Eur. J. Biochem. 207, 259 - 263 (1992)
$3 FEBS 1992
ADP receptor-induced activation of guanine-nucleotide-binding proteins in human platelet membranes Christian GACHET ’, Jean-Pierre CAZENAVE’, Philippe OHLMANN ’, Gerhard HILF2, Thomas WIELAND’ and Karl H. JAKOBS’ ’ INSERM U.311, Biologie et Pharmacologie des Interactions du Sang avec les Vaisseaux et les Biomateriaux, Centre Regional de Transfusion Sanguine, Strasbourg, France ’ Pharmakologisches Institut, Universitat Heidelberg, Federal Republic of Germany Institut fur Pharmakologie der Universitat G H Essen, Federal Republic of Germany (Received March 5, 1992)
-
EJB 92 0306
ADP receptor-regulated binding of the labeled GTP analog, guanosine 5’-0-(3-[35S]thiotriphosphate) ([3 5S]GTP[yS]),to guanine-nucleotide-binding proteins (G proteins) was studied in human platelet membranes. The potent ADP receptor agonist, 2-methyl-thio-adenosine 5’-diphosphate (2MeSADP), a non-hydrolyzable analog of ADP, increased the binding of [3sS]GTP[yS] without apparent lag phase. Under optimal conditions, i.e. in the presence of GDP (1 - 10 pM), 2MeSADP increased the binding up to about threefold, with half-maximal and maximal increase observed at 10 nM and 1 pM 2MeSADP, respectively. ADP itself increased the binding of [35S]GTP[yS] by maximally about twofold, with half-maximal increase occurring at 0.1 pM ADP. The agonist-induced stimulation was competitively antagonized by the ADP receptor(s) antagonist, (1s)-adenosine 5‘-0(1-thiotriphosphate) { (Sp)-ATP[aS]). Other platelet receptor agonists known to act through receptors coupled to G proteins also increased binding of [35S]CTP[yS]in human platelet membranes, but without being inhibited by (Sp)-ATP[aS]. The data presented indicate that the platelet ADP receptor(s) can interact with and efficiently activate G proteins. the nature of which remains to be identified.
About 30 years ago, it was discovered that ADP was responsible for platelet adhesion to glass and platelet aggregation [I - 31. The development of a simple turbidimetric method [4] to follow quantitatively the aggregation of platelets in vitro made possible the study of the effects of a variety of agonists and antagonists on platelet function. Addition of ADP to washed human platelets results in shape change, exposure of the fibrinogen binding site on the GPIIb-IIIa complex and reversible aggregation in the presence of fibrinogen and physiological concentrations of Ca2’ [5, 61. At the intracellular level, platelet activation following ADP binding to its receptor leads to a transient rise in free cytoplasmic Ca2+ concentration, resulting both from C a 2 + influx across the plasma membrane and from C a 2 + mobilization from internal stores [7], and myosin-light-chain phosphorylation. In addition, ADP inhibits stimulated adenylyl cyclase [8 - 1I], but this is not the cause of platelet aggregation [12]. On the other hand, agents that increase platelet cyclic AMP levels, such as prostaglandin I2 and El (PGE1), are very potent Correspondence to C. Gachet, Service d’Hemostase et de Thrombose, Centre Regional de Transfusion Sanguine, 10, rue Spielmann, F-67085 Strasbourg Cedex, France Abbreviations. [35S]GTP[yS], guanosine 5’-0-(3-[35S]thiotriphosphate); 2MeSADP. 2-methylthioadenosine 5’-diphosphate; (Sp)ATP[&], (1 S)-adenosine 5’-O-(1 -thiotriphosphate); PGE,, prostaglandin E l ; G protein, guanine-nucleotide-bindingprotein; Gi, guanine nucleotide binding protein(s) involved in inhibition of adenylyl cyclase. Enzyme. Apyrase (EC 3.6.1.5). Note. This work was presented in part at the 33‘d Annual Meeting of the American Society of Hematology, Denver, USA, December 1991 [Gachet et al. (1991) Blood 78 Suppl. 1, 140al.
and efficient inhibitors of platelet activation and aggregation [12, 131. When washed human platelets are used, ADP has the same effects as thrombin in simultaneously inhibiting activated adenylyl cyclase and inducing platelet aggregation, whereas other agonists, such as platelet-activating factor and vasopressin, do not inhibit adenylyl cyclase in intact platelets [lo, 14, 151 or, like adrenalin [16], do not induce aggregation although inhibiting adenylyl cyclase. Thrombin activates phospholipase C via a so-far-unidentified guanine-nucleotidebinding protein (G protein) and inhibits adenylyl cyclase via the G protein, Gi [15, 17-22]. At the present time, there is limited information as far as the ADP receptor signal transduction pathway is concerned, probably because of unresolved technical problems which occur when ADP is studied in the assay systems commonly used. Recent investigations seem to have conclusively established that phospholipase C is not involved when ADP induces human platelet activation in a medium containing physiological concentrations of Ca2’. Under these conditions, aggregation occurs without evidence of degradation of phosphatidylinositol bisphosphate 1233. In one study [24], it was shown that ADP inhibited, like adrenalin, the adenylyl cyclase of platelet membrane preparations. The results of this work were not in agreement with the requirement for GTP in the expression of inhibitory activity of adenylyl cyclase induced by ADP as reported in another study [25], thus leaving still unresolved the question of the involvement of putative GTP binding protein(s) in the ADP signal transduction pathway(s). The aim of the present study was to determine the requirements for receptor-dependent Gprotein activation after ADP stimulation. It has been shown previously that agonist-stimulated binding of the radiolabe-
260 led GTP analog, guanosine 5’-0-(3-thiotriphosphate) ([3sS]GTP[yS]) can be used to analyse receptor - G-protein interactions in different cell membrane preparations [26, 271. Using the non-hydrolyzable analog of ADP, 2-methyl-thioadenosine 5‘-diphosphate (2MeSADP), as agonist and the specific ADP receptor antagonist, the S p diastereoisomer of adenosine 5’-0-(l-thiotriphosphate){(Sp)-ATP[aS]}[28 - 321, we were able to demonstrate that the agonist-activated ADP receptor stimulates binding of [35S]GTP[yS]to human platelet G proteins, an effect specifically blocked by a competitive antagonist of the ADP receptor.
perature. These platelets were loaded with glycerol (Prolabo, Paris, France) by centrifugation through a 0 - 30% (mass/ vol.) glycerol gradient and lysed in a hypotonic Tris buffer containing leupeptin (50 pg/ml), diisopropylfluorophosphate (1 mM), aprotinin (20 Ujml) (SanoG-Choay, Paris, France) and EDTA (2 mM). After lysis, the broken platelets were layered onto a 30% (massivol.) sucrose cushion and centrifuged for 4 h at 60000 g in a Beckmann ultracentrifuge. The ‘floating’ membranes were removed carefully with a plastic pipette, washed and pelleted by centrifugation at I00000 g. These membranes were biochemically characterized by SDS/ PAGE analysis and their functional integrity was checked by measuring adenylyl cyclase activity [36].
