Nitric Oxide Inactivates NADPH Oxidase in Pig Neutrophils by Inhibiting Its Assembling Process*

The effects of nitric oxide (NO) on superoxide (O·̄2) generation of the NADPH oxidase in pig neutrophils were studied. NO dose-dependently suppressed O·̄2 generation of both neutrophil NADPH oxidase and reconstituted NADPH oxidase. Effects of NO on NADPH-binding site and the redox centers including FAD and low spin heme in cytochromeb 558 and the electron transfer rates from NADPH to heme via FAD were examined under anaerobic conditions. Both reaction rates and the K m value for NADPH were unchanged by NO. Visible and EPR spectra of cytochrome b 558showed that the structure of heme was unchanged by NO, indicating that NO does not affect the redox centers of the oxidase. In reconstituted NADPH oxidase system, NO did not inhibit O·̄2 generation of the oxidase when added after activation. The addition of NO to the membrane component or the cytosol component inhibited the activity by 24.0 ± 5.3 or 37.4 ± 7.1%, respectively. The addition of NO during the activation process or to the cytosol component simultaneously with myristate inhibited the activity by 74.0 ± 5.2 or 70.0 ± 8.3%, respectively, suggesting that cytosol protein(s) treated with myristate becomes susceptible to NO. Peroxynitrite did not interfere with O·̄2 generation.

electron transport system, in which activation requires the assembly of three cytosolic regulatory proteins (p47 phox , p67 phox , and Rac1/Rac2) to membrane-bound cytochrome b 558 (10,11). Cytochrome b 558 is postulated to be a membranebound flavocytochrome with six-coordinated low spin heme and FAD as redox centers. The electrons provided by NADPH are thought to be transferred in a linear sequence, NADPH 3 FAD 3 heme (Fe 3ϩ ) 3 O 2 . The heme in cytochrome b 558 is assumed to be the terminal electron donor in the production of O 2 . from molecular oxygen due to its unusually low redox potential, Ϫ245 mV (12). Although the most plausible site of NADPH oxidase attacked by NO was suggested to be in membrane protein(s) (6), a detailed analysis of these effects has not been performed. Considering that nitrosyl-iron complex easily forms in heme-containing enzymes (3), the heme structure of cytochrome b 558 and the electron flux from substrate (NADPH) to redox centers, FAD and low spin heme, in NADPH oxidase should be examined to clarify the effects of NO on its O 2 . -generating activity.
In the present study, we examined the effects of NO on electron fluxes in neutrophil NADPH oxidase. Under aerobic conditions the effects of NO on O 2 . -generating activity of NADPH oxidase (reaction 1) was examined by the cytochrome c reduction method. In this study, we also employed the solubilized NADPH oxidase obtained from stimulated cells and measured its O 2 . -generating activity in the presence of NO.
Under anaerobic conditions the effects of NO on the electron transfer reaction in each redox center was examined: NADPH 3 FAD 3 exogenous electron acceptor, cytochrome c (reaction 2) and NADPH 3 FAD 3 cytochrome b 558 (reaction 3). Under both aerobic and anaerobic conditions, the binding ability of NO to the six-coordinated low spin heme (His-Fe 3ϩ -His) of cytochrome b 558 (reaction 4) was examined by both visible absorption and EPR spectroscopy. We also studied the effects of NO on the activation of the oxidase (reaction 5), i.e. assembly of cytosolic and membrane components using the reconstituted NADPH oxidase system. Finally, we studied the effects of peroxynitrite (ONOO Ϫ ) on NADPH oxidase to confirm that the results obtained above were caused by NO (and not by ONOO Ϫ ) because addition of NO in the presence of O 2 . produces ONOO Ϫ at nearly diffusion-limited rates (13) and because ONOO Ϫ is known to be a potent oxidant.

EXPERIMENTAL PROCEDURES
Materials-Myristic acid and arachidonic acid from Wako Pure Chemical (Tokyo, Japan) were dissolved in ethanol. Heptylthioglucoside was purchased from Dojindo Laboratories (Kumamoto, Japan). NADPH was from Oriental Yeast (Tokyo, Japan). Superoxide dismutase, cytochrome c (type VI from horse heart), and phorbol myristate acetate (PMA) 1 were purchased from Sigma. Other reagents were of * This work was supported in part by grants from the Ministry of Education, Science and Culture of Japan. 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.
NO Solution-Saturated NO solution was prepared according to the methods published elsewhere (7,14). Briefly, nitrogen gas was bubbled through 50 mM phosphate buffer (pH 7.0) for 20 min to remove dissolved oxygen. Then authentic NO gas (99%, Nippon Sanso Co. Ltd., Tokyo) that had passed through 1 M KOH to remove nitrogen dioxide was bubbled into the solution for 20 min. The concentration of NO was measured spectrophotometrically as previously reported (15).
