Modulation by Peroxynitrite of Akt- and AMP-activated Kinase-dependent Ser1179 Phosphorylation of Endothelial Nitric Oxide Synthase*

Peroxynitrite (ONOO−), a nitric oxide-derived oxidant, uncouples endothelial nitric oxide synthase (eNOS) and increases enzymatic production of superoxide anions (O 2 ⨪ ) (Zou, M. H., Shi, C., and Cohen, R. A. (2002) J. Clin. Invest. 109, 817–826). Here we studied how ONOO−influences eNOS activity. In cultured bovine aortic endothelial cells (BAEC), ONOO− increased basal and agonist-stimulated Ser1179 phosphorylation of eNOS, whereas it decreased nitric oxide production and bioactivity. However, ONOO−strongly inhibited the phosphorylation and activity of Akt, which is known to phosphorylate eNOS-Ser1179. Moreover, expression of an Akt dominant-negative mutant did not prevent ONOO−-enhanced eNOS-Ser1179 phosphorylation. In contrast to Akt, ONOO− significantly activated 5′-AMP-activated kinase (AMPK), as evidenced by its increased Thr172 phosphorylation as well as increased Ser92 phosphorylation of acetyl-coenzyme A carboxylase, a downstream target of AMPK. Associated with the increased release of O 2 ⨪ , ONOO− significantly increased the co-immunoprecipitation of eNOS with AMPK. Further, overexpression of the AMPK-constitutive active adenovirus significantly enhanced ONOO− up-regulated eNOS-Ser(P)1179. In contrast, overexpression of a dominant-negative AMPK mutant attenuated the ONOO−-enhanced eNOS-Ser1179phosphorylation as well as O 2 ⨪ release. We conclude that ONOO− inhibits Akt and increases AMPK-dependent Ser1179 phosphorylation of eNOS resulting in enhanced O 2 ⨪ release.

nitric oxide (NO), is formed during sepsis, inflammation, diabetes, ischemia-reperfusion, and atherosclerosis and contributes to these pathophysiological processes (13)(14)(15)(16)(17). In a recent study we showed that ONOO Ϫ oxidizes the zinc-thiolate cluster of eNOS, inhibiting its NO synthetic activity but increasing the NADPH oxidase activity and O 2 . production by the enzyme (17).
In the present study, we further examined how ONOO Ϫ regulates eNOS activity. We show that in BAEC cells, ONOO Ϫ enhances O 2 . release partially by increasing eNOS-Ser 1179 phosphorylation while inhibiting NO release from the uncoupled enzyme. ONOO Ϫ inhibits Akt/PKB (protein kinase B) activity but, in contrast, activates AMPK as shown by increased Thr 172 phosphorylation of AMPK and Ser 92 phosphorylation of ACC, a downstream target of AMPK. In addition, ONOO Ϫ significantly increases the association of eNOS and AMPK. Furthermore, by expressing an AMPK dominant-negative mutant, the enhanced activity of AMPK was demonstrated to be required for the increased release of O 2 . by the eNOS.
Adenoviral Transfection-BAEC were transfected with adenovirus expressing green fluorescent protein, a dominant-negative mutant Akt (Akt-DN) (18), constitutively active AMPK (19), or dominant-negative AMPK (mutation of Lys 45 to Arg) (20). BAEC were transfected in medium with 2% fetal calf serum overnight. The cells were then washed and incubated in fresh EGM medium with 2% fetal calf serum for an additional 24 h prior to experimentation. Using these conditions, transfection efficiency was typically Ͼ80% as determined by green fluorescence protein expression.
Peroxynitrite Synthesis-ONOO Ϫ was synthesized using a quenchedflow reaction as previously described (17). The concentrations of ONOO Ϫ were determined spectrally in 0.1 M NaOH (⑀ 302 ϭ 1670 M Ϫ1 S Ϫ1 ). ONOO Ϫ was diluted in 0.1 M NaOH before use to avoid a sharp shift of pH.
