Introduction

Top Abstract Introduction Methods Results Discussion References

Stimulation of human basophils by either antigens that cross-link receptor-bound IgE or by non-IgE dependent agonists (such as fMLP) elicits the release of inflammatory mediators. LTC₄ is the major eicosanoid product released from human basophils as no prostaglandins or other leukotriene products have been detected to date (Warner et al., 1987; Warner et al., 1989). The release of AA from phospholipids and neutral lipids is required for the synthesis of eicosanoids and in many tissues the rate limiting step for the synthesis of eicosanoids is the hydrolysis of AA from other lipids. In recent years, there has been controversy both over the source of the AA and the enzymes responsible for generating the free AA used for eicosanoid synthesis. Previous studies in basophils demonstrated that free AA was accumulated during stimulation. These studies also demonstrated that there were problems using radiolabeled AA to measure these events and that measuring mass by GC/NICIMS was a more accurate method for following the production of free AA (MacGlashan and Hubbard, 1993). This technique was used to address specific questions about C5a-induced LTC4 and AA generation but these experiments raised some additional questions about the production of free AA during these reactions. In particular, the generation of free AA after stimulation with fMLP, in the presence or absence of IL-3, did not appear to change in accordance with the production of LTC4. This and the lack of detailed studies of free AA generation during these reactions prompted questions about the role of different sources of free AA in LTC4 generation as well as questions about the enzymes involved in its generation.

There are several enzymes capable of generating free AA from cellular lipids. There is interest in the role of two unique PLA₂ enzyme groups found associated with mast cells and several other cell types. One class comprises the group IV (or type IV) PLA₂ and this high molecular weight (85-kDa) cytosolic enzyme has a substrate preference for phospholipids with AA in the sn-2 position (Clark et al., 1991; Dennis, 1994; Marshall and Roshak, 1993; Mayer and Marshall, 1993; McCord et al., 1994). Group IV PLA₂ or cPLA₂ has been shown to be translocated from the cytosol to the plasma membrane in response to increases in cytosolic Ca⁺⁺ (Clark et al., 1991; Hirasawa et al., 1995) and the phosphorylation of cPLA₂ correlates with increases in the activity of the enzyme (Currie et al., 1994; Hirasawa et al., 1995; Lin et al., 1993).

sPLA₂ represents another class of PLA₂ enzymes that have a smaller molecular weight (14-kDa), can be cell associated and have been identified in rheumatoid synovial fluid, neutrophils and mast cells (Fonteh et al., 1994; Kramer et al., 1989; Marshall et al., 1994; Murakami et al., 1992). The release of sPLA₂ from stimulated human platelets, rat peritoneal mast cells, mouse bone marrow-derived mast cells, P388D macrophage-like cell line, and other cell types has been reported in several studies (Barbour and Dennis, 1993; Fonteh et al., 1994; Kramer et al., 1989; Murakami et al., 1992). It is believed that sPLA₂ can bind to the cell surface or is in close proximity to the surface possibly through association with a heparan sulfate binding site (Suga, et al., 1993). In addition, membrane receptors for sPLA2 have been described and a membrane receptor for one form of sPLA₂ has been cloned (Lambeau et al., 1994). More recently it has also become evident that there may be confusion over which sPLA2 is present in cells, type II or type V (Balboa et al., 1996; Chen et al., 1994a, 1994b). These are similar proteins that react equally with many of the reagents initially developed for studying type II sPLA₂.

There is evidence for the involvement of both type IV and type II/V PLA₂ enzymes in mast cell secretion. In antidinitrophenyl IgE-sensitized bone marrow mast cell cultures, exposure to antigen elicits the phosphorylation of cPLA₂, an increase in cPLA₂ activity and an increase in the amount of cPLA₂ is apparent after several hours (Nakatani et al., 1994). In the mast cell line RBL-2H3(ml) stimulation with antigen, Ca⁺⁺-ionophores, or carbachol induced the phosphorylation of cPLA₂ and increased AA release (Hirasawa et al., 1995). These studies suggest that cPLA₂ has the predominate role in mediating AA release for eicosanoid synthesis. Early evidence for the involvement of sPLA₂ in AA release for eicosanoid synthesis was presented by Murakami et al., (1991), showing that the addition of an exogenous 14-kD PLA₂ to stimulated rat peritoneal mast cells increased PGD₂ synthesis in a dose dependent manner. Later it was found that activated mast cells released a 14-kD PLA₂ into culture (Murakami et al., 1992). sPLA₂ has also been identified in the secretory granules of mast cells (Chock et al., 1994). In mouse bone marrow mast cells IgE-mediated stimulation also elicits the release of a sPLA₂ and the addition of exogenous PLA₂ increases eicosanoid synthesis (Fonteh et al., 1994).