MATERIALS AND METHODS Materials
Binding of [35S]GTP[yS]
ADP, prostaglandin El (pGE1), adrenalin and diisopropylfluorophosphate were from Sigma Chemical co.(St Louis, MO, USA). Leupeptin was from Fluka Chemie AG (Buchs, Switzerland). GDP, GTP[yS] and (Sp)-ATP[aS] were from Boehringer (Mannheim, FRG). [3’S]GTP[yS] (12003400 Ci/mmol) was from New England Nuclear (Dreieich, FRG). 2MeSADP was obtained from Sanofi Recherche (Toulouse, France). A synthetic peptide (14 residues, > 98% purity), corresponding to the newly exposed N-terminal sequence of the thrombin receptor following proteolysis by thrombin [33], from Neosystem (Strasbourg, France). All other products were commercial reagents of analytical grade.
Aliquots of human platelet membranes were thawed and diluted in 10 mM triethanolamine/HCl, pH 7.4, to a protein concentration of 100 pg/ml. The reaction mixture for measuring [3’S]GTP[yS] binding contained, if not otherwise indicated, 50 mM Tris/HCl pH 7.5, 5 mM MgC12, 1 mM EDTA, 1 mM dithiothreitol, 1 pM GDP, 0.3 -0.5 nM [3sS]GTP[yS] (about 50000 cpm) and 1.5 -4 pg membrane protein in a total volume of 100 pl. Incubation was started by addition of the membrane suspension to the prewarmed reaction mixture and carried out in triplicate for 30 min at 25°C. The reaction was terminated by rapid filtration through nitrocellulose filters (45 pm porosity) under vacuum. The filters were washed with 12 ml50 mM TrisiHCI pH 7.5 containing 5 mM MgC12. Radioactivity bound to the membranes on the filters was determined in scintillation fluid containing toluene and Triton X-100. Non-specific binding determined in the presence of 10 pM unlabeled GTP[yS] amounted to about 0.4% of added [35S]GTP[yS]and was subtracted from the total bound radioactivity to determine the specific radioactivity. The intraassay variation of triplicates was less than 5% of the mean.
Preparation of washed human platelets Blood was collected from a forearm vein, six volumes of blood into one volume of acid/citrate/dextrose anticoagulant, and twice-washed platelet suspensions were prepared as previously described [34]. The final resuspending medium (pH 7.35) was Tyrode’s buffer containing 2 mM Ca2+,I mM Mg2+, 0.35% (massivol.) human serum albumin (Centre Regional de Transfusion Sanguine, Strasbourg, France) and apyrase (2 pg/ml, a concentration that converted 0.25 pM ATP to AMP within 2 min at 37°C). Platelets were stored at 37 “C throughout the experiments. Platelet count was adjusted in the final suspension to 300000/pl using a Baker 810 platelet counter (Baker Instruments, Allentown, PA, USA). Platelet aggregation studies Aggregation was measured at 37 “C by a turbidimetric method in a dual-channel Payton aggregometer (Payton Associates, Scarborough, Ontario, Canada). A 0.4-ml aliquot of platelet suspension was stirred at 1100 rpm and activated by addition of 50 pl ADP or 2MeSADP in the presence of 50 p1 (Sp)-ATP[xS] (10 pM final concentration) or 50 p1 diluent and diisopropylfluorophosphate-treated human fibrinogen (0.8 mg/ml) [34]. The extent of aggregation was estimated quantitatively by measuring the maximum curve height above the baseline level. Preparation of human platelet membranes Platelet membranes were prepared essentially as described by Barber and Jamieson [35]. Briefly, blood was obtained from healthy donors who had not taken any drug for at least two weeks. Platelets were washed as described [34] and resuspended in Tyrode’s buffer containing no Ca2+, in the presence of 2 mM EDTA and apyrase (2 pg/ml) at room tem-
RESULTS Washed human platelet aggregation 2MeSADP was a powerful aggregating agent, inducing half-maximal and maximal platelet aggregation at 0.05 pM and 0.5 pM, respectively, while ADP induced half-maximal and maximal response at 1 pM and 10 pM, respectively. Responses to both 2MeSADP and ADP were competitively inhibited by (Sp)-ATP[aS] (Fig. 1). This effect of (Sp)-ATP[aS] was specific, since it did not inhibit platelet aggregation induced by thrombin, platelet-activating factor or arachidonic acid (data not shown). Binding of [35S]GTP[yS]to platelet membranes As illustrated in Fig. 2, GDP inhibited the binding of [3sS]GTP[./S] to platelet membranes with half-maximal and maximal effects at approximatively 1 pM and 0.1 mM, respectively, when no agonist was added. In the presence of 2MeSADP (1 pM), the GDP inhibition curve shifted significantly to the right from the control curve, with maximal 2MeSADP-induced increase in [35S]GTP[yS]binding being observed at 10 pM GDP. Depending on the different membrane preparations used, maximal 2MeSADP-induced increase in [35S]GTP[yS]binding occurred at 1-10 pM added GDP.
261
10 20 30 40 50 60 incubation time (min)
log[agonist]/M
Fig. 1. ADP receptor-induced platelet aggregation. Washed human platelets were stimulated with increasing concentrations of 2MeSADP ( A , A) or increasing concentrations of ADP (0, O), in the absence ( A , 0) or presence (A,0 ) of 10 pM (Sp)-ATP[aS]. Results are expressed as mean values f SEM from three separate experiments.
Fig. 3. Kinetics of [35S]GTP[yS]binding to human platelet membranes. Human platelet membranes were incubated at 25‘C with 0.3 -0.5 nM [35S]GTP[yS]in the presence (0, 0) and absence (m, 0 ) of 1 pM 2MeSADP without (0, D) or with (0, 0 ) 1 pM GDP. At indicated times, samples were analyzed for bound [35S]GTP[yS].Results are from one typical experiment representative for more than three separate experiments. Each point is the mean SEM of triplicate determinations.
I
-8
-7 -6 -5 log[GDP]/M
-4
T
-3
Fig. 2. Influence of GDP and 2MeSADP on the binding of [35S]GTP[rS] to human platelet membranes. Specific binding of [35S]GTP[yS]to human platelet membranes was determinated as described in Materials and Methods in the absence (0) or presence of 1 pM 2MeSADP ( O ) , at increasing concentrations of GDP. Results are from one typical experiment representative for more than ten separate experiments. Each point is the mean j= SEM of triplicate determinations.
The effects of 2MeSADP and GDP on the kinetics of [3’S]GTP[yS] binding to membranes of human platelets are presented in Fig. 3. In the absence of GDP, addition of 2MeSADP (1 pM) led to a marginal increase in [35S]GTP[yS] binding (< 10% at any time point up to 60 min measured). In contrast, when GDP (1 pM) was present, 2MeSADP caused a large increase in binding of [35S]GTP[yS].This agonist effect occurred without apparent lag phase, increased continuously and amounted to almost 5 pmoljmg protein at 60 min. All subsequent experiments were performed for an incubation time of 30 min. Stimulation of [”S]GTP[yS] binding to human platelet membranes by 2MeSADP was concentration-dependent (Fig. 4A). Under standard assay conditions (1 pM GDP, 30 min at 25 “C), half-maximal and maximal stimulation were observed at 10 nM and 1 pM 2MeSADP, respectively. The increase in [3’S]GTP[yS] binding to human platelet G proteins induced by 2MeSADP was antagonized by (Sp)-ATP[aS] (Fig. 4A). In a similar fashion to its stable analog 2MeSADP,
-9
-8
-7
-6
-5
log[ZMeSADP] / M
Fig. 4. ADP receptor mediated stimulation of [35S]GTP[yS]binding to human platelet membranes. Binding of [35S]GTP[yS]to human platelet membranes was determined with 1 pM G D P at increasing concentrations of 2MeSADP (A) or ADP (B), in the absence (0)or presence ( 0 )of 10 pM (Sp)-ATP[aS]. Results are from one typical experiment representative for more than three separate experiments. Each point is the mean f SEM of triplicate determinations.
