Activation of the Leukocyte NADPH Oxidase by Protein Kinase C in a Partially Recombinant Cell-free System*

The leukocyte NADPH oxidase is an enzyme present in phagocytes and B lymphocytes that when activated catalyzes the production of O⨪2 from oxygen at the expense of NADPH. A correlation between the activation of the oxidase and the phosphorylation of p47 PHOX , a cytosolic oxidase component, is well recognized in whole cells, and direct evidence for a relationship between the phosphorylation of this oxidase component and the activation of the oxidase has been obtained in a number of cell-free systems containing neutrophil membrane and cytosol. Using superoxide dismutase-inhibitable cytochrome c reduction to quantify O⨪2 production, we now show that p47 PHOX phosphorylated by protein kinase C activates the NADPH oxidase not only in a cell-free system containing neutrophil membrane and cytosol, but also in a system in which the cytosol is replaced by the recombinant proteins p67 PHOX , Rac2, and phosphorylated p47 PHOX , suggesting that neutrophil plasma membrane plus those three cytosolic proteins are both necessary and sufficient for oxidase activation. In both the cytosol-containing and recombinant cell-free systems, however, activation by SDS yielded greater rates of O⨪2 production than activation by protein kinase C-phosphorylated p47 PHOX , indicating that a system that employs protein kinase C-phosphorylated p47 PHOX as the sole activating agent, although more physiological than the SDS-activated system, is nevertheless incomplete.

The NADPH oxidase is a membrane-associated enzyme that catalyzes the one electron reduction of oxygen to O 2 . at the expense of NADPH (1). The oxidase comprises multiple protein components present in both the cytosol and the plasma membrane. The enzyme is dormant in resting cells but becomes activated when the cells are exposed to appropriate stimuli. Upon activation, a cytosolic complex consisting of the oxidase components p47 PHOX , p67 PHOX , and p40 PHOX associates with the membrane-bound cytochrome b 558 to assemble the active oxidase (2)(3)(4)(5)(6)(7). The phosphorylation of p47 PHOX is a well recognized concomitant of oxidase activation in whole cells, but the mechanism of activation of the oxidase is not fully understood (8 -14). One key to understanding the activation of the oxidase emerged with the discovery of the cell-free activation system (15)(16)(17) in which it was shown that NADPH oxidase activity could be induced in a mixture of membrane and cytosol by the addition of amphiphiles like arachidonic acid (15,17) and SDS (16,18). Recently, increasing attention has been paid to cell-free systems in which the oxidase is activated without using anionic amphiphiles (19 -22). Our studies showed that the oxidase can be activated by p47 PHOX phosphorylated by protein kinase C in a cell-free system containing neutrophil membrane and cytosol (21). In addition, these studies also revealed a that the phosphorylation of p47 PHOX was not the only ATP-dependent step in the activation of the oxidase by protein kinase C. A preceding phosphorylation event occurs in the membranes rendering them capable of supporting oxidase activation. The target of this event has yet to be determined. Although these experiments showed a direct relationship between the phosphorylation of p47 PHOX and the activation of the oxidase, the use of whole cytosol made it difficult to recognize whether cytosolic factors other than those necessary for activation by SDS are required for oxidase activation by a kinase. In this paper we report studies of a recombinant cell-free system containing only membrane and cytosolic oxidase components (p47 PHOX , p67 PHOX , and Rac2). Our findings suggest that the cytosolic components phosphorylated p47 PHOX , p67 PHOX , and Rac2 are sufficient for partial activation of the oxidase.
Preparation of Neutrophil Fractions-Neutrophil cytosol and membrane were prepared as described by Borregaard et al. (23). Neutrophils were prepared from normal subjects by dextran sedimentation and Ficoll-Hypaque fractionation of freshly drawn citrate-anticoagulated blood. The neutrophils were suspended at 10 8 cells/ml in a modified relaxation buffer (0.1 M KCl, 3 mM NaCl, 3.5 mM MgCl 2 , 10 mM PIPES buffer, pH 7.3). Plasma membrane and cytosol were prepared by nitrogen cavitation followed by centrifugation through a Percoll gradient. Both cytosol and membrane were divided into aliquots and stored at Ϫ70°C until use.
