INTRODUCTION

5-Lipoxygenase-activating protein (FLAP)¹ and leukotriene (LT) C₄ synthase are both proteins involved in the biosynthesis of leukotrienes. Leukotrienes are biologically active compounds that function as mediators of various inflammatory processes, such as leukocyte chemotaxis, increased vascular permeability, smooth muscle constriction, and increased mucus secretion (1, 2). Cellular leukotriene biosynthesis is initiated by a rise in intracellular calcium, which induces the release of arachidonic acid from different phospholipids by various classes of phospholipases (3, 4, 5, 6). Calcium also initiates the translocation of arachidonate 5-lipoxygenase to the nuclear membrane (7, 8). Subsequently, 5-lipoxygenase catalyzes a two-step reaction from arachidonic acid to the unstable epoxide LTA₄ via 5-hydroperoxyeicosatetraenoic acid (9). This reaction in intact cells requires the presence of FLAP (10) an 18-kDa protein localized to the nuclear membrane (8, 11). FLAP has been demonstrated to bind arachidonic acid and to increase the efficiency of 5-lipoxygenase conversion of 5-hydroperoxyeicosatetraenoic acid to LTA₄ (12, 13); however, no enzymatic function has been described for FLAP. The epoxide LTA₄ can be hydrolyzed to LTB₄ by the cytosolic protein LTA₄ hydrolase or conjugated with reduced glutathione by LTC₄ synthase to LTC₄ (1). LTC₄ synthase is a microsomal 16.6-kDa polypeptide that is enzymatically active as a homodimer (14) and was recently cloned independently by two groups (15, 16).

5-Lipoxygenase and FLAP expression is restricted to various myeloid cells, B lymphocytes, and pancreatic acinar cells (1, 17, 18). LTC₄ synthase activity has been described in eosinophils, basophils, mast cells, and certain phagocytic mononuclear cells (19, 20, 21, 22). Also, human endothelial cells, vascular smooth muscle cells, and platelets all express LTC₄ synthase activity without concomitant expression of 5-lipoxygenase. The formation of LTC₄ in these cells is therefore dependent on the transcellular metabolism of LTA₄, e.g. by interaction with activated neutrophils (23, 24, 25, 26, 27, 28). LTC₄ synthase activity has also been described in certain leukemic cell lines such as KG-1 cells, THP-1 cells, U-937 cells, and HL-60 cells (14, 29, 30, 31). Furthermore, an increased LTC₄ formation has been reported in leukocytes from patients with chronic myelogenous leukemia (32) and in experimental glomerulonephritis (33). Although various cytosolic glutathione S-transferases (GSTs) may conjugate LTA₄ with glutathione to form LTC₄ (34, 35, 36), LTC₄ synthase has been defined as a microsomal protein distinct from human cytosolic and microsomal GSTs (37, 38, 39). Successful attempts to purify LTC₄ synthase in the KG-1 myeloid cell line as well as dimethyl sulfoxide-differentiated U937 cells have confirmed LTC₄ synthase as being a distinct membrane protein with no activity toward 1-chloro-2,4-dinitrobenzene (substrate for the , µ, , and microsomal classes of GSTs) or p-nitrobenzylchloride (substrate for the  class of GSTs) as well as lack of recognition by specific antisera raised against , µ, , and microsomal GSTs (29, 30). Cytosolic GSTs are active as homo- or heterodimers of subunits of ~25 kDa, whereas microsomal GST is active as a trimer with a subunit size of 17 kDa (40, 41). Both cytosolic and microsomal GSTs are heavily expressed in the liver but are also found in various tissues such as kidney, lung, skeletal muscle, intestine, adrenals, heart, pancreas, and testes (40, 41). The 1-chloro-2,4-dinitrobenzene conjugation activity of both rat and human microsomal GSTs is activated by N-ethylmaleimide (42, 43, 44). The cDNAs of both the rat and the human microsomal GST contain an open reading frame encoding a 154-amino acid polypeptide, and the two proteins show 85% amino acid identity (45). These enzymes will be referred to as human or rat microsomal glutathione S-transferase I (microsomal GST-I). The microsomal GST-I from both species has a wide specificity for lipophilic and electrophilic substrates; however, LTA₄ is a poor substrate for microsomal GST-I (39, 44, 46, 47), which therefore should not contribute to the LTC₄ synthase activity reported in various tissues and cells. The biological functions of GSTs are attributed to detoxification of xenobiotics and metabolism of drugs as well as protection from oxidative stress caused by lipid peroxidation (40, 41, 48). In this report we describe a novel protein with characteristics in common with FLAP, LTC₄ synthase, and microsomal GST-I.

