Arachidonyl Trifluoromethyl Ketone, a Potent Inhibitor of 85-kDa Phospholipase 4, Blocks Production of Arachidonate and 12-Hydroxyeicosatetraenoic Acid by Calcium Ionophore-challenged Platelets*

Arachidonyl trifluoromethyl ketone (AACOCF,) is a potent and selective slow binding inhibitor of the 85-kDa cytosolic phospholipase Biochemistry 32,5935-5940).AACOCFS and a number of its structural analogues have been used to inves- tigate the role of c P w in the cellular generation of free arachidonic acid (AA) and in eicosanoid biosynthesis. AACOCF, inhibited the release of AA from calcium iono-phore-challenged U937 cells (IC5o = 8 p ~ , 2 x 10’ cells ml-’) and from platelets (IC5o = 2 p

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ''aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 To whom correspondence should be addressed. Tel.: 514-428-8542; types, the rate-determining step in the generation of eicosanoids is the release of arachidonic acid (AA)' from its cellular store in the phospholipid pool. A number of different pathways for the mobilization of AA have been proposed: phospholipase C in concert with glycerol lipases (11, lysophospholipase (2), and phospholipase 4 (PW). Detailed studies of the distribution and stoichiometry of metabolites have demonstrated the importance of P& as a major mediator of agonist-induced AA release in many cell types (3-5). Over the past decade a number of distinct types of PLA, have been isolated and characterized. The best known of these are a family of 14-kDa calcium-dependent secreted enzymes (sPLA,), an 85-kDa cytosolic calcium-dependent enzyme (cP&), and intracellular calcium-independent enzymes.
There has been a great deal of debate about the relative importance of each type of P& to the overall process of cellular AA mobilization. The 14-kDa P W s require millimolar concentrations of calcium for activity, and they do not exhibit selectivity toward the fatty acid at the sn-2 position of the phospholipid. Therefore, based on these properties it appears unlikely that the role of this type of PLA, is to initiate AA release from inside the cell. However, evidence has been provided that sPLA, might play a role in the production of prostaglandins in certain cell types. Barbour and Dennis (6) reported that the antisense inhibition of the 14-kDa PLA, expression blocked the production of prostaglandin E, by P388D cells. It has also been proposed that the 14-kDa P& may be involved in AA release and prostanoid production in human umbilical vein endothelial cells (7) and mesangial cells Calcium-independent PLAp have been isolated from human and canine myocardium (9, 10) and P388D cells (6). The compound (E)-6-(bromomethylene)-tetrahydro-3-(l-naphthalenyl)-ZH-pyran-2-one (HEL) is a potent suicide inhibitor of the myocardial calcium-independent P q (11) and selectively inhibits this enzyme uersus other known P&s (12). Thus HEL has lipase 4; c P W , 85-kDa cytosolic phospholipase &; s P w , 14

15619
been useful in determining the contribution made by this type of PIA& to AA mobilization by pancreatic islet cells (13) and aortic smooth muscle cells (14).
The cytosolic Pis thought to be involved in the production of AA for eicosanoid production since it preferentially hydrolyzes phospholipids containing AA at the sn-2 position (15,161, responds to physiological changes in calcium concentration by translocating to membranes (15,17), and is activated by hormonal signaling through phosphorylation of a serine residue (18,19). However, no direct evidence for the involvement of this enzyme in the production of eicosanoids has been provided. Recently, we have described the first potent and selective inhibitor of the cPLA, (20). An analogue of AA in which the -COOH functionality is replaced by -COCF, (AACOCF,) is a potent slow tight binding inhibitor of cP&.
AACOCF, also shows 500-fold greater potency against cPLA, than sPLA, (20). Through use of AACOCF, and a number of its close structural analogues we report that cPLpL, is the major mediator of AA release in calcium ionophore-stimulated platelets and U937 cells. Furthermore, we demonstrate that in platelets the AA mobilized by the action of the cPLA, is subsequently utilized by the 12-lipoxygenase pathway for production of 12-HETE.

MATERIALS AND METHODS
Preparation of Platelets and U937 Cells"u937 cells were grown and differentiated with Me,SO as described by Tremblay et al. (21). Platelets were prepared from human venous blood obtained from healthy volunteers. The collected blood was immediately mixed with 0.1 volume of anticoagulant solution (65 mM citric acid, 85 mM sodium citrate, and 2% glucose) and centrifuged a t 200 x g for 10 min. The supernatant was mixed with 50% (by volume) Hanks' balanced salt solution buffered with 15 mM HEPES pH 7.4 (HHBSS) and a 30% volume of the anticoagulant solution. This mixture was centrifuged at 750 x g for 10 min, and the pellet was resuspended in HHBSS. Platelet concentration was determined with a Coulter counter.
