- Email: [email protected]
Functional association of nox1 with p22phox in vascular smooth muscle cells
Free Radical Biology & Medicine, Vol. 37, No. 10, pp. 1542–1549, 2004 Copyright D 2004 Elsevier Inc. Printed in the USA. All rights reserved 0891-5849/$-see front matter
doi:10.1016/j.freeradbiomed.2004.08.011
Original Contribution FUNCTIONAL ASSOCIATION OF NOX1 WITH P22PHOX IN VASCULAR SMOOTH MUSCLE CELLS IBRAHIM R. HANNA,* LULA L. HILENSKI,* ANNA DIKALOVA,* YOSHIHIRO TANIYAMA,* SERGEY DIKALOV,* ALICIA LYLE,* MARK T. QUINN,y BERNARD LASSE` GUE,* and KATHY K. GRIENDLING* *Division of Cardiology, Department of Medicine, Emory University, Atlanta, GA 30322; and yDepartment of Veterinary Molecular Biology, Montana State University, Bozeman, MT 59717, USA (Received 19 May 2004; Revised 27 July 2004; Accepted 12 August 2004) Available online 27 August 2004
Abstract — The vascular NAD(P)H oxidases constitute important sources of ROS in the vessel wall and have been implicated in vascular disease. Vascular smooth muscle cells (VSMCs) from conduit arteries express two gp91phox homologs, Nox1 and Nox4, of which Nox1 is agonist-sensitive. Because p22phox has been shown to be functionally important in vascular cells stimulated with vasoactive hormones, the relationship of Nox1 and p22phox was investigated in VSMCs from rat and human aortas. Coimmunoprecipitation studies demonstrated that p22phox and hemagglutinintagged Nox1 associate in unstimulated VSMCs. These findings were confirmed by confocal microscopy, showing colocalization of the two proteins in their native states in the plasma membrane and submembrane areas of the cell. NADPH-driven superoxide production, as measured by electron spin resonance using 1-hydroxy-3-carboxypyrrolidine as a spin probe, is dependent on the coexpression of both subunits, suggesting the importance of the association for the functional integrity of the enzyme. These results indicate that in contrast to the neutrophil enzyme, VSMCs can use Nox1 rather than gp91phox as a catalytic center in the p22phox-based oxidase and that these two proteins are preassembled at or near the plasma membrane and submembrane vesicular structures in unstimulated cells. D 2004 Elsevier Inc. All rights reserved. Keywords — NAD(P)H oxidase, Nox, Vascular smooth muscle, p22phox, Superoxide, Reactive oxygen species, Free radicals
G-protein Rac translocate to the membrane and assemble with cytochrome b558, resulting in enzyme activation and burst production of ROS. The h-chain of the enzyme, gp91phox, comprises the electron transfer activity, whereas p22phox acts as a stabilizing and regulatory subunit [10]. The NAD(P)H oxidase subunits expressed in vascular cells vary depending on cell type and vessel size. Vascular smooth muscle cells (VSMCs) derived from large conduit vessels possess a functional p22phox-containing flavin oxidase despite the absence of the catalytic center gp91phox [10]. In recent years, novel structural homologs of gp91phox have been identified in VSMCs, namely Nox1 and Nox4 [11,12]. Nox1 mediates NAD(P)Hdependent production of ROS in rat aortic smooth muscle cells (RASMs) stimulated with angiotensin II and PDGF [12], whereas the function of Nox4 in vascular cells remains unclear. Current research is centered on identifying the binding partners of these catalytic subunits, based
INTRODUCTION
Reactive oxygen species (ROS) play a central role in vascular pathobiology. They are produced by and affect all cell types in the vessel wall [1–5]. A major source of these highly reactive molecules is the NAD(P)H oxidase family of enzymes distributed throughout the artery. The composite structures of these oxidases have been modeled after the well-studied neutrophil enzyme, with some important differences [6]. The phagocytic oxidase is composed of at least five subunits, with two transmembrane proteins, p22phox and gp91phox, that associate in a one-to-one stoichiometry to form cytochrome b558 [7–9]. Upon neutrophil stimulation, the three cytosolic components p47phox, p67phox, and the small
Address correspondence to: Kathy K. Griendling, Division of Cardiology, Emory University, 319 WMB, 1639 Pierce Drive, Atlanta, GA 30322, USA. Fax: (404) 727 3585; E-mail: [email protected]. 1542
Association of Nox1 and p22phox
on the assumption that the holoenyzmes will be assembled similarly to the neutrophil oxidase. The 65 kDa protein Nox1 shares 56% amino acid sequence identity with gp91phox, with a similar hydrophobicity profile, and conserved NAD(P)H, flavin, and heme binding sites [11]. Recent work has shown that Nox1 activity is enhanced by interaction with two cytosolic factors, NoxO1 and NoxA1, which have significant homology to p47phox and p67phox, respectively [13–15]. Although an interaction of Nox1 with p22phox is commonly assumed, experimental proof of such an association is lacking. Takeya et al. [14] showed that transfection of p22phox enhances Nox1-mediated S superoxide (O2 ) production in CHO cells, but it is unclear whether this occurs because of a functional interaction between Nox1 and p22phox or through a p22phox-mediated stabilization of Nox1, independent of a direct association between the two proteins. This latter point has not been addressed experimentally, but is of vital importance in cells for which Nox1 represents the endogenous oxidase. VSMCs represent an ideal system in which to test this concept, because both Nox1 and p22phox have been demonstrated to participate in NADPH oxidase activity [12,16]. In this study, we hypothesized that Nox1 and p22phox interact to form a functional catalytic center for NADPH oxidase activity in VSMCs. EXPERIMENTAL PROCEDURES
Cell culture RASMs were isolated from rat aortas by enzymatic digestion as previously described [17]. Cells were grown in Dulbecco’s modified Eagle’s medium (DME) containing 4.5 g/l glucose, 2 mM glutamine, 100 U/ml penicillin, and 100 Ag/ml streptomycin and supplemented with 10% calf serum. Cells between passages 6 and 14 were used in experiments. In some experiments, RASMs stably expressing a 465 bp fragment of antisense p22phox were used. These cells have a complete ablation of p22phox expression [16]. Human aortic smooth muscle cells (HASMs) were obtained from Clonetics and cultured in SmGM-2 medium (Clonetics) supplemented with gentamycin, insulin (5 Ag/ml), human recombinant EGF (0.5 ng/ml), hFGF-B (2 ng/ml), and 5% fetal bovine serum. Cells between passages 7 and 10 were used in experiments. Adenoviruses The AdEasy System [18], which contains the green fluorescent protein (GFP) gene, was used to prepare viruses with no additional insert (AdGFP), hemagglutinin (HA)-tagged Nox1 (AdNoxHA), V5-tagged p22phox (Adp22phoxV5), or Nox1 (AdNox1).
1543
The addition of the small hemagglutinin tag to the carboxyl terminal of Nox1 was designed to minimize interference with function while avoiding the possibility of loss of the tag due to propeptide cleavage. To prepare HA-tagged Nox1, rat Nox1 was cloned from a RASM cDNA library. The HA tag was added at the 3V end of the coding region using PCR with the following primers: Forward, AAAGGTACCAGCATCTGCTTTGTGCCTG; Reverse, CCCCAAGCTTCAGGCGTAGTCTGGGACGTCGTATGGGTAGAACGTTTCTTTGTTGAAGTAGAATTGAACCT. The reverse primer includes the 3V nox1 coding region and the HA tag (corresponding protein sequence YPYDVPDYA), followed by a stop codon. The 1.5 kb PCR fragment was cloned into pAdTrackCMV and sequenced. V5-tagged p22phox was prepared by first amplifying the full coding region of rat p22phox with the following primers: Forward, CGCGGATCCGCATGGGGCAGATCGAGTGGGC; Reverse, CCGGAATTCCACGACCTCATCTGTCACTGGAAT. This reverse primer was designed to remove the stop codon and replace it with an EcoRI site for cloning. The 598 bp PCR product was first cloned into pcDNA3.1/V5-His to generate a fusion protein of p22phox tagged at the carboxyl terminus with V5 and 6His and sequenced before subcloning into pAdTrack CMV. Cultured RASMs were exposed to viral particles in DME with 0.1% CS for 12 h, washed, and incubated for 48 to 72 h in the same medium without virus before additional experimentation. Cell lysis and protein extraction Adenovirus-infected RASMs were washed and scraped off culture plates in ice-cold phosphate-buffered saline (PBS) and then centrifuged at 800 g for 10 min at 48C. The cellular pellet was resuspended in 500 Al of lysis buffer containing 38 mM Hepes, 38 mM NaCl, 3.8 mM EDTA, 7.6 mM Na pyrophosphate, 38 mM NaF, and 0.76 mM Na orthovanadate, supplemented with the protease inhibitors PMSF, leupeptin, and aprotinin to final concentrations of 1 mM, 10 Ag/ml, and 10 Ag/ml, respectively. The cell suspension was sonicated at 10 W for 10 s (using Microson 2425 from Misonix, Inc., Farmingdale, NY, USA) and proteins were extracted with 2% Triton X-100. The detergent-soluble fraction was separated by centrifugation at 20,000 g for 20 min at 48C. Immunoblotting Protein concentration in the Triton X-100 soluble fraction was measured using the Bradford assay, and 50 Ag of total protein was loaded in each well after being boiled in Laemmli buffer. After separation by SDS–PAGE and transfer to nitrocellulose membranes, signals were detected with rabbit anti-HA antibody
