Essential requirement of cytosolic phospholipase A(2) for stimulation of NADPH oxidase-associated diaphorase activity in granulocyte-like cells.

We have previously established a model of cytosolic phospholipase A(2) (cPLA(2))-deficient differentiated PLB-985 cells (PLB-D cells) and demonstrated that cPLA(2)-generated arachidonic acid (AA) is essential for NADPH oxidase activation. In this study we used this model to investigate the physiological role of cPLA(2) in regulation of NADPH oxidase-associated diaphorase activity. A novel diaphorase activity assay, using 4-iodonitrotetrazolium violet as an electron acceptor, was used in permeabilized neutrophils and PLB-985 cells differentiated toward the granulocytic or monocytic phenotypes. Phorbol 12-myristate 13-acetate, guanosine 5'-3-O- (thio)triphosphate (GTP gamma S), or FMLP stimulated a similar diphenylene iodonium-sensitive diaphorase activity pattern in neutrophils and in differentiated parent PLB-985 cells. This diaphorase activity was not detected in undifferentiated cells, but developed during differentiation. Furthermore, diaphorase activity could not be stimulated in permeabilized neutrophils from X-linked CGD patients and in differentiated gp91(phox)-targeted PLB-985 cells that lacked normal expression of gp91(phox), but was restored to these cells following transduction with retrovirus encoding gp91(phox). The differentiated PLB-D cells showed no diaphorase activity when stimulated by either GTP gamma S or FMLP, and only partial activation when stimulated with phorbol 12-myristate 13-acetate. Diaphorase activity in response to either agonists was fully restored by the addition of 10 microm free AA. The permeabilized cell 4-iodonitrotetrazolium violet reduction assay offers a unique tool for the evaluation of NADPH oxidase-associated diaphorase activity in stimulated whole cells. These results establish an essential and specific physiological requirement of cPLA(2)-generated AA in activation of electron transfer through the FAD reduction center of NADPH oxidase.

The phagocyte NADPH oxidase is a multicomponent transport chain that transfers electrons from NADPH to molecular oxygen and generates superoxide, a precursor of microbicidal oxidants important for host defense. NADPH oxidase subunits include three cytoplasmic components, p47 phox , p67 phox , and Rac2, and a membrane flavocytochrome b 558 composed of gp91 phox and p22 phox (1,2). The flavocytochrome b 558 contains three redox centers (FAD and 2 hemes) and an NADPH binding site (3). In addition, a H ϩ channel was identified within the N-terminal 230 amino acids of gp91 phox (2,4). In phagocytes, stimulation results in translocation of cytosolic NADPH oxidase components to the membrane, where they interact with the flavocytochrome to form the activated oxidase. This activated enzyme transfers electrons from cytosolic NADPH to extracellular or phagosomal molecular oxygen through the electron transport chain. The flavin center transfers electrons from the physiological electron donor, NADPH, to two membrane-imbedded hemes of the cytochrome. The latter serves as the terminal electron donor to oxygen (2,5). It has been demonstrated that the single electron transfer from internal NADPH, through the oxidase complex, to external oxygen is an electrogenic process and that the efflux of H ϩ through the H ϩ channel is necessary for charge compensation (6 -9). Several reports have demonstrated a diaphorase activity associated with NADPH oxidase (10 -14). Cross et al. (5,(15)(16)(17)(18) have recently demonstrated this novel NADPH oxidase-associated diaphorase activity, present in the membrane fraction of activated neutrophils, using the artificial electron acceptor iodonitrotetrazolium violet (INT), 1 which interacts directly with the reduced flavin center of the flavocytochrome.
We have previously established a model of cytosolic phospholipase A 2 -deficient differentiated PLB-985 cells (PLB-D cells) and demonstrated that cPLA 2 -generated arachidonic acid (AA) is essential for activation of NADPH oxidase through undefined mechanisms (19). The model of differentiated PLB-D cells offers a unique tool to study the physiological mechanism by which AA generated by cPLA 2 regulates NADPH oxidase activity. In our most recent work, we have shown that cPLA 2generated AA is essential for activation of the oxidase-associated H ϩ channel using this model (20). Although H ϩ channel inhibitors decreased NADPH oxidase activity, as expected from the behavior of coupled reactions, they had no effect on the metabolic acidification following oxidase activation, but inhibited only the second alkalization phase. It was concluded that the H ϩ channel regulates oxidase activity but is not responsible for initiation of oxidase activity. Thus, AA likely acts at other sites mediating oxidase activation.