Release and Measurement ofAA in Platelets and U937 Cells-Human platelets were resuspended to a final density of 4 x 10' ml" in HHBSS and ETYA added from a Me,SO stock solution to give a final concentration of 10 p~. U937 cells were resuspended to a final density of 2 x lo6 ml" in HHBSS supplemented with 1.4 mM CaCI, and 0.7 mM MgSO,. To 250 pl of cell suspension, 1 pl of a Me,SO solution of the inhibitor or of Me,SO only was added, and the mixtures were incubated at 37 "C for 2 min. Calcium ionophore (A23187, 10 p~ for U937 cells, 2 p~ for platelets) was added, and incubation continued for 10 min in the case of the platelets and 3 min for the U937 cells. The cells were extracted with 1.25 ml of Dole's solution (heptane, isopropyl alcohol, 1 N sulfuric acid; 10:40:1), followed by 0.75 ml of heptane, 0.5 ml of water, and vortexing. The heptane phase was removed and dried (Na,SO,), and the pentafluorobenzyl esters of the fatty acids were prepared and quantified by gas chromatography-mass spectometry with reference to an internal standard of d,-AA as described by Li et al. (22).
Release and Measurement of TxB, and 12-HETE from Platelets-Platelets at a final concentration of 4 x lo' ml" (0.5 ml) were preincubated for 2 min at 37 "C before stimulation with 2 p~ calcium ionophore (A23187) or 10 FAA. After another 10-min incubation at 37 "C, 0.25 ml of cold methanol was added to stop the reaction, and TxB, levels were measured by enzyme immunoassay (Cayman Chemical Co., Inc.) For 12-HETE measurements, 150 pl of the mixture was acidified with 10 pl of 3 N acetic acid, and the sample was analyzed by reverse-phase HPLC on a C,, Nova-Pak column with acetonitrile/water/acetic acid (60:40:0.1) as the solvent (2 mllrnin).
Phospholipase A, Assays-Samples of the purified recombinant cPLk, were prepared as described by Street et al. (201, and the purified recombinant 14-kDa synovial fluid-type P& was prepared as described by Tremblay et al. (23). A suspension of platelets was prepared from 60 ml of human blood as described above and centrifuged at 750 x g for 10 min. The platelets were resuspended in 20 ml of a buffer containing 10 mM HEPES pH 7.5, 1 mM EDTA, 100 mM KCl, 1 l l l~ phenylmethylsulfonyl fluoride, leupeptin (1 pg ml-l)), and the cells were broken by sonication. The suspension was centrifuged a t 110,000 X g for 60 min; the cytosolic and the microsomal fractions separated, and each was assayed for Pm activity. The assay procedure previously described for the recombinant 14-kDa PLA, was used (23)  Also shown is AA released by 10' platelets after stimulation with 2 p~ A23187; 0, ETYA-treated platelets; 0, platelets without ETYA treatment.
Tris-HC1 (pH 8.U, and 2.5 mM CaCl, in a volume of 100 4. Assays were incubated for 30 min at 37 "C and quenched by the addition of 20 mM EDTA and 0.9 rnl of acetonitrile. The assay buffers for calcium-independent Pcontained 5 m~ EDTA and 50 mM Tris-HC1 (pH 8.1). The c P m activity was measured using the assay conditions described by Inhibitors were added directly to the assay mixture from a Me,SO stock solution.
Metabolism ofAACOCF, by Cells and Hepatic Microsomes-A JEOL HXllOA mass spectrometer operating in continuous flow liquid secondary ion mass spectrometric mode (CF-LSIMS) was used to characterize the reduction ofAACOCF, in human platelets and U937 cells. A Waters 600"s HPLC was operated at a flow of 0.9 ml min". The flow was split just prior to a Rheodyne injector to produce a flow of 3 pl min" through a Spherisorb 0.32 x 100-mm capillary column then flowing to the frit probe of the mass spectrometer. A linear gradient of 70-95% A in 25 min was used for the separation ( A 1.5% glycerol in CH,CN; B: 20 rn NH,OAc (pH 4.3), 1.5% glycerol). The mass spectrometer was operated in positive ion mode. Ions were produced by bombardment with primary Cs' ions (20 keV); the accelerating voltage was 10 kV.