Production and Purification of Recombinant p67 PHOX from Baculovirus-infected Sf9 Cells-Purified recombinant p67 PHOX was produced by means of the baculovirus system described by Leto et al. (1991), using a p67 PHOX -expressing recombinant virus generously provided by T. L. Leto. Large scale production of pure recombinant p67 PHOX was achieved by infecting monolayer cultures of Sf9 cells in 150 cm 2 flasks at a density of 1-2 ϫ 10 6 cells/ml (24). Cells were harvested 72 h postinfection, washed twice in phosphate-buffered saline by centrifugation at 400 ϫ g for 10 min, and then resuspended to 5 ϫ 10 7 /ml in lysis * This work was supported in part by United States Public Health Service Grants AI-24227, AI-28479, and RR-00833. 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.
§ Postdoctoral fellow of the Fundaçã o de Amparo à Pesquisa do Estado de Sã o Paulo (Brazil).
¶ Postdoctoral fellow of the Arthritis Foundation.
buffer (50 mM KCl, 3 mM NaCl, 2 mM MgCl 2 , 0.1 mM dithiothreitol, 1 mM EDTA, 10 g/ml leupeptin, 10 g/ml pepstatin, 10 g/ml aprotinin, 2 mM phenylmethylsulfonyl fluoride, 5.4 mM PIPES, pH 7.5). All subsequent work was conducted at 4°C. Cells were disrupted by sonication (4 ϫ 10 s) and centrifuged at 400 ϫ g for 10 min. The supernatant fraction containing p67 PHOX was brought to 45% saturation with solid ammonium sulfate. The resulting precipitate was isolated by centrifugation (1200 ϫ g at 4°C for 30 min), then dissolved in 10 ml of buffer A (20 mM Tris, pH 7.5, 0.1 mM dithiothreitol, 1 mM EDTA, 2 mM EGTA, 0.15 mM phenylmethylsulfonyl fluoride) and dialyzed overnight against the same buffer. The dialyzed solution was applied to a Mono Q-Sepharose column (Amersham Pharmacia Biotech) previously equilibrated with buffer A and washed with 5 volumes of the same buffer. Proteins were eluted from the column by fast protein liquid chromatography with a 0.1-0.3 M NaCl gradient in the same buffer at a flow rate of 0.8 ml/min. The fractions containing purified p67 PHOX were pooled and stored at Ϫ70°C. Preparation of Recombinant GST-p47 PHOX and Rac2 Fusion Proteins-Recombinant fusion proteins composed of glutathione S-transferase (GST) linked downstream to p47 PHOX or Rac2 were isolated from Escherichia coli transformed with pGEX-1T plasmids containing cDNA inserts encoding the downstream proteins as described by Park et al. (3). The fusion proteins were purified by affinity chromatography on glutathione-agarose beads. Initially the culture was grown overnight at 37°C in 100 ml of "Terrific Broth" containing 0.1% ampicillin, then diluted into 1 liter of fresh Terrific Broth/ampicillin. The diluted cultures were grown for an additional hour at 37°C (for GST-p47 PHOX expression) or an additional 2.5 h at 37°C (for GST-Rac2 expression). Isopropyl-␤-D-thiogalactopyranoside (0.1 mM) was then added, and the cultures were grown with vigorous agitation for an additional 3 h at 37°C for GST-p47 PHOX expression or 30°C for GST-Rac2 expression. At the conclusion of the incubations, the cells were recovered by centrifugation at 2000 ϫ g for 10 min at 4°C. The GST-p47 PHOX pellet was suspended in 10 ml of ice-cold phosphate-buffered saline containing a 1 ϫ mixture of protease inhibitors (Roche Molecular Biochemicals), while the GST-Rac2 pellet was resuspended in a lysis buffer (50 mM Tris-HCl, pH 7.6, 50 mM NaCl, 5 mM MgCl 2 , 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride). The cells were then disrupted by sonication. The sonicates were clarified by centrifugation at 14,000 ϫ g for 15 min at 4°C. The fusion proteins were isolated from the supernatant by purification over glutathione-agarose as described by Smith and Johnson (25). Before use, excess glutathione was removed from the solution of purified recombinant protein by dialysis against relaxation buffer. The concentrations of all proteins (95-99% pure) were determined with a Bio-Rad assay kit using bovine serum albumin as a standard.