Our investigation further examines the characteristics of AA generation in stimulated basophils, the relationship of its generation to LTC4 and histamine release and the role of sPLA₂ in the generation of AA for the synthesis of LTC₄ in human basophils. We have used inhibitors of sPLA₂ to determine if the synthesis of leukotrienes in human basophils is coupled to the release of AA from phospholipids hydrolyzed specifically by 14-kDa PLA₂.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. The following were purchased: PIPES, fMLP, BSA, PMSF, leupeptin, soybean trypsin inhibitor, aprotinin (Sigma Chemical Co., St. Louis, MO); crystallized HSA (Miles Laboratories, Elkhart, IN); FCS and RPMI 1640 containing 25 mM HEPES (Gibco, Grand Island, NY); Percoll (Pharmacia, Piscataway, NJ). Zileuton (SKF 108008), SB 203347 were obtained from SmithKline Beecham Pharmaceuticals, King of Prussia, PA and mAb 3F10 and recombinant human type II 14-kDa PLA₂ were prepared as described previously (Marshall et al., 1994; Roshak et al., 1994). Triacsin C was obtained from BIOMOL Research Laboratories, Inc. Plymouth Meeting, PA. [³H]AA was purchased from Du Pont, Wilmington, DE. BPO(7)-HSA (benzylpenicilloyl conjugated to human serum albumin with a valency of 7) and BPO-EACA (benzylpenicilloyl-e-aminocaproic acid) were prepared previously (MacGlashan et al., 1986).

Buffers. PIPES-albumin-glucose (PAG) buffer contained 25 mM PIPES, 100 mM NaCl, 5 mM KCl, 0.1% glucose and 0.003% HSA, pH 7.3, PIPES buffer excluding the albumin and glucose. Release experiments were carried out in PAG supplemented with 1 mM CaCl₂ and 1 mM MgCl₂. Counter-current elutriation was carried out in PAG containing 0.25% BSA in place of 0.003% HSA.

Basophil preparations. Human basophils were isolated from leukocyte enriched preparations from residual blood after plateletpheresis. Fractions of this preparation were further enriched for basophils by counter-current elutriation and Percoll gradient centrifugation as previously described (MacGlashan et al., 1994). Basophil preparations were 85% basophils or greater, as determined by alcian blue staining (Gilbert and Ornstein, 1975).

AA and 5-HETE measurements. The masses of AA and 5-HETE were determined via GC/(NICI)MS with modifications of described procedures (Hubbard et al., 1986). This method provides measurements of the mass of endogenous AA and 5-HETE release as opposed to observing the release of radiolabeled AA incorporated into cellular lipids. Basophils were incubated with or without stimuli at 37°C, in polypropylene microfuge tubes (1.7 ml, PGC Scientifics Gaithersburg, MD) at a final volume of 100 µl. For AA and 5-HETE mass measurements, reactions were terminated with the addition of 1 ml acetone followed immediately by the addition of 15 to 20 ng of each 5,6,8,9,11,12,14,15-octadeuterated (²H₈) AA (d₈-AA) and 5-HETE (d₈-5-HETE) as internal standards. The acetone/buffer mixtures were transferred to silanized glass vials and after evaporation under N₂, the carboxyl moieties of AA and 5-HETE were converted to pentafluorobenzyl esters as described (MacGlashan and Hubbard, 1993). After evaporation of volatile reagents under N₂, the dried residue was treated with 50 µl of BSTFA in the presence of 25 µl of acetonitrile for conversion of the 5-hydroxyl moiety of 5-HETE to the trimethylsilyl ether derivative as described in (Hubbard et al., 1986). Subsequent to the evaporation of volatile reagents under N₂, the derivatized AA and 5-HETE were dissolved in 30 µl of heptane for injection into the gas chromatograph. GC/(NICI)MS conditions for measurement of released AA and 5-HETE are described in earlier reports (Hubbard et al., 1986; MacGlashan and Hubbard, 1993). 5-HETE mass was determined by ion monitoring at m/z(mass-to-charge ratio) 391 for the 5-HETE TMS derivative vs. m/z 399 for the TMS derivative of the d₈-5-HETE internal standard. The amount of AA contributed by contaminating cells (monocytes and lymphocytes) was assessed by stimulation of cell preparations ranging from 39 to 89% basophils with fMLP. These data suggested that, on a per cell basis, the contaminating cells contributed an equal amount of AA (data not shown). Thus, it is important to note that all data is based on basophil preparations of >85% purity and that contaminating cells contributed only a small portion of the AA mass measured. For the measurements of free AA associated with the cell pellet vs. the buffer, the reaction was stopped by centrifugation for 5 sec at 14,000G, the top 50 µl removed into 1 ml acetone and 1 ml of acetone added to the remaining buffer plus cell pellet. After AA analysis by GC/(NICI)MS, the fraction in the buffer was calculated by doubling the amount determined in the top half of the buffer and dividing by the total AA, which was the top half buffer determination plus the amount found in the remaining buffer plus cell pellet.

