Calmodulin-independent inhibition of platelet phospholipase A2 by calmodulin antagonists

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 246, No. 2, May 1, pp. 699-709,1986

Calmodulin-Independent Inhibition of Platelet Phospholipase A2 by Calmodulin Antagonists TSUYOSHI WATANABE,*,’ YOSHIAKI SHOJI KUME,t3 CHIKAYUKI *First

HASHIMOTO,* TAM10 TERAMOTO,* NAITO,$ AND HIROSHI OKA*

Department of Internal Medicine and tCentm1 Clinical Laboratory, Tokyo, and $Department of Internal Medicine, TokyeTeishin

Faculty Hoqital,

Received July 22,1985, and in revised form November

of Medicine,

University

of

Tokyo, Japan

13,1985

We tested the effects of calmodulin, two types of calmodulin antagonists, and various phospholipids on the phospholipase Az activities of intact platelets, platelet membranes, and partially purified enzyme preparations. Trifluoperazine, chlorpromazine (phenothiazines) and N-(6-amino-hexyl)-5-chloro-l-naphthalenesulfonamide (W-7), at concentrations which antagonize the effects of calmodulin, significantly inhibited (a) thrombin- and Ca2+ ionophore-induced production of arachidonic acid metabolites by suspensions of rabbit platelets and (b) Ca2+-induced arachidonic acid release from phospholipids of membrane fractions, but not (c) phospholipase A2 activity in purified enzyme preparations. The addition of acidic phospholipids, but not calmodulin, stimulated phospholipase A2 activity in purified enzyme preparations while decreasing its Km for Ca’+. The dose-response and kinetics of inhibition by calmodulin antagonists of acidic phospholipid-activated phospholipase A2 activity in purified preparations were similar to those of Ca2+-induced arachidonic acid release from membrane fractions. Calmodulin antagonists were also found to inhibit Ca2’ binding to acidic phospholipids in a similar dose-dependent manner. Our results suggest (a) that the platelet phospholipase AZ is the key enzyme involved in arachidonic acid mobilization in platelets and is regulated by acidic phospholipids in a Ca2+-dependent manner and (b) that calmodulin antagonists 0 19%Academic inhibit phospholipase Az activity via an action on acidic phospholipids. Press,Inc.

Platelets activated by various stimuli form arachidonic acid (AA)3 metabolites

such as thromboxane (TX) AZ, 12-hydroxy5,8,10-heptadecatrienoic acid (HHT), and 12-hydroxy-5,8,10,14-eicosatetraenoic acid (HETE) (1). The first step in AA metabolism is reported to be the liberation of free AA mainly from phosphatidylcholine (PC) and phosphatidylinositol (PI) (Z-4). There are two distinct mechanisms for AA release, but it is unclear which of two plays the major role in AA mobilization in stim-

’ To whom correspondence should be addressed at the Department of Biochemistry, Michigan State University, East Lansing, Mich. 48824. * Present address: The Central Clinical Laboratory, Yamanashi Medical College, Yamanashi, Japan. a Abbreviations used: AA, arachidonic acid, TX, thromboxane; HHT, 12-hydroxy-5,8,10-heptadecatrienoic acid, HETE, 12-hydroxy-5,8,10,14-eicosatetraenoic acid, PL, phospholipase; PC, phosphatidylcholine; PI, phosphatidylinositol; PE, phosphatidylethanolamine; PS, phosphatidylserine; PA, phosphatidic acid, SM, sphingomyelin; DG, diglyceride; TMB-8, 8(N~-dimethylamino)-octyl-3,4,5-trimethoxyben~ate; CAL, ealmodulin; BSA, bovine serum albumin; SDS,

sodium dodecyl sulfate; W-7, N-(B-aminohexyl)-Bchloro-1-naphthalenesulfonamide; W-5, N-(&aminohexyl)-1-naphthalenesulfonamide, HCI; PG, prostaglandin. 699

0003-9861/86 $3.00 Copyright All rights

B 1986 by Academic Press. Inc. of reproduction in any form reserved.

700

WATANABE

ulated platelets. In one pathway phospholipase (PL) A2 acts directly on PC, PI, and other phospholipids to release AA. In an alternate pathway the sequential activities of PI-specific PL C, diglyceride (DG) lipase, and monoglyceride lipase acting on PI effect release of AA. Both PL A2 and PI-specific PL C have been isolated from human and rabbit platelets (5-15). Both enzymes require Ca2+ when assayed in vitro (7, 9, 12-15). The Ca2+ionophore A23187 evokes AA release from membrane phospholipids of platelets. The effects of A23187 and of physiological stimuli such as thrombin are blocked by %(N,N-dimethylamino)-octyl3,4,5-trimethoxybenzoate (TMB-8), a putative intracellular Ca2+ antagonist (1618). These results suggest that Ca2+ mobilization is important in initiating AA release in stimulated platelets. Calmodulin (CAL) is a Ca2+-binding protein found in most eucaryotic cells including platelets (19-21). CAL mediates many of the effects of Ca2+on various enzymes and cellular reactions (19). Phenothiazines and naphthalenesulfonamides are reported to be CAL antagonists and have been shown to inhibit stimulus-induced platelet reactions including AA liberation (22-26). Conflicting reports have been published regarding the involvement of CAL in Ca2+-dependent stimulation of PL A2 activities from snake venom, pancreas, and crude platelet membranes (2729). Most recently, Withnall et al. reported that partially purified platelet PL A2 could not be stimulated by exogenous CAL (30). In contrast, Walenga et al. showed that in thrombin-treated human platelets agents that inhibit the effects of CAL inhibit the formation of AA metabolites but not the formation of phosphatidic acid. (PA). These authors suggested that PL A2 accounted for most of the AA liberated and that CAL’was required for its activation (31-32). However, phenothiazines bind proteins other than CAL in a Ca2+-dependent manner (33) and have other effects on membrane phospholipids (34). Thus, phenothiazines appear not to be specific for CAL. Recently, many Ca2+-dependent reactions in stimulated cells, including plate-