ADP stimulated [35S]GTP[yS] binding to human platelet membranes in a concentration-dependent manner, with halfmaximal and maximal stimulation observed at about 0.1 pM and 10 pM, respectively. This effect was also antagonized by (Sp)-ATP[aS] (Fig. 4B). Other platelet receptor agonists that act through G-protein-dependent transduction pathways also stimulated [3’S]GTP[yS] binding to purified platelet membranes (Table 1). Under standard assay conditions, PGEl (1 pM) stimulated [35S]GTP[yS]binding 1.2- 1.5-fold and adrenalin (10 pM) 1.4 - 2-fold. Thrombin (1 Ujml) stimulated the binding of [3’S]GTP[yS] very poorly (1 - 1.2-fold as compared to controls). In contrast, a synthetic peptide, corresponding to the newly exposed N-terminal sequence of the recently described platelet thrombin receptor [33] following proteolysis by thrombin, stimulated [35S]GTP[yS]binding to human platelet membranes from 1.6- 2-fold as compared to controls.
262 Table 1. lnfluence of (Sp)-ATP[aSJon [35S]GTPlySl binding to human platelet membranes stimulated by different agonists. Human platelet membranes were incubated for 1 h at 25°C in the absence or presence of 10 pM (Sp)-ATP[aS]. [3'S]GTP[yS] binding induced by adrenaline (10 pM), PGE, (1 yM). the thrombin peptide (10 pM), 2MeSADP (I y M ) and ADP ( 1 0 p M ) was measured. Data are the mean f SEM from four separate experiments. Statistical analysis using the Kruskall- Wallis test showed a significant decrease of agoniststimulated [35S]GTP[yS] binding only for 2MeSADP and ADP ( * P < 0.05, **P< 0.01).
Agonist
[35S]GTPIyS]binding
___ control
__ (Sp)-ATP[aS] (10 pM)
~~
Adrenalin PGE, Peptide 2MeSADP ADP
41+ 2 23+ 5 61 i 11 120+ 2 56+ 6
39 f 13 21+ 6 45+ 6 5 2 + 5* 1 8 + 2**
In order to assess the specificity of (Sp)-ATP[aS] to inhibit 2MeSADP-stimulated [3sS]GTP[yS]binding, human platelet membranes were incubated in the presence and absence of 10 pM (Sp)-ATP[aS] and [3sSs]GTP[yS]binding was stimulated with adrenalin, PGEl and the thrombin receptor peptide (Table 1). The stimulation of [35S]GTP[yS]binding to platelet membranes by any of these agonists was not reduced when measured in the presence of (Sp)-ATP[aS] (10 pM), a concentration that inhibited specifically the effect of 1 pM 2MeSADP (> 50%) and of 10 pM ADP ( > 60%) on [3sS]GTP[yS]binding. D1SCUSSJON
The stable, non-hydrolyzable analog of ADP, 2MeSADP, is at submicromolar concentrations a powerful aggregating agent. 2MeSADP is also 100-fold more potent than ADP in inhibiting CAMP accumulation in human platelets [32]. In addition, (Sp)-ATP[aS] is a specific antagonist of ADP-induced aggregation [29] and binds to a single class of receptors, displacing ADP analogs with an affinity of 17 nM [31]. In the present experiments, 2MeSADP induced platelet aggregation with half-maximal and maximal response at 0.05 pM and 0.5 pM, respectively, an effect competitively inhibited by (Sp)ATP[aS]. ADP-induced platelet aggregation with half-maxima1 and maximal effect at 1 pM and 10 pM, respectively, and was also competitively inhibited by (Sp)-ATP[aS]. These results are in agreement with other studies [28-321. Since ADP is metabolically rapidly degraded while 2MeSADP is not hydrolyzed and since (Sp)-ATP[aS] has the lowest Kd for putative ADP receptor(s), we decided to use these compounds in an attempt to determine whether ADP induces platelet activation through a signal transduction pathway involving activation of a G protein. Quantitative measurement of [35S]GTP[yS]binding to G proteins in membrane preparations from different cell systems is a sensitive tool to study the activation mechanisms of Gprotein-coupled receptors [26, 271. AS discussed in references [26, 271, the presence of GDP allows detection of agoniststimulated binding of [3sS]CTP[yS]by a mechanism which is not yet fully understood. It seems from studies using intact membranes, purified G proteins and receptors, that the
agonist-liganded receptor reduces the affinity of the interacting G protein for bound GDP, allowing [3sS]GTP[1/S]to bind to the G protein [26]. As shown in Figs 2 and 3, addition of GDP to platelet membranes increased not only the relative stimulation but also the absolute increase in GTP[yS] binding induced by 2MeSADP, with a maximal effect observed at 110 pM GDP, varying from one membrane preparation to another. The effect of 2MeSADP on [3sSs]GTP[yS]binding to human platelet membranes was concentration-dependent, reaching half-maximal and maximal effect at 10 nM and 1 pM, respectively, and was competitively inhibited by the ADP receptor antagonist (Sp)-ATP[aS]. These results strongly suggest that the stimulated ["S]GTP[yS] binding observed was the result of agonist-liganded receptor activation. Similar data were obtained for ADP, with half-maximal and maximal stimulation of [3'S]GTP[yS] binding at 0.1 pM and 10 pM, respectively. One striking observation was that in the presence of (Sp)-ATP[aS], control values of [3 'S]GTP[yS] binding were decreased. This may result from the presence of endogenous ADP in the platelet membrane preparations, although it cannot be excluded that antagonist liganded receptors may exert a negative influence on the G protein activation state. In addition, (Sp)-ATP[aS] may directly interfere with binding of [3sS]GTP[yS]to G proteins. All these reasons are not mutually exclusive. Nevertheless, by increasing the concentrations of 2MeSADP or ADP (up to 1 mM), a full response could be restored (data not shown), indicating that the inhibitory effect of (Sp)-ATP[ctS] is competitive. Other agents known to act in platelets through activation of G proteins, such as adrenalin, PGE, and thrombin, were also studied for their ability to stimulate [35Ss]GTP[yS]binding to G proteins in human platelet membranes. These compounds significantly increased GTP[yS] binding with apparently different efficacies, except for thrombin which, surprisingly, was only weakly effective in our system. In contrast, a synthetic peptide corresponding to the newly exposed Nterminal sequence of the recently described thrombin receptor after thrombin proteolysis [33] was strongly effective. The question of the poor efficiency of thrombin at present remains unresolved. We can only speculate that the thrombin receptor is somehow altered during membrane preparation or is not readily accessible to thrombin in the membrane vesicles or that proteolytic activity of thrombin is lost. Nevertheless, we were able to demonstrate that agents known to induce G protein activation stimulated [35S]GTP[yS]binding to human platelet membranes and that their effects were not impaired by (Sp)-ATP[aS], indicating that this platelet aggregation inhibitor is indeed specific for the ADP receptor (Table 1). Similar results have been obtained with rat platelet membranes [37]. The question which now arises is to identify the putative G protein(s) coupled to the ADP receptor(s). Treatment of human platelet membranes with N-ethylmaleimide (10 300 pM) resulted in a concentration-dependent decrease of [35S]GTP[yS]binding induced by ADP, 2MeSADP, adrenalin or the thrombin receptor peptide, whereas PGE,-stimulated binding of [35S]GTP[yS]to platelet membranes was unaffected (data not shown). N-Ethylmaleimide has been previously shown to inhibit adrenalin and thrombin effects on adenylyl cyclase and GTPase activities related to Gi proteins [38]. Although this effect of N-ethylmaleimide on ADP-stimulated [35S]GTP[yS]binding does not provide direct evidence for the involvement of a heterotrimeric G protein, such a transduction mechanism is highly feasible. The effect of pertussis toxin treatment of human platelet membranes would be helpful to
263 characterize the G proteins activated by ADP under our assay conditions. Unfortunately, probably because of the lability of our membrane material, we were unable to obtain conclusive data following pertussis toxin treatment and the question of the nature of the GTP-binding protein(s) activated by ADP and 2MeSADP is still unresolved at the present time. Finally, we cannot exclude the possibility that low-molecular-weight G proteins could be involved in our system. In conclusion, we have demonstrated that (a) platelet ADP receptor(s) are coupled to G protein(s), the nature of which remains to be identified, and (b) that agonist-activated ADP receptors stimulate the binding of GTP[yS] to G proteins as do other receptor agonists in this and other systems. The following questions arise. (a) Is the G-protein-coupling system linked to ADP-induced aggregation or to adenylyl cyclase inhibition or to both? (b) Which G protein(s) couple ADP receptor(s) to their effectors ?
