Production of leukotriene C4 in different human tissues is attributable to distinct membrane bound biosynthetic enzymes.

Microsomal glutathione S-transferase-II (GST-II) has recently been discovered and characterized as a member of the 5-lipoxygenase-activating protein (FLAP)/5(S)-hydroxy-6(R)-S-glutathionyl-7,9-trans-11, 14-cis-eicosatetraenoic acid (LTC4) synthase gene family, which also includes microsomal glutathione S-transferase-I (GST-I) as a distant member of this gene family. This new enzyme is unique as it is the only member of this family capable of efficiently conjugating reduced glutathione to both 5,6-oxido-7,9,11,14-eicosatetraenoic acid (LTA4) and 1-chloro-2,4-dinitrobenzene. Although microsomal GST-II has been demonstrated to display both general glutathione S-transferase (GST) and specific LTC4 synthase activities, its biological function remains unknown. In this study, we investigated the physiological location of microsomal GST-II as well as the relative importance of this enzyme versus LTC4 synthase for the production of LTC4 in various human tissues and cells that have been previously demonstrated to possess LTC4 synthase activity. As determined by Western blot, microsomal GST-II was predominantly expressed in human liver microsomes, human endothelial cell membranes, and sparsely detected in human lung membranes. In contrast, LTC4 synthase was prevalent in human lung membranes, human platelet homogenates, and human kidney tissue. Concomitant to the formation of LTC4, microsomal GST-II also produces a new metabolite of LTA4, a postulated LTC4 isomer. This isomer was used to distinguish between microsomal GST-II and LTC4 synthase activities involved in the biosynthesis of LTC4. Based on the relative production of LTC4 to the LTC4 isomer, microsomal GST-II was demonstrated to be the principal enzyme responsible for LTC4 production in human liver microsomes and human endothelial cells and played a minor role in the formation of LTC4 in human lung membranes. In comparison, LTC4 synthase was the main enzyme capable of catalyzing the conjugation of reduced glutathione to LTA4 in human lung membranes and human platelet homogenates. Therefore, microsomal GST-II appears to be an integral component in the detoxification of biological systems due to its marked presence in human liver, in accordance with its known GST activity. Microsomal GST-II, however, may also be pivotal for cysteinyl leukotriene formation in endothelial cells, and this could change our current understanding of the regulation of leukotriene biosynthesis in inflammatory disorders such as asthma.

Microsomal glutathione S-transferase (GST)-II 1 is the newest member of the FLAP/LTC 4 synthase gene family discovered to date. This novel enzyme displays 33% identity to FLAP, 44% identity to LTC 4 synthase, and limited sequence identity to microsomal GST-I at the amino acid level (1). These four proteins may also have similar structural configurations based on analogous hydrophobicity plots. Microsomal GST-II is a 16.6-kDa protein with a calculated pI of 10.4, possesses both LTC 4 synthase as well as conventional GST activities as shown through its ability to conjugate reduced glutathione (GSH) with both LTA 4 and 1-chloro-2,4-dinitrobenzene, and was consequently characterized as a member of the membrane bound GSTs (1). This new enzyme, accordingly, may have important roles in both leukotriene biosynthesis and in cellular detoxification by GST activity.
GSTs are a family of enzymes that catalyze the conjugation of reduced GSH to a variety of electrophilic substrates. The GSTs belong to a gene superfamily in which four different gene families encode the cytosolic GSTs (␣, , , and ) and two encode the microsomal forms of the enzyme (1,2). Biologically, these enzymes are responsible for detoxification of xenobiotics by catalyzing GSH conjugation to generated metabolites and for protection from endogenous hydroperoxides produced during oxidative stress via their GSH-dependent peroxidase activity (3).
LTC 4 synthase is a unique membrane bound enzyme that has been distinguished from all previously known cytosolic and microsomal GSTs (4, 5) through its narrow substrate specificity for LTA 4 . LTC 4 synthase and FLAP are important components of the leukotriene biosynthetic pathway. FLAP is required for cellular LTA 4 formation, possibly through the presentation of arachidonate to 5-lipoxygenase (6 -9). LTC 4 synthase is the first committed enzyme for the conversion of LTA 4 to the cysteinyl leukotrienes, LTC 4 , LTD 4 , and LTE 4 , which have significant roles in immediate hypersensitivity reactions (see reviews in Refs. 10 -15). * This work was supported in part by grants from The Wenner-Gren Foundation, The Heart and Lung Foundation, The Hellmuth Hertz Foundation, The Swedish Society of Medicine, Ulla and Gustaf af Uggla's Foundation, and The Swedish Foundation for International Cooperation in Research and higher education (to P-J. J.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  Prior to the molecular characterization of LTC 4 synthase and microsomal GST-II, a number of studies have described LTC 4 synthase activity in various cell types and tissues (15)(16)(17)(18)(19)(20)(21)(22)(23)(24). The present study addresses the relative importance of these two enzymes in the synthesis of LTC 4 in various human tissues.
