Adventitial delivery of dominant-negative p67phox attenuates ...

Am J Physiol Heart Circ Physiol 290: H1933–H1941, 2006; doi:10.1152/ajpheart.00690.2005.

Adventitial delivery of dominant-negative p67phox attenuates neointimal hyperplasia of the rat carotid artery Mitchell Weaver,2 Jianhua Liu,4 David Pimentel,1 Daniel J. Reddy,2 Pamela Harding,1 Edward L. Peterson,3 and Patrick J. Pagano1 1

Hypertension and Vascular Research Division, 2Division of Vascular Surgery, and 3Biostatistics and Research Epidemiology Department, Henry Ford Health System, Detroit, Michigan; and 4 Myocardial Biology Unit, Boston University Medical Center, Boston, Massachusetts Submitted 27 June 2005; accepted in final form 31 October 2005

Weaver, Mitchell, Jianhua Liu, David Pimentel, Daniel J. Reddy, Pamela Harding, Edward L. Peterson, and Patrick J. Pagano. Adventitial delivery of dominant-negative p67phox attenuates neointimal hyperplasia of the rat carotid artery. Am J Physiol Heart Circ Physiol 290: H1933–H1941, 2006; doi:10.1152/ajpheart.00690.2005.— Several essential components of NADPH oxidase, including p22phox, gp91phox (nox2) and its homologs nox1 and nox4, p47phox, p67phox, and rac1, are present in the vasculature. We previously reported that p67phox is essential for adventitial fibroblast NADPH oxidase O⫺ 2 production. Thus we postulated that inhibition of adventitial p67phox activity would attenuate angioplasty-induced hyperplasia. To test this hypothesis, we treated the adventitia of carotid arteries with a control adenovirus (Ad-control), a virus expressing dominant-negative p67phox (Ad-p67dn), or a virus expressing a competitive peptide (gp91ds) targeting the p47phox-gp91phox interaction (Ad-gp91ds). Common carotid arteries (CCAs) from male Sprague-Dawley rats were transfected with Ad-control, Ad-p67dn, or Ad-gp91ds in pluronic gel. After 2 days, a 2-F (Fogarty) catheter was used to injure CCAs in vivo. After 14 days, CCAs were perfusion-fixed and analyzed. In 13 experiments, digital morphometry suggested a reduction of neointimal hyperplasia with Ad-p67dn compared with Ad-control; however, the reduction did not reach statistical significance (P ⫽ 0.058). In contrast, a significant reduction was achieved with Adgp91ds (P ⫽ 0.006). No changes in medial area or remodeling were observed with either treatment. Moreover, adventitial fibroblast proliferation in vitro was inhibited by Ad-gp91ds but not by Ad-p67dn, despite confirmation that Ad-p67dn inhibits NADPH oxidase in fibroblasts. These data appear to suggest that a multicomponent vascular NADPH oxidase plays a role in neointimal hyperplasia. However, inhibition of p47phox may be more effective than inhibition of p67phox at attenuating neointimal growth.

NEOINTIMAL HYPERPLASIA is a major response to vascular injury (33). Proliferation and migration of vascular smooth muscle cells (VSMCs) and, more recently, fibroblasts have been implicated in narrowing of the arterial lumen in response to injury, mimicking one of the hallmark characteristics of atherosclerosis (6, 32). Smooth muscle cell migration and proliferation are involved in narrowing of the arterial lumen in response to injury and atherosclerosis, a process promoted by protooncogenes (34) and various growth factors (11, 19). Recent studies have implicated superoxide anion (O⫺ 2 ) derived from NADPH oxidase in vascular cell proliferation and neointimal growth (17, 37), and processes involving reactive

oxygen species (ROS) have been included in a number of growth-related signaling pathways (14). NADPH oxidase is a well-characterized ROS-generating system that catalyzes the one-electron reduction of O2 to O⫺ 2, a precursor of a variety of other ROS. The classical phagocyte NADPH oxidase is known to be a multicomponent enzyme complex that includes the two membrane-spanning polypeptide subunits p22phox and anchoring component gp91phox, which are associated with the plasma membrane cytoskeleton (and together comprise flavocytochrome b558), along with three cytoplasmic polypeptide subunits, p40phox, p47phox, and p67phox (8, 43). The cytosolic guanine nucleotide-binding protein p21rac, a member of the Ras family of proteins, is required for oxidase activation (20). Exposure of the cell to agonists induces interaction of cytosolic and membrane-associated components and activates the dormant oxidase (8). Numerous studies have suggested a similar process of oxidase activation in the vasculature (16, 23, 26, 29, 43). Recently, NADPH oxidases in VSMCs and adventitial fibroblasts have been implicated in the proliferation and migration of cells from the vascular wall to the neointima (13, 18). VSMCs in large arteries express p22phox, p47phox, and homologs of gp91phox but do not appear to express p67phox (21, 22, 36). In contrast, we reported that the vascular adventitia expresses four major components of the oxidase and showed that classical p67phox is involved in adventitial fibroblast O⫺ 2 production (27, 40). In smooth muscle cells and adventitial fibroblasts, critical components of NADPH oxidase are upregulated in response to vascular injury and atherosclerosis, and ROS derived from NADPH oxidase appear to be involved in cell proliferation and migration (2, 38). We recently reported that delivery of a peptide targeting the interaction of p47phox and gp91phox (gp91ds) attenuates neointimal hyperplasia in vivo (18). Thus we postulated that therapies targeting another essential component of the NADPH oxidase system, p67phox, would prove useful in treating vascular injury. Moreover, in light of the growing interest in migrating adventitial fibroblasts playing an important role in neointimal hyperplasia, we examined the influence of p67phox targeting on fibroblast proliferation in vitro. In this study, we examined for the first time the ability of dominant-negative p67phox (p67dn) to attenuate neointimal hyperplasia in response to balloon angioplasty and compared it with the effects of gp91ds. Focusing on the ability of adven-