Treatment of Bovine Aortic Endothelial Cells with ONOO Ϫ -Confluent BAEC were treated with ONOO Ϫ as described previously (17). 950 l of 100 mmol/liter HEPES buffer, pH 7.4, was added to confluent cells in 6-well plates. 50 l of concentrated ONOO Ϫ in 0.1 mol/liter NaOH was evenly but quickly added into 6-well plates in rapidly rotating orbital shakers at room temperature. There was no pH shift during treatment with ONOO Ϫ . The same volumes of 0.1 mol/liter NaOH or decomposed ONOO Ϫ (ONOO Ϫ was first decomposed in 1 mol/liter Tris buffer, pH 7.4, and kept for 5 min or overnight) were used as controls.
Detection of Ser 1179 Phosphorylation of eNOS in BAEC-eNOS is a homodimeric enzyme. Under native conditions, eNOS dimers, which are sensitive to temperature but resistant to SDS, run as dimers (ϳ270 kDa) in low temperature SDS-PAGE (17). When denatured by boiling, eNOS dimers become dissociated and run as monomers with a molecular mass of ϳ135 kDa in room temperature SDS-PAGE. Unless otherwise indicated, to distinguish the effects of ONOO Ϫ on SDS-resistant eNOS dimers, unboiled samples were examined by low temperature SDS-PAGE.
The low temperature SDS-PAGE was performed according to Ref. 17. After being washed twice with ice-cold PBS buffer, ONOO Ϫ -treated BAEC cells were lysed and sonicated twice. Protein lysates were mixed with 3-fold loading buffer and loaded on 6% gels without boiling. Proteins were separated either with low temperature SDS-PAGE under reducing conditions (with ␤-mercaptoethanol). Proteins were blotted onto nitrocellulose membranes and incubated with a polyclonal antibody against phospho-Ser 1179 of eNOS (eNOS-Ser(P) 1179 , 1:1000, 4°C overnight). Ser 1179 phosphorylation of eNOS was visualized by using the appropriate horseradish peroxidase-linked secondary antibodies and ECL reagents. To test the specificity of antibody to eNOS-Ser(P) 1179 for native proteins, bovine recombinant eNOS purified from SF9 cells were treated with ONOO Ϫ , and no staining was found with ONOO Ϫtreated or non-treated recombinant eNOS, indicating that ONOO Ϫ did not increase nonspecific binding of eNOS with the antibody.
For room temperature SDS-PAGE, cell extracts were mixed with ␤-mercaptoethanol-containing Laemmli buffer (3ϫ) and boiled for 10 min. Proteins were separated at room temperature and Western blotted onto nitrocellulose. eNOS-Ser(P) 1179 was detected as described above.
Assays of L-Arginine Uptake and eNOS Activity-The nitric oxide synthase activity was assayed as described previously (18). 5 min after being treated with ONOO Ϫ , BAEC cells in 6-well plates were washed twice with 2 ml PBS buffer, pH 7.5, and 1 ml of PBS buffer with 0.1 mmol/liter CaCl 2 was added. The cells requiring L-NAME were preincubated with L-NAME (500 mol/liter) for 60 min. The eNOS activity was assayed by incubating BAEC with 10 M L-arginine plus 5 Ci of L-[ that, the cells were lysed with 250 l of 100% ethanol for 3 min and were added with 2 ml of ice-cold stop buffer (20 mM sodium acetate, pH 5.5, 1 mM L-citrulline, 2 mM EDTA, 2 mM EGTA). The lysate in stop buffer was then subjected to anion exchange chromatography using 2 ml of Dowex AG50W-X8 columns (0.4 g/ml, Bio-Rad) pre-equilibrated with stop buffer. L-Citrulline was eluted three times with 1 ml of stop buffer, and the eluent was collected for the determination of L-[ 3 H]citrulline by liquid scientillation counting. Data is reported as the extent of the conversion of L-[ 3 H]arginine to L-[ 3 H]citrulline that is sensitive to pre-treatment of the BAEC for 30 min with L-NAME and expressed as percent inhibition.
Cyclic GMP Assay-Confluent BAEC cells were treated with ONOO Ϫ (50 mmol/liter). After being washed twice with 3 ml of PBS buffer, cells were stimulated with agonists for 15 min. Cells were scraped with cell scrapers on ice and quickly frozen Ϫ80°C before assay. The cellular cGMP contents were assayed by using ELISA kits obtained from Cayman Chemicals (Ann Arbor, MI) as described previously (16).
