated with PBS alone and AdLacZ arteries (EC50 to acetylcholine=7.2±0.2 –log M, 6.9±0.1 –log M, and 6.9±0.1 –log M; maximum relaxation=97±3%, 96±4%, and 90±4%, respectively; n=5).
Figure 2. Concentration response curves to phenylephrine (A), acetylcholine (B) in AdLacZ or AdVEGF arteries, and acetylcholine with or without MnTBAP (C) in AdVEGF arteries. A, Contractions to phenylephedrine in AdVEGF arteries were significantly reduced as compared with AdLacZ arteries (*P<0.05). B, Relaxations to acetylcholine in AdVEGF arteries were significantly impaired as compared with AdLacZ arteries (*P<0.01). C, Treatment with MnTBAP significantly improved relaxations to acetylcholine in AdVEGF arteries (*P<0.01). All results are shown as mean±SEM (n=5) and were analyzed by ANOVA, followed by Bonferroni method.
Vascular Superoxide Anion Production
After loading with the oxidation-sensitive dye dihydroethidium, a marked increase in ethidium bromide fluorescence was found throughout the vascular wall of AdVEGF arteries, which reflected an increase in superoxide anion compared with AdLacZ arteries (Figure 3A) or those treated with PBS alone (data not shown). The increase in ethidium bromide fluorescence was observed mainly in endothelial cells, and to a lesser degree in vascular smooth muscle cells (Figure 3A). Lucigenin-enhanced chemiluminescence showed significantly higher superoxide anion levels in AdVEGF arteries compared with AdLacZ arteries (Figure 3B). Increased production of superoxide anion was significantly reduced in the presence of 100 μmol/L apocynin (Figure 3B). In contrast, 30 μmol/L wortmannin had no effect (Figure 3B).
Figure 3. A, Fluorescent photomicrographs showing in situ detection of superoxide anion with use of confocal microscopic sections of AdLacZ or AdVEGF arteries. High ethidium bromide fluorescence was found in the vascular wall of AdVEGF arteries compared with AdLacZ arteries and was localized mainly in endothelial cells, and to a lesser degree in vascular smooth muscle cells. Original magnification, x10 (upper panels); x63 (lower panels). B, Bar graphs showing superoxide anion levels in AdLacZ arteries, AdVEGF arteries, and AdVEGF arteries treated with apocynin or wortmannin. (*P<0.05 versus AdLacZ arteries or AdVEGF arteries with apocynin). Superoxide anion levels in AdVEGF arteries were not affected by wortmannin. All results are shown as mean±SEM (n=5 to 10) and were analyzed by ANOVA, followed by Dunnett method.
Expressions of Protein by Western Blot Analysis
Expressions of p47phox (Figure 4) and phosphorylated eNOS (Figure 5A) protein were increased in AdVEGF arteries compared with AdLacZ arteries. In contrast, expressions of iNOS and MnSOD protein were not different between AdLacZ and AdVEGF arteries (n=4, data not shown).
Figure 4. Representative Western blot analysis demonstrated high expression of p47phox protein expression in AdVEGF arteries compared with AdLacZ arteries. Bar graphs indicate the results of the relative densitometry compared with actin. Results are shown as mean±SEM (n=8) and were analyzed by an unpaired Student t test (*P<0.05).
Figure 5. A, Representative Western blot analysis demonstrated high expression of phosphorylated eNOS (phospho-eNOS) protein in AdVEGF arteries compared with AdLacZ arteries (n=5). B, Bar graphs showing basal levels of cGMP in AdLacZ or AdVEGF arteries. Results are shown as mean±SEM (n=5) and were analyzed by an unpaired Student t test (*P<0.05).
cGMP Levels
Production of cGMP was significantly increased in AdVEGF arteries as compared with AdLacZ arteries (Figure 5B).