REFERENCES 1. Hellem, A. J. (1960) Scand. J . Clin. Lab. Invest. 12 (suppl. 51),
1-117. 2. Gaardner, A., Jonsen, J., Hellem, A. J. & Owren, P. A. (1961) Nature 191, 513-532. 3. Ollgaard, E. (1961) Thromb. Diath. Haemorrh. 6, 86-97. 4. Born, G . V. R. (1962) Nuture 194,927-929. 5. Mustard, J. F., Packham, M. A,, Kinlough-Rathbone, R. L., Perry, D. W. & Regoeczi, E. (1978) Blood52,453-466. 6. Marguerie, G. A., Edgington, T. S. & Plow, E. F. (1980) J . Biol. Chem. 255, 154-161. I. Sage, S. 0. & Rink, T. J. (19x6) Biochem. Biophys. Res. Commun. 136, 1124-1129. 8. Cole, B., Robison, A. & Hartmann, R. C. (1971) Ann. N . Y . Acud. Sci. 185, 477-487. 9. Mills, D. C . B. & Smith, J. B. (1972) Ann. N. Y . Acad. Sci. 201, 391 - 399. 10. Haslam, R. J. (1973) Ser. Haematol. VZ, 333 - 350. 11. Haslam, R. J. & Cusack, N. J. (1981) in Purinergic receptors (Burnstock, G., ed.) Ch. 7, Chapman and Hall, London. 12. MacFarlane, D. E. (1987) in Platelet responses and metabolism (Holmsen, H., ed.) vol. 11, pp. 19-36, CRC Press, Boca Raton, FL. 13. Hawiger, J., Parkinson, S. & Timmons, S . (1980) Nature 283, 195- 197.
14. Williams, K. A., Murphy, W. & Haslam, R. J. (1987) Biochem. J . 243, 667 - 678. 15. Brass, L. F., Woolkalis, M. & Manning, D. R. (1988) J . Bid. Chem. 263, 5348-5355. 16. Lanza, F., Beretz, A,, Stierlt, A,, Hanau, D., Kubina, M. & Cazenave, J.-P. (19x8) Am. J . Physiol. 255 1276-1288. 17. Brass, L. F., Manning, D. R . & Shattil, S. J. (1991) in Progr. Hemostasis Thromh. 10, 27 - 174. 18. Aktories, K. & Jakobs, K. H. (1984) Eur. J. Riochrm. 145, 333 338. 19. Houslay, M. D., Bojanic D. &Wilson, A. (1986) Biochem. J . 234, 737 -740. 20. Houslay, M. D., Bojanic, D., Gansler, D., O’Hagan, S. &Wilson, A. (1986) Biochem. J . 238, 109-113. 21. Lapetina, E. G . (1990) FEBS Lett. 268,400-404. 22. Williams, A. G., Woolkalis, M. J., Poncz, M., Manning, D. R., Gewirtz, A. M. & Brass L. F. (1990) Blood 76, 721 -730. 23. Vickers, J. D., Kinlough-Rathbone, R. L., Packham, M. A. & Mustard, J. F. (1990) Eur. J . Biochem. 193, 521 -528. 24. Mellwig, K. P. & Jakobs, K. H. (1980) Thromb. Res. 18, 7- 17. 25. Cooper, D. M. F. & Rodbell, M. (1979) Nature 282, 517-518. 26. Hilf, G., Gierschik, P. & Jakobs, K . H. (1989) Eur. J . Biochem. 186,725-732. 27. Gierschik, P., Moghtader, R., Straub, C., Dieterich, K. & Jakobs, K. I-I. (1991) Eur. J. Biochem. 197,725-732. 28. Cough, G., Maguire, M. H. & Penglis, F. (1972) Mol. Pharmacol. 8, 170-177. 29. Cusack, N. J. & Hourani, S . M. 0. (1982) Br. J . Pharmacol. 76. 221 -227. 30. Jefferson, J. R., Harmon, J . T. & Jamieson, G. A. (1988) B l m d 71, 110-116. 31. Agarwal, A. K., Tandon, N. N., Greco, N. J., Cusack, H. J. & Jamieson, G. A. (1989) Thromb. Huemostus. 62, 1103- I 1 06. 32. MacFarlane, D. E., Srivasta, P. C. & Mills, D. C. B. (1983) J . Clin.Invest. 71,420-428. 33. Vu, T. K., Hung, D. T., Wheaton, V. I. & Coughlin, S. R. (1991) Cell 64, 1057- 1068. 34. Cazenave, J.-P., Hemmendinger, S., Beretz, A,, Sutter-Bay A. & Launay, J. (1983) Ann. Biol. Clin. 41, 167-179. 35. Barber, A. J. & Jamieson, G. A. (1970) J . Bid. Chem. 245,63576365. 36. Salomon, Y. (1979) Adv. Cyclic Nucleotide Res. 10, 35-55. 37. Cachet, C., Savi, P., Ohlmann, P., Maffrand, J.-P., Jakobs, K. H. & Cazenave, J.-P. (1992) Thromh. Huemostus., in the press. 38. Grandt, R., Aktories, K. & Jakobs, K. H. (1986) Biochem. J . 237, 669 - 674.