Isolation of Human Peripheral Blood Platelets and Preparation of Homogenates-Human peripheral blood was collected from healthy volunteers who had not taken any medication for the previous 7 days. One-tenth the volume of 3.8% (w/v) trisodium citrate, in 0.9% (w/v) saline, was immediately added to the collected blood as anticoagulant, and the mixture was centrifuged at 200 ϫ g for 10 min at 22°C. The platelet-rich plasma was transferred to a tube containing 50% (v/v) Tris-buffered saline (TBS) containing 6.2 mM EDTA and 30% (v/v) of 3.8% (w/v) trisodium citrate. After gentle mixing by inversion, the diluted platelet-rich plasma was centrifuged at 650 ϫ g for 10 min at 22°C. The pellet was washed twice with TBS containing 6.2 mM EDTA, and cells were counted using a Coulter counter. The final concentration was adjusted to 1.0 ϫ 10 8 platelets/ml in cold lysis buffer containing 10 mM Hepes/KOH (pH 7.4), 2 mM EDTA, 2% (w/v) taurocholate, with freshly added dithiothreitol (5 mM), PMSF (1 mM, from a fresh 200 mM stock in EtOH), leupeptin (20 g/ml), pepstatin A (10 g/ml, from a fresh 1 mg/ml stock in EtOH), and aprotinin (10 g/ml). Lysates were shaken for 20 min at 4°C, sonicated for 5 s at 0°C, frozen in liquid nitrogen, and stored at Ϫ80°C. Bradford (28) protein assay (Bio-Rad) was performed on the homogenate.
Preparation of Human Endothelial Cell, Liver, and Lung Membranes-A frozen human venous endothelial cell pellet containing 1 ϫ 10 7 cells was thawed on ice and resuspended in PBS containing 2 mM EDTA and 2 mM PMSF. The cells were sonicated three times for 15 s each on ice and subjected to differential centrifugation at 1,000 ϫ g for 15 min and 100,000 ϫ g for 1 h, both at 4°C. The pellets were resuspended in 50 l of PBS containing 2 mM EDTA, frozen in liquid nitrogen, and stored at Ϫ80°C. Subcellular fractions of frozen human liver and lung tissues were prepared according to standard procedures (see Refs. 29 and 30, respectively).
Western Blot Analyses of Microsomal GST-I, Microsomal GST-II, LTC 4 Synthase, and FLAP Expression-Western blot analyses were performed similarly to those described previously (26). Briefly, SDScontaining sample buffer (31) was added to all samples. The samples were subsequently heated for 5 min at 95°C, electrophoresed through SDS-polyacrylamide gels (Novex), and electroblotted onto nitrocellulose. Ponceau S staining was used to visualize the efficiency of transfer. Membranes were then soaked for 1 h at 25°C in Tris-buffered saline containing 0.1% (v/v) Tween 20 (0.1% T-TBS) (20 mM Tris/HCl (pH 7.5), 0.5 M NaCl) containing 5% (w/v) Bio-Rad blotting grade blocker non-fat dry milk. Blots were washed twice for 5 min each with 0.1% T-TBS and subsequently treated for 1 h at 25°C with the indicated specific primary polyclonal antibody (dilution 1:500) in 0.05% T-TBS containing 5% dry milk. After washing the blots 3 times for 5 min each with 0.1% T-TBS, the membranes were incubated for 1 h at 25°C with a horseradishperoxidase-linked donkey anti-rabbit antibody (dilution 1:3,000) in 0.05% T-TBS containing 1% dry milk. The blots were washed 3 times for 5 min each with 0.3% T-TBS and then 3 times for 5 min each with 0.1% T-TBS and subsequently developed using enhanced chemiluminescence (Renaissance Western blot chemiluminescence reagent, DuPont NEN) according to the manufacturer instructions.

