Formation of Murine Macrophage-derived 5-Oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG 7 ) Is Catalyzed by Leukotriene C 4 Synthase*

5-Oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG 7 ), a biologically active glutathione (GSH) adduct of the eicosanoid 5-oxo-eicosatrienoic acid (5-oxoETE), is the major metabolite formed within the murine peritoneal macrophage. The conjugation of GSH to electrophilic 5-oxoETE in vitro was found to be catalyzed by both solu-ble glutathione S -transferase and membrane-bound leukotriene C 4 (LTC 4 ) synthase. The cytosolic glutathione S -transferase-catalyzed products were not biologically active; however, the adduct formed from recombinant LTC 4 synthase had identical mass spectrometric proper- ties and biological activity to the macrophage-derived FOG 7 . The biosynthesis of FOG 7 in the macrophage was inhibited by MK-886, a known inhibitor of LTC 4 synthase, suggesting that this nuclear membrane-bound enzyme might be responsible for GSH conjugation to 5-oxoETE in the intact cell. Subcellular fractionation revealed that the microsomal fraction from the murine macrophage con-tained the enzyme responsible for FOG 7 biosynthesis. Western blot analysis confirmed the presence of LTC 4 synthase in the microsomal fraction that did not catalyze conjugation of GSH to 1-chloro-2,4-dinitrobenzene, The free acid of LTA 4 was synthesized by the hydrolysis of LTA 4 methyl ester as previously described (24). Collection of Elicited Peritoneal Macrophages— Elicited macrophages were obtained by injecting 1 ml of thioglycolate 4% (10%) into the peritoneum of ICR mice. After 3 days, the mice were euthanized in a CO 2 atmosphere. The peritoneum was then lavaged once with 10 ml of Dulbecco’s modified Eagle’s medium, 10% fetal bovine serum, 100 units/ml penicillin, and 100 (cid:1) g/ml streptomycin with 1% heparin. The peritoneal lavage fluid obtained was centrifuged at 600 (cid:2) g for 8 min for the separation of cells from fluid. Isolation of Neutrophils— Leukocytes were prepared from human peripheral blood in EDTA anticoagulant by dextran 60 sedimentation, contaminating erythrocytes were removed by brief (45 s) hypotonic lysis, and neutrophils were purified on Ficoll-Hypaque, yielding 96– 97% neutrophils, 2–3% eosinophils, 0–1% mononuclear cells. All exper-iments were done within a 2-h in vitro age of the cell. Preparation of Human Platelets— Peripheral blood was collected from healthy volunteers. The blood (20 ml) was treated with 2 ml of 77 m M EDTA (in saline). After two 15-min centrifugation steps at 100 (cid:2) g at room temperature to remove contaminating red blood cells, the platelet-rich plasma was acidified by the addition of 1 ⁄ 10 volume of ACD (100 m M sodium citrate, 41 m M citric acid, and 136 m M glucose) and centrifuged at 1,000 (cid:2) g for 15 min at room temperature. The platelets were washed using the method of Patscheke (25) in citrate buffer, pH 6.5, containing 0.4% bovine serum albumin, 100 n M PGE 1 . Platelets were resuspended in Hanks’

5-Oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG 7 ), a biologically active glutathione (GSH) adduct of the eicosanoid 5-oxo-eicosatrienoic acid (5-oxoETE), is the major metabolite formed within the murine peritoneal macrophage. The conjugation of GSH to electrophilic 5-oxoETE in vitro was found to be catalyzed by both soluble glutathione S-transferase and membrane-bound leukotriene C 4 (LTC 4 ) synthase. The cytosolic glutathione S-transferase-catalyzed products were not biologically active; however, the adduct formed from recombinant LTC 4 synthase had identical mass spectrometric properties and biological activity to the macrophage-derived FOG 7 . The biosynthesis of FOG 7 in the macrophage was inhibited by MK-886, a known inhibitor of LTC 4 synthase, suggesting that this nuclear membrane-bound enzyme might be responsible for GSH conjugation to 5-oxoETE in the intact cell. Subcellular fractionation revealed that the microsomal fraction from the murine macrophage contained the enzyme responsible for FOG 7 biosynthesis. Western blot analysis confirmed the presence of LTC 4 synthase in the microsomal fraction that did not catalyze conjugation of GSH to 1-chloro-2,4-dinitrobenzene, indicating an absence of microsomal glutathione Stransferase activity. These results suggest that LTC 4 synthase, thought to be specific for the conjugation of GSH to LTA 4 , can also recognize 5-oxoETE as an electrophilic substrate.
