Endothelial nitric oxide synthase is regulated by tyrosine phosphorylation and interacts with caveolin-1.

The regulation of endothelial nitric oxide synthase (eNOS) by phosphorylation is poorly understood. Here, we demonstrate that eNOS is tyrosine-phosphorylated in bovine aortic endothelial cells (BAEC) using 32P metabolic labeling followed by phosphoamino acid analysis and by phosphotyrosine specific Western blotting. Treatment of BAEC with hydrogen peroxide and the protein tyrosine phosphatase inhibitor, sodium orthovanadate, increases eNOS tyrosine phosphorylation. Utilizing a novel immunoNOS assay, the increase in tyrosine phosphorylation is associated with a 50% decrease in the specific activity of the enzyme. Because eNOS is localized in plasmalemma caveolae, we examined if tyrosine phosphorylated eNOS interacts with caveolin-1, the coat protein of caveolae. Immunoprecipitation of eNOS from bovine lung microvascular endothelial cells resulted in the co-precipitation of caveolin-1. Conversely, immunoprecipitation of caveolin-1 resulted in the co-precipitation of tyrosine-phosphorylated eNOS. Thus, tyrosine phosphorylation is a novel regulatory mechanism for eNOS and caveolin-1 is the first eNOS-associated protein. Collectively, these observations provide a novel regulatory mechanism for eNOS and suggest that tyrosine phosphorylation may influence its activity, subcellular trafficking, and interaction with other caveolin-interacting proteins in caveolae.

cal processes, including cellular proliferation, migration, and differentiation. More specifically, phosphorylation of tyrosine residues is an essential step in cellular activation by many external signals, including growth factors, cytokines, and cellular stress (1,2).
In the cardiovascular and nervous systems, activation of cell surface receptors triggers the immediate synthesis of nitric oxide (NO). 1 In the cardiovascular system, NO is derived from one of the chemically equivalent guanidino nitrogens of L-arginine, in a reaction catalyzed by endothelial NO synthase (eNOS or NOS 3). In intact blood vessels and in cultured endothelial cells (EC), eNOS is rapidly activated by agonists that mobilize intracellular calcium. Increases in intracellular calcium facilitate interactions with calmodulin and the activated calciumcalmodulin complex can stimulate NADPH-dependent electron flux through eNOS to produce NO (3). Recent evidence suggests that eNOS can also be activated in a calcium-independent manner by fluid shear stress (4 -6) and insulin-like growth factor (7), presumably mediated through a tyrosine kinase cascade, since inhibitors of tyrosine kinases, but not chelation of intracellular calcium, inhibits NO release. Whether eNOS is directly tyrosine-phosphorylated or indirectly linked to a tyrosine kinase cascade is not known. Once NO is produced by either mechanism, it rapidly diffuses and mediates endotheliumdependent relaxation of blood vessels, inhibition of platelet aggregation and leukocyte adhesion, and vascular smooth muscle growth (8).
eNOS has been shown to reside in plasmalemmal caveolae (9,10). Recent data suggests that plasmalemmal caveolae are cellular signal processing centers. Activation of fibroblasts with platelet-derived growth factor or epidermal growth factor results in the rapid recruitment of signal transducing proteins (Syp, Shc, Ras/Raf-1, and mitogen-activated protein kinase) into caveolae microdomains (11,12). The direct interaction between caveolin-1, the major coat protein of caveolae, with inactive G␣ and Ras supports the concept that not only the recruitment, but the inactivation of signaling molecules, occurs in caveolae (13,14). The presence of eNOS in caveolae suggests that resident caveolar proteins can potentially interact with eNOS and that NO can influence signal transduction through caveolae.
Here, we report that eNOS is tyrosine-phosphorylated in EC. Treatment of EC with hydrogen peroxide (H 2 O 2 ) or inhibition of endogenous protein tyrosine phosphatases increases eNOS tyrosine phosphorylation and hyperphosphorylation is associated with a decrease in NOS specific activity. Additionally, tyrosine-phosphorylated eNOS interacts with the caveolar coat protein, caveolin-1, based on co-immunoprecipitation assays. These findings demonstrate a novel regulatory mechanism for eNOS and suggest that tyrosine phosphorylation may influence its activity, subcellular trafficking, and interaction with other caveolin-interacting proteins in caveolae. * This work was supported by Grants HL 51948 (to W. C. S.), F32-HL09224 (to J. L.), and RO1CA-45708 (to D. F. S.) from the National Institutes of Health, an Established Investigator Award, a grant-in aid from the American Heart Association (to W. C. S.), the Patrick and Catherine Weldon Donaghue Medical Research Foundation (to W. C. S.), and the Government of Mexico/Yale University (to G. G. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

EXPERIMENTAL PROCEDURES
Materials and Antibodies-Tissue culture reagents were from Life Technologies, Inc., and chemicals were from Sigma. n-Octyl-␤-D-glucopyranoside was from Calbiochem. [lsqb] 32 P]Orthophosphoric acid was from Amersham Corp. eNOS monoclonal antibody (H32, mAb) was kindly provided by J. S. Pollock (Medical College of Georgia), and the caveolin-1 polyclonal antibody and horseradish peroxidase conjugate anti-phosphotyrosine antibody (RC20) were from Transduction Laboratories. Bovine lung microvascular endothelial cells (BLMVEC, 10 cell doublings) were provided by Peter Del Vecchio. Stock solutions of sodium orthovanadate (vanadate, assuming a hydration number of 10) were prepared in sterile distilled water and adjusted to pH 10. The solution was then heated to boiling until translucent and the pH readjusted to 10.