Measurement of LTC 4 Synthase and Microsomal GST-II Enzymatic Activities by Reverse-phase HPLC-LTC 4 synthase and microsomal
GST-II activities were assayed as described previously (1) by measuring the amount of LTC 4 and an isomer of LTC 4 produced in incubations at 25°C from various samples in 0.1 M potassium phosphate buffer (pH 7.4) containing reduced glutathione (5 mM) and 60 M LTA 4 (free acid) stabilized by the presence of 0.05% (w/v) bovine serum albumin in a final volume of 100 l. After 15 min, the reaction was terminated by the addition of an equivalent volume of acetonitrile/methanol/acetic acid at 50:50:1, and the precipitated proteins were removed by centrifugation at 16,000 ϫ g for 15 min at 4°C. The amount of LTC 4 and an isomer of LTC 4 synthesized were resolved by isocratic reverse-phase HPLC on a Waters Associates Novapak C 18 column (3.0 ϫ 150 mm, 4 m particle size) with the mobile phase (acetonitrile/methanol/water/acetic acid at 28:14:54:1 (pH 5.6)) at a flow rate of 1.2 ml/min. Quantification of the amount of products formed was based on the measurement of the peak absorbance at 280 nm from known amounts of injected LTC 4 . The LTC 4 and the LTC 4 isomer peaks were identified by comparison to retention time of synthetic LTC 4 and on-line analysis of the UV absorbance spectra of the eluted compounds using a Waters 991 diode-array spectrophotometer.
Preparation of Microsomal GST-II, LTC 4 Synthase, and FLAP Membranes from Baculovirus-infected Sf9 Cells-Microsomal GST-II, LTC 4 synthase, and FLAP proteins were obtained from baculovirus-infected Sf9 cell membranes as described previously (1,8). Briefly, Sf9 cells (Invitrogen) were infected with recombinant or wild-type virus and cultured for 72 h at 28°C in Grace's insect media (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (v/v), gentamycin (50 g/ml), and fungizone (2.5 g/ml). The cells were then harvested, washed, resuspended in PBS, and sonicated three times for 10 s each on ice. The sonicates were subsequently subjected to differential centrifugation at 500 ϫ g for 10 min and 100,000 ϫ g for 1 h, both at 4°C. The pellets were resuspended in PBS and stored at Ϫ80°C. Coomassie protein (Pierce) assay was performed according to the manufacturer instructions.

RESULTS
Microsomal GST-I, microsomal GST-II, and FLAP polyclonal antibodies are specific, whereas, the LTC 4 synthase antibody displays cross-reactivity-Polyclonal antibodies were used as tools to determine the relative contribution of LTC 4 synthase versus microsomal GST-II for the production of LTC 4 in biological systems. To determine the presence of either of these proteins, the specificity of these antibodies with respect to one another and to the other members of this family, FLAP and microsomal GST-I, were tested. The antibodies used to detect microsomal GST-II, LTC 4 synthase, and FLAP had been raised against peptides in the region displaying the highest identity to one another, the FERV region (Fig. 1). The microsomal GST-I polyclonal antibody was raised against the fulllength protein, which also contains a similar region of homology to the other peptides. As the different peptides contain homologous regions, the four antibodies were tested for possible cross-reactivity by Western blot analyses. Fig. 2 demonstrates that microsomal GST-I, microsomal GST-II, and FLAP antibodies were specific. However, the LTC 4 synthase antibody displayed some cross-reactivity by detecting microsomal GST-II and FLAP proteins. These results indicate that we have specific polyclonal antibodies for the detection of microsomal GST-I, microsomal GST-II, and FLAP and a polyclonal antibody that cannot distinguish between LTC 4 synthase and microsomal GST-II.
Expression of Microsomal GST-I, Microsomal GST-II, LTC 4 Synthase, and FLAP in Various Human Tissues-After establishing the specificity of the above polyclonal antibodies, they were subsequently used to search for their respective protein targets in various human tissues (Fig. 3). As expected, microsomal GST-I was markedly detected in liver tissue, as well as in adrenal gland, kidney, and to a lesser extent mammary gland tissues. Microsomal GST-II was highly expressed in liver ( Fig. 3) and in lower amounts in adrenal gland, kidney, pancreas, and thymus tissues (after prolonged exposure, data not shown). The LTC 4 synthase antibody detected ϳ17-kDa proteins intensely in kidney, adrenal gland, liver, and lymph and weakly in skeletal and stomach samples, whereas, the FLAP antibody displayed strong ϳ18-kDa bands in lymph and thymus tissues and weak bands in adrenal gland, liver, and skeletal tissues. These results have identified many important human tissues in which these proteins may be found.
