Functional coupling between secretory phospholipase A₂ and cyclooxygenase-2 and its regulation by cytosolic group IV phospholipase A₂

Abstract

Secretory phospholipase A₂ (sPLA₂) is the major effector involved in arachidonic acid (AA) mobilization and prostaglandin E₂ (PGE₂) production during stimulation of P388D₁ macrophages with the inflammatory stimuli bacterial lipopolysaccharide and platelet-activating factor. We herein demonstrate that PGE₂ in stimulated P388D₁ cells is accounted for by the inducible cyclooxygenase (COX)-2. COX-1, though present, appears not to participate significantly in stimulus-induced PGE₂ production in P388D₁ macrophages. Reconstitution experiments utilizing exogenous recombinant sPLA₂ demonstrate that activation of the sPLA₂ at the plasma membrane is highly dependent on previous activation of the cytosolic phospholipase A₂ (cPLA₂). Collectively these results demonstrate (i) that functional coupling exists between sPLA₂ and COX-2 in activated cells, (ii) the critical role that cPLA₂ plays in lipid mediator production, and (iii) that there is crosstalk between cPLA₂ and sPLA₂ in the cell.

The phospholipase A₂ (PLA₂) superfamily is composed of a number of lipolytic enzymes whose common feature is to hydrolyze the fatty acid present in the sn-2 position of glycerophospholipids (1). At least three different PLA₂s exist in mammalian cells, namely a secretory PLA₂ (sPLA₂), the 85-kDa cytosolic group IV PLA₂ (cPLA₂), and a Ca²⁺-independent group VI PLA₂ (iPLA₂) (1, 2). Increasing evidence suggests that all three PLA₂ forms may participate in some manner in arachidonic acid (AA) turnover and mobilization with their relative contribution being dependent on the cell type and agonist used.

cPLA₂ has been shown to act on intracellular membranes—i.e., those of the perinuclear envelope and endoplasmic reticulum (3, 4). The plasma membrane of the cells appears to be the major site of action of the sPLA₂, which is consistent with the fact that sPLA₂ requires millimolar Ca²⁺ for activity (5–11). sPLA₂ appears to be responsible for the majority of free AA produced by stimulated P388D₁ cells (7, 12–14).

Once liberated from phospholipids, the AA is converted to prostaglandins by the cyclooxygenases (COXes). Two different COX enzymes have been identified in cells (15). The first one, COX-1, is constitutively expressed by most cell types, whereas the other one, COX-2, is generally induced after cell stimulation (15). As is the case for the PLA₂s, the relative contribution of each of these COX isoenzymes to overall prostaglandin synthesis appears to be strikingly dependent on cell type and stimulus used (16). However the question remains as to whether a specific coupling of each COX isoenzyme to a particular PLA₂ exists. Although supportive evidence for that view is emerging (15, 16), identification of the particular COX and PLA₂ isoenzymes involved has proven to be difficult and often contradictory. For example, Herschman and colleagues (11, 17–19) have suggested a coupling between sPLA₂ and COX-1 for the immediate response of mast cells; however, Arm and colleagues (20, 21) have suggested that the sPLA₂ is coupled to COX-2 in the same cells. In a recent report, Murakami et al. (22) have shown that the cPLA₂ couples to both COX-1 and COX-2 for the two phases of prostaglandin production in mouse osteoblasts.

There is a second important unanswered question regarding the cellular biochemistry of sPLA₂. This enzyme appears to act poorly against intact cellular membranes. Thus, some sort of physical perturbation or “membrane rearrangement” is required for sPLA₂ to act on a cellular membrane (5, 6). However, the biochemical signals that cause such a perturbation are largely unknown. Fourcade et al. (23) have suggested that loss of membrane asymmetry due to phospholipid flip-flop could be one of these signals. We have recently found that accumulation of membrane destabilizers such as diacylglycerol or ceramide may also serve to perturb plasma membrane packing and hence favor sPLA₂ action (24). More importantly, we found as well that increasing the intracellular levels of AA by exogenous addition of unlabeled fatty acid to [³H]AA-labeled cells (which mimics cPLA₂ activation) was able to trigger release of [³H]AA by a mechanism dependent on sPLA₂. These data suggested that cPLA₂ activation may be another key event that leads to perturbation of the membrane, rendering it susceptible to sPLA₂ attack (13).