Histamine and leukotriene C₄ measurements. Incubations were as described above. For the measurement of histamine and LTC₄, the reactions were terminated with the addition of 400 µl ice-cold normal saline containing 1 mM EDTA and reactions were centrifuged in a Beckman 12 Microfuge (Beckman Instruments, Inc., Palo Alto, CA) for 1 min at 12,000 × g. For histamine measurements 200-µl aliquots of the supernatant were diluted in normal saline. Histamine release was determined by an automated method of (Siraganian, 1974). LTC₄ release was determined using 100-µl aliquots of the supernatants and an in-house LTC₄ assay (Hayes et al., 1983; MacGlashan et al., 1986) which has a sensitivity of 50 pg/ml.

Measurement of 5-lipoxygenase activity. Human basophils (0.5-1 × 10⁶ basophils in 100 µl PAGCM) were incubated with or without 4 µM SB 203347 for 15 min at 37°C which was followed by the addition of 0.5 µCi of [³H]AA in the presence or absence of fMLP or ionomycin as will be noted in the results. Labeled lipids and AA metabolites were extracted using a modification of Rouzer (Rouzer et al., 1982) as follows, reactions were terminated with the addition of 200 µl ethanol, followed by 20 µl 88% formic acid, samples were extracted with three washes of 400 µl CHCl₃ [with 0.01% (w/v) butylatedhydroxytoulene], dried under N₂, resuspended in 250 µl CHCl₃ and then 250 µl 0.1% (v/v) glacial acetic acid pH 3.7 are added. The extracts were analyzed by reverse phase HPLC on an ultrasphere ODS column (4.6 mm × 25 cm) (Beckman) after the protocol of (Peters et al., 1983).

Measurement of [Ca⁺⁺]_i. fMLP-stimulated changes in basophil [Ca⁺⁺]_i concentration were measured by digital video microscopy of Fura-2 loaded basophils as described previously (MacGlashan and Warner, 1991). The basophils were labeled with Fura-2 by incubation with 1 µM Fura-2AM (Molecular Probes, Eugene, OR) for 20 min at 37°C in RPMI-1640 (Gibco) containing 0.32 mM EDTA and 2% FCS. After washing, the basophils (300,000-500,000 cells) were suspended in 200 µl of PAG buffer and were held on ice until use. For each analysis, a 15-µl aliquot (approximately 25,000 basophils) was placed onto a siliconized glass coverslip in the microscope chamber. After allowing them to settle, the cells were overlayed with 1 ml of PAG containing 1 mM CaCl₂ and 1 mM MgCl₂. The cells were allowed to warm to 37°C, and ratio (352/380 nm) images of four fields were collected at 30-sec intervals to establish the baseline. The fMLP then was added, and the [Ca⁺⁺]_i was monitored by 10 to 20 single wavelength (380 nm) images at 2-sec intervals and then by ratio images taken at 5-sec intervals. The fluorescence values were converted to [Ca⁺⁺]_i as described previously (MacGlashan and Warner, 1991). At the end of each analysis, 1 ml of supernatant was collected, and histamine content was measured using the automated fluorometric technique or LTC₄ was measured as described above. Histamine release was expressed as the percentage of total histamine content, which was determined using supernatant of a cell aliquot lysed with 2% perchloric acid.

