226 ( 1992) 281- 285 ¢0 1992 Ebevier Science Publishers B.V. All tights reserved 0922-4106/92/$05.(}0 Eur~;~c,,, g mmal o f Pharmacology - Moleculc,'rPharmuco!o£~ Section.

~.Sl

EJPMOL 90324

Serotonin inhibits adenylate cyclase in human platelet membranes M i c h a e l E. N e w m a n Yaacot. Herzog Centre ]or Brain and Psychiat o" Research, Ezrath Nashim-He~og lfo~pi~al. Jer;~s'Jh'm, Israel

Received 27 November 1991, revised MS received i8 Marcia 1992, accepted 7, ~ori11992

Sero~.onin (5-HT), the non-specific 5-HT~ receptor agonist methysergide, and the 5-HT,~ receptor agonist CGE 12¢.166 ,~ose-~_e~endc,:~tl~ ii~hibitcd foi~ko~;,-~i~ul~ted ~denylatc cyclase activity in membranes dcrive,-i f-rim h~m~.n plateie,,s. Bt~spirone and ipsapirnne showed partia] agonis~ activity. The effect of 5-HT to inhibit forskolin-stimulzqed a:tivity wa.; partially antagonised by spiperone, yohimbine and pindol¢:,l, with each of these compounds showing some partial agonist activity. It is concluded that human platelet membranes possess receptors for 5-HT coupled negatively to adenylate cyclase. The relationship of these receptors to other binding sites which have bccn ~hown tot 5-HT on platelct membranes remains to be finally elucidate 1. Adenylate cyclase; Ptatelets (human); 5-HT (5-hydrox~yt~'p!aminc, serotonin)

1. Introduction Blood platelets possess various elements similar to those in neurotrat~smitter systems, including several receptors which mediate the aggregatory response (Zucker and Nachmias, 1985). These include receptors to thrombiJl and vasopressin, and adrenergic receptor:~ of both the/3- and ex2 subtypes. Human platelet merebranes also possess an adenylate cyclase enzyme which is stimulated by prostaglandin E~ and inhibited by epinephrine, acting at c~2-adrenoceptors. These agents are"presumed to act via stimulatory or inhibitory guanine nucleotide binding proteins Gs and Gi, and indeed both stimulation and inhibition of adenylate ~ : y clase in platelet membranes by agents acting distal to ~he receptor such as N a F or forskolin have been shown. Serotonin has been shown to stimulate adenylate cyclase in rat and guinea-pig hippocampus via receptors of the 5-HTjA type (Markstein et al., 1986; Shenker et al., 1987), although this effect has not been observed by all investigators. In addition, inhibition of forskolinstimulated adenylate cyclase by 5-HT has been shown to take place at 5-HTIA receptors in the hippocampus (De Vivo and Maayani, 1986; Bockaert et al., 1987; Sehoeffter a~ld Hoyer~ 1988), at 5-HT m receptors in rat substantia nigra (Bouhelal et al., 1987; Schoeffter and

Correspondence to (present address): Michael E. Newman. Biological PsychiatryLabo~atoD',Department of Psychiatry,Hadassah U n i versity Hospital, P.OB. 12000, Jerusalem 91120, Israel.

Hoyer, 1989a) and at 5-H'I'~D re~_cptors in caif substantia nigra (Schoeffter et al., t988). Human platelet membranes possess a serotonin upo take site and 5 - H T 2 receptors linked to the polyphosphoinositide hydrolysis transducing system, which can be labelled with the ligands [3H]sp;mperidol, [3H]LSD, [12~I]LSD and [3H]ketanserin (McBride et al., 1983, i987; Geaney et al., 1984; Biegon et al., 1987). A binding site for [~H]8-hydro×y-2-(di-n-propyla;~lino)tetralin ([3H]8-OH-DPAT), a ligand with high affinity for 5-HT~A receptors, was recently demonstratcd on human platelet membranes by Ieni and Meyerson (1988). These authors, however, concluded thw this site was :~harmacotogically different from the neuronal 5-HT~A site and probably represented a co~:~ponent of the 5.HT transporter. A further binding site on the human p!atelet membrane is that labelled by [3H]SCI-t 23390, a compound with dopamine D~ agonist activity (de Keyser et al., 19S8, 1989). This site had a high (5.7 nM) affinity for 5-HT but not for 8-OH-DPAT or other 5-HT~A receptor agonists, and was not coupled to stimulation of adenylate cyclase. We have recently (Newman and Lerer, 1990) provided preliminary evidence that in human platelet membranes 5-HT both stimulates adenylate cyclase under basal conditions and inhibits adenylate cyclase activity when the enzyme is stimulated by forskolin. These activities parallel those observed with 5-HT and certain of its analogs at the 5-HT~A receptor in rat or guinea-pig hippocampal membranes (de Vivo and Maayani, 1986; Markstein et al., 1986; Shenker et al.,

