- Email: [email protected]
Vasopressin inhibition of human platelet adenylate cyclase: variable responsiveness between donors and involvement of a G-protein different from Gi
European Journal of Pharmacology, 150 (1988) 367-372
367
Elsevier EJP 50288
Vasopressin inhibition of human platelet adenylate cyclase: variable responsiveness between donors and involvement of a G-protein different from Gi Daniel Vittet *, Marie-No,lie Mathieu, Bernard Cantau 1 and Claude Chevillard I N S E R M U. 300, Facult~ de Pharmacie, Avenue Charles Flahaut, 34060 Montpellier, and 1 Centre CNRS-INSERM de Pharmacologie-Endocrinologie, 34033 Montpellier C~dex, France
Received 26 November 1987, revised MS received 15 February 1988, accepted 8 March 1988
There is controversy concerning the inhibitory effect of arginine-vasopressin (AVP) on human platelet adenylate cyclase activity, which putatively involves G i as the G-protein. To clarify this point, the effects of AVP on human platelet membranes were studied by measuring the activities of the high-affinity GTPase, as an index of G-protein involvement, and of adenylate cyclase. AVP stimulated GTPase activity in a dose-dependent fashion (KAct = 1.1 + 0.2 nM) and caused a parallel adenylate cyclase inhibition (KAct = 1.3 + 0.7 nM). The extent of these AVP-induced responses varied considerably from one subject to another but they were linearly related, suggesting a causal relationship between the two activities. Moreover, a difference in responsiveness to the inhibitory effects to epinephrine on adenylate cyclase was also observed between donors. Since the AVP- and epinephrine-stimulated GTPase activities were additive at their respective maximal effect, and in view of the lack of linear relationship between AVP- and epinephrine-induced adenylate cyclase inhibition, our results suggest, that in spite of the AVP inhibitory action on platelet adenylate cyclase, the G-protein involved in this effect is different from G i. Adenylate cyclase; GTPase; Vasopressin; Epinephrine; Platelets
I. Introduction
Arginine-vasopressin (AVP) has been shown to cause aggregation of human blood platelets (Haslam and Rosson, 1972). This effect appears to be mediated by activation of the V~-vasopressinreceptor subtype. Indeed, AVP stimulates phosphoinositide metabolism (Siess et al., 1986; Pollock and Mclntyre, 1986) and subsequently increases cytosolic free Ca 2+ (Hallam et al., 1984; Pollock and Mclntyre, 1986). The ligand specificity of a series of AVP-specific analogues (Vittet et al., 1986) also corresponds to the Vl-subtype of the AVP-platelet receptor. Consequently, stimula-
* To whom all correspondence should be addressed.
tion of an AVP-enhanced GTPase activity in platelet membranes (Houslay et al., 1986) is thought to reflect the activity of the G-protein, GpI, which is involved in the regulation of the phosphoinositide turnover that controls calciumsensitive processes (Spiegel, 1987). Another report (Vanderwel et al., 1983) shows that AVP is also able to inhibit platelet adenylate cyclase in the same concentration range as shows dose dependence for specific binding. Thus, the AVP action also appeared to be mediated by G i , i.e. the G-protein involved in the inhibition of adenylate cyclase (for review: Spiegel, 1987). Nonetheless, other investigators were only able to reproduce a significant AVP-induced adenylate cyclase inhibition with high concentrations of the peptide (Thibonnier and Roberts, 1985), or were
0014-2999/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
368 unable to demonstrate inhibition by AVP (Houslay et al., 1986). The present studies were designed to clarify this point and to characterize receptor-effector coupling mechanisms more fully by measuring simultaneously the GTPase and adenylate cyclase responses in platelet membrane preparations from a population of 10 different healthy subjects. Several subjects were used because considerable variations between donors have been observed with regard to platelet aggregatory responses to AVP (Pollock and Mclntyre, 1986; Roos et al., 1986).
