Serine Esterase Inhibitors Block Stimulus-induced Mobilization of Arachidonic Acid and Phosphatidylinositide-specific Phospholipase C Activity in Platelets*

Serine esterase inhibitors (phenylmethanesulfonyl fluoride, 5-dimethylaminonaphthalene-1-sulfonyl (dan-syl) fluoride, 2-nitro-4-carboxyphenyl-N,N-diphenyl- carbamate, or p-nitrophenyl anthranilate) blocked the production of malonyldialdehyde by platelets induced with a variety of stimuli (including thrombin, trypsin, collagen, and A23187). These inhibitors did not block malonyldialdehyde production by platelets from exogenous arachidonic acid. Those inhibitors studied in greater detail (phenylmethanesulfonyl fluoride and 2-nitro-4-carboxyphenyl-N, N-diphenylcarbamate) were shown to inhibit the release of [l-‘4C]arachidonic acid from phosphatidylinositol and phosphatidylcholine in intact platelets but not the conversion of arachidonic acid to thromboxanes, prostaglandins, or hydroxyfatty acids. These inhibitors also blocked the stimulus-in- duced production of [32P]phosphatidic acid in intact platelets. Both arachidonic acid release from phosphatidylinositol and phosphatidic acid production have been reported to depend on the production of digylcer-ide by the action of a phosphatidylinositol-specific phospholipase C. That enzyme in the soluble fraction from disrupted platelets was inhibited at concentra- tions of serine esterase inhibitors which block arachidonic the Dipentadecanoin was added as an internal standard and diglyceride was isolated from the lipid extract by chromatography on Silica Gel G thin laver plates in ether/petroleum ether/acetic acid (100:50:1, v/v) and quantitated by gas-liquid chromatography of fatty acid methvl esters produced by transesterification with 10% boron tri-chloride m methanol.


2-nitro-4-carboxyphenyl-N,N-diphenyl-
carbamate, or p-nitrophenyl anthranilate) blocked the production of malonyldialdehyde by platelets induced with a variety of stimuli (including thrombin, trypsin, collagen, and A23187). These inhibitors did not block malonyldialdehyde production by platelets from exogenous arachidonic acid. Those inhibitors studied in greater detail (phenylmethanesulfonyl fluoride and 2nitro-4-carboxyphenyl-N, N-diphenylcarbamate) were shown to inhibit the release of [l-'4C]arachidonic acid from phosphatidylinositol and phosphatidylcholine in intact platelets but not the conversion of arachidonic acid to thromboxanes, prostaglandins, or hydroxyfatty acids. These inhibitors also blocked the stimulus-induced production of [32P]phosphatidic acid in intact platelets. Both arachidonic acid release from phosphatidylinositol and phosphatidic acid production have been reported to depend on the production of digylceride by the action of a phosphatidylinositol-specific phospholipase C. That enzyme in the soluble fraction from disrupted platelets was inhibited at concentrations of serine esterase inhibitors which block arachidonic acid release in intact platelets. These results indicate that serine esterase inhibitors block the stimulus-induced mobilization of arachidonic acid in platelets at least in part by their action on the phosphatidylinositol-specific phospholipase c.
Metabolites of arachidonic acid are important to platelet function. Thromboxane Ae, as well as PGGZ' and PGH, can * This work was supported by National Institutes of Health Grant HL18937. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adrlertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. induce platelet aggregation and the release of serotonin from storage granules (1). Thromboxane A2 is also a potent vasoconstrictor and the hydroxy acids formed by the enzymatic action of lipoxygenase (z.e. HETE) and thromboxane synthetase ( i e . HHT) may be chemotactic agents ( 2 ) . Since there is little or no free arachidonic acid in the platelet, regulation of the production of these metabolites is presumed to be exerted on the enzymatic release of arachidonic acid from the platelet phospholipids. Mechanisms for release of arachidonic acid involving phospholipase An (3,4) or the sequential action of a phosphatidylinositol-specific phospholipase C and diglyceride lipases (5-8) have been proposed.
The release of arachidonic acid f r o p platelet phospholipids is stimulated by a variety of platelet aggregating agents, such as collagen, thrombin, and certain other proteases (e.g., trypsin and papain), basic peptides and mellitin, certain thiol reagents, and the divalent cation ionophore A23187. We previously reported that in platelets stimulated by collagen or thrombin, the serine esterase inhibitor PMSF blocked formation of malonyldialdehyde and thromboxane An as well as the antimycin A-resistant burst in oxygen consumption which is due to oxidation of arachidonic acid (9). In this communication, we extend those findings to a number of other active site serine esterase inhibitors @e. acetylbenzenesulfonyl fluoride, tosyl chloride and fluoride, dansyl fluoride, 2-nitro-4carboxyphenyl-N,N-diphenylcarbamate andp-nitrophenylanthranilate) and suggest the phosphatidylinositol-specific phospholipase C as a site for their action.

