(Received for publication, October 17, 1996)
From the Departments of Biological Chemistry and Molecular and Medical Pharmacology, and the Molecular Biology Institute, UCLA Center for the Health Sciences, Los Angeles, California 90095
Aggregation of IgE cell surface receptors on MMC-34 cells, a murine mast cell line, induces the synthesis and secretion of prostaglandin D2 (PGD2). Synthesis and secretion of PGD2 in activated MMC-34 cells occurs in two stages, an early phase that is complete within 30 min after activation and a late phase that reaches a maximum about 6 h after activation. The early and late phases of PGD2 generation are mediated by prostaglandin synthase 1 (PGS1) and prostaglandin synthase 2 (PGS2), respectively. Arachidonic acid, the substrate for both PGS1 and PGS2, is released from membrane phospholipids by the activation of phospholipases. We now demonstrate that in activated mast cells (i) secretory phospholipase A2 (PLA2) mediates the release of arachidonic acid for early, PGS1-dependent synthesis of PGD2; (ii) secretory PLA2 does not play a role in the late, PGS2-dependent synthesis of PGD2; (iii) cytoplasmic PLA2 mediates the release of arachidonic acid for late, PGS2-dependent synthesis of PGD2; and (iv) a cytoplasmic PLA2-dependent step precedes secretory PLA2 activation and is necessary for optimal PGD2 production by the secretory PLA2/PGS1-dependent early pathway.
Mast cells, an important cell type in allergic diseases, are widely distributed throughout vascularized tissue and epithelia. Activation of mast cells by aggregation of high affinity IgE receptors causes degranulation, releasing stored mediators of inflammation such as histamine and serotonin. Mast cell activation also induces the synthesis and release of leukotrienes and prostaglandin D2 (PGD2). Ligand stimulation in most cells elicits a relatively slow production of prostaglandins, peaking only after 4-6 h (1, 2). In contrast, PGD2 synthesis in activated mast cells occurs in two stages, a rapid, early phase and a late, delayed phase (3, 4).
Prostaglandin production is regulated by both phospholipases A2 (PLA2)1 and prostaglandin synthases (PGS). PLA2 enzymes release arachidonic acid from membrane phospholipids. Free arachidonate is converted to prostaglandin H2 (PGH2), a common precursor for all prostanoids, by prostaglandin synthases. Several PLA2 enzymes have been implicated in arachidonate release following ligand stimulation of various cell types (5, 6). Many cells also express two distinct prostaglandin synthases: PGS1, a primarily constitutively expressed form, and PGS2, an inducible PGS expressed following appropriate ligand stimulation in different cell types (1, 2). Experiments using antisense oligonucleotide inhibition (7) and NS-398, a PGS2-specific inhibitor (1, 8), demonstrated that ligand-induced prostaglandin production in fibroblasts and macrophages requires induced PGS2 expression, despite the presence of active PGS1 enzyme. In contrast, the rapid, early phase of PGD2 synthesis in activated mast cells is mediated by pre-existing PGS1 (3, 4). The second, delayed phase of PGD2 synthesis in activated mast cells is similar to prostaglandin production in growth factor-induced fibroblasts and endotoxin-induced macrophages, requiring activation-induced PGS2 expression (3, 4).
Following activation, mast cells secrete a low molecular weight PLA2, sPLA2 (9). Fonteh et al. (9) suggest that the sPLA2 released following activation plays a role in eicosanoid biosynthesis by activated mast cells. Activation of mast cells also induces the activation, translocation, and expression of cytoplasmic cPLA2 (10-13). In this report we investigate the roles of sPLA2 and cPLA2 in early, PGS1-dependent PGD2 synthesis and late, PGS2-dependent PGD2 synthesis following mast cell activation by aggregation of high affinity IgE receptors.
Mouse MMC-34 cells (14) were grown in RPMI 1640 medium (ICN, Cleveland, OH) supplemented with 10% fetal calf serum (Gemini Bioproducts Inc., Calabasas, CA).
ReagentsMurine IgE and monoclonal anti-IgE were purchased
from Pharmingen (San Diego, CA). PGD2 assay kits were from
Amersham Corp. (UK); aminopropyl solid phase columns No. 9070 (100 mg/ml) were from Burdick and Jackson (Muskegon, MI).