Address for reprint requests and other correspondence: P. J. Pagano, Hypertension and Vascular Research Div., Rm. 7044, E & R Bldg., Henry Ford Hospital, 2799 West Grand Blvd., Detroit, MI 48202-2689 (e-mail: [email protected]).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

NADPH oxidase; p47phox; adventitia; restenosis

http://www.ajpheart.org

0363-6135/06 $8.00 Copyright © 2006 the American Physiological Society

H1933

H1934

ADVENTITIAL OXIDASE AND NEOINTIMA FORMATION

titial fibroblasts to participate in this response, we also compared the dose-dependent effects of each inhibitor on fibroblast proliferation in vitro. Our data suggest that p67dn attenuates angioplasty-induced increases in vascular neointima formation, but not to the same degree as gp91ds. Interestingly, p67dn had no effect on adventitial fibroblast proliferation in vitro, whereas gp91ds inhibited proliferation. These data appear to suggest that p47phox plays a greater role in adventitial fibroblast proliferation and neointimal hyperplasia than p67phox. MATERIALS AND METHODS

Viral Oxidase Inhibitory Constructs Ad-gp91ds. We designed a bicistronic adenoviral expression vector, Ad-PDGF␤R-gp91ds-IRES-eGFP (Ad-gp91ds) to coexpress gp91ds inhibitory peptide and enhanced green fluorescent protein (eGFP) under the control of the platelet-derived growth factor-␤ receptor (PDGF␤R). The pPDGF␤R-luciferase plasmid (3) was generously provided by Dr. K. Funa (Go¨teborg University, Go¨teborg, Sweden). A 1,468-bp portion of the pPDGF␤R-luciferase plasmid was removed by Sac I digestion (3) (PDGF␤R promoter sequence) and inserted into a replication-deficient human adenovirus serotype 5 (Ad5) vector. This 1,468-bp portion of the pPDGF␤R-luciferase vector was placed upstream from a Kozac sequence (GCC-ACC-ATG), followed by the nucleotide sequence (TGC-TCG-ACA-AGG-ATTCGA-AGA-CAA-CTG) encoding the gp91ds peptide, an internal ribosome entry site (IRES) sequence, a 719-bp sequence encoding eGFP, and a sequence for simian virus 40 pRep8 poly(A). Ad-p67dn. Adenovirus expressing p67dn (Ad-p67dn) was constructed at the Myocardial Biology Unit, Boston University, by placement of a cytomegalovirus (CMV) promoter upstream from the transcript encoding p67dn (substitution of alanine for valine at amino acid 204) as described by Han et al. (15). Ad-control. Ad-CMV-eGFP (University of Iowa Gene Transfer Core) was used as a control. eGFP expression was under the control of the CMV promoter. Animals and Viral Infection Male Sprague-Dawley rats (10 –12 wk old, 300 – 400 g; Charles River) were anesthetized with ketamine (80 mg/kg ip) and xylazine (7 mg/kg ip), and the neck was dissected to expose the common carotid artery (CCA) as well as the internal and external carotid arteries. Ad-control, Ad-p67dn, or Ad-gp91ds in 15% pluronic gel [Poloxamer 407 NF, BASF; 3.5 ⫻ 108 plaque-forming units (pfu)/ml] was gently spread around the outside of the left CCA. The incision was closed and the animals were allowed to recover, with free access to water and food. All protocols were approved by the Institutional Animal Care and Use Committee of Henry Ford Hospital and are consistent with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. In Vivo Balloon Injury and Morphometric Analysis of Neointima Formation At 2 days after viral infection, the animals were anesthetized and the incision was reopened. A 4-0 silk ligature was tied around the distal portion of the left external carotid artery, and another was passed around the proximal portion but not tied. A small transverse arteriotomy was made between the ligatures, and a 2-Fr (Fogarty) embolectomy catheter (Edwards Life Sciences) was passed through the left CCA into the thoracic aorta. The balloon was distended with 0.9% saline to increase CCA diameter by ⬃100% and withdrawn through the CCA to cause arterial injury. Distension and withdrawal were repeated five times. After CCA injury, the balloon catheter was removed and the proximal external carotid artery suture was tied. After the incision was closed, the animals were allowed to recover, AJP-Heart Circ Physiol • VOL