Synthesis of ONOO Ϫ -Peroxynitrite was synthesized by the reaction of acidified H 2 O 2 and NaNO 2 in a quenched flow reactor with a subsequent stabilization induced by 1.5 M NaOH as described previously (16). The purity and concentration of ONOO Ϫ was checked by its absorbance at 302 nm. The production of ONOO Ϫ from NO and O 2 . was also confirmed by the increase in absorbance at 302 nm. Preparation of Neutrophils-Neutrophils were obtained from pig blood as reported previously (17) by the Conray-Ficoll differential density configuration method.
Isolation of Membrane and Cytosolic Components from Neutrophils-The cell pellet was frozen and thawed in the presence of phenylmethylsulfonyl fluoride at a final concentration of 1 mM and sonicated in ice-cold Krebs-Ringer phosphate buffer containing 0.34 M sucrose. Membrane vesicles and cytosol were obtained from sonicated cells by centrifugation (100,000 ϫ g for 60 min) (18). The membrane component of the NADPH oxidase was solubilized from the membrane vesicle with heptylthioglucoside (19) and used for further purification of cytochrome b 558 . The cytosol and solubilized NADPH oxidase from resting membrane vesicles were used in the reconstituted NADPH oxidase. All procedures were applied to both resting and stimulated cells. Cell stimulation was induced by myristate as reported previously (19).
Purification of Cytochrome b 558 -Cytochrome b 558 was purified from the membrane component (20) with slight modifications to avoid denaturation of heme in cytochrome b 558 . A buffer composed of 50 mM phosphate buffer (pH 7.0), 50 mM NaCl, 10% glycerol, and 0.6% heptylthioglucoside was used in the heparin-Sepharose column.
Assay of Superoxide Generation in Cells, Solubilized NADPH Oxidase, and Reconstituted NADPH Oxidase System (Reactions 1 and 5)-O 2 .
-generating activities of both neutrophils and solubilized NADPH oxidase were measured at the rate of cytochrome c reduction subtracting the rate in the presence of superoxide dismutase from that in the absence of superoxide dismutase in an assay medium containing 1.5 mM MgCl 2 and 50 mM phosphate buffer (pH 7.4) (21). The reaction was started by adding PMA for cells (37°C) or 0.1 mM NADPH for both the stimulated and reconstituted NADPH oxidase (25°C) as reported previously (22). The increase in absorption at 550 -540 nm was monitored in a Hitachi model 556 spectrophotometer. The measurements were performed in neutrophils, reconstituted NADPH oxidase, and solubilized NADPH oxidase obtained from stimulated cells in the presence or the absence of NO or ONOO Ϫ . In reconstituted NADPH oxidase, NO was added to membrane or cytosol or to both components before and after activation of the oxidase with myristate.

Measurement of Electron Flux through Redox Centers under Anaerobic Conditions (Reactions 2 and 3)-
The electron fluxes through two redox centers, FAD and cytochrome b 558 , in the electron transport chain of the NADPH oxidase system were examined. The electron flux from NADPH to FAD was evaluated by measuring the reduction rate of cytochrome c, which was utilized as an exogenous electron acceptor (reaction 2) according to the previously reported method (19). The electron flux to cytochrome b 558 from NADPH catalyzed by FAD was measured by following the absorbance at 558 nm spectrophotometrically (reaction 3) (19). The extent of reduction of cytochrome b 558 was calculated using the extinction coefficient at 558 -540 nm, 21.6 ϫ 10 3 liter mol Ϫ1 cm Ϫ1 (23). Spectral changes of cytochrome b 558 were observed every minute over the 400 -600 nm range. Strictly anaerobic conditions were achieved by using glucose (10 mM)/glucose oxidase (40 units/ml) in an airtight cuvette (19).
Calculations and Statistics-All data were presented as the means Ϯ S.D. of at least three experiments. Nonlinear least squares regression was used to calculate K m value.
Measurement of Oxygen Consumption-Oxygen consumption was measured with a Clark-type oxygen electrode (Yellow Spring Instrument, Yellow Spring, OH) in neutrophils (1 ϫ 10 7 cells in 1 ml of Krebs-Ringer phosphate buffer with 5 mM glucose) and solubilized NADPH oxidase obtained from stimulated neutrophils (30 g of protein in 1 ml of Krebs-Ringer phosphate buffer).