[3H]Arachidonate-labeled Escherichia coli
suspension was from DuPont NEN. Methyl arachidonylfluorophosphonate
(MAFP) was obtained from Cayman Chemical Co. (Ann Arbor, MI). The
cytosolic phospholipase A2 assay kit used for the data
shown in Fig. 10, right panel, was also from Cayman. NS-398
was a gift from Taisho Corp (Japan). Monoclonal antibody F10 (mAbF10)
directed against recombinant PLA2 (9), recombinant
sPLA2, and SB203347, a specific inhibitor of
sPLA2 (15), were the gifts of Dr. Lisa Marshall (SmithKline Beecham Pharmaceuticals, King of Prussia, PA). Recombinant
cPLA2 was the gift of Dr. Michael H. Gelb (University of
Washington, Seattle).
Prostaglandin Determinations
Cell cultures were treated with IgE and anti IgE as described in figure legends and text. Medium was collected by centrifugation and analyzed for PGD2 using the Amersham kit.
Assay of PLA2 Enzymatic ActivityFor the assay
of cPLA2 activity from extracts of activated mast cells
(Fig. 10, right panel), we used the cPLA2 kit
from Cayman Chemical Co., according to the manufacturer's
instructions. Briefly, samples of mast cell extracts (see legend to
Fig. 10 for details) were incubated with
arachidonylthiophosphatidylcholine (ATPC), the substrate. Enzymatic
hydrolysis of ATPC releases free thiol, which is then converted into
5-thio-2-nitrobenzoic acid by Ellman's reagent
[5,5-dithiobis(2-nitrobenzoic acid)]. 5-Thio-2-nitrobenzoic acid
concentration is determined by spectrophotometric analysis at 412 nm.
All other phospholipase assays were performed as described previously
(16, 17). Briefly, supernatants were incubated in a reaction mixture
(150 µl) containing 25 mM HEPES, pH 7.4, 150 mM NaCl, 5 mM CaCl2, and 0.1 µCi
of [3H]arachidonate-labeled E. coli membranes
for 1 h at 37 °C. Free arachidonic acid was separated by
elution of the sample over aminopropyl solid phase silica columns and
quantitated by scintillation counting.
The early phase of PGD2 synthesis, completed
within the first 10-30 min after mast cell activation, is mediated by
PGS1 (3, 4). The late phase of PGD2 production, mediated by
PGS2, does not peak until 4-6 h after activation (3, 4). We began our studies on the relationships between phospholipases and prostaglandin synthases in mast cells by analyzing the effect of inhibiting sPLA2 activity on PGD2 production after
activation by aggregation of IgE receptors. If mast cells are treated
with the sPLA2 inhibitor SB203347 (15), the early burst of
PGD2 synthesis is completely blocked (Fig.
1), suggesting that sPLA2 mediates the early
burst of PGD2 production in response to activation by
cross-linking of IgE receptors.
It is possible that the inhibitory effects of SB203347 described above
might be due to inactivation of PGS1, since this enzyme is necessary
for the early burst of PGD2 production following mast cell
activation (3, 4). To examine this question, we asked whether mast
cells activated in the presence of SB203347 can utilize exogenous
arachidonic acid as substrate. Cells activated in the presence of
SB203347, although unable to produce the early burst of prostaglandin
from endogenous arachidonate, are capable of converting exogenous
arachidonate to PGD2 (Fig. 2). The PGS1 present in these cells is, therefore, not inactivated and is
enzymatically active in the presence of SB203347. The data in Fig. 1
and Fig. 2 suggest that SB203347 blocks the early burst of
PGD2 synthesis in activated mast cells by preventing
sPLA2 from mobilizing arachidonic acid from phospholipids
and providing substrate for PGS1.
As an alternative means of demonstrating the role of sPLA2
in prostaglandin synthesis in activated mast cells, we examined the
effect of a monoclonal antibody to this enzyme on PGD2
production. In this experiment NS-398, a specific inhibitor of PGS2
(18), was added to prevent late phase, PGS2-dependent
PGD2 production. Treatment of MMC-34 mast cells with either
(i) a monoclonal antibody, mAbF10, against sPLA2 (9) or
(ii) the specific sPLA2 inhibitor, SB203347, blocked
activation-induced PGD2 production (Fig. 3, left panel), again suggesting that early,
PGS1-dependent PGD2 production in activated
mast cells requires sPLA2.