ET AL.

lets, have been reported to depend upon the presence of acidic phospholipids such as phosphatidylserine (PS) (35-36). PL A2 isolated from both rabbit and human platelets has been reported to be stimulated by the addition of PA, PS, or PI (10, 37). PA has also been reported to be an endogenous Ca2’ ionophore (38). However, the relationships between these phospholipids and CAL antagonists in the Ca2+-dependent regulation of AA mobilization in stimulated platelets remain to be clarified. In this study, we tested the effects of CAL, acidic phospholipids, and two types of CAL antagonists on the PL A2 activities of rabbit platelets. Various preparations of PL A2 and PC substrates were tested in different combinations. The results suggest (a) that PL A2 is the key enzyme responsible for AA release in stimulated platelets and is selectively inhibited by CAL antagonists, (b) that platelet PL A2 depends on acidic phospholipids, but not on CAL, for Ca2+-dependent activation, and (c) that the inhibition of PL A2 by CAL antagonists is caused by their interaction with acidic phospholipids and not by their interaction with CAL. MATERIALS

AND

METHODS

Afuterials. [“CjAA (52.1 mCi/mmol), ‘%aCla, and Calmodulin ‘%I-RIA Kits were purchased from New England Nuclear. CAL (from hog brain) was from Boehringer Mannheim. Trifluoperazine was a gift from Yoshitomi Pharmaceutical Company, Tokyo, Japan. N-(6-aminohexyl)-&chloro-l-naphthalenesulfonamide (W-7) and N-(6-aminohexyl)-l-napthalenesulfonamide, HCl (W-5) were obtained from Rikaken Company, Tokyo, Japan. Chlorpromazine, L-IX-PA (from egg yolk), L-a-lyso-PC (from egg yolk), L-a-phosphatidylethanolamine (PE) (from egg yolk), L-ol-lyso-PE (from egg yolk), sphingomyelin (SM) (from bovine brain), cholesterol, AA, diolein, PL Az (bee venom), and essentially fatty acid-free bovine serum albumin (BSA) were products of Sigma. L-~-PC (from egg yolk), L-(Y-PS (from bovine brain), and ~-a-P1 (from soybean) purchased from Sigma were further purified by silicic acid column chromatography. Each purified sample migrated as a single spot on thin-layer chromatography. Thrombin (bovine) was from Mochida Pharmaceutical Company, Tokyo, Japan. Caa+ ionophore, A23187, was from Calbiochem. Silica gel G and H thinlayer chromatography plates were from Merck. Sephadex G-100 was from Pharmacia. Unlabeled pros-

PLATELET

PHOSPHOLIPASE

AZ AND

taglandins (PGs) were gifts from Ono Pharmaceutical Company, Osaka, Japan. All other chemicals were of reagent grade and used without further purifications. Thrombin- or A2.9187~induced AA metabolism in [14CjAA-labelecl platelet suspensions. Platelet-rich plasma was prepared from fresh titrated rabbit blood obtained from the carotid artery (39). Washed platelets were prepared as described previously (40) in the presence of 1 mM EDTA and then were resuspended in one-fifth the original volume of platelet resuspension buffer as described by Baenziger and Majerus (41). To prepare [‘“CIAA-labeled platelets, washed platelets were suspended in platelet resuspension buffer containing 0.25% BSA and incubated with [“C]AA (5 &i/l0 ml) at 3’7°C for 60 min. The labeled platelets were then washed three times with cold platelet washing buffer containing 1 mM EDTA as described by Baenziger and Majerus (41). The cells were collected and suspended in platelet resuspension buffer. The labeled platelet suspension (1 ml) containing 1.75 X lo9 platelets and 5 ~1 of an aqueous solution of the agent to be tested were preincubated for 2 min and then stimulated with 25 ~1 of thrombin (200 U/ml of resuspension buffer) or 25 ~1 of A23187 (80 pmol/ml of 0.1% dimethyl sulfoxide in resuspension buffer) at 37°C. Incubations were stopped by adding 3.75 ml of CHC&:methanol (1:2, v/v). Next, 1.25 ml of CHC& and 1.25 ml of 2 M KC1 containing 5 mM EDTA were added. The CHCl, phase was carefully removed, and the aqueous phase was acidified to pH 3 with 1 N HCI and reextracted twice with 2.5 ml of CHCls. The combined CHCla phases were dried under Nz. The extracted materials were applied to silica gel G thin-layer plates and the plates were developed in the upper phase of ethyl acetate:isooctane:acetic acid: water (90:50:20:100, v/v/v/v) to separate AA metabolites and PA (7,31). Phospholipids were analyzed on silica gel H plates in the two-dimensional solvent system described by Billah and Lapetina (7) or in a onedimensional chromatographic system consisting of CHCls:methanol:acetic acid:water (100:60:15:8, v/v/v/ v). Standards were made visible with Iz vapor. The positions of radioactive bands were determined by autoradiography. Each spot was scraped into a scintillation vial, and the radioactivity in each band was determined with Packard 400CL/DL liquid scintillation counter (ca. 92% efficiency). of platelet &sates and membrane fracPreparatim tions. Washed platelets were resuspended in 5 mM potassium phosphate, pH 7.0, and sonicated at maximum power four times; 15 s each with a Supersonic Vibrator Tomy Model UR-150 P, Tominaga Work Company, Tokyo, Japan. The resultant homogenate was used as the platelet lysate. To prepare platelet membrane fractions, platelet lysates were centrifuged at 10,500g for 60 min, and the resultant pellet was resuspended in 5 mM potassium phosphate, pH 7.0. Evaluation of endogenous PL Ae activity toward

