Phosphatidylserine-specific phospholipase A1 stimulates histamine release from rat peritoneal mast cells through production of 2-acyl-1-lysophosphatidylserine.

Lysophosphatidylserine (1-acyl-2-lyso-PS) has been shown to stimulate histamine release from rat peritoneal mast cells (RPMC) triggered by FcepsilonRI (high affinity receptor for IgE) cross-linking, although the precise mechanism of lyso-PS production has been obscure. In the present study we show that phosphatidylserine-specific phospholipase A(1), PS-PLA(1), stimulates histamine release from RPMC through production of 2-acyl-1-lyso-PS in the presence of FcepsilonRI cross-linker. The potency of 2-acyl-1-lyso-PS was almost equal to that of 1-acyl-2-lyso-PS. A catalytically inactive PS-PLA(1), in which an active serine residue (Ser(166)) was replaced with an alanine residue did not show such activity. sPLA(2)-IIA, another secretory PLA(2) that is capable of producing lyso-PS in vitro, was also a poor histamine inducer against RPMC. PS-PLA(1) significantly stimulated histamine release from crude RPMC, indicating that lyso-PS is mainly derived from cells other than mast cells. In agreement with this phenomenon, the enzyme stimulated the histamine release more efficiently when RPMC were mixed with apoptotic Jurkat cells. Under these conditions, lyso-PS with unsaturated fatty acid was released from the apoptotic cells treated with PS-PLA(1). Finally, heparin, which has affinity for PS-PLA(1), completely blocked the stimulatory effect of the enzyme. In conclusion, PS-PLA(1) may bind to heparan sulfate proteoglycan, efficiently hydrolyze PS appearing on plasma membranes of apoptotic cells, and stimulate mast cell activation mediated by 2-acyl-1-lyso-PS.

serine (lyso-PS; 1-acyl-2-lyso-PS) (2, 3) strongly enhances the secretory response of rat peritoneal mast cells (RPMC), which is initiated by binding of antigen-IgE complexes to high affinity receptors for IgE (Fc⑀RI). Lyso-PS is ϳ1,000 times more active at inducing this response than PS (2,3). The action of lyso-PS is highly specific; all other lysophospholipids (including lysophosphatidyl-D-serine, an optical isomer of lysophosphatidyl-L-serine) have been reported to be ineffective (2,3). Thus, it has been suggested that a specific receptor for lyso-PS should exist on the plasma membrane of RPMC (4,5). Enhancement of mast cell degranulation can be observed at a submicromolar (Ͻ10 Ϫ7 M) concentration of lyso-PS. Lyso-PS is also reported to potentiate nerve growth factor-induced differentiation of PC12 cells (6) and regulate proliferation of human T cells (7). In addition, lyso-PS, like other lysophospholipid mediators such as lysophosphatidic acid, sphingosine 1-phosphate, and sphingosylphosphorylcholine, induces a transient increase in cytosolic free calcium ([Ca 2ϩ ] i ) in several cell lines including Jurkat cells (8) and ovarian cancer cells (9). Although lyso-PS (1-acyl-2-lyso-PS) is thought to be produced from PS by phospholipase A 2 (PLA 2 ) reaction (10), the physiological role of lyso-PS is unclear as the enzyme involved in lyso-PS production has not been fully elucidated.
Recently, PS-specific PLA 1 , which was first discovered in the culture medium of activated rat platelets, has been purified and cloned (11). This secretory enzyme, called PS-PLA 1 , belongs structurally to a lipase family and acts specifically on PS, hydrolyzing fatty acid from the sn-1 position to produce 2-acyl-1-lyso-PS, another lyso-PS that has not been fully characterized (11)(12)(13)(14). PS appears to be the only phospholipid that is hydrolyzed by this enzyme, although we do not know what type of membrane is a target of the enzyme. PS-PLA 1 does not hydrolyze phosphatidyl-D-serine, an optical isomer of phosphatidyl-L-serine. Thus it is reasonable to assume that 2-acyl-1-lyso-PS has the same activity as 1-acyl-2-lyso-PS and that PS-PLA 1 is involved in mast cell activation by producing 2-acyl-1-lyso-PS, although its function has not been characterized yet.
