Receptor-regulated Dynamic S-Nitrosylation of Endothelial Nitric-oxide Synthase in Vascular Endothelial Cells

The endothelial isoform of nitric-oxide synthase (eNOS) is regulated by a complex pattern of post-translational modifications. In these studies, we show that eNOS is dynamically regulated by S-nitrosylation, the covalent adduction of nitric oxide (NO)-derived nitrosyl groups to the cysteine thiols of proteins. We report that eNOS is tonically S-nitrosylated in resting bovine aortic endothelial cells and that the enzyme undergoes rapid transient denitrosylation after addition of the eNOS agonist, vascular endothelial growth factor. eNOS is thereafter progressively renitrosylated to basal levels. The receptor-mediated decrease in eNOS S-nitrosylation is inversely related to enzyme phosphorylation at Ser1179, a site associated with eNOS activation. We also document that targeting of eNOS to the cell membrane is required for eNOS S-nitrosylation. Acylation-deficient mutant eNOS, which is targeted to the cytosol, does not undergo S-nitrosylation. Using purified eNOS, we show that eNOS S-nitrosylation by exogenous NO donors inhibits enzyme activity and that eNOS inhibition is reversed by denitrosylation. We determine that the cysteines of the zinc-tetrathiolate that comprise the eNOS dimer interface are the targets of S-nitrosylation. Mutation of the zinc-tetrathiolate cysteines eliminates eNOS S-nitrosylation but does not eliminate NO synthase activity, arguing strongly that disruption of the zinc-tetrathiolate does not necessarily lead to eNOS monomerization in vivo. Taken together, these studies suggest that eNOS S-nitrosylation may represent an important mechanism for regulation of NO signaling pathways in the vascular wall. The endothelial isoform of nitric-oxide synthase (eNOS) is regulated by a complex pattern of post-translational modifications. In these studies, we show that eNOS is dynamically regulated by S-nitrosylation, the covalent adduction of nitric oxide (NO)-derived nitrosyl groups to the cysteine thiols of proteins. We report that eNOS is tonically S-nitrosylated in resting bovine aortic endothelial cells and that the enzyme undergoes rapid transient denitrosylation after addition of the eNOS agonist, vascular endothelial growth factor. eNOS is thereafter progressively renitrosylated to basal levels. The receptor-mediated decrease in eNOS S-nitrosylation is inversely related to enzyme phosphorylation at Ser1179, a site associated with eNOS activation. We also document that targeting of eNOS to the cell membrane is required for eNOS S-nitrosylation. Acylation-deficient mutant eNOS, which is targeted to the cytosol, does not undergo S-nitrosylation. Using purified eNOS, we show that eNOS S-nitrosylation by exogenous NO donors inhibits enzyme activity and that eNOS inhibition is reversed by denitrosylation. We determine that the cysteines of the zinc-tetrathiolate that comprise the eNOS dimer interface are the targets of S-nitrosylation. Mutation of the zinc-tetrathiolate cysteines eliminates eNOS S-nitrosylation but does not eliminate NO synthase activity, arguing strongly that disruption of the zinc-tetrathiolate does not necessarily lead to eNOS monomerization in vivo. Taken together, these studies suggest that eNOS S-nitrosylation may represent an important mechanism for regulation of NO signaling pathways in the vascular wall. Nitric oxide (NO) 1The abbreviations used are: NO, nitric oxide; BAEC, bovine aortic endothelial cells; DEA/NO, diethylamine-NONOate; NOS, nitric-oxide synthase; eNOS, endothelial nitric-oxide synthase; HRP, horseradish peroxidase; l-NMA, N-methyl-l-arginine; HA, hemagglutinin; NOS, NO synthase; EGFP, enhanced green fluorescent protein; VEGF, vascular endothelial growth factor; ANOVA, analysis of variance. is a reactive free radical gas that plays a central role in diverse signaling pathways (reviewed in Ref. 1Stuart-Smith K. Mol. Pathol. 