CALMODULIN

ANTAGONISTS

701

platelet-derived labeled phospholipids. [‘*C]AA-labeled platelets were prepared and then used to prepare platelet lysates and membrane fractions as described above in the presence of 1 mM EDTA. An aliquot (50 ~1) of the [‘QAA-labeled platelet lysate or membrane fraction was added to 900 ~1 of 100 mEd Tris-HCl, pH 7.4, 1 mM EDTA. The suspension was preincubated with 5 ~1 of an aqueous solution of the agent to be tested and the assay was initiated by adding 50 ~1 of 100 mM CaClz and was incubated at 37°C for 60 min. Assays were stopped by adding CHCls:methanol (1:2, v/v). Extractions of lipids from the reaction mixtures, thin-layer chromatography, and quantitation of radioactivity were performed as described above. of platelet PL AS. Platelet memPartial purification branes were prepared from 350-400 ml blood and suspended in 5 ml of 1.15% KCI. PL AZ activity was solubilized by the pellet extraction procedure described by Jesse and Franson (12). The enzyme activity was further purified by column chromatography on a Sephadex G-100. Partial purifications of 435- to 651-fold with recoveries of 56-78% were attained using this procedure. Assay of PL A, activity with [‘4CjAA-labeled PC as the substrate. [“C]AA-labeled PC was enzymatically synthesized from lyso-PC by using human red cell lysate according to the method of Robertson and Lands (42) and was then purified by thin-layer chromatography. The specific activity of radiolabeled PC was adjusted to 2 X lo5 cpm/rmol with purified nonisotopic PC. The standard assay mixture contained 100 mM Tris-HCl, pH 7.4,200 nmol of labeled PC, 2.5 mM CaClz, and the enzyme preparation in a total volume of 0.5 ml. Phospholipid substrate and activators were dispersed by four 1-min sonic pulses at the maximum power. This mixture, from which Ca’+ was omitted, was preincubated for 2 min with an aqueous solution of the agent to be tested. The reaction was started by the addition of Ca”, continued at 37°C for 1 h, then stopped by the addition of 1.88 ml of CHC&: methanol (1:2). The extraction of lipids from the reaction mixture, separation of the extracted materials by thin-layer chromatography, and quantitative measurement of radioactivity in each band were performed as described above. of CAL Extractions of CAL Radia’mmunoassay from platelet lysates, membrane fractions, and the purified enzyme preparations were performed essentially according to the method described for rabbit polymorphonuclear leukocytes by Chafouleas et al. (43). Radioimmunoassay of CAL was performed at four different dilutions according to the manual for the Calmodulin ‘%I-RIA Kit. “Ca’+ binding to authentic lipids. tiCa2f binding to lipids were measured according to the procedure of Blaustein and Goldman (44). One milliliter of a solution containing 116 mM NaCl, 2.5 mM KCl, 1.0 mM oCaC1, (1 &i), and 2.0 mM Tris-HCl, pH 7.4, was

702

WATANABE

combined with 2 ml of CHCls:methanol (21, v/v) containing 1.0 mg of an authentic lipid. After 5 ~1 of water or an aqueous solution of antagonist was added, the mixture was shaken for 10 min and centrifuged at 1OOOgfor 5 min. %a’+ binding to lipid was calculated from the percentage of radioactivity in the CHCl, phase. MisceL!aneous. SDS-polyacrylamide slab gel electrophoresis was performed according to the method of Laemmli (45) with slight modifications. Gels were subjected to silver staining by using the method of

ET AL. Oakley et al (46). Lipid phosphorus in each band on the thin-layer chromatographic plate was determined by the method of Chen et aL (47). Protein concentrations were determined according to the method of Lowry et al (48) with BSA as the standard. RESULTS

E.&e&s of CAL antagksts on AA metabolism in [lbCJAA-labeled platelets. When

suspensions

of [14C]AA-labeled

intact

4

Time

hd

Time hid

Time

Time

hid

hid

FIG. 1. Effects of trifluoperazine on AA metabolism in thrombin-activated (A, B) and A2318’7activated (C, D) platelets labeled with [14C]AA. Experiments were performed as described under Materials and Methods. Before the addition of thrombin or A23187, about 88% of the total radioactivity (101,010 cpm/ml of suspension) was present in phospholipid fractions. Platelet suspensions were preincubated with (B, D) or without (A, C) 50 MM trifluoperazine and stimulated by adding thrombin (5 U/ml) (A, B) or A23187 (2 1~) (C, D). After the indicated incubation times, the amounts of radioactivities in fractions corresponding to total phospholipids (A), free AA (O), HETE (O), TX & + HHT (0), DG (X), and PA (0) were determined. All the values are the means of triplicate samples.