The enzymatic oxidation of arachidonic acid plays an important role in biology, leading to the production of a diverse family of biologically active eicosanoids, which typically carry information between cells, acting as intracellular chemical communicators of cellular activation. One pathway of arachidonic acid oxidation involves the addition of molecular oxygen to carbon-5 of arachidonic acid to afford 5-hydroperoxyeicosatetraenoic acid (1, 2), a reaction catalyzed by 5-lipoxygenase, but also a product of free radical oxidation of arachidonate. Leukotrienes are derived from the chemically reactive intermediate leukotriene A 4 (LTA 4 ), 1 which is the product of a second 5-lipoxygenase-mediated reaction that utilizes 5-hydroperoxyeicosatetraenoic acid as substrate. LTA 4 is transformed either into the neutrophil chemotactic leukotriene B 4 (LTB 4 ) (3) through the LTA 4 hydrolase-catalyzed addition of water to LTA 4 (4) or by conjugation of the tripeptide glutathione (GSH) by LTC 4 synthase to yield leukotriene C 4 (LTC 4 ) (5). LTC 4 is rapidly metabolized through a series of peptidic cleavage reactions by ectoenzymes to the cysteinyl-glycine leukotriene D 4 (LTD 4 ) and the cysteine leukotriene E 4 (LTE 4 ) (6). These three cysteinyl leukotrienes were previously known as slow reacting substances of anaphylaxis (7) and are synthesized by several inflammatory cell types including the eosinophils, mast cells, basophils, macrophages, and platelets (8 -10). As a family, cysteinyl leukotrienes possess potent biological activities causing contraction of various smooth muscles and have been implicated as mediators of acute hypersensitivity reactions including asthma (11).
Although considerable interest has focused attention on the leukotriene pathway of arachidonic acid metabolism within cells, it is now recognized that another family of equally potent eicosanoids is formed through the metabolism of 5-hydroperoxyeicosatetraenoic acid via the substrate 5-hydroxyeicosatetraenoic acid, which itself is metabolized into 5-oxoETE by a NADP ϩ -dependent dehydrogenase in the neutrophil (12,13). The discovery that 5-oxoETE is a potent chemotactic factor for the eosinophil has raised interest in this eicosanoid because of a suggested role for eosinophils in the pathogenesis of asthma (14,15). Recently, a new biologically active cysteinyl 5-lipoxygenase product was structurally characterized as 5-oxo-7-glutathionyl-8,11,14-eicosatrienoic acid (FOG 7 ), which was shown to be chemotactic for both the eosinophil and neutrophil (16). FOG 7 and LTC 4 are the only known biologically active GSH adducts of arachidonic acid and are, interestingly, isobaric, with a molecular mass of 625 daltons.
The conjugation of GSH to various endogenous and exogenous electrophiles is not uncommon in biological systems due to the presence of numerous glutathione S-transferases (GSTs). The GSTs make up a complex multigene family of proteins that play a central role in detoxifying electrophilic xenobiotics in nearly all species studied (17). The primary function of these proteins is to catalyze the nucleophilic conjugation of GSH to exogenous and endogenous electrophiles by effectively increasing the concentration of the thiolate anion near the substrate when held in the active site (18). GSTs are typically cytosolic enzymes; however, there are four known microsomal GSTs including microsomal GST-I (19), microsomal GST-II (20), mi-crosomal GST-III (21), and LTC 4 synthase (5). LTC 4 synthase, the enzyme responsible for the biosynthesis of LTC 4 , differs from conventional GSTs by its selectivity for LTA 4 and closely related analogs and failure to conjugate GSH to xenobiotics (22). LTC 4 synthase also exhibits differential susceptibility to inhibitors (24) and lacks immunoreactivity to antibodies for known GSTs (23). The purpose of the present investigation was to determine which of these enzymes was responsible for the 1,4-Michael addition of GSH to 5-oxoETE 2 in vivo to afford biologically active FOG 7 in the murine peritoneal macrophage. [5][6]8,9,11,12,14,]oxoETE (greater than 99 atom % D 7 ), LTA 4 methyl ester, and LTC 4 were purchased from the Cayman Chemical Co. (Ann Arbor, MI). Hanks' balanced salt solution (HBSS) was purchased from Invitrogen. Indo-1/AM was obtained from Calbiochem. Reduced glutathione (GSH), 1-chloro-2,4-dinitrobenzene (CDNB), lysophosphatidylcholine, human placental (89 enzyme units/mg of protein), and rat liver GST (81 enzyme units/mg of protein) were purchased from Sigma. NBD-phallicidin (N-(7-nitrobenz-2-oxa-1,3-diazal-4-yl)phallicidin) was obtained from Molecular Probes (Eugene, OR). All solvents were HPLC grade and obtained from Fisher. MK-886 was a kind gift from Anthony Ford-Hutchinson (Merck). Recombinant human LTC 4 synthase and LTC 4 synthase polyclonal antibody were kind gifts from K. F. Austen and B. Lam (Harvard, Boston, MA). FOG 7 was synthesized from peritoneal macrophages as previously described (16). The free acid of LTA 4 was synthesized by the hydrolysis of LTA 4 methyl ester as previously described (24).

