Activation of Cytosolic Phospholipase A2α through Nitric Oxide-induced S-Nitrosylation

Cytosolic phospholipase A2α (cPLA2α) is the rate-limiting key enzyme that cleaves arachidonic acid (AA) from membrane phospholipids for the biosynthesis of eicosanoids, including prostaglandin E2 (PGE2), a key lipid mediator involved in inflammation and carcinogenesis. Here we show that cPLA2α protein is S-nitrosylated, and its activity is enhanced by nitric oxide (NO). Forced expression of inducible nitric-oxide synthase (iNOS) in human epithelial cells induced cPLA2α S-nitrosylation, enhanced its catalytic activity, and increased AA release. The iNOS-induced cPLA2α activation is blocked by the specific iNOS inhibitor, 1400W. The addition of the NO donor, S-nitrosoglutathione, to isolated cell lysates or purified recombinant human cPLA2α protein induced S-nitrosylation of cPLA2α in vitro. Incubation of cultured cells with the iNOS substrate l-arginine and NO donor significantly increased cPLA2α activity and AA release. These findings demonstrate that iNOS-derived NO S-nitrosylates and activates cPLA2α in human cells. Site-directed mutagenesis revealed that Cys-152 of cPLA2α is critical for S-nitrosylation. Furthermore, COX-2 induction or expression markedly enhanced iNOS-induced cPLA2α S-nitrosylation and activation, leading to 9-, 23-, and 20-fold increase of AA release and 100-, 38-, and 88-fold of PGE2 production in A549, SG231, and HEK293 cells, respectively, whereas COX-2 alone leads to less than 2-fold change. These results indicate that COX-2 has the ability to enhance iNOS-induced cPLA2α S-nitrosylation and that maximal PG synthesis is achieved by the synergistic interaction among iNOS, cPLA2α, and COX-2. Since COX-2 enhances the formation of cPLA2α-iNOS binding complex, it appears that COX-2-induced augmentation of cPLA2α S-nitrosylation is mediated at least in part through increased association between iNOS and cPLA2α. These findings disclose a novel link among cPLA2α, iNOS, and COX-2, which form a multiprotein complex leading to cPLA2α S-nitrosylation and activation. Therefore, therapy aimed at disrupting this interplay may represent a promising strategy to effectively inhibit PGE2 production and prevent inflammation and carcinogenesis.

NO. Kim et al. (27) reported that iNOS-derived NO S-nitrosylates COX-2 and thus enhances COX-2 catalytic activity in macrophages, which provides mechanistic explanation for NOinduced COX-2 activation. Recently, some NO donors combined with several nonsteroidal anti-inflammatory drugs (e.g. nitroaspirin and nitroflurbiprofen) have been synthesized and used in the treatment of inflammatory and noninflammatory disorders, suggesting the possible beneficial effect of drugs acting on NO and PGs simultaneously. On the other hand, fatty acid has been shown to act as the storage of NO in lipid bilayers and micelles and as the regulator of NO release (28). In human blood plasma and urine, nitrofatty acids are abundantly present (29,30), and nitrolinoleic acid is a potent endogenous ligand for peroxisome proliferator-activated receptor-␥ (31). Despite these achievements, however, the potential effect of NO on the regulation of PLA 2 has not been studied.
This study was designed to investigate whether nitric oxide signaling might modulate cPLA 2 ␣ activation. Our data indicate that cPLA 2 ␣ is S-nitrosylated and activated by iNOS-derived NO in human epithelial cells. This effect is mediated by Cys-152 of cPLA 2 ␣. Thus, S-nitrosylation of cPLA 2 ␣ may represent an important mechanism for regulation of AA release and eicosanoid biosynthesis. Furthermore, we show that COX-2 facilitates the association between iNOS and cPLA 2 ␣, thereby dramatically enhancing iNOS/NO-induced cPLA 2 ␣ S-nitrosylation and PGE 2 synthesis. These results reveal a novel iNOS-mediated functional coupling between cPLA 2 ␣ and COX-2 for efficient PG synthesis in human cells.