PLATELET

0

50 100 Antagonist

250 ($4)

PHOSPHOLIPASE

0

50 100 Antopnist

Az AND

250 (pM)

FIG. 2. Effects of varying concentrations of CAL antagonists on the formation of AA metabolites (A) and PA (B) by thrombin-treated platelet suspensions, PL A2 activity toward synthetic PC substrate (C) and “Ca*’ binding to authentic PA (D). (A, B) Experiments were performed as described under Materials and Methods. Before the addition of thrombin, about 93% of the total radioactivity of labeled platelets (93,020 cpm/ml of suspension) was present in phospholipid fractions. After 5 min incubation, radioactivities in fractions corresponding with HETE, TX Bz. HHT + total PGs (A) and PA (B) were determined. The control values of radioactivity (lOO%)in total AA metabolites (A) and PA (B) without CAL antagonists after 5 min incubation with thrombin were 7,342 and 1,650 cpm, respectively. All the values are the means of triplicate samples. (C) The purified enzyme preparations (2 pg of protein) were used for the assay of PL A2 activity toward sonicated PC substrate without PA (solid lines) or PC substrate sonicated with 400 PM PA (dotted lines). Control activities (lOO%)toward PC substrate without PA and toward PA-added PC substrate were 4.2 and 19.1 pmol/mg of protein/h, respectively. All the values are the means of duplicate samples. (D) “Ca2+ binding to authentic PA was measured in the presence of indicated final concentrations of CAL antagonists. When CAL antagonists were not present (lOO%)or when PA was omitted from the organic phase, 0.519 or 0.000 mol of &Ca’+ was bound per mole of PA, respectively. Almost the same results were obtained when PS or PI was used instead of PA. All the values are the means of triplicate samples.

CALMODULIN

ANTAGONISTS

703

platelets were treated with thrombin (5 U/ ml) or A23187 (2 PM), the amount of radioactivity in phospholipids decreased and the amount of radiolabeled AA metabolites (i.e., TX Bz, HHT, and HETE) increased. Increases in these metabolites were timedependent, reaching plateau levels in 2 min. Significant transient accumulations of DG and time-dependent accumulation of PA were observed when rabbit platelets were stimulated with thrombin but not A23187 (Figs. 1A and C); similar effects have been seen by using human platelets (49). Maximal changes in lipid turnover and eicosanoid production occurred with 0.5-1.0 mM Ca’+. All subsequent experiments were performed using 1 mM Ca2+ in the assay mixtures. As shown in Figs. 1B and D, 50 I.LM trifluoperazine strongly inhibited both thrombin- and A23187-induced formation of AA metabolites. Interestingly, the accumulation of PA induced by thrombin was increased about twofold in the presence of 50 PM trifluoperazine (Fig. 1B). Both phenothiazines (trifluoperazine and chlorpromazine) and a naphthalenesulfonamide derivative (W-7) inhibited the production of AA metabolites induced by thrombin and by A23187 in a dose-dependent manner (Fig. 2A); another naphthalenesulfonamide derivative, W-5, which has been reported to be a much weaker CAL antagonist than W-7, had only a slight effect (Fig. 2A). Dose-inhibition curves were essentially the same for both A23187- and thrombin-activated platelets. As shown in Fig. 2B, concentrations of CAL antagonists that completely inhibited formation of AA metabolites stimulated the incorporation of radioactivity into PA. In order to determine the effect of CAL antagonists on the rates of biosynthesis of PA and DG, we determined the early time course of AA metabolism in platelets labeled with [14C]AA and then activated with thrombin. As shown in Fig. 3B, 50 PM trifluoperazine inhibited the rates of AA release from PI

W-5 (A), W-7 (O), chlorpromazine (0), and trifluoperazine (0) were used as CAL antagonists for all experiments.

WATANABE

(A)

ET AL.

(B)

0I

Time (se-z)

Time (set)

Time (SC)

FIG. 3. Effect of trifluoperazine on the early time course of AA metabolism in thrombin-treated platelet suspensions labeled with [‘%]AA. The experimental conditions were the same as those used for Fig. 1 except that the total radioactivity incorporated in platelets was 103,300 cpm/ml of suspension. The concentrations of trifluoperazine used are 0 PM (A), 50 PM (B), and 250 PM (C). PC (Q), PI (:), PA (O), DG (X), free AA (O), TX Bz + HHT (0), HETE (0).

and PC and the rates of formation of oxygenated AA metabolites; however, trifluoperazine did not significantly change the rates of PA or DG formation. A higher dose (250 @MI) of trifluoperazine almost completely inhibited PA accumulation and AA release from PC but only partially inhibited AA release from PI and subsequent DG formation (Fig. 3C).