with free access to water and food. After 14 days, the vessels were perfusion-fixed with 10 ml of PBS followed by 10% formaldehyde in PBS under pressure (⬃120 mmHg). CCAs were harvested (with care taken not to disrupt the adventitia) and embedded in paraffin, and the middle third was serially cut into 6-␮m sections. Masson’s trichrome staining (Sigma), performed following the manufacturer’s instructions, revealed several layers of elastic laminae. Digital morphometric analysis was performed using image analysis software (Spot Diagnostics) by an examiner with no knowledge of the treatment group. When cross-sectional area was measured, medial area was determined by subtraction of the area defined by the internal elastic lamina (IEL) from the area defined by the external elastic lamina (EEL). Intimal area was determined by subtraction of the luminal area from the area defined by the IEL. We compared elastolysis visually at high magnification by examining whether the IEL and EEL were intact. Radial thicknesses of the neointima and media were measured at four points around the circumference and averaged. Myointimal proliferation of the injured carotid artery was expressed as the ratio of neointimal area and thickness to medial area and thickness, respectively. The circumferences of the IEL and EEL were measured to examine the remodeling response. Each section was examined several times, and the values were averaged. Immunohistochemistry for Carotid Artery eGFP and p67phox Expression Monoclonal mouse anti-eGFP (Clontech) was used to detect eGFP expression, and monoclonal mouse anti-p67phox (generously supplied by Dr. Mark Quinn) (7) was used to detect p67phox expression. Sections were heated at 58°C for 2 h, deparaffinized and hydrated using standard techniques, and then boiled in 10 mM citric acid buffer for 10 min for antigen retrieval. The sections were incubated in 0.3% H2O2 in 80% methanol for 30 min. Immunostaining was performed using a mouse IgG kit (Vector Laboratories). The sections were blocked with 10% normal horse serum for 30 min and then incubated overnight at 4°C with primary antibody (1:250 dilution for anti-eGFP and 1:20 dilution for anti-p67phox in PBS containing 2% normal horse serum). Negative controls were incubated with the same concentration of matching IgG isotype (IgG2a). Sections were incubated with biotinylated secondary antibody (1:400 dilution in PBS containing 2% normal horse serum) for 30 min and then with ABC reagent for 30 min (Vector). The reaction was visualized using diaminobenzidine tetrahydrochloride (Vector Laboratories). The sections were counterstained with hematoxylin (Sigma), and the slides were dehydrated with alcohol, cleared in xylene, and mounted. Preparation of Rat Aortic Adventitial Fibroblasts Aortic fibroblasts were prepared as described previously (27). Briefly, aortas from anesthetized rats were removed using sterile techniques and placed in DMEM containing 100 U/ml penicillin and 100 ␮g/ml streptomycin. Vessels were cleaned of perivascular adipose tissue and cut longitudinally, and the endothelium was scraped free. The tissue was incubated in DMEM containing 1 mg/ml collagenase and 0.125 mg/ml elastase for 10 –15 min at room temperature. Medial smooth muscle cells were peeled from the adventitia, which was then digested in DMEM containing 1 mg/ml collagenase and 0.125 mg/ml elastase for 4 h at 37°C. The resulting solution was centrifuged, and the pellet was resuspended in DMEM containing 20% fetal bovine serum (FBS) and then seeded onto 100-mm culture dishes. These cells were considered passage 0. Comparison of the Effect of Ad-p67dn and Ad-gp91ds on Adventitial Fibroblast Proliferation Fibroblasts at passage 7 were seeded in six-well plates (Corning) at a density of 50,000 cells/well and grown to 60% confluence. The cells were washed extensively with Hanks’ balanced salt solution to ensure 290 • MAY 2006 •

www.ajpheart.org


total removal of serum and placed in 0% FBS-DMEM for 20 h (quiescence). Ad-control, Ad-p67dn, or Ad-gp91ds was added to individual wells for the dose-response assay (3.5 ⫻ 107, 1.0 ⫻ 108, and 3.5 ⫻ 108 pfu/ml) and allowed to transfect the cells for 4 h. The medium was removed, and 0% FBS-DMEM or 10% FBS-DMEM was added to the wells. The cells were harvested at 66 h and counted in a Coulter counter. Cell number for each treatment group was calculated as a percentage and compared with Ad-control-treated cells. Effect of Ad-p67dn on Adventitial Fibroblast NADPH Oxidase Activity Fibroblasts at passage 7 were seeded in 60-mm culture plates (Corning), grown to 60% confluence, and brought to quiescence, and Ad-control or Ad-p67dn (3.5 ⫻ 108 pfu/ml) was added for 4 h. The medium was replaced with 0% FBS-DMEM for 72–96 h. Vehicle (0.01 N acetic acid in saline) or ANG II (10⫺7 M) was added, and the plates were incubated at 37°C for 3 h. NADPH oxidase activity was measured as described previously (30) by comparison of ANG IIstimulated NADPH-dependent lucigenin (5 ␮M) chemiluminescence in Ad-control with that in Ad-p67dn-transfected cells. Statistical Analyses Values are means ⫾ SE, with n indicating the number of experiments in vivo. Morphometric measurements were compared for differences between groups by one-way ANOVA. Body weight, IEL and EEL circumference, and cell proliferation were compared by ANOVA with repeated measures. To determine differences, pairwise testing was implemented using Hochberg’s method to set the criteria for significance and adjust for multiple testing. NADPH oxidase activity was examined using Student’s two-sample t-test. RESULTS

Body Weight Body weight rose over the course of the study in all treatment groups. However, we observed no difference in body weight among groups at the beginning or end of the study. Transgene Expression by Adenoviral Constructs Figure 1A shows eGFP in a cross section of the rat CCA treated with Ad-control. Staining was prominent in both adventitia and neointima. Figure 1B shows a negative control from the Ad-control group in which primary antibody for eGFP was replaced by IgG2a. In Fig. 1C, eGFP staining in a CCA treated with Ad-gp91ds demonstrates substantially lighter staining, primarily in the adventitia. At higher magnification (Fig. 1D), staining was associated primarily with fusiform adventitial fibroblasts. We also confirmed an effect of the p67dn construct by immunohistochemistry using monoclonal p67phox (7). As shown in Fig. 1, E and F, reactivity for p67phox is reduced in Ad-p67dn-treated (n ⫽ 4) compared with Adcontrol-treated carotid arteries (n ⫽ 5). A negative control, in which primary antibody was replaced with IgG2a, showed no reactivity (not shown). These data are consistent with reduced expression of the wild-type protein. Alternatively, reduced specific binding of the antibody as a consequence of overexpressed mutated p67dn may have resulted in lower overall reactivity. In both cases, p67phox activity would be reduced, resulting in lower ROS production. AJP-Heart Circ Physiol • VOL