To confirm that the inhibition of PGD2 synthesis in response to either SB203347 or mAbF10 is due to a reduction in sPLA2 activity, the culture media from this experiment were also analyzed for secreted phospholipase activity (Fig. 3, right panel). Activated MMC-34 cells accumulated increased PLA2 activity in the medium. Accumulation of this PLA2 activity, like PGD2 accumulation, was blocked by treatment with the sPLA2 inhibitors mAbF10 or SB203347. We conclude that the sPLA2 released by activated mast cells is required for early, PGS1-dependent PGD2 synthesis.
Secretory PLA2 Does Not Mediate Late, PGS2-dependent PGD2 Synthesis in Activated Mast CellsThe late phase of PGD2 synthesis in activated mast cells appears to be unaffected by the sPLA2 inhibitor SB203347 at 1 µM concentration (Fig. 1). To more accurately assess the level of late phase, PGS2-dependent PGD2 accumulation in SB203347-treated MMC-34 cells following activation, we included in this experiment (Fig. 1) a set of cells in which PGS1 activity was eliminated by preincubation with aspirin. We have shown that aspirin preincubation and washing irreversibly inhibits both PGS1 activity and the early phase of PGD2 synthesis in MMC-34 cells, without interfering with either (IgE + anti-IgE)-induced PGS2 induction or the PGS2-dependent late phase of PGD2 production (3). PGD2 production in activated mast cells in which sPLA2 activity was inhibited by SB203347 is identical to that observed in cells in which the PGS1 activity was blocked by aspirin preincubation (Fig. 1), suggesting that sPLA2, like PGS1 (3), plays no role in the late phase of PGD2 synthesis in activated mast cells.
To demonstrate that the late burst of PGD2 synthesis
occurring in the presence of SB203347 was not due to insufficient
levels of inhibitor, we carried out an SB203347 dose-response analysis in which we measured PGD2 production at 6 h. Higher
concentrations of SB203347 were unable to further inhibit
PGD2 accumulation (Fig. 4), confirming that
the late phase of PGD2 synthesis in activated mast cells
does not require sPLA2 activity.
To extend these observations, we demonstrated that inhibiting
sPLA2 activity with either sPLA2-directed
monoclonal antibody mAbF10 or the sPLA2 inhibitor SB203347
does not increase the inhibition of total (i.e. early phase
and late phase) PGD2 production in activated mast cells
beyond that seen by aspirin pretreatment alone (Fig. 5,
lanes 3-5). Inhibition of PGS1 alone (lane 3) or inhibition of both PGS1 and sPLA2 (lanes 4 and
5) result in the same reduction in total PGD2
production at 6 h by activated MMC-34 cells. We conclude that
sPLA2 activity, like PGS1 activity, is required for the
early phase of PGD2 production in activated mast cells but
not for the late phase of PGD2 production.
The late phase of PGD2 production following IgE + anti-IgE activation of MMC-34 cells is mediated entirely by PGS2 and can be completely suppressed by the PGS2-specific nonsteroidal anti-inflammatory drug NS-398 (3). In contrast, NS-398 has no effect on the PGS1-dependent early phase of PGD2 production in activated mast cells (3). Although NS-398 treatment blocks only a portion (i.e. the late phase) of total PGD2 production in activated MMC-34 cells (Fig. 5, lane 6), the combination of either SB203347 or mAbF10, both of which inhibit the early phase of PGD2 production, with NS-398 can completely suppress mast cell PGD2 production (Fig. 5, lanes 7 and 8).
MAFP, a cPLA2 Inhibitor, Blocks Late PGD2 Production in Activated Mast CellsOur data suggest that
sPLA2 does not play a role in late,
PGS2-dependent PGD2 production in activated
mast cells (Figs. 1, 4, and 5). Another phospholipase must, therefore,
provide arachidonate for the late phase of PGD2 production.
The enzyme most likely to fulfill this role is the type IV cytoplasmic
PLA2, or cPLA2, enzyme (see "Discussion").
MAFP has been reported to preferentially inhibit cPLA2 and
to have relatively little effect on sPLA2 (19). We also
find that MAFP preferentially inhibits recombinant cPLA2, whereas SB203347 preferentially inhibits recombinant sPLA2
(Fig. 6). We used MAFP to determine whether
cPLA2 might play a role in either early
PGS1-dependent PGD2 synthesis or late
PGS2-dependent PGD2 synthesis following mast
cell activation.