MATERIALS AND METHODS

CellsSpodoptera frugiperda Sf9 cells were obtained from Invitrogen and cultured in Grace's insect media supplemented with fetal bovine serum (10%), gentamycin (50 µg/ml), and fungizone (2.5 µg/ml). The cells were cultured at 28 °C, and the stock cell concentration was maintained between 0.5 and 3 × 10⁶ cells/ml. Isolation of granulocytes and mononuclear cells from human blood was performed essentially as described previously (49). Briefly, 50 ml of whole blood, obtained from a healthy donor was mixed with dextran (0.5% w/v, final concentration). After sedimentation of the erythrocytes for 30 min, the resulting plasma was centrifuged at 200 × g for 15 min. The cell pellet was resuspended in phosphate-buffered saline, pH 7.4 (Dulbecco's formula) and washed twice at 200 × g for 10 min. Erythrocytes were removed by hypotonic lysis with distilled water. Subsequently, the leukocytes were applied on a discontinuous density gradient (Ficoll-Isopaque) and centrifuged at 600 × g for 30 min (49). Mononuclear cells and granulocytes were collected and washed once. Viability was better than 98% as determined by trypan blue exclusion. The cell concentration was adjusted to 1.5 × 10⁷/ml.

A TBLASTN search of the GenBank^TM data base using the FLAP peptide sequence revealed similarity with the sequence deposited by the WashU-Merck EST project with an accession number of H59143[GenBank]. The sequence of the novel cDNA (termed microsomal GST-II) was confirmed on both strands according to the Sanger dideoxy chain termination method (50), using the PRISM^TM ready reaction dyedeoxy^TM terminator cycle sequencing kit and an ABI model 373 DNA sequencer. Oligonucleotides for sequencing and RT-PCR were obtained from Research Genetics (Huntsville, AL). The insert cDNA sequence was released from the pT7T3D vector by an EcoRI/HindIII double digest, end-filled with Klenow, and blunt end-ligated into the StuI-cut multiple cloning site of the pFastBac vector (Life Technologies, Inc.). Virus was constructed according to the Bac-to-Bac Baculovirus expression systems, as described by the manufacturer's instructions (Life Technologies, Inc.). Also, a cDNA for LTC₄ synthase was kindly provided by K. Scoggan.

Hybridization screening of a P1 artificial chromosome (PAC) library followed by fluorescence in situ hybridization was performed by Bios Laboratories, Inc. The initial PAC library screening was performed using a labeled microsomal GST-II cDNA probe obtained by random priming. Four positive PAC clones were obtained and confirmed by Southern blotting of the HaeIII-digested DNA. The clone with the best yield (PAC clone 180H14) was used in the subsequent chromosomal localization. The PAC clone 180H14 was labeled with digoxigenin dUTP by nick translation. The labeled probe was combined with human DNA and salmon sperm DNA and hybridized to prometaphase chromosomes obtained from PHA-stimulated peripheral blood lymphocytes in a solution containing 50% formamide, 10% dextran sulfate, and 2 × SSC. A biotin-labeled probe specific for the centromere of chromosome 4 was cohybridized with clone 180H14. Hybridization signals were detected with antidigoxigenin antibodies conjugated with rhodamine or fluoresceinated avidin, followed by counterstaining with 4,6-diamidino-2-phenylindole.