Biopterin Levels and GTPCH I Activity
BH4 (Figure 6A), BH2 (Figure 6B) levels, and the BH4 as a percentage of total biopterin (Figure 6C) did not differ between AdLacZ and AdVEGF arteries. In addition, GTPCH I activity was not significantly affected by AdVEGF delivery to the carotid arteries (n=7; data not shown).
Figure 6. BH4 (A), BH2 (B) levels, and the BH4 as a percentage of total biopterin (C) were not different between AdLacZ and AdVEGF arteries. All results are shown as mean±SEM (n=7) and were analyzed by an unpaired Student t test (P=0.78, 0.46, and 0.47, respectively).
Discussion
This is the first in vivo study to demonstrate stimulatory effect of recombinant VEGF on generation of superoxide anion in arterial wall. Pharmacological analysis of vasomotor function in arteries expressing recombinant VEGF demonstrates that impairment of endothelium-dependent relaxations to acetylcholine is mediated by increased formation of superoxide anion. Consistent with this observation, we provide evidence that recombinant VEGF upregulates expression of p47phox protein, essential subunit of NAD(P)H oxidase, and that pharmacological inhibition of NAD(P)H oxidase normalizes production of superoxide anion. These findings strongly suggest that NAD(P)H oxidase may be the major source of superoxide anion in arteries exposed to high local concentration of VEGF. Our findings are also the first to our knowledge to demonstrate that recombinant VEGF does not affect vascular metabolism of tetrahydrobiopterin, an essential co-factor required for enzymatic activity of eNOS, suggesting that uncoupled eNOS is an unlikely source of superoxide anion in the VEGF-transduced rabbit carotid artery.
Consistent with our previous studies,28,34 adenovirus-mediated gene delivery into endothelial cells of carotid artery resulted in significant increase of recombinant VEGF expression in arterial wall. Increase in VEGF expression was associated with high expression of phosphorylated eNOS protein, as well as high levels of cGMP. Previous studies demonstrated that in vascular endothelium VEGF stimulates production of NO by phosphorylating eNOS via activation of phosphatidylinositol-3-OH-kinase (PI3K) and protein kinase B (Akt).35,36 Although we did not directly measure production of NO, high levels of cGMP most likely reflect increased production of NO in VEGF transduced arteries. We did not detect expression of iNOS, further suggesting that high local production of NO in carotid artery was caused by stimulatory effect of VEGF on eNOS activity. This effect of VEGF may explain reduced vasoconstriction in response to phenylephrine. In our previous studies, we demonstrated that elevation of NO production and its second messenger cGMP had inhibitory effect on vasomotor function, including vasoconstrictor effects to endothelin-1 and uridine 5'-triphosphate.37,38 Overexpression of VEGF reduced endothelium-dependent relaxations to acetylcholine. The inhibitory effect of VEGF was prevented by the presence of superoxide anion scavenger, MnTBAP, strongly suggesting that high production of superoxide anion is responsible for impairment of endothelial function. MnTBAP normalized endothelium-dependent relaxations in VEGF-transduced arteries but did not affect relaxations to acetylcholine in LacZ-transduced blood vessels, demonstrating selectivity of superoxide anion scavenger.