2002; 55: 360-366Crossref PubMed Scopus (42) Google Scholar). In mammals, NO is synthesized by three synthases responsible for (patho)physiologic NO signaling in the immune, neurological, and cardiovascular systems (reviewed in Ref. 2Andrew P.J. Mayer B. Cardiovasc. Res. 1999; 43: 521-531Crossref PubMed Scopus (584) Google Scholar). Vascular endothelial cells robustly express the endothelial isoform of NO synthase (eNOS), a constitutive 135-kDa protein that is central to homeostatic mechanisms such as vasorelaxation, regulation of myocardial contractility, and blood platelet aggregation (reviewed in Ref. 3Loscalzo J. Welch G. Prog. Cardiovasc. Dis. 1995; 38: 87-104Crossref PubMed Scopus (510) Google Scholar). Similar to all NO synthase (NOS) isoforms, eNOS functions as an obligate homodimer through an association mediated by a cysteine-complexed Zn2+ (zinc-tetrathiolate) at the dimer interface (4Raman C.S. Li H. Martasek P. Kral V. Masters B.S. Poulos T.L. Cell. 1998; 95: 939-950Abstract Full Text Full Text PDF PubMed Scopus (577) Google Scholar, 5Hemmens B. Goessler W. Schmidt K. Mayer B. J. Biol. Chem. 2000; 275: 35786-35791Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar, 6Kolodziejski P.J. Rashid M.B. Eissa N.T. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 14263-14268Crossref PubMed Scopus (32) Google Scholar). eNOS activity is regulated by several dynamic phosphorylations and protein-protein interactions (reviewed in Ref. 7Shaul P.W. Annu. Rev. Physiol. 2002; 64: 749-774Crossref PubMed Scopus (475) Google Scholar) that can be modulated following the activation of cell surface receptors by agonists such as the vascular endothelial growth factor (VEGF) (8He H. Venema V.J. Gu X. Venema R.C. Marrero M.B. Caldwell R.B. J. Biol. Chem. 1999; 274: 25130-25135Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar). In its canonical signaling role, NO acts as an intracellular messenger that increases cGMP concentration in target cells by activating guanylate cyclases (9Davis K.L. Martin E. Turko I.V. Murad F. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 203-236Crossref PubMed Scopus (506) Google Scholar). However, NO is also believed to signal through a guanylate cyclase-independent mechanism mediated by the interaction of reactive nitrogen oxides with the reduced cysteines of proteins, forming cysteine S-nitrosothiols that can effect changes in protein function (10Martinez-Ruiz A. Lamas S. Cardiovasc. Res. 2004; 62: 43-52Crossref PubMed Scopus (217) Google Scholar). This process has been termed “S-nitrosylation” and appears to be a physiologically significant post-translational modification with functional consequences for signal protein activity (reviewed in Ref. 11Stamler J.S. Lamas S. Fang F.C. Cell. 2001; 106: 675-683Abstract Full Text Full Text PDF PubMed Scopus (1133) Google Scholar). Dynamic receptor-modulated S-nitrosylation of a signal protein has not been reported previously, and the role of reversible S-nitrosylation in cellular signal transduction remains incompletely understood (11Stamler J.S. Lamas S. Fang F.C. Cell. 2001; 106: 675-683Abstract Full Text Full Text PDF PubMed Scopus (1133) Google Scholar). In this paper, we report that eNOS is tonically S-nitrosylated in bovine aortic endothelial cells (BAEC) and that the enzyme undergoes rapid transient denitrosylation after the addition of the eNOS agonist, VEGF or insulin. eNOS is thereafter progressively renitrosylated on a time course parallel to that of its return to resting activity