1544
I. R. Hanna et al.
(Santa Cruz Biotechnologies), rabbit anti-p22phox antibody (R3179) [19], or rabbit anti-V5 (Medical and Biological Laboratories, Japan) and visualized with enhanced chemiluminescence. Activated rat macrophage protein extracts were used as positive controls for p22phox signals. Coimmunoprecipitation Triton X-100 soluble protein extract (1.5 mg) in a final volume of 1 ml lysis buffer was used. Proteins were immunoprecipitated with mouse monoclonal antip22phox (m44.1) [20] or mouse monoclonal anti-HA (Covance/BAbCO) antibodies (3 Ag/mg) in the presence of protein-L agarose (Santa Cruz Biotechnologies) overnight at 48C. Immune complex-bound beads were washed three times in washing buffer (43 mM Hepes, 43 mM NaCl, 4.3 mM EDTA, 8.6 mM Na pyrophosphate, 4.3 mM NaF, 0.86 mM Na orthovanadate, 1 mM PMSF, 10 Ag/ml aprotinin, 10 Ag/ml leupeptin, and 1% Triton X-100), centrifuged, and resuspended in SDS sample buffer. Samples were boiled for 10 min and proteins separated by SDS–PAGE for immunoblotting using the antibodies specified in the figure legends. Immunofluorescence confocal microscopy HASMs grown on coverslips were washed with icecold PBS and fixed in 4% paraformaldehyde in PBS at room temperature for 10 min. After being washed, cells were permeabilized in 0.05% Triton X-100 in PBS for 5 min and rinsed sequentially in PBS, 50 mM NH4Cl, and PBS for 10 min each. After incubating 1 h in blocking buffer (3% bovine serum albumin in PBS (PBS/ BSA)), the cells were incubated with primary antibodies (rabbit anti-p22phox (R5554)) [21] and goat anti-Nox1 (Santa Cruz Biotechnologies) for 1 h. Samples were rinsed three times in PBS/BSA and then incubated for 1 h with corresponding secondary antibodies conjugated to either rhodamine red X (anti-goat) or FITC (anti-rabbit) (Jackson ImmunoResearch Laboratories). Cells on coverslips were mounted onto glass slides in Vectashield (Vector Laboratories, Burlingame, CA, USA) and FITC and rhodamine red X (RRX) images were scanned with the multitracking mode on a Zeiss LSM 510 using the 488- and 543-nm laser lines for the detection of FITC and RRX, respectively. Controls with no primary antibody showed no fluorescence labeling and single-label controls were performed in double-labeling experiments. Antibody specificity was confirmed using blocking peptides [22].
S Measurement of NADPH-dependent O 2
production
Membrane samples from RASMs were prepared as described previously [11]. RASMs were harvested, washed twice with ice-cold PBS, scraped, centrifuged at
400 g (10 min), and resuspended in 1 ml of lysis buffer (50 mM phosphate (treated for 2 h with 5 g/100 ml Chelex100 and filtered) containing the protease inhibitors aprotinin (10 Ag/ml), leupeptin (0.5 Ag/ml), pepstatin (0.7 Ag/ml), and PMSF (0.5 mM) (pH 7.4)). Cells were sonicated (power 4 W, using Microson 2425 from Misonix, Inc.) for 10 s on ice and centrifuged at 28,000 g for 15 min at 48C. The membrane pellet was resuspended in 150 Al of lysis buffer and protein concentration was measured using the Bradford method. Ten micrograms of protein was added to 1 mM 1hydroxy-3-carboxypyrrolidine (CPH), 200 AM NADPH, and 0.1 mM diethylenetriaminepentaacetic acid in a total volume of 100 Al of Chelex-treated PBS. Hydroxylamine spin probes such as CPH provide quantitative measureS ments of O2 radicals with high sensitivity [23]. ESR spectroscopy was used for quantitative measurements of O2S production. In duplicate samples, NADPH was omitted. Superoxide formation was assayed as NADPHdependent, superoxide dismutase (SOD)-inhibitable forS mation of 3-carboxyproxyl (CP ). Samples were placed in 50 Al glass capillaries (Corning, New York, NY). The ESR spectra were recorded using an EMX ESR spectrometer (Bruker) and a super-high-Q microwave cavity. The ESR instrumental settings were as follows: field sweep 50 G, microwave frequency 9.78 GHz, microwave power 20 mW, modulation amplitude 2 G, conversion time 656 ms, time constant 656 ms, 512 points resolution, and receiver gain 1 105. Kinetics were recorded using a 1312 ms conversion time and a 5248 ms time constant by monitoring the ESR amplitude of the low-field component S of the ESR spectrum of CP . It is important to note that S both basal NADPH-dependent CP production in memS branes and CP production in response to transfection of Nox1 and p22phox were inhibited by 95–98% by SOD (50 U/ml) added directly to the sample. Similar results were found after incubating the samples with the flavin oxidase inhibitor diphenylene iodonium (DPI; 50 AM) for 30 min on ice (96 F 3% inhibition). Statistics Results are expressed as means F SE. Statistical significance was assessed by ANOVA. A value of p b .05 was considered to be statistically significant. RESULTS
Coimmunoprecipitation of p22phox and Nox1 In order to study the association between Nox1 and p22phox in VSMCs, and in the absence of an immunoprecipitating antibody for Nox1, we expressed HA-tagged Nox1. As previously demonstrated [12], infection of RASMs with HA-tagged Nox1 led to the expression of 65 and 50 kDa protein products (Fig. 1, top). The
Association of Nox1 and p22phox
1545
p22phox in lysates from AdNoxHA-infected cells, but not in cells infected with AdGFP. Specificity was confirmed using isotype-matched mouse IgG as a negative control for immunoprecipitation. No HA signal was detected with IgG alone (data not shown). In the reverse experiment, p22phox coprecipitated with Nox1-HA only in RASM lysates expressing Nox1-HA (Fig. 1, bottom). These experiments strongly suggest that Nox1 and p22phox can physically interact in VSMCs. Colocalization of Nox1 and p22phox by confocal microscopy
Fig. 1. Coimmunoprecipitation of p22phox and Nox1-HA. Triton X-100 soluble lysates from RASMs infected with AdGFP or AdNoxHA were used for immunoprecipitation. Top: Mouse anti-p22phox antibodies (m44.1) were used for immunoprecipitation and rabbit anti-HA antibodies or rabbit anti-p22phox (R3179) antibodies were used for immunoblotting. In this case, p22phox runs as a dimer. Bottom: The reverse experiment was performed. Immunoprecipitation was carried out using mouse anti-HA antibodies, and immunoblotting was performed using rabbit anti-p22phox (R3179) or anti-HA antibodies. Immunoblots are representative of three identical experiments. LB, lysis buffer.
additional middle band is nonspecific, as determined using nox1 antisense [12]. Proteins from infected RASMs were precipitated with anti-p22phox (m44.1, which has been previously shown to specifically immunoprecipitate p22phox [20,24–28]) and blotted for Nox1-HA. As shown in Fig. 1, top, Nox1-HA coimmunoprecipitated with
Because coimmunoprecipitation disrupts the integrity of the membrane and overexpression can alter protein localization, the association of endogenous Nox1 and p22phox in VSMCs was studied using immunofluorescence and confocal microscopy to localize the two proteins. Fig. 2 shows that Nox1 and p22phox colocalize to the plasma membrane and to vesicle-like structures in proximity to the cell membrane. Interestingly, this colocalization pattern was seen in unstimulated cells, suggesting that the two proteins associate in resting VSMCs, as is true of the binding of gp91phox and p22phox in the membrane of resting neutrophils.
S
NAD(P)H-dependent O2 production in 0VSMC membranes requires coexpression of p22phox and Nox1 To investigate whether the association of Nox1 and S p22phox is required for O2 production, we took advantage of RASMs stably transfected with p22phox antisense cDNA that completely lack p22phox expression ([16] and Fig. 3A). These cells were infected with AdGFP, Adp22phoxV5, AdNoxHA, or both AdNoxHA and Adp22phoxV5 and analyzed for protein expression S (Fig. 3B) and O2 production (Figs. 3D and 3E). Importantly, infection of RASMs with AdNox1 or
Fig. 2. Colocalization of p22phox and Nox1 using confocal microscopy. HASMs were fixed and then permeabilized using Triton X-100. Goat anti-Nox1 and rabbit anti-p22phox (R5554) antibodies were used for the detection of native Nox1 and p22phox proteins, respectively. Rhodamine red X-labeled anti-goat and FITC-labeled anti-rabbit antibodies were then added. Signals were detected by immunofluorescence confocal microscopy. Left: Nox1 (red) localizes to the plasma membrane and to submembrane vesicle-like structures. Middle: p22phox (green) shows a more diffuse pattern of distribution including the plasma membrane and submembrane area. Right: Colocalization of the two proteins is detected as the superimposition of red and green signals with resulting yellow-orange color. Nox1 and p22phox colocalize in the resting cells at the level of the plasma membrane and in vesicular structures (arrows).