The aim of the present study was to investigate the involvement of cPLA 2 and its metabolite AA in regulation of NADPH oxidase-associated diaphorase activity, using the cPLA 2 -defi-cient cell line model. In order to achieve this goal, we first adapted the INT reduction assay for measuring oxidase diaphorase activity in permeabilized neutrophils and differentiated PLB-985 cells.

EXPERIMENTAL PROCEDURES
Cell Culture and Induction of Differentiation-PLB-985 cells, selected PLB-D clones, and gp91 phox -targeted PLB-985 cells, lacking the expression of normal gp91 phox (21) (kindly provided by M. C. Dinauer, James Whitcomb Riley Hospital for Children, Indianapolis, IN), were grown in stationary suspension culture in RPMI 1640 medium as described previously (19). The results presented were derived from four individual clones.
Optimal concentrations of 1.25% Me 2 SO were added to PLB cells or PLB-D cells (2 ϫ 10 5 cells/ml) at their logarithmic growth phase to induce differentiation toward granulocyte-like cells (20). Differentiation was induced for 4 days and determined by Mac-1 antigen expression detected by indirect immunofluorescence, as described previously (22).
Neutrophil Isolation-Granulocytes from healthy subjects and from X-linked CGD patients were purified by Ficoll/Hypaque centrifugation, dextran sedimentation, and hypotonic lysis of erythrocytes (23).
Superoxide Anion Measurements in Intact Cells-The production of superoxide anion (O 2 Ϫ ) by intact cells was measured as the superoxide dismutase-inhibitable reduction of ferricytochrome c as described previously (19). Briefly, cells were suspended (5 ϫ 10 5 cells/well) in 100 l of Hanks' balanced salts solution containing 150 M ferricytochrome c and activated by the addition of the appropriate stimulus. The reduction of ferricytochrome c was followed by a change of absorbance at 550 nm every 2 min over a 20-min time course, using a Thermomax microplate reader (Molecular Devices, Menlo Park, CA). The maximal rates of superoxide generation were determined using an extinction coefficient ⑀ 550 ϭ 21 mM Ϫ1 cm Ϫ1 .
Cell Permeabilization-Cells were permeabilized according to Lu et al. (24), with slight modifications. Cells were counted, centrifuged and then resuspended in cold phosphate-buffered saline to a concentration of 1 ϫ 10 7 cells/ml. Samples of 0.8 ϫ 10 7 cells were then centrifuged and resuspended in cold electroporation buffer containing 140 mM KCl, 1 mM MgCl 2 , 100 nM CaCl 2 , 1 mM EGTA, 10 mM glucose, and 10 mM Hepes titrated to pH 7. The cells were then electroporated in a Bio-Rad Pulser cuvette with two to four discharges of 5 kV/cm 2 of a 25-microfarad capacitor using a Bio-Rad Gene Pulser. Between pulses the samples were incubated for 30 s on ice. The cells were centrifuged (10 s at 6000 ϫ g) and instantly resuspended in the electroporation buffer with the addition of 1 mM ATP, 100 M GTP, 2 mM NADPH, and 1 mM INT or 150 M cytochrome c, for the immediate measurement of INT reduction or superoxide production, respectively. 60 g/ml SOD, 0.5 g/ml rotenone, or 6.5 M diphenylene iodonium (DPI) were added to the sample when indicated. More than 95% of the cells were permeabilized by this method, as determined by trypan blue staining, and more than 85% were still permeabilized 15 min after electroporation.

Measurement of Cytochrome c or INT Reduction in Permeabilized
Cells-Measurement of cytochrome c or INT reduction in permeabilized cells were performed in a 96-well microtiter plate; 10 6 cells were plated in each well (5). Inhibitors and activators were added as indicated, and the reaction was monitored for 10 min in 30-s intervals, using the kinetic microtiter plate spectrophotometer. INT reduction was followed by an increase in absorbance at 490 nm, and the maximal rates of INT reduction in nanomoles of e Ϫ /10 6 cells/min were determined using an extinction coefficient ⑀ ϭ 10.5 mM Ϫ1 , cm Ϫ1 , assuming INT is a twoelectron acceptor. Cytochrome c reduction was followed by an increase in absorbance at 550 -650 nm.