levels after receptor-mediated activation. These studies also provide evidence that eNOS S-nitrosylation reversibly attenuates enzyme activity and identify a potentially important pathway for eNOS regulation. Materials—Fetal bovine serum was from Hyclone (Logan, UT). All of the other cell culture reagents were from Invitrogen. VEGF, diethylamine-NONOate (DEA/NO), and wortmannin were from Calbiochem. Anti-eNOS monoclonal antibody was from Transduction Laboratories (Lexington, KY), and anti-phospho-eNOSSer-1179 antibody (phospho-Ser1177 in the human eNOS sequence) was from Cell Signaling Technologies (Beverly, MA). Polyclonal antibodies against eNOS phospho-Ser116, eNOS phospho-Thr497 (phospho-Thr495 in the human sequence), and the HA epitope were from Upstate Biotechnology (Lake Placid, NY). Polyclonal eNOS anti-serum raised in rabbits in this laboratory was described previously (12Busconi L. Michel T. J. Biol. Chem. 1993; 268: 8410-8413Abstract Full Text PDF PubMed Google Scholar). Monoclonal antibody 12CA5 used for immunoprecipitation of HA epitope-tagged eNOS was from Roche Applied Science. Anti-HA monoclonal antibody for immunoblot, and protein-A/G-agarose was from Santa Cruz Biotechnology (Santa Cruz, CA). Alexa Fluor 568 anti-rabbit IgG secondary antibody was from Molecular Probes (Eugene, OR). SuperSignal substrate for chemiluminescence detection, Immunopure IgG elution buffer, biotin-HPDP, and horseradish peroxidase conjugated to secondary antibodies and to avidin (avidin-HRP) were from Pierce. PCR primers were synthesized by Integrated DNA Technologies (Coralville, IA). Streptavidin-agarose, protein-A-agarose, and all of the other reagents were from Sigma. Cell Culture and Drug Treatment—BAEC were obtained from Cambrex (Walkersville, MD) and maintained in culture on gelatin-coated 100-mm culture dishes with Dulbecco's modified Eagle's medium supplemented with fetal bovine serum (10% v/v) and penicillin-streptomycin (2%). Experiments were performed with cells between passages 5 and 8. Cultures were serum-starved overnight before experiments and drug treatments were performed as described (13Igarashi J. Erwin P.A. Dantas A.P.V. Chen H. Michel T. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 10664-10669Crossref PubMed Scopus (173) Google Scholar). COS-7 cells were obtained from ATCC (Manassas, VA) and maintained on uncoated culture dishes in the same medium as BAEC (14Igarashi J. Michel T. J. Biol. Chem. 2000; 275: 32363-32370Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). Plasmid Construction and Transfection—The plasmid pKENH encoding HA epitope-tagged wild-type bovine eNOS cDNA (GenBank™ accession number M89952 (46Lamas S. Marsden P.A. Li G.K. Tempst P. Michel T. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 6348-6352Crossref PubMed Scopus (921) Google Scholar)), its parent plasmid pBKCMV (Stratagene), and eNOS cDNA constructs containing mutations at Ser116 and Ser1179 have been described previously (15Robinson L.J. Busconi L. Michel T. J. Biol. Chem. 1995; 270: 995-998Abstract Full Text Full Text PDF PubMed Scopus (211) Google Scholar, 16Kou R. Greif D. Michel T. J. Biol. Chem. 2002; 277: 29669-29673Abstract Full Text Full Text PDF PubMed Scopus (160) Google Scholar, 17Bernier S.G. Haldar S. Michel T. J. Biol. Chem. 2000; 275: 30707-30715Abstract Full Text Full Text PDF PubMed Scopus (164) Google Scholar). Plasmids encoding eNOS mutants deficient in myristoylation (eNOSmyr–) and palmitoylation (eNOSpalm–) have been previously characterized in detail (12Busconi L. Michel T. J. Biol. Chem. 1993; 268: 8410-8413Abstract Full Text PDF PubMed Google Scholar, 18Robinson L.J. Michel T. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11776-11780Crossref PubMed Scopus (131) Google Scholar, 19Gonzalez E. Kou R. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2002; 277: 39554-39560Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). pKENH served as the template for creation of eNOS constructs where Cys96 and/or Cys101 were mutated to Ser by PCR mutagenesis using a two-step approach (20Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Current Protocols in Molecular Biology Online. John Wiley & Sons, Inc., New York2003Google Scholar). In the first step, one fragment was generated with the forward primer for the specific mutation site (described below) plus the reverse primer that subtends the eNOS BglII site (5′-TGAGGCAGAGATCTTCACCG-3′). The second fragment was generated with the reverse mutant primer (see below) plus the forward primer subtending the EcoRI site in the polylinker of the plasmid (5′-CGCGAATTCGAAGGAGCCAC-CATGGGCAACTTGAAGAG-3′). The two fragments were then mixed, and the final product was amplified using the EcoRI and BglII primers. The primer sequences for PCR-directed mutagenesis of eNOS Cys96 to Ser were: forward, 5′-GGCCCAGCACTCCCAGGTGCTGCCTGGG-3′; reverse, 5′-CCCAGGCAGCACCTGGGAGTGCTGGGCC-3′. For mutagenesis of eNOS Cys101 to Ser, the primer sequences were: forward, 5′-GGCCCTGCACTCCCAGGTGCAGCCTGGG-3′; reverse, 5′-CCCAGGCTGCACCTGGGAGTGCAGGGCC-3′. To create the double-mutant, eNOSC96S/C101S, eNOSC96S was used as the template for the primers: forward, 5′-GGCCCAGCACTCCCAGGTGCAGCCTGGG-3′; reverse, 5′-CCCAGGCTGCACCTGGGAGTGCTGGGCC-3′. After molecular cloning of the fragments back into the EcoRI/BglII-restricted parental plasmid, the nucleotide sequence was confirmed by standard dideoxynucleotide-sequencing methods (University of Maine DNA Sequencing Laboratory). BAEC and COS-7 cells were transfected with the plasmids using FuGENE 6 (Roche Applied Science) according to the manufacturer's protocol. The plasmid encoding a fusion of eNOS to the enhanced green fluorescent protein (eNOS-EGFP) has been described previously (19Gonzalez E. Kou R. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2002; 277: 39554-39560Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). Preparation of Cellular Lysates and Immunoprecipitation—100-mm dishes of cells were washed twice with ice-cold phosphate-buffered saline and harvested by scraping cells into 1 ml of lysis buffer comprised of Tris (20 mm, pH 7.4), Nonidet P-40 (1% v/v), deoxycholate (2.5% w/v), NaCl (150 mm), Na3VO4 (2 mm), EDTA (1 mm), NaF (1 mm), and neocuproine (100 μm) supplemented with a mixture of protease inhibitors (17Bernier S.G. Haldar S. Michel T. J. Biol. Chem. 2000; 275: 30707-30715Abstract Full Text Full Text PDF PubMed Scopus (164) Google Scholar). Lysates were rocked at 4 °C for 30 min and centrifuged at 14,000 × g for 15 min. A 1:100 dilution of eNOS polyclonal anti-serum or 4 μg/ml anti-HA monoclonal antibody 12CA5 was incubated with 1 ml of the supernatants at 4 °C with rocking for 1–16 h. After the addition of protein-A/G-agarose and another hour of rocking at 4 °C, the beads were washed extensively with lysis buffer. eNOS was then eluted from the beads and biotinylated according to the biotin switch method essentially as described (21Kim J.-E. Tannenbaum S.R. J. Biol. Chem. 2003; 279: 9758-9764Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar) or released for SDS-PAGE and immunoblot by boiling in SDS-sample buffer containing β-mercaptoethanol as described previously (19Gonzalez E. Kou R. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2002; 277: 39554-39560Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). The biotin switch-processed eNOS was then separated by low temperature non-reducing SDS-PAGE for Western blot with avidin-HRP (21Kim J.