1546
I. R. Hanna et al.
AdNoxHA at similar multiplicities of infection (m.o.i.) resulted in a comparable increase in O2S production over baseline (Fig. 3C), indicating that the addition of the HA tag to the carboxyl terminal of Nox1 did not alter protein function. As shown in Fig. 3B, incubation of RASMs with these adenoviruses resulted in appropriate protein expression. Of interest, coexpression of Nox1 and p22phox seemed to enhance the expression of each individual protein, suggesting that, like gp91phox and p22phox in neutrophils [29], the two proteins might stabilize each other. The expression of either Nox1 or p22phox alone did not result in a significant increase in S O2 production over baseline (Figs. 3D and 3E). S Interestingly, basal O2 production is both SOD- and DPI-inhibitable (data not shown), suggesting that an S additional oxidase is responsible for constitutive O2 formation in these cells. In contrast, co-infection with both AdNoxHA and Adp22phoxV5 resulted in a S significant 1.7-fold increase in O2 production (Figs. 3D and 3E), which was completely abolished by SOD. This suggests that both subunits are required for NAD(P)H oxidase activity. Interestingly, the reintroduction of p22phox into p22phox-deficient cells did not S result in an increase in O2 production, perhaps because Nox1 expression is low in unstimulated cells [12]. DISCUSSION
The vascular NAD(P)H oxidases modulate the function of all cell types in the vessel wall, mediating a wide range of processes leading to endothelial dysfunction, vascular smooth muscle proliferation and migration, LDL oxidation, and matrix degradation [30–33]. The
S
Fig. 3. NADPH-dependent O2 production in RASM membranes requires the coexpression of p22phox and Nox1. (A) Western blot showing the absence of p22phox protein expression in p22phoxdeficient RASMs. ASp22, RASMs permanently transfected with antisense p22phox [16]; vector, RASMs permanently transfected with empty vector. (B) p22phox-deficient RASMs were infected with AdGFP, AdNoxHA, Adp22phoxV5, or both AdNoxHA and Adp22phoxV5. Western blot shows appropriate expression of tagged proteins. (C) Superoxide production measured by ESR, using NADPH as the electron donor and CPH as a spin probe, in membrane fractions of wildtype RASMs either uninfected (Con) or infected with AdGFP alone, AdNox1, or AdNoxHA. *p b .05 compared to control. (D) Representative tracing of superoxide production measured in membrane fractions by ESR, using NADPH as the electron donor and CPH as a spin probe, in p22phox-deficient RASMs treated as described for B. The ability of SOD (50 U/ml) to inhibit CP production in AdNoxHAand Adp22phoxV5-infected cells is shown in the lower trace as indicated. Inhibition by SOD was nearly complete in all samples. Inset shows typical ESR spectrum of CP. (E) Bar graphs representing averaged data, measured over 10 min, from experiments such as that shown in D. Results are the means F SE of three independent experiments. *p b .05 compared to AdGFP-infected p22phox-deficient RASMs.
Association of Nox1 and p22phox
structural basis of vascular NAD(P)H oxidase activity is poorly defined. One unique feature of nonphagocytic cells compared to neutrophils is the constitutive activity present in resting cells. In this study, using two different experimental designs, we show for the first time that Nox1 and p22phox associate in the plasma membrane and in intracellular vesicle-like structures in the resting state of aortic VSMCs. Moreover, we show the interaction to be necessary for oxidase function, as quantified by O2S production using NADPH as an electron donor. Using coimmunoprecipitation techniques in VSMCs derived from rat aortas, we show that p22phox coimmunoprecipitated with HA-tagged Nox1, regardless of whether proteins were precipitated with anti-p22phox or anti-HA antibodies. It is conceivable that the overexpression of a specific protein could modify its interaction with other proteins in the cell, resulting in otherwise nonphysiologic associations. In our experiments, the viral load added to RASMs in the coimmunoprecipitation reactions corresponded to the m.o.i. used for the measurement of membrane NAD(P)H-dependent S O2 production. In the latter experiments, the overS expression of Nox1 resulted in a 72% increase in O2 production over baseline as measured by ESR, a value that is significantly lower than the 101% increase in ESR signal observed in membranes of RASMs exposed to angiotensin II, which upregulates endogenous Nox1 [34]. As such, it is unlikely that the overexpression of Nox1HA at the m.o.i. used resulted in a supraphysiologic level of Nox1-HA protein, forcing an artifactual interaction with p22phox. The inherent limitations of coimmunoprecipitation reactions include the disruption of cell compartments and the creation of artificial molecular interactions otherwise prevented by the segregation of different proteins to different areas in the cell. To address this possible issue, we also studied the localization of Nox1 and p22phox in their native state in intact HASMs using immunofluorescent probes and confocal microscopy. In these experiments, endogenous Nox1 colocalized with endogenous p22phox in a pattern suggestive of plasma membrane and vesicular distribution. Interestingly, this association was demonstrated in the resting state, an observation that is true for the endothelial NADPH oxidase (composed of gp91phox and p22phox) as well [35]. The pattern of distribution of p22phox in HASMs overlapped, but was not restricted to, that of Nox1, raising the possibility of an interaction between p22phox and other members of the Nox family, namely Nox4. We have previously shown that different vascular NAD(P)H oxidases localize to different areas of the cell [22], suggesting that their activities are modulated by different stimuli, resulting in localized changes in ROS production within the cell.
1547
The functional implication of the association between Nox1 and p22phox was studied in the membrane fraction of RASMs lacking p22phox, after the overexpression of one or both proteins. In the absence of S p22phox, overexpression of Nox1 did not increase O2 production over that in vector-infected cells. However, overexpression of Nox1 in combination with the reintroduction of p22phox resulted in a significant S increase in O2 production in these cells, similar to results found in CHO cells transfected with Nox1, p22phox, and the cytosolic subunits [14]. Interestingly, restitution of p22phox expression in resting RASMs was S not accompanied by a change in O2 production, consistent with the notion that unstimulated RASMs express very little Nox1 [12] and that growth-arrested S cells have minimal O2 production. Moreover, in wildtype RASMs, overexpression of Nox1 or both Nox1 and p22phox resulted in increased O2S production, whereas p22phox overexpression did not (unpublished observations). This suggests that Nox1 may act as a rateS limiting factor in NADPH-dependent O2 production in these cells. Based on current knowledge about the phagocyte flavin oxidase, a number of roles can be postulated for p22phox, including stabilization of the catalytic subunit Nox1 or association with the cytosolic components of the oxidase to form the multisubunit functional enzyme. Our data are supportive of both possibilities. As seen in Fig. 3B, coexpression of Nox1 and p22phox enhanced the expression of both proteins, compared to expression of either alone. Figs. 3D and 3E clearly show that p22phox S is necessary for O2 production, as discussed above. Previously, we have shown that either p22phox or nox1 antisense prevented the ability of angiotensin II to S increase O2 production [12,16]. However, because VSMCs express more than one catalytic subunit (Nox1, 4, and 5), it was unclear how these two observations were related. The present data clearly show that Nox1 and p22phox interact to form a functional oxidase and that S this interaction is capable of producing O2 even in the absence of a stimulus. Whether this constitutive activity can be increased by interaction with specific cytosolic subunits remains to be determined. In conclusion, as predicted based on homology with gp91phox, we show that Nox1 associates with p22phox in the resting state of VSMCs to form a functional NADPHdependent oxidase. We expect this finding to simplify the characterization of the remaining structural components of this important enzyme complex, leading to a more thorough understanding of its regulation in response to agonist/antagonist substances. The differential regulation of the various vascular NAD(P)H oxidases, their patterns of distribution, and their effects on target cells are fertile areas of future investigation that will expand our under-
1548
I. R. Hanna et al.
standing of this new class of molecules, with the hope that targeted pharmacological interventions will allow selective modulation of one enzyme system, without affecting critical components of the body’s innate immunity. Acknowledgments — This work was supported by Postdoctoral Fellowship AHA Grant 0120351B to Dr. Hanna and NIH Grants HL38206, HL58000, and AR42426. We thank Dr. Margaret Tarpey and QiQin Vin for their help in preparing the adenoviruses.