Cell Culture, Transfection, and Transduction-Four types of human epithelial cells were utilized in this study (liver epitheliaderived SG231 and CCLP1, lung epithelia-derived A549, and kidney epithelia-derived HEK293). SG231 cells were cultured in ␣-minimal essential medium supplemented with 10% fetal bovine serum (FBS) as we described (32,33). CCLP1 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% FBS (32,33). A549 cells (obtained from ATCC) were cultured in Dulbecco's modified Eagle's medium containing 10% FBS. HEK293 cells (obtained from ATCC) were cultured in Eagle's minimum essential medium containing 10% FBS. All of the cells were cultured at 37°C in a humidified CO 2 incubator. For overexpression of iNOS in these cells, the iNOS virus or control Y5 virus (provided by the University of Pittsburgh Vector Core Facility) at different titers were added into the culture medium without FBS for 2 h, followed by 5% FBS overnight. The cPLA 2 ␣ expression plasmid was transiently transfected into the cells with low expression of endogenous cPLA 2 ␣ (CCLP and HEK293). For this purpose, the cells were seeded on the 6-well plate with 10% FBS the day before transfection. On the following day, the cells in each well (80% confluence) were transfected with 1 g of cPLA 2 ␣ or control MT2 plasmid using Lipofectamine Plus reagent (6 l of Plus reagent, 4 l of Lipofectamine) in serum-free medium. For CCLP1 cells, the transfection medium was replaced with culture medium containing 5% FBS 3 h later. For HEK293 cells, the transfection reagents were removed after 2 h and replaced with regular medium with 5% FBS. After an additional 16-h incubation, the culture medium was collected for PGE 2 production assay, whereas the cells were washed twice with cold PBS and subjected to further analysis as described under "Results." Cytokine Treatment-SG231, A549, and HEK293 cells were seeded onto 6-well plates. After 1 day in culture, the cells at ϳ70 -80% confluence were incubated with vehicle or cytokine mixture (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) in 0.5% FBS medium for 24 h. For adenovirus infection plus cytokine mixture treatment, the cells were incubated with vector for 2 h in 0.5% FBS medium, followed by incubation with cytokine mixture in 0.5% FBS medium overnight.
In Vitro cPLA 2 Activity Assay-2-Deoxy-2-thioarachidonoylphosphatidylcholine was used as the substrate to measure cPLA 2 activity in vitro. The assay was performed by using the cPLA 2 assay kit purchased from Cayman Chemical (catalog number 765021). The secretary PLA 2 and calcium-independent PLA 2 were excluded or deactivated from the whole cell lysate to improve the assay specificity. All procedures were performed according to the manufacturer's instructions with minor modifications. In brief, after transfection with plasmids or infection by adenovirus vector or treatment with different reagents, the cells were scraped down from plates and homogenized in Hepes-EDTA buffer. The homogenates were then centrifuged to remove cell debris at 14,000 rpm for 20 min. To minimize contamination of secretary PLA 2 and calcium-independent PLA 2 , supernatant was concentrated by an Ambion Y30 filter, followed by incubation for 20 min with bromoenol lactone, a calcium-independent PLA 2 inhibitor. 10 l of cell lysate was finally subjected to the assay in the 96-well mode, and the OD value was measured at 414 nm.
[ 3 H]Arachidonic Acid Release-The cells seeded on 6-well plates were treated with various reagents or infected with adenovirus in the presence or absence of cytokine mixture for 24 h. [5,6,8,9,11,12,14, H]arachidonic acid (0.5 Ci/ml) was simultaneously added into culture medium containing 1% serum and left for incorporation into cell membranes overnight. On the next day, the culture media were removed, and the cells were rinsed five times with medium, followed by incubation with medium containing 0.5% serum for 1-2 h. The culture medium was then collected and centrifuged at 14,000 rpm for 20 min to remove suspension cells. 3 H activity was measured using a liquid scintillation counter. PGE 2 Production Assay-The cells were treated with different reagents or infected with adenovirus for 24 h. The culture medium was then collected and centrifuged at 14,000 rpm for 20 min to exclude suspension cells. 50 l of supernatant, with triplicate wells for each sample, was utilized for enzyme immunoassay. Triplicate blank wells and nonspecific binding wells were included to minimize the interference of background color. The binding of PGE 2 with antibody lasted for 2 h at room temperature, which was followed by five rinses with wash buffer to maximally eliminate unbound or nonspecifically bound antibody. 150 l of enzyme substrate was added into each well for 30 min, and the OD value was measured at 630 nm using an enzyme-linked immunosorbent assay reader. The actual PGE 2 concentration was calculated according to the standard curve ranging from 0 to 320 pg/50 l. The samples with too high concentration of PGE 2 were diluted appropriately for further assay.