-E. Tannenbaum S.R. J. Biol. Chem. 2003; 279: 9758-9764Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar) or concentrated further with streptavidinagarose, eluted by boiling with SDS-sample buffer, and then subjected to SDS-PAGE under reducing conditions for immunoblot against eNOS as described previously (22Jaffrey S.R. Snyder S.H. Sci. STKE 2001. 2001; : PL1Google Scholar). All of the steps before biotinylation were performed under low light conditions. Analysis of eNOS Dimer Stability—The thermal stability profile of recombinant eNOS expressed in COS-7 cells was determined using low temperature SDS-PAGE essentially as described (23Venema R.C. Ju H. Zou R. Ryan J.W. Venema V. J. Biol. Chem. 1997; 272: 1276-1282Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar). The lysates were prepared from COS-7 cells transfected with plasmids encoding recombinant eNOS by scraping the cells into lysis buffer as described above, except that rocking of the lysates at 4 °C was done for 15 min and centrifugation was conducted for 5 min (17Bernier S.G. Haldar S. Michel T. J. Biol. Chem. 2000; 275: 30707-30715Abstract Full Text Full Text PDF PubMed Scopus (164) Google Scholar). The supernatants were then mixed with SDS-sample buffer and incubated for 30 min at 0, 20, 30, 40, and 50 °C before low temperature SDS-PAGE (23Venema R.C. Ju H. Zou R. Ryan J.W. Venema V. J. Biol. Chem. 1997; 272: 1276-1282Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar). Western Blots—After separation by SDS-PAGE, proteins were electroblotted onto nitrocellulose membranes. To detect biotinylated eNOS, these nitrocellulose membranes were blocked with 5% bovine serum albumin in Tris-buffered saline containing 0.1% Tween 20 before incubation with avidin-HRP in 5% bovine serum albumin/Tris-buffered saline containing 0.1% Tween 20 according to the manufacturer's instructions. Because the gels that are used to detect biotinylated proteins are electrophoresed under non-reducing conditions, the protein bands tend to bend and blur more than with standard reducing SDS-PAGE. To detect total eNOS after blotting with avidin-HRP, membranes were stripped using the Immunopure IgG elution buffer for 24 h at room temperature. All of the immunoblots were performed as described previously (13Igarashi J. Erwin P.A. Dantas A.P.V. Chen H. Michel T. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 10664-10669Crossref PubMed Scopus (173) Google Scholar). Stripping of immunoblots for reblotting was performed with ReBlot antibody stripping solution from Chemicon International (Temecula, CA). Densitometric analyses of autoradiographs were done using a ChemiImager 400 (Alpha Innotech, San Leandro, CA). Subcellular Fractionation and Measurement of eNOS Activity— eNOS activity was analyzed either in lysates or subcellular fractions prepared from 100-mm dishes of transfected COS-7 cells. Subcellular fractionation was performed essentially as described with the exception that 10 μm tetrahydrobiopterin was added to the cell harvest buffer (buffer 1) comprised of Tris (50 mm, pH 7.4), EDTA (0.1 mm), EGTA (0.1 mm), and β-mercaptoethanol (2 mm) with a mixture of protease inhibitors (24Prabhakar P. Cheng V. Michel T. J. Biol. Chem. 2000; 275: 19416-19421Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar). After centrifugation of lysates (2,000 × g for 5 min at 4 °C), resuspension in buffer 1, and 10-s sonication using a Branson 450 sonifier (Branson Ultrasonic, Danbury, CT) at 20% nominal converter amplitude, the particulate and soluble fractions were resolved by ultracentrifugation (100,000 × g for 30 min) and the particulate fraction was resuspended in an equal volume of buffer 1 (24Prabhakar P. Cheng V. Michel T. J. Biol. Chem. 