When 10 µM MAFP is added at the same time that IgE
receptors are aggregated on MMC-34 cells, the early phase of
PGD2 accumulation is not substantially inhibited (Fig.
7, left panel). In contrast, the late phase
of PGD2 accumulation is suppressed to an extent similar to
that observed with NS-398, the PGS2-specific inhibitor. These data
suggest that the late PGS2-dependent phase of
PGD2 production in activated mast cells requires
cPLA2 activity to provide arachidonic acid as substrate for
PGS2. To demonstrate that the degree of inhibition of late phase
PGD2 production following activation is maximal at the 10 µM concentration of MAFP used, we carried out a
dose-response curve for MAFP (Fig. 7, right panel). Even at
a concentration 5-fold greater than that used in the time course
experiment, no additional inhibition of late phase PGD2 accumulation by MAFP was observed.
MAFP might exert its effect on the late phase of PGD2
accumulation in part by inhibiting the induction and/or enzymatic
activity of PGS2. To investigate this question, we examined the ability of MAFP-treated, activated MMC-34 cells to utilize exogenous
arachidonic acid as substrate for PGD2 synthesis. Untreated
MMC-34 cells do not produce PGD2 from endogenous
arachidonic acid but can convert exogenous arachidonate to
PGD2 because they express constitutive PGS1 (Fig.
8). Treatment with aspirin inactivates PGS1 and prevents unstimulated MMC-34 cells from converting exogenous arachidonate to
PGD2 (Fig. 8). If MMC-34 cells are first pretreated with
aspirin, then activated by aggregation of IgE receptors, they produce
PGD2 from endogenous arachidonic acid during a 6-h
incubation, as a result of the activity of cPLA2 and the
induction of PGS2 (Fig. 8, left panel). MAFP can completely
prevent this PGD2 production from endogenous sources of
arachidonic acid in cells activated following pretreatment with aspirin
(Fig. 8, left panel). However, these same cells are able to
produce PGD2 from exogenous arachidonate, demonstrating
that induction of functional PGS2 occurs in the presence of MAFP in
response to mast cell activation (Fig. 8, right panel). Thus
MAFP inhibition of the late, PGS2-dependent component of
PGD2 production in activated mast cells is not due to
either inhibition of PGS2 production or to inhibition of PGS2 enzyme
activity.
cPLA2 Plays a Role in the Early Phase of PGD2 Production in Activated Mast Cells
Balsinde and
Dennis (20) have recently provided compelling evidence that
cPLA2 activation is necessary for, and precedes, sPLA2 activation in murine macrophages stimulated with
endotoxin and platelet activating factor. In these experiments, MAFP
treatment was able to block a substantial proportion of the
sPLA2-dependent production of arachidonic acid
in macrophages (20). However, this inhibition required preincubation of
cells with MAFP prior to ligand activation. To investigate the role of
cPLA2 in the early sPLA2-dependent
phase of PGD2 production as well as in the late
sPLA2-independent phase of PGD2 production in
activated mast cells, we first compared the effect of (i) MAFP
preincubation on total PGD2 production in activated mast
cells with (ii) PGD2 production when MAFP was added at the
time of aggregation of IgE receptors. If MAFP is added at the time of
activation, the inhibition of PGD2 accumulation at 6 h
is identical to that observed with NS-398, the inhibitor of PGS2
activity (Fig. 9, top panel, lanes 4 and
5). These data are consistent with our conclusions that cPLA2 and PGS2 mediate the late phase of PGD2
production in activated mast cells. However, if MMC-34 cells are
preincubated with MAFP prior to IgE receptor aggregation, an additional
increment of PGD2 accumulation is inhibited (Fig. 9,
top panel, lane 6), suggesting that cPLA2 may
also play a role in the early
sPLA2/PGS1-dependent phase of PGD2
production in activated mast cells.
To more completely characterize the role of cPLA2 on the early sPLA2/PGS1-dependent phase of PGD2 production in activated mast cells, we examined the time- and dose-dependent influence of MAFP. Maximal inhibition of the early phase of PGD2 production in activated mast cells occurred at a concentration of 10 µM MAFP (Fig. 9, middle panel) and was complete within 15 min of preincubation (Fig. 9, lower panel). These data, like those of Balsinde and Dennis (20), suggest that cPLA2 activation must precede sPLA2 activation for optimal early phase PGD2 production in response to mast cell IgE receptor aggregation.