REFERENCES [1] Miller, F. J. J.; Gutterman, D. D.; Rios, C. D.; Heistad, D. D.; Davidson, B. L. Superoxide production in vascular smooth muscle contributes to oxidative stress and impaired relaxation in atherosclerosis. Circulation Res. 82:1298 – 1305; 1998. [2] Mohazzab-H, K. M.; Kaminski, P. M.; Wolin, M. S. NADH oxidoreductase is the major source of superoxide anion in bovine coronary artery endothelium. Am. J. Physiol. 266:H2568 – H2572; 1994. [3] Pagano, P. J.; Clark, J. K.; Cifuentes-Pagano, M. E.; Clark, S. M.; Callis, G. M.; Quinn, M. T. Localization of a constitutively active, phagocyte-like NADPH oxidase in rabbit aortic adventitia: enhancement by angiotensin II. Proc. Natl. Acad. Sci. USA 94; 1997. [4] Rajagopalan, S.; Kurz, S.; Mqnzel, T.; Tarpey, M.; Freeman, B. A.; Griendling, K. K.; Harrison, D. G. 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. 97:1916 – 1923; 1996. [5] Griendling, K. K.; Sorescu, D.; Ushio-Fukai, M. NAD(P)H oxidase. role in cardiovascular biology and disease. Circ. Res. 86: 494 – 501; 2000. [6] Lassegue, B.; Clempus, R. E. Vascular NAD(P)H oxidases: specific features, expression, and regulation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 285:R277 – R297; 2003. [7] Parkos, C. A.; Allen, R. A.; Cochrane, C. G.; Jesaitis, A. J. Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J. Clin. Invest. 80:732 – 742; 1987. [8] Wallach, T. M.; Segal, A. W. Stoichiometry of the subunits of flavocytochrome b558 of the NADPH oxidase of phagocytes. Biochem. J. 320:33 – 38; 1996. [9] Babior, B. M.; Lambeth, J. D.; Nauseef, W. M. The neutrophil NADPH oxidase. Arch. Biochem. Biophys. 397:342 – 344; 2002. [10] Hanna, I. R.; Taniyama, Y.; Szocs, K.; Rocic, P.; Griendling, K. K. NAD(P)H oxidase-derived reactive oxygen species as mediators of angiotensin II signaling. Antioxid. Redox Signaling 4:899 – 914; 2002. [11] Suh, Y.-A.; Arnold, R. S.; Lassegue, B.; Shi, J.; Xu, X.; Sorescu, D.; Chung, A. B.; Griendling, K. K.; Lambeth, J. D. Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79 – 82; 1999. [12] Lassegue, B.; Sorescu, D.; Szocs, K.; Akers, M.; Zhang, Y.; Grant, S. L.; Lambeth, J. D.; Griendling, K. K. 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. [13] Geiszt, M.; Lekstrom, K.; Witta, J.; Leto, T. L. 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. [14] Takeya, R.; Ueno, N.; Kami, K.; Taura, M.; Kohjima, M.; Izaki, H.; Nunoi, H.; Sumimoto, H. Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J. Biol. Chem. 278:25234 – 25246; 2003.
[15] Banfi, B.; Clark, R. A.; Steger, K.; Krause, K.-H. Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J. Biol. Chem. 278:3510 – 3513; 2003. [16] Ushio-Fukai, M.; Zafari, A. M.; Fukui, T.; Ishizaka, N.; Griendling, K. K. p22phox is a critical component of the superoxide-generating NADH/NADPH oxidase system and regulates angiotensin IIinduced hypertrophy in vascular smooth muscle cells. J. Biol. Chem. 271:23317 – 23321; 1996. [17] Griendling, K. K.; Taubman, M. B.; Akers, M.; Mendlowitz, M.; Alexander, R. W. Characterization of phosphatidylinositol-specific phospholipase C from cultured vascular smooth muscle cells. J. Biol. Chem. 266:15498 – 15504; 1991. [18] He, T. C.; Zhou, S.; da Costa, L. T.; Yu, J.; Kinzler, K. W.; Vogelstein, B. A simplified system for generating recombinant adenoviruses. Proc. Natl. Acad. Sci. USA 95:2509 – 2514; 1998. [19] Jesaitis, A. J.; Buescher, E. S.; Harrison, D.; Quinn, M. T.; Parkos, C. A.; Livesey, S.; Linner, J. Ultrastructural localization of cytochrome b in the membranes of resting and phagocytosing human granulocytes. J. Clin. Invest. 85:821 – 835; 1990. [20] Burritt, J. B.; Quinn, M. T.; Jutila, M. A.; Bond, C. W.; Jesaitis, A. J. Topological mapping of neutrophil cytochrome b epitopes with phage-display libraries. J. Biol. Chem. 270:16974 – 16980; 1995. [21] Quinn, M. T.; Mullen, M. L.; Jesaitis, A. J. Human neutrophil cytochrome b contains multiple hemes: evidence for heme associated with both subunits. J. Biol. Chem. 267:7303 – 7309; 1992. [22] Hilenski, L. L.; Clempus, R. E.; Quinn, M. T.; Lambeth, J. D.; Griendling, K. K. Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler. Thromb. Vasc. Biol. 24:677 – 683; 2004. [23] Dikalov, S. I.; Dikalova, A. E.; Mason, R. P. New non-invasive diagnostic tool for inflammation-induced oxidative stress using electron spin resonance spectroscopy and cyclic hydroxylamine. Arch. Biochem. Biophys. 402:218 – 226; 2001. [24] Touyz, R. M.; Chen, X.; Tabet, F.; Yao, G.; He, G.; Quinn, M. T.; Pagano, P. J.; Schiffrin, E. L. Expression of a functionally active gp91phox-containing neutrophil-type NAD(P)H oxidase in smooth muscle cells from human resistance arteries. Circ. Res. 90:1205 – 1213; 2002. [25] Burritt, J. B.; DeLeo, F. R.; McDonald, C. L.; Prigge, J. R.; Dinauer, M. C.; Nakamura, M.; Nauseef, W. M.; Jesaitis, A. J. Phage display epitope mapping of human neutrophil flavocytochrome b558. J. Biol. Chem. 276:2053 – 2061; 2001. [26] DeLeo, F. R.; Burritt, J. B.; Yu, L.; Jesaitis, A. J.; Dinauer, M. C.; Nauseef, W. M. Processing and maturation of flavocytochrome b558 include incorporation of heme as pre-requisite for heterodimer assembly. J. Biol. Chem. 275:13986 – 13993; 2000. [27] Regier, D. S.; Waite, K. A.; Wallin, R.; McPhail, L. C. A phosphatidic acid-activated protein kinase and conventional protein kinase C isoforms phosphorylate p22phox, an NADPH oxidase component. J. Biol. Chem. 274:36601 – 36608; 1999. [28] Regier, D. S.; Greene, D. G.; Sergeant, S.; Jesaitis, A. J.; McPhail, L. C. Phosphorylation of p22phox is mediated by phospholipase D-dependent and -independent mechanisms. J. Biol. Chem. 275:28406 – 28412; 2000. [29] Yu, L.; Zhen, L.; Dinauer, M. C. Biosynthesis of the phagocyte NADPH oxidase cytochrome b558: role of heme incorporation and heterodimer formation in maturation and stability of gp91phox and p22phox subunits. J. Biol. Chem. 272:27288 – 27294; 1997. [30] Mugge, A.; Elwell, J. H.; Peterson, T. E.; Hofmeyer, T. G.; Heistad, D. D.; Harrison, D. G. Chronic treatment with polyethylene-glycolated superoxide dismutase partially restores endothelium-dependent vascular relaxations in cholesterol-fed rabbits. Circ. Res. 69:1293 – 1300; 1991. [31] Aviram, M.; Rosenbalt, M.; Etzioni, A.; Levy, R. Activation of NADPH oxidase required for macrophage-mediated oxidation of low-density lipoprotein. Metabolism 45:1069 – 1079; 1996. [32] Rajagopalan, S.; Meng, X. P.; Ramasamy, S.; Harrison, D. G.; Galis, Z. S. Reactive oxygen species produced by macrophagederived foam cells regulate the activity of vascular matrix metalloproteinases in vitro. J. Clin. Invest. 98:2572 – 2579; 1996.