Northern blot of human multiple tissue blots (Clontech) was performed according to the instructions of the manufacturer. The blots containing 2 µg of poly (A)⁺ RNA/lane were hybridized with the following radiolabeled oligonucleotide probes: microsomal GST-II, 5-CAG TCG GAA ACC GGT GAT CCG TTT TTT AGC-3; FLAP, 5-AAA TAT GTA GCC AGG GGT GCT CTG CGT TCT-3; LTC₄ synthase, 5-TCG CGT ACA GCG GTG CCA GCC TGA GCT GCG C-3. These probes were selected from areas with minimal DNA sequence identity between FLAP, LTC₄ synthase, and microsomal GST-II. For the labeling of the oligonucleotide probes we used [-³²P]ATP (DuPont) and polynucleotide kinase (T4 PNK Pharmacia Biotech Inc.). Prehybridization/hybridization was performed in the buffer described by the manufacturer (Clontech) at 47 °C. After hybridization and washing, the blots were exposed to x-ray film (Kodak Biomax^TM MR) at 70 °C. The exposure time for microsomal GST-II and FLAP was 36 and 72 h, respectively.

Total RNA was isolated from granulocytes and mononuclear cells using Trizol^TM Reagent (Life Technologies, Inc.) according to the instructions of the manufacturer. Subsequently, cDNA was prepared from 2 µg of total RNA in a 40-µl incubation volume using a first strand cDNA synthesis kit obtained from Boehringer Mannheim. PCR was carried out in 100 µl of incubation mixtures consisting of 2 µl of cDNA, 0.2 mM dNTPs, 0.5 µM each primer, and 2 units of Taq DNA polymerase (Boehringer Mannheim) in PCR buffer (10 mM Tris-HCl, 1.5 mM MgCl₂, 50 mM KCl, pH 8.3). The conditions of the reaction were 1) 94 °C for 4 min, 2) 94 °C for 45 s, 60 °C for 45 s, 72 °C for 45 s, 3) 72 °C for 10 min; 25 cycles were carried out for both -actin and microsomal GST-II. Primers for microsomal GST-II were 5-ATT CTC TCG GCC TGT CAG CAA AGT TAT-3 and 5-CAG TCG GAA ACC GGT GAT CCG TTT TTT AGC-3 for regions which were specific for microsomal GST-II. -actin primers were obtained from Clontech (human -actin 838-base pair control amplimer set). The expected size of the microsomal GST-II DNA fragment was 294 base pairs. Aliqouts of the PCR mixtures (10 µl) were analyzed by electrophoresis using 1% agarose gels containing 0.5 µg/ml ethidium bromide. The identity of the PCR product corresponding to 294 bp from both granulocytes and mononuclear cells was confirmed by subcloning into pBBIII vector followed by DNA sequencing as described earlier.

Sf9 cells were infected with microsomal GST-II, LTC₄ synthase, or mock virus at a density of 1.5 × 10⁶ cells/ml. 72 h postinfection, the cells were harvested, washed, and pelleted by centrifugation at 300 × g. The cell viability was estimated by trypan blue stain exclusion and was reduced to approximately 30% viabilty compared with noninfected Sf9 controls (98% viability). The pellet was resuspended in phosphate-buffered saline, pH 7.4 (Dulbecco's formula) and sonicated on ice for 3 × 10 s. After centrifugation (500 × g) for 10 min, the supernatant was centrifuged for 1 h at 100,000 × g. The pellet (microsomal fraction) was resuspended in phosphate-buffered saline. Protein concentration was 10-15 mg/ml as determined by the Coomassie protein assay according to the manufacturer's instructions (Pierce). In order to measure LTC₄ synthase activity, the protein concentration was adjusted to 1 mg/ml using potassium inorganic phosphate buffer (0.1 M, pH 7.4). Then 50 µl of protein was mixed with 50 µl of potassium inorganic phosphate buffer containing 10 mM reduced glutathione and 0.1% (w/v) bovine serum albumin. The reaction was started by the addition of LTA₄ (2 µl of 1.5 mM LTA₄ in EtOH). The reaction was terminated after 15 min by the addition of 100 µl of acetonitrile:methanol:acetic acid (50:50:1, v/v/v). Precipitated protein was removed by centrifugation at 14,000 × g for 10 min. Subsequently, 150 µl of the sample was analyzed by reverse-phase HPLC equipped with a Novapak C18 column (3.0 × 150 mm, 4-µm particle size) obtained from Waters. The mobile phase was acetonitrile:MeOH:H₂O:acetic acid at 29:19:52:1 (v/v/v/v, adjusted to pH 5.6 with 30% NaOH (w/v), and the flow rate was 1.2 ml/min. Qualitative analysis was performed by comparison with the retention time of synthetic LTC₄ and on-line analysis of UV spectra of eluted compounds using a Waters 991 diode-array spectrophotometer. Amounts were calculated based on the peak area at 280 nm from known amounts of injected LTC₄.