Materials
Collection of Elicited Peritoneal Macrophages-Elicited macrophages were obtained by injecting 1 ml of thioglycolate 4% (10%) into the peritoneum of ICR mice. After 3 days, the mice were euthanized in a CO 2 atmosphere. The peritoneum was then lavaged once with 10 ml of Dulbecco's modified Eagle's medium, 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin with 1% heparin. The peritoneal lavage fluid obtained was centrifuged at 600 ϫ g for 8 min for the separation of cells from fluid.
Isolation of Neutrophils-Leukocytes were prepared from human peripheral blood in EDTA anticoagulant by dextran 60 sedimentation, contaminating erythrocytes were removed by brief (45 s) hypotonic lysis, and neutrophils were purified on Ficoll-Hypaque, yielding 96 -97% neutrophils, 2-3% eosinophils, 0 -1% mononuclear cells. All experiments were done within a 2-h in vitro age of the cell.
Preparation of Human Platelets-Peripheral blood was collected from healthy volunteers. The blood (20 ml) was treated with 2 ml of 77 mM EDTA (in saline). After two 15-min centrifugation steps at 100 ϫ g at room temperature to remove contaminating red blood cells, the platelet-rich plasma was acidified by the addition of 1 ⁄10 volume of ACD (100 mM sodium citrate, 41 mM citric acid, and 136 mM glucose) and centrifuged at 1,000 ϫ g for 15 min at room temperature. The platelets were washed using the method of Patscheke (25) in citrate buffer, pH 6.5, containing 0.4% bovine serum albumin, 100 nM PGE 1 . Platelets were resuspended in Hanks' buffer containing 1 mg/ml bovine serum albumin, pH 7.4, without Ca 2ϩ and Mg 2ϩ at 6 ϫ 10 8 platelets/ml.
Preparation of Cytosolic and Microsomal Macrophage Fractions-All steps below were carried out at 4°C. Macrophages (80 ϫ 10 6 cells) in HBSS were centrifuged at 1,500 ϫ g for 5 min. The pellet was washed with 10 mM HEPES buffer, pH 6.7, containing 137 mM NaCl, 2.6 mM KCl, 0.36 mM NaH 2 PO 4 , and 1 mM EDTA containing aprotinin (5 g/ml) and leupeptin (5 g/ml). The cell suspension was lysed by nitrogen cavitation (700 p.s.i., 20 min), and the lysis solution was centrifuged at 1,500 ϫ g for 20 min to pellet cellular debris. The supernatant fraction was collected and centrifuged at 100,000 ϫ g for 60 min. The pellet, containing microsomal enzymes (total protein 1.1 mg) was separated from the cytosolic enzymes in the supernatant (total protein 7.2 mg). The microsomal fraction was resuspended in HEPES buffer containing Triton X-100 (0.3%).
GST-catalyzed Synthesis of GSH-5-oxoETE Adducts-The addition of GSH (2 mM) to 5-oxoETE (10 M) was carried out in HBSS with the presence of either human placental GST (5 units), rat liver GST (5 units), or recombinant human LTC 4 synthase (1.2 g, partially purified from SF9 expression cells) for 15 min at 37°C. GSH-5-oxoETE adducts were collected by centrifugation and separation using solid phase extraction. The above procedure was also carried out at pH 9 with D 7 -5-oxoETE (10 M) in the presence of human placental GST (5 units) to afford the GSH-D 6 -5-oxoETE adduct internal standard (greater than 95% D 6 ).