In the current study, we have studied the expression of COX isoforms in P388D₁ cells, and we have found that a functional link exists between sPLA₂ and COX-2 for eliciting prostaglandin production in activated cells. Furthermore, reconstitution experiments using recombinant sPLA₂ further support the notion that crosstalk exists between the cPLA₂ and the sPLA₂ during stimulation of AA release in P388D₁ cells. Thus the cPLA₂ plays a dual role in arachidonate signaling as it (i) participates in an intermediate signaling step that is required for sPLA₂ activation and (ii) acts as a producer of AA itself.

EXPERIMENTAL PROCEDURES

Materials.

Mouse P388D₁ macrophage-like cells were obtained from the American Type Culture Collection. Iscove’s modified Dulbecco’s medium (endotoxin <0.05 ng/ml) was from BioWhittaker. Fetal bovine serum was from HyClone. Nonessential amino acids were from Irvine Scientific. [5,6,8,9,11,12,14,15-³H]AA (specific activity 100 Ci/mmol; 1 Ci = 37 GBq) was from New England Nuclear. Platelet-activating factor (PAF), bacterial lipopolysaccharide (LPS; Re 595 from Salmonella minnesota), actinomycin D, cycloheximide, and AA were from Sigma. Okadaic acid was from Calbiochem or Biomol (Plymouth Meeting, PA). Methyl arachidonyl fluorophosphonate (MAFP), NS-398, indomethacin, and aspirin were from Cayman Chemicals (Ann Arbor, MI). Human recombinant synovial PLA₂ and purified IgG from anti-human synovial PLA₂ rabbit antiserum (BQY-113A) were generously provided by J. L. Bobbitt and R. M. Kramer (Eli Lilly, Indianapolis, IN). Antibodies against murine COX isoforms were generously provided by W. L. Smith (Department of Biochemistry, Michigan State University, East Lansing, MI).

Cell Culture and Labeling Conditions.

P388D₁ cells were maintained at 37°C in a humidified atmosphere at 90% air and 10% CO₂ in Iscove’s modified Dulbecco’s medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, and nonessential amino acids. P388D₁ cells were plated at 10⁶ per well, allowed to adhere overnight, and used for experiments the following day. All experiments were conducted in serum-free Iscove’s modified Dulbecco’s medium. When required, radiolabeling of the P388D₁ cells with [³H]AA was achieved by including 0.5 μCi/ml [³H]AA during the overnight adherence period (20 h). Labeled AA that had not been incorporated into cellular lipids was removed by washing the cells four times with serum-free medium containing 5 mg/ml albumin.

Stimulation of the Cells.

The standard regimen for activating P388D₁ cells with LPS and PAF has been described previously (7, 13, 25). Briefly, P388D₁ cells were placed in serum-free medium for 30–60 min before the addition of LPS (200 ng/ml) for 1 h, after which they were challenged with 100 nM PAF for the time indicated.

Western Blot Analyses.

The cells were washed twice with serum-free medium and homogenized by 25 strokes in a Dounce homogenizer in a buffer consisting of 20 mM Tris⋅HCl, 2 mM EDTA, 10 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 20 μM leupeptin, 20 μM aprotinin, and 0.1% 2-mercaptoethanol, pH 7.5. The homogenates were centrifuged at 500 × g for 5 min at 4°C to separate nuclei. The resulting supernatant was centrifuged at 100,000 × g for 90 min at 4°C to separate cytosolic and membrane fractions. Samples from cell homogenates (50 μg) and their equivalent cytosolic and membrane fractions were separated by SDS/PAGE (10% acrylamide gel) and transferred to Immobilon-P (Millipore). Nonspecific binding was blocked by incubating the membranes with 5% nonfat milk in phosphate-buffered saline for 18 h. Membranes were incubated with anti-COX-1 or anti-COX-2 antiserum and then treated with horseradish peroxidase-conjugated protein A (Amersham). Bands were detected by enhanced chemiluminiscence (ECL; Amersham).

Northern Blot Analyses.