2. Materials and methods
2.1. Platelet membrane preparation Platelets were isolated from platelet-rich plasma prepared as previously described (Launay et al., 1987) by centrifugation at 600 x g for 15 min at room temperature. Crude membranes were prepared from the platelet pellet as previously described (Vittet et al., 1986) with 5 m M E G T A present during the preparation procedure. Protein concentrations were measured according to the method of Lowry et al. (1951). 2.2. GTPase assay GTPase activity was measured by monitoring the release of [32p]Pi from [7-32p]GTP essentially as described by Cassel and Selinger (1976). The incubation medium in a final volume of 100 /~1 contained: 10 m M MgC12, 5 mM phosphocreatine, 70 U / m l creatine kinase, 0.2 m M ATP, 0.1 m M EGTA, 1 m g / m l bovine serum albumin, 50 m M Tris/HC1, p H 7,4, and the appropriate drugs. The reaction was initiated by addition of membranes (10-20/zg of protein/tube) was allowed to proceed for 10 rain at 30°C, during which time AVP binding was near equilibrium (Vittet et al., 1986). The test tubes were then transferred to a water bath and incubated for 2 min at 20 ° C; 0.15 /~M [y-32p]GTP (Amersham France S.A., Les Ulis, 20-30 C i / m m o l ) was then added for a further 3-min incubation, over which time linear time
courses were obtained and initial rates were assessed. The reaction was terminated by the addition of 500 /~1 of cold 50 m M K H 2 P O 4 buffer, p H 7.4, containing 5% ( w / v ) Norit-A charcoal. The tubes were vortexed, centrifuged at 2500 x g for 10 min, and 300/zl of supernatant was taken for scintillation counting. The reactions were performed in quadruplicate and controls containing no added membranes were included. 2.3. Adenylate cyclase assay Adenylate cyclase activity was measured from the rate of conversion of [Or-32 P]ATP into labelled cyclic AMP, with essentially the same reaction mixture as described above for GTPase assay with 1 mM 3-isobutyl-l-methylxanthine (IBMX) included and 0.3/~M [a-32p]ATP ( 1 / i C i / t u b e ) (New England Nuclear, 35 C i / m m o l ) substituted for ['/- 32p]GTP. Furthermore, 10 /~M G T P and 100 m M NaC1 were added to obtain an optimal AVP effect on adenylate cyclase activity (Vanderwel et al., 1983). We have checked that the presence of these agents did not significantly affect the AVPinduced GTPase response. The reaction was initiated by the addition of 10-40 /~g of membrane protein per assay and was allowed to proceed for 15 min at 30 o C; it was stopped by the addition of 1 ml of a solution containing: 70 m M sodium dodecyl sulfate, 50 m M Tris HC1, p H 7.4, 1 mM ATP and 1 m M cAMP. Labelled cyclic AMP was separated according to Salomon et al. (1974) and was counted by liquid scintillation spectrometry. All values were corrected for the cAMP recovery estimated from the recovery of [3H]cyclic AMP (0.01 /~Ci/tube) (New England Nuclear, 30-50 C i / m m o l ) added to the incubation medium, and from a blank value determined in the absence of membranes.
3. Results
AVP, like epinephrine (Aktories and Jakobs, 1981), stimulated a high-affinity GTPase activity in human platelet membranes with a K M of 0.15 /~M for G T P (results now shown). The AVPstimulated GTPase activity was fully additive with
369 3O
60.
(A)
5Q
•
20
40. m
"~
"O C
IP ~
2
8
10_
._~ |
.m
2o.
,.
~
10.
_10.
o
o
-20_ AVP
EPI
110
AVP
I 9
÷
I 9
AVP,
I 7
. log M
EPI
Fig. 1. Ligand-stimulated human platelet GTPase activity. Activities are given as % of basal values. The values shown are from a representative experiment. Saturating concentrations of each ligand were used, such that no further increase of GTPase activity was elicited by any increase in concentration. The concentrations employed were AVP (100 nM) and epinephrine (100/~M). When epinephrine was used, propranolol (400/~M) was added to eliminate any possible fl-adrenergic effect.
15_
(B)
e3
+
=+
•
;o_ l
"~
the e p i n e p h r i n e - s t i m u l a t e d o n e w h e n s a t u r a t i n g c o n c e n t r a t i o n s of either of these ligands were used (fig. 1). A V P e n h a n c e d this G T P a s e activity i n a d o s e - d e p e n d e n t m a n n e r (fig. 2A) yielding a KAct value of 1.1 + 0.2 n M A V P (S.E., n = 5). I n addition, there was a parallel d o s e - d e p e n d e n t i n h i b i tion of a d e n y l a t e cyclase activity. H a l f - m a x i m a l i n h i b i t i o n was observed with 1.3 + 0.7 n M A V P (S.D., n = 2). Both A V P - s t i m u l a t e d G T P a s e activity (0-28%) a n d A V P - i n d u c e d a d e n y l a t e cyclase i n h i b i t i o n (015%) differed widely from one subject to another. There was a linear relationship (r = 0.93) b e t w e e n A V P - i n d u c e d m a x i m a l G T P a s e activity a n d A V P i n d u c e d m a x i m a l a d e n y l a t e cyclase i n h i b i t i o n in 10 different subjects (fig. 2B). Platelets from the highly A V P - r e s p o n s i v e subjects displayed a m a x i m a l i n h i b i t o r y or stimulatory adenylate cyclase response to e p i n e p h r i n e or p r o s t a g l a n d i n El, respectively. Conversely, plate-
•
-
"T
•
•
•
#e
'o
;,
AVP-induced ( %
of
GTPase basal
i
;o
stimulation
value
)
Fig. 2. Effects of AVP on adenylate cyclase activity and on high-affinity GTPase activity in human platelet membranes. (A) Dose-resiaonsecurves for AVP-induced changes in GTPase (e) and adenylate cyclase (o) activities in platelet membranes. Activities are expressed as % of the corresponding basal activity. The values shown are from an experiment representative of at least two experiments performed on different subjects. The basal GTPase activity was 14.5 pmol GTP hydrolyzed/mg protein per min, and the basal adenylate cyclase activity was 0.57 pmol cyclic AMP formed/mg protein per min. Arrows indicate the respective PKAct values. (B) Relationship between AVP-induced GTPase stimulation and AVP-induced adenylate cyclase inhibition. The maximal AVP-induced GTPase activity and corresponding maximal AVP-induced adenylate cyclase inhibition (both expressed as % of basal values) were measured on the same day in the same platelet membrane preparation. Thirteen values are shown, corresponding to 10 different healthy subjects.