GSH-5-oxoETE Adduct Isolation and Purification-The methanol supernatant, after solid phase extraction, was evaporated to dryness by vacuum rotary evaporation and redissolved in 60 l of the initial HPLC mobile phase. Reverse phase HPLC was used to separate the GSH-5-oxoETE adducts by gradient elution with mobile phase A containing 8.3 mM acetic acid buffered at pH 5.2 with NH 4 OH and mobile phase B composed of CH 3 CN:methanol (65:35, v/v). GSH-5-oxoETE adducts were separated on a 150 ϫ 2.00-mm Columbus 5-m C18 RP HPLC column (Phenomenex, Rancho Palos Verdes, CA), and fractions were collected at 1-min intervals from the column eluted at 200 l/min with a linear gradient from 15% to 55% B in 10 min to 80% B in 25 min. Measurement of Cytosolic Calcium Levels-Intracellular cytosolic calcium was assessed by incubation of neutrophils (10 7 cells/ml) loaded with the acetoxymethyl ester of Indo-1 (Indo-1/AM) as described previously (26). Before the addition of each test substance, CaCl 2 and MgCl 2 were added to the cell suspensions at 1 mM final concentration to 3 ϫ 10 6 neutrophils in a 4-ml cuvette. The K D of 250 nM for the Indo-1/AM Ca 2ϩ complex was used to calculate the intracellular calcium concentration, a F max was determined by the addition of digitonin at 0.1%, and F min was determined by the addition of 7.8 mM EGTA in Tris buffer.
F-actin Polymerization Determination-Actin polymerization was assessed by flow cytometry as described previously (27). Briefly, neutrophils (0.9 ml of 1 ϫ 10 6 cell/ml) were incubated at 37°C in HBSS in the presence of the test substance. After a 30-s incubation, cells were fixed, permeabilized, and stained in a single step by the addition of 0.1 ml of 37% phosphate-buffered formalin containing 1.65 ϫ 10 Ϫ6 M NBD-phallicidin and 100 g of lysophosphatidylcholine. The stain mixture plus cells was incubated for 10 min at 37°C. Cells were centrifuged at 400 ϫ g for 5 min at room temperature and resuspended in HBSS (1 ml) before analysis.
Enzyme Assays-GST activity was measured using GSH and CDNB as substrates (28). The activity of the enzyme was determined in a 0.1 M potassium phosphate buffer, pH 6.5, containing 1 mM GSH and 1 mM CDNB using an extinction coefficient of 9.6 mM Ϫ1 cm Ϫ1 . The rate of product formation was monitored by measuring the change in absorbance at 340 nm. LTC 4 synthase activity was determined as previously described by measuring the formation of LTC 4 methyl ester by reversed phase HPLC after incubation of samples with LTA 4 methyl ester (10 M) and GSH (2 mM) in HBSS (29). Protein concentrations were determined using the method described by Bradford (30). Gel electrophoresis and antibody recognition of LTC 4 synthase was carried out essentially as previously described (31) using a peptide antibody directed against GPPEFERVYRAQVN in the sequence for LTC 4 synthase. 3 Electrospray Mass Spectrometry (Negative Ions)-Analysis of GSH-5-oxoETE and LTC 4 production was carried out using a Sciex API 3000 triple quadrupole mass spectrometer (PE Sciex, Thornhill, Ontario, Canada). Multiple reaction monitoring of the specific transitions m/z 624 3 306, m/z 630 3 306, and m/z 624 3 272 were used to detect the elution of GSH-5-oxoETE adducts including FOG 7 , the D 6 -GSH-5-ox-oETE adduct internal standard, and LTC 4 eluting from the HPLC 2 Nomenclature is according to the following. The conjugation of GSH to 5-oxoETE can be catalyzed by various GSTs in vitro. These molecules will be referred to as GSH-5-oxoETE adducts as opposed to FOG 7 , which is the GSH-5-oxoETE adduct biosynthesized in the intact murine macrophage and is biologically active. Formation of the GSH-D 6 -5-oxoETE adduct used as an internal standard for these studies was catalyzed by human placental GST. 3 B. Lam, personal communication.
column, respectively. Each sample was chromatographed using a 1-mm ϫ 150-mm Ultremex 3 C18 reverse phase HPLC column (Phenomenex) with the same gradient and solvent system for the GSH-5-oxoETE adduct isolation and purification but with a flow rate of 50 l liters/min using air as the nebulizing gas and nitrogen as curtain gas. The electrospray ionization spray voltage was Ϫ3500 V, the orifice potential was maintained at Ϫ50 V, and the collisional offset potential 20 eV. The quantity of GSH-5-oxoETE adduct produced in various studies was calculated from the abundance of the ion transition m/z 624 3 306 for the adduct relative to the abundance for the same transition (m/z 630 3306) from D 6 -GSH-5-oxoETE (2 ng), which was added as internal standard at a constant amount to each sample before work-up and LC/MS analysis. Determination of the quantity of D 6 -GSH-5-oxoETE added as internal standard was carried out using a standard curve derived from the abundance of the LTC 4 molecular anion (m/z 624), measured by LC/MS from 1-50 ng injected on the column and injection of three different dilutions of synthetic D 6 -GSH-5-oxoETE to yield molecular anion (m/z 630) abundance within the linear region of the LTC 4 calibration curve. This method assumed equal electrospray ionization efficiencies between LTC 4 and the D 6 -GSH-5-oxoETE adduct. The quantity of LTC 4 produced in the peritoneal macrophage was calculated from the ratio of the abundance of the ion transition for LTC 4 , m/z 624 3 272 with the transition m/z 630 3 306 for the D 6 -GSH-5-oxoETE adduct internal standard calibrated against the standard curve.