To characterize the source of superoxide anion in arterial wall, we examined expression of p47phox protein, an important component of NAD(P)H oxidase.39 In vivo transduction of carotid artery with recombinant VEGF caused significant increase of p47phox protein expression, suggesting that increased activity of NAD(P)H oxidase is the most likely source of superoxide anion. Our conclusion was reinforced by the fact that an inhibitor of NAD(P)H oxidase, apocynin, normalized production of superoxide anion. These findings are in agreement with reported ability of VEGF to increase formation of ROS and activity of NAD(P)H oxidase in cultured endothelial cells.25 Interestingly, Abid et al proposed that NAD(P)H oxidase mediated increase in concentration of ROS is an important signaling mechanism required for upregulation of MnSOD expression. In vivo, we could not detect higher expression of MnSOD in rabbit carotid artery transduced with recombinant VEGF. The reasons for discrepancy between in vitro and in vivo findings are unclear, but could be caused by inherent difference in superoxide anion production of cultured endothelium versus endothelium of intact carotid artery. Expression of VEGF did not affect metabolism of BH4. This conclusion was supported by the fact that VEGF did not affect BH4 levels or GTPCH I enzymatic activity. Most importantly, despite detected increase in superoxide anion formation in VEGF-transduced arteries, ratio between reduced (BH4) and oxidized (BH2) biopterin remained normal. This is an important observation demonstrating that level of oxidative stress induced by expression of recombinant VEGF does not cross the threshold needed for oxidation of BH4 reported in cardiovascular diseases associated with high production of ROS and subsequent uncoupling of eNOS.40 Consistent with this conclusion, wortmannin, a selective inhibitor of PI3K, did not affect production of superoxide anion, ruling out the possibility that uncoupled eNOS activated by VEGF (via PI3K/Akt pathway) contributes to production of superoxide anion. Whether recombinant VEGF expression in diseased arteries chronically exposed to high oxidative stress may contribute to vascular dysfunction by stimulating superoxide anion production is unknown. However, our observations regarding ability of VEGF to activate NAD(P)H oxidase and increase production of superoxide anion may help to explain controversial findings reported in patients treated with VEGF.9,10,41 It is conceivable that in patients with cardiovascular disease, differences in redox background may have decisive impact on therapeutic effect of VEGF.
The results of the present study demonstrate that in the rabbit carotid artery, adenovirus-mediated delivery of recombinant VEGF stimulates production of superoxide anion most likely caused by upregulation of NAD(P)H oxidase expression and enzymatic activity. Our findings suggest that redox state of the vascular wall may be an important determinant of the beneficial effect of VEGF in treatment of cardiovascular disease.
Acknowledgments
This work was supported in part by National Heart, Lung, and Blood Institute grants HL-53524, HL-58080, and HL-66958, the American Heart Association Bugher Foundation Award for the Investigation of Stroke, and by the Mayo Foundation. The authors thank Janet Beckman for editorial assistance.
Received August 3, 2004; accepted November 22, 2004.
References
Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med. 1995; 1: 27–31.
Ferrara N. Role of vascular endothelial growth factor in the regulation of angiogenesis. Kidney Int. 1999; 56: 794–814.
Risau W. Mechanisms of angiogenesis. Nature. 1997; 386: 671–674.
Isner JM, Losordo DW. Therapeutic angiogenesis for heart failure. Nat Med. 1999; 5: 491–492.
Asahara T, Chen D, Tsurumi Y, Kearney M, Rossow S, Passeri J, Symes JF, Isner JM. Accelerated restitution of endothelial integrity and endotheliumdependent function after phVEGF165 gene transfer. Circulation. 1996; 94: 3291–3302.
Bauters C, Asahara T, Zheng LP, Takeshita S, Bunting S, Ferrara N, Symes JF, Isner JM. Recovery of disturbed endothelium-dependent flow in the collateralperfused rabbit ischemic hindlimb after administration of vascular endothelial growth factor. Circulation. 1995; 91: 2802–2809.
Thurston G, Suri C, Smith K, McClain J, Sato TN, Yancopoulos GD, McDonald DM. Leakage-resistant blood vessels in mice transgenically overexpressing angiopoietin-1. Science. 1999; 286: 2511–2514.
Thurston G, Rudge JS, Ioffe E, Zhou H, Ross L, Croll SD, Glazer N, Holash J, McDonald DM, Yancopoulos GD. Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat Med. 2000; 6: 460–463.
Baumgartner I, Pieczek A, Manor O, Blair R, Kearney M, Walsh K, Isner JM. Constitutive expression of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation. 1998; 97: 1114–1123.