Glutathione S-transferase activity was measured spectrophotometrically by measuring formation of the conjugate of reduced GSH and 1-chloro-2,4-dinitrobenzene (CDNB) at 340 nm (51). Microsomes of either Sf9 cells or Sf9 cells expressing microsomal GST-II or LTC₄ synthase were diluted to 5 mg of protein/ml using sodium phosphate (0.11 M, pH 6.5). 5 µl of this sample (25 µg) was transferred into a 195-µl incubation mixture consisting of 1 mM GSH, 1 mM CDNB in sodium phosphate buffer (0.11 M, pH 6.5). The product formation was continuously measured at 340 nm for 5 min on a SpectraMac 250 (Fisher). In experiments where the effect of N-ethylmaleimide was investigated, a 50-µl aliquot of microsomes containing either microsomal GST-II or LTC₄ synthase (at 2.5 mg/ml) was treated with 1 mM N-ethylmaleimide for 3 min. Thereafter, 10 µl of this protein was added to the incubation mixture (190 µl) containing 3 mM glutathione and 1 mM CDNB in sodium phosphate buffer (0.11 M, pH 6.5).

RESULTS

A TBLASTN search of the GenBank^TM data base using the FLAP peptide sequence revealed that an expressed sequence tag (EST) clone with the accession number H59143[GenBank] displayed significant sequence identity with FLAP. The clone corresponding to this sequence, human clone 204168, was obtained from the IMAGE Consortium, and sequencing confirmed its identity. The cDNA insert of clone 204168 contains an open reading frame encoding a polypeptide of 147 amino acids with a predicted molecular mass of 16.6 kDa, which we have termed microsomal GST-II (Fig. 1). The amino acid identity was 44% to LTC₄ synthase and 33% to FLAP (Fig. 2). The completely conserved amino acids between all three proteins are displayed as a consensus sequence (Fig. 2). Also, amino acids 70-86 of human microsomal GST-I (ERVrrAhlNdleniipF) displayed a limited sequence identity to FLAP (amino acids 51-67), LTC₄ synthase, and microsomal GST-II (amino acids 47-63) as shown in Fig. 2. Six of 17 amino acids in this region of human microsomal GST-I (amino acids 70-86) were identical to the area of consensus (Fig. 2, amino acids 51-67), but 8 of 17 were identical to the corresponding amino acid sequence in both LTC₄ synthase and microsomal GST-II (amino acids 47-63). The overall sequence identity of human microsomal GST-I to microsomal GST-II was only 11%, whereas the similarity was 31%.

Hydropathy plot analysis of FLAP and LTC₄ synthase has demonstrated that these proteins contain three hydrophobic regions, which have been suggested to represent transmembrane domains (10, 15, 16). Fig. 3 shows the hydropathy plot analysis of LTC₄ synthase, microsomal GST-II, FLAP, and human microsomal GST-I. The first three proteins display a very similar pattern, but the C terminus is neutral for LTC₄ synthase, hydrophilic for microsomal GST-II, and hydrophobic for FLAP. Human microsomal GST-I displays a similar pattern; however, the first hydrophobic and hydrophilic stretches are longer and more prominent, whereas the second hydrophobic domain is divided by a short neutral part. The third hydrophobic domain is also shorter, and its C-terminal sequence is neutral.

The calculated isoelectrical point of microsomal GST-II is 10.4 as compared with 8.7 for FLAP, 11.1 for LTC₄ synthase, and 10.2 for human microsomal GST-I.

A PAC DNA library was first screened using the cDNA of microsomal GST-II in order to obtain genomic microsomal GST-II for chromosomal localization by fluorescence in situ hybridization. Based on cohybridization with another probe specific for the centromere of chromosome 4 and fractional length measurements of 10 chromosomes, it was concluded that 180H14 is located at a position that is 74% of the distance between the p and q telomeres on chromosome arm 4q, an area that corresponds to bands 4q28-31. In comparison, the FLAP gene has been localized to chromosome 13q12 (52), the LTC₄ synthase gene on chromosome 5q35 (53), and the gene for human microsomal GST-I on chromosome 12 (54).