Association of Nox1 and p22phox [33] Szocs, K.; Lassegue, B.; Sorescu, D.; Hilenski, L. L.; Valppu, L.; Couse, T. L.; Wilcox, J. N.; Lambeth, J. D.; Griendling, K. K. Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler. Thromb. Vasc. Biol. 22:21 – 27; 2002. [34] Sorescu, D.; Somers, M. J.; Lassegue, B.; Grant, S.; Harrison, D. G.; Griendling, K. K. Electron spin resonance characterization of the NAD(P)H oxidase in vascular smooth muscle cells. Free Radic. Biol. Med. 30:603 – 612; 2001. [35] Li, J. M.; Shah, A. M. Intracellular localization and preassembly of the NADPH oxidase complex in cultured endothelial cells. J. Biol. Chem. 277:19952 – 19960; 2002. ABBREVIATIONS
AdGFP — adenovirus expressing green fluorescent protein only AdNox1 — adenovirus expressing Nox1 AdNoxHA — adenovirus expressing hemagglutinin-tagged Nox1 Adp22phoxV5 — adenovirus expressing V5-tagged p22phox
BSA — bovine serum albumin CPH — 1-hydroxy-3-carboxypyrrolidine DME — Dulbecco’s modified Eagle’s medium DPI — diphenylene iodonium ESR — electron spin resonance GFP — green fluorescent protein HA — hemagglutinin HASMs — human aortic smooth muscle cells m.o.i. — multiplicity of infection Nox — NAD(P)H oxidase S O2 — superoxide PBS — phosphate-buffered saline RASMs — rat aortic smooth muscle cells ROS — reactive oxygen species SOD — superoxide dismutase VSMCs — vascular smooth muscle cells
1549
doi:10.1016/j.freeradbiomed.2004.08.011
Original Contribution FUNCTIONAL ASSOCIATION OF NOX1 WITH P22PHOX IN VASCULAR SMOOTH MUSCLE CELLS IBRAHIM R. HANNA,* LULA L. HILENSKI,* ANNA DIKALOVA,* YOSHIHIRO TANIYAMA,* SERGEY DIKALOV,* ALICIA LYLE,* MARK T. QUINN,y BERNARD LASSE` GUE,* and KATHY K. GRIENDLING* *Division of Cardiology, Department of Medicine, Emory University, Atlanta, GA 30322; and yDepartment of Veterinary Molecular Biology, Montana State University, Bozeman, MT 59717, USA (Received 19 May 2004; Revised 27 July 2004; Accepted 12 August 2004) Available online 27 August 2004
Abstract — The vascular NAD(P)H oxidases constitute important sources of ROS in the vessel wall and have been implicated in vascular disease. Vascular smooth muscle cells (VSMCs) from conduit arteries express two gp91phox homologs, Nox1 and Nox4, of which Nox1 is agonist-sensitive. Because p22phox has been shown to be functionally important in vascular cells stimulated with vasoactive hormones, the relationship of Nox1 and p22phox was investigated in VSMCs from rat and human aortas. Coimmunoprecipitation studies demonstrated that p22phox and hemagglutinintagged Nox1 associate in unstimulated VSMCs. These findings were confirmed by confocal microscopy, showing colocalization of the two proteins in their native states in the plasma membrane and submembrane areas of the cell. NADPH-driven superoxide production, as measured by electron spin resonance using 1-hydroxy-3-carboxypyrrolidine as a spin probe, is dependent on the coexpression of both subunits, suggesting the importance of the association for the functional integrity of the enzyme. These results indicate that in contrast to the neutrophil enzyme, VSMCs can use Nox1 rather than gp91phox as a catalytic center in the p22phox-based oxidase and that these two proteins are preassembled at or near the plasma membrane and submembrane vesicular structures in unstimulated cells. D 2004 Elsevier Inc. All rights reserved. Keywords — NAD(P)H oxidase, Nox, Vascular smooth muscle, p22phox, Superoxide, Reactive oxygen species, Free radicals
G-protein Rac translocate to the membrane and assemble with cytochrome b558, resulting in enzyme activation and burst production of ROS. The h-chain of the enzyme, gp91phox, comprises the electron transfer activity, whereas p22phox acts as a stabilizing and regulatory subunit [10]. The NAD(P)H oxidase subunits expressed in vascular cells vary depending on cell type and vessel size. Vascular smooth muscle cells (VSMCs) derived from large conduit vessels possess a functional p22phox-containing flavin oxidase despite the absence of the catalytic center gp91phox [10]. In recent years, novel structural homologs of gp91phox have been identified in VSMCs, namely Nox1 and Nox4 [11,12]. Nox1 mediates NAD(P)Hdependent production of ROS in rat aortic smooth muscle cells (RASMs) stimulated with angiotensin II and PDGF [12], whereas the function of Nox4 in vascular cells remains unclear. Current research is centered on identifying the binding partners of these catalytic subunits, based
INTRODUCTION
Reactive oxygen species (ROS) play a central role in vascular pathobiology. They are produced by and affect all cell types in the vessel wall [1–5]. A major source of these highly reactive molecules is the NAD(P)H oxidase family of enzymes distributed throughout the artery. The composite structures of these oxidases have been modeled after the well-studied neutrophil enzyme, with some important differences [6]. The phagocytic oxidase is composed of at least five subunits, with two transmembrane proteins, p22phox and gp91phox, that associate in a one-to-one stoichiometry to form cytochrome b558 [7–9]. Upon neutrophil stimulation, the three cytosolic components p47phox, p67phox, and the small
Address correspondence to: Kathy K. Griendling, Division of Cardiology, Emory University, 319 WMB, 1639 Pierce Drive, Atlanta, GA 30322, USA. Fax: (404) 727 3585; E-mail: [email protected]. 1542
Association of Nox1 and p22phox
on the assumption that the holoenyzmes will be assembled similarly to the neutrophil oxidase. The 65 kDa protein Nox1 shares 56% amino acid sequence identity with gp91phox, with a similar hydrophobicity profile, and conserved NAD(P)H, flavin, and heme binding sites [11]. Recent work has shown that Nox1 activity is enhanced by interaction with two cytosolic factors, NoxO1 and NoxA1, which have significant homology to p47phox and p67phox, respectively [13–15]. Although an interaction of Nox1 with p22phox is commonly assumed, experimental proof of such an association is lacking. Takeya et al. [14] showed that transfection of p22phox enhances Nox1-mediated S superoxide (O2 ) production in CHO cells, but it is unclear whether this occurs because of a functional interaction between Nox1 and p22phox or through a p22phox-mediated stabilization of Nox1, independent of a direct association between the two proteins. This latter point has not been addressed experimentally, but is of vital importance in cells for which Nox1 represents the endogenous oxidase. VSMCs represent an ideal system in which to test this concept, because both Nox1 and p22phox have been demonstrated to participate in NADPH oxidase activity [12,16]. In this study, we hypothesized that Nox1 and p22phox interact to form a functional catalytic center for NADPH oxidase activity in VSMCs. EXPERIMENTAL PROCEDURES
Cell culture RASMs were isolated from rat aortas by enzymatic digestion as previously described [17]. Cells were grown in Dulbecco’s modified Eagle’s medium (DME) containing 4.5 g/l glucose, 2 mM glutamine, 100 U/ml penicillin, and 100 Ag/ml streptomycin and supplemented with 10% calf serum. Cells between passages 6 and 14 were used in experiments. In some experiments, RASMs stably expressing a 465 bp fragment of antisense p22phox were used. These cells have a complete ablation of p22phox expression [16]. Human aortic smooth muscle cells (HASMs) were obtained from Clonetics and cultured in SmGM-2 medium (Clonetics) supplemented with gentamycin, insulin (5 Ag/ml), human recombinant EGF (0.5 ng/ml), hFGF-B (2 ng/ml), and 5% fetal bovine serum. Cells between passages 7 and 10 were used in experiments. Adenoviruses The AdEasy System [18], which contains the green fluorescent protein (GFP) gene, was used to prepare viruses with no additional insert (AdGFP), hemagglutinin (HA)-tagged Nox1 (AdNoxHA), V5-tagged p22phox (Adp22phoxV5), or Nox1 (AdNox1).
1543
The addition of the small hemagglutinin tag to the carboxyl terminal of Nox1 was designed to minimize interference with function while avoiding the possibility of loss of the tag due to propeptide cleavage. To prepare HA-tagged Nox1, rat Nox1 was cloned from a RASM cDNA library. The HA tag was added at the 3V end of the coding region using PCR with the following primers: Forward, AAAGGTACCAGCATCTGCTTTGTGCCTG; Reverse, CCCCAAGCTTCAGGCGTAGTCTGGGACGTCGTATGGGTAGAACGTTTCTTTGTTGAAGTAGAATTGAACCT. The reverse primer includes the 3V nox1 coding region and the HA tag (corresponding protein sequence YPYDVPDYA), followed by a stop codon. The 1.5 kb PCR fragment was cloned into pAdTrackCMV and sequenced. V5-tagged p22phox was prepared by first amplifying the full coding region of rat p22phox with the following primers: Forward, CGCGGATCCGCATGGGGCAGATCGAGTGGGC; Reverse, CCGGAATTCCACGACCTCATCTGTCACTGGAAT. This reverse primer was designed to remove the stop codon and replace it with an EcoRI site for cloning. The 598 bp PCR product was first cloned into pcDNA3.1/V5-His to generate a fusion protein of p22phox tagged at the carboxyl terminus with V5 and 6His and sequenced before subcloning into pAdTrack CMV. Cultured RASMs were exposed to viral particles in DME with 0.1% CS for 12 h, washed, and incubated for 48 to 72 h in the same medium without virus before additional experimentation. Cell lysis and protein extraction Adenovirus-infected RASMs were washed and scraped off culture plates in ice-cold phosphate-buffered saline (PBS) and then centrifuged at 800 g for 10 min at 48C. The cellular pellet was resuspended in 500 Al of lysis buffer containing 38 mM Hepes, 38 mM NaCl, 3.8 mM EDTA, 7.6 mM Na pyrophosphate, 38 mM NaF, and 0.76 mM Na orthovanadate, supplemented with the protease inhibitors PMSF, leupeptin, and aprotinin to final concentrations of 1 mM, 10 Ag/ml, and 10 Ag/ml, respectively. The cell suspension was sonicated at 10 W for 10 s (using Microson 2425 from Misonix, Inc., Farmingdale, NY, USA) and proteins were extracted with 2% Triton X-100. The detergent-soluble fraction was separated by centrifugation at 20,000 g for 20 min at 48C. Immunoblotting Protein concentration in the Triton X-100 soluble fraction was measured using the Bradford assay, and 50 Ag of total protein was loaded in each well after being boiled in Laemmli buffer. After separation by SDS–PAGE and transfer to nitrocellulose membranes, signals were detected with rabbit anti-HA antibody
