NADPH oxidase activity is independent of p47 in vitro

The neutrophil superoxide generating NADPH oxidase is activated by the assembly of cytosolic protein components with a membrane-associated flavocytochrome. The activity can be reconstituted in vitro using purified cytosolic factors p47, p67, and Rac plus the phospholipid-reconstituted flavocytochrome b558. Here, we demonstrate that activity is reconstituted in the absence of p47 when high concentrations of p67 and Rac are used. Vmax values were the same in the presence or absence of p47, yet p47 increases the affinity of both p67 and Rac for the oxidase complex by nearly 2 orders of magnitude. p67-(1–246), a truncated form of the protein which eliminates SH3 domains involved in binding to p47, also supports superoxide generation, both in the presence and absence of p47, providing further evidence for p47 independent activity. In the absence of p47, p67-(1– 246) binds to the NADPH oxidase complex 3-fold more tightly than does native p67, indicating that the C terminus contains a region which masks binding to the oxidase complex. Results indicate that p47 does not play a direct role in regulating electron transfer. Rather, its function is to serve as an adaptor protein to enhance the assembly of the other cytosolic components with the flavocytochrome and possibly to unmask a binding region in the N terminus of p67 by binding to its C-terminal domains. p67 and/or Rac play a more direct role in regulating electron transfer.

During the respiratory burst, neutrophils and other phagocytic cells produce superoxide and other reduced oxygen species that participate in microbial killing (1). The multicomponent NADPH oxidase catalyzes the enzymatic reduction of oxygen to produce superoxide. Upon activation, cytosolic components p47 phox , 1 p67 phox (2), and the small GTPase Rac translocate (3,4) to membrane where they bind directly or indirectly in a multicomponent complex with flavocytochrome b 558 . The flavocytochrome b 558 contains all the prosthetic groups necessary for activity, FAD, two hemes, a consensus sequence for pyridine nucleotide binding (5,6), but catalyzes the reduction of oxygen to superoxide only when assembled with the cytosolic components. Individuals with genetic deficiencies or mutations in p47 phox , p67 phox , or one of the subunits of cytochrome b 558 (gp91 phox and p22 phox ) exhibit chronic granulomatous disease, a condition marked by frequent infections as a result of the inability of neutrophils to generate superoxide and kill bacteria (7,8). In addition to the above components which were identified based on their absence in chronic granulomatous disease, Rac was purified from macrophages (9) and neutrophils (10) as the GTP-dependent factor which is required for in vitro NADPH oxidase activation.
NADPH oxidase activity can be reconstituted using purified p47 phox , p67 phox , Rac, and cytochrome b 558 along with an anionic amphiphile such as arachidonate plus the stable GTP analog GTP␥S (11,12). Both in vivo and in vitro, p47 phox translocates to the cytochrome independent of other cytosolic factors (13)(14)(15). However, significant p67 phox translocation requires p47 phox , both in vivo and in vitro (14 -16). p47 phox has binding sites for both cytochrome b 558 (17,18) and p67 phox (18 -20), and it has been suggested that p67 phox binding to the oxidase complex occurs indirectly through its binding to p47 phox . Such a model would account for the failure of p67 phox to translocate to the membrane in the absence of p47 phox . However, in vitro kinetic evidence suggests that p67 phox can bind independently of p47 phox (21), implying the presence of one or more additional binding sites. In support of this interpretation, deletion of a p47 phox binding region on p67 phox does not eliminate the ability of p67 phox to activate superoxide generation (22). Also, p67 phox binds directly to Rac, based on blotting and yeast two-hybrid assays (20,23,24).
The function of the cytosolic factors in regulating electron flow from NADPH to oxygen is poorly understood. According to one proposal, p67 phox and p47 phox have distinct roles in controlling electron flux; it has been proposed that p67 phox regulates the transfer of electrons from NADPH for reduction of the flavin, while p47 phox controls electron flow between the flavin and heme groups (25). Here, we demonstrate that when high concentrations of the other cytosolic factors are used, p47 phox is not required to elicit high rates of superoxide generation. Rather, p47 phox serves as an adaptor protein to enhance the binding of other cytosolic components to the oxidase complex and does not itself have a role in regulating electron flux within the oxidase complex.