$3 FEBS 1992
ADP receptor-induced activation of guanine-nucleotide-binding proteins in human platelet membranes Christian GACHET ’, Jean-Pierre CAZENAVE’, Philippe OHLMANN ’, Gerhard HILF2, Thomas WIELAND’ and Karl H. JAKOBS’ ’ INSERM U.311, Biologie et Pharmacologie des Interactions du Sang avec les Vaisseaux et les Biomateriaux, Centre Regional de Transfusion Sanguine, Strasbourg, France ’ Pharmakologisches Institut, Universitat Heidelberg, Federal Republic of Germany Institut fur Pharmakologie der Universitat G H Essen, Federal Republic of Germany (Received March 5, 1992)
-
EJB 92 0306
ADP receptor-regulated binding of the labeled GTP analog, guanosine 5’-0-(3-[35S]thiotriphosphate) ([3 5S]GTP[yS]),to guanine-nucleotide-binding proteins (G proteins) was studied in human platelet membranes. The potent ADP receptor agonist, 2-methyl-thio-adenosine 5’-diphosphate (2MeSADP), a non-hydrolyzable analog of ADP, increased the binding of [3sS]GTP[yS] without apparent lag phase. Under optimal conditions, i.e. in the presence of GDP (1 - 10 pM), 2MeSADP increased the binding up to about threefold, with half-maximal and maximal increase observed at 10 nM and 1 pM 2MeSADP, respectively. ADP itself increased the binding of [35S]GTP[yS] by maximally about twofold, with half-maximal increase occurring at 0.1 pM ADP. The agonist-induced stimulation was competitively antagonized by the ADP receptor(s) antagonist, (1s)-adenosine 5‘-0(1-thiotriphosphate) { (Sp)-ATP[aS]). Other platelet receptor agonists known to act through receptors coupled to G proteins also increased binding of [35S]CTP[yS]in human platelet membranes, but without being inhibited by (Sp)-ATP[aS]. The data presented indicate that the platelet ADP receptor(s) can interact with and efficiently activate G proteins. the nature of which remains to be identified.
About 30 years ago, it was discovered that ADP was responsible for platelet adhesion to glass and platelet aggregation [I - 31. The development of a simple turbidimetric method [4] to follow quantitatively the aggregation of platelets in vitro made possible the study of the effects of a variety of agonists and antagonists on platelet function. Addition of ADP to washed human platelets results in shape change, exposure of the fibrinogen binding site on the GPIIb-IIIa complex and reversible aggregation in the presence of fibrinogen and physiological concentrations of Ca2’ [5, 61. At the intracellular level, platelet activation following ADP binding to its receptor leads to a transient rise in free cytoplasmic Ca2+ concentration, resulting both from C a 2 + influx across the plasma membrane and from C a 2 + mobilization from internal stores [7], and myosin-light-chain phosphorylation. In addition, ADP inhibits stimulated adenylyl cyclase [8 - 1I], but this is not the cause of platelet aggregation [12]. On the other hand, agents that increase platelet cyclic AMP levels, such as prostaglandin I2 and El (PGE1), are very potent Correspondence to C. Gachet, Service d’Hemostase et de Thrombose, Centre Regional de Transfusion Sanguine, 10, rue Spielmann, F-67085 Strasbourg Cedex, France Abbreviations. [35S]GTP[yS], guanosine 5’-0-(3-[35S]thiotriphosphate); 2MeSADP. 2-methylthioadenosine 5’-diphosphate; (Sp)ATP[&], (1 S)-adenosine 5’-O-(1 -thiotriphosphate); PGE,, prostaglandin E l ; G protein, guanine-nucleotide-bindingprotein; Gi, guanine nucleotide binding protein(s) involved in inhibition of adenylyl cyclase. Enzyme. Apyrase (EC 3.6.1.5). Note. This work was presented in part at the 33‘d Annual Meeting of the American Society of Hematology, Denver, USA, December 1991 [Gachet et al. (1991) Blood 78 Suppl. 1, 140al.
and efficient inhibitors of platelet activation and aggregation [12, 131. When washed human platelets are used, ADP has the same effects as thrombin in simultaneously inhibiting activated adenylyl cyclase and inducing platelet aggregation, whereas other agonists, such as platelet-activating factor and vasopressin, do not inhibit adenylyl cyclase in intact platelets [lo, 14, 151 or, like adrenalin [16], do not induce aggregation although inhibiting adenylyl cyclase. Thrombin activates phospholipase C via a so-far-unidentified guanine-nucleotidebinding protein (G protein) and inhibits adenylyl cyclase via the G protein, Gi [15, 17-22]. At the present time, there is limited information as far as the ADP receptor signal transduction pathway is concerned, probably because of unresolved technical problems which occur when ADP is studied in the assay systems commonly used. Recent investigations seem to have conclusively established that phospholipase C is not involved when ADP induces human platelet activation in a medium containing physiological concentrations of Ca2’. Under these conditions, aggregation occurs without evidence of degradation of phosphatidylinositol bisphosphate 1233. In one study [24], it was shown that ADP inhibited, like adrenalin, the adenylyl cyclase of platelet membrane preparations. The results of this work were not in agreement with the requirement for GTP in the expression of inhibitory activity of adenylyl cyclase induced by ADP as reported in another study [25], thus leaving still unresolved the question of the involvement of putative GTP binding protein(s) in the ADP signal transduction pathway(s). The aim of the present study was to determine the requirements for receptor-dependent Gprotein activation after ADP stimulation. It has been shown previously that agonist-stimulated binding of the radiolabe-
260 led GTP analog, guanosine 5’-0-(3-thiotriphosphate) ([3sS]GTP[yS]) can be used to analyse receptor - G-protein interactions in different cell membrane preparations [26, 271. Using the non-hydrolyzable analog of ADP, 2-methyl-thioadenosine 5‘-diphosphate (2MeSADP), as agonist and the specific ADP receptor antagonist, the S p diastereoisomer of adenosine 5’-0-(l-thiotriphosphate){(Sp)-ATP[aS]}[28 - 321, we were able to demonstrate that the agonist-activated ADP receptor stimulates binding of [35S]GTP[yS]to human platelet G proteins, an effect specifically blocked by a competitive antagonist of the ADP receptor.
perature. These platelets were loaded with glycerol (Prolabo, Paris, France) by centrifugation through a 0 - 30% (mass/ vol.) glycerol gradient and lysed in a hypotonic Tris buffer containing leupeptin (50 pg/ml), diisopropylfluorophosphate (1 mM), aprotinin (20 Ujml) (SanoG-Choay, Paris, France) and EDTA (2 mM). After lysis, the broken platelets were layered onto a 30% (massivol.) sucrose cushion and centrifuged for 4 h at 60000 g in a Beckmann ultracentrifuge. The ‘floating’ membranes were removed carefully with a plastic pipette, washed and pelleted by centrifugation at I00000 g. These membranes were biochemically characterized by SDS/ PAGE analysis and their functional integrity was checked by measuring adenylyl cyclase activity [36].