FOG 7 in Vitro Synthesis-Cytosolic
GSTs were examined as catalysts for conjugating GSH to 5-oxoETE via a 1,4-Michael addition reaction to afford product. For these studies, 5-ox-oETE (10 M) and GSH (2 mM) were incubated for 10 min at 37°C in the presence of both commercial rat liver and human placental GSTs. The adducts were purified by solid phase extraction after the addition of D 6 -GSH-5-oxoETE adduct internal standard and analyzed by LC/MS/MS (Fig. 1). Unique ion transitions were monitored for the GSH adducts and the inter-nal standard such that production of adducts could be quantitated. In addition to investigating the catalytic properties of these cytosolic GSTs for 5-oxoETE conjugation, recombinant human LTC 4 synthase was also investigated for this activity. Multiple reaction monitoring revealed that all of the tested enzymes could catalyze the conjugation of GSH to 5-oxoETE; however, not all GSH addition products had the biological activity of FOG 7 , as measured by F-actin polymerization using flow cytometry.
Recombinant LTC 4 synthase (Fig. 1A) produced a single product, observed by monitoring the transition m/z 624 3 306 by LC/MS/MS. This product eluted from the HPLC column slightly after that of the D 6 -GSH-5-oxoETE adduct internal standard (m/z 630 3 306). This was the observed trend for FOG 7 biosynthesized by the peritoneal murine macrophage. The product catalyzed by LTC 4 synthase was observed to coelute with macrophage-derived FOG 7 (data not shown).
The products catalyzed by both human placental (Fig. 1B) and rat liver cytosolic GSTs (Fig. 1C) produced similar results. Both the human and rat GSTs catalyzed the addition of GSH to 5-oxoETE to afford compounds that co-eluted with the authentic FOG 7, suggesting that these GSH adducts might be FOG 7 . Furthermore, an additional minor adduct that eluted before that of D 6 -GSH-5-oxoETE internal standard was observed using rat liver GST (Fig. 1C). FOG 7 has been observed to profoundly activate chemotaxis and chemokinesis of both the human neutrophil and eosinophil by initiating polymerization of F-actin in these cells (16). The biological activity of the GSH-5-oxoETE adducts catalyzed by the cytosolic GSTs and the recombinant LTC 4 synthase was compared with FOG 7 for inducing F-actin polymerization in the human polymorphonuclear leukocyte. F-actin polymerization in neutrophils was induced by the LTC 4 synthase GSH-5-ox-oETE adduct ( Fig. 2A), but the human placental GST conjugate product did not initiate the polymerization of F-actin in the human neutrophil (Fig. 2B), indicating that this latter GSH-5-

FIG. 1. Mass spectrometric analysis by multiple reaction monitoring (LC/MS/MS) of the production of GSH-5-oxoETE adducts, catalyzed by recombinant human LTC 4 synthase (1.2 g) (A), human placental GST (5 units) (B), and rat liver GST (5 units) (C).
The abundance of each ion transition for m/z 624 3 306 and retention times were compared with FOG 7 by the use of the GSH-D 6 -5-oxoETE adduct internal standard by monitoring the transition m/z 630 3 306 (inset).

FIG. 2. Measurement of F-actin polymerization using flow cytometry and NBD-phallicidin for the GSH-5-oxoETE adducts catalyzed by recombinant human LTC 4 synthase (A), human placental GST (B), and rat liver GST (C and D).
oxoETE adduct most likely was not chemotactic for these cells, contrary to that of macrophage-derived FOG 7 . The two products catalyzed by the rat GST enzyme were separated by RP HPLC and analyzed for activity toward F-actin polymerization. Neither the major nor minor GSH-5-oxoETE adducts initiated polymerization of F-actin (Fig. 2, C and D). Flow cytometry revealed that the F-actin polymerization observed for FOG 7 (50 nM) afforded a shift in fluorescence almost identical to that of the GSH-5-oxoETE adduct (50 nM) catalyzed by LTC 4 synthase, again suggesting that this product might be FOG 7 and that LTC 4 synthase was perhaps the key enzyme in the biosynthesis of FOG 7 .