Preparation of Cell Fractions and Recombinant Proteins-Human neutrophils were isolated from healthy donors and purified according to Bowman et al. (26) after informed consent was obtained. Plasma membranes were isolated as described previously (27). Cytochrome b 558 was solubilized with cholate and octyl glucoside and purified to a specific heme content greater than 6 nmol of heme/mg of protein using several column chromatographic steps as described previously (6). Cytochrome b 558 was incorporated into phospholipid vesicles composed of phosphatidylcholine:phosphatidylethanolamine:PI:SM:cholesterol in a 4:2:1:3:3 mass ratio (28) and dialyzed against two changes of 1 liter of 100 nM Tris acetate, pH 7.4, 100 mM KCl, 20% glycerol, 1 mM dithiothreitol, 1 mM EGTA, 1 mM phenylmethanesulfonyl fluoride, 1 g/ml leupeptin, 1 * This work was supported by National Institutes of Health Grant AI22809. 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.
g/ml pepstatin, and 5 M FAD for a total of 24 h at 4°C. Rac was expressed in Escherichia coli as a glutathione S-transferase protein and purified using glutathione-coupled Sephadex followed by thrombin cleavage as described previously (27). Recombinant p47 phox and p67 phox were produced in insect cells and purified as described by Uhlinger et al. (14,29). The cDNA for the truncated p67 phox (amino acids 1-246)glutathione S-transferase fusion protein was a gift from Alan Hall. Protein was expressed as with Rac above, except p67 phox -(1-246) was eluted with glutathione according to Diekmann et al. (23) and dialyzed to remove free glutathione. All proteins were purified to Ͼ90% homogeneity. Protein concentration was determined according to Bradford (30).
Assay for Cell-free NADPH Oxidase Activity-Superoxide generation was measured by superoxide dismutase-inhibitable reduction of cytochrome c as described previously (31) using a Thermomax Kinetic Microplate reader (Molecular Devices, Menlo Park, CA). Rac, preloaded with a 5-fold molar excess of GTP␥S for 15 min at 25°C in the absence of MgCl 2 (27), was combined with p47 phox , p67 phox , 10 nM cytochrome b 558 , and 1 M FAD followed by activation with 40 M arachidonate in 50 mM NaCl, 4 mM MgCl 2 , 1.25 mM EGTA, 20 mM Tris-HCl, pH 7.0, as described in Rotrosen et al. (12). The mixture was incubated at 25°C for 5 min followed by addition of 200 M NADPH and 200 M cytochrome c. An extinction coefficient at 550 nm of 21 mM Ϫ1 cm Ϫ1 was used to calculate the quantity of cytochrome c reduced (32). The theoretical lines through the data shown in the figures were calculated using a nonlinear least squares fit of the data using the Michaelis-Menten equation and plotted with Sigma Plot. Kinetic parameters were calculated using Enzfitter with the data in Figs. 2, 3, and 4 and reported as V max and EC 50 (effective concentration at 50% of V max ).

RESULTS AND DISCUSSION
Previous reports have indicated that p47 phox is necessary for NADPH oxidase activity in vivo (33)(34)(35) as well as in vitro (11,12). As shown in the left half of Fig. 1 (filled bars), when relatively low concentrations of cytosolic proteins are used (i.e. submicromolar concentrations similar to those used in previous studies), p47 phox , p67 phox , and Rac all are needed to obtain a high rate of superoxide generation. These experiments used concentrations similar to those used in earlier studies and confirm published results. In contrast, when high concentrations of p67 phox (6 M) and Rac (2 M) were used, a high rate of superoxide generation was seen even in the absence of p47 phox (Fig. 1, open bars). p67 phox and Rac are required, however. Even in the presence of very high concentrations of the other two components, little or no activity was seen when either of these components was omitted. Activity was specific for these NADPH oxidase components even at high concentrations. RhoA (2 M), another member of the Rho family of small GTPases which is 59% identical to Rac, failed to replace Rac1 in this system. It was also possible that at high concentrations, p67 phox was substituting for p47 phox . Although these two proteins show rather low sequence identity overall (18%), they both contain two SH3 domains. It was therefore possible that the SH3 domains of p67 phox at high concentrations were functioning in place of those of one of the cytosolic factors. Like p47 phox and p67 phox , Grb2 contains two SH3 domains. High concentrations of Grb2 (10 M) failed to substitute for either p67 phox or p47 phox when normal concentrations of the remaining cytosolic factors were used (data not shown). p67 phox -(1-246), a truncated version of p67 phox lacking both SH3 domains, was also active at high concentrations in the NADPH oxidase assay system in the absence of p47 phox (see below), indicating that the SH3 domains of p67 phox were not substituting for those of p47 phox .