Purification of basophils for isolation and assay of PLA₂. To obtain basophils at nearly 100% purity for the measurement of sPLA₂, basophils already purified by the above procedures were further enriched by positive selection. The method used here further improves on a method that has been previously published (MacGlashan et al., 1994). The enriched basophils were incubated with a monoclonal anti-IgE antibody (TES-19, 2 µg/ml, Tanox Corp., Houston, Texas) for 10 min on ice and in PAG containing 50 µM EGTA. Following 2 washes, the cells were incubated with MACS anti-mouse IgG 2a+b (Millentyl Biotec, Auburn, CA) at a ratio of one part bead preparation to six parts cells in PAG/EGTA. After a 20-min incubation in an ice bath and two washes, the cells were resuspended in PAG/EGTA containing 0.06% HSA, the cells were run through a mini-MACS column. The contaminating cells were collected and after their removal the basophils were collected. Typically these preparations are greater than 99% basophils, for the cells used to prepare sPLA₂, there were 0.3 and 0.5% nonbasophilic leukocytes. For comparative purposes, the contaminating cells were obtained from the upper layer of the last Percoll procedure (see above). These cells are typically of the same general composition as the contaminating cells that accompany the basophils. Since the basophils were handled in ice cold buffers and in EGTA, there should have been little, if any, IgE-mediated activation during preparation. Cells were pelleted from ice-cold PAG and stored at 70°C. The frozen pellets were resuspended in homogenization buffer containing 0.34 M sucrose, 10 mM HEPES, 1 mM EGTA, 1 mM PMSF, 200 µM leupeptin, 20 µg/ml soybean trypsin inhibitor and 20 µg/ml aprotinin, pH 7.4, and PLA₂ was isolated from the suspension after disruption by nitrogen cavitation. The amount of sPLA₂ was measured using an ELISA as described in (Bolognese et al., 1995).

Assay for extracellular PLA₂ activity. PLA₂ activity was assayed using a modification of a protocol suggested by Du Pont technical service. Reactions were in 1.7-ml microfuge tubes containing 5 × 10⁵ basophils stimulated with fMLP in the presence of 0.1 µCi of a [³H]AA-labeled Escherichia coli suspension (from Du Pont) in a final reaction volume of 200 µl. Basophils were incubated at 37°C for 30 min. The assay was terminated with the addition of 100 µl 2 N HCl and 100 µl 2% BSA to each tube and tubes were then incubated at 4°C for 30 min. After 30 min, particulate material was pelleted by centrifugation in a microfuge, 300 µl of each reaction supernatant were collected and counted in a scintillation counter.

	Results

Top Abstract Introduction Methods Results Discussion References

AA release, metabolism and LTC₄ generation in fMLP-stimulated basophils. Before testing the effects of sPLA₂ inhibitors on AA or LTC₄ generation, several preliminary studies were conducted. These are summarized in figures 1 and 2 and table 1. Previous studies profiled the kinetics of AA generation and the resulting curves indicated a near plateau in its generation within a few minutes of stimulation with fMLP (MacGlashan and Hubbard, 1993). While there was a slight decline in the measured AA after several minutes, these results did not discriminate between two possible interpretations, AA generation had reached a steady state of production and metabolism or AA production ceased after several minutes and the measured AA was not further metabolized. Both conditions could lead to a plateau of measured AA. Ideally, the experiment to discriminate between the two possible interpretations is to dissociate the ligand at the plateau of the kinetic curve. A subsequent decrease in free AA would suggest the steady state interpretation is correct while no change in free AA would suggest AA generation was complete with no further metabolism. This is not straightforward for fMLP although it can be done for an IgE-mediated stimulus as will be discussed below. For the fMLP-induced reaction, EGTA was added to stop the reaction after 10 min. From studies of cytosolic calcium elevations, it is known that this will immediately return cytosolic calcium levels to resting conditions as well as reducing the free Ca⁺⁺ extracellularly. Figure 1A demonstrates that there is no significant decrease in the measured free AA after EGTA addition. This time course is compared to the time course for cells followed without the addition of EGTA. There was only a slight decrease in the amount of measured AA in the presence of EGTA. Although there may be some problems with this approach, at face value the data indicate that the plateau of measured AA represents the completion of AA generation.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Kinetics of AA generation after stimulation with fMLP or antigen, BPO(7)HSA. In A (n = 2), purified basophils were stimulated with 1 µM fMLP and reactions stopped with acetone at the times indicated. After 10 min (indicated by the arrow), 5 mM EGTA (final concentration) was added to one set of tubes () and PAGCM (containing 1 µM fMLP) was added to another set () and the reaction stopped with acetone at various times after these additions. In B (n = 1, errors indicate standard deviations of duplicate measurements) sensitized basophils were stimulated with 1 µg/ml BPO(7)HSA and reactions stopped with acetone for the times indicated. After 15 min (indicated by the arrow), 0.1 mM BPO-EACA (final concentration) was added to one set of tubes () and PAGCM (containing BPO(7)HSA at 1 µg/ml) was added to another set () and the reactions stopped with acetone at various times after these additions. At the 10-min point, a set of tubes received BPO-EACA, to test the efficacy of the BPO-EACA.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   The effects of zileuton and triacsin C on the accumulation of AA and release of LTC4 in human basophils stimulated with fMLP. Basophils were preincubated with zileuton (10 µM) and triacsin C (1 µM) for 15 min before stimulation with 1 µM fMLP for 10 min. A, Treatment with zileuton alone and B, treatment with both zileuton and triacsin C. Controls (open columns), zileuton (black columns), triacsin C (open columns with strips), zileuton and triacsin C (black columns with strips). Data are corrected for the AA or LTC4 found in unstimulated cells which were 2 to 5 pmol/106 and 0.0 to 0.5 pmol/106 for AA and LTC4, respectively. All data are the means ± S.E.M. of experiments conducted on different days with different cell donors (n = 6). The asterisk indicates data significantly different from control at P < .05.