Detection of Nitrosyl Heme by EPR Spectrometry (Reaction 4)-To determine whether NO binds to the heme of cytochrome b 558 , i.e. formation of nitrosyl heme, purified cytochrome b 558 was reduced with dithionite, and its spectrum was observed in the presence or the ab-sence of NO. EPR spectra were recorded in a JEOL JES-FE X-band ESR spectrometer at 77-100 K. To examine the effects of the change in the spin state of the heme on NO binding, intact five-coordinated low spin cytochrome b 558 was denatured at 40°C for 120 min to form six-coordinated high spin heme. Typical EPR conditions were: microwave power, 5 mW; modulation amplitude, 10 gauss at 100 kHz; response, 0.3 s; sweep time, 4 min.

RESULTS AND DISCUSSION
Effect of Nitric Oxide on the Superoxide-forming Activity of Neutrophils (Reaction 1)- Fig. 1 shows the time course of cytochrome c reduction by PMA-stimulated neutrophils in the absence (trace B) or the presence of 25 M NO (trace C). As a control, trace A was measured in the absence of neutrophils. NO was added 2 min before the addition of cytochrome c to avoid the possible binding of NO to cytochrome c, because NO is reported to bind to a variety of hemoproteins such as hemoglobin (24) or iron-sulfur centers (25). As shown in Fig. 1A, cytochrome c reduction was not observed when NO was added to the assay buffer, indicating that there is no direct interaction of NO with cytochrome c. (These results were consistent with the report that the rate of NO reaction with cytochrome c is severalfold slower than that with oxygen or proteins under aerobic conditions (26).) 2 min after the addition of NO to the cell suspension, the residual concentration of NO was less than 1% of its initial concentration, which excludes the quenching of O 2 . First, we examined whether NO impairs the binding ability of NADPH to oxidase (reaction 1). K m values for NADPH of the solubilized NADPH oxidase were determined in the presence or the absence of NO. As shown in Table I Effect of NO on Electron Flux through Redox Centers (Reactions 2 and 3)-Electron transfer reactions from NADPH to FAD (reaction 2) and from NADPH to cytochrome b 558 (reaction 3) were studied using solubilized NADPH oxidase from stimulated neutrophils under anaerobic conditions (Table II). The cytochrome c-reducing activities of the NADPH oxidase obtained from both stimulated and resting cells were in the same range, either in the presence or the absence of 25 M NO, indicating that the electron flux from NADPH to FAD is not affected by NO. The interdomain electron transfer from NADPH through FAD to heme in cytochrome b 558 was measured by the reduction of cytochrome b 558 . The reduction rates of cytochrome b 558 of both stimulated and resting NADPH oxidase did not differ in the presence or the absence of NO, indicating that the electron flux between FAD and the low spin heme of cytochrome b 558 is not the site of impairment by NO.
The Detection of a Complex between NO and the Heme of Cytochrome b 558 (Reaction 4)-To examine whether NO binds to the heme of cytochrome b 558 or changes the spin state of the heme, nitrosyl-heme formation was measured by both visible absorption and EPR spectra of solubilized NADPH oxidase. Fig. 3 shows the difference spectra of cytochrome b 558 in NADPH oxidase from resting cells obtained by subtracting the oxidized spectrum from the dithionite-reduced spectrum in the absence (Fig. 3A) and the presence (Fig. 3, B and C) of NO. Spectra were measured after incubating the oxidase with NO for 5 and 15 min at 4°C (Fig. 3, B and C, respectively). No apparent difference was found in the ␣-peak (558 nm) or ␥-peak (427 nm) in the presence of NO. The same results were ob-tained again, when solubilized NADPH oxidase from stimulated cells was examined, suggesting that the structure of heme was not affected by NO. However, a very high and unphysiological concentration of NO (1 mM) changed the cytochrome b 558 conformation slightly; the spectral intensity was reduced by about 10%, and the ␥-peak was shifted about 0.6 nm.
To investigate whether NO is trapped at the heme site, nitrosyl-iron complex formation was examined by EPR for purified cytochrome b 558 reduced with dithionite in the presence of NO at 77-100 K. As shown in Fig. 4A, signals characteristic of nitrosyl-iron complex were not observed. On the other hand, when five-coordinated high spin cytochrome b 558 , i.e. denatured cytochrome b 558 , was incubated with NO, a triplet hyper-    fine signal characteristic of heme group nitrosylation was observed (Fig. 4B). The results indicate that NO does not bind nor is trapped at the six-coordinated low spin heme site in intact cytochrome b 558 and that neither the fifth or sixth histidine ligand (27) Table II, we can conclude that the redox centers in the electron transfer chain of the NADPH oxidase remain intact after incubation of the oxidase with NO.