2000; 275: 19416-19421Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar). Protein concentrations were determined using the Bradford reagent (Bio-Rad), and eNOS activity was determined by measuring the conversion of l-[3H]arginine to l-[3H]citrulline by anion-exchange chromatography, essentially as described with the exception that the reaction mixture did not contain dithiothreitol and incubation in the reaction mixture was at 30 °C for 30 min (25Balligand J.L. Kobzik L. Han X. Kaye D.M. Belhassen L. O'Hara D.S. Kelly R.A. Smith T.W. Michel T. J. Biol. Chem. 1995; 270: 14582-14586Abstract Full Text Full Text PDF PubMed Scopus (361) Google Scholar). Activity assays of purified eNOS isolated from a baculovirus expression system were performed as described previously (26Busconi L. Michel T. Mol. Pharmacol. 1995; 47: 655-659PubMed Google Scholar). Confocal Fluorescence Microscopy—BAEC grown on coverslips were transfected with plasmids encoding eNOS-EGFP and HA-tagged eNOSC96S/C101S at 50–60% confluence using FuGENE 6. Cell fixing, permeabilization, incubation with polyclonal anti-HA antibody (1:250 dilution), Alexa Fluor 568 anti-rabbit IgG secondary antibody (1:850 dilution), mounting, and imaging at the Nikon Imaging Center at Harvard Medical School were performed as described previously (27Gonzalez E. Nagiel A. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2004; 279: 40659-40669Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Images were processed using two-dimensional no-neighbors deconvolution by MetaMorph (Universal Imaging, Downington, PA) before three-dimensional reconstruction. Other Methods—Mean values for individual experiments are plotted ± S.E. Data plotting and statistical analyses (ANOVA) were performed using Origin software (OriginLab, Northampton, MA). A p value <0.05 was considered statistically significant. Dynamic eNOS S-Nitrosylation in Endothelial Cells—Our initial experiments explored the hypothesis that eNOS is S-nitrosylated in endothelial cells. We exploited a method developed to detect endogenously S-nitrosylated proteins in cell lysates by biotinylation. After blocking free protein sulfhydryls with methyl methanethiosulphonate, the cell lysates are treated with ascorbate to reduce S-nitrosothiols, which are then labeled with a biotinylated active site reagent (28Jaffrey S.R. Erdjument-Bromage H. Ferris C.D. Tempst P. Snyder S.H. Nat. Cell Biol. 2001; 3: 193-197Crossref PubMed Scopus (1226) Google Scholar). This “biotin switch” approach allows for reliable detection of protein S-nitrosothiols by Western blot analyses. We found that when eNOS from resting BAEC is processed by the biotin switch method, a robust signal is detected in Western blots probed with avidin-HRP. We then performed characterizations of agonist-induced changes in eNOS S-nitrosylation using VEGF because of its robust eNOS agonist activity (29Fulton D. Gratton J.P. McCabe T.J. Fontana J. Fujio Y. Walsh K. Franke T.F. Papapetropoulos A. Sessa W.C. Nature. 1999; 399: 597-601Crossref PubMed Scopus (2232) Google Scholar). We had initially anticipated that eNOS S-nitrosylation might increase when the enzyme was activated, but we found that when BAEC are treated with VEGF (10 ng/ml for 5 min) and that eNOS S-nitrosylation reproducibly disappears (n = 7). As shown in Fig. 1A, we performed a time course of VEGF treatments to further analyze changes in eNOS S-nitrosylation. We found that eNOS in resting BAEC is S-nitrosylated and that the enzyme undergoes rapid denitrosylation after 5 min of VEGF-treatment. After this nadir of S-nitrosylation, eNOS is progressively renitrosylated back to basal levels over ∼60 min. Treatment of BAEC with insulin, a less robust eNOS agonist (30Takahashi S. Mendelsohn M.E. J. Biol. Chem. 2003; 278: 30821-30827Abstract Full Text Full Text PDF PubMed Scopus (170) Google Scholar), produces comparable results (Fig. 1B). We chose to focus on analyses of eNOS S-nitrosylation responses to VEGF, because this agonist has been characterized extensively as a potent eNOS agonist in these cells (31Dimmeler S. Fleming I. Fisslthaler B. Hermann C. Busse R. Zeiher A.M. Nature. 