Isner JM, Baumgartner I, Rauh G, Schainfeld R, Blair R, Manor O, Razvi S, Symes JF. Treatment of thromboangiitis obliterans (Buerger’s disease) by intramuscular gene transfer of vascular endothelial growth factor: preliminary clinical results. J Vasc Surg. 1998; 28: 964–975.
Celletti FL, Waugh JM, Amabile PG, Brendolan A, Hilfiker PR, Dake MD. Vascular endothelial growth factor enhances atherosclerotic plaque progression. Nat Med. 2001; 7: 425–429.
Celletti FL, Hilfiker PR, Ghafouri P, Dake MD. Effect of human recombinant vascular endothelial growth factor165 on progression of atherosclerotic plaque. J Am Coll Cardiol. 2001; 37: 2126–2130.
Griendling KK, Ushio-Fukai M. NADH/NADPH oxidase and vascular function. Trends Cardiovasc Med. 1997; 7: 301–307.
Suzuki YJ, Ford GD. Redox regulation of signal transduction in cardiac and smooth muscle. J Mol Cell Cardiol. 1999; 31: 345–353.
Rao GN, Berk BC. Active oxygen species stimulate vascular smooth muscle cell growth and proto-oncogene expression. Circ Res. 1992; 70: 593–599.
Katusic ZS. Superoxide anion and endothelial regulation of arterial tone. Free Radic Biol Med. 1996; 20: 443–448.
Lander HM. An essential role for free radicals and derived species in signal transduction. Faseb J. 1997; 11: 118–124.
Abe J, Berk BC. Reactive oxygen species as mediators of signal transduction in cardiovascular disease. Trends Cardiovasc Med. 1998; 8: 59–64.
Hattori Y, Kawasaki H, Abe K, Kanno M. Superoxide dismutase recovers altered endothelium-dependent relaxation in diabetic rat aorta. Am J Physiol. 1991; 261: H1086–1094.
Nakazono K, Watanabe N, Matsuno K, Sasaki J, Sato T, Inoue M. Does superoxide underlie the pathogenesis of hypertension? Proc Natl Acad Sci U S A. 1991; 88: 10045–10048.
White CR, Brock TA, Chang LY, Crapo J, Briscoe P, Ku D, Bradley WA, Gianturco SH, Gore J, Freeman BA, et al. Superoxide and peroxynitrite in atherosclerosis. Proc Natl Acad Sci U S A. 1994; 91: 1044–1048.
Keaney JF, Jr., Xu A, Cunningham D, Jackson T, Frei B, Vita JA. Dietary probucol preserves endothelial function in cholesterol-fed rabbits by limiting vascular oxidative stress and superoxide generation. J Clin Invest. 1995; 95: 2520–2529.
Rajagopalan S, Kurz S, Munzel T, Tarpey M, Freeman BA, Griendling KK, Harrison DG. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation. Contribution to alterations of vasomotor tone. J Clin Invest. 1996; 97: 1916–1923.
Laursen JB, Rajagopalan S, Galis Z, Tarpey M, Freeman BA, Harrison DG. Role of superoxide in angiotensin II-induced but not catecholamine-induced hypertension. Circulation. 1997; 95: 588–593.
Abid MR, Tsai JC, Spokes KC, Deshpande SS, Irani K, Aird WC. Vascular endothelial growth factor induces manganese-superoxide dismutase expression in endothelial cells by a Rac1-regulated NADPH oxidase-dependent mechanism. FASEB J. 2001; 15: 2548–2550.
Ushio-Fukai M, Tang Y, Fukai T, Dikalov SI, Ma Y, Fujimoto M, Quinn MT, Pagano PJ, Johnson C, Alexander RW. Novel role of gp91(phox)-containing NAD(P)H oxidase in vascular endothelial growth factor-induced signaling and angiogenesis. Circ Res. 2002; 91: 1160–1167.