Detection of S-Nitrosylation of cPLA 2 ␣-A biotin switch assay was utilized to measure the S-nitrosylation of cPLA 2 ␣ according to the previously described methods (34,35) with modifications. For this assay, the unnitrosylated thiol groups are first blocked by S-methyl methanethiosulfonate; the targeted nitrothiols are then reduced to free thiols by ascorbate; the free thiols are subsequently labeled with biotin through a synthesized chemical, N-(6-(biotinamido)hexyl)-3Ј-(2Ј-pyridyldithio)propionamide; finally, the biotinylated protein is immunoprecipitated by streptavidin beads and subjected to visualize by regular Western blotting. Specifically, the cells after cytokine treatment or transfection (as specified) were scraped down from plates in PBS buffer and centrifuged to obtain cell pellets, which were then homogenized in 200 l of HEN buffer (250 mM Hepes-NaOH, pH 7.7, 1 mM EDTA, 0.1 mM neocuproine). The mixture was centrifuged at 2000 ϫ g for 10 min at 4°C, and supernatant was recovered as crude cell lysate. CHAPS (final concentration of 0.4%) solution was added to further dissolve membrane proteins. For measurement of S-nitrosylation induced by NO donor, isolated cell lysate was incubated with GSNO (20 M) for 20 min at room temperature, followed by removal of NO donor through P-6 column (from Bio-Rad). For measurement of S-nitrosylation in cells related to iNOS overexpression, cell lysate was directly blocked in 600 l of blocking buffer (HEN buffer, 2.5% SDS, 20 mM S-methyl methanethiosulfonate) for 20 min at 50°C with frequent vortex. To precipitate proteins, 5 ml chilled acetone was added, and the mixture was placed at Ϫ20°C for 30 min, followed by centrifugation at 5000 ϫ g for 5 min. The protein pellet was then recovered, washed five times with acetone/H 2 O solution (70% acetone, 30% double-distilled H 2 O), and dissolved in 100 l of HENS buffer (HEN buffer, 1% SDS). To reduce and label nitrothiol groups, the solution was rotated with sodium L-ascorbate (100 mM) and N-(6-(biotinamido)hexyl)-3Ј-(2Јpyridyldithio)propionamide (1.5 mM) at room temperature for 1 h, followed by the addition of chilled acetone (1 ml) to precipitate proteins. Protein pellets were washed five times with acetone/H 2 O solution and suspended in 100 l of HENS buffer. To pull down biotin-labeled proteins, 200 l of neutralization buffer (20 mM Hepes-NaOH, 100 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, pH 7.7) and 30 l of immobilized streptavidin beads were added into the solution, and the mixture was rotated for 1 h at room temperature. The formed streptavidin beadprotein complex was then washed five times with neutralization buffer plus NaCl (neutralization buffer with 500 mM NaCl) and two times with neutralization buffer. 30 l of elution buffer (20 mM Hepes-NaOH, 100 mM NaCl, 1 mM EDTA, 100 mM 2-mercaptoethanol, pH 7.7) was applied to the complex to elute proteins from streptavidin beads for 20 min at room temperature. The obtained proteins were then subjected to Western blotting analysis. All of the procedures, except Western blotting, were performed in the absence of direct light. The absence of ascorbate was applied as the assay negative control.
Purification of Human Recombinant cPLA 2 ␣ Protein-Affi-Gel 10-activated affinity support (catalog number 153-6099; Bio-Rad) was applied to purify recombinant human cPLA 2 ␣ protein. This affinity chromatography method utilizes monoclonal anti-cPLA 2 ␣ antibody, which is covalently coupled to the Affi-10 gel. In brief, the cPLA 2 ␣ expression plasmid was transfected into HEK293 cells (without endogenous cPLA 2 ␣ protein) for 24 h and the crude cell lysate was collected. During this period, Affi-10 gel (500 l in 100 mM Hepes, pH 6.8) was mixed with monoclonal cPLA 2 ␣ antibody (500 g, SC-454; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for 2-4 h at 4°C. The coupled gel was then washed three times with Hepes buffer (100 mM, pH 6.8) before the crude lysate was applied to the column. The lysate was passed through the column repeatedly five times, followed by three washes of the column with Hepes buffer (pH 6.8). To elute the specific protein, 100 l of glycine-HCl (100 mM, pH 2.7) was added to the gel, and the protein was collected in a tube containing 100 l of Hepes (100 mM, pH 8.0). The obtained protein was immediately desalted by passing through a Micro Bio-spin P-6 column (pre-equilibrated with 100 mM Hepes, pH 6.8) and then subjected to SDS-PAGE for characterization of its purity. 50 g of purified cPLA 2 ␣ protein was applied to a biotin switch assay or in vitro cPLA 2 activity assay.