2~2 1987). The possibility thus e~:isted that a receptor similar to the 5-HT~A receptor found in several regions of human and animal brain may also exist in human ptatelets. In the present work, we have attempted to characterise the inhibitory effects of 5-HT and its analogs on adenylate cyclase in human platelet membranes.

2.3. Data analysis ECso values and maximal degrees of stimulation or inhibition were calculated from dosc-re~izon~e eur,'e~ by computerised curve fitting according to the Michaelis-Menten equation using the program ENZF I T (Elsevier Biosoft)

Z 4. Materials 2. Materials and methods

2.1. Preparation of platelet membranes Platelet membranes were prepared either from freshly collected blood or from platelc~ concentrates obtained from the blood b a n k at Hadassah iSaiversity Hospital. Tile freshly collected blood samples (50 ml) were mixed with 7.5 ml of 0.11 M glucose/85 mM trisodium citrate/71 mM citric acid to give a final concentration of 15% by volume. The blood was centrifuged at 100 × g for 15 rain and the supernatant, consisting of plateletrich plasma, collected using plastic pipette tips arid centrifuged at 400 × g for 20 min. The supernatant x~as then discarded. If red blood ceils were present in the pellet after tiffs centrifugation, the pellet was re-suspended in a small volume of 50 mM Tris-HCl, 2 mM EGTA, 1 mM dithiothreitol and centrifuged for 2 min only at 1 0 0 × g . The pellet containing the red blood cells was discarded. The supernatant from this centrifugation was then diluted with more buffer and re-centrifuged at 400 × g for 20 rain. The pellet was then homogenised using a Polytron apparatus (5 s at position 5), diluted with approx. 60 ml buffer and centrifuged for 20 min at 48,000 × g. The pellet from this centrifugation was resuspended in 2 ml buffer with Polytron homogenisation and frozen in liquid nitrogen in 1 ml aliquots,

2.2 Adenylate cyclase assay The reaction was performed essentially as described by De Vivo and Maayani (1986) in an assay volume of 0.2 ml containing 25 mM Tris-HCI, pH 7.4 at 18°C; 2 mM MgCI2; 0.5 mM ATP; 10 IxM GTP; !00 mM NaC!; 1 /.tCi [a-~~P]ATP; 1.5 m g / m l creatine phosphate; 0.2 m g / m l creatine phosphokinase and 1 mM 3-isobutyl1-methylxanthine. Incubations were started by addition of ~ 50 /.tg protein and terminated after 10 rain at 30°C by addition of 0.1 ml of a solution containing 0.5 mM cyclic AMP and 4 mM ATP. and transferring the tubes to a boiling water bath. [3~'P]Cyclic A M P was separated from ATP by the double column method. All concentration points were tested in duplicate, each observation corresponding to the mean of duplicate samples.

The following drugs were generously donated: methysergide and mesulergine (Sandoz, East Hanover, N J, USA); buspirone (Bristol-Myers, Evansville, IN, USA); ipsapirone (Troponwerke, Cologne, Germany). 8-OH-DPAT, trifluoromethylphenylpiperazine (TFMPP), 1-[2-(4-aminophenyl)ethyl]-4-(3-trifluoromethylphenyl)piperazine (PAPP) and CGS 12066B were purchased from Research Biochemicals, Natick, MA, USA. Forskolin was from Calbiochem, Lucerne, Switzerland, and 5-HT, (-)-pindolol, spiperone and yohimbine from Sigma Chemical Co., St. Louis, MO, USA.