H1935

Effect of Oxidase Inhibitor Constructs vs. Control Virus on Injury-Induced Changes in Neointimal Area in Carotid Artery Figure 2 shows representative cross sections of injured CCAs transfected with Ad-gp91ds and Ad-p67dn vs. Adcontrol. At 14 days after injury, in the control group, growth in neointimal and medial areas was approximately equal (Fig. 2A). Neointimal-to-medial area ratio appeared partially reduced in CCAs treated with Ad-p67dn (Fig. 2B) and markedly reduced in those treated with Ad-gp91ds (Fig. 2C). There appeared to be a substantial decrease in neointimalto-medial area ratio in CCAs treated with Ad-p67dn compared with adenoviral control, as changes closely approached statistical significance (45%, P ⫽ 0.058, n ⫽ 13 vs. 25). By comparison, gp91ds-expressing virus caused a statistically significant reduction in neointimal-to-medial area ratio (67%, P ⫽ 0.006, n ⫽ 13 vs. 25; Fig. 3B). Digital morphometric analysis suggested that neointimal area was smaller in CCAs from Ad-p67dn than from Ad-control (39%), but the difference was not significant; in contrast, neointimal area was significantly decreased by Ad-gp91ds (64%; Fig. 3A). Similarly, neointimal thickness appeared reduced, but comparisons between Adcontrol and Ad-p67dn did not reach statistical significance (49%); on the other hand, thickness was significantly lower in CCAs treated with Ad-gp91ds than in those treated with Ad-control (63%; Fig. 4A). Similarly, neointimal-to-medial thickness ratios appeared lower in the Ad-p67dn group (45%) but did not reach significance, whereas mean thickness ratio was significantly lower in the gp91ds group (64%, P ⫽ 0.015; Fig. 4B). Overall, data analysis of all comparisons of Adcontrol vs. Ad-p67dn approached statistical significance, suggesting a reduction in neointimal growth. No difference in elastolysis was observed between groups (data not shown). Although there was an apparent trend toward increased luminal area in the gp91ds group, it did not vary significantly across treatment protocols (data not shown). Effect of Oxidase Inhibitor Constructs vs. Control Virus on CCA Medial Area Contrary to the changes in neointimal area, neither inhibitor construct had any effect on medial area. Digital quantification of medial area (Fig. 5A) and medial thickness (Fig. 5B) did not vary across treatment groups. Effect of Oxidase Inhibitor Constructs vs. Control Virus on Vessel Size To examine whether either treatment had any effect on vessel size, we digitally determined IEL and EEL circumference. Figure 6, A and B, suggests a slight trend toward larger IEL and EEL circumference in the p67dn and gp91ds groups than in the control group; however, these data did not reach significance. Concentration-Response Effects of Adenoviral Oxidase Inhibitors on Adventitial Fibroblast Proliferation To examine the ability of each construct to inhibit cell proliferation more thoroughly, we tested the effect of a range of concentrations of each virus on adventitial fibroblast proliferation in vitro. We did not observe an inhibitory effect of 290 • MAY 2006 •

www.ajpheart.org

H1936


Fig. 1. Examination of enhanced green fluorescent protein (eGFP) and p67phox crossreactivity in rat common carotid arteries (CCAs). A: representative cross section of left CCA 14 days after injury and 16 days after treatment with control adenovirus (Ad-control) stained for eGFP. Magnification ⫻400. B: negative control for A, with substitution of IgG2a as primary antibody. Magnification ⫻400. C: representative cross section of left CCA 14 days after injury and 16 days after treatment with virus expressing competitive peptide targeting p47phox-gp91phox interaction (Ad-gp91ds) stained for eGFP. Magnification ⫻400. D: C at a higher magnification (⫻1,000). E: representative cross section of Ad-control stained for p67phox. Magnification ⫻1,000. F: cross section of virus expressing dominant-negative p67phox (Ad-p67dn) stained for p67phox. Magnification ⫻1,000. Positive staining appears brown. L, lumen; M, media; A, adventitia.

either virus on fibroblast proliferation at 42 h after 10% serum treatment at any concentration tested. However, we observed a 35% reduction in cell number at 66 h of serum treatment with 3.5 ⫻ 108 pfu/ml Ad-gp91ds (Fig. 7), the same concentration that proved effective in vivo. Interestingly, Ad-p67dn did not affect adventitial fibroblast proliferation at any concentration. Effect of Ad-p67dn on Adventitial Fibroblast NADPH Oxidase Activity We previously established that gp91ds inhibits whole vessel and adventitial fibroblast ROS production (9, 18, 25, 30). Thus we examined whether Ad-p67dn inhibits fibroblast NADPH oxidase activity, despite its lack of effect on proliferation. Figure 8 shows that ANG II-stimulated NADPH oxidase activity in adventitial fibroblasts was inhibited by 47% by treatment with 3.5 ⫻ 108 pfu/ml Ad-p67dn vs. Ad-control. AJP-Heart Circ Physiol • VOL

DISCUSSION

Our present findings suggest that a vascular cell oxidase resembling the classical phagocyte NADPH oxidase is critically involved in neointimal growth arising from balloon angioplasty in the rat. Our data suggest for the first time that p67phox may be involved in perivascular activation of the oxidase, which precedes neointimal hyperplasia; however, targeting of the p47phox-gp91phox interaction appears to inhibit this process more effectively. We and others previously established the presence of four major components of classical NADPH oxidase, gp91phox, p22phox, p47phox, and p67phox, in the perivascular region of rabbit and rat vessels and showed that p67phox is involved in adventitial fibroblast oxidase activity (26, 27, 41). Our present findings are consistent with partial involvement of this component in the vascular migratory and hyperplastic response to injury, because blocking it with novel adenoviral constructs partially reduced neointimal growth. In290 • MAY 2006 •

www.ajpheart.org


H1937

Fig. 3. Neointimal area is reduced more by Ad-gp91ds than by Ad-p67dn. A: neointimal area in CCAs 14 days after balloon injury. B: neointimal-to-medial area ratios. Values are means ⫾ SE; n ⫽ 25 (Ad-control) and 13 (Ad-p67dn and Ad-gp91ds). *P ⬍ 0.05 vs. Ad-control.