The expression of microsomal GST-II, LTC₄ synthase, and FLAP mRNA was investigated in various tissues by Northern blot analysis using specific oligonucleotides as probes (see ``Materials and Methods''). Using the microsomal GST-II probe, a 0.6-kb mRNA was detected in human tissues including heart, liver, skeletal muscles, pancreas, spleen, prostate, testis, ovary, small intestine, and colon (Fig. 4, A and B). These same human tissue Northern blots were probed with a FLAP-specific probe, and a 0.7-kilobase mRNA was detected in lung, spleen, thymus, and peripheral blood leukocytes (Fig. 4, A and B). Fetal liver and fetal kidney both contained microsomal GST-II mRNA, which is in line with the fact that microsomal GST-II was cloned from a fetal liver spleen cDNA library (Fig. 4C). Microsomal GST-II was also heavily expressed in the adrenals, especially the cortex (Fig. 4E). We also investigated the expression of FLAP and microsomal GST-II in different organs of the immune system and cancer cell lines (Fig. 4, F and G). Microsomal GST-II was expressed in the human promyelocytic leukemia HL60 and HeLa cell S3, but not in the T or B cell lines Molt4 and Raji, respectively. Microsomal GST-II was found also in chronic myelogenous leukemia cell line K-562, adenocarcinoma SW480, and melanoma G361. FLAP was expressed in lymph node tissue, thymus, appendix, peripheral blood leukocytes, and bone marrow. FLAP was also detected in HL60 cells and Raji and at lower amounts in Molt4 and the adenocarcinoma SW480. The expression of microsomal GST-II and FLAP in different brain tissues is shown in Fig. 4D. The only significant RNA detected in brain was FLAP in the medulla and spinal cord. We also investigated the expression of LTC₄ synthase using several human tissue blots (Fig. 4, A, D, F, and G); however, utilizing the hybridization described, no detection of LTC₄ synthase mRNA was obtained after 3 days of exposure (data not shown). The expression of microsomal GST-II in human peripheral blood leukocytes was analyzed by RT-PCR. Fig. 5 shows that total RNA isolated from mononuclear or polymorphonuclear leukocytes contains microsomal GST-II. The identity of the PCR fragment was confirmed by subcloning and full-length sequencing (data not shown).

The cDNA insert of clone 204168 microsomal GST-II and the cDNA for LTC₄ synthase were subcloned into the pFastBac plasmid followed by creation and isolation of bacmid DNA. Also, a bacmid mock DNA was created. 72 h after transfection of Sf9 cells with bacmid DNA, mRNA from the Sf9 cells was isolated and analyzed by Northern blot and RT-PCR. Microsomal GST-II mRNA was specifically detected using both methods. The corresponding transfection viral stocks were amplified once and subsequently used for infection of Sf9 cells. Following infection, the effects of the viruses on cell viability and growth were observed and compared with noninfected cells.