370 group I (maximum adenylate cyclase responses: - 1 . 5 + 0.6% and +250 ___15% of basal values for epinephrine and prostaglandin El, respectively; S.E., n = 7). However, no relationship could be found between platelet adenylate cyclase responsiveness to AVP and to epinephrine or prostaglandin E 1 (fig. 3).
a
7I0_
5_ J= A
)lk 0
•
I
I
I
I
i
I0
20
30
40
50
EPI-induced
( % of
IS_
basal value)
B
e
•
I
10_
5_ o a o
0
|
!
100
200
PGEi.induced
4. D i s c u s s i o n
Adenylate cyciase inhibition
"
-!
I
i
z~
400
509
Adenylate cyclase s t i m u l a t i o n
( % of
basal
value)
Fig. 3. Relationship between AVP-induced maximal adenylate cyclase inhibition and adenylate cyclase responsivenessto epinephrine (A) or prostaglandin E1 (B) in 10 different healthy subjects. All ligands were used at saturating concentrations which elicited a maximal adenylate cyclase response; these were prostaglandin E1 (1/~M), epinephrine(100/~M) and AVP (100 nM). Propranolol (400 ttM) was added together with epinephrine to block any possible fl-adrenoceptoraction.
lets from poorly AVP-responsive subjects lacked the inhibitory adenylate cyclase response to epinephrine and had a weaker stimulatory adenylate cyclase response to prostaglandin E 1 (fig. 3). Thus, two groups of platelet donors could be distinguished (fig. 3): group I, whose platelet adenylate cyclase was fully responsive to epinephrine and to prostaglandin E 1 (maximum adenylate cyclase responses: - 4 2 + 3% and + 396 + 17% of basal values, respectively; S.E., n---5) and group II, whose platelet adenylate cyclase was unresponsive to epinephrine and exhibited a weaker stimulatory response to prostaglandin E l, compared to
The present investigation shows that AVP induced a dose-dependent stimulation of high-affinity GTPase activity, confirming the results of Houslay et al., 1986, and the inhibition of adenylate cyclase activity. The observed decrease in the cyclic AMP concentration was indeed the result of a direct adenylate cyclase inhibition by AVP, as IBMX, a potent cyclic nucleotide phosphodiesterase inhibitor was present during the assay. Both of these effects i.e. GTPase activation and adenylate cyclase inhibition showed identical dose-effect curves, with KAct of 1.1 and 1.3 nM, respectively, corresponding to the dose-dependent curve for AVP-specific binding on platelet membranes which exhibited a K d of 1.8 nM (Vittet et al., 1986). This identity suggests a causal relationship between these effects and a lack of effect of peptide binding to its sites and the first events subsequent to activation of this receptor, i.e. GTPase activation and adenylate cyclase inhibition. The causal relationship between these two events was confirmed by the linear correlation between AVP-induced GTPase stimulation and AVP-induced adenylate cyclase inhibition in different subjects. The inhibitory action of the peptide on platelet adenylate cyclase has been the object of controversy. Vanderwel et al., 1983 have shown an AVP-induced platelet membrane adenylate cyclase inhibition at low AVP concentrations ( K A c t = 1.2 nM) but others found such a significant inhibition only at high AVP concentrations, i.e. far higher than the K a value (Thibonnier and Roberts, 1985) or found no AVP inhibitory action on platelet adenylate cyclase (Houslay et al., 1986). The present results would appear to explain the discrepancies between the various findings. Although AVP-induced inhibition of platelet adenylate
371 cyclase was observed, this effect did not occur with platelets from all donors, and when present, its absolute value was always small. This could account for the non-significant results obtained by some authors, and be ascribable to the random selection of AVP-responsive and unresponsive subjects. The considerable variations in AVP-dependent GTPase activation as well as in AVP-induced adenylate cyclase inhibition measured in different subjects may reflect a homologous or heterologous regulation of the AVP-receptor coupling mechanisms. Considering the respective inhibitory and stimulatory effects of epinephrine and prostaglandin E 1 o n platelet membrane adenylate cyclase, the existence of two groups of platelet donors with different responsiveness to these agonists also suggests heterologous regulations of their receptor coupling mechanisms. Hormonal stimulation and inhibition of adenylate cyclase are mediated by the G-proteins G S and Gi, respectively, whereas the putative Gp~ protein mediates the agonist-induced activation of phospholipase C. G s and G i are inhibited by cholera and pertussis toxins, respectively. Like all G-proteins, they display high-affinity GTPase