terestingly, when we examined the effects of Ad-gp91ds vs. those of Ad-p67dn on adventitial fibroblast proliferation in vitro, we found that, although Ad-p67dn reduced NADPH oxidase activity, only Ad-gp91ds was able to reduce cell growth. Taken together, these data appear to suggest that a multicomponent NADPH oxidase is involved in the fundamental processes leading to balloon angioplasty-induced neointimal hyperplasia. However, the data also appear to suggest that adventitial p47phox is more critical to neointimal hyperplasia than classical p67phox, especially with regard to the contribution by vascular fibroblasts, and may suggest that a homolog of p67phox is involved in this process. Immunohistochemistry for eGFP revealed that adenoviral constructs delivered the marker gene to the perivascular cells and neointima of the rat CCA. Because there are no available means of immunologically distinguishing endogenous p67phox Fig. 2. Representative cross sections obtained 14 days after balloon injury of CCAs transfected with Ad-control (A), Ad-p67dn (B), and Ad-gp91ds (C). Images represent cross sections from 13 to 25 rats in each group, showing all 3 segments of the CCA: adventitia, media, and neointima (left to right). Masson’s trichrome stain. Magnification ⫻400. AJP-Heart Circ Physiol • VOL

290 • MAY 2006 •

www.ajpheart.org

H1938


however, a subpopulation of macrophage-like cells also appeared to express the inhibitor (25). Practically speaking, the lower overall eGFP expression in Ad-gp91ds vs. Ad-control raised the question of whether higher eGFP expression in the controls may have contributed artificially to heightened neointimal growth and thus enhanced the effect of the inhibitory construct. This does not appear to be the case, because neointimal size and neointimal-to-medial area ratios were lower in Ad-eGFP-treated CCAs than in non-virus-treated controls (18). Neointimal hyperplasia is characterized by VSMC proliferation and migration to the neointima during injury. Recent studies suggest that adventitial myofibroblasts derived from fibroblasts also proliferate and migrate to the site of injury (24, 42). Because fibroblast proliferation is ROS dependent (17), inhibition of ROS production in adventitial fibroblasts may be an important means of attenuating the direct contribution of fibroblasts to injury- and atherosclerosis-related hyperplasia. Perhaps more importantly, perivascular oxidases appear to have an indirect paracrine effect on the vascular wall via the release of tissue-permeant ROS such as H2O2, which can stimulate medial smooth muscle cells to undergo a change

Fig. 4. Neointimal thickness is reduced more by Ad-gp91ds than by Adp67dn. A: neointimal thickness in CCAs 14 days after balloon injury. B: neointimal-to-medial thickness ratios. Values are means ⫾ SE; n ⫽ 25 (Ad-control) and 13 (Ad-p67dn and Ad-gp91ds). *P ⬍ 0.05 vs. Ad-control.

from p67dn (a 1-amino acid substitution from the endogenous form), we were unable to assess p67dn delivery directly. Thus, as a control, we chose a vector expressing eGFP under the direction of the same CMV promoter, allowing us to use eGFP to determine p67dn distribution. When we performed immunohistochemistry for p67phox, we found a marked decrease in p67phox reactivity in Ad-p67dn- vs. Ad-control-treated carotid arteries. These data are consistent with reduced expression of the wild-type p67phox. Alternatively, our data may suggest reduced specific binding of the antibody to the overexpressed mutated p67dn, resulting in decreased reactivity under the stringency of the immunohistochemical assay. In either case, p67phox activity would be reduced in the Ad-p67dn group, resulting in lower ROS production. Expression of eGFP in CCAs infected with Ad-gp91ds seems to be localized primarily to adventitial fibroblasts but is expected to include other proliferating cells. Limited adventitial eGFP expression early after injury (1 day) is consistent with the presence of its fibroblast-active PDGF␤R promoter, which has been reported to direct expression in cells of a proliferative phenotype (3). In fact, in our recent report examining the mouse carotid artery, we showed that Ad-gp91ds caused the inhibitor to be expressed primarily in fibroblasts; AJP-Heart Circ Physiol • VOL

Fig. 5. Ad-gp91ds and Ad-p67dn have no effect on medial area (␮m2 ⫻ 103) (A) and thickness (B) in CCAs 14 days after balloon injury. Values are means ⫾ SE; n ⫽ 25 (Ad-control) and 13 (Ad-p67dn and Ad-gp91ds). 290 • MAY 2006 •

www.ajpheart.org


H1939

Fig. 7. Ad-gp91ds, but not Ad-p67dn, inhibits adventitial fibroblast proliferation. Concentration-response effects of infection with Ad-gp91ds and Adp67dn compared with Ad-control on cell count after 66 h of treatment with 10% FBS. pfu, Plaque-forming units. Values are means ⫾ SE. *P ⬍ 0.05 vs. Ad-control.