Akt Activity Assay-Akt activity was assayed by using an Akt kinase assay kit obtained from Cell Signaling Inc. (Beverly, MA), as described by the supplier. An antibody to Akt was used to selectively immunoprecipiate Akt from cell lysates. The resulting immunoprecipitates were then incubated with a GSK fusion protein in the presence of ATP and kinase buffer. This allows immunoprecipitated Akt to phosphorylate GSK-3. Phosphorylation of GSK was measured by Western blotting using a phospho-GSK-3␣/␤ (Ser21/9) antibody and used as an index of Akt activity.
Detection of O 2 . Release in Cultured Endothelial Cells-The release of O 2 . in cultured BAEC cells was assayed by superoxide dismutase-inhibitable cytochrome c reduction by measuring the absorbance at 550 nm (⑀ 550 ϭ 21 mM Ϫ1 s Ϫ1 ) as described previously (16,17). Immunoprecipitation and Western Blotting of eNOS and AMPK-Coimmunoprecipitation studies for determining the interaction of AMPK with eNOS were performed as described previously (16,17). Confluent BAEC cells in 6-well dishes were pretreated with ONOO Ϫ as described above. The cells were washed three times with cold PBS buffer (pH 7.4) immediately after being treated with ONOO Ϫ , and 100 l of lysis buffer was added. Cell lysates (1 mg/ml) were incubated with the antibodies against eNOS (15 g/ml) or AMPK (15 g/ml) overnight. Cell proteins were loaded with 3-fold sample buffer and boiled for 5 min. The proteins were loaded on 6% SDS-PAGE, transferred onto nitrocellulose membranes, and blotted with primary antibody overnight at 4°C. The proteins were visualized by using the appropriate horseradish peroxidaselinked secondary antibodies and ECL reagents.

RESULTS AND DISCUSSION
Increase of eNOS-Ser 1179 Phosphorylation by ONOO Ϫ -Ser 1179 phosphorylation of eNOS has been widely considered as an important mechanism for increasing NO production under conditions such as fluid shear stresses (21,22,23), growth factors such as vascular endothelial growth factor (24,25), insulin-like growth factor-1 (IGF-1) (24), estrogen (26,27), or oxidants such as hydrogen peroxide (H 2 O 2 ) (18). Mutation of Ser 1179 to alanine attenuates agonist-induced NO release (21) because phosphorylation of Ser 1179 in the reductase domain of eNOS enhances the rate of electron flux from the reductase to the oxygenase domain of the enzyme, thus increasing the rate of NO synthesis (28).
All three NOS isoforms are dimeric enzymes comprised of two identical monomers that are bridged by a zinc tetrathiolate (Zn-S 4 ) cluster. Under native conditions, eNOS, which is sensitive to temperature but resistant to SDS, runs as a dimer under reducing conditions in low temperature SDS-PAGE (17). When denatured, eNOS dimers become dissociated and run as monomers with a molecular mass of ϳ135 kDa in normal temperature SDS-PAGE.
ONOO Ϫ significantly increased O 2 . release in BAEC (17), but the mechanism remained unknown. We first addressed whether or not eNOS Ser 1179 phosphorylation contributed to ONOO Ϫ -induced O 2 . release. BAEC were treated with ONOO Ϫ , and samples were boiled in the presence of ␤-mercaptoethanol for 10 min, and eNOS was separated by room temperature SDS-PAGE. As shown in Fig. 1a, ONOO Ϫ , but not decomposed ONOO Ϫ , significantly up-regulated eNOS-Ser(P) 1179 in BAEC.