activity (Spiegel, 1987). The fact that AVP was able to inhibit adenylate cyclase in platelets suggests that the peptide acts through Gi, in addition to its stimulating action on G pi, leading to an increased turnover of inositol phospholipids (Siess et al., 1986). Nonetheless, three observations eliminate this possibility. First, as previously described by Houslay et al. (1986) when saturating concentrations of either AVP or epinephrine were used to stimulate GTPase activity, additive results were obtained when these agonists were added together, suggesting the implication of distinct G-proteins in these two ligand-receptor coupling mechanisms. Secondly, treatment of platelet membranes with pertussis toxin has been shown to have no effect on AVP-stimulated GTPase activity whereas it partially inhibits thrombin-stimulated GTPase activity (Houslay et al., 1986), the receptors for which are related two distinct G-proteins, namely G~ and Gp~ (Grandt et al., 1986; Houslay et al., 1986). Thirdly, if Gi were responsible for the AVP-evoked inhibition of adenylate cyclase in
platelet membrane, there would be a linear relationship between AVP- and epinephrine-induced adenylate cyclase inhibition, which did not occur in our experiments. Therefore, the G-protein involved in the AVP-induced inhibition of platelet membrane adenylate cyclase appears to be different from G i, which mediate epinephrine or thrombin-induced adenylate cyclase inhibition. The occurrence of distinct forms of cDNA coding for G i (for review, Spiegel, 1987) might result in the expression of different Gi-proteins, sensitive or not to pertussis toxin and which would mediate agonist-specific adenylate cyclase inhibition. The pertussis toxin effect on AVP-induced adenylate cyclase inhibition in comparison to ADP-ribosylation of specific Gi-proteins must be investigated before a conclusion can be reached. The vasopressin receptor subtype present on human platelet membrane is V1 (Thibonnier and Roberts, 1985; Vittet et al., 1986, Launay et al., 1987). V1-Activation mediates the stimulation of inositol phospholipid metabolism in many tissues. In the case of platelets, AVP has been shown to induce increases in the intracellular concentrations of diacylglycerol and inositol trisphosphate (Siess et al., 1986; Vittet et al., unpublished observation), suggesting the involvement of G Pi in the action of AVP. Nevertheless, the KAct value reported by Siess et al. (1986) for diacylglycerol formation, like the KAc t value we measured for inositol trisphosphate accumulation was about 50 nM AVP, which is far higher than both the K d value for AVP-specific binding on platelet membranes (1.8 nM) and the KAe t value now observed for AVP-induced inhibition of adenylate cyclase. Therefore, the low concentrations of AVP sufficient to inhibit adenylate cyclase activity and the higher concentrations required to stimulate phospholipase C suggest that the latter phenomenon may be indirect, or that platelet functional responsiveness is impaired under the conditions used to monitor diacylglycerol or inositol phosphate recovery. Nevertheless, further studies are needed to determine which of these transduction processes occurs initially or whether they occur concomitantly to induce platelet activation. The physiological significance of AVP-induced platelet aggregation nevertheless remains unclear
372 even though other aggregating agents such as epin e p h r i n e or A D P can p o t e n t i a t e it ( L a u n a y et al., 1987). Indeed, i n all these cases, the A V P conc e n t r a t i o n required to lead to the s a t u r a t i o n of receptors is far higher t h a n the h u m a n b l o o d levels of A V P ( H a s l a m a n d Rosson, 1972) a n d n o m a r k e d a m p l i f i c a t i o n was seen with the t r a n s d u c t i o n m e c h a n i s m s studied. T o summarize, the p r e s e n t investigation showed that A V P c a n i n h i b i t adenylate cyclase activity i n platelet m e m b r a n e s from h u m a n subjects i n a c o n c e n t r a t i o n - d e p e n d e n t m a n n e r , with a low KAct. This i n h i b i t i o n appears to b e subject to regulation i n view of the m a r k e d differences b e t w e e n the A V P responses of different donors, a n d given the linear relationship which occurs b e t w e e n A V P - i n duced G T P a s e s t i m u l a t i o n a n d adenylate cyclase i n h i b i t i o n . I n spite of the observed a d e n y l a t e cyclase i n h i b i t i o n , the G - p r o t e i n involved does n o t a p p e a r to be Gi, which mediates the e p i n e p h r i n e i n d u c e d a d e n y l a t e cyclase i n h i b i t i o n .