3. Results In preliminary experiments, the integrity of the platelet preparations derived from the two sources were compared by examining stimulation of adenylate cylase activity via a receptor (PGEj), distal to the receptor at the G protein (NaF) and distal to the receptor at the catalytic unit (forskoliu)in membranes obtained from both freshly derived bloods and blood bank preparations. Table i shows that there was no difference in the degree of activation by NaF, PGE~ or forskolin between the two preparations. Fig. 1 show~ the effects of 5-HT and a number of 5-HTt receptor agonists on platelet cyclase activity in the presence of 30 ~ M forskolin and !00 mM NaCI. 5-HT and CGS 12066 inhibited forskolin-stimulated activity, by about 20% with ECs0 values of approximately 15 nM and 125 nM respectively. The other compounds, including TFMPP, which is about equiootent at 5-HT~A and 5oHT m receptors, showed no acti',-

TABLE1 Comparisonof adenylate cyclase activities in platelet membranes derivedfrom freshly processed blood and blood bank preparations. Resultsare exoressed in omol/min/mg protein and are mean_+ S.E.~ of the number of samples indicated in parentheses, each sample being :~ssayedin duplicate. Activating agent Fresh blood Blood bank None (basal activit~'~ 24.8_+ 4.7 (7) 27.0_+ 5.2 (13) 10mM NaF 67.0± 16.7(7) 59.1_+ 9.2(12) t0ttMPGEt 439.9±128.0(4) 487.0+131.5(7) 30 p.M forskolin 802.8+ 186.9 (7) 686.0_+176.4 (12)

283 >.. I-.

,,~

I

100-~,~--------Ik .~ 95 t

@ ,~

]

5-HT

CGS 12066

•

• PAPP

* TFMPP

ECst ~ values and maximal activities for agonis~s on inhibition of adenylate cyclase activity' in human platelet membranes.

~ ~_...~

Values are means_+ S.E.M. of the number of observations indicated in parentheses.

--"-tk

"¢~.r ---.. ' -.. "-

90

~

"-"~"

",,.*".:"

.......... . .... > , . "~""

~, "~'~L

-~

85

,,

I.o') I

80

Z "4

O3

,

Agonist

EC~{~, nM

Max. inhibition, f;

5-HT CGS 12066 Methysergide Buspironc l~s~pirone.

15.6+_ 10.5 124.8+82.5 17.2 ± 23.1 18.4± 3.0 89.8 + 53.3

33.8± 3.4 (13) 22 34 6 5 (4) 14.7 + 3.3 (4) 12.7±3.3 (~) 0.(} __+0,9 (.~

~,

75

,'V

O '.a

TABLE 2

• DPAT

t

70 ~ ' 0

-9

-8

-7

[AGONIST],

..... -5

-6

-4

log M

Fig. 1. Effects of various concentations of 5-HT and 5-HT~ recelr,tor agonists on forskolin-stimulated adenylate cyclase activity in human platelet membrane.~. Each point represen,~s the mean of at least four determinations. Standard errors were < 8% and are omitted for clarity.

>. p.. ~, v. (9 " I-_~ ~

II

5-HT ALONE

100 ' l ~ - ~ , - , , , e . ~ T 1 % ~-~"O--~'~

i

5-HT + 10 uM PINDOLOL 5-HT + 10 ~M MES~JLERGINE

*..-..

o3 80

~

0 K 0

5-HT ALONE

.......... .....

(¢) -J

o~t z -. o t,~ O ~

•

68

0 75

~" 0 0

-9

-8

-7

[AGONIST],

-6

-5

-4

log M

Fig. 2. Effects of the 5-!'tT l agonist methysergide and the 5-HT~a partial agonists buspirone and ipsapirone on adenylate cyc!asc aetivity in human platelet membranes. Each point repre.~ents the mean o! at least four determinations. Standard errors were < 6% and are omitted for clarity,

60 -9

-8

-7 [5-HT],

-6

-5

-4

log M

Fig. 4. Effects of l0 #M pindotol and 10 ~M mesuler~ine on 5-HT-mediated inhibition of forskolin-stimutatcd adcnylate cvcI:,.~e activity in human platetet membranes, Each point represents the mean of at ]east four determinations, Standard errors were < 4c* and arc omitted for clarity.