To determine the extent to which the affinity for p67 phox is influenced by p47 phox , the concentration dependence of p67 phox was determined in the presence and absence of p47 phox (Fig. 2). The EC 50 for p67 phox was decreased 66-fold by p47 phox (Table I). These results are consistent with our earlier report in which the EC 50 value for p67 phox varied inversely with p47 phox concentration (21), except that the earlier study failed to achieve sufficiently high concentrations of p67 phox to observe activity that was independent of p47 phox . Despite the large change in p67 phox EC 50 values, the V max values were the same within experimental error (Table I). The effect of p47 phox on the EC 50 for p67 phox is in agreement with previous studies indicating a role for p47 phox in localizing p67 phox to the NADPH oxidase complex (18). However, the generation of superoxide in the absence of p47 phox clearly demonstrates that p67 phox has one or more additional binding sites for the oxidase complex which is (are) independent of p47 phox .
Further evidence for p47 phox independent assembly of p67 phox to the NADPH oxidase complex is seen in Fig. 3. p67 phox -(1-246), which lacks both SH3 domains, was used in place of full-length p67 phox . One or both of the SH3 domains of p67 phox are known to mediate binding to a proline-rich region on p47 phox (18 -20). As shown, p67 phox -(1-246) partially substituted for the full-length p67 phox in supporting superoxide generation. Interestingly, in the absence of p47 phox , truncated p67 phox had a 3-to 4-fold lower EC 50 than did full-length p67 phox (see Table I). This may indicate that part of the function of the C-terminal half of the molecule is to mask or inhibit a binding domain in the N-terminal half. When this region is removed, the affinity of truncated p67 phox for the NADPH oxidase is increased. Thus, in addition to acting directly to local- ize p67 phox in the oxidase complex, we propose that another function of p47 phox is to unmask a binding site(s) in the N terminus of p67 phox by binding to regions in the C terminus of p67 phox . Although both truncated and full-length p67 phox both supported superoxide generation in the absence of p47 phox , the V max for the truncated form was only about 25% of that of the full-length form. Thus, for unknown reasons, truncated p67 phox is partially defective in its regulatory function. Surprisingly, when p47 phox was included, there was a modest (4 -5-fold) enhancement of the apparent binding of the truncated p67 phox . This should be compared with the roughly 70-fold binding enhancement when intact p67 phox was used. This is consistent with the known role of the SH3 domains in p67 phox as a major determinant mediating binding to p47 phox . However, the small enhancement of p67 phox -(1-246) apparent binding by p47 phox can probably best be explained by the presence of an additional binding region for p47 phox within the N-terminal half of p67 phox . This binding might also explain the modest increase in V max induced by p47 phox . It is also possible that p47 phox might induce its effects indirectly by altering the conformation of cytochrome b 558 .
To investigate the influence of p47 phox on Rac, the concentration dependence for Rac was determined in the presence and absence of p47 phox . Results (Fig. 4) were similar to those seen with p67 phox ; p47 phox caused a 35-fold decrease in the EC 50 for Rac, but did not alter the V max significantly (Table I). This decrease in EC 50 could be due to a direct interaction between p47 phox and Rac, or it could be mediated indirectly through p47 phox binding to and altering the Rac association with another component of the oxidase. No direct interaction between Rac and p47 phox has been observed in vitro (20, 23), 2 but this remains a theoretical possibility. Alternatively, p47 phox binding to cytochrome b 558 or p67 phox could induce a conformational change in one of these components, creating a higher affinity binding site for Rac. There is no direct evidence for binding of 2 Y. Nisimoto, personal communication. In both experiments, Rac concentrations were sufficient to achieve greater than 85% of the V max . This experiment is representative of 3.