Expression of LTC 4 Producing Enzymes in Human Liver, Lung, Endothelial Cells, and Platelets- Fig. 3 demonstrates that microsomal GST-II was primarily present in human liver tissue. To confirm and extend this finding, the subcellular localization of microsomal GST-II was investigated in human liver tissue by probing liver samples prepared by differential centrifugation with the microsomal GST-II and LTC 4 synthase polyclonal antibodies (Fig. 4). Both microsomal GST-II and LTC 4 synthase antibodies displayed an ϳ17-kDa band in the liver homogenate, the 10,000 ϫ g supernatant, and the , microsomal GST-II from baculovirus-infected Sf9 cell membranes (5 g and 0.5 g of protein), FLAP from baculovirus-infected Sf9 cell membranes (50 g and 5 g of protein), partially purified LTC 4 synthase from THP-1 cell extracts (0.013 g of LTC 4 synthase protein), and wild-type (wt) baculovirus-infected Sf9 cell membranes (50 g of protein) were resolved by SDS-polyacrylamide gel electrophoreses and electroblotted onto nitrocellulose. Western blot analyses were performed as described under "Experimental Procedures" using the polyclonal antibodies that were raised against the various peptides in Fig.  1 and were detected using enhanced chemiluminescence. The different blots: ␣-microsomal GST-I (␣MGST-I), -II (␣MGST-II), ␣LTC 4 synthase, and ␣FLAP were exposed to film for 1 min, 15 s, 15 s, and 1 s, respectively.
FIG. 3. Comparison of human tissue distributions of microsomal GST-I, microsomal GST-II, LTC 4 synthase, and FLAP. Equivalent amounts of various human tissue homogenates (75 g of protein), LTC 4 synthase, and microsomal GST-II from baculovirus-infected Sf9 cell membranes (50 and 5 g of protein, respectively) were electrophoresed through polyacrylamide gels and electroblotted onto nitrocellulose. Western blot analyses were performed using the polyclonal antibodies raised against the peptides in Fig. 1 and were detected using enhanced chemiluminescence as described under "Experimental Procedures." The various blots: ␣-microsomal GST-I (␣MGST-I), -II (␣MGST-II), ␣LTC 4 synthase, and ␣FLAP had exposure times of 15 s, 5 min, 5 min, and 1 min, respectively. 100,000 ϫ g pellet fractions. The intensity of detection increased with increasing fractionation of the membrane samples, and no bands were observed in the liver 100,000 ϫ g supernatant fraction. These results indicate that microsomal GST-II is present in human liver as a membrane bound protein.
Also, the presence of microsomal GST-II versus LTC 4 synthase was investigated in human lung, platelets, and endothelial cells. The LTC 4 synthase antibody recognized an ϳ17-kDa band in the lung 10,000 ϫ g pellet and a more intense band in the 100,000 ϫ g pellet, as well as an ϳ17-kDa band in platelet homogenates. The LTC 4 synthase antibody also displayed a band in endothelial cell membranes after a prolonged exposure time (20 min, data not shown). In contrast, microsomal GST-II was only slightly detected in lung membranes, not observed in platelet homogenates, and detected in endothelial cell membranes. These results indicate that LTC 4 synthase is the predominant enzyme present in human lung tissue and human platelets, whereas, microsomal GST-II appears to be the principal enzyme found in human endothelial cell membranes. 4 isomer have retention times of 9.0 and 7.6 min, respectively, and distinct maximum UV absorbencies of 281 and 283 nm, respectively (1) (Fig. 5). LTC 4 synthase, however, mainly catalyzes the synthesis of LTC 4 . Therefore, the ratio of the production of LTC 4 to the postulated LTC 4 isomer was used to distinguish between the presence of microsomal GST-II activity and LTC 4 synthase activity (Table I). To confirm, as well as discern, the presence of either microsomal GST-II or LTC 4 synthase as demonstrated by Western blot (Fig. 4), activity assays were performed. Lung 100,000 ϫ g pellet, platelet homogenates, and LTC 4 synthase derived from both THP-1 cell extracts and Sf9 cell membranes all displayed an LTC 4 :LTC 4 isomer ratio of Ͼ50. In contrast, liver 100,000 ϫ g pellet, endothelial 100,000 ϫ g pellet, and microsomal GST-II from Sf9 cell membranes all demonstrated ratios of Ͻ5. These ratios indicate that LTC 4 synthase activity is mainly in human lung and platelets, whereas, microsomal GST-II activity re-sides predominantly in human liver and endothelial cells.