Translocation of Oxidase Cytosolic Components to the Plasma Membranes-Membrane and cytosol fractions were prepared as described previously (19). Stimulated or resting permeabilized cells were centrifuged, sonicated in the presence of 5 mM EGTA, 100 mM KCl, 3 mM NaCl, 3.5 mM MgCl 2 , 10 mM PIPES (pH 7.4), 1 mM PMSF, and 100 M leupeptin, and membrane and cytosol fractions were separated. Similar results were obtained in the presence or absence of preincubation with 5 g/ml cytochalasin B (data not shown), which inhibits actin polymerization induced by GTP␥S or FMLP in permeabilized neutrophils (25)(26)(27)(28), indicating that the translocation of the oxidase cytosolic components is to the membrane, rather than a cytoskeleton-associated movement. Protein (150 g) from cell membrane were separated by electrophoresis on 12% SDS-polyacrylamide gels and blotted to nitrocellulose. Immunoblot detection of NADPH oxidase components was performed as described previously (19). The relative changes of the proteins was quantitated using densitometry in a reflectance mode (Hoefer Scientific Instruments, San Francisco, CA).
Retroviral Transduction of gp91 phox -deficient PLB-985 Cells-NADPH oxidase activity was restored to targeted, gp91 phox -deficient PLB-985 cells (21) following three successive overnight transductions with MFGS-gp91 phox /RD114 retrovirus, produced in a FLYRD18 packaging line by methods described previously (29,30). Following day 3 of transduction, the cells were maintained in culture for in culture for 1 day and induced to differentiate with 1.25% Me 2 SO for an additional 4 days before assaying NADPH oxidase and diaphorase activities.

RESULTS AND DISCUSSION
INT Reduction in Permeabilized Human Neutrophils-Several studies have detected NADPH oxidase-associated diaphorase activity in a cell-free system and in membrane fractions of activated human neutrophils and PLB-985 cells using INT as a substrate (5,(15)(16)(17)(18). The first set of experiments in the present study was designed in order to develop a system for detecting diaphorase activity in permeabilized whole cells, since diaphorase activity resides within the cytoplasmic surface of the membrane. The system for detection of INT reduction was based on a permeabilized cell technique detecting cytochrome c reduction (24,31). The INT reduction assay was established in permeabilized human neutrophils (Table I) and then applied to permeabilized PLB-985 cells differentiated toward a granulocytic phenotype (Table II). Two stimuli were chosen to develop the assay: PMA, which is a potent activator of NADPH oxidase; and GTP␥S, which can activate the oxidase only in permeabilized cells. As presented in Table I, stimulation of permeabilized human neutrophils with the optimal concentrations of PMA (50 ng/ml) or GTP␥S (25 M) resulted in linear rates of INT reduction, which were significantly higher than observed in non-activated cells (p Ͻ 0.003 or p Ͻ 0.027, respectively). The optimum INT concentration for maximum detection of diaphorase activity was found to be 1 mM (data not shown). INT INT reduction in permeabilized neutrophils was detected as described under "Experimental Procedures" in the presence of 2 mM NADPH. Where indicated 60 g/ml SOD, 0.5 g/ml rotenone, or 6.5 M DPI were added. reduction could not be detected in stimulated permeabilized neutrophils in the absence of NADPH (Table I), which further indicated a close association with NADPH oxidase activation.
In order to determine whether superoxide released by the stimulated oxidase causes any direct reduction of INT in stimulated permeabilized neutrophils, diaphorase activity was examined in the presence of SOD and compared with cytochrome c reduction. SOD was used at its optimal concentration of 60 g/ml, which totally inhibits cytochrome c reduction in intact neutrophils (data not shown). Addition of SOD caused inhibition of INT reduction in permeabilized neutrophils stimulated with either PMA or GTP␥S by 27.1 Ϯ 4.72% or 40.4 Ϯ 3.49%, respectively (Table I). SOD had no effect on INT reduction in non-activated cells (Table I). Under the same conditions, SOD inhibited 71 Ϯ 1.89% and 55 Ϯ 4.41% of cytochrome c reduction stimulated by PMA or GTP␥S, respectively, indicating that under these conditions superoxide is not solely responsible for cytochrome c reduction. Similar effects of SOD on these two activities were shown in cell-free assays (5). To inhibit the fraction of INT reduction caused by superoxide, SOD was added to the reaction mixture of all INT reduction assays, similar to the previous described methods for cell free system (5,18). To determine whether INT can be reduced by mitochondria, INT reduction was measured in the presence of the mitochondrial electron transfer chain inhibitor, rotenone. Addition of 0.5 g/ml rotenone had no effect the rate of INT reduction measured in permeabilized neutrophils activated by PMA or GTP␥S (Table I). However, to prevent any INT reduction by the mitochondrial electron transfer chain, all experiments were carried out in the presence of 0.5 g/ml rotenone.