SB203347 and MAFP Inhibit Distinct Components of Phospholipase A2 Activity in Activated Mast CellsWe analyzed the abilities of the sPLA2 inhibitor, SB203347, and the cPLA2 inhibitor, MAFP, to inhibit the phospholipase activity present in the culture media and in the homogenates of activated MMC-34 cells. The bulk of the phospholipase activity present in the media from activated mast cells is inhibited by SB203347. Little or no phospholipase activity secreted into the culture media following activation is inhibitable by MAFP (Fig. 10, left panel). In contrast, the bulk of the phospholipase activity in extracts prepared from activated mast cells is inhibited by MAFP. None of the phospholipase activity remaining associated with cells after activation is inhibited by SB203347 (Fig. 10, right panel).
MAFP can also inhibit iPLA2, the calcium-independent phospholipase A2 (21). To determine whether iPLA2 activity is present in these mast cell extracts, we examined the level of phospholipase activity in the extracts in the presence of EGTA. Greater than 80% of the MAFP-inhibitable phospholipase A2 activity present in the activated mast cell extract is calcium-dependent (Fig. 10, right panel; compare lanes 2 and 4).
Aggregation of mast cell IgE receptors results in two phases of PGD2 production, an early burst completed within 10-30 min and a second phase peaking after 5-6 h (3, 4). Pharmacologic studies demonstrated that the early PGD2 burst is due solely to activity of pre-existing PGS1 enzyme, while delayed PGD2 production in activated mast cells requires induced PGS2 synthesis (3, 4). Arachidonic acid, released from membrane phospholipids by PLA2, is a substrate for both PGS enzymes. Mast cells contain at least two distinct arachidonic acid precursor pools, whose arachidonate products remain segregated after release from cellular phospholipids (9, 22, 23). Moreover, mast cells have at least three distinct PLA2 isoforms (6, 24). The temporal separation of PGD2 production by PGS1 and PGS2 in activated mast cells suggested that distinct phospholipid arachidonate pools and distinct phospholipases might provide arachidonate to the two prostaglandin synthase enzymes.
The Late Phase of PGD2 Production in Activated Mast CellsThe inability of either antibody to sPLA2 or SB203347, the sPLA2 inhibitor, to reduce PGS2-dependent PGD2 production demonstrates that sPLA2 plays no role in the late phase of PGD2 synthesis. MAFP was initially described as a cytoplasmic type IV cPLA2 inhibitor, with no effect on type II sPLA2 (19). The late, PGS2-dependent phase of PGD2 production in activated mast cells is completely blocked by MAFP, suggesting that cPLA2 is required to provide arachidonate for this second component of PGD2 production. MAFP, however, also inhibits a cytosolic, calcium-independent PLA2 (iPLA2) (21). When assayed previously, iPLA2 could not be detected in bone marrow-derived murine mast cells (25). In our experiments, we found only a small fraction of MAFP-inhibitable phospholipase A2 activity present in extracts of activated mast cells to be calcium-independent (Fig. 10). In addition, specific iPLA2 inhibition in stimulated P388D1 macrophages enhances, rather than inhibits, arachidonate release (20). MAFP inhibition of late, PGS2-dependent PGD2 production in activated mast cells is, therefore, likely to be due to type IV cPLA2 inactivation. cPLA2 and PGS2 appear to be metabolically coupled in activated mast cells for the delayed phase of PGD2 synthesis. However, conclusive proof of this hypothesis will require the specific suppression of cPLA2 synthesis by antisense cPLA2, development of pharmacologic agents that more specifically inhibit cPLA2, or mast cells derived from animals in which the cPLA2 gene has been disrupted.
The Early Phase of PGD2 Production in Activated Mast CellsPrevious experiments suggested that sPLA2 plays a role in PGD2 production following mast cell activation (9). The ability, demonstrated here, of both a monoclonal antibody to sPLA2 (9) and a specific sPLA2 inhibitor, SB203347 (15), to completely block early, PGS1-dependent PGD2 production following mast cell activation supports this conclusion and demonstrates that sPLA2 and PGS1 are metabolically coupled. sPLA2 activity is required for the early phase of PGD2 production in activated mast cells. The low molecular weight, secreted PLA2 from mast cells has not been molecularly characterized. Three related murine genes encoding low molecular weight, secreted PLA2 enzymes of related structure are known, each with a cell type-specific distribution pattern (26, 27). It will be of great interest to identify the PLA2 isoform(s) secreted by activated mast cells, since this enzyme(s) may represent an important target in mast cell eicosanoid production.