Properties of d$Fermt preparations of PL A,. As shown in Fig. 4, similar pH profiles were found for Caz+-induced AA release from phospholipids of platelet lysates as well as for PL A2 activities toward PC substrate of membrane fractions and partially purified enzyme preparations. Phenothiazines inhibited PL Az activities at or below pH 8, but not at the optimal pH (Fig. 4). W-7 inhibited the activities at all pH values (data not shown). For this reason, we tested the effects of these compounds on platelet PL A2 activities at the physiological pH 7.4.

Eflects of exogenous CAL and lipids cm PL A2 activities. The removal of CAL from the partially purified preparations of PL Az was verified by SDS-gel electrophoresis. No protein band with a relative molecular mass of 17,000 was present when 10 pg of protein was applied to the gel and the gel was stained with A,NO,. Moreover, no CAL could be detected by radioimmunoassay (co.5 ng/mg of membrane protein). The amounts of CAL in platelet lysates and membrane fractions were determined by radioimmunoassay to be 188 and 41 ng/mg of protein, respectively. Phospholipid contents were estimated to be 0.22, 0.45, and CO.03 mg/mg of protein in platelet lysates, membrane fractions, and partially purified PL AZ preparations, respectively. We first tested the effect of added CAL on PL Az activities of platelet membrane fractions and purified enzyme preparations. CAL had no effect on PL AZ activities with either platelet-derived or synthetic

PLATELET

PHOSPHOLIPASE

Az AND

30 t

6

7

8

9

IO

II

PH

FIG. 4. Effect of pH on PL Az activities in platelet lysates and partially purified enzyme preparations. Incubations were performed as described under Materials and Methods except for the buffers used. Each assay mixture contained 0.5 mg of protein for Ca2+induced release of labeled AA from phospholipids of platelet lysates in the presence (0) or absence (0) of 50 PM trifluoperazine, 1.0 mg of protein for PL Az activity toward synthetic PC in platelet lysates (Cl), and 2 pg of protein for PL Az activity toward synthetic PC in purified enzyme preparations (A). The buffers used were 100 mM sodium phosphate (pH 6 and 7), 100 mM Tris-HCl (pH S-9.5), and 100 rnM glycineNaOH (pH 10-11). Activities are expressed as percentage hydrolysis of total labeled phospholipids per hour for Caz+-induced AA release from phospholipids of platelet lysates, nmol/mg of protein per hour for PL Az activity toward synthetic PC in platelet lysates, and rmol/mg of protein per hour for PL A2 activity in purified enzyme preparations. All the values are the means of duplicate samples.

phospholipid substrates at any Ca2+ concentrations (1 ~~-10 mM) in the presence or absence of (a) acidic phospholipids (Figs. 5A-C) or (b) CAL antagonists (data not shown). We subsequently tested the effects of various lipids (i.e., cholesterol, PE, PI, PS, PC, SM, lyso-PC, PA, and diolein) on the purified PL A2 preparations. PA and PS caused a dose-dependent activation of the enzyme. These results are similar to those of Kannagi et al. (lo), except that PI had only a slight activating effect in our experiments (data not shown).

CALMODULIN

ANTAGONISTS

705

Half-maximal activation of PL Az activity with platelet-derived phospholipids occurred at about 50 PM Ca2+; in contrast, about 500 PM Ca2’ was required for halfmaximal activation of both the partially purified PL A2 and PL A2 of platelet membranes when these enzymes were analyzed with exogenous PC (Figs. 5A and B). Sodium deoxycholate (0.1%) activated the partially purified PL A2 but did not change its K, for Ca2+(Fig. 5C). However, the addition of PA and/or PS (400 PM) to the purified enzyme preparations along with the PC substrate dramatically lowered the & for Ca2+ to a value similar to that found for the PL A2 activity of membrane fractions measured using platelet-derived phospholipids (Fig. 5B). Effects of CAL antagonists on PL A, activities. CAL antagonists inhibited PL A2

activity in platelet lysates and membrane fractions, as measured by release of labeled AA from endogenous phospholipids (data not shown). This inhibition exhibited almost the same dose-dependence as seen with intact platelets (Fig. 2A). However, these CAL antagonists caused less inhibition of PL A2 activity, as measured by using exogenous PC as the substrate (Fig. 2C). This was true whether the purified enzyme, platelet lysates, or membrane fractions were used as the source of PL A2. With exogenous PC as the substrate, inhibition of PL A2 activity when two phenothiazines were used was never greater than 50% even at 1 mM. Only W-7 (500 wM) completely inhibited PL A2 activity. However, when PA and/or PS (400 PM) were added to the radioactive PC substrate suspension prior to use in the assay, both types of CAL antagonists inhibited the activity of the purified enzyme preparations in the same dose-dependent manner observed for inhibition of AA release from labeled phospholipids (Figs. 2A and C). Trifluoperazine appears to exert a mixed type of inhibition of PL A2 activity, as measured by Ca2+-induced release of labeled AA from platelet membrane fractions (Fig. 5A). In the absence of PA or PS, high concentrations of trifluoperazine seem to compete with Ca2+ to inhibit the partially purified PL A2 (Fig. 5C). However,

706

WATANABE

ET AL.