Another secretory phospholipase A that may have an ability to produce lyso-PS is group IIA sPLA 2 (sPLA 2 -IIA) (15,16). Like PS-PLA 1 , sPLA 2 -IIA was first identified in culture supernatants from activated rat platelets (17). This enzyme is often detected at various sites of inflammation and has been implicated in the inflammation process and eicosanoid production (18 -20). In vitro, this enzyme is able to produce 1-acyl-2-lyso-PS from PS, although it also hydrolyzes other phospholipids such as phosphatidylethanolamine and phosphatidylcholine.
In this study, the role of two phospholipase As, PS-PLA 1 and sPLA 2 -IIA, in the mast cell activation was investigated. The results indicate that PS-PLA 1 is indeed capable of producing lyso-PS from intact membranes and is involved in mast cell activation. Our findings indicate that PS-PLA 1 plays a key role in the production of bioactive lysophospholipid, lyso-PS.
Mast Cells-RPMC from the peritoneal cavity of male Wistar rats, weighing 200 -250 g, were purified with Percoll (Amersham Pharmacia Biotech, Uppsala, Sweden) as described previously (23). Purified RPMC were suspended in HEPES-buffered Tyrode solution (137 mM NaCl, 2.7 mM KCl, 12 mM HEPES, 1 mM MgCl 2 , 2 mM CaCl 2 , 5.6 mM dextrose, and 0.01% bovine serum albumin, pH 7.4). The purity of the RPMC in the final preparation was Ͼ95%, as estimated by toluidine blue staining. We used the cells recovered from the peritoneal cavity as "crude RPMC." Using Wright-Giemsa staining we confirmed that the crude peritoneal cells (about 2 ϫ 10 7 cells per rat) consisted mainly of mononuclear leukocytes (macrophages and monocytes) with a minor population of other cells such as mast cells and polymorphonuclear leukocytes (neutrophils). The ratio of mononuclear leukocytes, polymorphonuclear leukocytes, and other cells was typically 94:1:5. Thus in the crude RPMC preparation there are about 20 times as many other cells as mast cells.
Cell Culture-Human Jurkat T cells were maintained in RPMI 1640 medium containing 5% fetal calf serum in an atmosphere of 5% CO 2 . To induce apoptosis, the cells were incubated with serum-free RPMI 1640 medium in the presence of anti-Fas monoclonal antibody (CH-11, 20 ng/ml) for 4 h at 37°C. In the co-culture system, we mixed 1 ϫ 10 4 of purified RPMC and 2 ϫ 10 5 Jurkat cells.
Determination of Apoptotic Cells-PS exposure was measured by the binding of FITC-labeled annexin V using the Annexin V-FITC Kit (Immunotech). Briefly, 5 ϫ 10 5 cells (apoptotic cells or rat peritoneal cells) were washed with ice-cold phosphate-buffered saline and resuspended in 495 l of binding buffer. Annexin V-FITC (5 l) was added to the suspension and incubated for 10 min on ice in the dark. The cells were analyzed by flow cytometry using EPICS XL flow cytometer (Beckman Coulter).
Assay of Histamine Release from RPMC-The RPMC were washed once, suspended at a cell density of 5 ϫ 10 4 RPMC/ml (in 0.2 ml) in HEPES-buffered Tyrode solution and sensitized with 10 g/ml monoclonal IgE anti-DNP-As for 30 min at 37°C. After two washes in HEPES-buffered Tyrode solution, the cells were stimulated with 60 g/ml DNP-As at 37°C in the presence of lyso-PS or recombinant PS-PLA 1 . After 15 min, 1.2 ml of ice-cold HEPES-buffered Tyrode solution was added to terminate the reaction, and the reaction mixture was centrifuged at 800 ϫ g for 5 min at 4°C. Histamine in the super-natant was determined by the fluorometric assay of Shore et al. (24). Histamine release was calculated as a percentage of the total cell content. Values for histamine release are presented as the means Ϯ S.E. for several replicate experiments on different samples of pooled cells. Spontaneous histamine release in the absence of compounds was 3.4 Ϯ 3.3%. In some cases, concanavalin A (ConA; 100 g/ml), which is known to cross-link the Fc⑀RI receptor, was used to prime the RPMC instead of IgE-antigen. HEPES-buffered Tyrode solution used in this study contains 0.01% fatty acid-free bovine serum albumin. Albumin is an essential factor in the assay system to evaluate lyso-PS activity, although high concentration (Ͼ1%) affected the activity (data not shown). We determined optimal concentration of bovine serum albumin to be a range from 0.01 to 0.1% and used 0.01% in this study (data not shown).