Immunoprecipitation and Western Blotting-Cells were lysed in Nonidet P-40 buffer (50 mM Tris-HCl, 2 mM EDTA, 150 mM NaCl, 1% Nonidet P-40, pH 7.2) and homogenized on ice. Cell debris was removed by centrifugation at 14,000 rpm for 20 min at 4°C. Whole cell lysate (ϳ500 g in 100 l) was recovered and incubated with mouse anti-human cPLA 2 ␣ antibody (5 l, 200 g/ml) or mouse IgG for 1 h at 4°C, followed by the addition of protein A/G-agarose beads (30 l) for an extra 1 h. The protein-bead complex was washed four times with Nonidet P-40 buffer and subjected to regular Western blotting. 20 l of sample buffer was added into protein-bead complex, and the mixture was boiled at 90°C for 5 min. All of the supernatant was loaded onto polyacrylamide gel for electrophoresis. The separated proteins in polyacrylamide gels were transferred onto nitrocellulose membrane (Bio-Rad), and the blot was blocked in 5% milk PBS-T (0.5% Tween 20 in buffer PBS) for 1 h at room temperature. The membrane was then incubated overnight with primary antibodies (1:1000) in 5% milk PBS-T. Following repeated washing with PBS-T the next day, the membranes were incubated with the horseradish peroxidase-conjugated secondary antibody (1:10,000 dilution) for 1 h at room temperature. After washing, the blots were developed using the ECL Western blotting detection system and exposed to Eastman Kodak Co. MR radiographic films.
Statistical Analysis-Statistical analysis was performed using Microsoft Excel 2003 software. Comparisons were performed using two-tailed unpaired Student's t test. Values of p Ͻ 0.05 were considered statistically significant.

Overexpression of iNOS Induces cPLA 2 ␣ S-Nitrosylation and
Activation-Given the documented role of cPLA 2 ␣ in AA metabolism and the known cross-interactions between iNOS and prostaglandin signaling pathways in human cells (25,26), we sought to determine whether iNOS could regulate AA metabolism through S-nitrosylation of cPLA 2 ␣ protein. For this purpose, two human epithelial cell lines with high constitutive cPLA 2 ␣ expression (SG231 and A549) were transduced with the iNOS adenovirus expression vector or the control Y5 vector to determine the occurrence of cPLA 2 ␣ S-nitrosylation. As shown in Fig. 1A, forced expression of iNOS induced S-nitrosylation of cPLA 2 ␣ in both SG231 cells and A549 cells. Overexpression of iNOS had no effect on cPLA 2 ␣ protein level (Fig.  1A) or phosphorylation (see below). The level of cPLA 2 ␣ S-nitrosylation is directly proportional to the amount of iNOS pro-tein expressed in those cells. Furthermore, iNOS overexpression significantly increased the release of AA from both SG231 and A549 cells (Fig. 1B). These data suggest that modulation of cPLA 2 ␣ by iNOS-mediated S-nitrosylation probably increases cPLA 2 ␣ enzymatic activity. Indeed, in vitro cPLA 2 activity assays confirmed that iNOS overexpression enhanced cPLA 2 activity by 3-fold in SG231 cells and 2-fold in A549 cells (Fig.  1C). These results provide the first evidence that iNOS is able to S-nitrosylate and activate cPLA 2 ␣ in human cells. The observation that the iNOS-induced increase of cPLA 2 activity is blocked by the specific iNOS inhibitor, 1400W, suggests that iNOS-mediated production of NO is required for cPLA 2 ␣ S-nitrosylation (Fig. 1C).
