Interaction of endothelial and neuronal nitric-oxide synthases with the bradykinin B2 receptor. Binding of an inhibitory peptide to the oxygenase domain blocks uncoupled NADPH oxidation.

Endothelial nitric-oxide synthase (type III) (eNOS) was reported to form an inhibitory complex with the bradykinin receptor B2 (B2R) from which the enzyme is released in an active form upon receptor activation (Ju, H., Venema, V. J., Marrero, M. B., and Venema, R. C. (1998) J. Biol. Chem. 273, 24025-24029). Using a synthetic peptide derived from the known inhibitory sequence of the B2R (residues 310-329) we studied the interaction of the receptor with purified eNOS and neuronal nitric-oxide synthase (type I) (nNOS). The peptide inhibited formation of L-citrulline by eNOS and nNOS with IC(50) values of 10.6 +/- 0.4 microM and 7.1 +/- 0.6 microM, respectively. Inhibition was not due to an interference of the peptide with L-arginine or tetrahydrobiopterin binding. The NADPH oxidase activity of nNOS measured in the absence of L-arginine was inhibited by the peptide with an IC(50) of 3.7 +/- 0.6 microM, but the cytochrome c reductase activity of the enzyme was much less susceptible to inhibition (IC(50) >0.1 mM). Steady-state absorbance spectra of nNOS recorded during uncoupled NADPH oxidation showed that the heme remained oxidized in the presence of the synthetic peptide consisting of amino acids 310-329 of the B2R, whereas the reduced oxyferrous heme complex was accumulated in its absence. These data suggest that binding of the B2R 310-329 peptide blocks flavin to heme electron transfer. Co-immunoprecipitation of B2R and nNOS from human embryonic kidney cells stably transfected with human nNOS suggests that the B2R may functionally interact with nNOS in vivo. This interaction of nNOS with the B2R may recruit the enzyme to allow for the effective coupling of bradykinin signaling to the nitric oxide pathway.

Nitric oxide (NO) 1 is a signaling molecule that is synthesized from L-arginine and O 2 by nitric-oxide synthases (NOS, EC 1.14.13.39) that exist in the three following isozymes (for a review see Refs. 1 and 2): nNOS (type I), eNOS (type III), and inducible NOS (type II). nNOS and eNOS are constitutively expressed and Ca 2ϩ -dependent, whereas inducible NOS is cytokine-inducible and Ca 2ϩ -independent. All three isozymes are homodimers, with each subunit consisting of two domains, an amino-terminal oxygenase domain containing a heme group and binding sites for the substrate L-arginine and the cofactor BH 4 and a carboxyl-terminal reductase domain with binding sites for NADPH and the flavins FAD and FMN. The two domains are joined by a binding region for CaM that has a dual function, activating electron transfer both from NADPH to the flavins and from the reductase to the oxygenase domain (3). Although inducible NOS binds CaM in nominally Ca 2ϩ -free solutions (Ͻ30 nM), free Ca 2ϩ in the micromolar range is essential for CaM binding to the constitutive isozymes, so that a rise in intracellular free Ca 2ϩ is a key signal for ligand-induced activation of nNOS and eNOS (for review see Ref. 1). In vascular endothelial cells, fluid shear stress triggers an alternative, apparently Ca 2ϩ -independent pathway of eNOS activation (for review see Refs. 4 and 5). Recent reports suggest that shear stress leads to activation of protein kinase Akt (protein kinase B) (6), which catalyzes phosphorylation of a specific serine residue of eNOS (S1179 in the human isoform) resulting in Ca 2ϩ -independent activation of the enzyme (7,8).
The Ca 2ϩ -dependent activation of eNOS in response to ligands such as bradykinin also appears to involve a fairly complex signaling cascade (for review see Ref. 9). Post-translational amino-terminal fatty acylation targets eNOS to caveolae in the plasma membranes of vascular endothelial cells and cardiac myocytes through a direct inhibitory interaction of the enzyme with caveolin-1 and -3, respectively (10 -12). Ligandinduced rises of intracellular Ca 2ϩ appear to promote a reversible Ca 2ϩ /CaM-mediated dissociation of the inhibitory eNOScaveolin complex resulting in transient eNOS activation (13). The ligand-induced dissociation of eNOS from caveolin might explain the translocation of the enzyme from the plasma membrane to cytosolic stuctures in response to stimulation of endothelial cells with bradykinin (14 -16). However, a recent report demonstrated that eNOS forms an inhibitory complex with the B2R from which the active enzyme is released in a Ca 2ϩ -dependent manner upon receptor activation (17), presumably * This work was supported in part by Grants 13586-MED, 13013-MED (to B. M.), 12191-MED (to K. S.), and 13793-MOB (to E. R. W.) from the Fonds zur Förderung der Wissenschaftlichen Forschung in Austria. 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.