Preparation of PS-PLA 1 and sPLA 2 -IIA-Recombinant rat PS-PLA 1 was prepared as described previously (11). Briefly Sf9 cells (8 ϫ 10 5 cells/ml) in a spinner flask were infected with recombinant rat PS-PLA 1 baculovirus (multiplicity of infection ϭ 10) and incubated for 96 h at 27°C. The proteins were purified from the culture supernatant of the infected cells by sequential passages through a DEAE ion exchange column, a heparin-Sepharose column, and a blue Sepharose column as described previously (11). Approximately 400 g of recombinant PS- PLA 1 was recovered from 1 liter of infected culture fluid. The specific activity of purified recombinant PS-PLA 1 was 2.8 nmol/min⅐g of protein for PS. PS-PLA 1 belongs to the lipase family, and alignment of amino acid sequences of PS-PLA 1 with other members of the lipase family, such as lipoprotein lipase and hepatic lipase, indicated that Ser 166 is important for the activity of the enzyme (11). Catalytically inactive mutant PS-PLA 1 , in which the active serine residue is replaced with an Ala residue, was produced as follows: two oligonucleotides, 5Ј-GGGGGGAATTCATGTGTCCTGGCCTCTGGGGGACA-3Ј (nucleotide positions 124 of rat PS-PLA 1 cDNA (11)) and 5Ј-GAGCCCCCA-GAGCGACACCAATGA-3Ј (nucleotide positions 485-508 of rat PS-PLA 1 cDNA), were used as primers in a polymerase chain reaction (PCR) to introduce a mutation. The amplified PCR product and oligo-nucleotide, CCCCCAAGCTTCTACACACAGGCCATTTTCAGGTC (nucleotide positions 1348 -1371 of rat PS-PLA 1 ), were then used as primers for a second PCR. The resulting PCR product was introduced into EcoRI/HindIII sites of the baculovirus transfer vector pFASTBAC1 plasmid (Invitrogen), and production of recombinant baculovirus was performed according to the manufacturer's instructions (Bac-to-Bac system). The mutant PS-PLA 1 proteins were expressed and purified as described above. sPLA 2 -IIA protein was purified from culture medium of activated rat platelets as described previously (17). The specific activity of purified sPLA 2 -IIA was 32 nmol/min⅐g of protein.
Monoclonal Antibodies against Rat PS-PLA 1 -Rat PS-PLA 1 cDNA (encoding amino acids Val 27 -Val 456 ) was ligated into the BamHI/Hin-dIII sites of the pET21c vector (pET system, Novagen). After the plasmid had been introduced into Escherichia coli strain BL21 (DE3) (Novagen), protein expression was induced with 1 mM isopropyl-␤-Dthiogalactopyranoside. Rat PS-PLA 1 polypeptides were recovered in the insoluble fraction. The cell pellet was sonicated for 20 min on ice using a tip-type sonicator. After sonication, protein was recovered by ultracentrifugation at 100,000 ϫ g. The insoluble pellet was dissolved in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer and subjected to 10% acrylamide SDS-PAGE. After staining with Coomassie Brilliant Blue, the corresponding protein band was excised and mixed with Freund's adjuvant. BALB/c mice were immunized with the protein, and monoclonal antibodies were produced using the PAI myeloma cell line. Eight monoclonal antibody-producing hybridoma cell lines (clones 7F10, 8H7, 10G4, 12B12, 12D9, 15D12, 16F1, and 20D4) were established. In this study, monoclonal antibody from clone 15D12 (mouse IgG 1 ) was used for Western blotting.