Phosphorylation of GST-p47 PHOX -Phosphorylation of recombinant GST-p47 PHOX was carried out using 200 g of fusion protein in final volume of 200 l. The reaction mixture contained 1 mM ATP, 10 mM magnesium acetate, 1.0 mM CaCl 2 , 10 g phosphatidylserine, 1 g diolein, and 0.5 unit of protein kinase C in 200 l of relaxation buffer (0.1 M KCl, 3 mM NaCl, 3.5 mM MgCl 2 , 10 mM PIPES buffer, pH 7.3). The lipids were added as mixed liposomes prepared by dissolving 2.5 mg/ml phosphatidylserine and 1 mg/ml diacylglycerol in chloroform, removing the chloroform under a stream of nitrogen, and then sonicating the dried lipids for 2 min on ice in 0.8 ml of 20 mM Tris buffer, pH 7.4. Incubations were carried out for 30 min at 37°C. The phosphorylated protein, designated p47 PHOX P 6 , 2 was separated from the reaction mixture as described elsewhere (21).
Cell-free Activation of the NADPH Oxidase with p47 PHOX P 6 -Activation of the NADPH oxidase in the cell-free system was directly measured by following the superoxide dismutase-inhibitable reduction of cytochrome c at 550 nm in a dual beam recording spectrophotometer. The complete reaction mixture contained 5 ϫ . production is expressed as moles of O 2 . /moles of cytochrome b 558 /minute. Where indicated, cytosol (10 7 cell equivalents) was added to the system instead of the recombinant proteins GST-p47 PHOX , p67 PHOX , and GST-Rac2. Experiments were also performed in which the recombinant proteins (p67 PHOX , GST-p47 PHOX , and GST-Rac2) were individually omitted to determine which of these proteins was required for the production of O 2 . by the oxidase.
Cell-free Activation of the NADPH Oxidase with SDS-In these experiments, 5 ϫ 10 6 cell equivalents of membrane were incubated for 10 min at 30°C with 105 pmol of GST-Rac2, 75 pmol of p67 PHOX and 70 pmol of GST p47 PHOX plus 50 M GTP␥S as described previously, with the exception that unphosphorylated p47 PHOX was used in place of p47 PHOX P 6 . After the preincubation, SDS (90 M final concentration) was added to the reaction mixture and incubated for 1 min. In some experiments, cytosol (10 7 cell eq) was used in place of the recombinant proteins as described above.
Cytochrome b 558 Determination-Neutrophil membranes (1 ϫ 10 7 cell eq) were resuspended in 400 l of Triton buffer (0.1 M potassium phosphate buffer, pH 7.25, containing 2% Triton (v/v). Cytochrome b 558 content was measured as the dithionite-reduced minus oxidized absorption assuming ⌬E 559 -540 ϭ 21.6 mM Ϫ1 cm Ϫ1 (27). In some spectra the height of the peak was estimated by interpolation to correct for baseline drift.