(8)

.I/(Coz+,

mM)

(Iz=i=G

25

._ Ca’+ (mM)

- ‘0

IO I/(Co”,

mM)

FIG. 5. Effects of trifluoperazine and CAL on Ca2+ requirements of (A) Ca’+-induced AA release from platelet-derived phospholipids of membrane fractions and (B, C) PL A2 activity toward synthetic PC in purified enzyme preparations. Incubations were performed as described under Materials and Methods except at the indicated Cap+ concentrations; 0.5 mg of platelet membrane protein or 4 pg of purified enzyme protein was used to measure the activities, (A) The concentrations of trifiuoperazine

PLATELET

PHOSPHOLIPASE

Az AND

CALMODULIN

ANTAGONISTS

trifluoperazine caused a mixed type of inhibition of the partially purified PL A2 when PA was added to the PC substrate suspension (Fig. 5B). Acidic phospholipids and CAL antagonists did not change the time course of PL A2 activity, as rates remained linear for at least 60 min. The inhibitory effects of CAL antagonists on PL A2 activities could not be duplicated by PE, another amphipathic amine. It provided the evidence that not all amphipathic amines are specific PL A2 inhibitors. %‘a’+ binding to lipids. We measured 45Ca2+binding to PE, PI, PS, PC, SM, lysoPC, lyso-PE, PA, cholesterol, and diolein. PA, PS, and PI, but not other lipids, bound significant amounts of 45Ca2+.As shown in Fig. ZD, binding of 45Ca2+to PA, PS, and PI was inhibited by CAL antagonists in a dose-dependent manner similar to that seen for inhibition of eicosanoid formation by stimulated platelets.

707

tent of PA accumulation in thrombintreated platelets (Fig. 2B). However, the initial rates of PA and DG formation were not affected by CAL antagonists (Fig. 3B). Moreover, PI breakdown and DG formation were observed even at concentrations of CAL antagonists which inhibited PA accumulation (Fig. 3C). We conclude from these observations that CAL antagonists, at concentrations that purportedly inhibit the effects of CAL, inhibit PL A2 but not PI-specific PL C in activated platelets. The increased accumulation of PA caused by CAL antagonists in thrombin-activated platelets may be the result of the inhibition of PL A2 acting on PA (15). Both PL A2 and PI-specific PL C are reported to be dependent upon Ca2+for activity when assayed in vitro (7,9,12,13). PL A2 is a membranebound enzyme with a pH optimum of 8.511.5. PI-specific PL C is a cytosolic enzyme and has a neutral pH optimum (6-9). The PL AZ activities measured (a) with exogenous PC and the partially purified PL AZ, DISCUSSION platelet lysates, or membrane fractions and PL A2 and PI-specific PL C appear to be (b) Ca2+-dependent release of AA from enimportant in mobilizing AA following dogenous phospholipids of platelet memstimulation of platelets (2-7). As shown in branes showed the same alkaline pH opFigs. 1A and C, significant increases in DG timum reported for PL A2 (Fig. 4). CAL and PA levels were observed in thrombinantagonists caused similar patterns of inactivated platelets but not in A2318’7-ac- hibition for (a) production of AA metabtivated platelets. This suggests that in olites by intact platelets, (b) PL A2 activity A23187-activated platelets, both PC and PI measured by release of AA from plateletare hydrolyzed by PL A2 activities, while derived phospholipids, and (c) PL A2 activin thrombin-activated platelets both PL A2 ity toward synthetic substrate mixtures and PI-specific PL C seem to be activated supplemented with PS or PA (Figs. 2A and (49). As shown in Figs. 1 and 2A, CAL an- C). Taken together, these results suggest tagonists inhibit both thrombin- and that most of the radiolabeled AA liberated A23187-induced AA release from phospho- and metabolized in stimulated platelets lipids in intact platelets in the same dose- results from the action of a PL AZ-type acdependent manner. CAL-antagonists at the tivity, not via the pathway involving PIconcentrations which almost completely specific PL C, DG lipase, and monoglyceride inhibit AA mobilization increased the ex- lipase.

used were 0 p~ (0; CAL not added; A, 1 &M CAL added), 25 PM (0, CAL not added; *, 1 PM CAL added), 50 PM (X, CAL not added), and 100 PM (0, CAL not added). (B) The substrate for purified enzyme preparations was dispersed by sonication (a, CAL not added, 0,l PM CAL added) without PA or by sonieation with 400 NM PA (0, trifluoperazine not added; *, 25 PM trifluoperazine added; Cl, 100 PM trifluoperazine added). (C) The substrate was dispersed by sonication (0, CAL not added; 0,l PM CAL added) or solubilized with 0.1% sodium deoxycholate (A, CAL not added; A, 1 pM CAL added; X, 250 pM trifluoperazine added; +, 1 PM CAL and 250 pM trifluoperazine added). All the values are the means of duplicate samples. Inserts show double reciprocal plots of data.