‡ Recipient of a doctoral scholarship from the Austrian Academy of Sciences.
ʈ Supported by Grants HL57201 and HL62152 from the National Institutes of Health and by a grant-in-aid from the American Heart Association.
** To whom correspondence should be addressed. because of tyrosine phosphorylation of the eNOS interacting region of the B2R (18). In vitro binding assays with glutathione S-transferase fusion proteins corresponding to various domains of the B2R and functional studies with synthetic peptides identified the inhibitory eNOS binding site of the receptor as amino acid residues 310 -329, in intracellular domain 4 (17). Although eNOS dissociation from the receptor was Ca 2ϩ -dependent, the inhibitory B2R peptides had no effect on Ca 2ϩ /CaM binding (17), excluding a reciprocal regulation of the enzyme by B2R/ CaM as described for the caveolin cycle (19). More recently, similar interactions of eNOS have been described with intracellular domain 4 of other G-protein-coupled receptors, such as the angiotensin II AT 1 and endothelin-1 ET B receptors but not with the ATP P 2 Y 2 receptor (18). Thus, the reversible inhibitory interaction of eNOS with G-protein-coupled receptors appears to be a widespread signaling mechanism linking receptor activation to endothelial NO formation.
The present study was designed to elucidate the molecular mechanisms of NOS inhibition by B2R binding. For this purpose, we studied how a synthetic peptide corresponding to the sequence of the B2R interaction site affects the enzymatic activities and structural properties of purified eNOS and nNOS. It is shown that enzyme inhibition is most likely a consequence of peptide binding in close proximity to the prosthetic heme group, resulting in impaired reductive activation of molecular oxygen. The B2R peptides inhibited eNOS and nNOS with similar potency, indicating that both isoforms might interact with the receptor in cells. This hypothesis was corroborated in co-immunoprecipitation experiments performed with HEK 293 cells stably overexpressing human nNOS.

EXPERIMENTAL PROCEDURES
Materials-Rat nNOS was purified from recombinant baculovirusinfected Sf9 cells as described previously (20,21 (22). Human nNOS and eNOS cDNA clones were generous gifts from Dr. John Parkinson (Berlex Biosciences, Richmond, CA). NADPH was purchased from Pharma Waldhof GmbH (Dü sseldorf, Germany). Monoclonal anti-B2R antibody (Clone 20, mouse) was obtained from Transduction Laboratories (Lexington, KY) and peroxidase-conjugated anti-IgG antibodies and the enhanced chemiluminescence Western blotting detection kit were from Amersham Pharmacia Biotech. Polyclonal anti-nNOS antibodies were raised in rabbits against purified recombinant rat nNOS. Synthetic peptides were obtained from the Medical College of Georgia Biochemistry Core Facility and were Ͼ95% pure as determined by high performance liquid chromatography (17). The EasySelect Pichia expression kit was from Invitrogen, purchased through Bio-Trade (Vienna, Austria). Other materials for molecular biology were from New England Biolabs, Life Technologies, Inc., and Qiagen. Myoglobin (product number M1882), carbonic anhydrase (product number C4831), and all other chemicals were from Sigma.