AACOCF, (10 p~) was incubated with the cells a t 37 "C (250 pl total volume) for the stated times, and then an equal volume of acetonitrile was added. The suspension was centrifuged and the supernatant removed. An aliquot (10 pl) of this was then injected directly onto the CF-LSIMS system. The hydrate ofAACOCF, eluted at 16 min, while the reduction product (AACH(OH)CF,) eluted at 20.5 min.
Hepatic microsomes were prepared from male Sprague-Dawley rats (350 g) by standard procedure (24), and aliquots were frozen a t -80 "C. Microsomal incubations were conducted with 1 mg of thawed microsomal protein; 400 pl of cofactor solution containing 2.5 mM MgCl,, 2.5 m~ NADP, and 25 mM glucose 6-phosphate in 125 mM phosphate buffer (pH 7.4); 2 units of glucose-6-phosphate dehydrogenase; and 12.5 pl of water. After a 2-min incubation at 37 "C in a water bath, 12.5 pl of a Me,SO solution of AACOCF, was added to give a final concentration of 200 p~ AACOCF, in a total volume of 500 pl. Blank incubations contained no AACOCF,, and control incubations were conducted with boiled microsomes. After 5 min, the mixtures were quenched by the addition of 500 p1 of acetonitrile. The precipitated protein was removed by centrifugation, and the resulting supernatant was used for CF-LSIMS.
Synthesis of Inhibitors-AACOCF,, AACH(OH)CF,, and AACOCH, were prepared according to the method of Street et al. Human platelets or Me,SO-differentiated U937 cells were incubated with arachidonyl analogue or vehicle and activated with calcium ionophore, and the products were measured as described under "Materials and Methods." A, stimulated level of AA released by differentiated U937 cells 3 min after the addition of 10 p~ A23187. 0, AACOCH,-treated cells; 0, AACH(OH)CF,-treated cells; A, AACOCF,-treated cells. B, inhibition of AA release and 12-HETE production by AACOCF, in A23187-stimulated platelets. To obtain the maximum stimulated levels of 12-HETE and AA, both ETYA-treated AACOCH,. The treated platelets were then incubated for 2 min with the and normal platelets (4 x lo7 ml-l) were incubated for 10 min with 10 p~ A23187. 0, AA released by ETYA/AACOCH,-treated platelets; 0, 12-stated concentrations of AACOCF, and then challenged with 2 p~ HETE produced by AACOCH,-treated platelets.
the cells decreased, most likely due to the reincorporation of the free acid back into the phospholipid pool and to its utilization in the production of prostaglandins. Incubation of the cells with AACOCF, for 2 min prior to the addition of A23187 produced a concentration-dependent decrease in the stimulated level of free AA (Fig. 2.4). At a cell concentration of 2 x lo6 ml-', complete inhibition of AA release was observed a t 30 p~ AACOCF,, and 8.5 p~ inhibitor was required t o reduced the stimulated level by 50%. AACOCH, and AACH(OH)CF, are close structural analogues of AACOCF,; however, neither of these compounds caused detectable inhibition of the purified cPLA, (20), nor did they inhibit the release of AA in the U937 cells a t concentrations up to 40 p~. At all concentrations tested, both AACOCH, and AACH(OH)CF, increased the level of free AA in the stimulated U937 cells (Fig. 2 A ) .
In order to measure AA release by human platelets, it is necessary to inactivate both the cyclooxygenase and the 12lipoxygenase pathways by treatment with 10 p~ ETYA (25). Challenge of the treated platelets with 2 p~ A23187 resulted in a rapid burst ofAA release that reached a peak level of approximately 70-120 ng ofAA/107 platelets 1-2 min after the addition of the calcium ionophore (Fig. 1). In contrast to the U937 cells, the level of free AA in the ETYA-treated platelets did not decline at longer times. The effect of a 2-min preincubation of the platelets with 15 p~ of the arachidonyl analogues before addition of the ionophore was determined. AACOCF, completely inhibited AArelease at this concentration, while AACOCH, and AACH(OH)CF, both raised the levels of AA by approximately 2-fold. The inhibition of AA release by AACOCF, was dosedependent, 2 p~ inhibitor producing a 50% reduction in the Effect of arachidonate analogues on 18-HETE and TxB, production Samples were preincubated for 2 min with human platelets (4 X lo7 platelets ml)-l at 37 "C before challenge with eitherA23187 (2 p) or AA (10 p~) .