The Effects of NO on the Activation and Assembly Processes of the Reconstituted NADPH Oxidase System (Reaction 5)-
The effects of NO on the activation and assembly processes were examined in reconstituted NADPH oxidase (Fig. 5) Fig. 5). When a mixture of membrane and cytosol components was activated with myristate simultaneously with NO, the O 2 . -forming activity was decreased to 26.0 Ϯ 5.3% of control (C in Fig. 5). In contrast, when NO was added after the activation, its inhibitory effect was much weaker (D and E in Fig.  5). These results suggest that the main site of NO-induced inhibition is the impairment of activation, such as the assembly of cytosol protein(s) and the membrane. The inhibition of C in Fig. 5, which was more pronounced than the sum of inhibitions of each component (A and B in Fig. 5), suggests that myristate is the important factor for NO-induced inhibition in addition to the impairment of each component. To further clarify the role of myristate, the cytosol was pretreated with NO in the presence of myristate before mixing with membrane components (D in Fig. 6). The O 2 . -generating activity was markedly inhibited to the level when myristate and NO were simultaneously added to membrane and cytosol components (E in Fig. 6). Membrane components were pretreated with NO in a similar way as cytosol, but the inhibitory effect of NO was not changed in the presence of myristate (A and B in Fig. 6). From the results in Table II and Figs. 3 and 4, it becomes clear that NO does not impair redox centers in NADPH oxidase, suggesting that sulfhydryl groups in proteins are candidates to be attacked by NO, because NO is known to react with tissue sulfhydryls to form S-nitrosothiol compounds, such as S-nitrosocysteine (28). Some insight for an explanation of the effect of myristate and plausible impaired site appears to be shown by a recent report on the role of Src homology 3 (SH3) domain in p47 phox during activation (29). It demonstrated that specifically SH3 domain is folded, masked, and localized at its C-terminal region in dor-mant cells or in the resting cytosol but opens up upon activation and then binds to the membrane protein(s), mainly cytochrome b 558 . There is one cysteine residue  in the SH3 domain that is masked in the resting state of the oxidase, but once the SH3 domain opens up upon activation by myristate, the thiol group of Cys-196 may become accessible to NO. Because there are three other cysteine residues in the SH3 domain of p47 phox and 8 cysteine residues in the cytosolic p67 phox protein, NO-induced oxidation of these sulfhydryl groups may be attributable to the 30 -35% inhibition seen in the treatment of cytosol with NO (B in Fig. 5). Similarly the 20 -25% inhibition upon treatment of the membrane component with NO may also be a result of sulfhydryl group oxidation (A in Fig. 5). This answered the observation shown in Fig. 2 in which the inhib- . generation in reconstituted NADPH oxidase system.
Step A, neutrophil membrane (15 g) incubated with 25 M NO for 5 min at 0°C before activation.
Step C, mixture of membrane and cytosol incubated with NO and myristate.
Step D, NO added with cytochrome c (cyt c) after activation and then incubated for 5 min before the addition of NADPH.
Step E, NO was added with 0.1 mM NADPH.
FIG. 6. Effect of NO on myristate pretreated cytosol or membrane in reconstituted NADPH oxidase system. The amounts of cytosol, membrane, myristate, and NO used in this experiment were the same as in Fig. 5. A, membrane was incubated with NO without pretreatment with myristate. B, membrane was pretreated with myristate and then incubated with NO for 5 min. C, cytosol was incubated with NO without pretreatment with myristate. D, cytosol was pretreated with myristate and then incubated with NO for 5 min. E, same as C in Fig. 5. itory effect of NO in neutrophils was more pronounced than that in the solubilized NADPH oxidase, because, in solubilized NADPH oxidase from stimulated cells, the activation process was already completed and all components were correctly assembled, whereas in neutrophils the activation process was initiated in the presence of NO.
Effect  Fig. 7, respectively). The second explanation is also unlikely, because the oxidation of reduced cytochrome c by ONOO Ϫ is as slow as that with hydrogen peroxide (30). Therefore, we conclude that the temporary lag in cytochrome c reduction is caused by the "quenching" of O 2 . by NO. Actually the reaction of O 2 . with NO is much faster than the reduction reaction of cytochrome c by O 2 . . Note that after the lag time the reduction rate of cytochrome c returned to nearly the initial value. The slight decrease in the reduction rate of cytochrome c after NO addition is consistent with the slight decrease in the O 2 .
generation shown in the reconstituted NADPH oxidase (D and E in Fig. 5). This decrease in the slopes in Fig. 7