MATERIALS AND METHODS
Platelets were obtained less than 24 h after collection from the Connecticut American Red Cross Blood Center and were concentrated to 20 to 30 mg/ml of platelet protein in their own plasma.
Where indicated, platelets were washed as previously described (10) and suspended a t a concentration of 20 to 30 mg of platelet protein/ ml in 25 mM Tris, pH 7.4, 1 mM ethylene glycol bis(p-aminoethy1 ether)N,N,N',N'-tetraacetic acid (EGTA), 137 mM NaC1,5.4 mM KCI, 0.2% dextrose. Platelets were stimulated after dilution to a concentration of 1 to 3 mg of platelet protein/ml at 37°C.
Platelet phospholipids were prelabeled with [''C]arachidonic acid essentially as described by Bills et al. (11). Exogenous ["Clarachidonic acid metabolism was measured at a total concentration of 50 p~ arachidonic acid with 1 to 3 mg/ml of washed platelets in 1 ml of the buffer. In all experiments where radioactive products were determined, the reactions were stopped by the addition of IO ml of chloroform/methanol (1:1, v/v) plus 0.5 ml of 0.5 M citric acid, pH 3. Fractions of the lipid extract were analyzed by thin layer chromatography for arachidonic acid metabolites and phospholipids (12). Radioactivity in hands from thin layer plates was determined by direct counting in Budget-Solve (RPI), 12% water, 6% methanol. prelabeled for 60 min with 5 pCi of [ "PP]phosphoric acid in 5 ml of Phosphatidic acid formation was measured in platelets which were plasma. They were then harvested, washed, and resuspended. Platelets were stimulated in pH 7.4 buffer (plus 1.0 mM CaCId with collagen (70 pg/ml), thrombin (10 units/ml), or trypsin (4 M ) either without inhibitor or in the presence of either 1 mM 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate or 2 mM I'MSF. After 2 min at 37"C, lipids were extracted with chloroform/methanol (2:l, v/v) and analyzed by TLC in the organic phase of ethyl acetate/trimethylpentane/acetic acid/water (110:50:20:100, v/v), a svstem in which phosphatidic acid is the only phospholipid which migrates (13).
Platelet supernatant fractions from sonicated cells were prepared for assay of phosphatidylinositol-specific phospholipase C activity essentially as described by Mauco et al. (7): 0.75 mg of supernatant protein was incubated with 2 0 0 p~ phosphatidylinositol (yeast, Avanti Biochemical), 1 mM CaCIL in 50 mM Tris/acetate buffer, pH 6.5, in a total volume of 0.5 ml. After 30 min at 37"C, the reaction was stopped bv the addition of 2.5 ml of chloroform/methanol (21, v/v). Dipentadecanoin was added as an internal standard and diglyceride was isolated from the lipid extract by chromatography on Silica Gel G thin laver plates in ether/petroleum ether/acetic acid (100:50:1, v/ v/v) and quantitated by gas-liquid chromatography of fatty acid methvl esters produced by transesterification with 10% boron trichloride m methanol. Malonvldialdehyde and platelet oxygen consumption were determined as previously described (9).

RESULTS AND DISCUSSION
Malonyldialdehyde production was chosen for the initial assessment of the effect of serine esterase inhibitors since this colorimetrically assayable compound arises from the metabolism of prostaglandin endoperoxides, probably largely by the action of thromboxane synthetase on PGHL (14). Several different chemical types of serine esterase inhibitors were used in these experiments. These inhibitors are compounds with well defined specificity for the active site serines of certain esterases and proteases. For example, 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate in common with other p-nitrophenyl carbamates (15), reacts stoichiometrically with chymotrypsin to form inactive diphenylcarbamyl-chymotrypsin (16); I'MSF and other sulfonyl fluorides sulfonate the active site serine of trypsin, chymotrypsin, and thrombin (17,18); dansyl fluoride reacts specifically a t the active sites of chymotrypsin and subtilisin (19); and p-nitrophenylanthranilate has been used as a specific reporter group for the active site of chymotrypsin (20). The results presented in Table I illustrate that each of these inhibitors, as well as others related to them, was able to antagonize the stimulus-induced production of malonyldialdehyde by platelets, but had little or no effect on the production of malonyldialdehyde from 25 to 300 KM exogenous arachidonic acid. The inhibition of stimulus-induced malonyldialdehyde formation by the compounds listed in Table I was concentration-dependent. Near1.v complete inhibition of stimulus-induced malonyldialdehyde production could usually be attained when the inhibitor concentrations were 2 to 3 times higher than shown in Table I. The inhibitors were also effective in blocking malonyldialdehyde production TABLE: I Effects ofserine esternse inhlbitors on mnlonyldlaldehyde production from e.rog.enou.s and endogenous arachidonic acid Platelets concentrated in plasma to 'LO to 30 mg of platelet protein/ ml were diluted 5 t o IO-fold in Tris-buffered saline at pH 8.0, 37°C. Inhibitors were added for 2 min followed either by arachidonic acid or one o f the platelet stimulants. After ti min. the reactions were stopped and malonyldialdehyde measured as previously described (9). I he control production of malonyldialdehyde in nmol/mg of platelet protein is set at 100 to normalize all results which were conducted at different times and with platelets from different donors. Each result is expressed relative t o its own control.  " Control platelets (1 to 2 mg of protein/ml) were incubated w t h 4 M control trvpsin (no inhibitor present) for 5 min at 37°C and malonyldialdehvde production was measured. This value was set at 100.