The results are expressed as the percentage of the amount of 12-HETE and TxB, produced by the stimulated platelets after 10 min in the absence of arachidonate analogue. The levels of 12-HETE produced by A23187-and AA-stimulated platelets were 53 ng/10' cells and 133 ng/lO' cells, respectively. The levels of T X B , increased from 2.3 to 32 ng/107 cells following A23187 stimulation and from 2.8 to 107 ng/107 platelets for AA-stimulated platelets. stimulated level of free AA (Fig. 2B 1. HEL was used to test if a myocardial-type calcium-independent P& might also play a role in the release ofAAin ionophore-stimulated platelets. Preincubation of platelets with 10 p~ HEL before stimulation resulted in less than 20% inhibition of AA release. Inhibition of 18-HETE and TxB, Production by Platelets-To determine the effect of inhibition of AA release on eicosanoid biosynthesis, the production of 12-HETE and TxB, by activated platelets (no ETYA) in the presence of AACOCF, and its analogues was measured. The results are shown in Table I. 12-HETE and TxB, are two major metabolites of arachidonic acid produced by platelets, and on stimulation with 2 1.1~ A23187, approximately 50 ng of 12-HETE/107 platelets and 30 ng of TxB2/107 platelets were produced. Treatment of the platelets with the potent cyclooxygenase inhibitor, flurbiprofen, resulted in almost complete inhibition of TxB, production and a 2-fold increase in the amount of 12-HETE produced. Both AACOCH, and AACH(OH)CF, behaved in a similar manner to flurbiprofen in that addition of either compound to the platelets at a concentration of 15 1.1~ resulted in a marked inhibition of TxB, synthesis (70-90%) and a 2.5-3-fold increase in the amount of 12-HETE produced. In contrast to the result obtained with AACOCH, and AACH(OH)CF,, AACOCF, at a concentration of 15 p~ completely inhibited both 12-HETE and TxB, biosynthesis. The inhibition of 12-HETE biosynthesis by AACOCF, was concentration-dependent (IC5,, = 2 p~) , and as shown in Fig. 2 B , the dose dependence is very similar to that of AA release in ETYA-treated platelets.
Challenge of Platelets with AA-Direct addition ofAA (10 p~) to platelets without A23187 challenge allows the production of 12-HETE and TxB, independent of the action of phospholipase 4. Under these conditions, flurbiprofen, AACOCH,, AA-CH(OH)CF,, and AACOCF, all inhibited the production of TxB,, and all increased the levels of 12-HETE (Table I). These results establish that these compounds can inhibit TxB, biosynthesis by a mechanism that does not directly involve the inhibition of AA release. The results also establish that none of the arachidonyl analogues significantly inhibit the platelet 12lipoxygenase. Selectivity ofAACOCF, for Inhibition of the cPLA2 in Platelet Subfractions-The selectivity of MCOCF, for inhibition of the cPLA, versus other platelet PLA,s was demonstrated using platelet membrane and cytosolic fractions. The assays used to measure the platelet P& activities were optimized t o differentiate between sP&, cP&, and calcium-independent PI,&, activities. The synthetic phospholipid derivative, 10-Py-PM, was used to measure the activity of sP&. The sPLA, requires

PLA, activities observed in platelet extracts and their susceptibility to inhibition by AACOCF,
The following amounts of protein were used in each assay: sPL4, 100 pg; cPLA,, 60 ng; platelet membrane fraction, G 1 2 pg; platelet cytosol, 4-11 pg. Each assay was incubated for 30 min at 37 "C, and the product was quantified as described under "Materials and Methods." free calcium for hydrolysis of 10-Py-PM, and no product was observed when EDTA was added to the assay buffer in excess of the calcium chloride ( Table 11). The specific activity of cP& is approximately 60,000-fold lower than that of sPLA, in the 10-Py-PM assay (Table 11). The activity of the cPLA, was measured using a mixed micelle substrate of ['4C]PAPC and Triton X-100 (20, 26). In the mixed micelle assay, the cPLA, required free calcium for activity (1834 pmol30 min" pg-l), and 5 m~ dithiothreitol did not inhibit the enzyme activity. The c P 4 was also much less active when the mixed micelles were prepared with 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (29.7 pmol 30 min" pg"). The mixed micelle assay was found to be specific for the c P q versus sPLA,, since no detectable product was formed by 0.6 pg of s P q over a 30-min incubation.