Cell Culture-Bovine aortic endothelial cells (BAEC) or BLMVEC were isolated and cultured in Dulbecco's modified Eagle's medium containing 10% (v/v) fetal bovine serum, penicillin, streptomycin, and Lglutamine as described previously (15,16). BAEC were use for these experiments between passages 2 and 4. In some experiments, BAEC were incubated in complete medium with vanadate (1 mM) for 2 h or H 2 O 2 (0.1-10 mM) for 20 min and processed for immunoprecipitation. These treatments were not associated with EC toxicity, presumably due to the presence of 10% serum, as assessed by overall cell morphology.
Labeling and Phosphoamino Acid Analysis-Confluent 100-mm dishes of BAEC were incubated in phosphate-free Dulbecco's minimum essential medium with 80 Ci/ml of 32 [P]orthophosphoric acid for 4 h, harvested, and immunoprecipitated as described below. Proteins were resolved by SDS-PAGE (7.5% acrylamide) transferred to PVDF, and identified by autoradiography. For two-dimensional phosphoamino acid analysis, proteins were blotted on PVDF membranes, and the eNOS band was isolated after autoradiography. In brief, PVDF-blotted eNOS was hydrolyzed in 6 N HCl for 45 min at 105°C. The aqueous phase was lyophillized, extensively washed, and resuspended in electrophoresis buffer as described previously (17).
Immunoprecipitations and Western Blotting-Cells were washed twice with TBS (50 mM Tris⅐HCl, 150 mM NaCl), lysed on ice in modified RIPA buffer (100 mM Tris⅐HCl, pH 7.4, 1%, v/v, Nonidet P-40 10 mM NaF, 1 mM, vanadate, 1 mM Pefabloc, 10 g/ml of aprotinin, 10 g/ml of leupeptin) and lysates transferred to an Eppendorf tube and rotated for 30 min at 4°C. Lysates were Dounce-homogenized (50 strokes) and insoluble material removed by centrifugation at 12,000 ϫ g for 10 min at 4°C. The protein concentration of the soluble material was determined using a Lowry assay (DC protein assay, Bio-Rad). Equal amounts of protein were precleared by incubation with Pansorbin for 1 h at 4°C and Pansorbin complexes removed by centrifugation. The cleared lysates were incubated with an excess of eNOS mAb for 2 h at 4°C, protein A-Sepharose added, and the mixture was incubated for an additional 1 h. The immune complexes were washed three times with RIPA buffer and boiled in SDS-PAGE sample buffer for 5 min.
For co-immunoprecipitation experiments, BLMVEC were washed twice with TBS and lysed in RG buffer (10 mM Tris, pH 8.0, 60 mM n-octyl-␤-D-glucopyranoside, 150 mM NaCl, 10 mM NaF, 1 mM vanadate, 1 mM Pefabloc, 10 g/ml of aprotinin and 10 g/ml of leupeptin). Cells were extracted and cleared by centrifugation at 12,000 ϫ g for 10 min. Lysates were precleared with protein A for 1 h and equal amounts of protein incubated with eNOS mAb or caveolin-1 mAb for 16 h. Protein A-Sepharose was added and samples processed as above.
Immunoprecipitates samples were separated by SDS-PAGE (7.5 or 12.5% gels), followed by overnight transfer of the proteins to nitrocellulose membranes. Membranes were blocked by incubation in Trisbuffered saline (10 mM Tris, pH 7.5, 100 mM NaCl) containing 5% non-fat dry milk for 2 h, followed by 2-h incubation in primary eNOS mAb (tissue culture supernatant H32) or caveolin-1 polyclonal antibody diluted in blocking buffer. Membranes were washed extensively in Tris-buffered saline before incubation for 30 min with goat anti-mouse or donkey anti-rabbit horseradish peroxidase-conjugated secondary Abs. Membranes were then washed and developed using enhanced chemiluminescence substrate (ECL, Amersham).