Effects of sPLA₂ inhibitors on stimulated AA release and LTC₄ generation. To examine the role of the secretory type II PLA₂ in the generation of AA for the synthesis of LTC₄, inhibitors of this enzyme were used. A monoclonal antibody developed by SKB, mAb 3F10, has been shown to neutralize 14-kD sPLA₂ activity in solution (Marshall et al., 1994). As shown in figure 3, fMLP-induced LTC₄ synthesis was also inhibited (IC₅₀  80 µg/ml) by preincubation of intact basophils with mAb 3F10 for 10 min (a control antibody used at the same concentrations did not have this effect and the 3F10 antibody had no effect on the measurement of LTC₄ in the radioimmunoassay itself). Although inhibiting LTC₄ synthesis, the antibody had no statistically significant effect on free AA and histamine release.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   The effects of mAb 3F10 (anti-PLA2) on LTC4, AA and histamine release from human basophils. Basophils were incubated with various concentrations of 3F10 for 15 min at 37°C before stimulation with 1 µM fMLP. Data are expressed as percent of stimulated controls. Control values were 38 ± 16 pmol per 106 basophils, 55 ± 13 pmol per 106 basophils, and 84 ± 7% for AA (), LTC4 (), and histamine release (), respectively. Nonspecific mouse IgG at 105 µg/ml was used in control incubations. All data are the means ± S.E.M. of experiments conducted on different days with different cell donors (n = 4).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   The effects of SB 203347 on LTC4, AA and histamine release from human basophils. Basophils were incubated with various concentrations of SB 203347 for 15 min at 37°C before stimulation with 1 µM fMLP. Cells were incubated with fMLP for 10 min before harvesting for mediators or AA analysis, () arachidonic acid, () LTC4, () histamine release. All data are the means ± S.E.M. of experiments conducted on different days with different cell donors (n = 4). Panel inset, the effects of SB 203347 on anti-IgE antibody (300 ng/ml for 10 min) stimulated LTC4, AA, and histamine release. Data are expressed as percent of anti-IgE stimulated controls. Control values for anti-IgE were 4.5 ± 0.4 pmol/106 basophils, 2.2 ± 0.2 pmol per 106 basophils, and 25 ± 5% for AA (), LTC4 () and histamine release (), respectively.

Inhibition of AA generation exposed by treatment with zileuton/triacsin C. From the preliminary experiments summarized in figure 2, we know that pretreatment of basophils with zileuton will partially reveal the mass of AA utilized for LTC₄ synthesis. Using the approach discussed above, where basophils were first treated with a combination of zileuton and triacsin C (or either inhibitor was used alone), we tested for the effect of mAb 3F10 on the increased AA observed under these conditions. Figure 5 shows the results of these experiments. Note that the results are expressed as the net AA generated after fMLP stimulation. As shown in figure 5 (or fig. 2), the presence of triacsin C alone increased the stimulated AA generation (fMLP-stimulated cells not treated with zileuton or 3F10, bar 1 vs. bar 5, counting from the left). We were interested in how mAb 3F10 affected AA release that was revealed by the inclusion of zileuton, i.e., inhibiting the pathway to LTC₄ generation. The effects of 3F10 were qualitatively similar for either method (zileuton alone or zileuton + triacsin C) of revealing this AA. As expected, the addition of zileuton (without 3F10) increased the measured AA generation. AA generation increased by 48 pmol/10⁶ basophils in the absence of triacsin C and by 92 pmol/10⁶ basophils in the presence of triacsin C. mAb 3F10 inhibited only this increase in free AA, i.e., the levels of AA generated in the presence of mAb 3F10 were the same as cells not treated with zileuton or mAb 3F10 (bar 4 vs. bar 2 or bar 8 vs. bar 6). mAb 3F10 alone had no statistically significant effects on AA generation in the presence or absence of triacsin C. In the absence of fMLP, triacsin C also increased the resting free AA but this increase was also not inhibited by 3F10 (data not shown). These results indicate that 3F10 could inhibit the fMLP-stimulated generation of a mass of free AA exposed by inhibiting the 5-LO pathway. These data also indicate that there were two pools of AA being generated, one that was inhibitable by anti-sPLA2 (3F10) and one that was not.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   The ability of mAb 3F10 to attenuate the increase in AA produced by zileuton in basophils stimulated with fMLP. Basophils were preincubated for 15 min with zileuton in the absence or in the presence of mAb 3F10 (105 µg/ml) before stimulation. Basophils were also preincubated with zileuton in combination with triacsin C in the presence or absence of mAb 3F10. The addition of the respective reagents is noted by the plus and minus characters, zileuton was used at 10 µM and triacsin C at 1 µM. All data are the means ± S.E.M. of net AA generation after a 15-min challenge with FMLP (1 µM) for experiments conducted on different days with different cell donors (n = 3). The asterisk indicates data significantly different from control at P < .05. Bars 1, 2, 4 (counting from the left) were not statistically different from one another, neither were bars 5, 7 and 8 different from one another (by analysis of variance).