Acknowledgements This work was supported by the Institut National de la Sant6 et de la Recherche M&licale.The authors are grateful to Colette Bellegarde for expert secretarial assistance.
References Aktories, K. and K.H. Jakobs, 1981, Epinephrine inhibits adenylate cyclase and stimulates a GTPase in human platelet membranes via a-adrenoceptors, FEBS Lett. 130, 235. Cassel, D. and Z. Selinger, 1976, Catecholamine-stimulated GTPase activity in turkey erythrocyte membranes, Biochim. Biophys. Acta 452, 538. Grandt, R., K. Aktories and K.H. Jakobs, 1986, Evidence for two GTPases activated by thrombin in membranes of human platelets, Biochem. J. 237, 669.
Hallam, T.J., N.T. Thompson, M.C. Scrutton and T.J. Rink, 1984, The role of cytoplasmic free calcium in the responses of quin-2 loaded human platelets to vasopressin, Biochem. J. 221, 897. Haslam, R.J. and G.M. Rosson, 1972, Aggregation of human blood platelets by vasopressin, Am. J. Physiol. 223, 958. Houslay, M.D., D. Bojanic, D. Gawler, S. O'Hagan and A. Wilson, 1986, Thrombin, unlike vasopressin appears to stimulate two distinct guanine nucleotide regulatory proteins in human platelets, Biochem. J. 238, 109. Launay, J-M., D. Vittet, M. Vidand, A. Rondot, M-N. Mathieu, C. Lalau-Keraly, B. Cantau and C. Chevillard, 1987, Vlavasopressin specific receptors on human platelets: potentiation by ADP and epinephrine and evidence for homologous down-regulation, Thromb. Res. 45, 323. Lowry, O.H., N.H. Rosebrough, A.L. Fan and R.J. Randall, 1951, Protein measurement with the folin phenol reagent, J. Biol. Chem. 193, 265. Pollock, W.K. and D.E. Mclntyre, 1986. Desensitization and antagonism of vasopressin-induced phosphoinositide metabolism and elevation of cytosolic free calcium concentration in human platelets, Biochem. J. 234, 67. Roos, J., F. Ferracin and A. Pletscher, 1986, Interaction of vasopressin with human blood platelets: dependency on Mg 2÷, Thromb, Haemostas. 56, 260. Salomon, Y., C. Londos, and M. Rodbell, 1974, A highly sensitive adenylate cyclase assay method, Anal. Biochem. 58, 541. Siess, W., M. Stifel, H. Binder and P.C. Weber, 1986, Activation of Vl-receptors by vasopressin stimulates inositol phospholipid hydrolysis and arachidonate metabolism in human platelets, Biochem. J. 233, 83. Spiegel, A.M., 1987, Signal transduction by guanine nucleotide binding proteins, Mol. Cell. Endocrinol. 49, 1. Thibonnier, M. and J.M. Roberts, 1985, Characterization of human platelet vasopressin receptors, J. Clin. Invest. 76, 1987. Vanderwel, M., D.S. Lum and R.J. Haslam, 1983, Vasopressin inhibits the adenylate cyclase activity of human platelet particulate fraction through Vl-receptors, FEBS Lett. 164, 340. Vittet, D., A. Rondot, B. Cantau, J.-M. Launay and C. Chevillard, 1986, Nature and properties of human platelet vasopressin receptors, Biochem. J. 233, 631.
367
Elsevier EJP 50288
Vasopressin inhibition of human platelet adenylate cyclase: variable responsiveness between donors and involvement of a G-protein different from Gi Daniel Vittet *, Marie-No,lie Mathieu, Bernard Cantau 1 and Claude Chevillard I N S E R M U. 300, Facult~ de Pharmacie, Avenue Charles Flahaut, 34060 Montpellier, and 1 Centre CNRS-INSERM de Pharmacologie-Endocrinologie, 34033 Montpellier C~dex, France
Received 26 November 1987, revised MS received 15 February 1988, accepted 8 March 1988
There is controversy concerning the inhibitory effect of arginine-vasopressin (AVP) on human platelet adenylate cyclase activity, which putatively involves G i as the G-protein. To clarify this point, the effects of AVP on human platelet membranes were studied by measuring the activities of the high-affinity GTPase, as an index of G-protein involvement, and of adenylate cyclase. AVP stimulated GTPase activity in a dose-dependent fashion (KAct = 1.1 + 0.2 nM) and caused a parallel adenylate cyclase inhibition (KAct = 1.3 + 0.7 nM). The extent of these AVP-induced responses varied considerably from one subject to another but they were linearly related, suggesting a causal relationship between the two activities. Moreover, a difference in responsiveness to the inhibitory effects to epinephrine on adenylate cyclase was also observed between donors. Since the AVP- and epinephrine-stimulated GTPase activities were additive at their respective maximal effect, and in view of the lack of linear relationship between AVP- and epinephrine-induced adenylate cyclase inhibition, our results suggest, that in spite of the AVP inhibitory action on platelet adenylate cyclase, the G-protein involved in this effect is different from G i. Adenylate cyclase; GTPase; Vasopressin; Epinephrine; Platelets
I. Introduction
Arginine-vasopressin (AVP) has been shown to cause aggregation of human blood platelets (Haslam and Rosson, 1972). This effect appears to be mediated by activation of the V~-vasopressinreceptor subtype. Indeed, AVP stimulates phosphoinositide metabolism (Siess et al., 1986; Pollock and Mclntyre, 1986) and subsequently increases cytosolic free Ca 2+ (Hallam et al., 1984; Pollock and Mclntyre, 1986). The ligand specificity of a series of AVP-specific analogues (Vittet et al., 1986) also corresponds to the Vl-subtype of the AVP-platelet receptor. Consequently, stimula-
* To whom all correspondence should be addressed.