284 ence of 10 # M spiperone and to 21.8 + 1.7% in the presence of 10 /xM yohimbine. K , values, calculated from the Cheng-Prusoff equation, were 72.5 nM for spiperone and 185 nM for yohimbine. Fig. 4 shows similar data for the two compounds (-)-pindolol, ctassifled as a 5-HTI,~ and 5-HT m receptor antagonist, and mesulergine, classified as a 5-HTIc receptor antagonist. Pindo!ol behaved as a competitive antagonist with partial agonist activity, reducing the maximal inhibition obtained with 5-HT from 22.4 + 3.1% to I0.7_+ 2.1% and yielding a K h value of 229 nM, whde mesulergine behaved as a non-competitive antagonist, reducing maximal inhibition by 5-HT to 10.8 + 2.6% while not affecting its ECs0.

4. Discussion The present experiments show, that ,he receptor mediating inhibition of forsko~in-stimulated adenylate cycla.~e in human platelet membranes has some features in common with the 5-HT~A receptor but also shows some important differences. While the 'classical' 5-HTtA receptor agonist 8-OH-DPAT was not particulary active, both buspirone and ipsapirone exhibited partial agonist activity. These compounds are considered to be inactive at either 5-HT m or 5-HT m sites, yet in the present experiments CGS 12066, a c o m pound with agonist activity at both these receptors (Schoeffter and Hoyer, 1989b), inhibited forskolinstimulated adenylate cyclase with an ECs0 of 125 nM, a value similar to that obtained at the 5-HT m receptor in rat substantia nigra (Bouhelal et al., 1987). However, PAPP, which has agonist activity at both 5-HT~A and 5-HTLo receptors, and TFMPP, which shows selectivity for the 5-HTIA and 5-HT m sites (Schoeffter and Hoyer, 1989b), only showed activity at high concentrations, The effects of antagonists were also studied in an attempt to classify the activity. Spiperone, a 'classical' antagonist at the 5-HT~A receptor, antagonized 5-HT inhibition with a K b of 72.5 nM, a value comparable with that obtained at the 5-HT~A receptor in rat or guinea-pig hippocampus, but increasing the 5-HT concentration did not completely reverse the effect of spiperone, indicating the compound to have some partial egonist activity. This was confirmed by the reduction iI: the maximal inhibition obtained with 5-HT in the presence of spiperone. Similarly, the az-adrenoceptor antagonist yohimbine also antagonised the effeet of 5-HT in a manner indicating it to have partial agonist activity. Yohimbine has been shown to have partial agonist activity at the 5-HT1D receptor in calf substantia nigra (Schoeffter et al., 1988). Recent evidenee (De Vos et al., 1991) indicates that rauwolscine, another compound initially characterised as an a 2adrenoceptor antagonist, also has agonist properties at

the 5-HTIA receptor in human frontal cortex. The effects of yohimbine or rauwolscine on inhibitioc of forskolin-stimulated adenylate cyclase activity mediated by the 5-HTtA receptor have not been determined. However, binding studies at the 5-HTtA receptor in pig frontal cortex (Hoyer, 1988) showed yohimbine to have a pKa of 6.86, corresponding to a K d of 138 nM, in close agreement with the K b of 185 nM reported here. Two other compounds were tested for antagonist activity. (-)-Pindolol, which was a competitive antagorqst at the 5-HT~A receptor both in calf (Sehoeffter and Hoyer, 1968) and rat hippocampus (Oksenberg and Peroa!t:a, 1988), and also at the 5-HT m receptor in rat substantia nigra (Schoeffter and Hoyer, 1989a), behaved as an antagonist with some partial agonist ae~iv~y fhus, maximal inhibition by 5-HT was reduced trom 22.4 + 3.1% to 10.7 _+ 2.!% in the presence of 10 /zM pindolol, while the calculated K b value for pindotol was 229 nM. Partial agonist activity for pindolol has been shown both at the 5-HT~A receptor in rat hippocampus (De Vivo and Maayani, 1990) and at the 5-HT~D receptor in calf substantia nigra (Schoeffler and Hoyer, 1989a). Mesulergine, however, behaved as a non-competitive antagonist. Virtually all the compounds tested for inhibition of forskolin-stimulated adenylate cyclase in the present st~'dy were active over a wide concentration range, fro~ 1 nM to 100/zM, without reaching a clear maximal ehc,~t. This suggests, that more than one site may be involved in the effect. In the case of TFMPP, the compound which showed clearest ~ndication of a biphasic dose-response curve, a single ECs0 value could not be calculated. Similarly the effects of spiperone, yohimbine and pindolol, which were classified as competitive antagonists with partial agonist activity, could also be explained by interaction with more than one receptor. An action of these compounds as non-competitive antagonists at one receptor and full agonists at the other could account for the apparent biphasic nature of the inhibition curves. However, it should be pointed out that the present data for pindolol are in accordance with the data of De Vivo and Maayani (1990), who concluded the compound to be a partial agonist with low intrinsic efficacy in rat hippocampal membranes. Mesulergine, on the other hand, appeared t~ be a pure non-competitive antagonist since it did not affect the EC50 for 5-HT-indueed inhibition. While the effect of 5-HT to inhibit forskolin-stimulated adenylate cyclase in human platelet membranes does thus seem to have some features in common with the activity mediated by 5-HT~A receptors in other tissues, differences in the relative potencies of agonists and antagonists indicate that another class of receptor may be involved. The human platelet site with a high affinity for the dopamine D~ receptor agonist SCH