View larger version (36K): [in this window] [in a new window]   Fig. 6.   The ability of SB 203347 to attenuate the increase in AA produced by zileuton in the presence or absence of triacsin C in basophils stimulated with fMLP. Basophils were preincubated for 15 min with zileuton in the absence or in the presence of SB 203347 (4 µM) before stimulation. Basophils were also preincubated with zileuton in combination with triacsin C in the presence or absence of SB 203347. A, The addition of the respective reagents is noted by the plus and minus characters, zileuton was used at 10 µM and triacsin C at 1 µM. All data are the means ± S.E.M. of net AA generation after a 15-min challenge with FMLP (1 µM) for experiments conducted on different days with different cell donors (n = 4) B, LTC4 release, the presence or absence of the various reagents is indicated. All data are the means ± S.E.M. The asterisk indicates data significantly different from control at P < .05. For A, bars 1, 2, 4 (counting from the left) were not statistically different from one another, neither were bars 5, 7 and 8 different from one another (by analysis of variance).

Side effects of SB203347. Although these two very different inhibitors of sPLA₂ activity resulted in similar data, there were concerns about the possible nonspecific effects of SB 203347. There were two concerns addressed, the first was related to the possibility that SB 203347 was acting as a pure or partial 5-LO, LTA₄ or LTC₄ synthase inhibitor and the second was whether SB 203347 had some competitive binding effects on the fMLP receptor as indirectly suggested by earlier studies in neutrophils.

sPLA₂ in basophils. The data suggest that a sPLA₂ is responsible for the generation of AA that is ultimately metabolized to LTC₄. Using an ELISA, we examined two basophil preparations, that were purified to nearly 100% (0.3-0.5% contaminants), for the presence of sPLA₂. These were compared to a similar number of contaminant cells from the same preparations. This assay indicated that there was 21.5 ± 1.9 pg/µg cell protein of sPLA₂ in basophils vs. 23.4 ± 1.8 pg/µg protein for the contaminating leukocytes. The similar levels for basophils and leukocytes (which translates to 130 pg/10⁶ cells for each, or 5000 molecules/cell) indicates that the sPLA₂ measured in the highly purified preparation of basophils was obtained from the basophils and not the minimal number of contaminating cells.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Effects of type II rhsPLA2 on free AA generation and LTC4 release from basophils. A shows the concentration dependence on free AA generation by rh PLA2 applied exogenously to basophils. The insert shows the results for the lower concentrations of rh PLA2. The release of free AA was measured under three conditions of stimulation () buffer (n = 3), () FMLP (n = 3) and () anti-IgE antibody (n = 2). Panel B shows the LTC4 release obtained from the same samples shown in A. The sample symbols apply to B and C. C shows histamine release obtained from the same samples shown in A.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The data presented indicate that not all AA generated during activation of basophils is made available to the generation of LTC₄. This is similar to a viewpoint that has emerged in the study of AA metabolism in a number of other cell types. The evidence for this perspective in human basophils comes from several types of experiments. First, there was an unpredictable relationship between the ability of basophils to generate free AA and to generate LTC₄ when different stimuli were compared (table 1). It could be argued that the apparent dissociation in the generation of two types of lipid products reflects differences in the signal transduction pathways for the different stimuli. However, for any stimulus, triacsin C caused marked increases in free AA without a concomitant increase in LTC₄ release (or histamine release). In addition, the free AA generated after the addition of exogenous rh-sPLA2 did not affect the generation of LTC4 in stimulated or unstimulated cells (if basophils are later shown to express type V sPLA2, the absence of an effect by exogenous group II sPLA on LTC4 release might be explained by our use of the wrong type).