1999; 399: 601-605Crossref PubMed Scopus (3047) Google Scholar). eNOS Renitrosylation Requires NO Synthase Activity—The renitrosylation of eNOS after agonist-modulated denitrosylation suggested to us that eNOS S-nitrosylation is dependent on eNOS activity. To test the dependence of eNOS S-nitrosylation on NOS activity, we pretreated BAEC with the NOS inhibitor N-methyl-l-arginine (l-NMA, 5 mm, 1 h) and then added VEGF (10 ng/ml) or its vehicle for 1 h before harvesting eNOS for biotin switch (32Griffith O.W. Kilbourn R.G. Methods Enzymol. 1996; 268: 375-392Crossref PubMed Google Scholar). Incubation of BAEC with l-NMA neither substantively decreases the level of basal eNOS S-nitrosylation nor blocks agonist-induced denitrosylation. However, as shown in Fig. 1C, l-NMA significantly attenuated the extent of renitrosylation of the enzyme after agonist treatment (n = 3, p < 0.05). eNOS Phosphorylation, Subcellular Targeting, and S-Nitrosylation—In parallel with our analyses of dynamic changes in eNOS S-nitrosylation, BAEC were also analyzed for VEGF-modulated changes of eNOS phosphorylation at Ser1179. Phosphorylation at Ser1179 is correlated with VEGF-modulated increases in eNOS activity and can be blocked by the phosphoinositide 3-kinase inhibitor wortmannin (29Fulton D. Gratton J.P. McCabe T.J. Fontana J. Fujio Y. Walsh K. Franke T.F. Papapetropoulos A. Sessa W.C. Nature. 1999; 399: 597-601Crossref PubMed Scopus (2232) Google Scholar, 31Dimmeler S. Fleming I. Fisslthaler B. Hermann C. Busse R. Zeiher A.M. Nature. 1999; 399: 601-605Crossref PubMed Scopus (3047) Google Scholar). As shown in Fig. 2A, the time course of eNOS phosphorylation at Ser1179 was inversely related to the extent of eNOS S-nitrosylation (Fig. 1A). Because agonist-induced changes in the pattern of eNOS phosphorylation are influenced by eNOS subcellular localization (19Gonzalez E. Kou R. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2002; 277: 39554-39560Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar), we decided to explore the relations among eNOS phosphorylation, subcellular targeting, and S-nitrosylation. We first explored whether mutation of important eNOS phosphoresidues to “phosphonull” Ala or “phosphomimetic” Asp had any effect on eNOS S-nitrosylation. After immunoprecipitation of the eNOS mutants from lysates prepared from transfected COS-7 cells, eNOS was biotinylated by the biotin switch method, concentrated using streptavidin-agarose, and eluted by boiling in SDS-sample buffer (22Jaffrey S.R. Snyder S.H. Sci. STKE 2001. 2001; : PL1Google Scholar). S-Nitrosylated eNOS was then detected in immunoblots probed with eNOS antibodies. As shown in Fig. 2B, biotin switch analyses of eNOS mutants at Ser116 and Ser1179 (sites of agonist-modulated dephosphorylation and phosphorylation, respectively) revealed no substantive differences in S-nitrosylation compared with wild-type eNOS (Fig. 2B) (16Kou R. Greif D. Michel T. J. Biol. Chem. 2002; 277: 29669-29673Abstract Full Text Full Text PDF PubMed Scopus (160) Google Scholar, 29Fulton D. Gratton J.P. McCabe T.J. Fontana J. Fujio Y. Walsh K. Franke T.F. Papapetropoulos A. Sessa W.C. Nature. 1999; 399: 597-601Crossref PubMed Scopus (2232) Google Scholar, 31Dimmeler S. Fleming I. Fisslthaler B. Hermann C. Busse R. Zeiher A.M. Nature. 