Electrophoresis and Immunoblotting-Protein samples were subjected to SDS-PAGE using the Laemmli buffer system (28). The gels were stained with Coomassie Blue. Alternatively, the separated proteins were electrophoretically transferred onto a nitrocellulose sheet (29) and probed with partially purified rabbit polyclonal antibodies raised against full-length Rac2 or the C-terminal decapeptides from p47 PHOX and p67 PHOX . These antibodies were used at 1:1000, 1:2000, and 1:5000 dilution for GST-Rac2, GST-p47 PHOX , and p67 PHOX , respectively. The proteins were visualized with a 1:2000 dilution of horseradish peroxidase-labeled goat anti-rabbit Ig antibody (Caltag) and the ECL enzymatic chemoluminescence detection system (Renaissance, DuPont).

Absence of Cytosolic Factors from the Neutrophil Mem-
brane-In the experiments described below, neutrophil membranes were mixed with recombinant cytosolic oxidase proteins in various combinations. SDS-PAGE gels of the recombinant proteins used for these experiments are shown in Fig. 1. Since the experiments involved the addition of cytosolic components to the assay mixtures, it was important to determine whether any of these components was present in the neutrophil membrane that was to be used in the experiments. For this purpose, membrane and cytosol in equal amounts (expressed as cell equivalents) were subjected to SDS-PAGE and immunoblotting, developing with antibodies against the cytosolic components that were to be added in recombinant form. The components were readily visible in the cytosol, but could not be detected in the membrane.
Activation of the Cell-free Recombinant Leukocyte NADPH Oxidase by p47 PHOX P 6 -In the cell-free system, the leukocyte NADPH oxidase has customarily been activated using certain anionic amphiphiles including SDS. In an earlier study in which we supplemented the cell-free system with p47 PHOX (added as the GST fusion protein), we found that the enzyme could be activated without detergent, provided the p47 PHOX was first phosphorylated by protein kinase C (21). We believe that the activation of the cell-free oxidase by protein kinase C may represent a more physiological process than activation by anionic amphiphiles, because in intact cells, as in the kinaseactivated cell-free system, oxidase activation is associated with the phosphorylation of p47 PHOX . It has been shown by others that in a cell-free system in which neutrophil cytosol has been replaced by the two recombinant cytosolic oxidase subunits p47 PHOX and p67 PHOX together with the small GTPase Rac2, O 2 . is produced upon the addition of SDS (30). In order to see if the same system could be activated by a kinase, we conducted experiments in which O 2 .
production was measured in a recombinant system that contained phosphorylated p47 PHOX (i.e. p47 PHOX P 6 ) instead of the unphosphorylated protein. The results (Fig. 2) showed that O 2 . was produced in the complete system, but that the omission of all or any one of the three recombinant cytosolic proteins or the omission of membrane (not shown) essentially eliminated oxidase activity. The use in the recombinant system of Rac2 preloaded with GDP␤S led to a marked reduction in O 2 . formation as compared with a system that employed Rac2 preloaded with GTP␥S, suggesting that Rac2 had to be in its active form for the oxidase to be activated (Fig. 3). These findings indicate that all four components (membrane, p67 PHOX , p47 PHOX P 6 , and Rac2) were required for activation of the NADPH oxidase, indicating that they are both necessary and sufficient for kinasedependent cell-free oxidase activation. Experiments with a cell-free system containing membrane and cytosol indicated that the phosphorylation of p47 PHOX was an essential prerequisite for O 2 . production by the system (31).
To see if the same situation prevailed in the recombinant system, the rate of O 2 . production using p47 PHOX P 6 was compared with the rate of O 2 . production using unphosphorylated p47 PHOX . The results (Fig. 4) show that O 2 . production in this system required the phosphorylation of p47 PHOX . It is possible that further phosphorylation of p47 PHOX by a membrane-associated kinase might also be necessary for the O 2 .