Three 250-ml Sf9 insect cell cultures were infected with recombinant baculoviruses expressing LTC₄ synthase, microsomal GST-II, and mock, respectively. A 10-ml aliquot of cells was removed every 24 h, and cells were washed once in PBS, pelleted, and frozen at 80 °C. After 5 days, the cells were thawed and sonicated (crude homogenate). An aliquot was removed to prepare 100,000 × g pellet and supernatant. LTC₄ activity was assayed in 100-µl incubation mixtures containing 0.5 mg/ml sample protein in 0.1 M potassium phosphate (pH = 7.4), 0.05% albumin, 5 mM glutathione, and 30 µM LTA₄. After 15 min the reaction was terminated by adding 100 µl of stop solution (AcN:MeOH:HAc, 50:50:1), and 150 µl was subjected to RP-HPLC analysis. Table I shows the formation of LTC₄ in Sf9 cells after infection. In both noninfected and mock-infected crude homogenates, only very small amounts of LTC₄ formation could be detected (<35 pmol/mg protein). However, cells infected with either LTC₄ synthase or microsomal GST-II both catalyzed the formation of LTC₄ (Table I). The enzymatic activity for LTC₄ synthase and microsomal GST-II was 3-5 times higher in the 100,000 × g pellet compared with the activity in the crude homogenates (Table I). The corresponding 100,000 × g cytosol fraction contained no significant activity (data not shown). The microsomal enzymatic activities were 1375 ± 405 pmol LTC₄/mg protein/15 min (mean ± S.D., n = 3) for cells infected with LTC₄ synthase and 3480 ± 528 pmol LTC₄/mg protein/15 min (mean ± S.D., n = 3) for cells infected with microsomal GST-II. The formation of LTC₄ was dependent on the presence of both LTA₄ and reduced glutathione. Also, the LTC₄ formation was abolished by boiling for 5 min prior to the assay. Fig. 6 shows LTC₄ formation as a function of time. Microsomal GST-II microsomes were incubated at both 0.5 mg/ml and 0.1 mg/ml. The formation of LTC₄ product increased rapidly over the first 3 min and reached a plateau at approximately 7 min after the start of the reaction. Microsomal GST-II at 0.1 mg/ml and LTC₄ synthase at 0.5 mg/ml led to similar time courses, and the formation of LTC₄ was considered linear up to 3 min for microsomal GST-II and up to 5 min for LTC₄ synthase (Fig. 6). In order to determine the apparent K_m for LTA₄ (at a constant GSH concentration of 5 mM), microsomal GST-II microsomes (0.1 mg/ml) and LTC₄ synthase microsomes (0.5 mg/ml) were incubated at various LTA₄ concentrations for 3 and 5 min respectively. Using hyperbolic regression analysis, the apparent K_m was 41 µM for microsomal GST-II and 7 µM for LTC₄ synthase (Fig. 7). Fig. 8A shows the RP-HPLC chromatograms of the products formed after incubation of LTC₄ synthase microsomes (0.5 mg/ml), microsomal GST-II microsomes (0.1 mg/ml) mock virus microsomes (0.5 mg/ml), and buffer alone with 30 µM LTA₄ and 5 mM glutathione for 15 min. Peak 2 coelutes with synthetic standard LTC₄ (40 pmol). However, microsomal GST-II also catalyzed the formation of another product (peak 1), eluting as a more polar compound on RP-HPLC. The UV absorbance spectra were compared, and Fig. 8B shows that the spectra corresponding to peak 2 in Fig. 8A all had UV maxima at 281 nm, whereas the spectra corresponding to peak 1 had UV maxima shifted + 2 nm. The formation of this product was also dependent on the presence of both LTA₄ and glutathione, abolished by boiling prior to the incubation, time-dependent, and saturable (data not shown).

Table I. Time course of microsomal GST-II and LTC4 synthase expression Sf9 cells Time postinfection Cell viability LTC4 Crude homogenate 100,000 × g pellet days pmol/mg protein Noninfected 0 99 21 <2 Noninfected 1 99 21 Mock-infected 1 99 29 <2 LTC4-synthase 1 98 404 1140 Microsomal GST-II 1 99 29 <2 Noninfected 2 99 34 Mock-infected 2 99 26 <2 LTC4-synthase 2 97 469 1205 Microsomal GST-II 2 75 494 811 Noninfected 3 98 6 Mock-infected 3 43 3 <2 LTC4-synthase 3 40 364 1206 Microsomal GST-II 3 38 708 3362

GST activity was measured spectrophotometrically by measuring formation of the conjugate of the reduced GSH and the CDNB at 340 nm. Microsomes from Sf9 cells expressing microsomal GST-II catalyzed the conjugation of CDNB and GSH at 68 ± 6 nmol/mg/min (mean ± S.D., n = 3). In comparison, the activity found in rat and human liver microsomes has been reported to be 94 and 76 nmol/mg/min, respectively (42). In purified preparation the specific activity for both purified rat and human microsomal GST-I has been reported to be 2 µmol/min/mg (43, 55). The corresponding activities in microsomes from Sf9 cells infected with LTC₄ synthase and mock viruses were 11 ± 6 and 10 ± 3 nmol/mg/min, respectively (mean ± S.D., n = 3). In buffer, the rate of nonenzymatic conjugation was 5 ± 2 nmol/mg/min (mean ± S.D., n = 3). Fig. 9 shows the time course of the conjugation of GSH and CDNB by microsomes from Sf9 cells infected with microsomal GST-II, LTC₄ synthase, and mock virus as well as buffer control. Also, the effect of N-ethylmaleimide was investigated. The activity in Sf9 cell microsomes containing microsomal GST-II was not affected by treatment of the protein with 1 mM N-ethylmaleimide.