DPI is a known potent inhibitor of NADPH oxidase (32)(33)(34)(35). Recent studies show that DPI inhibits the electron transfer through the oxidase by affecting both the FAD binding center (36) and heme components of the flavocytochrome (37-39). Cross et al. (5) reported that, in a cell-free system, NADPH oxidase-associated diaphorase activity and cytochrome c reduction could be inhibited by DPI. This inhibitor was used in stimulated permeabilized neutrophils to determine the fraction of INT reduction attributable to NADPH oxidase activity. The dose-dependent effect of DPI on INT and on cytochrome c reduction was examined in the range of 0 -8 M in stimulated permeabilized neutrophils. As shown for permeabilized neutrophils activated by GTP␥S (Fig. 1), maximal inhibition of either INT reduction or cytochrome c reduction was achieved in the presence of 5 M DPI. Similar results were obtained with PMA as a stimulant (data not shown). The presence of DPI caused a significant (p Ͻ 0.05) inhibition of diaphorase activity in permeabilized neutrophils stimulated with PMA or GTP␥S (Table  I), whereas it had no effect on the low level of INT reduction in unstimulated cells. Representative results demonstrating the kinetics of INT reduction and its inhibition by DPI are shown in Fig. 2. The fraction of DPI-inhibited INT reduction, which is insensitive to SOD, is therefore likely mediated by NADPH oxidase-associated diaphorase activity. To further support this suggestion, INT reduction was studied in neutrophils from a patient with X-linked CGD patient with gp91 phox deficiency. As shown in Fig. 2, INT reduction in X-linked gp91 phox -deficient neutrophils stimulated with PMA was identical in the absence or presence of DPI and in unstimulated cells. Stimulation with either GTP␥S or FMLP revealed similar results (data not shown). Of note is that INT reduction in permeabilized neutrophils from a patient with X-linked deficient gp91 phox is similar to INT reduction in unstimulated normal neutrophils and to the fraction of INT reduction that is not inhibited by DPI in normal neutrophils. These results clearly indicate that the fraction of INT reduction inhibited by DPI represents NADPH oxidase-associated diaphorase activity and, furthermore, that DPI is a specific inhibitor of this activity in this system. A scheme illustrating the topology of the flavocytochrome with regard to functional sites involved in electron transport, including sites of interactions with electron acceptors (INT) and inhibitors, is presented in Fig. 3. Representative kinetics of DPIinhibitable INT reduction in permeabilized neutrophils from healthy donors and from an X-linked gp91 phox -deficient patient stimulated by PMA, GTP␥S, or the physiological agonist FMLP, and in unstimulated cells, are presented in Fig. 4. The means Ϯ S.E. of the maximal rates of DPI-inhibitable INT reduction stimulated with PMA, GTP␥S, or FMLP are 12.48 Ϯ 1.9, 8.88 Ϯ 1.04 or 7.12 Ϯ 0.5 nmol of e Ϫ /10 6 cells/min, respectively, and is negligible in unstimulated cells, 0.44 Ϯ 0.66 nmol of e Ϫ /10 6 cells/min, (Fig. 4, Table I). The maximal rates of SOD-inhibitable cytochrome c reduction in permeabilized neutrophils stimulated with PMA, GTP␥S, or FMLP was found to be: 15.5 Ϯ 2.5, 10.25 Ϯ 1.5, or 7.57 Ϯ 1.49 nmol of O 2 Ϫ / 10 6 cells/min, respectively. Thus, the relative yields for cytochrome c and INT reduction were essentially identical in permeabilized neutrophils.