Treatment of basophils with the anti-sPLA₂ monoclonal antibody 3F10 both strengthened the suggestion that different pools of AA contribute to LTC₄ production as well as first suggesting that a sPLA₂ might be involved in the generation of AA used by the 5-LO pathway. In these studies, 3F10 decreased the amount of LTC₄ produced at concentrations that did not significantly change AA mass. The ability of 3F10 to inhibit LTC₄ synthesis in the absence of an effect on the mass of observable AA suggests a different hypothesis about the coupling of AA to the 5-LO pathway. The data suggest that there are two pools of AA generated. The proposed model would state that one pool of AA is tightly coupled to its metabolism by 5-LO so that it is not normally observed. Any free AA that is normally observed belongs to a second pool of generated AA that may serve other purposes in the cell response. This model explains the data in figure 3, inhibition of LTC4 release and inhibition of observed free AA can be dissociated because the antibody inhibits an enzyme responsible for generating the unobserved pool of AA. One prediction of this model is that if the AA normally metabolized to LTC4 could be revealed, inhibitors of sPLA2 should eliminate its appearance at concentrations similar to those found to inhibit LTC4 release. This was indeed the case, adding zileuton to cells stimulated with fMLP caused an increase in AA that did not exceed the expected gain from unsynthesized LTC₄. The addition of mAb 3F10 inhibited only this increase. Similar results were observed in cells treated with both zileuton and triacsin C. While the model doesn't directly predict the effects of mAb 3F10 on cells treated with triacsin C alone, these results were perhaps not unexpected, i.e., there was no significant inhibition of the appearance of this AA. The fact that triacsin C has an effect on this source of "noninhibitable" AA also indicates that this free AA is normally in a steady state of generation and reacylation during its production. These results support the view that there are two pools of AA generated during stimulation and that one pool results from the action of a process involving sPLA₂.

A second inhibitor of sPLA₂, SB 203347, replicated both of these critical results; the inhibition of LTC₄ release in the absence of an inhibition of observable AA and the inhibition of only the AA made visible by treating the cells with zileuton (± triacsin C). As with all pharmacological studies, the specificity of SB 203347 is an important consideration for an interpretation of these results although the results with 3F10 allay some concerns about the overall interpretation of the data. Most critically, inhibition of 5-LO or LTA₄/LTC₄ synthase might replicate the observations for SB 203347. However, as noted above, inhibition of 5-LO by zileuton results in free AA accumulation and this effect was not observed for SB 203347. Although previous studies have noted inhibition of 5-LO in microsomal preparations by SB 203347, there was at least a 10-fold difference in the IC₅₀s for sPLA₂ and 5-LO inhibition (Marshall et al., 1995). Nor did we observe enhanced 5-HETE (the metabolic product of 5-HPETE) accumulation. Indeed, the dose response curve for inhibition of 5-HETE precisely mirrored inhibition of LTC₄. The lack of enhancement of this precursor suggests that there was no inhibition of LTA₄ synthase and possibly LTC₄ synthase. Attempts to observe the effects of SB 203347 on movement of [³H]AA through the 5-LO pathway were inconclusive since exogenously added AA was rapidly incorporated into the phospholipids of intact basophils. This was especially true for fMLP-stimulated cells even in the presence of triacsin C. Nevertheless, the portion of these experiments where there was no stimulation or the cells were stimulated with ionomycin did suggest that SB 203347 had little effect on conversion of exogenous [³H]AA to [³H]LTC₄. Previous studies demonstrated that SB 203347 was a more potent inhibitor of sPLA₂ than an inhibitor of cPLA₂ suggesting that at a concentration that caused inhibition of LTC4 release, this drug was acting on a 14-kDa PLA2. Neither mAb 3F10 or SB203347 can discriminate between type II and type V sPLA₂ (Chen et al., 1994a, 1994b). As recent studies have indicated that secretory cells may actually express type V rather than type II (Balboa et al., 1996), it is possible that this will hold true for basophils as well.