DISCUSSION

A TBLASTN search of the GenBank^TM data base using the FLAP peptide sequence has revealed a new clone with significant sequence identity to FLAP and LTC₄ synthase and with limited sequence identity to microsomal GST. This novel protein retains LTA₄ and 1-chloro-2,4-dinitrobenzene-conjugating activity with reduced glutathione and was termed microsomal glutathione S-transferase II (Fig. 1).

Sequence comparison of microsomal GST-II, LTC₄ synthase, and FLAP shows that 33 of 147 amino acids are completely conserved in these proteins (Fig. 2). Also, microsomal GST-II and LTC₄ synthase (amino acids 47-63) and FLAP (amino acids 51-67) display a limited sequence identity to microsomal GST-I (amino acids 70-86). Furthermore, as shown in Fig. 3, these regions correspond to the carboxyl terminus of the first hydrophilic stretch in all four proteins. Interestingly, a series of deletion mutants of FLAP in this region (deletions 37-53, 52-58, and 59-61) have demonstrated that this part of FLAP is critical for binding of leukotriene biosynthesis inhibitors such as MK-886 and L-689,037 (56, 57). Also, the binding of arachidonic acid to FLAP has been shown to be competed for by MK-886 (12). Therefore, a proposed mechanism of action for these leukotriene biosynthesis inhibitors is to compete with arachidonic acid for its binding site on FLAP, FLAP acting as an arachidonic acid transfer protein for 5-lipoxygenase. In line with the fact that LTC₄ synthase and microsomal GST-II both bind LTA₄, whereas rat and human microsomal GST-I have been shown to catalyze the reduction of phospholipid hydroperoxides (41), this region is a candidate for being the binding site for the fatty acid backbone component of the various substrates/ligands.

Due to the hydropathy pattern of FLAP (three hydrophobic regions separated by two hydrophilic regions) it has been proposed that FLAP spans a membrane bilayer 3 times (10). Interestingly, all of these four membrane proteins of approximately the same length display a similar hydrophobicity pattern (Fig. 3.). The fact that human microsomal GST-1 displays this hydrophobicity pattern plus the limited sequence identity to the other three proteins indicate that they all are members of a family of membrane proteins with highly specialized functions.

Both LTC₄ synthase and microsomal GST-I conjugate glutathione with electrophilic substrates. LTC₄ synthase has a narrow substrate specificity (37, 38, 39) compared with microsomal GST-I (41). Also, purified LTC₄ synthase has 20,000 times higher specific activity compared with purified human microsomal GST-I to form LTC₄ from LTA₄ and glutathione (46). In Table I and Fig. 6, we show that Sf9 cells infected with recombinant baculovirus for microsomal GST-II became capable of catalyzing the formation of LTC₄ from LTA₄ and reduced glutathione. The specific activity was 3-5 times higher in the 100,000 × g pellet as compared with the total cellular extracts. The microsomal activity was about 5 times higher for microsomal GST-II than LTC₄ synthase. This may reflect a better expression efficiency for microsomal GST-II and/or a higher V_max for the formation of LTC₄. The activity obtained for LTC₄ synthase in this expression system was in the same range of activity as reported by others using baculovirus expression (15). The apparent K_m was determined for LTA₄, and the affinity of LTA₄ to microsomal GST-II was found to be lower compared with the affinity to LTC₄ synthase (Fig. 7). Also, microsomal GST-II produced another product that eluted as a more polar compound with a conjugated triene spectra that was shifted +2 nm compared with the UV spectra of synthetic LTC₄ (Fig. 8). Preliminary mass spectrometric analyses by liquid chromatography-mass spectrometry/mass spectrometry demonstrated that this more polar compound has a mass spectrum identical to that of LTC₄, suggesting that the compound is an isomer of LTC₄ (data not shown).