NO Donor Causes S-Nitrosylation and Activation of cPLA 2 ␣-We next utilized NO donor to evaluate the direct effect of NO on cPLA 2 ␣ S-nitrosylation and enzymatic activity. The addition of the NO donor GSNO to the cell lysates isolated from SG231 cells resulted in S-nitrosylation of cPLA 2 ␣, in vitro ( Fig. 2A). The assay specificity was confirmed by the observation that omission of ascorbic acid in the assay system completely abolished the detection of S-nitrosylated cPLA 2 ␣. Accordingly, incubation of cultured SG231 cells with GSNO significantly increased cPLA 2 activity in vitro (Fig. 2B). Increased cPLA 2 activity was also seen in the SG231 cells treated with the iNOS substrate L-Arg; this effect was blocked by the iNOS inhibitor 1400W (Fig. 2B). Furthermore, L-Arg, GSNO, and NOC-18 (another NO donor) also increased the release of AA from SG231 as well as from A549 cells (Fig. 2C). These observations document that NO donors are capable of inducing cPLA 2 ␣ S-nitrosylation and activation.
We further examined the effect of NO donor in cells with a low basal level of cPLA 2 ␣ protein (CCLP1). Whereas GSNO did not alter cPLA 2 activity in CCLP1 cells, it increased cPLA 2 activity when the CCLP1 cells were transfected with the cPLA 2 ␣ expression plasmid (Fig. 2D). These findings suggest that the S-nitrosylation and activation of cPLA 2 ␣ in cells is in direct proportion to the level cPLA 2 ␣ protein as well as the availability of NO.
To further characterize the direct effect of NO on cPLA 2 ␣ protein S-nitrosylation and to exclude the potential influence of other molecules, cPLA 2 ␣ expression plasmid was transfected into HEK293 cells, and the recombinant cPLA 2 ␣ protein was purified by affinity chromatography. The purified recombinant cPLA 2 ␣ protein was then incubated with the NO donor to determine its S-nitrosylation and enzyme activity, in vitro. As shown in Fig. 3A, GSNO induces the S-nitrosylation of purified cPLA 2 ␣ protein. Accordingly, the NO donors (GSNO and NOC-18) also increased the enzyme activity of purified cPLA 2 ␣ (Fig. 3B). These results demonstrate that NO is capable of directly S-nitrosylating and activating cPLA 2 ␣.
The iNOS-induced cPLA 2 ␣ S-Nitrosylation and AA Release Is Enhanced by COX-2-In light of the documented functional coupling between cPLA 2 ␣ and COX-2 in human cells (5,(21)(22)(23)(24), we further examined whether COX-2 might influence the iNOS-induced cPLA 2 ␣ S-nitrosylation. For this purpose, HEK293 cells were transfected with the expression vector for cPLA 2 ␣, COX-2 and iNOS, either alone or in combination, to determine the S-nitrosylation of cPLA 2 ␣. HEK293 cells were utilized because they do not constitutively express cPLA 2 ␣, COX-2, or iNOS and allow high transfection efficiency. As shown in Fig. 4A, although S-nitrosylation of cPLA 2 ␣ was observed in HEK293 cells cotransfected with the iNOS and cPLA 2 ␣ vectors, much more S-nitrosylated cPLA 2 ␣ is detected with the combination of COX-2 transfection. These findings indicate that COX-2 has the ability to enhance iNOS-induced cPLA 2 ␣ S-nitrosylation. To further examine the effect of endogenous COX-2 on cPLA 2 ␣ S-nitrosylation, HEK293 cells were treated with the cytokine mixture containing IL-1␤, TNF-␣, and IFN-␥ to induce COX-2 expression (these cytokines are known to induce COX-2 gene expression in other cells). As expected, treatment of the HEK293 cells with the cytokine mixture effectively induced the expression of COX-2, without alteration of cPLA 2 ␣ expression, cPLA 2 ␣ phosphorylation, or iNOS expression (Fig. 4B). Notably, the combination of iNOS expression with induction of endogenous COX-2 led to dramatic increase of cPLA 2 ␣ S-nitrosylation (Fig. 4C). This effect was blocked by the iNOS inhibitor, 1400W (Fig. 4C), further suggesting the requirement of iNOS activity for cPLA 2 ␣ S-nitrosylation. Accordingly, the release of AA was also dramatically enhanced by the combination of iNOS overexpression and COX-2 induction (ϳ20-fold), compared with a 4-fold increase in the absence of COX-2 (Fig. 4D). Collectively, these findings demonstrate that COX-2 facilitates iNOS-induced cPLA 2 ␣ S-nitrosylation.