Our previous studies (17) demonstrated that ONOO Ϫ oxidizes the ZnS 4 cluster, resulting in zinc release and formation of disulfide bonds between the monomeric units. The zincdepleted eNOS dimers were dissociated under reducing conditions as observed with low temperature SDS-PAGE. In BAEC under the conditions of this study both eNOS dimer, corresponding to a 260-kDa protein, and eNOS monomer (135 kDa) were observed, indicating that the ZnS 4 cluster of the enzyme is partially oxidized (Fig. 1b). We first addressed whether or not the integrated status of eNOS protein (i.e. dimer versus monomer) affected the phosphorylation of Ser 1179 . As shown in Fig. 1b, treatment of BAEC with ONOO Ϫ caused a dose-dependent decrease in eNOS dimers and increase in eNOS monomers. On the other hand, ONOO Ϫ caused a dose-dependent increase in eNOS-Ser(P) 1179 (Fig. 1b). Despite the fact that fewer eNOS dimers were detected by low temperature SDS-PAGE after treatment with ONOO Ϫ , the amount of phosphorylated eNOS dimer was significantly increased. ONOO Ϫ decomposed in 1 M Tris, pH 7.5, for 10 min before addition to the cells did not affect the Ser 1179 phosphorylation of eNOS (not shown). Thus, ONOO Ϫ has at least two effects on eNOS; it oxidizes the ZnS 4 cluster and increases the phosphorylation of Ser 1179 . As shown in Fig. 1b, eNOS-Ser(P) 1179 was mainly detected in eNOS dimers in contrast to weak staining in eNOS monomers. The reason eNOS monomers did not stain with the antibody against Ser 1179 -P is unknown. It might be due to a lowered affinity of eNOS monomers with the antibody against phosphorylated Ser 1179 compared with eNOS dimers. Similarly, we have previously found (17) that eNOS monomers have a lowered affinity with the antibody against eNOS.
As shown in Fig. 1c, the increased Ser 1179 phosphorylation of eNOS was detected in cells harvested immediately after being treated with ONOO Ϫ , but phosphorylation persisted for up to 3 h after treatment with ONOO Ϫ .
Inhibition of NO Production and Bioactivity by ONOO Ϫ -The catalytic mechanisms of NOS involve flavin-mediated electron transport from C-terminal-bound NADPH to the N-terminal heme center where oxygen is reduced and incorporated into the guanidine group of L-arginine giving rise to NO and Lcitrulline. Therefore, the formation of L-citrulline can be used as an index of NO release.
Because increased eNOS-Ser 1179 phosphorylation is thought to increase NO production, we next determined the effect of ONOO Ϫ on NO production and bioactivity in BAEC by monitoring the release of L-citrulline and cyclic GMP. As shown in Fig.  2a, ONOO Ϫ significantly inhibited the production of NO. ONOO Ϫ did not affect 3 H-arginine uptake (data not shown), and supplementation with exogenous L-arginine (up to 1 mmol/liter for 3 h) did not restore ONOO Ϫ -induced inhibition of NO release (data not shown), indicating that the effect of ONOO Ϫ is not due to decreased uptake or availability of L-arginine.

FIG. 1. ONOO ؊ up-regulates eNOS-Ser 1179 phosphorylation.
Cultured BAEC were treated with ONOO Ϫ (0 -50 mol/liter), decomposed ONOO Ϫ (ONOO Ϫ were added in 1 M Tris, pH 7.4, for 10 min before being added into samples), or the NaOH vehicle (100 mmol/liter) as described under "Materials and Methods." 10 min after treatment cells were harvested and lysed. Proteins in boiled samples were separated under reducing or non-reducing conditions by SDS-PAGE at room temperature (Fig. 1a) or in low temperature SDS-PAGE with unboiled samples (Fig. 1, b, c, and d). a, representative blot of eNOS-Ser(P) 1179 in room temperature SDS-PAGE with boiled samples. BAEC cells were boiled for 10 min in the presence of ␤-mercaptoethanol. Proteins were separated at room temperature by SDS-PAGE and Western blotted. eNOS-Ser(P) 1179 was detected at 135-kDa as described under "Materials and Methods." b, representative blots of reducing gels of eNOS dimer and monomers and Ser 1179 phosphorylation of eNOS obtained from cells immediately after exposure to ONOO Ϫ . eNOS dimers and monomers were separated by low temperature SDS-PAGE (6%) under reducing gels (ϩ␤-mercaptoethanol, ϩ␤-ME). The blot represents those from ten independent experiments. c, time-dependent phosphorylation of eNOS-Ser(P) 1179 by ONOO Ϫ in cultured BAEC. Increased phosphorylation was observed immediately after and for up to 180 min following exposure to ONOO Ϫ (50 mol/liter). The blot represents three individual experiments. D, decomposed ONOO Ϫ ; Di-eNOS, eNOS dimer; eNOS, eNOS monomer.  4). eNOS dimers and monomers were separated by low temperature SDS-PAGE (6%) under reducing (ϩ␤-ME) conditions. eNOS phosphorylation was observed both with and without ONOO Ϫ treatment following the agonists but was greater following ONOO Ϫ treatment. c, ONOO Ϫ treatment decreased cyclic GMP content following agonist stimulation of BAEC. Cyclic GMP was assayed as described under "Materials and Methods." ONOO Ϫ significantly decreased agonist-induced cyclic GMP production (n ϭ 9, *p Ͻ 0.01).