Muhlhauser J, Merrill MJ, Pili R, Maeda H, Bacic M, Bewig B, Passaniti A, Edwards NA, Crystal RG, Capogrossi MC. VEGF165 expressed by a replication-deficient recombinant adenovirus vector induces angiogenesis in vivo. Circ Res. 1995; 77: 1077–1086.
Hynes SO, Smith LA, Richardson DM, Kovesdi I, O’Brien T, Katusic ZS. In vivo expression and function of recombinant GTPCH I in the rabbit carotid artery. Am J Physiol Heart Circ Physiol. 2004; 286: H570–H574.
d’Uscio LV, Baker TA, Mantilla CB, Smith L, Weiler D, Sieck GC, Katusic ZS. Mechanism of endothelial dysfunction in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol. 2001; 21: 1017–1022.
Zanetti M, d’Uscio LV, Kovesdi I, Katusic ZS, O’Brien T. In vivo gene transfer of inducible nitric oxide synthase to carotid arteries from hypercholesterolemic rabbits. Stroke. 2003; 34: 1293–1298.
Li L, Fink GD, Watts SW, Northcott CA, Galligan JJ, Pagano PJ, Chen AF. Endothelin-1 increases vascular superoxide via endothelin(A)-NADPH oxidase pathway in low-renin hypertension. Circulation. 2003; 107: 1053–1058.
Zubkov AY, Rollins KS, Parent AD, Zhang J, Bryan RM, Jr. Mechanism of endothelin-1-induced contraction in rabbit basilar artery. Stroke. 2000; 31: 526–533.
Matsumoto T, D’Uscio LV, Eguchi D, Akiyama M, Smith LA, Katusic ZS. Protective effect of chronic vitamin C treatment on endothelial function of apolipoprotein E-deficient mouse carotid artery. J Pharmacol Exp Ther. 2003; 306: 103–108.
Sato J, Mohacsi T, Noel A, Jost C, Gloviczki P, Mozes G, Katusic ZS, O’Brien T, Mayhan WG. In vivo gene transfer of endothelial nitric oxide synthase to carotid arteries from hypercholesterolemic rabbits enhances endotheliumdependent relaxations. Stroke. 2000; 31: 968–975.
Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K, Franke TF, Papapetropoulos A, Sessa WC. Regulation of endothelium-derived nitric oxide production by the protein kinase Akt. Nature. 1999; 399: 597–601.
Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM. Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature. 1999; 399: 601–605.
Akiyama M, Eguchi D, Weiler D, O’Brien T, Kovesdi I, Scotland RS, Sessa WC, Katusic ZS. Expression and function of recombinant S1179D endothelial nitric oxide synthase in canine cerebral arteries. Stroke. 2002; 33: 1071–1076.
Eguchi D, D’Uscio LV, Wambi C, Weiler D, Kovesdi I, O’Brien T, Katusic ZS. Inhibitory effect of recombinant iNOS gene expression on vasomotor function of canine basilar artery. Am J Physiol Heart Circ Physiol. 2002; 283: H2560–H2566.
Jones SA, O’Donnell VB, Wood JD, Broughton JP, Hughes EJ, Jones OT. Expression of phagocyte NADPH oxidase components in human endothelial cells. Am J Physiol. 1996; 271: H1626–H1634.
Katusic ZS. Vascular endothelial dysfunction: does tetrahydrobiopterin play a role? Am J Physiol Heart Circ Physiol. 2001; 281: H981–H986.
Rajagopalan S, Mohler ER, 3rd, Lederman RJ, Mendelsohn FO, Saucedo JF, Goldman CK, Blebea J, Macko J, Kessler PD, Rasmussen HS, Annex BH. Regional angiogenesis with vascular endothelial growth factor in peripheral arterial disease: a phase II randomized, double-blind, controlled study of adenoviral delivery of vascular endothelial growth factor 121 in patients with disabling intermittent claudication. Circulation. 2003; 108: 1933–1938.(Osamu Suda; Leslie A. Smi)