Using the 10-Py-PM and the ['4C]PAPC/fl-iton X-100 mixed micelle assays, a number of distinct PLA, activities were observed in the crude cytosolic and membrane fractions derived from human platelets (Table 11). With the 10-Py-PM assay, most of the P L 4 activity was found in the cytosolic fraction. The cytosolic P q activities were derived from both calciumdependent and calcium-independent enzymes. The membrane fraction also showed significant P q activities when assayed with 10-Py-PM; in this case the observed activity was derived predominantly from calcium-dependent enzyme(s). When the [14C]PAPC/Triton X-100 mixed micelle assay was used, significant P L 4 activity was observed only in the cytosolic fraction. This PLA, activity was calcium-dependent, was not inhibited by 5 mM dithiothreitol, and showed much lower activity on mixed micelles of Triton X-100 and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine.
The susceptibility of the various platelet PLA,s to inhibition by AACOCF, was tested. The results are shown in Table 11. Addition of AACOCF, to the [14ClPAPC/Triton X-100 mixed micelle assay at a concentration of 1.6 mol % inhibited the recombinant cPLA, by 78% and the calcium-dependent PLA, activity in the cytosolic subfraction of the platelets by 77%. In contrast, addition of AACOCF, to the 10-Py-PM assay at a higher concentration of 10 mol % produced no significant inhibition of either the calcium-dependent or the calcium-independent P L 4 activities in the soluble extracts. At this inhibitor concentration, some inhibition of the recombinant sPLA, (25%) and the calcium-dependent membrane-associated PLA, (17%) was observed. The calcium-independent P L 4 activity observed with the 10-Py-PM assay was not inhibited by 10 mol % of HEL, suggesting that this enzyme might be different from the myocardial calcium-independent P q described previously (12, 14,27).
Metabolism of AACOCF, in Platelets and U937 Cells-The reduction of AACOCF, to AACH(OH)CF, in the cellular environment was followed using CF-LSIMS. The percentage of the AACOCF, reduction was estimated by comparing the relative

DISCUSSION
The potential role of P q in the mobilization ofAA from the cellular phospholipid pool has been the subject of a great deal of research. Human platelets produce large quantities of AA and eicosanoids upon stimulation with calcium ionophore and other agonists such as thrombin. Thus platelets provide a good system for the study of AA mobilizatiodutilization pathways. Platelets contain a number of different P q s , the cP& (17, 28, 291, the 14-kDa secreted PLA, (30)(31)(32), and others (33,34).
Recent data showed that a 70% depletion of the s P q in rabbit platelets does not affect TxB, production, suggesting that the sPLA, is not required for AA liberation during platelet activation (35). This is also supported by results from Gelb and coworkers' who have recently tested a number of potent competitive inhibitors of sPLA, and found that they did not inhibit the thrombin-stimulated release ofAAfrom platelets. The presence of the cPLA, in platelets and its translocation from the cytosol to the membranes in response to thrombin has recently been demonstrated (17,361, and these results have implicated the cPLA, in platelet activation. To attempt to elucidate the potential role of the cPL4 in the AA mobilization pathways of platelets, we have employed a recently described potent slow binding inhibitor of the cPLA,, AACOCF, (20). AACOCF, is a selective inhibitor of the cPLA, versus the sPLA, (20) and, as demonstrated here, versus other platelet PLA,s as well. A number of distinct calcium-dependent and calcium-independent P h activities were detected in platelet fractions, and only the activity showing the functional characteristics of the cPL4 was significantly inhibited by AACOCF,. In agreement with the subcellular location of the cPLA, in resting platelets (171, the AACOCF,-susceptible PLA, activity was found only in the soluble fraction. Assays for CoA-independent transacylase activity in U937 microsomes (37) showed that AACOCF, is a very weak inhibitor of this activity (ICbO > 50 p M h 3 AACOCF, produced a dose-dependent decrease in the calcium ionophore-stimulated production of AA from Me,SO-dif- ferentiated U937 cells and from platelets. In contrast, AA-COCH, and AACH(OH)CF, compounds, which are noninhibitory to the purified cPLA,, also did not inhibit AA release by U937 cells or platelets at any of the concentrations tested (up to 40 p~) . The arachidonyl analogues appear t o be relatively nontoxic to cells, since none of them caused increased leakage of lactate dehydrogenase when incubated with platelets. ' The results presented here also point to a number of limitations for the use of AACOCF, in cell-based studies. First, AA-COCF, is a slow binding inhibitor of c P q , and although it is a potent inhibitor (K,* < 5 x mol fraction) at low interfacial concentrations, the full inhibitory potency takes many minutes to develop (20). Consequently, in a cell-based assay where AA mobilization is over in less than a minute, relatively high concentrations of AACOCF, are required to inhibit cP& within this short period of time. Second, it was found that AACOCF, is reduced to its noninhibitory alcohol (AACH(OH)CF,) on incubation with U937 cells, platelets, or rat liver microsomes. The rate of reduction was sufficiently slow that much of the AA-COCF, could be recovered intact from the cells after short incubation times. However, the reduction of AACOCF, suggests that it could not be used successfully in cell-based systems where prolonged incubations are required. Despite these disadvantages, the results presented here show that there is a good correlation between the inhibitory potency of the arachidonyl analogues against the purified cPLA, and their ability to inhibit AA release in intact cells.