Anti-phosphotyrosine Blots-Nitrocellulose membranes were blocked by incubation in Tris-buffered saline-Tween 20 (TBS-T) containing 1% bovine serum albumin for 2 h, followed by 20-min incubation in RC20 diluted in blocking buffer. The membranes were washed extensively in TBS-T and developed by using ECL. To determine the specificity of RC20 for phosphotyrosine in eNOS, the Ab was preincubated overnight at 4°C with 1 mM solutions, pH 7.5, of either phosphoserine, phosphothreonine, and phosphotyrosine and added to nitrocellulose membranes containing equal amounts of immunoprecipitated eNOS.
ImmunoNOS Assay-eNOS immunocomplexes were obtained from control and vanadate-treated BAEC as described above. Samples were then split into two tubes, one for Western blotting and one for determination of NOS activity. eNOS immunoprecipitated and immobilized on protein A-Sepharose beads was resuspended in 1 ml of NOS assay buffer (50 mM Tris⅐HCl, pH 7.5, 0.1 mM EDTA, 0.1 mM EGTA, 2 M leupeptin, 1 M pepstatin, 1 M aprotinin, 1 mM phenylmethanesulfonyl fluoride, 10 mM NaF, 1 mM vanadate), and NOS activity was determined by measuring the conversion of L-[ 3 H]arginine into L-[ 3 H]citrulline as described previously (18). Briefly, beads were incubated (total volume 100 l) in assay buffer containing 1 mM NADPH, 3 M tetrahydrobiopterin, 100 nM calmodulin, 2.5 mM CaCl 2 , and 10 M L-arginine, and L-[ 3 H]arginine (0.2 Ci, 55 Ci/mol) for 30 min at 37°C. After the incubation period, the reaction was quenched by the addition of 1 ml of cold stop buffer (20 mM HEPES, pH 5.5, containing 2 mM or each EDTA and EGTA) and the reaction mix passed over a 1-ml column containing Dowex AG 50WX-8 resin (Na ϩ form; preequilibrated in stop buffer) washed with 1 ml of water and collected directly into a 20-ml liquid scintillation vial. NOS activity was completely inhibited by incubation of the immunoprecipitate with the NOS inhibitor, N G -monomethyl-L-arginine (L-NMMA, 1 mM), and NOS activity isolated from vanadate treated BAEC was expressed as a percent of control (vehicletreated BAEC). Immunoprecipitated NOS activity was in the range of 0.1-2.0 pmol of L-citrulline generated/min (n ϭ 5 independent experiments). In all experiments, similar amounts of enzyme and the phosphorylation state of NOS in each incubation was verified by eNOS and phosphotyrosine blotting, respectively.

RESULTS AND DISCUSSION
As demonstrated previously, biosynthetic labeling of BAEC with 32 [P]orthophosphate resulted in the incorporation of the label into immunoprecipitated eNOS (Fig. 1A, Refs. 19 -21). To identify the phosphorylated residues, immunoprecipitated 32 Plabeled eNOS was isolated from BAEC, hydrolyzed under acidic conditions, and phosphoamino acids analyzed by twodimensional TLC electrophoresis. As seen in Fig. 1B, a majority of the label was incorporated into phosphoserine as described previously (19), but clearly detectable amounts of phosphotyrosine were also found. Acid hydroylsis of labeled eNOS for longer times resulted in the reduced ability to detect phosphotyrosine (data not shown), consistent with the known acid lability of phosphotyrosine residues (22). Due to the qualitative nature of phosphoamino acid analysis, the stoichiometry of eNOS tyrosine phosphorylation is not yet known.
To corroborate the presence of tyrosine phosphorylated res- idues in eNOS, phosphotyrosine blotting was performed using a specific phosphotyrosine Ab, RC20. As seen in Fig. 1C, eNOS immunoprecipitated from BAEC was recognized by the antiserum; and preabsorption of the antiserum with phosphotyrosine, but not phosphoserine or threonine, completely prevented the immunoreactivity with eNOS. Identical results were obtained using another phosphotyrosine Ab, 4G10 (data not shown), and was also seen in human umbilical vein endothelial cells and in HEK 293 cells stably transfected with the eNOS cDNA. Thus, eNOS is tyrosine-phosphorylated as determined by two independent biochemical methods.