Western Blotting-Protein samples were separated by SDS-PAGE and transferred to nitrocellulose filters using the Bio-Rad protein transfer system. The filters were blocked with Tris-buffered saline containing 5% (w/v) skimmed milk and 0.05% (v/v) Tween 20, incubated with anti-rat PS-PLA 1 monoclonal antibodies (ascites prepared from clone 15D12, diluted 1:1000) in Tris-buffered saline containing 5% skimmed milk and 0.05% Tween 20, and then treated with anti-mouse IgG-horseradish peroxidase. Proteins bound to the antibodies were visualized with an enhanced chemiluminescence kit (ECL, Amersham Pharmacia Biotech).
Lipid Preparation-Apoptotic human Jurkat T cells (2.3 ϫ 10 8 cells) or rat peritoneal cells (2.4 ϫ 10 8 cells in the presence of ConA (100 g/ml)) were incubated with serum-free RPMI 1640 medium in the presence or absence of recombinant PS-PLA 1 (1.8 g/ml) for 30 min at 37°C. Phospholipids in the cell supernatants were extracted by the MS Analysis-Lipid extracts from the cell culture supernatant were analyzed with a Quattro II mass spectrometer (Micromass, Manchester, United Kingdom) equipped with an electrospray ion source (ESI) as described previously (27). Aliquots (2 l) of samples (100 -200 pmol/l) dissolved in chloroform/methanol (2:1) were introduced by means of a flow injector into the ESI chamber at a flow rate of 2 l/min. The eluting solvent used was acetonitrile/methanol/water (2:3:1) containing 0.1% ammonium formate, pH 6.4. The mass spectrometer was operated in the positive and negative ion modes. Nitrogen gas at a temperature of 80°C and a flow rate of 12 liter/min was used for drying. In most cases, the capillary voltage was set at 3.7 kV, and the cone voltage was set at 30 V, both in the positive and negative ion modes. For MS/MS experi- ments, 3-4 ϫ 10 Ϫ4 torr of argon was used as the collision gas, and a collision energy of 30 -40 V was used for obtaining fragment ions for precursor ions.

2-Acyl-1-lyso-PS Enhances Antigen-dependent Activation of RPMC-
The ability of 2-acyl-1-lyso-PS to stimulate antigendependent mast cell activation was determined. Purified RPMC primed with anti-DNP monoclonal IgE were incubated with DNP-As in the presence of lyso-PS. Like 1-oleoyl-2-lyso-PS, 2-oleoyl-1-lyso-PS was able to stimulate histamine release from IgE-antigen-stimulated mast cells (Fig. 1A). The dose-dependent curve showed that 2-oleoyl-1-lyso-PS was as potent as 1-oleoyl-2-lyso-PS at stimulating histamine release. Neither lyso-PS was able to enhance histamine release in the absence of antigen. Similar results were obtained when ConA was used instead of IgE-antigen complexes (Fig. 1B). 2-Acyl-1-lyso-PS and 1-acyl-2-lyso-PS were also prepared from bovine brainderived PS, with different acyl chains at the sn-1 and sn-2 positions. The activity of these lyso-PS was almost identical to that of 2-oleoyl-1-lyso-PS (data not shown), indicating that the fatty acid moiety does not affect the activity of lyso-PS. It is generally accepted that the acyl chain at the sn-2 position of 2-acyl-lysophospholipids easily migrates to the sn-1 position, resulting in production of 1-acyl-lysophospholipids (26). To determine whether 2-acyl-1-lyso-PS is really active against mast cells, ConA-primed mast cells were incubated with PS liposomes for 15 min in the presence or absence of PS-PLA 1 , and histamine release from the cells was determined. PS is known to stimulate mast cell activation at a concentration above 2 ϫ 10 Ϫ5 M. However, at lower concentrations, PS alone was unable to stimulate histamine release from ConA-primed RPMC. As shown in Fig. 2, PS stimulates histamine release from RPMC only in the presence of recombinant PS-PLA 1 in a dose-dependent manner. This shows that the 2-acyl-1-lyso-PS produced from PS by PS-PLA 1 is active against RPMC, since the migration of the acyl chain takes several hours (26).