LTC 4 Synthase and Microsomal GST-II Biosynthetic Activities in Human Tissues Can Be Distinguished by Product Profiles-Microsomal GST-II catalyzes the formation of LTC 4 and a possible LTC 4 isomer from LTA 4 and glutathione. LTC 4 and the putative LTC
Due to the similarities in structure between microsomal GST-II, LTC 4 synthase, and FLAP and their involvement in leukotriene biosynthesis, activity assays were also performed on microsomal GST-II in the presence of known inhibitors of leukotriene biosynthesis. MK-886 is a potent inhibitor of FLAP (nanomolar range), while both L-699 333 and MK-886 are capable of inhibiting LTC 4 synthase activity in the micromolar range. L-699 333 and MK-886, were both found to inhibit LTC 4 formation by microsomal GST-II expressed in Sf9 cell membranes with IC 50 values in the 10 M range (data not shown). The inhibition of microsomal GST-II activity by these leukotriene biosynthesis inhibitors suggests a similar active site for these three proteins. DISCUSSION Recently, a novel membrane bound protein, microsomal GST-II, was discovered that displayed homology at the amino acid level to FLAP, LTC 4 synthase, and to a lesser extent microsomal GST-I (1). These proteins all display similar hydrophobicity patterns and consequently may have similar structures. Due to the above similarities, these four proteins appear to be members of a gene family that encodes membrane bound proteins important for either leukotriene production or cellular detoxification by GSH conjugation. Microsomal GST-II is a unique member of this family due to its ability to efficiently conjugate reduced GSH to both LTA 4 and to 1-chloro-2,4-dinitrobenzene (1). Hence, microsomal GST-II may be more important than LTC 4 synthase as a catalyst for the formation of LTC 4 in certain tissues. To try to understand the biological function (importance) of this unique enzyme, its location in human cells and tissues compared with other family members was determined by Western blot analyses. In addition, a new method was developed for distinguishing between previously determined LTC 4 synthase activity and what may actually be microsomal GST-II activity.
Microsomal GST-I, microsomal GST-II, and FLAP polyclonal antibodies were determined to be specific for the recognition of their respective proteins even though the peptides used to raise FIG. 4. Human liver and endothelial cell membranes principally express microsomal GST-II, while, human lung and platelet homogenates primarily express LTC 4 synthase. Differential centrifugation fractions from human liver (75 g of protein), human lung membrane fractions (75 g of protein), human endothelial cell membranes (10 l of 1 ϫ 10 7 cell equivalents), human platelet homogenates (75 g of protein), microsomal GST-II from baculovirus-infected Sf9 cell membranes (5 g of protein), and partially purified LTC 4 synthase from THP-1 cell extracts (7.35 g of protein) were electrophoresed through polyacrylamide gels, electroblotted onto nitrocellulose, and immunoblotted using the polyclonal antibodies to microsomal GST-II or LTC 4 synthase and detected using enhanced chemiluminescence as described under "Experimental Procedures." The liver, lung, and control blots were exposed to film for 15 s, whereas, the endothelial and platelet blots had an exposure time of 1 min. ␣MGST-II, ␣-microsomal GST-II.  Table I due to the presence of either microsomal GST-II activity or LTC 4 synthase activity. The scale of the different chromatograms were in all cases normalized to the LTC 4 peak. The LTC 4 peak displayed a retention time of 9.0 min and a maximum UV absorbance at 281 nm, while the postulated LTC 4 isomer peak had a retention time of 7.6 min and a maximum UV absorbance at 283 nm. these antibodies contained regions of homology. Only the LTC 4 synthase antibody displayed cross-reactivity by recognizing FLAP, microsomal GST-II, and LTC 4 synthase. This nonspecific association of the LTC 4 synthase antibody to FLAP may be explained by a combination of high antibody concentration (1:500 dilution) and high expression of FLAP from baculovirusinfected Sf9 cells resulting in its high abundance on the blot. The recognition of FLAP by the LTC 4 synthase antibody could be discerned from LTC 4 synthase and microsomal GST-II due to the higher molecular weight of FLAP and was therefore not a problem in determining the presence of FLAP in various tissues. This antibody, however, could not discriminate between LTC 4 synthase and microsomal GST-II, thus, the presence of LTC 4 synthase was inferred based on the absence of specific microsomal GST-II recognition by the microsomal GST-II polyclonal antibody in identical tissues.