tion of an AVP-enhanced GTPase activity in platelet membranes (Houslay et al., 1986) is thought to reflect the activity of the G-protein, GpI, which is involved in the regulation of the phosphoinositide turnover that controls calciumsensitive processes (Spiegel, 1987). Another report (Vanderwel et al., 1983) shows that AVP is also able to inhibit platelet adenylate cyclase in the same concentration range as shows dose dependence for specific binding. Thus, the AVP action also appeared to be mediated by G i , i.e. the G-protein involved in the inhibition of adenylate cyclase (for review: Spiegel, 1987). Nonetheless, other investigators were only able to reproduce a significant AVP-induced adenylate cyclase inhibition with high concentrations of the peptide (Thibonnier and Roberts, 1985), or were
0014-2999/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
368 unable to demonstrate inhibition by AVP (Houslay et al., 1986). The present studies were designed to clarify this point and to characterize receptor-effector coupling mechanisms more fully by measuring simultaneously the GTPase and adenylate cyclase responses in platelet membrane preparations from a population of 10 different healthy subjects. Several subjects were used because considerable variations between donors have been observed with regard to platelet aggregatory responses to AVP (Pollock and Mclntyre, 1986; Roos et al., 1986).
2. Materials and methods
2.1. Platelet membrane preparation Platelets were isolated from platelet-rich plasma prepared as previously described (Launay et al., 1987) by centrifugation at 600 x g for 15 min at room temperature. Crude membranes were prepared from the platelet pellet as previously described (Vittet et al., 1986) with 5 m M E G T A present during the preparation procedure. Protein concentrations were measured according to the method of Lowry et al. (1951). 2.2. GTPase assay GTPase activity was measured by monitoring the release of [32p]Pi from [7-32p]GTP essentially as described by Cassel and Selinger (1976). The incubation medium in a final volume of 100 /~1 contained: 10 m M MgC12, 5 mM phosphocreatine, 70 U / m l creatine kinase, 0.2 m M ATP, 0.1 m M EGTA, 1 m g / m l bovine serum albumin, 50 m M Tris/HC1, p H 7,4, and the appropriate drugs. The reaction was initiated by addition of membranes (10-20/zg of protein/tube) was allowed to proceed for 10 rain at 30°C, during which time AVP binding was near equilibrium (Vittet et al., 1986). The test tubes were then transferred to a water bath and incubated for 2 min at 20 ° C; 0.15 /~M [y-32p]GTP (Amersham France S.A., Les Ulis, 20-30 C i / m m o l ) was then added for a further 3-min incubation, over which time linear time
courses were obtained and initial rates were assessed. The reaction was terminated by the addition of 500 /~1 of cold 50 m M K H 2 P O 4 buffer, p H 7.4, containing 5% ( w / v ) Norit-A charcoal. The tubes were vortexed, centrifuged at 2500 x g for 10 min, and 300/zl of supernatant was taken for scintillation counting. The reactions were performed in quadruplicate and controls containing no added membranes were included. 2.3. Adenylate cyclase assay Adenylate cyclase activity was measured from the rate of conversion of [Or-32 P]ATP into labelled cyclic AMP, with essentially the same reaction mixture as described above for GTPase assay with 1 mM 3-isobutyl-l-methylxanthine (IBMX) included and 0.3/~M [a-32p]ATP ( 1 / i C i / t u b e ) (New England Nuclear, 35 C i / m m o l ) substituted for ['/- 32p]GTP. Furthermore, 10 /~M G T P and 100 m M NaC1 were added to obtain an optimal AVP effect on adenylate cyclase activity (Vanderwel et al., 1983). We have checked that the presence of these agents did not significantly affect the AVPinduced GTPase response. The reaction was initiated by the addition of 10-40 /~g of membrane protein per assay and was allowed to proceed for 15 min at 30 o C; it was stopped by the addition of 1 ml of a solution containing: 70 m M sodium dodecyl sulfate, 50 m M Tris HC1, p H 7.4, 1 mM ATP and 1 m M cAMP. Labelled cyclic AMP was separated according to Salomon et al. (1974) and was counted by liquid scintillation spectrometry. All values were corrected for the cAMP recovery estimated from the recovery of [3H]cyclic AMP (0.01 /~Ci/tube) (New England Nuclear, 30-50 C i / m m o l ) added to the incubation medium, and from a blank value determined in the absence of membranes.