PS-PLA 1 Enhances Histamine Release from RPMC-Next we examined the effect of recombinant PS-PLA 1 on histamine release from IgE-antigen-stimulated mast cells in the absence of PS liposomes. As shown in Fig. 3, when the purified RPMC, which were pretreated with anti-DNP monoclonal IgE, were incubated with DNP-As in the presence of recombinant PS-PLA 1 (5 g/ml), the amount of histamine released increased. A similar result was obtained when ConA was used instead of IgE-antigen for Fc⑀RI cross-linking (Fig. 3). PS-PLA 1 alone did not show such an activity. This result suggests that lyso-PS is produced from PS on the RPMC membrane by PS-PLA 1 . When we analyzed the crude RPMC, we found that they contain many cells other than mast cells (see "Experimental Procedures"). Thus, it is possible that lyso-PS is produced from PS not only on mast cells but also on the other cells that reside near the mast cells. To test this possibility, we first used co-culture system in which purified RPMC were mixed with apoptotic cells. PS is located mainly in the inner leaflet of lipid bilayers in normal cells. Thus, the accessibility of PS-PLA 1 to the substrate on the plasma membrane seems to be limited. It is generally accepted that PS appears on the cell surface in apoptotic cells or cytokine-activated cells. As shown in Fig. 4C, exposure of PS is limited in normal Jurkat cells, but it is evident in apoptotic (anti-Fas antibody-treated) Jurkat cells. When the purified RPMC were mixed with apoptotic Jurkat cells, the ability of PS-PLA 1 to stimulate histamine release from the mast cells increased significantly (Fig. 4A). These results clearly indicate that PS-PLA 1 hydrolyzes PS exposed on apoptotic cells and produces lyso-PS. Next, we used crude RPMC instead of purified RPMC and subjected them to a histamine assay in the presence of ConA and PS-PLA 1 . Exposure of PS was limited in crude RPMC population but was enhanced after cells were treated with ConA (Fig. 4D). As shown in Fig. 4B, when the crude mast cells were used, the ability of PS-PLA 1 to stimulate ConA-induced histamine release from RPMC increased significantly. The crude mast cell preparation consisted mainly of mononuclear leukocytes (macrophages and monocytes) with a minor population of other cells such as polymorphonuclear leukocytes (neutrophils) (see "Experimental Procedures"), and together they were about 20 times more numerous than RPMC. Thus again in this system it was shown that PS-PLA 1 efficiently stimulates Fc⑀RI-dependent histamine release from RPMC in the presence of PS-exposing cells. It was also shown that a catalytically inactive mutant PS-PLA 1 has no enzyme activity (Fig. 5A) and is unable to stimulate histamine release from RPMC (Fig. 5B). This result confirmed that catalytic activity of PS-PLA 1 is required to enhance histamine release from RPMC. All these data strongly indicate that PS-PLA 1 can hydrolyze PS efficiently when it appears on the surface of apoptotic cells and that lyso-PS is actually derived from cells surrounding mast cells in vivo.
Detection of Lyso-PS Produced by PS-PLA 1 -We further examined whether lyso-PS itself is actually produced by PS-PLA 1 from the cell membranes. As shown in Fig. 6B, the ability of the cell supernatant from the apoptotic cells to stimulate ConA-induced histamine release from RPMC was greatly enhanced when the apoptotic cells or crude RPMC were treated with PS-PLA 1 . From the standard curve of lyso-PS (Fig. 6A), it is estimated that ϳ8 nmol or the equivalent of lyso-PS was recovered from 2.3 ϫ 10 8 cells. This corresponds to 4% of the total PS of the cells (data not shown). Lipids in the cell supernatant were further analyzed by ESI-MS. In good agreement with the activity data (Fig. 6B), signals with m/z values of 522.4, and 494.3, which correspond to lyso-PS with oleic acid (18:1, m/z 522.4), and palmitoleic acid (16:1, m/z 494.3), respectively, were detected in the lipid fraction from supernatant of apoptotic Jurkat T cells (Fig. 6C) and crude RPMC (ConA-treated, Fig. 6G) incubated with PS-PLA 1 . The identity of the peak (m/z 522.4) as 18:1lyso-PS was confirmed by MS/MS analysis of the daughter ions (data not shown). The signals were not detected in the absence of PS-PLA 1 treatment (Fig. 6, D and H). Interestingly the lipids were detected almost exclusively in the cell supernatant and not in the cell fraction (Fig. 6, E and I). It can be concluded that the lyso-PS detected is 2-acyl-1-lyso-PS, since acyl chains at the sn-2 position of phospholipids are rich in unsaturated fatty acids.