Total RNA was isolated from unstimulated or LPS/PAF-stimulated cells by the Trizol reagent method (Life Technologies), exactly as indicated by the manufacturer. Five micrograms of RNA was electrophoresed in an 1% formaldehyde/agarose gel and transferred to nylon filters (Hybond; Amersham) in 10× SSC buffer (1× SSC = 0.15 M NaCl, 0.015 M sodium citrate, pH 7.0). Hybridizations were performed in QuickHyb solution (Stratagene) by following the manufacturer’s instructions. ³²P-labeled probes for COX-1 or COX-2 (Cayman Chemicals) were coincubated with the filters for 1 h at 66°C, followed by three washes with 2× SSC containing 0.1% SDS at room temperature for 30 min. A final wash was carried out at 60°C for 30 min with 1× SSC containing 0.1% SDS. Bands were visualized by autoradiography.

Measurement of Prostaglandin E₂ (PGE₂) Production and of Extracellular [³H]AA Release.

For PGE₂ production, LPS-treated cells were stimulated with 100 nM PAF for 10 min. Afterward, the supernatants were removed and cleared of detached cells by centrifugation, and PGE₂ was quantitated by using a specific radioimmunoassay (PerSeptive Biosystems, Framingham, MA). For [³H]AA release experiments, the cells, labeled with [³H]AA as described above, were stimulated with 100 nM PAF for 10 min in the presence of 0.1 mg/ml BSA. The supernatants were removed, cleared of detached cells by centrifugation, and assayed for radioactivity by liquid scintillation counting.

Data Presentation.

Assays were carried out in duplicate or triplicate. Each set of experiments was repeated at least three times with similar results. Unless otherwise indicated, the data presented are from representative experiments.

RESULTS AND DISCUSSION

Expression of COX Isoforms.

When P388D₁ macrophages are treated with 200 ng/ml LPS for 1 h and then stimulated with 100 nM PAF, a rapid burst of PGE₂ is generated, reaching a peak 3–5 min after PAF addition (12). Immediate PGE₂ production occurs concomitantly with activation of both cPLA₂ and sPLA₂, and no further PGE₂ is produced in a delayed phase, even after several hours of incubation with LPS/PAF (12, 25). Immunoblot analysis of the COX isoforms expressed by resting P388D₁ cells revealed that COX-2 protein was increased by LPS treatment of the cells (Fig. 1A). COX-1, on the contrary, did not appear to change (Fig. 1A). LPS-induced expression of COX-2 was blocked by both actinomycin D and cycloheximide, inhibitors of RNA transcription and protein synthesis, respectively (Fig. 1A). COX-1 expression was not affected by either actinomycin D or cycloheximide (Fig. 1A).

Figure 1.

(A) Expression of COX isoforms in P388D₁ cells. The cells were incubated with 5 μM actinomycin D (A) or 10 μM cycloheximide (C) for 30 min before the LPS/PAF treatment (1 h with LPS plus 10 min with PAF). Cells were then lysed as described in the text. Protein (100 μg) was separated by SDS/PAGE and analyzed by immunoblotting using specific antibodies against COX-1 or COX-2. (B) Northern blot of COX-1 and COX-2 from total P388D₁ RNA. Total RNA from LPS/PAF-stimulated cells or nonstimulated (−) was isolated and hybridized with probes for COX-1 and COX-2, as indicated. As a control for RNA loading, an “in-gel” RNA picture is shown on the right.

The above results were confirmed by analyzing RNA message levels for both COX-1 and COX-2. When total RNA from resting cells was analyzed by Northern blotting using probes for COX-1, the signal detected was identical to that found when RNA from LPS/PAF-activated cells was used (Fig. 1B). Total RNA from resting cells also gave a positive signal when probes for COX-2 were used, albeit the signal was much weaker than that found for COX-1. A strong signal for COX-2 was detected, however, in RNA samples from LPS/PAF-activated cells (Fig. 1B).

LPS/PAF-induced PGE₂ production was completely abolished by NS-398, a specific COX-2 inhibitor (Fig. 2A). NS-398 completely inhibits COX-2 with an IC₅₀ <5 μM, whereas COX-1 remains unaffected at concentrations higher than 100 μM (26). Thus the above data suggest that COX-2 is the only enzyme responsible for stimulus-induced PGE₂ in these cells. Our previous results have demonstrated that sPLA₂ is the enzyme that provides most of the AA to be used for PGE₂ production in activated conditions in P388D₁ cells (7, 12, 14). Thus these previous data coupled with the results reported herein formally establish the existence of a link between sPLA₂ and COX-2 in activated cells.