If this interpretation of the results is correct, our data suggest that some aspect of basophil-associated sPLA₂ has an extracellular presentation. It had the unexpected attribute of being accessible to the 3F10 antibody without evidence that it was in the surrounding buffer; an exogenous substrate, ³H-AA labeled E. coli, was not hydrolyzed when included in the buffer of stimulated cells. Neither was ³H-AA generated from ³H-AA labeled E. coli after the addition of harvested buffers. Free AA generated by sPLA2 added to the buffer did not result in enhanced LTC4 release suggesting that this location was not appropriate for proper functioning of the basophil-associated sPLA2. The data also suggested that under normal conditions this coupling between sPLA₂/AA generation and its conversion by 5-LO was closely linked. One speculative image that fits these pieces of data is of a macromolecular complex with sPLA₂ having an external exposure that is accessible to the antibody mAb 3F10 but is otherwise positioned in a way not mimicked by the simple addition of sPLA₂. This sPLA₂ might then be closely linked to a 5-LO such that the free AA it generates is efficiently passed to the 5-LO. These results with mAb 3F10 suggest a model of the orientation of sPLA2 that is not unlike a model for CD45. Anti-CD45 antibodies inhibit IgE-mediated release in human basophils (Hook et al., 1991). CD45 has an extracellular region, to which the antibody binds, and an intracellular catalytic domain responsible for dephosphorylating src-related proteins. The mechanism by which anti-CD45 antibodies inhibit IgE-mediated release is unknown by could occur by sequestering the CD45 away from FcRI and associated tyrosine kinases. A similar model could be proposed for the action of mAb 3F10 for its effects on sPLA₂ except for the fact that the amino acid sequence of group II or V sPLA2 does not appear to include a transmembrane region. It is therefore more difficult to visualize the nature of the spatial arrangement for sPLA₂. This model also ignores recent evidence that leukotriene synthesis takes place on the nuclear membrane rather than the plasma membrane (Brock et al., 1994; Woods et al., 1993, 1995). Studies localizing the 5-LO or LTC₄ synthase enzymes to the nuclear membrane have not been carried out yet in human basophils, leaving open the possibility that metabolism occurs in the plasma membrane for these cells. Furthermore, even if 5-LO actually resides on the nuclear membrane, there may also be many points of juxtaposition between plasma and nuclear membranes. The fact that exogenously supplied sPLA₂ does not enhance LTC₄ release also indicates an important subtlety about this AA coupling to 5-LO.

In previous studies, we found that IL-3 would induce a qualitative shift in the response of basophils to C5a; without IL-3, C5a-stimulated basophils do not secrete LTC₄, although a brief treatment with IL-3 permitted C5a to induce the generation of LTC₄ (Kurimoto et al., 1989; MacGlashan and Hubbard, 1993). In these studies we noted that IL-3 also changed the level of AA generation. Free AA, above resting levels, was nearly absent after C5a-stimulation in untreated cells although it was generated in significant quantities after IL-3 treatment. At that time, we concluded that this AA was responsible for the appearance of LTC₄ synthesis. However, in light of the current studies, the AA that appears after IL-3 treatment might not represent sPLA₂-derived AA but that coming from an alternate pathway (cPLA₂ or triglyceride lipase). Because LTC₄ synthesis is also augmented we would tentatively conclude that either both sPLA₂ and alternate pathway enzymes were upregulated by IL-3 treatment or IL-3 treatment shifts the basophil's dependence on sPLA₂ for AA/LTC4 generation. This aspect of the effects of IL-3 on AA generation in basophils will require a reexamination.

As noted in the introduction, it is intriguing that a receptor for a sPLA₂ has been identified and cloned (Lambeau et al., 1994). Although the existence of such a sPLA₂ binding receptor could alter our perspective on how AA generation is regulated in basophils, this particular receptor does not appear to have a good affinity for the human sPLA₂ that we are studying. The studies using 3F10 antibody indicate that the basophil sPLA₂ regulating LTC₄ release is not one of those with high affinity for this potential receptor. The fact that exogenously added rhPLA₂ also did not influence LTC₄ release might also indicate that such a receptor is not present in basophil membranes. The caveat in this instance is that basophils might have the receptor but it was already occupied, leaving the exogenously added sPLA₂ to operate in another manner. An alternate view is that derived from the studies of Chock (Chock et al., 1994), where sPLA₂ was localized to the granules of mast cells. LTC₄ synthesis follows degranulation in all receptor-stimulated reactions in human basophils (e.g., fMLP and anti-IgE antibody) so it is plausible that degranulation must occur first, allowing sPLA₂ to find it way to the plasma membrane or the outer face of the cell.

In summary, these data suggest that leukotriene synthesis in fMLP-stimulated human basophils results from the generation of free AA by an extracellularly accessible sPLA₂. The data further show that not all AA generation has an impact on the level of LTC₄ synthesis, suggesting that the source of AA is a critical determinant of its use by various metabolic pathways.