Fig. 6. Ad-gp91ds and Ad-p67dn have no effect on parameters of remodeling, i.e., internal elastic lamina (IEL) and external elastic lamina (EEL) circumference, in CCAs 14 days after balloon injury. Values are means ⫾ SE; n ⫽ 25 (Ad-control) and 13 (Ad-p67dn and Ad-gp91ds).

from the contractile to the proliferative and migratory phenotype (31). Regardless of a direct or indirect effect on proliferation, adventitial NADPH oxidase is increasingly being seen as an early mediator in this response (9). Our present findings are consistent with both a paracrine and a direct effect of fibroblast oxidase inhibition on vascular cell hyperplasia. Several reports have shown an association between various forms of vascular NADPH oxidase and neointimal growth (28, 38). Moreover, studies have clearly shown that phosphorylation and translocation of the classical cytosolic component p47phox to the plasma membrane participate in the process of vascular NADPH oxidase activation (23, 29). We recently confirmed that interaction between p47phox and gp91phox in NADPH oxidase is involved in the production of vascular O⫺ 2 and neointimal growth by showing that gp91ds-tat (a novel peptide inhibitor of p47-gp91 interaction and precursor of Ad-gp91ds) inhibits stretch-induced O⫺ 2 production and neointimal growth of the rat CCA (18). We infused gp91ds-tat or vehicle control into rats for 2 days and then examined O⫺ 2 after stretch. We observed a significant reduction in ROS and neointimal proliferation by gp91ds-tat (18); in subsequent studies, we demonstrated that the active sequence gp91ds was AJP-Heart Circ Physiol • VOL

capable of markedly reducing whole vessel and fibroblast ROS levels (9, 25) and neointimal proliferation (9). In the present study, on the basis of substantial evidence showing that vascular NADPH oxidase is a complex system composed of at least five essential components (31), we compared the antihyperplastic effect of delivering Ad-gp91ds vs. Ad-p67dn to the adventitia. We and others showed that classical NADPH oxidase components gp91phox, p47phox, and p67phox are present in the aortic adventitia of the rabbit and rat (27, 40) and postulated that adventitial oxidase plays a role in the promotion of neointimal growth (31). For this reason, we examined the potential contribution of adventitial p67phox per se to vascular neointimal growth compared with the p47phoxgp91phox interaction. We previously showed that p67phox is involved in adventitial fibroblast NADPH oxidase activity (21, 22, 36). However, consistent with a lesser role of adventitial

Fig. 8. Effect of Ad-p67dn vs. Ad-control on particulate NADPH-dependent superoxide anion (O⫺ 2 ) generation. Ad-p67dn inhibits adventitial fibroblast NADPH oxidase activity. Values are means ⫾ SE, expressed as percentage. *P ⬍ 0.05 vs. Ad-control. 290 • MAY 2006 •

www.ajpheart.org

H1940


p67phox in stretch-induced neointimal growth, the present morphometric data suggest a reduction in neointimal growth in adventitially applied Ad-p67dn-treated CCAs compared with viral controls, as changes closely approached statistical significance (45%, P ⫽ 0.058). On the other hand, Ad-gp91ds caused statistically significant inhibition compared with control (67%, P ⫽ 0.001). These data indicate that adventitial p47phox is critically involved in NADPH oxidase-mediated neointimal growth and suggest a role for classical p67phox, particularly because gp91ds expression seemed more limited than p67dn expression (Fig. 1B vs. 1A). Moreover, the lesser effect of Ad-p67dn may be attributed in part to the presence of the p67phox homolog NOXA1 (12). It is unclear whether NOXA1 is important for adventitial or medial NADPH oxidase activity. Because inhibitors of this homolog are unavailable, we are currently unable to address the contribution of NOXA1 relative to that of p67phox. Because a number of previous studies suggested a role for migrating adventitial fibroblasts per se in neointimal hyperplasia (35), we compared the effectiveness of Ad-p67dn vs. Ad-gp91ds to reduce fibroblast proliferation in vitro. Interestingly, we observed no effect of Ad-p67dn over a range of concentrations, including the highest concentration that showed a partial effect in vivo. In contrast, Ad-gp91ds caused a 35% reduction in fibroblast proliferation (compared with a 67% reduction in neointimal hyperplasia in vivo). Indeed, the greater effect of these constructs in vivo is consistent with the expected participation of multiple cell types in neointimal hyperplasia. That is, the enhanced effect of both adenoviral constructs in vivo appears to suggest that ROS derived from adventitial cell types other than fibroblasts play an important role in vascular cell hyperplasia and migration in this model. It is also noteworthy that Ad-p67dn did not inhibit fibroblast proliferation, despite its ability to decrease NADPH oxidase activity. Thus it is tempting to speculate that these findings suggest a difference in the function of Ad-p67dn and gp91ds; i.e., their antiproliferative actions may depend on the particular isoforms and/or subcellular localizations of NADPH oxidase that are affected. In conclusion, our data are consistent with the involvement of a multicomponent adventitial phagocyte-like NADPH oxidase in neointimal growth. The data indicate that classical p47phox is critically involved in this response, whereas they appear to suggest only partial involvement of classical p67phox. However, future studies examining the relative contribution of their homologs, as well as other vascular cells, are necessary. Our findings also shed light on the possible clinical utility of perivascular application of viral inhibitors to reverse deleterious processes involving vascular oxidases. Perspectives NADPH oxidases are increasingly being recognized as important mediators in a complex signaling cascade leading to vascular growth. Studies show that the vascular growth response involves c-Src-mediated NADPH oxidase assembly and includes activation of a host of tyrosine and serine/threonine kinases, leading to activation of transcription factors, cell hypertrophy, and proliferation (5, 10, 14, 39). With the discovery of homologs of the various NADPH oxidase components in the vasculature, including gp91phox homologs nox1 and nox4, AJP-Heart Circ Physiol • VOL