Both AACOCH, and AACH(OH)CF, produced an increase in the stimulated level of AA in the cell-based assays. These compounds might inhibit enzymes in the utilization and reacylation pathways of the cell, producing an increase in the steadystate level of free AA. Alternatively, the increase could be due to a direct effect of these compounds on the activity of cP-, since AACOCH, and AACH(OH)CF, have been shown to increase the activity of the purified cPLA, in a mixed micelle assay (20). In this case, the increased activity of the purified cPLA, was ascribed to a more favorable partitioning of the enzyme to the lipid-water interface in the presence of these arachidonyl-like compounds (20, 38). That there is some direct stimulation of CPLA, activity in the cells is likely since both AACOCH, and AACH(OH)CF, also increased the simulated level ofAAin ionophore-challenged, ETYA-treated platelets. In this case the utilization and reacylation pathways of the platelet are blocked by ETYA, consequently, the increased AA release observed in the presence of the arachidonyl analogues is probably due to the increased activity of the AA release pathway.
AACOCF, inhibited 12-HETE biosynthesis in a dose-dependent manner in calcium ionophore-stimulated platelets but not in platelets where 12-HETE production was induced by the addition of AA. This observation strongly suggests that AA-COCF, inhibits 12-HETE biosynthesis at the level of AA mobilization rather than by inhibition of 12-lipoxygenase. The similar dose dependences for the inhibition of 12-HETE production and AA release also suggest that AACOCF, inhibits 12-HETE biosynthesis by decreasing the amount of free AA available for the 12-lipoxygenase. In contrast t o its effect on 12-HETE production, AACOCF, inhibited biosynthesis of TxB, in both calcium ionophore-and "challenged platelets. Thus it is likely that in this case AACOCF, inhibits TxB, production both at the level of AA release and a t a point further down the TxA, biosynthetic pathway. This is substantiated by the results obtained with structural analogues of AACOCF,.
Both AA-CH(OH)CF, and AACOCH, are noninhibitory to the purified c P W , but both inhibited the production of TxB, in calcium ionophore-and "challenged platelets. This again suggests that this class of compounds can inhibit prostanoid formation at a point in the biosynthetic pathway other than AA mobilization.
Interestingly, AACH(OH)CF,, and AACOCH, did not inhibit 12-HETE production in ionophore-stimulated platelets but increased the levels of this eicosanoid by greater than 2-fold. The stimulation of 12-HETE production by these compounds was also seen in AA-challenged platelets. It should be noted that AA was added to the platelets at a concentration that does not saturate 12-HETE production. The stimulatory effect of these compounds is likely mostly due to their inhibitory effect on the prostanoid biosynthetic pathway. It has been shown previously that the AA for both 12-HETE and TxB, biosynthesis is derived from a single phospholipid pool (39) and that AAcan be shunted from one pathway to the other. In agreement with this, flurbiprofen, a potent inhibitor of cyclooxygenase, also increased the level of 12-HETE produced in calcium ionophore-and AA-challenged platelets in a manner similar to the arachidonyl analogues. The stimulation of 12-HETE by AACOCH, and AA-CH(OH)CF, was more pronounced than by flurbiprofen; this may be related to the ability of the arachidonyl analogues to directly activate the cPLA, as previously observed in a mixed micelle assay (20).
In summary, AACOCF, is a selective inhibitor of the cPLA, uersus other platelet P w s , and this compound inhibits the release of AA in calcium ionophore-challenged U937 cells and the release of AA and 12-HETE in calcium ionophore-challenged platelets. AACOCF, also inhibits the production of TxB, in platelets, although this is likely due to a combination of its inhibitory effects on AA mobilization and at a point further along in the prostanoid biosynthetic pathway. These results are consistent with a major role for cPLA, in the liberation of AA from the phospholipid pool for eicosanoid biosynthesis.