-forming activity in this system. The addition of the protein kinase inhibitor GF-109203X to the reaction mixture, however, had no effect on O 2 . production, indicating that phosphorylation at least by GF-109203X-inhibitable membrane-associated kinases such as activated protein kinase C was not a factor in these experiments. The effect of protein concentration on O 2 . production in the recombinant system was next examined. In these experiments activity was monitored at varying concentrations of the three using the same concentration of Rac2 and p67 PHOX at different concentrations of p47 PHOX P 6 was determined. At the concentration of p47 PHOX P 6 used in this study, the activity of the recombinant and cytosolic cell-free systems were the same (Fig.  6). Curiously, the addition of more than 70 pmol of p47 PHOX P 6 reduced O 2 . production in the recombinant system but not in the cytosolic cell-free system. When the oxidase is activated, p47 PHOX , p67 PHOX , and Rac2 translocate to the membrane in equimolar quantities (32). Therefore we employed approximately equimolar concentrations of the three recombinant cytosolic components (actual stoichiometry 1.5/1/1 for Rac2/ p47 PHOX /p67 PHOX ). Altogether, these results indicate that an excess of either membrane or cytosolic components inhibits O 2 . production in the recombinant cell-free system. Why this same relationship doesn't prevail in the cytosol-containing system is a mystery. Activation of the NADPH Oxidase by SDS Versus Activation by p47 PHOX P 6 -p47 PHOX P 6 produced similar rates of production of O 2 . in the recombinant and cytosolic cell-free systems ( Fig. 7A). In contrast when SDS was used as the stimulus, the rate of O 2 . production in the recombinant cell-free system was approximately 50% of the rate seen with cytosol (Fig. 7B), and both rates were considerably greater than the rates obtained in the p47 PHOX P 6 -activated systems. These results strongly suggest that while p47 PHOX P 6 is sufficient to activate the oxidase at a low level, something from the cytosol that is missing in the recombinant cell-free system is required for maximal activation of the oxidase.

DISCUSSION
It has been known for many years that p47 PHOX becomes heavily phosphorylated on serine residues when the oxidase is activated (8 -14, 33). Here we present evidence that p47 PHOX phosphorylated by protein kinase C is capable of activating the leukocyte NADPH oxidase in a recombinant cell-free system consisting of neutrophil membrane, p67 PHOX and Rac2, therefore identifying the minimum cytosolic components necessary for kinase-dependent activation of the oxidase. These findings strongly suggest that the phosphorylation of p47 PHOX that occurs in whole cells during the activation of the leukocyte oxidase is functionally significant and that protein kinase C is a kinase capable of activating p47 PHOX by phosphorylation. In addition, our results with GTP␥S and GDP␤S confirm that the activation of the enzyme by protein kinase C also requires the activation of Rac2, as shown previously by others (34,35). The mechanism of activation of the NADPH oxidase by anionic amphiphiles is still not clear, but our results confirm previous studies demonstrating that the activation of the NADPH oxidase by SDS is not kinase dependent (20). Furthermore, the finding that oxidase activation is associated with the phosphorylation of p47 PHOX both in intact cells and in the kinase-dependent cell-free system suggests that as compared with amphiphiles, the activation of the oxidase by protein kinase C may represent a more physiological process. The fact that p47 PHOX S379A is nonfunctional both in whole cells and in the kinase-dependent cell-free system, yet is capable of participating in O 2 . production in the amphiphile-dependent cell-free system (21,33), further supports the physiological role played by protein kinase C-dependent activation of the oxidase.
The foregoing experiments also showed that the addition of p47 PHOX P 6 to a cell-free oxidase activating system is not enough to activate the oxidase to its full extent. A number of cytosolic components can be postulated as candidate factor(s) that allow the oxidase to become fully activated. These include lipids (e.g. arachidonic acid (36)), proteins (e.g. p40 PHOX (37)), possibly other kinases, or perhaps a hitherto undiscovered oxidase component. Nevertheless, our findings show that neutrophil membrane, p47 PHOX , p67 PHOX , and Rac2 are sufficient for protein kinase C-mediated activation of the oxidase, albeit at a relatively low level. The participation of other components in the kinase-dependent activation of NADPH oxidase is the subject of an ongoing investigation.