Heparin Blocks the Effect of PS-PLA 1 -Like other lipases, PS-PLA 1 has an affinity for heparin (11). We next examined whether such affinity for heparin is important for mast cell activation by PS-PLA 1 . As shown in Fig. 7, 50 ng/ml of heparin completely blocked the histamine release from crude RPMC. However, heparin did not inhibit the ability of PS-PLA 1 to hydrolyze PS in vitro nor lyso-PS-induced histamine release from RPMC (data not shown). This result indicates that an association between PS-PLA 1 and the cell surface membrane via heparan sulfate proteoglycan is important for the cellular function of the enzyme.
PS-PLA 1 Stimulates Mast Cell Activation More Efficiently than sPLA 2 -IIA-We finally examined the ability of sPLA 2 -IIA to stimulate RPMC activation and compared it with that of PS-PLA 1 , since sPLA 2 -IIA is capable of producing lyso-PS in vitro. sPLA 2 -IIA efficiently hydrolyzes anionic phospholipids such as phosphatidylethanolamine and PS in vitro, and it does not hydrolyze phospholipids on intact membranes. When the same amounts of each enzyme were separately added to crude RPMC in the presence of Fc⑀RI cross-linker, PS-PLA 1 was much more potent than sPLA 2 -IIA in inducing histamine release (Fig. 8A). A similar result was obtained when purified RPMC were co-cultured with apoptotic Jurkat cells in the presence of Fc⑀RI cross-linker and PS-PLA 1 , although at high concentrations, sPLA 2 -IIA significantly stimulated histamine release (Fig. 8B).

DISCUSSION
The present study was undertaken to determine whether PS-PLA 1 participates in the activation of RPMC. Three results of this study indicate that PS-PLA 1 stimulates histamine release from RPMC through the production of 2-acyl-1-lyso-PS: 1) 1-acyl-2-lyso-PS has almost the same ability as 2-acyl-1lyso-PS to stimulate IgE-antigen-induced histamine release from RPMC; 2) recombinant PS-PLA 1 , but not mutant PS-PLA 1 , stimulated IgE-antigen-induced histamine release from RPMC; and 3) lyso-PS with unsaturated fatty acids was detected by MS analysis after apoptotic cells or crude RPMC were incubated with recombinant PS-PLA 1 . In addition, the present results, summarized below, provide some insights into how lyso-PS production by the enzyme is regulated and how the enzyme acts on cell membranes.
Lyso-PS Production by PS-PLA 1 and Its Regulation-PS-PLA 1 was effective in stimulating histamine release from purified RPMC co-cultured with apoptotic Jurkat T cells (Fig. 4A) or from crude RPMC (Fig. 4B). In addition, lyso-PS with unsaturated fatty acids was actually detected in the cell supernatant by mass spectrometry. These results clearly show that PS-PLA 1 produces lyso-PS by hydrolyzing PS exposed on the surface of peritoneal cells or apoptotic cells, although, like other phospholipases, it does not act on intact cell membranes. They also suggest that a key factor regulating lyso-PS production is the availability of PS on the cell surface. It is well accepted that apoptotic cells expose PS on the surface. In addition, PS exposure is enhanced in activated platelets and cytokine-stimulated or ConA-treated cells (this study). Thus it can be concluded that PS exposed on such cells may be a target for PS-PLA 1 in vivo. In fact, we previously detected lyso-PS with unsaturated fatty acids in activated rat platelets, which abundantly express PS-PLA 1 (13). We also observed that the activated rat platelets stimulated IgE-antigen-or nerve growth factor-induced histamine release from RPMC. 2 Lyso-PS has also been reported to be present in vivo: it has been detected in the aqueous humor of the eyes (28) and is produced by rat peritoneal cells or at the site of tissue injury (10). It is possible that PS-PLA 1 is responsible for the production of such lyso-PS in vivo. In this study, we showed that lyso-PS, produced by PS-PLA 1 , was almost exclusively recovered from cell supernatant. In our previous work we have found that lyso-PS applied to cells is immediately converted to PS by acylation (5). The result may explain why lyso-PS was not detected in the cell fractions.