NADPH Oxidase-associated Diaphorase Activity in Permeabilized Differentiated PLB-985 Cells-The oxidase-associated  In all experiments, both the parent PLB-985 cells and G418-resistant clones transfected with the empty pcDNA 3 vector were used. Since no differences were observed between the two, for simplicity we present the results of the clones transfected with empty pcDNA 3 vector only (PLB cells). First, the stimulation of cytochrome c reduction was analyzed in permeabilized PLB cells. Significant superoxide production was elicited by either PMA or GTP␥S in permeabilized differentiated parental PLB cells, whereas no superoxide production could be detected in response to these stimuli in permeabilized differentiated PLB-D cells (lacking cPLA 2 ) (Fig. 5). These results are in accordance with our previous studies demonstrating the inability of intact differentiated PLB-D cells to produce superoxide (19,20). Comparison of SOD-inhibitable cytochrome c reduction in permeabilized neutrophils with that of permeabilized differentiated PLB-985 cells revealed that the fraction of SOD-inhibitable cytochrome c reduction is lower in PLB cells, reflecting significantly lower levels of oxidase activity detected in these cells (20).    (Fig. 6C). Undifferentiated leukemic cells do not express oxidase components, which are acquired during myeloid differentiation (19,40); thus, the absence of INT reduction in undifferentiated cells supports the notion that INT reductase activity represents NADPH oxidaseassociated diaphorase activity. More importantly, this association between INT reduction and NADPH oxidase activity was supported by experiments demonstrating that neither PMA, GTP␥S, nor FMLP stimulated DPI-inhibitable INT reduction in permeabilized differentiated PLB cells with a targeted deficiency in gp91 phox (Fig. 6D). Furthermore, diaphorase activity was fully restored following expression of retroviral gp91 phox protein in these cells (Fig. 6D), thereby establishing the functional oxidase as the source of diaphorase activity. Thus, NADPH oxidase-associated diaphorase activity can be determined in whole, differentiated PLB cells as in mature human neutrophils, using the electropermeabilization technique and INT as a substrate. Consistent with the lower cytochrome c reduction detected in differentiated PLB cells in comparison with neutrophils, these cells also express lower diaphorase activity.
cPLA 2 Is an Essential Activator of NADPH Oxidase Diaphorase Activity-Although stimulation of differentiated PLB-D cells with PMA did not elicit any superoxide production, it resulted in partial activation of the NADPH oxidase-associated diaphorase activity, although significantly lower (p Ͻ 0.001) than that observed in normal differentiated PLB cells (Fig. 7A). Addition of GTP␥S or FMLP to permeabilized differentiated PLB-D cells did not stimulate any diaphorase activity (Fig. 7A), consistent with the inability of these cells to produce superoxide. Fig. 7B depicts the means Ϯ S.E. of the maximum rates of NADPH oxidase-associated diaphorase activity in permeabilized differentiated PLB-D cells in comparison with permeabilized differentiated PLB cells to emphasize the difference. The mean Ϯ S.E. of the maximal rates of NADPH oxidase-associated diaphorase activity stimulated by PMA in permeabilized differentiated PLB-D was 1.96 Ϯ 0.56 nmol of e Ϫ /10 6 cells/min, significantly lower than that induced in permeabilized parent PLB cells. Addition of 10 M free AA to the differentiated PLB-D cells stimulated with PMA, GTP␥S, or FMLP fully restored the INT reducing capacity to these cells to levels that were comparable to INT reductase activity observed in the parent stimulated PLB cells (Fig. 8A). Not shown is that addition of free AA alone, without stimulation with PMA, GTP␥S, or FMLP, did not induce any INT reduction. In order to further support the involvement of cPLA 2 in NADPH oxidase-associated diaphorase activity, cPLA 2 activity was analyzed in cell lysates using 1-stearoyl-2-[1-14 C]arachidonyl phosphatidylcholine as a substrate (Fig. 8B). PMA, GTP␥S, or FMLP stimulated DTT-resistant cPLA 2 activity in permeabilized differentiated parent PLB cells. Permeabilized differentiated PLB-D cells, which lack cPLA 2 protein, exhibited no detectable DTTresistant cPLA 2 activity either before or after stimulation with PMA, GTP␥S, or FMLP. We have previously demonstrated that PMA-induced translocation of p47 phox , p67 phox , and Rac2 to the membrane is independent of AA released by cPLA 2 in intact stimulated PLB-D cells (19). In accordance with these results, the present study demonstrates that PMA, GTP␥S, or FMLP induced comparable translocation of the three cytosolic factors to the membranes in both permeabilized PLB cells and PLB-D cells (Fig. 8C). Although addition of free AA to stimulated cells restored the activity, it did not cause any augmentation of the translocation of the cytosolic components to the membranes. Thus, the results indicate that AA acts downstream on the membrane-associated oxidase components in intact and permeabilized cells.