'' Trypsin at 400 p~ was incubated with indicated concentratlons o f ' 2-nitro-4-carboxyphe1~vl-N,N-diphen.?lcarl~amate (NCDC) or dansyl fluoride for 5 min at 37°C. Aliquots o f the pretreated trypsin were then diluted 1C"fold into suspensions of normal platelets resulting in inhibitor concentrations of 50 p~ N C I X and 10 p~ dansyl fluoride experienced bv the platelets. Note that little inhibition of malonyldialdehvde production occurred under these conditions. ' Platelets were pretreated with the indicated concentration 01' inhibitor for 2 min at 37°C. Untreated trypsin (4 p~) was then added for 5 min and malonyldialdehyde was measured. Note that under these conditions (i.e. preincubation of the inhibitors with the platrlets rather than with trypsin), more than 90'; inhibition o f malonyldialdehvde production was attained.
We have been able to rule out interactions between the serine esterase inhibitors and the platelet-stimulating agents as a trivial cause for the observed effects. For example, I'MSF sulfonates the active site serine of thrombin (2l), and I'MSthrombin is inactive as a stimulant of platelet aggregation (22). We were able to obviate this problem by employing supramaximal concentrations of thrombin (10 to 20 units/ml) so that, within the time required to carry out an experiment (i.e. 5 min), only partial inactivation of thrombin occurred (9) and sufficient thrombin activity remained intact to produce maximal activation of platelet arachidonic acid metabolism when added to normal platelets. Papain is partially inhibited by I'MSF, but can be completely protected by thiols such as cysteine ( 2 3 ) . Cysteine-activated ( 1 mM) papain treated with PMSF, ABSF, or 2-nitro-4-carboxyphenvl-N,IV-diphenylcarhamate retained its activity against the chromogenic substrate tu-N-benzoyl-DL-arginine-p-nitroanilide. When trypsin was used as the platelet stimulant. neither 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate nor dansvl fluoride inactivated the enzyme under the conditions which led to strong inhibition of trypsin-induced malonvldialdehyde production in platelets (Table 11). T h e lack of direct effect o f dansyl fluoride on trypsin was expected since this compound is known to be a very slow inactivator of that enzyme (19). Ijansyl fluoride did not covalently react with A23187; no dansylated adduct of A23187 was found by thin layer chromatography.
Collagen incubated with dansyl fluoride or I'MSF remained fully active in stimulating platelets after dialysis to rerno\'e free dansyl fluoride or PMSF. We can conclude, therefore. that the effects of these serine esterase inhibitors can he attributed to their action on platelets rather than to any inactivation of the stimulatory agents.
Since the serine esterase inhibit.ors blocked inalonyldialdehyde production elicited by platelet .stimulants but did not affect its production from exogenous arachidonic acid. it was apparent that the release of arachidonic acid rather than any particular step in its metabolism was the site of action. This was substantiated by experiments in platelets whose phospholipids were prelabeled with radioactive arachidonic acid. Upon subsequent stimulation by collagen (24), by proteases such as thrombin or papain (Fig. l), or by A23187 (Fig. 2), 20 to 50% of the [14C]arachidonic acid in phosphatidylcholine (the major pool of labeled arachidonic acid) and a similar proportion of that in phosphatidylinositol was released.
[ lJC]Arachidonic acid in phosphatidylethanolamine and phosphatidylserine was essentially unchanged.