708

WATANABE

Consistent with the report of Withnall et al. (30), we found that exogenous CAL failed to stimulate PL AZ activity in partially purified enzyme preparations. In contrast, addition of acidic phospholipids such as PA or PS stimulated PL A2 activity. This stimulation was accompanied by a decrease in the K, for Ca2+(Fig. 5B). Deoxycholate also stimulated PL A2 but without changing the K, for Ca2+(Fig. 5C). Acidic phospholipids, especially PA formed from DG, might be important in vivo in regulating PL A2 activity and AA metabolism in platelets. A physiological role for acidic phospholipids in this system is consistent with the observations that the inhibition by CAL antagonists of PL A2 activity measured when synthetic substrate is supplemented with an acidic phospholipid parallels the inhibition by CAL antagonists of AA metabolites released by activated platelets (Figs. 2A and C). Although CAL did not stimulate PL A2, both classes of CAL antagonists inhibited PL Az activity toward platelet-derived phospholipids in platelet lysates and membrane fractions. This occurred at the same concentrations of CAL antagonists which inhibit formation of AA metabolites by stimulated platelets (Fig. 2A). In contrast, CAL antagonists had only a slight inhibitory effect on CAL-free, purified PL A2 (straight lines in Fig. 2C). As shown in Fig. 5C, high concentrations of trifluoperazine and chlorpromazine appear to compete with Ca2+for binding to PL A2. W-7 seemed to have the more potent direct effect on PL A2. Withnall and Brown reported that the competitive inhibition of pancreatic PL A2 by trifluoperazine was not overcome by excess Ca2+(28). It is probable that the differences between their results and ours are due to the difference between secreted PL A2 and membrane-bound PL A2. However, purified PL As, when assayed in the presence of PA or PS, was inhibited by CAL antagonists in almost the same dose-dependent and kinetic fashion (a mixed type of inhibition) as Ca2+-induced AA release from platelet-derived phospholipids (dotted lines in Figs. 2C and 5A and B).

ET AL.

CAL antagonists modulated the %a2+binding activity of authentic acidic phospholipids in a dose-dependent manner. The order of potency of these agents with respect to inhibition of 45Ca2f binding to acidic phospholipids was the same as that for inhibition of AA release from plateletderived phospholipids (Fig. 2D). Another naphthalenesulfonamide derivative, W-5, which is only one-tenth as active a CAL antagonist as W-7, had only slight effects on 45Ca2+binding to acidic phospholipids; moreover, W-5 caused little inhibition of PL AZ activities measured in enzyme preparations by hydrolysis of phospholipid or by release of AA metabolites from stimulated platelets. These results suggest that these CAL antagonists have direct effects on the Ca2+-binding site of acidic phospholipids. This may be related to the inhibitory effects of CAL antagonists on platelet PL A2 activities in the presence of exogenous or endogenous acidic phospholipids. PA has been reported as a putative endogenous Ca2+ ionophore (38). Thus, CAL antagonists may act by inhibiting Ca2+ mobilization by PA in thrombin-activated platelets. CAL antagonists may also directly inhibit PL A2 activity, especially at high concentrations. However, it is more probable that in the presence of Ca2+,acidic phospholipids stimulate PL A2 activity and that CAL antagonists inhibit PL A2 activity by interfering with the stimulatory capacity of acidic phospholipids. ACKNOWLEDGMENTS We are grateful to Dr. G. Toda and Dr. H. Kato of our department for their suggestions and to Dr. W. Smith and Dr. D. Dewitt of the Department of Biochemistry, Michigan State University, for their critical reading of the manuscript.

REFERENCES 1. SAMUELSSON, B., GOLDYNE, M., GRANSTR~M, E., HAMBERG, M., HAMMARSTR~M, S., AND MALMSTEN, C. (19’78) Anna Rev. Bioch.em 47.997-1029. 2. KUNZE, H., AND VOGT, W. (1971) Ann N.Y. Acad Sci. 180,123-125.