To examine if increases in tyrosine phosphorylation induced by blockade of protein tyrosine phosphatases regulate NOS activity, we developed an immunoprecipitation-NOS assay (im-munoNOS) to assess the amount of eNOS, its tyrosine phosphorylation state, and the corresponding activity of the enzyme from the same sample isolated from BAEC. Fig. 3 demonstrated that vanadate pretreatment of BAEC increased the phosphotyrosine content of eNOS (A) with equal amounts of eNOS present in control and treated cells (B). ImmunoNOS assay of an equal proportion of the immunoprecipitated protein in A and B showed that enhanced tyrosine phosphorylation of eNOS was associated with a 50% decrease in NOS activity (n ϭ 5 different experiments; similar results were obtained in total cell lysates). The NOS activity isolated from both control and vanadate-treated BAEC was attenuated with the NOS inhibitor, L-NMMA. Since vanadate was also present in the lysis buffer of control and treated cells to preserve phosphotyrosine during the immunoprecipitation, the inhibition of NOS activity in pretreated cells was most likely not due to a direct inhibitory effect of the drug on NOS per se. The mechanism by which vanadate increases eNOS phosphorylation is presumably due to inhibition of endogenous eNOS-specific phosphotyrosine phosphatases, however, we cannot rule out the possibility that vanadate is indirectly activating cellular tyrosine kinases, thereby increasing the tyrosine phosphorylation of eNOS. Thus, increases in eNOS tyrosine phosphorylation in BAEC decrease eNOS catalytic activity. eNOS has been shown to reside in the plasmalemmal microdomain, caveolae (9, 10). Caveolae may play a role in the organiza-  tion of signal processing centers suggesting that eNOS has the propensity to interact with the caveolar coat protein caveolin-1, as recently described for the G-protein ␣ subunits G s , G o , G i2 , and Ras (13,14). To examine if tyrosine-phosphorylated eNOS interacts with caveolin-1, we performed co-immunoprecipitation experiments with eNOS and caveolin antisera. Immunoprecipitation of eNOS from BLMVEC resulted in the co-precipitation of caveolin-1 (Fig. 4A). The amount of caveolin-1 co-precipitated with eNOS antiserum was not stoichiometric relative to the total pool of cellular caveolin (data not shown). Conversely, immunoprecipitation of caveolin-1 resulted in the co-precipitaton of tyrosine-phosphorylated eNOS (Fig. 4B). As seen above, the amount of eNOS that co-precipitated with caveolin-1 was not stoichiometric. Immunoprecipitation with non-immune antisera did not result in the detection of either eNOS or caveolin-1 (data not shown). The data are consistent with our previous findings in BLMVEC, demonstrating that a pool of eNOS co-localizes with caveolin-1 based on purification of plasmalemma caveolae and confocal microscopy (9).
These results provide a novel mechanism to regulate the catalytic activity and potentially the subcellular trafficking of eNOS. In general, oxidative stress induced by pro-inflammatory cytokines or oxidants activates tyrosine kinase signaling pathways. Increases in oxidative stress brought about by H 2 O 2 or by inhibition of endogenous protein-tyrosine phosphatases increase eNOS tyrosine phosphorylation, and in the latter case, reduce NOS activity. Interestingly, several cardiovascular diseases with diverse etiologies, such as atherosclerosis, vascular complications of diabetes, ischemia-reperfusion injury, and hypertension are associated with the common hallmarks of increased oxidative stress and endothelial cell dysfunction (24). Dysfunction is manifested by the inability to vasodilate in response to endothelium-dependent vasodilators, in the face of normal or increased expression of eNOS. Perhaps, increased eNOS tyrosine phosphorylation can contribute to dysfunction in these disease processes.
The co-immunoprecipitation of caveolin-1 with eNOS raises several interesting possibilities relevant to the trafficking of both proteins and to the activation and phosphorylation of eNOS. Both caveolin-1 and eNOS localize on the Golgi complex and in caveolae (9,15,26). However, the mechanisms of anterograde traffick-ing for proteins modified by fatty acylation, including caveolin-1, Src family members, G-proteins, and eNOS, from Golgi to caveolae are poorly understood. If eNOS directly interacts with caveolin-1, then a complex of signaling proteins may associate and be co-transported to plasmalemmal caveolae, presumably via vesicular transport. In addition, once in caveolae, eNOS may be tyrosine-phosphorylated by resident caveolar Src family members (14,27). By analogy to the GDP dissociation inhibitor function of caveolin-1 for the G ␣ subunit (13), it is possible that caveolin directly modulates NOS activity. An alternative model is that caveolin-1 indirectly immunoprecipitates with eNOS through interactions with other proteins.
Thus, tyrosine phosphorylation is a novel regulatory mechanism for eNOS and possibly for other NOS family members. Analogous to the interaction of dystrophin-complex proteins and post-synaptic density protein-95 with neuronal NOS (28,29), caveolin-1 is the first eNOS-associated protein, suggesting that protein-protein interactions can regulate NO production in diverse cell types. Collectively, these observations provide a novel regulatory mechanism for eNOS and suggest that tyrosine phosphorylation may influence its activity, subcellular trafficking, and interaction with other caveolin-interacting proteins in caveolae. The identification of the site(s) of tyrosine phosphorylation in eNOS and understanding the nature of interactions with caveolin-1 will undoubtedly shed light on the role of NO in signal transduction.