MATERIALS AND METHODS Materials
Binding of [35S]GTP[yS]
ADP, prostaglandin El (pGE1), adrenalin and diisopropylfluorophosphate were from Sigma Chemical co.(St Louis, MO, USA). Leupeptin was from Fluka Chemie AG (Buchs, Switzerland). GDP, GTP[yS] and (Sp)-ATP[aS] were from Boehringer (Mannheim, FRG). [3’S]GTP[yS] (12003400 Ci/mmol) was from New England Nuclear (Dreieich, FRG). 2MeSADP was obtained from Sanofi Recherche (Toulouse, France). A synthetic peptide (14 residues, > 98% purity), corresponding to the newly exposed N-terminal sequence of the thrombin receptor following proteolysis by thrombin [33], from Neosystem (Strasbourg, France). All other products were commercial reagents of analytical grade.
Aliquots of human platelet membranes were thawed and diluted in 10 mM triethanolamine/HCl, pH 7.4, to a protein concentration of 100 pg/ml. The reaction mixture for measuring [3’S]GTP[yS] binding contained, if not otherwise indicated, 50 mM Tris/HCl pH 7.5, 5 mM MgC12, 1 mM EDTA, 1 mM dithiothreitol, 1 pM GDP, 0.3 -0.5 nM [3sS]GTP[yS] (about 50000 cpm) and 1.5 -4 pg membrane protein in a total volume of 100 pl. Incubation was started by addition of the membrane suspension to the prewarmed reaction mixture and carried out in triplicate for 30 min at 25°C. The reaction was terminated by rapid filtration through nitrocellulose filters (45 pm porosity) under vacuum. The filters were washed with 12 ml50 mM TrisiHCI pH 7.5 containing 5 mM MgC12. Radioactivity bound to the membranes on the filters was determined in scintillation fluid containing toluene and Triton X-100. Non-specific binding determined in the presence of 10 pM unlabeled GTP[yS] amounted to about 0.4% of added [35S]GTP[yS]and was subtracted from the total bound radioactivity to determine the specific radioactivity. The intraassay variation of triplicates was less than 5% of the mean.
Preparation of washed human platelets Blood was collected from a forearm vein, six volumes of blood into one volume of acid/citrate/dextrose anticoagulant, and twice-washed platelet suspensions were prepared as previously described [34]. The final resuspending medium (pH 7.35) was Tyrode’s buffer containing 2 mM Ca2+,I mM Mg2+, 0.35% (massivol.) human serum albumin (Centre Regional de Transfusion Sanguine, Strasbourg, France) and apyrase (2 pg/ml, a concentration that converted 0.25 pM ATP to AMP within 2 min at 37°C). Platelets were stored at 37 “C throughout the experiments. Platelet count was adjusted in the final suspension to 300000/pl using a Baker 810 platelet counter (Baker Instruments, Allentown, PA, USA). Platelet aggregation studies Aggregation was measured at 37 “C by a turbidimetric method in a dual-channel Payton aggregometer (Payton Associates, Scarborough, Ontario, Canada). A 0.4-ml aliquot of platelet suspension was stirred at 1100 rpm and activated by addition of 50 pl ADP or 2MeSADP in the presence of 50 p1 (Sp)-ATP[xS] (10 pM final concentration) or 50 p1 diluent and diisopropylfluorophosphate-treated human fibrinogen (0.8 mg/ml) [34]. The extent of aggregation was estimated quantitatively by measuring the maximum curve height above the baseline level. Preparation of human platelet membranes Platelet membranes were prepared essentially as described by Barber and Jamieson [35]. Briefly, blood was obtained from healthy donors who had not taken any drug for at least two weeks. Platelets were washed as described [34] and resuspended in Tyrode’s buffer containing no Ca2+, in the presence of 2 mM EDTA and apyrase (2 pg/ml) at room tem-
RESULTS Washed human platelet aggregation 2MeSADP was a powerful aggregating agent, inducing half-maximal and maximal platelet aggregation at 0.05 pM and 0.5 pM, respectively, while ADP induced half-maximal and maximal response at 1 pM and 10 pM, respectively. Responses to both 2MeSADP and ADP were competitively inhibited by (Sp)-ATP[aS] (Fig. 1). This effect of (Sp)-ATP[aS] was specific, since it did not inhibit platelet aggregation induced by thrombin, platelet-activating factor or arachidonic acid (data not shown). Binding of [35S]GTP[yS]to platelet membranes As illustrated in Fig. 2, GDP inhibited the binding of [3sS]GTP[./S] to platelet membranes with half-maximal and maximal effects at approximatively 1 pM and 0.1 mM, respectively, when no agonist was added. In the presence of 2MeSADP (1 pM), the GDP inhibition curve shifted significantly to the right from the control curve, with maximal 2MeSADP-induced increase in [35S]GTP[yS]binding being observed at 10 pM GDP. Depending on the different membrane preparations used, maximal 2MeSADP-induced increase in [35S]GTP[yS]binding occurred at 1-10 pM added GDP.
261
10 20 30 40 50 60 incubation time (min)
log[agonist]/M
Fig. 1. ADP receptor-induced platelet aggregation. Washed human platelets were stimulated with increasing concentrations of 2MeSADP ( A , A) or increasing concentrations of ADP (0, O), in the absence ( A , 0) or presence (A,0 ) of 10 pM (Sp)-ATP[aS]. Results are expressed as mean values f SEM from three separate experiments.
Fig. 3. Kinetics of [35S]GTP[yS]binding to human platelet membranes. Human platelet membranes were incubated at 25‘C with 0.3 -0.5 nM [35S]GTP[yS]in the presence (0, 0) and absence (m, 0 ) of 1 pM 2MeSADP without (0, D) or with (0, 0 ) 1 pM GDP. At indicated times, samples were analyzed for bound [35S]GTP[yS].Results are from one typical experiment representative for more than three separate experiments. Each point is the mean SEM of triplicate determinations.
I
-8
-7 -6 -5 log[GDP]/M
-4
T
-3
Fig. 2. Influence of GDP and 2MeSADP on the binding of [35S]GTP[rS] to human platelet membranes. Specific binding of [35S]GTP[yS]to human platelet membranes was determinated as described in Materials and Methods in the absence (0) or presence of 1 pM 2MeSADP ( O ) , at increasing concentrations of GDP. Results are from one typical experiment representative for more than ten separate experiments. Each point is the mean j= SEM of triplicate determinations.
The effects of 2MeSADP and GDP on the kinetics of [3’S]GTP[yS] binding to membranes of human platelets are presented in Fig. 3. In the absence of GDP, addition of 2MeSADP (1 pM) led to a marginal increase in [35S]GTP[yS] binding (< 10% at any time point up to 60 min measured). In contrast, when GDP (1 pM) was present, 2MeSADP caused a large increase in binding of [35S]GTP[yS].This agonist effect occurred without apparent lag phase, increased continuously and amounted to almost 5 pmoljmg protein at 60 min. All subsequent experiments were performed for an incubation time of 30 min. Stimulation of [”S]GTP[yS] binding to human platelet membranes by 2MeSADP was concentration-dependent (Fig. 4A). Under standard assay conditions (1 pM GDP, 30 min at 25 “C), half-maximal and maximal stimulation were observed at 10 nM and 1 pM 2MeSADP, respectively. The increase in [3’S]GTP[yS] binding to human platelet G proteins induced by 2MeSADP was antagonized by (Sp)-ATP[aS] (Fig. 4A). In a similar fashion to its stable analog 2MeSADP,
-9
-8
-7
-6
-5
log[ZMeSADP] / M
Fig. 4. ADP receptor mediated stimulation of [35S]GTP[yS]binding to human platelet membranes. Binding of [35S]GTP[yS]to human platelet membranes was determined with 1 pM G D P at increasing concentrations of 2MeSADP (A) or ADP (B), in the absence (0)or presence ( 0 )of 10 pM (Sp)-ATP[aS]. Results are from one typical experiment representative for more than three separate experiments. Each point is the mean f SEM of triplicate determinations.