PLATELET

PHOSPHOLIPASE

Az AND

3. BILLS, T. K., SMITH, J. B., AND SILVER, M. J. (1976) Biochim. Biophys. Acta 424,303-314. 4. BILLS, T. K., SMITH, J. B., AND SILVER, M. J. (1977) J. Clin Inwst. 60,1-6. 5. RI~ENHOUSE-SIMMONS, S. (1979) J. Clin. Invest. 63,580~587. 6. BELL, R. L., STANFORD, N., KENNERLY, D. A., AND MAJERUS, P. W. (1980) Adv. Prostaglandin Thrwmboxane Res. 6,219-224. 7. BILLAH, M. M., LAPETINA, E. G., AND CUATRECASAS, P. (1980) J. Biol C&m. 255,10227-10231. 8. APITZ-CASTRO, R. J., MAS, M. A., CRUZ, M. R., AND JAIN, M. K. (1979) Biochem Biophys. Res. Cummun. 91,63-71. 9. KANNAGI, R., AND KOIZUMI, K. (1979) Arch B&hem Biophys. 196,534~542. 10. KANNAGI, R., AND KOIZUMI, K. (1979) Biochim. Biophys. Acta 556,423-433. 11. KANNAGI, R., KOIZUMI, K., AND MASUDA, T. (1981) J Biol. Chem. 256,1177-1184. 12. JESSE, R. L., AND FRANSON, R. C. (1979) B&him. Biophys. Acta 575,467-470. 13. MACUO, G., CHAP, H., AND DOUSTE-BLAZY, L. (1979) FEBS Lett 100,367-370. 14. BILLAH, M. M., AND LAPETINA, E. G. (1982) J. Biol. Chem. 257,5196-5200. 15. BILLAH, M. M., LAPETINA, E. G., AND CUATRECASAS, P. (1981) J. Biol. Chem 256,5399-5403. 16. PICKET, W. C., JESSE, R. L., AND COHEN, P. (1977) Biochim. Biophys. Aeta 486,209-213. 17. RITTENHOUSE-SIMMONS, S., AND DEYKIN, D. (1977) J. Clin Invest. 60, 495-498. 18. RI~ENHOUSE-SIMMONS, S., AND DEYKIN, D. (1978) B&him Biophys. Acta 543,409-422. 19. CHEUNG, W. Y. (ed.) (1980) Calcium and Cell Function Vol. I. Calmodulin, Academic Press, New York. 20. WHITE, G. C., LEVINE, S. N., AND STEINER, A. N. (1981) Amer. .J. Hematol. 10,359-367. 21. KAKIUCHI, S., SOBUE, K., YAMAZAKI, R., KAMBAYASHI, J., SAKON, M., AND KOSAKI, G. (1981) FEBS L&t 126,203-207. 22. LEVINE, R. M., AND WEISS, B. (1978) B&him. Bi+ phys. Ada 540,197-204. 23. KINDNESS, G., WILLIAMSON, F. B., AND LONG, W. F. (1980) Thromb. Res. 17,549-554. 24. WHITE, G. C., AND RAYNOR, S. T. (1980) Thromb. Res. l&279-284. 25. KAMBAYASHI, J., MORIMOTO, K., HAKATA, N., KoBAYASHI, T., AND KOSAKI, G. (1981) Thromb. Res 22.553-558. 26. NISHIKAWA, M., AND HIDAKA, H. (1982) J. Cl& Invest. 69,1348-1355. 27. WONG, P. Y.-K., AND CHEUNG, W. Y. (1979) Biochem. Biophys. Res. Commun 90,473-480.

CALMODULIN

ANTAGONISTS

709

28. WITHNALL, M. T., AND BROWN, T. J. (1982) Biochem Biophgs. Res. Commun 106,1049-1055. 29. MOSKOWITZ, N., SHAPIRO, L., SCHOOK, W., AND PUSZKIN, S. (1983) B&hem. Biqvhys. Res. Commum 115,94-99. 30. WITHNALL, M. T., BROWN, T. J., AND DIOCEE, B. K. (1984) B&hem. Biophys. Res. Cmmun. 121, 507-513. 31. WALENGA, R. W., OPAS, E. E., AND FEINSTEIN, M. B. (1981) J. Biol. C&m. 256.12523-12528. 32. LAPETINA, E. G. (1982) J. BioL Chem. 257, 73147317. 33. MOORE, P. B., AND DEDMAN, J. R. (1982) J. Biol Chem. 257,9663-9667. 34. JAIN, M. K., ESKOW, K., KUCHIBHOTLA, J., AND COLMAN, R. W. (1978) Thromb. Res. 13, 10671075. 35. TAKAI, Y., KISHIMOTO, A., IWASA, Y., KAWAHARA, Y., MORI, T., AND NISHIZUKA, Y. (1979) J. Biol. Chem 254,3692-3695. 36. SANO, K., TAKAI, Y., YAMANISHI, J., AND NISHIZUKA, Y. (1983) J. Biol. Chem 258,2010-2013. 37. APITZ-CASTRO, R., CRUZ, M., MAS, M., AND JAIN, M. K. (1981) Thumb. Res. 23,347-3X 38. NAYAR, R., MAYER, L. D., HOPE, M. J., AND CULLIS, P. R. (1984) Biochim Biqphys. Ada 777, 343346. 39. WATANABE, T., NARUMIYA, S., SHIMIZU, T., AND HAYAISHI, 0. (1982) J. Biol. Chem 257, 1484714853. 40. WATANABE, T., SHIMIZU, T., NARUMIYA, S., AND HAYAISHI, 0. (1982) Arch. Biochem. Biophys. 216,372-379. 41. BAENZIGER, N. L., AND MAJERUS, P. W. (1974) in Methods in Enzymology (Fleischer, S., and Packer, L., eds.), Vol. 31, pp. 149-155, Academic Press, New York. 42. ROBERTSON, A. F., AND LANDS, W. E. M. (1964) J. Lipid Res. 5,88-93. 43. CHAFOULEAS, J. G., DEDMAN, J. R., MUNJAAL, R. P., AND MEANS, A. R. (1979) J. Biol. Chem. 254,10262-10267. 44. BLAUSTEIN, M. P., AND GOLDMAN, D. E. (1966) Science 153,429-432. 45. LAEMMLI, U. K. (1970) Nature (Loncltm) 227,680685. 46. OAKLEY, B. R., KIRSCH, D. R., AND MORRIS, N. R. (1980) An& Biochem. 105,361-363. 47. CHEN, P. S. JR., TORIBARA, T. Y., AND WARNER, H. (1956) Anal Chem 28,1756-1758. 48. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, R. J. (1951) J. Biol. Chem. 193, 265-275. 49. RITTENHOUSE-SIMMONS, S. (1981) J. Biol Chem 256,4153-4155.