as well as interaction with classical p22phox (1) and novel cytosolic components (4), clearly the cellular effects of oxidase activation are expected to be multifactorial. Thus a multifaceted, combinatorial pharmacological approach to oxidase inhibition will likely be necessary to attenuate cell growth. ACKNOWLEDGMENTS We thank Dr. Beverly Davidson and the Gene Transfer Vector Core at the University of Iowa for assistance in constructing the virus. We also thank Dr. K. Funa for the generous supply of the PDGF␤R promoter, as well as Dr. Mark Quinn for the supply of the p67phox antibody. GRANTS This work was supported by National Heart, Lung, and Blood Institute Grants HL-55425 and HL-28982, American Heart Association (AHA) Grant 0151407Z, and the Fund for Henry Ford Hospital. P. J. Pagano is a recipient of an Established Investigator Award from the AHA. REFERENCES 1. Ambasta RK, Kumar P, Griendling KK, Schmidt HHHW, Busse R, and Brandes RP. Direct interaction of the novel Nox proteins with p22phox is required for the formation of a functionally active NADPH oxidase. J Biol Chem 279: 45935– 45941, 2004. 2. Azumi H, Inoue N, Takeshita S, Rikitake Y, Kawashima S, Hayashi Y, Itoh H, and Yokoyama M. Expression of NADH/NADPH oxidase p22phox in human coronary arteries. Circulation 100: 1494 –1498, 1999. 3. Ballagi AE, Ishizaki A, Nehlin J-O, and Funa K. Isolation and characterization of the mouse PDGF-␤-receptor promoter. Biochem Biophys Res Commun 210: 165–173, 1995. 4. Banfi B, Clark RA, Steger K, and Krause K-H. Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278: 3510 –3513, 2003. 5. Cifuentes ME and Pagano PJ. c-Src and smooth muscle NAD(P)H oxidase: assembling a path to hypertrophy. Arterioscler Thromb Vasc Biol 23: 919 –921, 2003. 6. Clowes AW, Reidy MA, and Clowes MM. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 49: 327–333, 1983. 7. De Leo FR, Ulman KV, Davis AR, Jutila KL, and Quinn MT. Assembly of the human neutrophil NADPH oxidase involves binding of p67phox and flavocytochrome b to a common functional domain in p47phox. J Biol Chem 271: 17013–17020, 1996. 8. DeLeo FR and Quinn MT. Assembly of the phagocyte NADPH oxidase: molecular interaction of oxidase proteins. J Leukoc Biol 60: 677– 691, 1996. 9. Dourron HM, Jacobson GM, Park JL, Liu J, Reddy DJ, Scheel ML, and Pagano PJ. Perivascular gene transfer of NADPH oxidase inhibitor suppresses angioplasty-induced neointimal proliferation of rat carotid artery. Am J Physiol Heart Circ Physiol 288: H946 –H953, 2005. 10. Finkel T. Oxygen radicals and signaling. Curr Opin Cell Biol 10: 248 –253, 1998. 11. Forney Prescott M, Webb RL, and Reidy MA. Angiotensin-converting enzyme inhibitor versus angiotensin II, AT1 receptor antagonist. Effects on smooth muscle cell migration and proliferation after balloon catheter injury. Am J Pathol 139: 1291–1296, 1991. 12. Geiszt M, Lekstrom K, Witta J, and Leto TL. Proteins homologous to p47phox and p67phox support superoxide production by NAD(P)H oxidase 1 in colon epithelial cells. J Biol Chem 278: 20006 –20012, 2003. 13. Griendling KK, Minieri CA, Ollerenshaw JD, and Alexander RW. Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res 74: 1141–1148, 1994. 14. Griendling KK, Sorescu D, Lasse`gue B, and Ushio-Fukai M. Modulation of protein kinase activity and gene expression by reactive oxygen species and their role in vascular physiology and pathophysiology. Arterioscler Thromb Vasc Biol 20: 2175–2183, 2000. 15. Han C-H, Freeman JLR, Lee T, Motalebi SA, and Lambeth JD. Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67phox. J Biol Chem 273: 16663–16668, 1998. 16. Hwang J, Saha A, Boo YC, Sorescu GP, McNally JS, Holland SM, Dikalov S, Giddens DP, Griendling KK, Harrison DG, and Jo H. Oscillatory shear stress stimulates endothelial production of O⫺ 2 from 290 • MAY 2006 •

www.ajpheart.org


17.

18.

19.

20. 21. 22.

23. 24.

25. 26.

27.

28.

29.