Balsinde and Dennis (20) have demonstrated, in activated P388D1 macrophages, that "a functionally active cPLA2 appears to be necessary for sPLA2 to act." Their results prompted us to examine the effects of preincubation of mast cells with MAFP on the early, sPLA2/PGS1-dependent phase of PGD2 synthesis in activated mast cells. An MAFP-sensitive step is required for one-half to two-thirds of the early phase of PGD2 production following the activation of mast cells by aggregation of their IgE receptors (Fig. 9). Since we have verified the inability of MAFP to block the enzymatic activity of either recombinant sPLA2 (Fig. 6) or the sPLA2 activity present in mast cell supernatants (Fig. 10), we conclude that the MAFP-sensitive step in early PGD2 production in response to aggregation of mast cell IgE receptors is likely to be mediated by cPLA2. Like the data of Balsinde and Dennis (20), our results suggest that a cPLA2-mediated function is required prior to the action of sPLA2. It should be emphasized that, although there exists a cPLA2-dependent event required for the early phase of PGD2 synthesis in activated mast cells, cPLA2 does not release arachidonic acid that is then available for PGD2 production by PGS1. Inhibition of sPLA2 by either SB203347 or mAbF10, which do not inhibit cPLA2, prevents all PGD2 production in the early phase of mast cell activation (Figs. 1, 2, 3).
Spatial Separation of PGD2 Production in Mast CellsIn most cells, PGS1 is associated with the endoplasmic reticulum (1, 2). In contrast, biochemical and ultrastructural analysis suggested that mast cell granules contain phospholipids (28), sPLA2 (29), and PGS1 (30). Degranulation and exposure of granule contents to high extracellular calcium following aggregation of IgE receptors may activate sPLA2, leading to production of arachidonic acid substrate for PGS1 and the synthesis of the early PGD2 burst. Additional investigation of sPLA2 and PGS1 localization in resting and activated mast cells will be of great interest, now that two temporal phases of mast cell PGD2 production have been demonstrated.
PGS2 is detected in both the endoplasmic reticulum and the nuclear envelope of mitogen-stimulated fibroblasts (31). Subcellular PGS2 localization has not been reported in mast cells. Following mast cell activation, cPLA2 moves from the cytoplasm to the nuclear envelope (13). Ligand treatment also stimulates translocation of cPLA2 to the nuclear fraction in Chinese hamster ovary cells (32) and in macrophages (33). Assuming PGS2 is also localized predominantly at the nuclear envelope in mast cells after activation-induced synthesis, the data are consistent with the following model. Activated mast cells translocate cPLA2 to the nuclear envelope, where this enzyme releases arachidonate from phospholipids. Newly synthesized PGS2, also present in the nuclear envelope, utilizes this arachidonate to catalyze the late phase of PGD2 production. Mast cell PGS1 may not be able to use the arachidonate produced by cPLA2 for several reasons. It seems most likely that physical sequestration of PGS1 in mast cells may make cPLA2-dependent arachidonate, produced at the nuclear envelope, inaccessible to PGS1. Alternatively, PGS1 may be inactivated by a suicide reaction (34) during the early PGD2 synthesis phase, prior to adequate activation-induced cPLA2 translocation.
Identification of the PLA2 and PGS isoforms that mediate the two phases of PGD2 production in activated mast cells may permit development of additional pharmacologic agents that can discriminate between these two PGD2 production pathways. In addition to mediating PGS1-dependent PGD2 production by mast cells following activation, sPLA2 released by activated mast cells can also initiate trans-cellular prostaglandin production by mobilizing arachidonate in a distal cell as substrate for PGS1 in that cell (16). Specific inhibition of mast cell sPLA2 may, therefore, be an important pharmacologic target for several modes of prostaglandin production in chronic and acute inflammatory responses.
We thank Raymond Basconcillo and Arthur Catapang for technical assistance and the members of the Herschman lab for helpful discussions. We also thank Lisa Marshall (SmithKline Beecham) for the gifts of mAbF10, recombinant sPLA2, and SB203347 and Michael Gelb (University of Washington) for recombinant cPLA2.