1999; 399: 601-605Crossref PubMed Scopus (3047) Google Scholar). We also explored the influence of eNOS targeting on enzyme S-nitrosylation by transfecting COS-7 cells with cDNA encoding wild-type eNOS, which is membrane-associated, with an acylation-deficient eNOS mutant (eNOSmyr–) that is targeted exclusively to the cytosol or with a palmitoylation-deficient eNOS (eNOSpalm–) that has an intermediate phenotype (19Gonzalez E. Kou R. Lin A.J. Golan D.E. Michel T. J. Biol. Chem. 2002; 277: 39554-39560Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). As shown in Fig. 2B, when eNOSmyr– is immunoprecipitated from transfected COS-7 cell lysates and processed by the biotin switch method, almost no S-nitrosylation signal is detected, in contrast to the robust signal of wild-type eNOS (91 ± 0.1% decrease in S-nitrosylation for the eNOSmyr– mutant relative to wild-type eNOS, n = 3, p < 0.001). The palmitoylation-deficient eNOS (eNOSpalm–) mutant showed levels of S-nitrosylation that were not substantively different from the wild-type eNOS in transfected COS-7 cells (Fig. 2B). There is no substantive difference between the NO synthase activity levels of wild-type eNOS and eNOSmyr– in lysates prepared from transfected COS-7 cells (data not shown and Refs. 12Busconi L. Michel T. J. Biol. Chem. 1993; 268: 8410-8413Abstract Full Text PDF PubMed Google Scholar, 33Lin S. Fagan K.A. Li K.X. Shaul P.W. Cooper D.M. Rodman D.M. J. Biol. Chem. 2000; 275: 17979-17985Abstract Full Text Full Text PDF PubMed Scopus (155) Google Scholar). S-Nitrosylation Inhibits eNOS Activity—We next performed experiments using eNOS purified to homogeneity from a baculovirus expression system to explore whether S-nitrosylation regulates eNOS activity (26Busconi L. Michel T. Mol. Pharmacol. 1995; 47: 655-659PubMed Google Scholar). As shown in Fig. 3, incubation of purified eNOS with 10 mm ascorbate (which reduces nitrosothiols) enhances eNOS activity to 120 ± 0.1% control sample activity (n = 4, p < 0.001) while reducing eNOS S-nitrosylation intensity to 51 ± 0.1% control (n = 3, p < 0.02). The NO donor DEA/NO (50 μm) attenuates eNOS enzyme activity to 65 ± 0.1% control levels (n = 4, p < 0.001), whereas eNOS S-nitrosylation levels increased to 117 ± 0.1% control sample intensity (n = 3, p < 0.02). The addition of ascorbate (10 mm) to the reaction containing S-nitrosylated eNOS significantly relieved the enzyme inhibition induced by DEA/NO, returning it to 81 ± 0.1% of the control (eNOS plus aged DEA/NO) activity level (n = 4, p < 0.001) and reducing S-nitrosylation levels to 89 ± 0.2% of the control (n = 3, p < 0.02). Identification of eNOS S-Nitrosylated Cysteines—In our next series of experiments, we sought to identify the specific sites of eNOS S-nitrosylation. The zinc-tetrathiolate motif present in eNOS has been proposed as an S-nitrosylation target (4Raman C.S. Li H. Martasek P. Kral V. Masters B.S. Poulos T.L. Cell. 1998; 95: 939-950Abstract Full Text Full Text PDF PubMed Scopus (577) Google Scholar), and a recent report (34Ravi K. Brennan L.A. Levic S. Ross P.A. Black S.M. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 2619-2624Crossref PubMed Scopus (199) Google Scholar) has confirmed this possibility, showing that eNOS releases Zn2+ and is inactivated by treatment with exogenous NO donors in vitro. We hypothesized that the cysteines comprising eNOS zinc-tetrathiolate complex (Cys96 and Cys101 of each subunit) are the targets of endogenous S-nitrosylation and used PCR-directed mutagenesis to construct site-specific mutants of eNOS, changing the cysteines individually and collectively to serine. We used the biotin switch method to test for S-nitrosylation of eNOS mutants at Cys96 and/or Cys101. As shown in Fig. 4A,