Although 5-oxoETE had been observed to increase intracellular calcium in both the neutrophil and eosinophil as assessed using Indo-1/AM fluorescence, FOG 7 did not induce calcium mobilization in these cells (16). The GSH-5-oxoETE adducts produced by LTC 4 synthase and the cytosolic GSTs also showed no activity in regard to the elevation of intracellular calcium in the neutrophil (data not shown).
Effect of MK-886 on GSH conjugation to 5-OxoETE-The inhibitor MK-886 was tested as an antagonist for the conjugation of GSH to 5-oxoETE, catalyzed by recombinant LTC 4 synthase and both human and rat cytosolic GSTs. As observed for the LTC 4 synthase conjugation of GSH to LTA 4 , MK-886 dosedependently inhibited the addition of GSH to 5-oxoETE to afford the chemotactic adduct with an apparent IC 50 of 7 M (Fig. 3). The GSH-5-oxoETE adducts catalyzed by human and rat cytosolic GSTs were unaffected by MK-886 at concentrations up to 100 M (Fig. 3).
When peritoneal murine macrophages were incubated for 15 min in the presence of 5-oxoETE (10 M), a single GSH-5-oxoETE adduct resulted (previously referred to as FOG 7 ), with a retention time on RP HPLC identical to that of the biologically active GSH-5-oxoETE adduct catalyzed by recombinant human LTC 4 synthase. To investigate whether or not LTC 4 synthase, expressed in the macrophage, catalyzed the formation of FOG 7 , the inhibitory effects of MK-886 were investigated in this cell. Elicited macrophages were incubated with 5-oxoETE (10 M) at 37°C for 15 min in the presence of various concentrations of MK-886. As shown in a typical experiment (Fig. 4), MK-886 inhibited the formation of FOG 7 in the intact macrophage in a dose-related manner. The mean IC 50 value for the mouse enzyme was 7.1 Ϯ 1 M (mean Ϯ S.E., n ϭ 3).
For comparison, elicited macrophages incubated with LTA 4 (10 M) at 37°C for 15 min afforded LTC 4 , as identified by multiple reaction monitoring of the LTC 4 -specific transitions m/z 624 3 272. The production of LTC 4 in the elicited macrophage was inhibited by MK-886 in a dose-related manner (Fig.  4). The mean IC 50 value for this enzyme was 5.1 Ϯ 1 M (mean Ϯ S.E., n ϭ 3).
Subcellular Localization of Enzymatic Activity toward FOG 7 Biosynthesis-The location of FOG 7 biosynthesis in the murine macrophage was determined in crude cell lysate, cytosol, and membrane fractions. Each subcellular fraction was incubated with 5-oxoETE (10 M) and GSH (2 mM) for 15 min at 37°C, and the extent of GSH conjugation to 5-oxoETE was analyzed by LC/MS/MS.
The total amount of protein recovered in both the cytosol and microsomal preparations was 83%. Both the cytosolic and microsomal fractions could catalyze the conjugation of GSH to 5-oxoETE, as evidenced by multiple reaction monitoring of aliquots from each of these samples. The microsomal fraction displayed a higher specific activity for the production of a GSH-5-oxoETE adduct than that observed from both the cell lysate and the isolated cytosolic fraction (Fig. 5). In addition to the higher specific activity, the GSH adduct from the microsomal fraction had a retention time identical to that of FOG 7 , as witnessed by co-elution of these two compounds during RP HPLC (data not shown). Also this compound was capable of initiating actin polymerization (Fig. 6A), and the production of this compound was inhibited by MK-886 (Fig. 5). The GSH adduct catalyzed by macrophage cytosolic GSTs eluted before the D 6 -GSH-5-oxoETE internal standard and as such, did not co-elute with authentic FOG 7 (data not shown) and did not initiate actin polymerization (Fig. 6B). In addition to this lack of activity, MK-886 had little or no effect on the production of this compound (Fig. 5), a result consistent with that for the cytosolic GSTs tested previously.
Identification of LTC 4 Synthase in the Microsomal Fraction-It appeared that the enzyme or enzymes responsible for the catalysis of GSH conjugation to 5-oxoETE to produce FOG 7 were membrane-bound proteins isolated in the microsomal fraction by fractional centrifugation. Whether LTC 4 synthase was the enzyme responsible for this process or whether additional microsomal GSTs were present in these cells and could catalyze this reaction was next examined.