ADP stimulated [35S]GTP[yS] binding to human platelet membranes in a concentration-dependent manner, with halfmaximal and maximal stimulation observed at about 0.1 pM and 10 pM, respectively. This effect was also antagonized by (Sp)-ATP[aS] (Fig. 4B). Other platelet receptor agonists that act through G-protein-dependent transduction pathways also stimulated [3’S]GTP[yS] binding to purified platelet membranes (Table 1). Under standard assay conditions, PGEl (1 pM) stimulated [35S]GTP[yS]binding 1.2- 1.5-fold and adrenalin (10 pM) 1.4 - 2-fold. Thrombin (1 Ujml) stimulated the binding of [3’S]GTP[yS] very poorly (1 - 1.2-fold as compared to controls). In contrast, a synthetic peptide, corresponding to the newly exposed N-terminal sequence of the recently described platelet thrombin receptor [33] following proteolysis by thrombin, stimulated [35S]GTP[yS]binding to human platelet membranes from 1.6- 2-fold as compared to controls.
262 Table 1. lnfluence of (Sp)-ATP[aSJon [35S]GTPlySl binding to human platelet membranes stimulated by different agonists. Human platelet membranes were incubated for 1 h at 25°C in the absence or presence of 10 pM (Sp)-ATP[aS]. [3'S]GTP[yS] binding induced by adrenaline (10 pM), PGE, (1 yM). the thrombin peptide (10 pM), 2MeSADP (I y M ) and ADP ( 1 0 p M ) was measured. Data are the mean f SEM from four separate experiments. Statistical analysis using the Kruskall- Wallis test showed a significant decrease of agoniststimulated [35S]GTP[yS] binding only for 2MeSADP and ADP ( * P < 0.05, **P< 0.01).
Agonist
[35S]GTPIyS]binding
___ control
__ (Sp)-ATP[aS] (10 pM)
~~
Adrenalin PGE, Peptide 2MeSADP ADP
41+ 2 23+ 5 61 i 11 120+ 2 56+ 6
39 f 13 21+ 6 45+ 6 5 2 + 5* 1 8 + 2**
In order to assess the specificity of (Sp)-ATP[aS] to inhibit 2MeSADP-stimulated [3sS]GTP[yS]binding, human platelet membranes were incubated in the presence and absence of 10 pM (Sp)-ATP[aS] and [3sSs]GTP[yS]binding was stimulated with adrenalin, PGEl and the thrombin receptor peptide (Table 1). The stimulation of [35S]GTP[yS]binding to platelet membranes by any of these agonists was not reduced when measured in the presence of (Sp)-ATP[aS] (10 pM), a concentration that inhibited specifically the effect of 1 pM 2MeSADP (> 50%) and of 10 pM ADP ( > 60%) on [3sS]GTP[yS]binding. D1SCUSSJON
The stable, non-hydrolyzable analog of ADP, 2MeSADP, is at submicromolar concentrations a powerful aggregating agent. 2MeSADP is also 100-fold more potent than ADP in inhibiting CAMP accumulation in human platelets [32]. In addition, (Sp)-ATP[aS] is a specific antagonist of ADP-induced aggregation [29] and binds to a single class of receptors, displacing ADP analogs with an affinity of 17 nM [31]. In the present experiments, 2MeSADP induced platelet aggregation with half-maximal and maximal response at 0.05 pM and 0.5 pM, respectively, an effect competitively inhibited by (Sp)ATP[aS]. ADP-induced platelet aggregation with half-maxima1 and maximal effect at 1 pM and 10 pM, respectively, and was also competitively inhibited by (Sp)-ATP[aS]. These results are in agreement with other studies [28-321. Since ADP is metabolically rapidly degraded while 2MeSADP is not hydrolyzed and since (Sp)-ATP[aS] has the lowest Kd for putative ADP receptor(s), we decided to use these compounds in an attempt to determine whether ADP induces platelet activation through a signal transduction pathway involving activation of a G protein. Quantitative measurement of [35S]GTP[yS]binding to G proteins in membrane preparations from different cell systems is a sensitive tool to study the activation mechanisms of Gprotein-coupled receptors [26, 271. AS discussed in references [26, 271, the presence of GDP allows detection of agoniststimulated binding of [3sS]CTP[yS]by a mechanism which is not yet fully understood. It seems from studies using intact membranes, purified G proteins and receptors, that the
agonist-liganded receptor reduces the affinity of the interacting G protein for bound GDP, allowing [3sS]GTP[1/S]to bind to the G protein [26]. As shown in Figs 2 and 3, addition of GDP to platelet membranes increased not only the relative stimulation but also the absolute increase in GTP[yS] binding induced by 2MeSADP, with a maximal effect observed at 110 pM GDP, varying from one membrane preparation to another. The effect of 2MeSADP on [3sSs]GTP[yS]binding to human platelet membranes was concentration-dependent, reaching half-maximal and maximal effect at 10 nM and 1 pM, respectively, and was competitively inhibited by the ADP receptor antagonist (Sp)-ATP[aS]. These results strongly suggest that the stimulated ["S]GTP[yS] binding observed was the result of agonist-liganded receptor activation. Similar data were obtained for ADP, with half-maximal and maximal stimulation of [3'S]GTP[yS] binding at 0.1 pM and 10 pM, respectively. One striking observation was that in the presence of (Sp)-ATP[aS], control values of [3 'S]GTP[yS] binding were decreased. This may result from the presence of endogenous ADP in the platelet membrane preparations, although it cannot be excluded that antagonist liganded receptors may exert a negative influence on the G protein activation state. In addition, (Sp)-ATP[aS] may directly interfere with binding of [3sS]GTP[yS]to G proteins. All these reasons are not mutually exclusive. Nevertheless, by increasing the concentrations of 2MeSADP or ADP (up to 1 mM), a full response could be restored (data not shown), indicating that the inhibitory effect of (Sp)-ATP[ctS] is competitive. Other agents known to act in platelets through activation of G proteins, such as adrenalin, PGE, and thrombin, were also studied for their ability to stimulate [35Ss]GTP[yS]binding to G proteins in human platelet membranes. These compounds significantly increased GTP[yS] binding with apparently different efficacies, except for thrombin which, surprisingly, was only weakly effective in our system. In contrast, a synthetic peptide corresponding to the newly exposed Nterminal sequence of the recently described thrombin receptor after thrombin proteolysis [33] was strongly effective. The question of the poor efficiency of thrombin at present remains unresolved. We can only speculate that the thrombin receptor is somehow altered during membrane preparation or is not readily accessible to thrombin in the membrane vesicles or that proteolytic activity of thrombin is lost. Nevertheless, we were able to demonstrate that agents known to induce G protein activation stimulated [35S]GTP[yS]binding to human platelet membranes and that their effects were not impaired by (Sp)-ATP[aS], indicating that this platelet aggregation inhibitor is indeed specific for the ADP receptor (Table 1). Similar results have been obtained with rat platelet membranes [37]. The question which now arises is to identify the putative G protein(s) coupled to the ADP receptor(s). Treatment of human platelet membranes with N-ethylmaleimide (10 300 pM) resulted in a concentration-dependent decrease of [35S]GTP[yS]binding induced by ADP, 2MeSADP, adrenalin or the thrombin receptor peptide, whereas PGE,-stimulated binding of [35S]GTP[yS]to platelet membranes was unaffected (data not shown). N-Ethylmaleimide has been previously shown to inhibit adrenalin and thrombin effects on adenylyl cyclase and GTPase activities related to Gi proteins [38]. Although this effect of N-ethylmaleimide on ADP-stimulated [35S]GTP[yS]binding does not provide direct evidence for the involvement of a heterotrimeric G protein, such a transduction mechanism is highly feasible. The effect of pertussis toxin treatment of human platelet membranes would be helpful to
263 characterize the G proteins activated by ADP under our assay conditions. Unfortunately, probably because of the lability of our membrane material, we were unable to obtain conclusive data following pertussis toxin treatment and the question of the nature of the GTP-binding protein(s) activated by ADP and 2MeSADP is still unresolved at the present time. Finally, we cannot exclude the possibility that low-molecular-weight G proteins could be involved in our system. In conclusion, we have demonstrated that (a) platelet ADP receptor(s) are coupled to G protein(s), the nature of which remains to be identified, and (b) that agonist-activated ADP receptors stimulate the binding of GTP[yS] to G proteins as do other receptor agonists in this and other systems. The following questions arise. (a) Is the G-protein-coupling system linked to ADP-induced aggregation or to adenylyl cyclase inhibition or to both? (b) Which G protein(s) couple ADP receptor(s) to their effectors ?