1544
I. R. Hanna et al.
(Santa Cruz Biotechnologies), rabbit anti-p22phox antibody (R3179) [19], or rabbit anti-V5 (Medical and Biological Laboratories, Japan) and visualized with enhanced chemiluminescence. Activated rat macrophage protein extracts were used as positive controls for p22phox signals. Coimmunoprecipitation Triton X-100 soluble protein extract (1.5 mg) in a final volume of 1 ml lysis buffer was used. Proteins were immunoprecipitated with mouse monoclonal antip22phox (m44.1) [20] or mouse monoclonal anti-HA (Covance/BAbCO) antibodies (3 Ag/mg) in the presence of protein-L agarose (Santa Cruz Biotechnologies) overnight at 48C. Immune complex-bound beads were washed three times in washing buffer (43 mM Hepes, 43 mM NaCl, 4.3 mM EDTA, 8.6 mM Na pyrophosphate, 4.3 mM NaF, 0.86 mM Na orthovanadate, 1 mM PMSF, 10 Ag/ml aprotinin, 10 Ag/ml leupeptin, and 1% Triton X-100), centrifuged, and resuspended in SDS sample buffer. Samples were boiled for 10 min and proteins separated by SDS–PAGE for immunoblotting using the antibodies specified in the figure legends. Immunofluorescence confocal microscopy HASMs grown on coverslips were washed with icecold PBS and fixed in 4% paraformaldehyde in PBS at room temperature for 10 min. After being washed, cells were permeabilized in 0.05% Triton X-100 in PBS for 5 min and rinsed sequentially in PBS, 50 mM NH4Cl, and PBS for 10 min each. After incubating 1 h in blocking buffer (3% bovine serum albumin in PBS (PBS/ BSA)), the cells were incubated with primary antibodies (rabbit anti-p22phox (R5554)) [21] and goat anti-Nox1 (Santa Cruz Biotechnologies) for 1 h. Samples were rinsed three times in PBS/BSA and then incubated for 1 h with corresponding secondary antibodies conjugated to either rhodamine red X (anti-goat) or FITC (anti-rabbit) (Jackson ImmunoResearch Laboratories). Cells on coverslips were mounted onto glass slides in Vectashield (Vector Laboratories, Burlingame, CA, USA) and FITC and rhodamine red X (RRX) images were scanned with the multitracking mode on a Zeiss LSM 510 using the 488- and 543-nm laser lines for the detection of FITC and RRX, respectively. Controls with no primary antibody showed no fluorescence labeling and single-label controls were performed in double-labeling experiments. Antibody specificity was confirmed using blocking peptides [22].
S Measurement of NADPH-dependent O 2
production
Membrane samples from RASMs were prepared as described previously [11]. RASMs were harvested, washed twice with ice-cold PBS, scraped, centrifuged at
400 g (10 min), and resuspended in 1 ml of lysis buffer (50 mM phosphate (treated for 2 h with 5 g/100 ml Chelex100 and filtered) containing the protease inhibitors aprotinin (10 Ag/ml), leupeptin (0.5 Ag/ml), pepstatin (0.7 Ag/ml), and PMSF (0.5 mM) (pH 7.4)). Cells were sonicated (power 4 W, using Microson 2425 from Misonix, Inc.) for 10 s on ice and centrifuged at 28,000 g for 15 min at 48C. The membrane pellet was resuspended in 150 Al of lysis buffer and protein concentration was measured using the Bradford method. Ten micrograms of protein was added to 1 mM 1hydroxy-3-carboxypyrrolidine (CPH), 200 AM NADPH, and 0.1 mM diethylenetriaminepentaacetic acid in a total volume of 100 Al of Chelex-treated PBS. Hydroxylamine spin probes such as CPH provide quantitative measureS ments of O2 radicals with high sensitivity [23]. ESR spectroscopy was used for quantitative measurements of O2S production. In duplicate samples, NADPH was omitted. Superoxide formation was assayed as NADPHdependent, superoxide dismutase (SOD)-inhibitable forS mation of 3-carboxyproxyl (CP ). Samples were placed in 50 Al glass capillaries (Corning, New York, NY). The ESR spectra were recorded using an EMX ESR spectrometer (Bruker) and a super-high-Q microwave cavity. The ESR instrumental settings were as follows: field sweep 50 G, microwave frequency 9.78 GHz, microwave power 20 mW, modulation amplitude 2 G, conversion time 656 ms, time constant 656 ms, 512 points resolution, and receiver gain 1 105. Kinetics were recorded using a 1312 ms conversion time and a 5248 ms time constant by monitoring the ESR amplitude of the low-field component S of the ESR spectrum of CP . It is important to note that S both basal NADPH-dependent CP production in memS branes and CP production in response to transfection of Nox1 and p22phox were inhibited by 95–98% by SOD (50 U/ml) added directly to the sample. Similar results were found after incubating the samples with the flavin oxidase inhibitor diphenylene iodonium (DPI; 50 AM) for 30 min on ice (96 F 3% inhibition). Statistics Results are expressed as means F SE. Statistical significance was assessed by ANOVA. A value of p b .05 was considered to be statistically significant. RESULTS
Coimmunoprecipitation of p22phox and Nox1 In order to study the association between Nox1 and p22phox in VSMCs, and in the absence of an immunoprecipitating antibody for Nox1, we expressed HA-tagged Nox1. As previously demonstrated [12], infection of RASMs with HA-tagged Nox1 led to the expression of 65 and 50 kDa protein products (Fig. 1, top). The
Association of Nox1 and p22phox
1545
p22phox in lysates from AdNoxHA-infected cells, but not in cells infected with AdGFP. Specificity was confirmed using isotype-matched mouse IgG as a negative control for immunoprecipitation. No HA signal was detected with IgG alone (data not shown). In the reverse experiment, p22phox coprecipitated with Nox1-HA only in RASM lysates expressing Nox1-HA (Fig. 1, bottom). These experiments strongly suggest that Nox1 and p22phox can physically interact in VSMCs. Colocalization of Nox1 and p22phox by confocal microscopy
Fig. 1. Coimmunoprecipitation of p22phox and Nox1-HA. Triton X-100 soluble lysates from RASMs infected with AdGFP or AdNoxHA were used for immunoprecipitation. Top: Mouse anti-p22phox antibodies (m44.1) were used for immunoprecipitation and rabbit anti-HA antibodies or rabbit anti-p22phox (R3179) antibodies were used for immunoblotting. In this case, p22phox runs as a dimer. Bottom: The reverse experiment was performed. Immunoprecipitation was carried out using mouse anti-HA antibodies, and immunoblotting was performed using rabbit anti-p22phox (R3179) or anti-HA antibodies. Immunoblots are representative of three identical experiments. LB, lysis buffer.
additional middle band is nonspecific, as determined using nox1 antisense [12]. Proteins from infected RASMs were precipitated with anti-p22phox (m44.1, which has been previously shown to specifically immunoprecipitate p22phox [20,24–28]) and blotted for Nox1-HA. As shown in Fig. 1, top, Nox1-HA coimmunoprecipitated with
Because coimmunoprecipitation disrupts the integrity of the membrane and overexpression can alter protein localization, the association of endogenous Nox1 and p22phox in VSMCs was studied using immunofluorescence and confocal microscopy to localize the two proteins. Fig. 2 shows that Nox1 and p22phox colocalize to the plasma membrane and to vesicle-like structures in proximity to the cell membrane. Interestingly, this colocalization pattern was seen in unstimulated cells, suggesting that the two proteins associate in resting VSMCs, as is true of the binding of gp91phox and p22phox in the membrane of resting neutrophils.
S
NAD(P)H-dependent O2 production in 0VSMC membranes requires coexpression of p22phox and Nox1 To investigate whether the association of Nox1 and S p22phox is required for O2 production, we took advantage of RASMs stably transfected with p22phox antisense cDNA that completely lack p22phox expression ([16] and Fig. 3A). These cells were infected with AdGFP, Adp22phoxV5, AdNoxHA, or both AdNoxHA and Adp22phoxV5 and analyzed for protein expression S (Fig. 3B) and O2 production (Figs. 3D and 3E). Importantly, infection of RASMs with AdNox1 or
Fig. 2. Colocalization of p22phox and Nox1 using confocal microscopy. HASMs were fixed and then permeabilized using Triton X-100. Goat anti-Nox1 and rabbit anti-p22phox (R5554) antibodies were used for the detection of native Nox1 and p22phox proteins, respectively. Rhodamine red X-labeled anti-goat and FITC-labeled anti-rabbit antibodies were then added. Signals were detected by immunofluorescence confocal microscopy. Left: Nox1 (red) localizes to the plasma membrane and to submembrane vesicle-like structures. Middle: p22phox (green) shows a more diffuse pattern of distribution including the plasma membrane and submembrane area. Right: Colocalization of the two proteins is detected as the superimposition of red and green signals with resulting yellow-orange color. Nox1 and p22phox colocalize in the resting cells at the level of the plasma membrane and in vesicular structures (arrows).