Expression and Purification of Human eNOS-Recombinant human eNOS was expressed in the yeast Pichia pastoris as described elsewhere in detail. 2 Briefly, human wild type eNOS cDNA was cloned into the expression vector pPICZA (EasySelect Pichia expression kit, Invitrogen). The final DNA construct was linearized with PmeI, and the DNA was transformed into the yeast P. pastoris KM71. An overnight culture (30°C) was inoculated from a single colony in 10 ml of buffered minimal glycerol medium, buffered minimal glycerol medium containing 100 mM potassium phosphate (pH 6), 13.4 g/liter yeast nitrogen base without amino acids, 0.4 mg/liter biotin, 40 mg/l L-histidine, 1% glycerol (v/v). The next day this culture was used to inoculate 750 ml of buffered minimal glycerol medium (1:200), and the culture was grown overnight at 30°C to an A 600 of 4 -6. The cells were then harvested and resus-pended in 150 ml of buffered minimal methanol medium, containing 100 mM potassium phosphate (pH 6), 13.4 g/liter yeast nitrogen base without amino acids, 400 g/liter biotin, 40 mg/liter L-histidine, 5% methanol (v/v)) in the presence of 4 mg/liter hemin chloride and incubated for 24 h at 30°C to induce protein expression. Yeast cells were then harvested, resuspended in 50 mM Tris/HCl buffer (pH 7.4) containing 1 mM EDTA, 5% glycerol, 12 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 1 mM CHAPS, and homogenized by vigorous vortexing with glass beads (0.5 mm). Following removal of the glass beads, the homogenate was centrifuged at 30,000 ϫ g for 15 min to yield a supernatant used for protein purification by affinity chromatography as described previously for bovine eNOS (23). The final elution performed was with 20 mM Tris/HCl buffer (pH 7.4) containing 100 mM NaCl and 4 mM EGTA. The enzyme was stored at Ϫ70°C in the presence of 1 mM CHAPS. Protein was determined with the Bradford method using bovine serum albumin as a standard protein (24).
Transfection of HEK 293 Cells with Human nNOS-A HEK 293 cell line stably expressing human nNOS was established according to a protocol that will be described elsewhere in detail. 3 Briefly, the pcDNA3 vector (Invitrogen), which contains the neomycin resistance cassette and the cDNA for human nNOS, was used. Plasmid DNA was prepared for transfection by linearization with PvuI, followed by phenol/chloroform extraction and ethanol precipitation. HEK 293 cells were transfected at 40% confluence with 5 g of linearized plasmid DNA using SuperFect (Qiagen) or Tfx-50 (Promega). The cells were replated 48 h after transfection at 0.5-1 ϫ 10 6 cells per plate in 100 mm plates in medium containing 1 mg/ml G418. After 10 -12 days, isolated colonies of resistant cells were tested for activity, and the HEK/nNOS line was maintained in 250 g/ml G418.
NOS Steady-state Absorption Spectra-The spectral identity of the nNOS steady state during NADPH oxidation was investigated by rapidscan spectroscopy with a Bio-Sequential SX-17MV Stopped-Flow spectrofluorimeter (Applied Photophysics, Leatherhead, UK). Samples were illuminated with an ozone-free 150-Watt xenon lamp and detected with a photodiode array, equipped with a 200 -730-nm grating. Final conditions were as follows: 1.0 M nNOS, 15 M NADPH, 7.5 M (125 g/ml) CaM, 0.5 mM CaCl 2 , 50 mM potassium P i (pH 7.4), 0.2 mM CHAPS, 2.4 mM 2-mercaptoethanol, Ϯ 0.3 mM B2R 310 -329, at 25°C. All reagents were premixed in one syringe except for potassium P i , which was present in both syringes, and CaM, which was used to start the reaction and was added to the second syringe only. To obtain difference spectra, the pre-mixing spectra, representing the average of the spectra recorded separately for the contents of the two syringes, were subtracted from the spectra recorded subsequent to the addition of CaM.
BH 4 Binding-Radioligand binding studies using [ 3 H]BH 4 as high affinity ligand were performed as described previously with minor modifications (28,29). For saturation binding, nNOS (0.5 g) was incubated for 10 min at 37°C in 0.1 ml of a 50 mM triethanolamine/HCl buffer (pH 7.4) with 10 nM [ 3 H]BH 4 (ϳ14 nCi) and increasing concentrations of unlabeled BH 4 (0 -10 M) in the absence and presence of 30 M B2R 310 -329 peptide, followed by rapid vacuum filtration with the MultiScreen assay system from Millipore and determination of the radioactivity retained on the filters by liquid scintillation counting. Data were corrected for nonspecific binding determined in the presence of 1 mM unlabeled BH 4 .