Both the macrophage cell lysate, the cytosolic and the microsomal fractions (50 g of protein), were assayed for LTC 4 synthase activity. The crude cell lysate and microsomal fraction catalyzed the conjugation of GSH to LTA 4 -methyl ester to produce LTC 4 -methyl ester. The cell lysate not only catalyzed the formation of the methyl ester, but in addition to this, significant formation of the free acid of LTC 4 was detected by UV analysis (Fig. 7), a reaction most likely catalyzed by cytosolic esterases present in the macrophage. The cytosolic fraction did not catalyze formation of either LTC 4 or LTC 4 methyl ester.
Both the cytosolic and the microsomal fractions were assayed for LTC 4 synthase by Western blot analysis using peptide antibody raised against amino acids 65-78 in human LTC 4 synthase. This antibody had specificity toward the mouse LTC 4 synthase and revealed that this enzyme was present in the microsomal fraction and, as expected, was not detected in the cytosolic fraction (data not shown). The apparent K m and V max for conjugation of glutathione to LTA 4 and 5-oxoETE was determined in microsomes isolated from the murine macrophage containing LTC 4 synthase. Various LTA 4 and 5-oxoETE concentrations were incubated with microsomal fractions (0.15 mg of protein/ml) after adding 2 mM GSH for 5 min before sample work-up essentially as previously described in studies of the LTC 4 synthase kinetics in platelets (10). Using hyperbolic regression analysis, the apparent K m was found to be 1.3 Ϯ 0.34 for LTA 4 and 1.6 Ϯ0.22 for 5-ox-oETE, with V max values for LTA 4 and 5-oxoETE of 89 Ϯ 4.7 and 130 Ϯ 3.7, respectively. These results represent the mean Ϯ S.E. of three separate experiments.
To distinguish between microsomal GST-mediated catalysis and LTC 4 synthase catalysis in the membrane fraction, the conjugation of GSH to CDNB was examined. Recombinant LTC 4 synthase had no catalytic activity toward the conjugation of GSH with CDNB, as was the case for the microsomal fraction (Fig. 8). The cytosolic fraction, however, did possess the ability to conjugate GSH to CDNB, indicating that there are cytosolic GSTs present in the murine macrophage (Fig. 8).
Platelet FOG 7 Production-Human platelets are known to express LTC 4 synthase and efficiently convert LTA 4 into LTC 4 by transcellular biosynthesis (10,32,33). Incubation of 5-oxoETE (5 M) carried out with isolated human platelets (6 ϫ 10 8 /ml) for 15 min at 37°C resulted in a robust production of 58.3 ng of FOG 7 /10 9 platelets. This production of FOG 7 by human platelets was inhibited by MK-886 (data not shown), a result consistent with that observed in the murine macrophage.

DISCUSSION
The macrophage is an efficient cell in processing arachidonic acid with the formation of both cyclooxygenase (34) and lipoxygenase products (35,36). Although the macrophage is also capable of producing the chemotactic eicosanoid 5-oxoETE, the synthesis of this eicosanoid is not significantly enhanced by cell stimulus in contrast to other 5-lipoxygenase products (37). There is evidence to suggest that oxygenation of arachidonate at carbon-5 with formation of 5-hydroperoxyeicosatetraenoic acid, a precursor of 5-oxoETE, may be a preferred free radical pathway of arachidonate metabolism (38). The major metabolite of 5-oxoETE in the macrophage is the GSH 1,4-Michael addition product FOG 7 , which also has chemotactic properties (16). The macrophage expresses the unique microsomal protein, LTC 4 synthase (39), long thought to be specific only for LTA 4 conversion to LTC 4 . It is now apparent that LTC 4 synthase recognized 5-oxoETE as substrate and is likely the principal enzyme involved in the formation of FOG 7 in the macrophage.
Recombinant human LTC 4 synthase as well as human and rat cytosolic GSTs all catalyzed the conjugation of GSH to the electrophilic 5-oxoETE, as identified by multiple reaction monitoring on a tandem quadrupole mass spectrometer. LTC 4 synthase catalyzed the production of a single GSH-5-oxoETE adduct, which displayed identical chromatographic properties and biological activity to that of macrophage-produced FOG 7 . Both human and rat GSTs catalyzed the conjugation of GSH to 5-oxoETE to produce an adduct that was not separable from FOG 7 under the RP HPLC conditions used; however, these adducts did not initiate F-actin polymerization, suggesting that these compounds were in fact not FOG 7 . Rat GST also produced an additional minor adduct that was separable from FOG 7 and the other GSH-5-oxoETE adducts. This minor GSH-5-oxoETE adduct also lacked the ability to initiate F-actin polymerization in the human neutrophil. MS/MS and MS 3 analysis of these cytosolic GST-catalyzed products afforded spectra identical to that of FOG 7 , suggesting that these compounds are isomers of FOG 7 . Because the stereochemistry of FOG 7 and these additional GSH-5-oxoETE adducts are not known, determination of the absolute stereochemistry of macrophage-derived FOG 7 will likely require total organic synthesis. The chemical addition of GSH to 5-oxo-ETE would likely form equal quantities of two diastereoisomers at carbon-7 (absolute stereochemistry 7R and S), which may not substantially alter lipophilicity and, thus, RP-HPLC retention times.