285

23390, classified as a 5-HTIE receptor by De Keyser et al. (1989), has some features in common with the activity d e s c r i b e d h e r e . Thus, 8 - O H - D P A T , which has a very low affinity at t h e 5-HTIE r e c e p t o r , did not i n d u c e inhibition o f forskolin-stimula~ed a d e n y l a m cy-

close in human platelet membranes (fig. 2). Similarly, t h e agonist methyser'gide h a d a h i g h e r affinity, at b o t h t h e 5-HTtE ~ite (71 n M ) a n d for inhibition o f cyclase in p l a t e l e t m e m b r a n e s (17 n M ) t h a n t h a t o b t a i n e d for

5-HTtA receptor-mediated inhibition of cycl~se in hippocampal membranes (guinea-pig 260 nM, De Vivo a n d Maayani, 1986; calf 480 nM, S c h o e f f t e r a n d H o y e r ,

1988). Finally mesulergine, normally considered a s p e cific 5 - H T t c r e c e p t o r antagonist, had a relatively high affinity at t h e 5 - H T m site (217 n M ) a n d also b e h a v e d as a p o t e n t a n t a g o n i s t in the e x p e r i m e n t s d e s c r i b e d h e r e . Although mesulergine, TFMPP and SCH 23390 are all p o t e n t ligands at the 5 - H T t c r e c e p t o r , t h e possibility t h a t this site is involved in the activity described in t h e p r e s e n t e x p e r i m e n t s s e e m s unlikely bec a u s e coupling of the 5-HTlc receptor to adenylate cyclase has n o t b e e n d e s c r i b e d in any tissue. It s h o u l d b e n o t e d t h a t while D e K e y s e r e t al. (1988) s h o w e d n o

effect of 5-HT on stimulation of adenylate cyclase in h u m a n platelets, its action o n inhibition o f forskotins t i m u l a t e d activity was n o t studied. F u r t h e r experim e n t s will b e r e q u i r e d to establish w h e t h e r t h e 5-HTtE site a n d t h e site m e d i a t i n g t h e inhibitory action of 5 - H T a n d its analogs on a d e n y l a t e cyclase activity are

indeed identical.

Acknowledgements This work was supported in part by grant MH 43873 from the National Institutes of Health, USA. I thank Professor Bernard Lerer for his constant support and encouragement.

References Biegon, A., A. Weizmann, L. Karp, A. Ram, S. Tiano and M. W o l f f , 1987, Serotonin 5-HT2 receplor binding on blood platelets - - a peripheral marker for depression'?, Life Sci. 41, 2485. Bockaert, J., A. Dumuis, R. Bouhelal, M. Sebben and R.N. C o r 3 ' , 1987, Piperazine derivatives including the putative anxiolytic drugs, buspirone and ipsapirone, are agonists at 5-HTtA receptots negatively coupled with adenylate cyclase in hippocampal neurons, Naunyn-Schmied. Arch. Pharmacol. 335, 588. Bouhelal, R., L. Smounya and J. Bockaert, 1988, 5-HTtB receptors are negatively coapled with adenylate cyciase in rat substantia nigra, Eur. J. Pharmacol. 151,189. De Keyser, J., M. de Waele, A. Convents, G. Ebinger and G. Vauquelin, 1988, Identification of Dl-like dopamine receptors on human blood platelets, Life Sci. 42, 1797.