PMSF or 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate inhibited the decrease of radioactive arachidonate in phosphatidylcholine and phosphatidylinositol induced by collagen, papain (Fig. l), A23187 (Fig. 2), or thrombin (20 units/ml) (data not shown). The increases in the radioactive metabolites of arachidonic acid (TXB2, HHT, and HETE) due to stimulation were also inhibited by PMSF (Fig.  1) and 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate (Fig.  2). These inhibitors did not reduce the conversion of exogenous [14C]arachidonic acid to thromboxane B2 and hydroxy acids, or the antimycin A-resistant oxygen burst elicited by exogenous arachidonic acid (data not shown). These results indicate that PMSF and 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate inhibit the release of arachidonic acid from platelet phospholipids rather than affecting the metabolism of arachidonic acid once it is released.
Platelets preincubated with [:'nP]P04 respond to stimulation by thrombin, trypsin or collagen with a large increase in [32P]phosphatidic acid (13). Phosphatidic acid appears to be formed by the phosphorylation of the diglyceride produced by the action of phospholipase C on phosphatidylinositol. PMSF and 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate (each at 1 mM) strongly inhibited the formation of ["'Plphosphatidic acid in intact, stimulated platelets (Fig. 3); inhibition was greater than 90% with collagen and 50 to 75% with thrombin (10 units/&) or trypsin (4 PM). In view of this finding, we investigated the effects of PMSF and 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate on phospholipase C activity in soluble fractions obtained from disrupted platelets. PMSF and 2-nitro-4-carboxyphenyl-N,N-diphenylcarbamate inhibited phospholipase C hydrolysis of phosphatidylinositol over the same concentration range found to be effective in intact platelets (Fig. 4).

DISCUSSION
The mechanisms by which various stimdi induce the release of arachidonic acid from platelet phospholipids have not been unequivocally demonstrated. It has been presumed for some time that arachidonic acid is released by the action of a phospholipase A2 on platelet phospholipids (3, 4). An alternative pathway for arachidonic acid mobilization has recently been proposed (6,8). The earliest event in lipid metabolism in stimulated platelets appears to be the hydrolysis of phosphatidylinositol to diglyceride by the action of a phosphatidylinositol-specific phospholipase C (5). Bell et al. (6) have demonstrated the presence of diglyceride lipase in platelets which acts on the arachidonate-rich diglyceride produced from phosphatidylinositol, resulting in net formation of free arachidonic acid. While this mechanism does not account for the hydrolysis of phosphatidylcholine, it may represent a critical early phase in arachidonic acid release. Another event in stimulated platelets which appears to precede arachidonic acid release is the production of phosphatidic acid (13). No specific inhibitors are known for either diglyceride lipase or the diglyceride kinase which catalyzes production of phosphatidic acid and the relative importance of these two possible fates of diglyceride in the platelet has not been conclusively established.
In this report, we have demonstrated that compounds which inhibit the common first enzyme of these pathways (Le. phosphatidylinositol-specific phospholipase C) block not only phosphatidic acid formation and release of arachidonic acid from phosphatidylinositol but also the release of arachidonic acid from phosphatidylcholine. While it remains to be demonstrated that the serine esterase inhibitors do not directly block other enzymes involved in these metabolic pathways h e . phospholipase Az, diglyceride lipase, or diglyceride kinase), our findings are consistent with the proposed central role of phospholipase C in the stimulus-induced mobilization of platelet arachidonic acid. Since arachidonate release from phosphatidylcholine is also blocked by these inhibitors, it is possible that some product of phosphatidylinositol metabolism (e.g. diglyceride, arachidonic acid or its metabolites, or phosphatidic acid) is an activator of the putative phospholipase Az. Furthermore, since the serine esterase inhibitors block arachidonic acid mobilization by a wide range of stimulating agents, the target(s) of the inhibitors must be common to all stimulatory pathways. Mobilization of intracellular Cas* is likely to be important in the release of arachidonic acid since this cation is an activator of both phospholipase A2 (3) and phospholipase C (5, 8). However, since serine esterase inhibitors block arachidonic acid release brought about by A23187, which presumably releases intracellular Cas+ by a direct action as an ionophore, it appears that an enzymatic step beyond Ca2' mobilization is blocked.
We conclude, therefore, that 2-nitro-4-carboxyphenyl-N,Ndiphenylcarbamate and PMSF can block formation of free arachidonic acid in platelets, in part at least, by inhibiting the phosphatidylinositol-specific phospholipase C. This is the first class of enzyme inhibitors which has been shown to affect this important enzyme. Further studies with the serine esterase inhibitors should be useful for investigating the enzymatic properties of the phosphatidylinositol-specific phospholipase C and for establishing in more detail the initial enzymatic pathways for arachidonic formation in platelets.