The Activation of Endogenous cPLA 2 ␣ Is Enhanced by Combination of iNOS Expression and COX-2 Induction-After the effect of iNOS and COX-2 on the S-nitrosylation and activation of transfected (exogenous) cPLA 2 ␣ was documented in the HEK393 cells, we further evaluated the effect of combined iNOS and COX-2 on the activation of endogenously expressed cPLA 2 ␣ in SG231 cells (with a high level of constitutive cPLA 2 ␣ expression). As shown in Fig. 5A, cytokine treatment of SG231 cells also induced the expression of COX-2 protein, with no effect on the level of cPLA 2 ␣ protein, cPLA 2 ␣ phosphorylation, or iNOS expression. Combination of iNOS expression and COX-2 induction resulted in an ϳ24-fold increase of AA release in SG231 cells, which was in contrast to the 3-fold increase with iNOS expression alone (Fig. 5B). These results further reveal the synergistic effect of COX-2 and iNOS on cPLA 2 ␣ activation.
Physical Association between iNOS and cPLA 2 ␣: Effect of COX-2-NO is a highly diffusible and short lived physically active gas. The proximity between NOS and targeted proteins adenovirus for 24 h, and cell lysates were collected to determine cPLA 2 ␣ S-nitrosylation by biotin switch as described under "Experimental Procedures." The same samples were also subjected to regular Western blotting for detection of total cPLA 2 ␣ and iNOS. ␤-Actin was utilized to normalize protein loading. B, iNOS overexpression increases AA release. SG231 and A549 cells infected with different titers of Y5 or iNOS adenovirus were incubated with [5,6,8,9,11,12 has been shown to help form protein S-nitrosylation (27,36). We utilized immunoprecipitation and Western blot analysis to determine the association between cPLA 2 ␣ and iNOS. In HEK293 cells, the cPLA 2 ␣-iNOS association complex was detected in cells with forced expression of both cPLA 2 ␣ and iNOS but not in cells with expression of either molecule alone (Fig. 6A). In SG231 cells, the cPLA 2 ␣-iNOS association complex was detected in cells transduced with the iNOS adenovirus vector but not in cells infected with the Y5 control vector (Fig.  6B). These findings indicate the proximity between cPLA 2 ␣ and iNOS in cells, which may explain the apparent effect of iNOS on cPLA 2 ␣ S-nitrosylation. Importantly, COX-2 induction increased the amount of cPLA 2 ␣ immunoprecipitated by iNOS in both HEK293 and SG231 cells (Fig. 6, A and B). Therefore, COX-2 enhances the association between iNOS and cPLA 2 ␣; this effect may explain the synergistic effect of COX-2 on iNOS-induced cPLA 2 ␣ S-nitrosylation. Since the COX-2 inhibitor, NS398, failed to prevent COX-2-induced association between iNOS and cPLA 2 ␣ (Fig. 6C), it is likely that COX-2 may facilitate iNOS-cPLA 2 ␣ association through COX-2 protein itself rather than through its enzymatic products.
In light of the reported binding of COX-2 to iNOS (27), we performed further immunoprecipitation and Western blot analysis to examine the potential association between cPLA 2 ␣ and COX-2 in our system. As shown in Fig. 6D, cPLA 2 ␣ is associated with COX-2 in SG231 cells transduced with the iNOS adenoviral expression vector plus cytokine treatment; this association is minimal in the absence of iNOS. These observations suggest an impor- The cell lysate was then collected and subjected to a cPLA 2 activity assay (a, p Ͻ 0.05 compared with vehicle; b, p Ͻ 0.05 compared with L-arginine treatment; n ϭ 3). C, the effect of NO donor and iNOS substrate on AA release. SG231 and A549 cells were incubated with [5,6,8,9,11,12,14,   Prior to the biotin switch assay, cPLA 2 ␣ protein was purified from the crude cell lysates and exposed to GSNO (20 M) for 30 min. B, the NO donors, GSNO and NOC-18, enhance the activity of purified cPLA 2 ␣, in vitro. HEK293 cells were transfected with the cPLA 2 ␣ expression plasmid, and the cPLA 2 ␣ protein was purified from the crude lysates as described under "Experimental Procedures." The purified cPLA 2 ␣ protein was then subjected to a cPLA 2 activity assay in vitro (*, p Ͻ 0.01 versus vehicle; n ϭ 3). tant role of iNOS for the association between cPLA 2 ␣ and COX-2. Therefore, it is possible that iNOS may recruit cPLA 2 ␣ and COX-2 to form a multiprotein complex.