Interestingly, microsomal GST-II expressed from Sf9 cells was detected as a doublet, which may indicate the phosphorylation of this enzyme or a downstream degradation product. Phosphorylation of microsomal GST-II could be a possible regulatory mechanism for this enzyme as it has also been postulated for LTC 4 synthase (27). LTC 4 synthase has two potential protein kinase C phosphorylation sites which are not, however, present in microsomal GST-II. Microsomal GST-II from Sf9 cells was also inhibited by leukotriene biosynthesis inhibitors with a similar IC 50 value as that for LTC 4 synthase (32). As these inhibitors have been shown to inhibit FLAP and LTC 4 synthase, these proteins may have a similar active site. FLAP amino acids in the region Ser 41 to Val 61 have been shown to be critical for binding of leukotriene biosynthesis inhibitors (33). This region is the most homologous region between the members of this family including microsomal GST-I and is, therefore, postulated to be the lipid binding site of these proteins.
As expected, human microsomal GST-I is mainly expressed in human liver tissue in accordance with previous findings of location and its function as a phase II detoxifying enzyme (3). The significant detection of microsomal GST-I in the adrenal gland and kidney substantiates previous findings of microsomal GST activity, protein expression, or mRNA expression and probably serves to protect these extrahepatic tissues from endogenous or exogenous toxicants (34,35). Microsomal GST-II is also primarily expressed in human liver membranes and to a lesser extent in endothelial cells and minimally in lung membranes, displaying a very narrow tissue distribution. The marked detection of microsomal GST-II in the liver along with its glutathione S-transferase activity suggests that it may have a similar biological function to microsomal GST-I. Microsomal GST-II also appears to be post-transcriptionally regulated since the protein was not significantly detected in many of the tissues that exhibit mRNA expression (1). However, this lack of protein detection may also be a matter of sensitivity in the Western blot analyses. FLAP and LTC 4 synthase were widely distributed with expression in overlapping tissues as might be anticipated for two proteins involved in the biosynthesis of cysteinyl leukotrienes. However, there were also tissues (kidney, stomach, and thymus) that only expressed either FLAP or LTC 4 synthase. Consequently, these tissues demonstrate either the requirement of transcellular metabolism for leukotriene biosynthesis or in the case of those tissues that only express FLAP, may solely produce LTB 4 . These observations may also indicate the recognition of other putative members of this gene family by the cross-reactive LTC 4 synthase antibody. Contamination of tissue preparations by infiltrating peripheral blood cells could also explain any discrepancies in tissue distributions. FLAP was distinctly found in human lymph and thymus tissues, confirming previous findings of FLAP in B-and T-lymphocytes (36), whereas, LTC 4 synthase is significantly detected in human kidney, lung, platelets, and to a lesser extent in skeletal tissue, identical to earlier observations of mouse LTC 4 synthase mRNA expression (37) and human LTC 4 synthase protein detection (38). The LTC 4 synthase antibody appears to detect a 17-kDa band in the liver with an equivalent intensity to that detected by the microsomal GST-II antibody. These band intensities are most likely saturated and have reached a plateau due to overexposure of the blots. Therefore, the band intensities displayed in the liver by different antibodies cannot be quantitatively compared.