Figure 2.

A) Effect of NS-398 on PGE₂ production in P388D₁ cells. The cells were incubated with the indicated concentrations of NS-398 for 30 min before and during the LPS/PAF treatment. PGE₂ in the supernatants was quantitated by using a specific radioimmunoassy. Basal PGE₂ production—i.e., that observed in the absence of LPS/PAF at the indicated NS-398 concentrations—is indicated by the open circles. (B) Effect of COX inhibitors on PGE₂ production from exogenous AA. The cells were treated with 5 μM NS-398, 25 μM indomethacin (Indo), 25 μM aspirin (Aspi), or neither (Ctrl) for 30 min before and during the 1-h treatment with LPS. Afterward, 1 μM AA was added and the amount of PGE₂ produced after 10 min was quantitated by using a specific radioimmunoassay.

From the data shown in Fig. 2A, it is apparent that basal unstimulated PGE₂ levels were not affected by NS-398. This finding could suggest that basal PGE₂ is produced by COX-1 in the resting cells. However it could also be possible that COX-1 is not functional in P388D₁ cells and that the basal PGE₂ produced corresponds to the fraction of PGE₂ that was produced during preincubation, just before the COX-2 became inactivated by NS-398. To investigate these possibilities, experiments were performed wherein PGE₂ production was analyzed in cells exposed to 1 μM exogenous AA (Fig. 2B). As well as being physiologically sound (under normal circumstances, the sPLA₂ provides the AA from outside), this approach has the advantage that the total amount of PGE₂ produced by the cells can be controlled (by adding more or less exogenous AA). PGE₂ production under these conditions was strongly, but not completely, inhibited by NS-398, indicating that the majority of the COX activity to which exogenous AA is accessible arises from COX-2 (≈80%). However, there is also an appreciable contribution by COX-1, as judged by the NS-398-resistant PGE₂ production (≈20%) (Fig. 2B). Aspirin and indomethacin, two nonspecific COX inhibitors, abolished PGE₂ production (Fig. 2B). Collectively, the results from these experiments suggest that the AA released by the sPLA₂ after cellular stimulation is accessible only to COX-2 and not to COX-1, even though the latter is present in active form in the macrophages.

cPLA₂ Regulates sPLA₂-Dependent AA and PGE₂ Production in Activated P388D₁ Macrophages.

To further assess the coupling of sPLA₂ to COX-2 in P388D₁ cells, a series of experiments was performed that involved adding exogenous recombinant sPLA₂ to the P388D₁ cells. The sPLA₂ present in P388D₁ cells belongs to group V (14). Unfortunately, no purification of group V sPLA₂ from either natural or recombinant sources has been reported. Thus for these experiments we used recombinant group IIA sPLA₂, which is structurally similar to the group V enzyme (16). Fig. 3A shows that exogenous sPLA₂ readily associated in a dose-dependent manner with the membrane of both resting and activated P388D₁ cells, as judged by flow cytometry under our previously described conditions (14). However, the ability of exogenous sPLA₂ to elicit lipid mediator production was strikingly dependent on the activation state of the cells. Exogenous sPLA₂ was almost inactive when added to nonstimulated cells, whereas the enzyme noticeably increased PGE₂ production when added to activated cells (Fig. 3B). These data are consistent with work by other laboratories in different cell types (8, 27, 28) and thus give additional support to the idea that, for the sPLA₂ to act on cellular membranes, membrane perturbation induced by a costimulator must occur first (5, 6).

Figure 3.

Effect of adding exogenous sPLA₂ to P388D₁ cells. (A) Either resting (open bars) or LPS/PAF-activated (crosshatched bars) cells were incubated with the indicated concentrations of exogenous sPLA₂, and sPLA₂ surface expression was determined by flow cytometry. (B) PGE₂ production by resting (○) or LPS/PAF-activated (•) P388D₁ cells in the presence of the indicated concentrations of exogenous sPLA₂. Exogenous sPLA₂ was added at the time the cells were exposed to PAF or vehicle, and was present for 10 min.