Another factor regulating lyso-PS production is the expression of PS-PLA 1 in vivo. We previously showed that rat platelets are a major source of PS-PLA 1 (11) and that PS-PLA 1 is detected in various human tissues (14). In addition, preliminary experiments in our laboratory have shown that expression of PS-PLA 1 can be induced in inflammatory sites and in several tissues by various inflammatory stimuli. 3 Our preliminary results show that the concentrations of PS-PLA 1 in the supernatant of activated rat platelets, rat plasma, and the peritoneal cavity of rats injected intraperitoneally with casein are 400, 5, and 150 ng/ml, respectively. Thus the amount of PS-PLA 1 used in this study is close to the physiological levels, and it is possible that the enzyme stimulates mast cells in vivo.
We showed that PS-PLA 1 , which has an affinity for heparin, completely lost its ability to stimulate mast cells in the presence of heparin (Fig. 7). Many lipases, such as hepatic lipase and lipoprotein lipase, have an affinity for heparin and function on the cell surface of hepatocyte or adipocyte by interacting with heparan sulfate proteoglycan. It has been also demonstrated that sPLA 2 -IIA is attached to the cell surfaces by its C-terminal heparin-binding domain and that the attachment is essential for prostaglandin biosynthesis (29). This may be also the case with PS-PLA 1 .
PS-PLA 1 Produces Lyso-PS More Efficiently than sPLA 2 -IIA-From the data presented here, we assume that sPLA 2 -IIA 2 K. Kawamoto, J. Aoki, A. Tanaka, A. Itakura, Y. Kiso, and H. Matsuda, manuscript in preparation. 3 Y. Nagai, J. Aoki, H. Arai, and K. Inoue, unpublished data. is less potent than PS-PLA 1 in producing lyso-PS (Fig. 8). Living cells are normally resistant but become susceptible during apoptosis to the enzyme. Unlike PS-PLA 1 , sPLA 2 -IIA acts on other phospholipids than PS. This is a possible reason why sPLA 2 -IIA is less potent than PS-PLA 1 in producing lyso-PS. sPLA 2 -IIA alone has been reported to stimulate histamine release from RPMC provided that the concentration of the enzyme is high enough (30). Indeed, a concentration of sPLA 2 -IIA Ͼ20 g/ml was required to induce histamine release from RPMC by sPLA 2 -IIA alone. Conversely, the effect of PS-PLA 1 was observed at a concentration of about 50 ng/ml in the presence of Fc⑀RI cross-linker (Fig. 8). The effect of sPLA 2 -IIA has also been reported to be independent of lyso-PS, because histamine release triggered by sPLA 2 -IIA can be observed in the absence of Fc⑀RI cross-linkers (30). Thus, the two PLAs act against mast cells in a very different manner, although we cannot exclude the involvement of another secretory PLA 2 that has recently been identified (31).
In summary, PS-PLA 1 was found to stimulate histamine release from RPMC. PS is normally present on the inner leaflet of lipid bilayers, and there is evidence to suggest that PS is exposed to the outer surface of apoptotic cells, dead cells, and cytokine-stimulated cells. PS-PLA 1 hydrolyzes PS on the outer membrane of such cells producing 2-acyl-1-lyso-PS, a lipid messenger of mast cells. The synthetic pathways for lipid mediators such as lysophosphatidic acid, platelet-activating factor, and sphingosine-1-phosphate have not been fully solved yet. In this sense it can be said that PS-PLA 1 is the first enzyme that has been shown to produce bioactive lysophospholipid, although further studies are definitely required to demonstrate the role of this enzyme in vivo.