Microsomal GST-II possesses both generalized GST and LTC 4 synthase activities thereby contributing to the formation of LTC 4 in biological systems. Microsomal GST-II concomitantly produces LTC 4 and a new LTA 4 metabolite, a possible isomer of LTC 4 , that can be separated from LTC 4 by reversephase HPLC. LTC 4 synthase stereoselectively produces LTC 4 , thus, the production ratio of LTC 4 to the LTC 4 isomer clearly TABLE I Microsomal GST-II activity is predominant in human liver membranes, whereas, LTC 4 synthase activity is primarily in human lung and platelets Differential centrifugation fractions from human liver (0.4 mg of protein), human endothelial 100,000 ϫ g membrane fractions (0.4 mg of protein), human lung 100,000 ϫ g membrane fractions (0.4 mg of protein), human platelet homogenates (0.2 mg of protein), partially purified LTC 4 synthase from THP-1 cell extracts (2.45 g of protein), microsomal GST-II from baculovirus-infected Sf9 cell membranes (0.05 mg of protein), LTC 4 synthase from baculovirus-infected Sf9 cell membranes (0.05 mg of protein), and wild-type baculovirus-infected Sf9 cell membranes (0.05 mg of protein) were prepared as described under "Experimental Procedures." Microsomal GST-II or LTC 4 synthase activity was assayed in the above samples by measuring the amount of LTC 4 and a postulated LTC 4 isomer produced in 15 min-incubations at 25°C in 0.1 M KPi, pH 7.4, containing reduced glutathione (5 mM) and 60 M LTA 4 (free acid) stabilized by the presence of 0.05% (w/v) bovine serum albumin, as resolved by isocratic reverse-phase HPLC. The relative production of LTC 4 to the postulated LTC 4 isomer was used to define microsomal GST-II activity versus LTC 4 synthase activity. The liver 100,000 ϫ g pellet samples represent 6 different human livers (n ϭ 6), including one sample representing the mean of a triplicate. The endothelial 100,000 ϫ g membrane fraction is a representative experiment of n ϭ 2. All other samples were performed in triplicate Ϯ S.E. differentiates microsomal GST-II activity versus LTC 4 synthase activity. Identification of microsomal GST-II activity corresponds directly to the detection of microsomal GST-II in Western blot analyses. There is significant microsomal GST-II activity in liver 100,000 ϫ g membrane samples. Microsomal GST-II is accordingly the predominant membrane bound enzyme responsible for LTC 4 formation in human liver and may subsequently be a fundamental enzyme responsible for detoxification of lipid epoxides. Based on the above data, however, the possibility that LTC 4 synthase still exists in the liver in an inactive form cannot be excluded. LTC 4 was selectively synthesized by lung membranes with a minor production of the LTC 4 isomer, demonstrating LTC 4 synthase as the favored enzyme for LTC 4 production in human lung tissue with a minimal contribution from microsomal GST-II. In the lung, microsomal GST-II may play a minor part in LTC 4 production, however, it may have another function such as protection from oxidative stress and inhaled xenobiotics. The limited formation of LTC 4 from liver cytosol most likely represents the ability of cytosolic GSTs (mostly family GSTs) to conjugate reduced GSH to LTA 4 (39).
Platelet homogenates produced significant amounts of LTC 4 in the absence of notable quantities of the LTC 4 isomer, which together with the Western blot data indicate that the LTC 4 synthase enzyme predominates for LTC 4 production in platelets. In contrast, microsomal GST-II was detected in endothelial cell membranes by Western blot and activity assays and may be the enzyme responsible for converting exogenous LTA 4 to LTC 4 during transcellular metabolism in these cells. The presence of microsomal GST-II as the main enzyme for LTC 4 production in endothelial cells could explain the findings of Habib and Maclouf (40). In these studies, endothelial cells and platelets were both demonstrated to produce LTC 4 from exogenously added LTA 4 , however, platelets were more efficient at this conversion. Endothelial cells displayed a higher apparent K m for LTA 4 in comparison to platelets. Similarly, microsomal GST-II has a higher K m for LTA 4 than LTC 4 synthase and is less stereoselective for the production of LTC 4 (1). Analogous to our results, the HPLC chromatogram demonstrating LTC 4 production by endothelial cells (40) also contains an extra peak where the LTC 4 isomer would be predicted to appear. Such a peak is absent or less pronounced in the chromatogram from platelets. These observations may support the present findings, that LTC 4 production is catalyzed by microsomal GST-II in endothelial cells and by LTC 4 synthase in platelets.
Overall, microsomal GST-II is an integral membrane bound protein in human liver tissue that is potentially involved in detoxification of biological systems. Importantly, we have developed a method to distinguish microsomal GST-II activity from LTC 4 synthase activity. This has led to the discovery that microsomal GST-II is predominantly responsible for LTC 4 production in liver membranes and may be a key enzyme in the transcellular metabolism of LTA 4 into cysteinyl leukotrienes in certain cells that lack all of the enzymes required for de novo leukotriene biosynthesis.