The nuclear membrane-bound protein 5-lipoxygenase-activating-protein (FLAP) shows close homology to that of LTC 4 synthase, and as a consequence, the FLAP antagonist, MK-886, has been shown to inhibit LTC 4 synthase and formation of LTC 4 from the reactive precursor LTA 4 (39). FOG 7 production in the intact macrophage was completely inhibited by MK-886 at concentrations of 100 M, with an effective IC 50 consistent with the inhibition of LTC 4 formation from LTA 4 in this cell type. The production of FOG 7 catalyzed by the microsomal preparation was also inhibited by MK-886 (IC 50 , 7 M). Even though this inhibition was not as efficient (80% inhibition at 100 M) as that observed for the intact macrophage, this result was consistent with inhibition of recombinant LTC 4 synthasedependent conversion of LTA 4 to LTC 4 and may be due to the detergent used to solubilize these membrane-bound proteins. The apparent K m determined for LTA 4 conversion to LTC 4 and 5-oxoETE conversion to FOG 7 by macrophage microsomes were essentially identical. This low K m for LTA 4 and 5-oxoETE (1.3 Ϯ 0.34 and 1.6 Ϯ 0.22, respectively) reinforces the suggestion that murine macrophage LTC 4 synthase can be involved in both the formation of LTC 4 and FOG 7 , therefore having a potential dual role in the metabolism of 5-lipoxygenase arachidonate products.
Western blot analysis using a polyclonal antibody for LTC 4 synthase revealed detection of this enzyme in the microsomal fraction but not in the cytosolic fraction. The antibody used, however, had been raised against peptides in the FERV region, a region almost identical in the human and murine sequence but having substantial amino acid sequence similarity in FLAP and microsomal GST-II (31), and as a consequence, this LTC 4 synthase antibody might also detect both FLAP and microsomal GST-II.
The conversion of LTA 4 -methyl ester to LTC 4 -methyl ester as well as LTC 4 in the macrophage indicated that LTC 4 synthase was expressed in this cell as expected (39). The closely related microsomal GST-II has been previously shown to catalyze the conjugation of GSH to LTA 4 but also mediates conjugation of GSH to CDNB (20,40). The absence of microsomal GST-II in the macrophage was confirmed by a failure of this preparation to form the CDNB-GSH adduct. LTC 4 synthase is the only known GST, both cytosolic and microsomal, that does not catalyze the conjugation of GSH to CDNB (20,29,41). Because the microsomal fraction did not conjugate GSH to CDNB, LTC 4 synthase was therefore most likely responsible for the biosynthesis of FOG 7 in the macrophage. The absence of microsomal GSTs in the peritoneal macrophage is consistent with the observation that LTC 4 synthase expression is not prevalent in tissues that express microsomal GST-II and/or GST-III, with the testis being the exception (42). However, an interesting question remains to be answered as to whether or not microsomal GSTs expressed within another cell type could catalyze conversion of 5-oxoETE into FOG 7 .
Stimulation of the murine peritoneal macrophage by the calcium ionophore A23187 results in the abundant production of LTC 4 due to generation of LTA 4 but very little 5-oxoETE (37) and no FOG 7 . 4 Because there have been no reports of stimulated production of both 5-oxoETE and LTA 4 within the same cell type, it is difficult to assess whether these two substrates ever compete for LTC 4 synthase. However, the generation of 5-oxoETE by nonenzymatic pathways (38)  macrophage or other cell types (e.g. platelet) without activation of 5-lipoxygenase and concomitant formation of LTA 4 .
In summary, the biosynthesis of FOG 7 from 5-oxoETE and GSH in the elicited peritoneal macrophage was found to be catalyzed by a membrane-localized enzyme rather than cytosolic GSTs. Biochemical and pharmacological evidence suggests that nuclear membrane-bound LTC 4 synthase, long thought to be specific for the formation of LTC 4 from LTA 4 , is responsible for this conjugation reaction.