1546
I. R. Hanna et al.
AdNoxHA at similar multiplicities of infection (m.o.i.) resulted in a comparable increase in O2S production over baseline (Fig. 3C), indicating that the addition of the HA tag to the carboxyl terminal of Nox1 did not alter protein function. As shown in Fig. 3B, incubation of RASMs with these adenoviruses resulted in appropriate protein expression. Of interest, coexpression of Nox1 and p22phox seemed to enhance the expression of each individual protein, suggesting that, like gp91phox and p22phox in neutrophils [29], the two proteins might stabilize each other. The expression of either Nox1 or p22phox alone did not result in a significant increase in S O2 production over baseline (Figs. 3D and 3E). S Interestingly, basal O2 production is both SOD- and DPI-inhibitable (data not shown), suggesting that an S additional oxidase is responsible for constitutive O2 formation in these cells. In contrast, co-infection with both AdNoxHA and Adp22phoxV5 resulted in a S significant 1.7-fold increase in O2 production (Figs. 3D and 3E), which was completely abolished by SOD. This suggests that both subunits are required for NAD(P)H oxidase activity. Interestingly, the reintroduction of p22phox into p22phox-deficient cells did not S result in an increase in O2 production, perhaps because Nox1 expression is low in unstimulated cells [12]. DISCUSSION
The vascular NAD(P)H oxidases modulate the function of all cell types in the vessel wall, mediating a wide range of processes leading to endothelial dysfunction, vascular smooth muscle proliferation and migration, LDL oxidation, and matrix degradation [30–33]. The
S
Fig. 3. NADPH-dependent O2 production in RASM membranes requires the coexpression of p22phox and Nox1. (A) Western blot showing the absence of p22phox protein expression in p22phoxdeficient RASMs. ASp22, RASMs permanently transfected with antisense p22phox [16]; vector, RASMs permanently transfected with empty vector. (B) p22phox-deficient RASMs were infected with AdGFP, AdNoxHA, Adp22phoxV5, or both AdNoxHA and Adp22phoxV5. Western blot shows appropriate expression of tagged proteins. (C) Superoxide production measured by ESR, using NADPH as the electron donor and CPH as a spin probe, in membrane fractions of wildtype RASMs either uninfected (Con) or infected with AdGFP alone, AdNox1, or AdNoxHA. *p b .05 compared to control. (D) Representative tracing of superoxide production measured in membrane fractions by ESR, using NADPH as the electron donor and CPH as a spin probe, in p22phox-deficient RASMs treated as described for B. The ability of SOD (50 U/ml) to inhibit CP production in AdNoxHAand Adp22phoxV5-infected cells is shown in the lower trace as indicated. Inhibition by SOD was nearly complete in all samples. Inset shows typical ESR spectrum of CP. (E) Bar graphs representing averaged data, measured over 10 min, from experiments such as that shown in D. Results are the means F SE of three independent experiments. *p b .05 compared to AdGFP-infected p22phox-deficient RASMs.
Association of Nox1 and p22phox
structural basis of vascular NAD(P)H oxidase activity is poorly defined. One unique feature of nonphagocytic cells compared to neutrophils is the constitutive activity present in resting cells. In this study, using two different experimental designs, we show for the first time that Nox1 and p22phox associate in the plasma membrane and in intracellular vesicle-like structures in the resting state of aortic VSMCs. Moreover, we show the interaction to be necessary for oxidase function, as quantified by O2S production using NADPH as an electron donor. Using coimmunoprecipitation techniques in VSMCs derived from rat aortas, we show that p22phox coimmunoprecipitated with HA-tagged Nox1, regardless of whether proteins were precipitated with anti-p22phox or anti-HA antibodies. It is conceivable that the overexpression of a specific protein could modify its interaction with other proteins in the cell, resulting in otherwise nonphysiologic associations. In our experiments, the viral load added to RASMs in the coimmunoprecipitation reactions corresponded to the m.o.i. used for the measurement of membrane NAD(P)H-dependent S O2 production. In the latter experiments, the overS expression of Nox1 resulted in a 72% increase in O2 production over baseline as measured by ESR, a value that is significantly lower than the 101% increase in ESR signal observed in membranes of RASMs exposed to angiotensin II, which upregulates endogenous Nox1 [34]. As such, it is unlikely that the overexpression of Nox1HA at the m.o.i. used resulted in a supraphysiologic level of Nox1-HA protein, forcing an artifactual interaction with p22phox. The inherent limitations of coimmunoprecipitation reactions include the disruption of cell compartments and the creation of artificial molecular interactions otherwise prevented by the segregation of different proteins to different areas in the cell. To address this possible issue, we also studied the localization of Nox1 and p22phox in their native state in intact HASMs using immunofluorescent probes and confocal microscopy. In these experiments, endogenous Nox1 colocalized with endogenous p22phox in a pattern suggestive of plasma membrane and vesicular distribution. Interestingly, this association was demonstrated in the resting state, an observation that is true for the endothelial NADPH oxidase (composed of gp91phox and p22phox) as well [35]. The pattern of distribution of p22phox in HASMs overlapped, but was not restricted to, that of Nox1, raising the possibility of an interaction between p22phox and other members of the Nox family, namely Nox4. We have previously shown that different vascular NAD(P)H oxidases localize to different areas of the cell [22], suggesting that their activities are modulated by different stimuli, resulting in localized changes in ROS production within the cell.
1547
The functional implication of the association between Nox1 and p22phox was studied in the membrane fraction of RASMs lacking p22phox, after the overexpression of one or both proteins. In the absence of S p22phox, overexpression of Nox1 did not increase O2 production over that in vector-infected cells. However, overexpression of Nox1 in combination with the reintroduction of p22phox resulted in a significant S increase in O2 production in these cells, similar to results found in CHO cells transfected with Nox1, p22phox, and the cytosolic subunits [14]. Interestingly, restitution of p22phox expression in resting RASMs was S not accompanied by a change in O2 production, consistent with the notion that unstimulated RASMs express very little Nox1 [12] and that growth-arrested S cells have minimal O2 production. Moreover, in wildtype RASMs, overexpression of Nox1 or both Nox1 and p22phox resulted in increased O2S production, whereas p22phox overexpression did not (unpublished observations). This suggests that Nox1 may act as a rateS limiting factor in NADPH-dependent O2 production in these cells. Based on current knowledge about the phagocyte flavin oxidase, a number of roles can be postulated for p22phox, including stabilization of the catalytic subunit Nox1 or association with the cytosolic components of the oxidase to form the multisubunit functional enzyme. Our data are supportive of both possibilities. As seen in Fig. 3B, coexpression of Nox1 and p22phox enhanced the expression of both proteins, compared to expression of either alone. Figs. 3D and 3E clearly show that p22phox S is necessary for O2 production, as discussed above. Previously, we have shown that either p22phox or nox1 antisense prevented the ability of angiotensin II to S increase O2 production [12,16]. However, because VSMCs express more than one catalytic subunit (Nox1, 4, and 5), it was unclear how these two observations were related. The present data clearly show that Nox1 and p22phox interact to form a functional oxidase and that S this interaction is capable of producing O2 even in the absence of a stimulus. Whether this constitutive activity can be increased by interaction with specific cytosolic subunits remains to be determined. In conclusion, as predicted based on homology with gp91phox, we show that Nox1 associates with p22phox in the resting state of VSMCs to form a functional NADPHdependent oxidase. We expect this finding to simplify the characterization of the remaining structural components of this important enzyme complex, leading to a more thorough understanding of its regulation in response to agonist/antagonist substances. The differential regulation of the various vascular NAD(P)H oxidases, their patterns of distribution, and their effects on target cells are fertile areas of future investigation that will expand our under-
1548
I. R. Hanna et al.
standing of this new class of molecules, with the hope that targeted pharmacological interventions will allow selective modulation of one enzyme system, without affecting critical components of the body’s innate immunity. Acknowledgments — This work was supported by Postdoctoral Fellowship AHA Grant 0120351B to Dr. Hanna and NIH Grants HL38206, HL58000, and AR42426. We thank Dr. Margaret Tarpey and QiQin Vin for their help in preparing the adenoviruses.