Immunoprecipitation and Western Blotting-10 confluent 100-mm Petri dishes of HEK 293 cells overexpressing nNOS were washed with phosphate-buffered saline, harvested, and centrifuged for 5 min at 1,300 revolutions/min. The pellet was washed with phosphate-buffered saline and homogenized in 1 ml of chilled Tris buffer (20 mM) (pH 7.4) containing 2.5 mM EDTA and 1% Triton X-100 for 15 min at 4°C. The homogenate was centrifuged at 10,000 ϫ g for 25 min at 4°C. Anti-nNOS antiserum (400 l) or anti-B2R antibody (50 l) were then added to the lysate (400 and 100 l for precipitation with anti-nNOS and anti-B2R antibodies, respectively) and rocked for 2 h at 4°C, followed by addition of Tris buffer-equilibrated protein A-Sepharose (400 and 50 l for precipitation with anti-nNOS and anti-B2R antibodies, respectively) and rocking for another 3.5 h at 4°C. Precipitates and supernatants were separated by centrifugation. The pellets were washed twice with chilled buffer (50 mM Tris/HCl (pH 8.0), 500 mM NaCl, 0.1% Triton X-100), suspended in 60 l of Laemmli buffer, and boiled for 5 min at 95°C, followed by electrophoresis on 8% SDS-polyacrylamide gels as described (30). The separated proteins were transferred to nitrocellulose membranes by electroblotting for 90 min at 240 mA, followed by immunodetection with anti-B2R (1:3,000 dilution) or anti-nNOS (1: 5,000 dilution) antibodies, using horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG, respectively, and the enhanced chemiluminescence detection kit from Amersham Pharmacia Biotech.
Statistics and Data Evaluation-All affinity constants (EC 50 , IC 50 , K m , and K D ) as well as B max and V max values were obtained by nonlinear regression analyses of individual experiments followed by the calculation of arithmetic mean values. Unless otherwise indicated, data are given as means Ϯ S.E. of three independent experiments.

RESULTS
A synthetic peptide corresponding to the sequence of the B2R-eNOS interaction site (B2R 310 -329) (17) inhibited the catalytic activity of purified recombinant eNOS and nNOS in the arginine to citrulline conversion assay with IC 50 values of 10.6 Ϯ 0.4 and 7.1 Ϯ 0.6 M, respectively (Fig. 1). Similar results were obtained with a slightly longer peptide corresponding to the same region of the receptor (B2R 310 -334), whereas two peptides derived from other sequences of the B2R (330 -341 and 348 -364) had no effects (data not shown). The effects of the synthetic peptide on enzyme kinetic parameters of eNOS and nNOS are summarized in Table 1. Although the maximal activities of both isozymes were significantly reduced by the B2R 310 -329 peptide (10 M), there was no significant effect on the K m values for L-arginine, indicating that the peptide did not compete with substrate binding.
In a previous study we observed that the potency of an inhibitory peptide decreased dramatically with increasing NOS concentrations, rendering it impossible to study the mechanism of enzyme inhibition using biochemical methods requiring protein concentrations Ͼ50 nM (31). However, this effect was much less pronounced with B2R 310 -329. As shown in Fig. 2A, the inhibitory effect of the peptide (ϳ75% inhibition at 30 M) was not affected when the concentration of eNOS was increased from 15 to 225 nM, although a significantly higher potency of the peptide was apparent at the lowest NOS concentration tested (7.5 nM). Similarly, inhibition of nNOS was not very sensitive to protein concentration. The enzyme inhibition produced by 30 M B2R 310 -329 decreased from about 80% to 60% with nNOS concentrations increasing from 6.3 to 190 nM (Fig. 2B) It appeared conceivable that inhibition of NOS was due to an interference of the peptide with the shuttle of electrons from NADPH to the heme. This electron transport chain, which is catalyzed by the flavins FAD and FMN upon binding of Ca 2ϩ / CaM, can be assayed as CaM-dependent cytochrome c reductase activity of both nNOS (3,27) and eNOS (23). At concentrations producing almost maximal NOS inhibition (50 -100 M), the B2R peptide had no effect on the reductase activities of either eNOS or nNOS. Although a significant inhibition of cytochrome c reduction was observed at higher peptide concentrations (data not shown), these results suggested that the inhibition of L-citrulline formation cannot be explained by an interference of the peptide with electron transport or CaM binding.
Based on these observations, it appeared likely that the B2R 310 -329 peptide interfered with NADPH oxidation and/or heme reduction. The NADPH oxidase activity of NOS, which becomes apparent when nNOS or eNOS are activated with Ca 2ϩ /CaM in the absence of L-arginine or BH 4 , results in reduction of O 2 to superoxide/H 2 O 2 at the cost of NADPH. This uncoupled NADPH oxidation is catalyzed at significant rates by nNOS (32,33) and human eNOS, 2 whereas the reaction is down-regulated in other isozymes (23,34,35), most likely because the heme reduction potential remains below the critical value of Ϫ291 mV in the absence of substrate and pterin (36). As shown in Fig. 3, nNOS exhibited an NADPH oxidase activity of 1.51 Ϯ 0.004 mol ϫ min Ϫ1 ϫ mg Ϫ1 , which was inhibited by the B2R peptide with an IC 50 of 3.7 Ϯ 0.6 M. Similarly, the NADPH oxidase activity of human eNOS (200 Ϯ 40 nmol ϫ mol ϫ min Ϫ1 ϫ mg Ϫ1 ) was inhibited to 28.9 Ϯ 5.31% of controls in the presence of 0.1 mM of the peptide.