De Keyser, J., H. Walraevens, A. Convems, G. Ebingcr and G. Vauquelin, 1989. [3II]SCH 23390 labels a novel 5-hydro~,tryptamine binding site in immar: blood platelet membranes. Eur. J. Pharmacol. 162, 437. De Vivo, M. and S. Maayani, 1986, Characterization of the 5-hydroxytryptamine-la receptor-mediated inhibition of forskolinstimulated adenylate cyclase activity in guinea-pig and rat hippocampal membranes, J. Piaarmacol. Exp. Ther. 238, 248. De Vivo, M. and S. M~ayani, 1990. Stirnula:ion ned inhibition of adcnylyt cyclase by distinct 5-hydrox~tryptamine receptors, Biochem. Pharmacol. 40, 1551. De V:-, H., E. Czerwiec, J.-P. De Backer. W. de Potter and G. Vauquelin, t991, [3H]Rauw~lscine behaves as an agonist riot the 5-IJrtA receptors in human frontal cortex membranes, Eur. 3. Ph~rmacol. 207, 1. Geaney, D.P., M. Schachter, J.M, Elliot and D.G. Grahame-Smith, 1934, Characterisation of [3HJly~ergic acid diethytamide binding to a 5-hydroxytr:cptamin¢-re~eptclron human platelet membranes, E~r. J. Phmmacol. 97, 87. Hoyeq D., 1988, Functional correlates of serotonin 5-HT1 recogniti~m sites, J. Receptor Res. 8, 59. leni, LR. and L.R. Meyerson, 1988, The 5-HTIA receptor probe [3r~,]8-OH-DPAT labels the 5-HT transporter in human platetets, L~,e Sci. 42, 3!1. Markstein, R., D. Hoyer and G. Engel, 1986, 5-HTbx receptors mediate stimulation of adenylate cyclase in rat hippocampus, Naunyn-Schmied. Arch. Pbarmacol. 333, 335. McBride, P.A., J.J. Mann, B. McEwen and A. Biegon, 1983, Characterization of serotonin binding sites on human platelets, Life Sci. 33, 2033. McBride, P.A., J.J. Mann, M.J. Policy, AJ. Wiley and J.A. Sweency, 1987, Assessment of binding indices and physiologicalresponsiveness of the 5-tfT2 receptor on human plateiets. Li~e Sci. 40. 1799. Newman, M.E. and B. Lerer, 19913, 5-HT~A receptor-mediated effects of antidepressants: relation to G protein function, Clin. Neuropbarmacol. 13 (Suppl. 2), 237. Oksenberg, D. and SJ. Peroutka, 1988, Antagonism of 5-hydroxytryptamin%A receptor-mediated modulation of adenytate cyciase activity by pindolol and prop':molol isomers, Biochem. Pharmacol. 37, 3429. Schoeffter, P. and D, Hoyer, 1988, Centrally acting hypotcnsive agents with affinity for 5-HTtA binding sites inhibit forskolinstimulated adenylate cyclase activity in caif hippocampus. Br. J. Pharmacol. 95, 975. Schoeffter, P. and D. Hoyer, 1989a, 5-Hydroxytryptamine 5-HT m and 5-HTID receptors mediating inhibition of adenylate cyclasc activity. Pharmacological comparison with special reference to the effec~*sof yohimbine, rauwolscine and some beta-adrenoceptor antagonists. Nauayn-Schmied. Arch. Pharmacol. 340, 285. Schueffter, P. and D. Hoycr, 1989b, Interaction of arylpipcrazines with 5-HTIA 5-HT u, 5-HTIc and 5-HTtD receptor:~:do db;criminatory 5-HT m receptor ligands exist?, Naunyn-Schmied. Arch. Pharmacol. 339, 675. Schoeffter, P., C, Waeber. J.M. Palacios and D, Hoyer, 1988, T{',e 5-hydrox3,'tryptamine 5-HT m receptor subtype is negatively conpied to adenylate cyciase in calf substantia nigra, NaunynSchmied. Arch. Pharmacol. 337, 602. Shenker, A., S. /vlaayani, H. Wcinstcin and J.P. Green, 1~87, Pbarmacological characterization of two 5-hydroxyt~'ptamine recep~ tors coupled to adenylate cyclase in guinea-pig hippocampal membranes, Mol. Pharmacol. 31,357. Zucker°M.S. and V.T. Nachmias, 1985, Platelet activation, Arteriosclerosis 5.2.