TABLE I
Effect of p47 phox on kinetic parameters of Rac and p67 phox Superoxide generation was measured by the reduction of cytochrome c as described under "Experimental Procedures." V max is the maximal velocity and EC 50 is the effective concentration at 50% V max . The mean Ϯ S.E. for three to four independent experiments are reported.  Rac to the flavocytochrome, although in vivo translocation data have been interpreted in terms of such a complex (4). p67 phox interacts with both p47 phox and Rac (20,23,24). Therefore, the binding of p47 phox to p67 phox could induce a conformational change in the Rac binding site on p67 phox , enhancing the binding of Rac. Based on current data, this seems to be the most likely mechanism, but this remains to be proven.
Since we were able to activate superoxide generation independently of p47 phox , we tested the hypothesis that p47 phox is the target of activation by arachidonate, as has been proposed by us and others (14,36). Fig. 5 demonstrates that when high levels of p67 phox and Rac are used, arachidonate is necessary for activation regardless of whether p47 phox is present. Therefore, p47 phox is either not the target of arachidonate activation or there is an additional component also influenced by arachidonate. Since the optimal concentration of arachidonate for activation is the same in the presence and absence of p47 phox (data not shown) we favor the latter model in which p47 phox is not the target of arachidonate. A possible target is cytochrome b 558 since evidence also indicates that the proton channel activity associated with this transmembrane protein is dependent upon arachidonate (37). p67 phox might also be the target for arachidonate, consistent with a kinetic study using SDS, another anionic amphiphilic activator of the oxidase (38).
An earlier study (25) indicated a role for p47 phox in facilitat-ing electron transfers within the NADPH oxidase complex, specifically at the level of the reduction of heme by flavin. However, the data reported herein rule out such a function, since p47 phox had no effect on the V max for superoxide generation when high concentrations of the other cytosolic components were used (Table I). We conclude that the role of p47 phox is to provide high affinity binding sites for p67 phox and (directly or indirectly) Rac, and that p47 phox does not itself function in modulating the catalytic properties of the NADPH oxidase complex. In this sense, it can be considered an "adaptor protein" which enhances the binding of the other cytosolic proteins to the flavocytochrome. p47 phox is phosphorylated at multiple sites in response to cell activation (39 -41), and this phosphorylation is presumed to play a role in regulating the interaction of this adaptor protein with the flavocytochrome. Presumably, p67 phox , Rac, or both proteins serve a direct regulatory role to enhance the electron flow within the NADPH oxidase complex.
The data indicate that both p67 phox and Rac bind weakly to the NADPH oxidase complex independently of p47 phox . This implies that either Rac, p67 phox , or both proteins have a binding site for cytochrome b 558 . With regard to Rac, evidence indicates that a site equivalent to the Ras effector site (residues 26 -45) is important for binding within the NADPH oxidase complex (23,(42)(43)(44), and a study suggests that this is the region which binds to p67 phox (23). We have recently described an additional effector site (residues 124 -135) in Rac which is important for binding to the NADPH oxidase complex (45). This region is a candidate binding site for interaction with the cytochrome. A third region, the C terminus, is isoprenylated and anchors Rac to the membrane. We find that the membrane interaction is essential for achieving high superoxide generating activity. 3 Less is known about candidate cytochrome binding sites on p67 phox . The present studies suggest the N-terminal half of the protein might contain such a binding site. Thus, Rac and/or p67 phox bind directly to the flavocytochrome in such a way as to modulate the electron flow from NADPH to oxygen to generate superoxide. p47 phox functions as an adaptor protein to bring the cytochrome and other cytosolic regulatory components into close contact.