3. Results
AVP, like epinephrine (Aktories and Jakobs, 1981), stimulated a high-affinity GTPase activity in human platelet membranes with a K M of 0.15 /~M for G T P (results now shown). The AVPstimulated GTPase activity was fully additive with
369 3O
60.
(A)
5Q
•
20
40. m
"~
"O C
IP ~
2
8
10_
._~ |
.m
2o.
,.
~
10.
_10.
o
o
-20_ AVP
EPI
110
AVP
I 9
÷
I 9
AVP,
I 7
. log M
EPI
Fig. 1. Ligand-stimulated human platelet GTPase activity. Activities are given as % of basal values. The values shown are from a representative experiment. Saturating concentrations of each ligand were used, such that no further increase of GTPase activity was elicited by any increase in concentration. The concentrations employed were AVP (100 nM) and epinephrine (100/~M). When epinephrine was used, propranolol (400/~M) was added to eliminate any possible fl-adrenergic effect.
15_
(B)
e3
+
=+
•
;o_ l
"~
the e p i n e p h r i n e - s t i m u l a t e d o n e w h e n s a t u r a t i n g c o n c e n t r a t i o n s of either of these ligands were used (fig. 1). A V P e n h a n c e d this G T P a s e activity i n a d o s e - d e p e n d e n t m a n n e r (fig. 2A) yielding a KAct value of 1.1 + 0.2 n M A V P (S.E., n = 5). I n addition, there was a parallel d o s e - d e p e n d e n t i n h i b i tion of a d e n y l a t e cyclase activity. H a l f - m a x i m a l i n h i b i t i o n was observed with 1.3 + 0.7 n M A V P (S.D., n = 2). Both A V P - s t i m u l a t e d G T P a s e activity (0-28%) a n d A V P - i n d u c e d a d e n y l a t e cyclase i n h i b i t i o n (015%) differed widely from one subject to another. There was a linear relationship (r = 0.93) b e t w e e n A V P - i n d u c e d m a x i m a l G T P a s e activity a n d A V P i n d u c e d m a x i m a l a d e n y l a t e cyclase i n h i b i t i o n in 10 different subjects (fig. 2B). Platelets from the highly A V P - r e s p o n s i v e subjects displayed a m a x i m a l i n h i b i t o r y or stimulatory adenylate cyclase response to e p i n e p h r i n e or p r o s t a g l a n d i n El, respectively. Conversely, plate-
•
-
"T
•
•
•
#e
'o
;,
AVP-induced ( %
of
GTPase basal
i
;o
stimulation
value
)
Fig. 2. Effects of AVP on adenylate cyclase activity and on high-affinity GTPase activity in human platelet membranes. (A) Dose-resiaonsecurves for AVP-induced changes in GTPase (e) and adenylate cyclase (o) activities in platelet membranes. Activities are expressed as % of the corresponding basal activity. The values shown are from an experiment representative of at least two experiments performed on different subjects. The basal GTPase activity was 14.5 pmol GTP hydrolyzed/mg protein per min, and the basal adenylate cyclase activity was 0.57 pmol cyclic AMP formed/mg protein per min. Arrows indicate the respective PKAct values. (B) Relationship between AVP-induced GTPase stimulation and AVP-induced adenylate cyclase inhibition. The maximal AVP-induced GTPase activity and corresponding maximal AVP-induced adenylate cyclase inhibition (both expressed as % of basal values) were measured on the same day in the same platelet membrane preparation. Thirteen values are shown, corresponding to 10 different healthy subjects.
370 group I (maximum adenylate cyclase responses: - 1 . 5 + 0.6% and +250 ___15% of basal values for epinephrine and prostaglandin El, respectively; S.E., n = 7). However, no relationship could be found between platelet adenylate cyclase responsiveness to AVP and to epinephrine or prostaglandin E 1 (fig. 3).
a
7I0_
5_ J= A
)lk 0
•
I
I
I
I
i
I0
20
30
40
50
EPI-induced
( % of
IS_
basal value)
B
e
•
I
10_
5_ o a o
0
|
!