Together with the lack of effect on electron transfer, inhibition of uncoupled NADPH oxidation suggested that the B2R peptide blocked reduction of the heme or prevented subsequent binding and reductive activation of O 2 (37). To address this issue, we studied the effect of the B2R 310 -329 peptide on the NOS steady-state absorbance spectrum during uncoupled NADPH oxidation by rapid-scan spectroscopy. Mixing 1 M nNOS and 15 M NADPH with 7.5 M CaM in the presence of CaCl 2 caused rapid bleaching of the NADPH absorbance at 340 nm, corresponding to an oxidation rate of 1.35 M NADPH s Ϫ1 . When the same experiment was performed in the presence of 0.3 mM B2R 310 -329, the oxidation rate was slowed down (by about 70%) to 0.40 M NADPH s Ϫ1 . The absolute spectra were dominated by the absorbances of NADPH and the peptide and thus did not conclusively reveal the identity of the nNOS steady state, although a shoulder at ϳ418 nm was apparent in the absence but not in the presence of the peptide (not shown). Indeed, a pronounced increase of the absorbance at 424 nm, presumably due to accumulation of the oxyferrous complex absorbing at 417 nm (38), was apparent in the difference spectra that were obtained by subtracting the pre-mixing spectra from the steady-state spectra recorded immediately (0.1 s) after the addition of CaM (Fig. 4). In the presence of 0.3 mM B2R peptide, accumulation of the oxyferrous complex was largely inhibited without appearance of other absorbing species. In both cases, the steady-state spectra remained unchanged until the NADPH was completely oxidized, followed by re-oxidation Pterin analogs (39 -41) and other compounds (25) competing with BH 4 binding act as potent NOS inhibitors. Therefore, several sets of experiments were carried out to test for a possible interference of the B2R peptide with pterin activation of NOS. First, NOS activity was measured in the presence of increasing concentrations of BH 4 . As shown in Fig. 5A, pterin-free human eNOS 2 was activated by BH 4 with an EC 50 (23,30) indicated that the peptide (0.1 mM) did not affect the stability of eNOS and nNOS dimers analyzed with or without L-arginine/BH 4 (data not shown).
Based on the observation that the B2R 310 -329 peptide inhibited nNOS with similar potency as eNOS, we speculated that nNOS might interact with the B2R in cells as described previously for eNOS (17). To address this issue, we performed co-immunoprecipitation experiments with a HEK 293 cell line constitutively expressing the B2R and stably overexpressing human nNOS. 3 With the anti-B2R antibody we did not detect the 69-kDa band present in endothelial cells (17), indicating that the receptor is predominantly expressed as non-glycosylated protein in HEK cells. Immunoprecipitation was performed with anti-nNOS or anti-B2R antibodies, followed by Western blotting with anti-B2R and anti-nNOS antibodies, respectively. Fig. 6 shows that immunoprecipitation of nNOS resulted in co-precipitation of the B2R (42 kDa; left panel), whereas immunoprecipitation of the B2R led to co-precipitation of nNOS (160 kDa; right panel). These bands were not detectable in precipitates obtained from identical samples precipitated in the presence of buffer instead of the antibodies (data not shown).