REFERENCES [1] Miller, F. J. J.; Gutterman, D. D.; Rios, C. D.; Heistad, D. D.; Davidson, B. L. Superoxide production in vascular smooth muscle contributes to oxidative stress and impaired relaxation in atherosclerosis. Circulation Res. 82:1298 – 1305; 1998. [2] Mohazzab-H, K. M.; Kaminski, P. M.; Wolin, M. S. NADH oxidoreductase is the major source of superoxide anion in bovine coronary artery endothelium. Am. J. Physiol. 266:H2568 – H2572; 1994. [3] Pagano, P. J.; Clark, J. K.; Cifuentes-Pagano, M. E.; Clark, S. M.; Callis, G. M.; Quinn, M. T. Localization of a constitutively active, phagocyte-like NADPH oxidase in rabbit aortic adventitia: enhancement by angiotensin II. Proc. Natl. Acad. Sci. USA 94; 1997. [4] Rajagopalan, S.; Kurz, S.; Mqnzel, T.; Tarpey, M.; Freeman, B. A.; Griendling, K. K.; Harrison, D. G. 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. 97:1916 – 1923; 1996. [5] Griendling, K. K.; Sorescu, D.; Ushio-Fukai, M. NAD(P)H oxidase. role in cardiovascular biology and disease. Circ. Res. 86: 494 – 501; 2000. [6] Lassegue, B.; Clempus, R. E. Vascular NAD(P)H oxidases: specific features, expression, and regulation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 285:R277 – R297; 2003. [7] Parkos, C. A.; Allen, R. A.; Cochrane, C. G.; Jesaitis, A. J. Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J. Clin. Invest. 80:732 – 742; 1987. [8] Wallach, T. M.; Segal, A. W. Stoichiometry of the subunits of flavocytochrome b558 of the NADPH oxidase of phagocytes. Biochem. J. 320:33 – 38; 1996. [9] Babior, B. M.; Lambeth, J. D.; Nauseef, W. M. The neutrophil NADPH oxidase. Arch. Biochem. Biophys. 397:342 – 344; 2002. [10] Hanna, I. R.; Taniyama, Y.; Szocs, K.; Rocic, P.; Griendling, K. K. NAD(P)H oxidase-derived reactive oxygen species as mediators of angiotensin II signaling. Antioxid. Redox Signaling 4:899 – 914; 2002. [11] Suh, Y.-A.; Arnold, R. S.; Lassegue, B.; Shi, J.; Xu, X.; Sorescu, D.; Chung, A. B.; Griendling, K. K.; Lambeth, J. D. Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79 – 82; 1999. [12] Lassegue, B.; Sorescu, D.; Szocs, K.; Akers, M.; Zhang, Y.; Grant, S. L.; Lambeth, J. D.; Griendling, K. K. 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. [13] Geiszt, M.; Lekstrom, K.; Witta, J.; Leto, T. L. 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. [14] Takeya, R.; Ueno, N.; Kami, K.; Taura, M.; Kohjima, M.; Izaki, H.; Nunoi, H.; Sumimoto, H. Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J. Biol. Chem. 278:25234 – 25246; 2003.
[15] Banfi, B.; Clark, R. A.; Steger, K.; Krause, K.-H. Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J. Biol. Chem. 278:3510 – 3513; 2003. [16] Ushio-Fukai, M.; Zafari, A. M.; Fukui, T.; Ishizaka, N.; Griendling, K. K. p22phox is a critical component of the superoxide-generating NADH/NADPH oxidase system and regulates angiotensin IIinduced hypertrophy in vascular smooth muscle cells. J. Biol. Chem. 271:23317 – 23321; 1996. [17] Griendling, K. K.; Taubman, M. B.; Akers, M.; Mendlowitz, M.; Alexander, R. W. Characterization of phosphatidylinositol-specific phospholipase C from cultured vascular smooth muscle cells. J. Biol. Chem. 266:15498 – 15504; 1991. [18] He, T. C.; Zhou, S.; da Costa, L. T.; Yu, J.; Kinzler, K. W.; Vogelstein, B. A simplified system for generating recombinant adenoviruses. Proc. Natl. Acad. Sci. USA 95:2509 – 2514; 1998. [19] Jesaitis, A. J.; Buescher, E. S.; Harrison, D.; Quinn, M. T.; Parkos, C. A.; Livesey, S.; Linner, J. Ultrastructural localization of cytochrome b in the membranes of resting and phagocytosing human granulocytes. J. Clin. Invest. 85:821 – 835; 1990. [20] Burritt, J. B.; Quinn, M. T.; Jutila, M. A.; Bond, C. W.; Jesaitis, A. J. Topological mapping of neutrophil cytochrome b epitopes with phage-display libraries. J. Biol. Chem. 270:16974 – 16980; 1995. [21] Quinn, M. T.; Mullen, M. L.; Jesaitis, A. J. Human neutrophil cytochrome b contains multiple hemes: evidence for heme associated with both subunits. J. Biol. Chem. 267:7303 – 7309; 1992. [22] Hilenski, L. L.; Clempus, R. E.; Quinn, M. T.; Lambeth, J. D.; Griendling, K. K. Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler. Thromb. Vasc. Biol. 24:677 – 683; 2004. [23] Dikalov, S. I.; Dikalova, A. E.; Mason, R. P. New non-invasive diagnostic tool for inflammation-induced oxidative stress using electron spin resonance spectroscopy and cyclic hydroxylamine. Arch. Biochem. Biophys. 402:218 – 226; 2001. [24] Touyz, R. M.; Chen, X.; Tabet, F.; Yao, G.; He, G.; Quinn, M. T.; Pagano, P. J.; Schiffrin, E. L. Expression of a functionally active gp91phox-containing neutrophil-type NAD(P)H oxidase in smooth muscle cells from human resistance arteries. Circ. Res. 90:1205 – 1213; 2002. [25] Burritt, J. B.; DeLeo, F. R.; McDonald, C. L.; Prigge, J. R.; Dinauer, M. C.; Nakamura, M.; Nauseef, W. M.; Jesaitis, A. J. Phage display epitope mapping of human neutrophil flavocytochrome b558. J. Biol. Chem. 276:2053 – 2061; 2001. [26] DeLeo, F. R.; Burritt, J. B.; Yu, L.; Jesaitis, A. J.; Dinauer, M. C.; Nauseef, W. M. Processing and maturation of flavocytochrome b558 include incorporation of heme as pre-requisite for heterodimer assembly. J. Biol. Chem. 275:13986 – 13993; 2000. [27] Regier, D. S.; Waite, K. A.; Wallin, R.; McPhail, L. C. A phosphatidic acid-activated protein kinase and conventional protein kinase C isoforms phosphorylate p22phox, an NADPH oxidase component. J. Biol. Chem. 274:36601 – 36608; 1999. [28] Regier, D. S.; Greene, D. G.; Sergeant, S.; Jesaitis, A. J.; McPhail, L. C. Phosphorylation of p22phox is mediated by phospholipase D-dependent and -independent mechanisms. J. Biol. Chem. 275:28406 – 28412; 2000. [29] Yu, L.; Zhen, L.; Dinauer, M. C. Biosynthesis of the phagocyte NADPH oxidase cytochrome b558: role of heme incorporation and heterodimer formation in maturation and stability of gp91phox and p22phox subunits. J. Biol. Chem. 272:27288 – 27294; 1997. [30] Mugge, A.; Elwell, J. H.; Peterson, T. E.; Hofmeyer, T. G.; Heistad, D. D.; Harrison, D. G. Chronic treatment with polyethylene-glycolated superoxide dismutase partially restores endothelium-dependent vascular relaxations in cholesterol-fed rabbits. Circ. Res. 69:1293 – 1300; 1991. [31] Aviram, M.; Rosenbalt, M.; Etzioni, A.; Levy, R. Activation of NADPH oxidase required for macrophage-mediated oxidation of low-density lipoprotein. Metabolism 45:1069 – 1079; 1996. [32] Rajagopalan, S.; Meng, X. P.; Ramasamy, S.; Harrison, D. G.; Galis, Z. S. Reactive oxygen species produced by macrophagederived foam cells regulate the activity of vascular matrix metalloproteinases in vitro. J. Clin. Invest. 98:2572 – 2579; 1996.
Association of Nox1 and p22phox [33] Szocs, K.; Lassegue, B.; Sorescu, D.; Hilenski, L. L.; Valppu, L.; Couse, T. L.; Wilcox, J. N.; Lambeth, J. D.; Griendling, K. K. Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler. Thromb. Vasc. Biol. 22:21 – 27; 2002. [34] Sorescu, D.; Somers, M. J.; Lassegue, B.; Grant, S.; Harrison, D. G.; Griendling, K. K. Electron spin resonance characterization of the NAD(P)H oxidase in vascular smooth muscle cells. Free Radic. Biol. Med. 30:603 – 612; 2001. [35] Li, J. M.; Shah, A. M. Intracellular localization and preassembly of the NADPH oxidase complex in cultured endothelial cells. J. Biol. Chem. 277:19952 – 19960; 2002. ABBREVIATIONS
AdGFP — adenovirus expressing green fluorescent protein only AdNox1 — adenovirus expressing Nox1 AdNoxHA — adenovirus expressing hemagglutinin-tagged Nox1 Adp22phoxV5 — adenovirus expressing V5-tagged p22phox
BSA — bovine serum albumin CPH — 1-hydroxy-3-carboxypyrrolidine DME — Dulbecco’s modified Eagle’s medium DPI — diphenylene iodonium ESR — electron spin resonance GFP — green fluorescent protein HA — hemagglutinin HASMs — human aortic smooth muscle cells m.o.i. — multiplicity of infection Nox — NAD(P)H oxidase S O2 — superoxide PBS — phosphate-buffered saline RASMs — rat aortic smooth muscle cells ROS — reactive oxygen species SOD — superoxide dismutase VSMCs — vascular smooth muscle cells
1549