The availability of BH 4 is essential for the NO synthetic activity of eNOS, although its role in NOS catalysis is not understood. There is evidence that ONOO Ϫ depletes BH 4 and thereby promotes eNOS uncoupling. To investigate whether oxidation of BH 4 was involved, sepiapterin (100 mol/liter), which is converted to BH 4 by the salvage pathway, or BH 4 (100 mol/liter) was immediately added to the cells after ONOO Ϫ treatment and incubated for an additional 3h. However, neither sepiapterin nor BH 4 restored the NO synthetic activity (L-citrulline formation) of eNOS in ONOO Ϫ -treated cells (data not shown), suggesting that depletion of BH 4 by ONOO Ϫ was unlikely to be involved in ONOO Ϫ -mediated eNOS dysfunction. Furthermore, BH 4 (100 mol/liter), which was added to cells immediately before ONOO Ϫ addition, did not block ONOO Ϫinduced inhibition on L-citrulline formation (data not shown). Because our previous results (17) have demonstrated that supplementation with BH 4 (100 mol/liter) did not prevent loss of recombinant eNOS dimers after ONOO Ϫ exposure (17), these results indicated that a zinc-thiolate cluster of eNOS is a preferential target for ONOO Ϫ .
We therefore addressed whether or not ONOO Ϫ activated Akt and thereby phosphorylated eNOS-Ser 1179 . Surprisingly, low concentrations of ONOO Ϫ significantly inhibited Akt activity as indicated by decreased Akt-dependent GSK-3 phosphorylation (Fig. 3a). Furthermore, overexpression of an Akt dominant-negative mutant, which prevents Akt activation (18,25), did not attenuate ONOO Ϫ -stimulated eNOS-Ser(P) 1179 (Fig.  3b). In addition, ONOO Ϫ decreased the phosphorylation of Akt-Ser 473 (Fig. 3c), which is involved in the regulation of Akt activity. Taken together, these results indicate that eNOS-Ser(P) 1179 caused by ONOO Ϫ was accompanied by decreased Akt activity and was therefore not dependent on Akt.
Role of Guanylyl Cyclase and Protein Kinase A and G in ONOO Ϫ -induced eNOS Phosphorylation-Although it is several-fold less potent than NO in doing so, ONOO Ϫ does activate guanylyl cyclase and consequently protein kinase G (31). We therefore addressed whether or not ONOO Ϫ increased eNOS-Ser(P) 1179 by activating protein kinase G. As shown in Fig. 3d, ODQ (1 mol/liter), an inhibitor for cyclic GMP production, did not decrease the ONOO Ϫ -induced eNOS-Ser(P) 1179 . Zaprinast (50 mol/liter), which inhibits phosphodiesterase, preventing the degradation of cyclic GMP, also did not influence the effect of ONOO Ϫ (data not shown). Furthermore, inhibition of protein kinase G by KT5823 (1 mol/liter) did not affect ONOO Ϫ induced eNOS-Ser(P) 1179 , indicating that Ser 1179 phosphorylation of eNOS by ONOO Ϫ is independent of protein kinase G.
Protein kinase A is reported to phosphorylate eNOS-Ser 1179 in response to increased shear stress (22,23). Preincubation with H89 (1 mol/liter), a potent inhibitor of protein kinase A, did not attenuate ONOO Ϫ -induced eNOS-Ser 1179 phosphorylation, suggesting that activation of protein kinase A is unlikely to be responsible.