iNOS Induces Maximal PGE 2 Production in the Presence of cPLA 2 ␣ and COX-2-Since cPLA 2 ␣ activation leads to the release of AA, providing substrate for COX-2-mediated PGE 2 synthesis, we further examine the effect of iNOS, cPLA 2 ␣, and COX-2 in different combinations on PGE 2 production. In HEK293 cells, combined iNOS expression and COX-2 induction in the absence of cPLA 2 ␣ induced only less than a 2-fold increase of PGE 2 (Fig. 7). When the cells were transfected with the cPLA 2 ␣ expression vector, either iNOS expression or COX-2 induction induced a 2-3-fold increase of PGE 2 . Interestingly, the combination of cPLA 2 ␣, COX-2, and iNOS dramatically enhanced PGE 2 production by more than 80-fold ( Fig. 7). It is noteworthy that overexpression of cPLA 2 ␣ and iNOS in combination with COX-2 induction achieved a much higher increase of PGE 2 production than either iNOS plus cPLA 2 ␣ or cPLA 2 ␣ plus COX-2. A similar synergistic effect was also observed in the cells with endogenous expression of cPLA 2 ␣ (SG231 and A549) (Fig. 8). iNOS overexpression plus COX-2 induction increased the production of PGE 2 by nearly 40-fold in SG231 cells (Fig. 8A) and ϳ100-fold in A549 cells (Fig. 8B). These results demonstrate that maximal PG synthesis is achieved by the synergistic interactions among iNOS, cPLA 2 ␣, and COX-2.

DISCUSSION
NO is a bioactive molecule with a single unpaired electron. The two major mechanisms whereby NO influences its intra-cellular targets are stimulation of guanylyl cyclase by direct binding of NO to iron in heme at the active site of guanylyl cyclase (37) and S-nitrosylation of protein targets on appropriate cysteines (34,36,38). S-Nitrosylation refers to the NO adduct with cysteine residue; it is increasingly recognized as an important posttranslational modification mechanism that mediates the diverse actions of NO. Our data in this study demonstrate that iNOS/NO S-nitrosylates and activates cPLA 2 ␣, which represents a novel mechanism for the regulation of cPLA 2 ␣ in human cells. This assertion is based on the following observations: 1) iNOS overexpression induces cPLA 2 ␣ S-nitrosylation, enhances its catalytic activity, and increases the  . cPLA 2 ␣, iNOS, and COX-2 are physically associated. A, co-immunoprecipitation (IP) of cPLA 2 ␣ with iNOS in HEK293 cells. HEK293 cells were transfected with cPLA 2 ␣ or MT2 plasmid, infected with iNOS (50 MOI) or Y5 adenovirus, or treated with cytokine mixture (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) for 24 h. The cell lysates were collected and incubated with anti-cPLA 2 ␣ antibody or mouse IgG (negative control) for 1 h at 4°C, followed by incubation with protein A/G-agarose beads for 40 min. After five washes with lysis buffer, the protein-bound beads were subjected to Western blotting (WB) for detection of iNOS. B, co-immunoprecipitation of cPLA 2 ␣ with iNOS in SG231 cells. SG231 cells were transfected with iNOS (100 MOI) or control Y5 adenovirus and/or treated with cytokine mixture (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) or vehicle Me 2 SO for 24 h. The cell lysates were then processed for immunoprecipitation and Western blot analysis. C, the effect of the COX-2 inhibitor, NS-398, on cPLA 2 ␣ and iNOS association in SG231 cells. SG231 cells were infected with iNOS adenovirus (100 MOI), with or without cytokines (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) or NS-398 (25 M) treatment for 24 h. The cell lysates were collected and subjected to immunoprecipitation and Western blotting. D, effect of iNOS on cPLA 2 ␣ and COX-2 association. SG231 cells were exposed to cytokines (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) with or without infection of iNOS adenovirus (100 MOI) for 24 h. The cell lysates were then collected and subjected to cPLA 2 ␣ immunoprecipitation and COX-2 Western blotting. Each figure is representative of three separated experiments. release of AA from cells; 2) mutation of Cys-152 of cPLA 2 ␣ abolishes iNOS-induced cPLA 2 ␣ S-nitrosylation and enzymatic activation; 3) the level of cPLA 2 ␣ S-nitrosylation is in direct proportion to the amount of iNOS and cPLA 2 ␣ proteins expressed in the cells; 4) the iNOS-induced increase of cPLA 2 activity is blocked by the specific iNOS inhibitor, 1400W; 5) the addition of the NO donor, GSNO, to isolated cell lysates induces S-nitrosylation of cPLA 2 ␣, in vitro; 6) incubation of cultured cells with GSNO significantly increases cPLA 2 activity; 7) exposure of purified cPLA 2 ␣ protein to NO donor causes its S-nitrosylation and activation; 8) the iNOS substrate L-arginine increases cPLA 2 activity, and the effect is blocked by the iNOS inhibitor; 9) L-arginine and NO donors also increase the release of AA from cells; and 10) cPLA 2 ␣ is physically associated with iNOS.