p47phox-dependent NAD(P)H oxidases, leading to monocyte adhesion. J Biol Chem 278: 47291– 47298, 2003. Irani K, Xia Y, Zweier JL, Sollott SJ, Der CJ, Fearon ER, Sundaresan M, Finkel T, and Goldschmidt-Clermont PJ. Mitogenic signaling mediated by oxidants in Ras-transformed fibroblasts. Science 275: 1649 – 1652, 1997. Jacobson GM, Dourron HM, Liu J, Carretero OA, Reddy DJ, Andrzejewski T, and Pagano PJ. Novel NAD(P)H oxidase inhibitor suppresses angioplasty-induced superoxide and neointimal hyperplasia of rat carotid artery. Circ Res 92: 637– 643, 2003. Kauffman RF, Bean JS, Zimmerman KM, Brown RF, and Steinberg MI. Losartan, a nonpeptide angiotensin II (Ang II) receptor antagonist, inhibits neointima formation following balloon injury to rat carotid arteries. Life Sci 49: PL-223–PL-228, 1991. Knaus UG, Morris S, Dong H-J, Chernoff J, and Bokoch GM. Regulation of human leukocyte p21-activated kinases through G proteincoupled receptors. Science 269: 221–223, 1995. Lasse`gue B and Clempus RE. Vascular NAD(P)H oxidases: specific features, expression, and regulation. Am J Physiol Regul Integr Comp Physiol 285: R277–R297, 2003. Lasse`gue B, Sorescu D, Szo¨cs K, Yin Q, Akers M, Zhang Y, Grant SL, Lambeth JD, and Griendling KK. Novel gp91phox homologues in vascular smooth muscle cells. Nox1 mediates angiotensin II-induced superoxide formation and redox-sensitive signaling pathways. Circ Res 88: 888 – 894, 2001. Lavigne MC, Malech HL, Holland SM, and Leto TL. Genetic demonstration of p47phox-dependent superoxide anion production in murine vascular smooth muscle cells. Circulation 104: 79 – 84, 2001. Li G, Chen S-J, Oparil S, Chen Y-F, and Thompson JA. Direct in vivo evidence demonstrating neointimal migration of adventitial fibroblasts after balloon injury of rat carotid arteries. Circulation 101: 1362–1365, 2000. Liu J, Ormsby A, Oja-Tebbe N, and Pagano PJ. Gene transfer of NAD(P)H oxidase inhibitor to the vascular adventitia attenuates medial smooth muscle hypertrophy. Circ Res 95: 587–594, 2004. Pagano PJ, Chanock SJ, Siwik DA, Colucci WS, and Clark JK. Angiotensin II induces p67phox mRNA expression and NADPH oxidase superoxide generation in rabbit aortic adventitial fibroblasts. Hypertension 32: 331–337, 1998. Pagano PJ, Clark JK, Cifuentes-Pagano ME, Clark SM, Callis GM, and Quinn MT. Localization of a constitutively active, phagocyte-like NADPH oxidase in rabbit aortic adventitia: enhancement by angiotensin II. Proc Natl Acad Sci USA 94: 14483–14488, 1997. Paravicini TM, Gulluyan LM, Dusting GJ, and Drummond GR. Increased NADPH oxidase activity, gp91phox expression, and endothelium-dependent vasorelaxation during neointima formation in rabbits. Circ Res 91: 54 – 61, 2002. Patterson C, Ruef J, Madamanchi NR, Barry-Lane P, Hu Z, Horaist C, Ballinger CA, Brasier AR, Bode C, and Runge MS. Stimulation of

AJP-Heart Circ Physiol • VOL

30.

31. 32. 33. 34. 35.

36.

37.

38.

39.

40.

41.

42. 43.

H1941

a vascular smooth muscle cell NAD(P)H oxidase by thrombin. Evidence that p47phox may participate in forming this oxidase in vitro and in vivo. J Biol Chem 274: 19814 –19822, 1999. Rey FE, Cifuentes ME, Kiarash A, Quinn MT, and Pagano PJ. Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O⫺ 2 and systolic blood pressure in mice. Circ Res 89: 408 – 414, 2001. Rey FE and Pagano PJ. The reactive adventitia. Fibroblast oxidase in vascular function. Arterioscler Thromb Vasc Biol 22: 1962–1971, 2002. Schwartz SM, Campbell GR, and Campbell JH. Replication of smooth muscle cells in vascular disease. Circ Res 58: 427– 444, 1986. Schwartz SM, deBlois D, and O’Brien ERM. The intima. Soil for atherosclerosis and restenosis. Circ Res 77: 445– 465, 1995. Schwartz SM and Murry CE. Proliferation and the monoclonal origins of atherosclerotic lesions. Annu Rev Med 49: 437– 460, 1998. Shi Y, O’Brien JE Jr, Fard A, Mannion JD, Wang D, and Zalewski A. Adventitial myofibroblasts contribute to neointimal formation in injured porcine coronary arteries. Circulation 94: 1655–1664, 1996. Sorescu D, Weiss D, Lasse`gue B, Clempus RE, Szo¨cs K, Sorescu GP, Valppu L, Quinn MT, Lambeth JD, Vega JD, Taylor WR, and Griendling KK. Superoxide production and expression of nox family proteins in human atherosclerosis. Circulation 105: 1429 –1435, 2002. Souza HP, Souza LC, Anastacio VM, Pereira AC, Junqueira MdL, Krieger JE, da Luz PL, Augusto O, and Laurindo FRM. Vascular oxidant stress early after balloon injury: evidence for increased NAD(P)H oxidoreductase activity. Free Radic Biol Med 28: 1232–1242, 2000. Szo¨cs K, Lasse`gue B, Sorescu D, Hilenski LL, Valppu L, Couse TL, Wilcox JN, Quinn MT, Lambeth JD, and Griendling KK. Upregulation of nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler Thromb Vasc Biol 22: 21–27, 2002. Touyz RM, Yao G, and Schiffrin EL. c-Src induces phosphorylation and translocation of p47phox. Role in superoxide generation by angiotensin II in human vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 23: 981–987, 2003. Wang HD, Hope S, Du Y, Quinn MT, Cayatte A, Pagano PJ, and Cohen RA. Paracrine role of adventitial superoxide anion in mediating spontaneous tone of the isolated rat aorta in angiotensin II-induced hypertension. Hypertension 33: 1225–1232, 1999. Wang HD, Pagano PJ, Du Y, Cayatte AJ, Quinn MT, Brecher P, and Cohen RA. Superoxide anion from the adventitia of the rat thoracic aorta inactivates nitric oxide. Circ Res 82: 810 – 818, 1998. Zalewski A and Shi Y. Vascular myofibroblasts. Lessons from coronary repair and remodeling. Arterioscler Thromb Vasc Biol 17: 417– 422, 1997. Zhan S, Vazquez N, Zhan SH, Wientjes FB, Budarf ML, Schrock E, Ried T, Green ED, and Chanock SJ. Genomic structure, chromosomal localization, start of transcription, and tissue expression of the human p40-phox, a new component of the nicotinamide adenine dinucleotide phosphate-oxidase complex. Blood 88: 2714 –2721, 1996.

290 • MAY 2006 •

www.ajpheart.org