REFERENCES 1. Hellem, A. J. (1960) Scand. J . Clin. Lab. Invest. 12 (suppl. 51),
1-117. 2. Gaardner, A., Jonsen, J., Hellem, A. J. & Owren, P. A. (1961) Nature 191, 513-532. 3. Ollgaard, E. (1961) Thromb. Diath. Haemorrh. 6, 86-97. 4. Born, G . V. R. (1962) Nuture 194,927-929. 5. Mustard, J. F., Packham, M. A,, Kinlough-Rathbone, R. L., Perry, D. W. & Regoeczi, E. (1978) Blood52,453-466. 6. Marguerie, G. A., Edgington, T. S. & Plow, E. F. (1980) J . Biol. Chem. 255, 154-161. I. Sage, S. 0. & Rink, T. J. (19x6) Biochem. Biophys. Res. Commun. 136, 1124-1129. 8. Cole, B., Robison, A. & Hartmann, R. C. (1971) Ann. N . Y . Acud. Sci. 185, 477-487. 9. Mills, D. C . B. & Smith, J. B. (1972) Ann. N. Y . Acad. Sci. 201, 391 - 399. 10. Haslam, R. J. (1973) Ser. Haematol. VZ, 333 - 350. 11. Haslam, R. J. & Cusack, N. J. (1981) in Purinergic receptors (Burnstock, G., ed.) Ch. 7, Chapman and Hall, London. 12. MacFarlane, D. E. (1987) in Platelet responses and metabolism (Holmsen, H., ed.) vol. 11, pp. 19-36, CRC Press, Boca Raton, FL. 13. Hawiger, J., Parkinson, S. & Timmons, S . (1980) Nature 283, 195- 197.
14. Williams, K. A., Murphy, W. & Haslam, R. J. (1987) Biochem. J . 243, 667 - 678. 15. Brass, L. F., Woolkalis, M. & Manning, D. R. (1988) J . Bid. Chem. 263, 5348-5355. 16. Lanza, F., Beretz, A,, Stierlt, A,, Hanau, D., Kubina, M. & Cazenave, J.-P. (19x8) Am. J . Physiol. 255 1276-1288. 17. Brass, L. F., Manning, D. R . & Shattil, S. J. (1991) in Progr. Hemostasis Thromh. 10, 27 - 174. 18. Aktories, K. & Jakobs, K. H. (1984) Eur. J. Riochrm. 145, 333 338. 19. Houslay, M. D., Bojanic D. &Wilson, A. (1986) Biochem. J . 234, 737 -740. 20. Houslay, M. D., Bojanic, D., Gansler, D., O’Hagan, S. &Wilson, A. (1986) Biochem. J . 238, 109-113. 21. Lapetina, E. G . (1990) FEBS Lett. 268,400-404. 22. Williams, A. G., Woolkalis, M. J., Poncz, M., Manning, D. R., Gewirtz, A. M. & Brass L. F. (1990) Blood 76, 721 -730. 23. Vickers, J. D., Kinlough-Rathbone, R. L., Packham, M. A. & Mustard, J. F. (1990) Eur. J . Biochem. 193, 521 -528. 24. Mellwig, K. P. & Jakobs, K. H. (1980) Thromb. Res. 18, 7- 17. 25. Cooper, D. M. F. & Rodbell, M. (1979) Nature 282, 517-518. 26. Hilf, G., Gierschik, P. & Jakobs, K . H. (1989) Eur. J . Biochem. 186,725-732. 27. Gierschik, P., Moghtader, R., Straub, C., Dieterich, K. & Jakobs, K. I-I. (1991) Eur. J. Biochem. 197,725-732. 28. Cough, G., Maguire, M. H. & Penglis, F. (1972) Mol. Pharmacol. 8, 170-177. 29. Cusack, N. J. & Hourani, S . M. 0. (1982) Br. J . Pharmacol. 76. 221 -227. 30. Jefferson, J. R., Harmon, J . T. & Jamieson, G. A. (1988) B l m d 71, 110-116. 31. Agarwal, A. K., Tandon, N. N., Greco, N. J., Cusack, H. J. & Jamieson, G. A. (1989) Thromb. Huemostus. 62, 1103- I 1 06. 32. MacFarlane, D. E., Srivasta, P. C. & Mills, D. C. B. (1983) J . Clin.Invest. 71,420-428. 33. Vu, T. K., Hung, D. T., Wheaton, V. I. & Coughlin, S. R. (1991) Cell 64, 1057- 1068. 34. Cazenave, J.-P., Hemmendinger, S., Beretz, A,, Sutter-Bay A. & Launay, J. (1983) Ann. Biol. Clin. 41, 167-179. 35. Barber, A. J. & Jamieson, G. A. (1970) J . Bid. Chem. 245,63576365. 36. Salomon, Y. (1979) Adv. Cyclic Nucleotide Res. 10, 35-55. 37. Cachet, C., Savi, P., Ohlmann, P., Maffrand, J.-P., Jakobs, K. H. & Cazenave, J.-P. (1992) Thromh. Huemostus., in the press. 38. Grandt, R., Aktories, K. & Jakobs, K. H. (1986) Biochem. J . 237, 669 - 674.