The observation that NO donor induces cPLA 2 ␣ S-nitrosylation in isolated cell lysates and purified recombinant protein in vitro suggests direct modification of cPLA 2 ␣ by NO adduct. This is also supported by the results from cultured cells showing that overexpression of iNOS or the addition of iNOS substrate or NO donor induced cPLA 2 ␣ S-nitrosylation and activation and that specific iNOS inhibitor blocked the iNOS overexpression or substrate-induced cPLA 2 ␣ activation. Our data show that S-nitrosylated cPLA 2 ␣ exhibits enhanced enzymatic activity. This observation is consistent with other studies showing that S-nitrosylated proteins display alterations in enzymatic activity, trafficking through intracellular compartments, stability in intracellular environment, binding with proteins, and other functions (34, 36, 39 -45).
The occurrence of S-nitrosylation and concomitant modulation of protein function are regulated by factors including NO concentration fluctuation, other redox signaling, proximity of NO with targeted protein, intracellular location of protein, iron concentration, and the presence of denitrosylation, thus far beyond control and prediction under random conditions (46). In this study, the association between iNOS and cPLA 2 ␣ proteins provides a mechanistic explanation for the noticeable effect of iNOS on cPLA 2 ␣ S-nitrosylation, which is consistent with the documented proximity between NOS and other targeted proteins for efficient S-nitrosylation (27,36). We  were infected with iNOS (100 MOI) or Y5 adenovirus and treated with cytokine mixture (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) for 24 h. The cell culture supernatant was collected and subjected to a PGE 2 assay (a, p Ͻ 0.01 compared with Y5 ϩ cytokines or iNOS ϩ vehicle; b, p Ͻ 0.05 compared with Y5 ϩ vehicle; n ϭ 3). B, effect of iNOS and COX-2 on PGE 2 production in A549 cells. A549 cells labeled with [ 3 H]AA were infected with iNOS (500 MOI) or Y5 (500 MOI) adenovirus with or without cytokine mixture (50 ng/ml IL-1␤, 50 ng/ml TNF-␣, 100 ng/ml IFN-␥) for 24 h. The culture supernatant was collected for the PGE 2 assay (a, p Ͻ 0.01 compared with Y5; b, p Ͻ 0.01 compared with Y5 ϩ cytokines; c, p Ͻ 0.01 compared with Y5 ϩ vehicle, n ϭ 3).
observed that transduction of iNOS adenovirus vector led to more S-nitrosylation of cPLA 2 ␣ than NO donors; this effect is most likely due to the higher concentration of NO accumulated at specific organelle or intracellular membrane in iNOS-overexpressed cells or due to the physical interaction between iNOS and cPLA 2 ␣, which allows efficient transfer of produced NO to cPLA 2 ␣.
sible that the illustrated interaction among iNOS, cPLA 2 ␣, and COX-2 may importantly contribute to the production of large amounts of PGs and NO that simultaneously drive and perpetuate inflammatory processes.
In summary, our results presented in this study demonstrate a novel iNOS/NO-mediated cPLA 2 ␣ S-nitrosylation and activation for prostaglandin synthesis and the striking synergistic effect of COX-2 in this process. Given the high magnitude of PGE 2 production generated by the synergistic interaction among iNOS, cPLA 2 ␣, and COX-2, it is possible that this mechanism may be implicated in the synthesis of bioactive prostaglandins during disease processes, such as inflammation and carcinogenesis. It is conceivable that disruption of the association among cPLA 2 ␣, iNOS, and COX-2 may achieve more effective inhibition of AA release and PGE 2 synthesis than conventional pharmacological inhibitors. This intervention may represent a highly effective therapeutic approach that merits further investigation.