Microsomes from Sf9 cells infected with recombinant baculovirus for microsomal GST-II also catalyzed the conjugation of glutathione and 1-chloro-2,4 dinitrobenzene at a rate of 68 nmol/mg/min. This activity corresponds to 3-4% of the activity reported for purified rat or human microsomal GST-I (43, 55). The rate of conjugation of glutathione and 1-chloro-2,4 dinitrobenzene by LTC₄ synthase was similar to that detected in microsomal preparations of mock-infected cells. This agrees with data showing that purified LTC₄ synthase did not catalyze this reaction (30). In addition, the effect of N-ethylmaleimide was investigated on the conjugating activity in microsomes containing microsomal GST-II. In contrast to microsomal GST-I (43, 44), no effect by N-ethylmaleimide was observed.

Northern blot analysis was used for analysis of the tissue distribution of the microsomal GST-II. Microsomal GST-II was detected in many different tissues (Fig. 4 and 5). This wide tissue distribution is also an attribute of microsomal GST-I (45, 58). Interestingly, FLAP expression was restricted to organs associated with the immune defense, i.e. peripheral blood leukocytes, spleen, lymph nodes, bone marrow, and lung. In these organs microsomal GST-II was only expressed in spleen. However, using the more sensitive method of RT-PCR, microsomal GST-II mRNA expression was detected also in both peripheral mononuclear cells and granulocytes (Fig. 5). The tissue distribution of FLAP correlates rather well with cells and tissues expected to possess 5-lipoxygenase activity, i.e. tissues with a high content of leukocytes. Based on these data it is unlikely that microsomal GST-II would be a 5-lipoxygenase-activating protein. A previous report has shown that coinfection of FLAP and 5-lipoxygenase, as compared with infection of 5-lipoxygenase alone, in a baculovirus insect cell system results in a stimulation of leukotriene production when these cells were stimulated with arachidonic acid and the calcium ionophore A23187 (13). This experiment was performed with microsomal GST-II coinfected with 5-lipoxygenase, and no significant increase of leukotriene formation was detected (data not shown).

The corresponding Northern blot analysis of LTC₄ synthase did not detect any mRNA. This suggests that LTC₄ synthase mRNA expression is highly regulated and possibly restricted in blood to various nonabundant leukocytes, in line with the report showing that enriched human eosinophils but not peripheral polymorphonuclear leukocytes express LTC₄ synthase (16). Interestingly, the data in Fig. 4 show that microsomal GST-II is highly expressed in the human cell line K-562, a chronic myelogenous leukemia cell line. This evokes the question as to whether or not microsomal GST-II might be the enzyme responsible for the reported increase of LTC₄ formation in leukocytes isolated from patients with chronic myelogenous leukemia (32). Also, it will be interesting to investigate the relative influence of LTC₄ synthase versus microsomal GST-II on LTC₄ formation in cells apparently devoid of 5-lipoxygenase such as platelets, endothelial cells, and smooth muscle cells.

These results indicate that microsomal GST-II is a microsomal protein with both LTC₄ synthase activity and the capacity to conjugate CDNB with glutathione. Its catalysis of LTC₄ seems to be more nonspecific compared with the catalysis performed by LTC₄ synthase. Also, its wide tissue distribution resembles microsomal GST-I (45, 58). Consequently, at this stage microsomal GST-II must be categorized as a microsomal glutathione S-transferase; however, its biological function should be further investigated. An intriguing possibility is that FLGST-H may also possess prostaglandin E synthase activity, since a membrane-associated prostaglandin E synthase has been reported as a 17.5-kDa protein requiring glutathione as a cofactor (59, 60). Other possibilities are that this protein represents a general metabolic system for detoxifying fatty acid epoxides such as those derived through cytochrome P₄₅₀ pathways, or as is the case for microsomal GST-I, the protein may possess glutathione peroxidase activity (61).

In summary, we have identified a novel microsomal glutathione S-transferase with significant amino acid identity to FLAP and LTC₄ synthase. This enzyme can catalyze the conjugation of both LTA₄ and CDNB with glutathione and therefore represents a unique microsomal glutathione S-transferase. The hydrophobicity pattern of microsomal GST-II, FLAP, LTC₄ synthase, and microsomal GST-I as well as their sequence homologies suggest an evolutionary relationship within this gene family.