We recently proposed that a transient accumulation of intracellular free AA produced by cPLA₂ may constitute another important factor leading to sPLA₂ activation in stimulated cells (13). This suggestion was based on the finding that artificially increasing the intracellular free AA levels by exogenous addition of fatty acid enabled phospholipid hydrolysis, apparently mediated by sPLA₂ (13). Because transient elevations in the intracellular AA concentration are caused by the cPLA₂, these data suggested that crosstalk may exist between the two PLA₂s during P388D₁ cell activation by LPS/PAF (13). Further evidence for such crosstalk was obtained by experiments determining the effect of MAFP on PGE₂ production by LPS/PAF-treated cells. MAFP inhibits cPLA₂ but not sPLA₂ (13), yet it strongly inhibited PGE₂ production (Fig. 4), which is mostly an sPLA₂-dependent process (7, 14).

Figure 4.

Effect of MAFP on PGE₂ production by P388D₁ cells. The cells were incubated with the indicated concentrations of MAFP, which was added 15 min before PAF. PGE₂ in the supernatants was quantitated by using a specific radioimmunoassy. Basal PGE₂ production—i.e., that observed in the absence of LPS/PAF at the indicated MAFP concentrations—is indicated by ○.

However, validation of the existence of crosstalk between the two PLA₂s requires proof that the AA is actually activating the sPLA₂ and not inducing phospholipid hydrolysis by some other, unexpected, mechanism. We have further addressed the specificity of AA in activating the sPLA₂ by carrying out AA-release reconstitution experiments utilizing exogenous sPLA₂. In agreement with our previous data (13), when [³H]AA-prelabeled LPS-treated cells were incubated with PAF in the presence of the cPLA₂ inhibitor MAFP (25 μM), [³H]AA release to the extracellular medium was markedly inhibited, but this inhibition was reversed when the cells were exposed to exogenous AA (1 μM) for 1 min before PAF addition (Fig. 5). Consistent with the data shown in Fig. 3B, addition of exogenous sPLA₂ (500 ng/ml) increased PAF-stimulated [³H]AA release (Fig. 5). However, exogenous sPLA₂ did not have any effect on PAF-stimulated [³H]AA release when MAFP-treated cells were used; in fact, the response still remained inhibited (Fig. 5).

Figure 5.

AA restores the enhancing effect of exogenous sPLA₂ on [³H]AA mobilization in activated P388D₁ cells. [³H]AA-labeled LPS-treated cells were preincubated with MAFP (25 μM) for 15 min where indicated. Subsequently, 1 μM exogenous unlabeled AA was added 1 min before treatment with PAF (100 nM), as indicated. Where indicated, exogenous sPLA₂ (500 ng/ml) was added together with PAF. The data are expressed as a percentage of the response observed in activated cells in the absence of any addition, which was taken as 100%.

The inability of exogenous sPLA₂ to enhance [³H]AA release in MAFP-treated cells cannot be explained by lack of membrane perturbation, because a costimulator—i.e., PAF—is also present. Moreover because, as mentioned above, MAFP does not have any inhibitory effect on sPLA₂ (13), these results demonstrate that MAFP specifically blocks generation of a signal that is required for the exogenous sPLA₂ to attack the membrane. Direct proof that this signal is provided by elevation of the intracellular free AA levels was given by the finding that coincubation with this fatty acid not only reversed the MAFP inhibition but even restored the enhanced [³H]AA release response to PAF plus exogenous sPLA₂ (Fig. 5). The enhancing effect of AA appears to be specific, because other fatty acids tested, whether saturated (palmitic acid, stearic acid, arachidic acid) or unsaturated (oleic acid, linoleic acid), failed to restore the [³H]AA release (13).