The mechanism by which AA acts on the assembled oxidase is not known. Several studies in cell-free systems using fulllength and C-terminally truncated forms of p47 phox and p67 phox suggested a function of AA as an activator that induces conformational changes in both p47 phox and p67 phox that promote their binding to the cytochrome b 558 or Rac2 (41)(42)(43)(44)(45). This model stands in contrast to conclusions drawn from the present study in permeabilized cells, as well as our earlier study in intact cells (19), where cPLA 2 -generated AA was not required for normal membrane translocation of the cytosolic factors, but rather appeared to be required at some later stage of activation of the oxidase complex following its assembly on the membrane. The assembly of membrane-associated oxidase components in whole cells is accompanied by enhanced phosphorylation of several cytosolic components (1, 2), which may induce similar conformational changes as observed in the cell-free systems by binding of soluble amphiphiles such as AA to cytosolic components.
The physiological stimulant, FMLP, which binds to specific receptors on the plasma membranes, and the impermeable stimulant GTP␥S did not elicit any diaphorase activity in differentiated PLB-D cells that lack cPLA 2 , whereas PMA, which acts directly on protein kinase C, induced partial activity in these cells. This difference is intriguing and may serve a useful tool for delineating the precise sites of action of AA on the oxidase-mediated electron transfer process. A series of studies provides evidence that electron transfer in the NADPH oxidase is regulated at several points: at the site where electrons enter the oxidase, at a site between the flavin center and the first heme, and at the level of the membrane-imbedded, heme-containing domain (16 -18, 46 -52). From the experiments performed in GTP␥Sor FMLP-stimulated PLB-D cells, it could be concluded that AA acts on the oxidase at a site prior to the FAD center, because no diaphorase activity was observed in these cells following stimulation with these agents. However, the experiments performed with PMA as a stimulus indicate that AA acts on two different sites on the activated oxidase, one located before the FAD center and the other located downstream from FAD reduction. Differentiated PLB-D cells stimulated with PMA fail to produce any superoxide, but partially reduce FAD, suggesting that the transfer of electrons through the heme center is completely dependent on free AA, whereas the reduction of FAD is less affected by this metabolite. A recent study (16) analyzing the activation of NADPH oxidase demonstrated that the electron transfer reaction was much more rapid when INT was used as electron acceptor than when oxygen alone was the electron acceptor. The authors proposed an intermediate activation state of the oxidase in which electron flow proceeds from NADPH to the enzyme flavin (and hence to INT), but not from flavin to heme. The electron transfer process in PMA-stimulated differentiated PLB-D cells observed in our present study lends further support to this model. In the absence of AA, the association of cytosolic factors with flavocytochrome b was normal and initially results in an intermediate state of activation of NADPH oxidase. In this state, electrons are transferred from NADPH to the FAD center of the enzyme, where electrons can be diverted to INT, although at a reduced rate, whereas transfer does not proceed to the heme center and no superoxide is produced. The intermediate state of activation is converted to the fully active state by AA, resulting in an increased rate of FAD reduction and promotion of electron transfer through the heme centers. Recent data presented by Diebold et al. (18) that analyzed the electron transfer through the oxidase in cell free systems using INT and cytochrome c as electron acceptors show that the electron flow through the FAD center is dependent on Rac2. The data suggests that, although an interaction between Rac2 and p67 phox is required for the complete electron flow through the oxidase electron transfer chain, this interaction is not necessary for electron transfer to the FAD center. These findings provide further evidence indicating the existence of a partially activated state of NADPH oxidase, which allows for electron flow to the FAD center but not beyond it, similar to that observed in differentiated cPLA 2 -deficient PLB cells stimulated by PMA (Fig. 7A). The interaction between Rac2 and p67 phox may serve as a possible target for the conversion of the oxidase to its fully activated state by cPLA 2 -generated AA.
In conclusion, the model of differentiated PLB-D cells lacking any cPLA 2 expression offers a unique tool for exploring the physiological role of cPLA 2 in events leading to activation of NADPH oxidase and its associated diaphorase activity. The development of methods to directly monitor reduction of the flavin center in permeabilized neutrophils and differentiated PLB cells provides an efficient system to study regulation of this process in whole cells. The present study demonstrates that cPLA 2 and its enzymatic production of AA are essential requirements for diaphorase activity induced by GTP␥S or FMLP and partial requirements for diaphorase activity induced by PMA in permeabilized differentiated PLB-D cells. Addition of free AA fully restored diaphorase activity induced by either of the agonists. By comparing the responses to various agonists in these cells, it could be concluded that AA acts on two distinct sites in the assembled oxidase to induce its fully active state, one that precedes reduction of the FAD center and the other located downstream from FAD reduction. The exact mechanism by which AA activates diaphorase activity and NADPH oxidase and its precise sites of action still remain to be elucidated.