DISCUSSION
The amino acid residues 310 -329 in the intracellular domain 4 of the B2R were shown to form an inhibitory complex with eNOS in endothelial cells (17). Using a synthetic peptide corresponding to the inhibitory sequence of the B2R, we have now elucidated the mechanism by which this interaction leads to NOS inhibition. We found that formation of L-citrulline by both recombinant human eNOS and rat nNOS was inhibited by similar concentrations of the peptide (3-30 M), suggesting that the target site of NOS has been conserved between the two constitutive isozymes. Our data indicate that the interaction with the receptor does not affect the transfer of electrons within the reductase domain. Because this electron shuttle is essentially dependent on Ca 2ϩ /CaM binding (27,42), it is also unlikely that the inhibitory effect is due to a competition of the peptide with CaM binding. However, we cannot exclude that such an effect is responsible for the inhibition of cytochrome c reduction observed at high (Ն0.1 mM) concentrations of the peptide. The inhibition was most likely not due to an interference with the binding or activity of BH 4 . Functional studies showed that formation of L-citrulline resulting from endogenously bound BH 4 was inhibited by the peptide to a similar degree as the maximal activity obtained with the pterin-saturated enzyme (cf. Fig. 5). Despite small effects on the apparent binding affinity of BH 4 , we obtained no conclusive evidence supporting a BH 4 -competitive type of inhibition by the B2R 310 -329 peptide. Moreover, the peptide had no effect on the stabilization of eNOS and nNOS dimers under denaturating conditions, further indicating that the inhibitory effect is unrelated to the binding or activity of BH 4 .
With respect to the mechanism of inhibition, the most relevant information was obtained from the effects of the B2R peptide on nNOS steady-state absorbance spectra recorded during uncoupled NADPH oxidation. During NOS catalysis the heme is thought to cycle between Fe(III), Fe(II), Fe(II)O 2 , and a number of unstable species (38). The steady-state absorbance spectra reflect the relative concentrations of these species, with the oxidized and reduced enzyme absorbing at 394 and 412 nm, respectively. For the oxyferrous complex, absorbance maxima of 417 and 427 nm were reported (37,38). Under control conditions, the steady-state absorbance spectrum exhibited a shoulder at 418 nm, whereas the absorbance difference spectrum was characterized by an increase in the aborbance at 424 nm, in accordance with previous observations (43). We ascribe these absorbance changes to the accumulation of the Fe(II)O 2 complex accumulating during uncoupled NADPH oxidation. In the presence of the B2R 310 -329 peptide (0.3 mM), which inhibited NADPH oxidation by about 70%, formation of the Fe(II)O 2 complex was suppressed without appearance of another species, indicating that the heme was not reduced under these conditions. These data strongly suggest that the B2R 310 -329 peptide affects NOS catalysis by interference with flavin-to-heme electron transfer.
Co-precipitation of the B2R and nNOS from HEK cell lysates under highly stringent conditions adds the B2R to a number of other functional and structural proteins that have been shown to interact with nNOS in both neurons and skeletal muscle. These proteins include ␣ 1 -syntrophin (44 -46), postsynaptic density proteins (45), the protein inhibitor of nNOS, PIN (47), caveolin-3 (48), the carboxyl-terminal PD2 ligand of nNOS (49), the muscle isoform of phosphofructokinase (50), and the Nmethyl-D-aspartate subtype of glutamate receptors (51). Except caveolin, which appears to interact with as yet unidentified sites in the reductase and oxygenase domains (52), most of these proteins bind to nNOS at its amino-terminal extension containing a PDZ motif, which is found in various structural proteins and may be important for scaffolding receptors, ion channels, and enzymes engaged in specific signaling processes (53). Thus, although the interacting site of NOS is not known, our data suggest that the B2R is the first example of a cellular protein interacting with nNOS via binding to the oxygenase domain in close proximity to the catalytic center of the enzyme.
Considering the widespread occurrence and biological actions of bradykinin in mammalian tissues, these results have considerable physiological implications. In addition to its actions as an endothelium-dependent vasodilator, bradykinin has several biological functions mediated by neuronal rather than endothelial NO. For example, activation of neuronal NO/cGMP signaling by bradykinin was suggested to be involved in the induction of pain (54,55), penile erection (56), noradrenaline release from cardiac sympathetic neurons (57), and neurogenic smooth muscle relaxation (58). Our results indicate that these processes are associated with a reversible interaction between the B2R and nNOS. This interaction may not only be important to recruit NOS at the site where it is needed but may additionally constitute an important cytoprotective mechanism. At suboptimal levels of L-arginine or BH 4 the uncoupling of NADPH oxidation from L-arginine oxidation results in the formation of the potent cytotoxins superoxide and peroxynitrite by both eNOS (59 -61) and nNOS (26,32,33,62,63). In light of the important pathophysiological role of the NO/superoxide reaction (64), formation of an inhibitory complex with the B2R, in which the uncoupled reaction is fully blocked, could be protective against endothelial dysfunction (65)(66)(67)(68), stroke (69 -72), and other disease states promoted by inflammatory oxidative stress.