Activation of AMPK by ONOO Ϫ -Chen et al. (11) have reported that AMPK phosphorylates eNOS-Ser 1179 in vitro and in ischemic cardiac myocytes. To investigate whether or not ONOO Ϫ causes phosphorylation of eNOS-Ser 1179 by activating AMPK, BAEC were treated with different concentrations of ONOO Ϫ , and cell proteins were stained with a specific antibody against phosphorylated Thr 172 of AMPK that is reported to be essential for AMPK activity (7). As shown in Fig. 4a, ONOO Ϫ dose-dependently increased the phosphorylation of AMPK-Thr 172 . Similar to phosphorylation of eNOS-Ser 1179 , ONOO Ϫ increased phosphorylation of AMPK-Thr 172 immediately after treatment, lasting at least 30 min (Fig. 4b). Moreover, ONOO Ϫ also increased the prolonged phosphorylation of Ser 79 of ACC (Fig. 4b), a downstream target that is known to be phosphoryl- ated by AMPK (2,9,10). This suggests that ONOO Ϫ treatment activates AMPK in BAEC.
AMPK-dependent eNOS Phosphorylation Following ONOO Ϫ -Further evidence for the role of AMPK-dependent eNOS-Ser 1179 phosphorylation was obtained in experiments in which constitutively active AMPK (AMPK-CA) and dominantnegative AMPK (AMPK-DN) mutants were expressed with adenoviral vectors. As shown in Fig. 5a, overexpression of AMPK-CA, which alone slightly increased e-NOS-Ser 1179 phosphorylation (data not shown), did not influence the effect of ONOO Ϫ on eNOS-Ser(P) 1179 (11) have demonstrated that AMPK co-immunoprecipitates with eNOS. As shown in Fig. 5c, ONOO Ϫ increased the association of AMPK and eNOS. Similar results were obtained either with immunoprecipitation of eNOS and staining for AMPK or with immunoprecipitation of AMPK and staining for eNOS.
Ser 1179 phosphorylation of eNOS has been widely considered to be an important mechanism for increased NO production under the influence of fluid shear stress (21)(22)(23)(24), growth factors (24 -27), or H 2 O 2 (18). Although mutation of Ser 1179 to alanine attenuates agonist-induced eNOS activity (21)  The cells, following exposure to ONOO Ϫ , were rinsed twice with 2 ml of PBS buffer, pH 7.4, and exposed to calcium ionophore A23187 (10 mol/liter) for 2 h. The O 2 . release was measured by superoxide dismutase-inhibitable cytochrome c reduction as described under "Materials and Methods" (n ϭ 6, # p Ͻ 0.05 versus control, *p Ͻ 0.05 versus ONOO Ϫ ). c, ONOO Ϫ increases the association of eNOS and AMPK. Increased staining was observed for AMPK in immunoprecipitates obtained with anti-eNOS antibody as well as for eNOS in immunoprecipitates obtained with the AMPK antibody. The blots represent those from 6 independent experiments. strated (16,17) for prostacyclin synthase. Because our previous studies (17) showed that in vivo, eNOS is partially oxidized in tissues of diabetic mice, it is highly likely that ONOO Ϫ regulates eNOS activity and its product formation in vivo. In addition, other regulators of eNOS function may be dependent on the redox status of the ZnS 4 cluster. For example, the mechanisms described here are likely important to understand the recent observation that inhibition of HSP90 with geldanamycin significantly increases O 2 . production while increasing Ser 1179 phosphorylation of eNOS (32). In summary, the main finding of the present study is that ONOO Ϫ inhibits Akt-dependent and increases AMPK-dependent phosphorylation of eNOS-Ser 1179 . ONOO Ϫ activates AMPK and increases the association of AMPK with eNOS, whereas it inhibits Akt-activity. AMPK-dependent phosphorylation of eNOS-Ser 1179 likely contributes to increased O 2 . and ONOO Ϫ production by eNOS. We conclude that phosphorylation of eNOS-Ser 1179 reflects only the activated state of eNOS rather than whether it can generate NO or O 2 . , which depends rather on the integrity of the ZnS 4 cluster of the protein.