P388D₁ cells manifest an extraordinarily high capacity to incorporate exogenous AA into membrane phospholipids (12, 29, 30). As much as 25% of the fatty acid at 1 μM—the concentration used in the experiment shown in Fig. 5—is incorporated into the phospholipids of 10⁶ cells in 10 min in the absence of albumin (12). This amount gives a rate of incorporation of about 25 pmol of AA per min per 10⁶ cells. However, in the presence of albumin—as employed in the experiment depicted in Fig. 5—the amount of available AA is reduced to one-fourth, as the majority is trapped by the albumin in the incubation medium (J.B. and E.A.D., unpublished data). Thus the rate of incorporation of exogenous unlabeled AA under the conditions of Fig. 5 is about 6 pmol per min per 10⁶ cells. At 1 μM AA, little fatty acid is incorporated into triacylglycerols, which indicates that the steady-state lysophospholipid pool is sufficient to account for this rate of incorporation of free AA even in the absence of albumin (30). Moreover, the specific activity of the reacylating enzymes arachidonoyl-CoA synthetase and lysophospholipid:acyl-CoA acyltransferase in P388D₁ cells are 104 and 33 nmol per min per 10⁶ cells (ref. 29; one million P388D₁ cells comprises ≈100 μg of protein). Therefore, because of the exceedingly high activity of the reacylating enzymes in P388D₁ cells it is highly unlikely that the restoring effect of exogenous unlabeled AA on [³H]AA release shown in Fig. 5 is because of inhibition of [³H]AA reacylation in the cells. Further experimental support for this interpretation is given by the finding that exogenous AA does not affect [³H]AA release in cells not treated with MAFP (Fig. 5). If the enhancing effect of exogenous AA on [³H]AA release in MAFP-treated cells was because of inhibition of reacylation, one would expect an enhanced [³H]AA release in the non-MAFP-treated cells as well, which was clearly not the case (Fig. 5).

AA and PGE₂ Release Mechanism in P388D₁ Cells.

Collectively, the data in this study establish that elevated levels of intracellular AA specifically lead to activation of sPLA₂ at the outer leaflet of the plasma membrane. Because elevated intracellular free AA levels are caused solely by cPLA₂ activation (13), the current results formally establish that crosstalk exists between cPLA₂ and sPLA₂ during P388D₁ cell activation by LPS/PAF. Therefore, the current data document a dual role for AA during macrophage activation by LPS/PAF. As well as serving itself as a substrate for the synthesis of other lipid mediators, the AA—or a metabolite—serves a signaling role by enabling activation of the sPLA₂ and thus allowing for a further amplification of the inflammatory response. These results, along with our previous data (12–14), support a mechanism for AA mobilization in P388D₁ cells whereby PAF receptor occupancy immediately leads to cPLA₂ activation, which, in turn, induces a transient elevation of intracellular free AA levels (Fig. 6). Elevated levels of free AA during the initial stages of P388D₁ cell activation may serve important metabolic and biochemical functions, one of which is to help activate the sPLA₂ that has appeared at the plasma membrane. Whether the activating effect of cPLA₂-derived AA on the sPLA₂ is a direct one or is mediated by accessory system(s) remains to be elucidated.

Figure 6.

Crosstalk between cPLA₂ and sPLA₂ in eliciting AA mobilization in activated P388D₁ macrophages. PAF activation of the LPS-primed macrophages triggers the “fast” activation of the cPLA₂ intracellularly, which in turn allows the “slow” action of sPLA₂ at the plasma membrane. The AA liberated by the sPLA₂ can be taken up and utilized by COX-2 to generate prostaglandins such as PGE₂. For further details see text.

Once the sPLA₂ is activated at the plasma membrane, it starts to hydrolyze phospholipids and release AA to the extracellular medium. Part of this AA will be recaptured by the cell, and a minor portion of it (see below) will be made accessible to COX-2, the enzyme responsible for producing PGE₂ during activation conditions. COX-1, though present and active, is spared from the process. Thus, the sPLA₂ specifically feeds COX-2, not COX-1. One possibility for the COX-1 not to participate in PGE₂ production in activated cells could be that this enzyme is located farther away from the sPLA₂ than is COX-2. Because of the exceedingly high reacylating capacity of the P388D₁ cells (29, 30), only less than 5% of the liberated AA is converted into prostaglandins (31). If the sPLA₂—and/or the cPLA₂—releases AA not in the immediate vicinity of COX-1, the fatty acid might be impeded in reaching the enzyme, as it would be acylated back into phospholipids well before reaching COX-1. According to this scheme, it could be envisioned that the active COX-2 pool is located in close proximity to the intracellular compartment wherein cellular uptake of AA takes place. Further work will be necessary to explore this intriguing possibility.

ABBREVIATIONS

PLA₂

phospholipase A₂

cPLA₂

85-kDa group IV cytosolic phospholipase A₂

sPLA₂

secretory phospholipase A₂

AA

arachidonic acid

COX

cyclooxygenase

PGE₂

prostaglandin E₂

LPS

bacterial lipopolysaccharide

PAF

platelet-activating factor

MAFP

methyl arachidonyl fluorophosphonate