100
200
PGEi.induced
4. D i s c u s s i o n
Adenylate cyciase inhibition
"
-!
I
i
z~
400
509
Adenylate cyclase s t i m u l a t i o n
( % of
basal
value)
Fig. 3. Relationship between AVP-induced maximal adenylate cyclase inhibition and adenylate cyclase responsivenessto epinephrine (A) or prostaglandin E1 (B) in 10 different healthy subjects. All ligands were used at saturating concentrations which elicited a maximal adenylate cyclase response; these were prostaglandin E1 (1/~M), epinephrine(100/~M) and AVP (100 nM). Propranolol (400 ttM) was added together with epinephrine to block any possible fl-adrenoceptoraction.
lets from poorly AVP-responsive subjects lacked the inhibitory adenylate cyclase response to epinephrine and had a weaker stimulatory adenylate cyclase response to prostaglandin E 1 (fig. 3). Thus, two groups of platelet donors could be distinguished (fig. 3): group I, whose platelet adenylate cyclase was fully responsive to epinephrine and to prostaglandin E 1 (maximum adenylate cyclase responses: - 4 2 + 3% and + 396 + 17% of basal values, respectively; S.E., n---5) and group II, whose platelet adenylate cyclase was unresponsive to epinephrine and exhibited a weaker stimulatory response to prostaglandin E l, compared to
The present investigation shows that AVP induced a dose-dependent stimulation of high-affinity GTPase activity, confirming the results of Houslay et al., 1986, and the inhibition of adenylate cyclase activity. The observed decrease in the cyclic AMP concentration was indeed the result of a direct adenylate cyclase inhibition by AVP, as IBMX, a potent cyclic nucleotide phosphodiesterase inhibitor was present during the assay. Both of these effects i.e. GTPase activation and adenylate cyclase inhibition showed identical dose-effect curves, with KAct of 1.1 and 1.3 nM, respectively, corresponding to the dose-dependent curve for AVP-specific binding on platelet membranes which exhibited a K d of 1.8 nM (Vittet et al., 1986). This identity suggests a causal relationship between these effects and a lack of effect of peptide binding to its sites and the first events subsequent to activation of this receptor, i.e. GTPase activation and adenylate cyclase inhibition. The causal relationship between these two events was confirmed by the linear correlation between AVP-induced GTPase stimulation and AVP-induced adenylate cyclase inhibition in different subjects. The inhibitory action of the peptide on platelet adenylate cyclase has been the object of controversy. Vanderwel et al., 1983 have shown an AVP-induced platelet membrane adenylate cyclase inhibition at low AVP concentrations ( K A c t = 1.2 nM) but others found such a significant inhibition only at high AVP concentrations, i.e. far higher than the K a value (Thibonnier and Roberts, 1985) or found no AVP inhibitory action on platelet adenylate cyclase (Houslay et al., 1986). The present results would appear to explain the discrepancies between the various findings. Although AVP-induced inhibition of platelet adenylate
371 cyclase was observed, this effect did not occur with platelets from all donors, and when present, its absolute value was always small. This could account for the non-significant results obtained by some authors, and be ascribable to the random selection of AVP-responsive and unresponsive subjects. The considerable variations in AVP-dependent GTPase activation as well as in AVP-induced adenylate cyclase inhibition measured in different subjects may reflect a homologous or heterologous regulation of the AVP-receptor coupling mechanisms. Considering the respective inhibitory and stimulatory effects of epinephrine and prostaglandin E 1 o n platelet membrane adenylate cyclase, the existence of two groups of platelet donors with different responsiveness to these agonists also suggests heterologous regulations of their receptor coupling mechanisms. Hormonal stimulation and inhibition of adenylate cyclase are mediated by the G-proteins G S and Gi, respectively, whereas the putative Gp~ protein mediates the agonist-induced activation of phospholipase C. G s and G i are inhibited by cholera and pertussis toxins, respectively. Like all G-proteins, they display high-affinity GTPase activity (Spiegel, 1987). The fact that AVP was able to inhibit adenylate cyclase in platelets suggests that the peptide acts through Gi, in addition to its stimulating action on G pi, leading to an increased turnover of inositol phospholipids (Siess et al., 1986). Nonetheless, three observations eliminate this possibility. First, as previously described by Houslay et al. (1986) when saturating concentrations of either AVP or epinephrine were used to stimulate GTPase activity, additive results were obtained when these agonists were added together, suggesting the implication of distinct G-proteins in these two ligand-receptor coupling mechanisms. Secondly, treatment of platelet membranes with pertussis toxin has been shown to have no effect on AVP-stimulated GTPase activity whereas it partially inhibits thrombin-stimulated GTPase activity (Houslay et al., 1986), the receptors for which are related two distinct G-proteins, namely G~ and Gp~ (Grandt et al., 1986; Houslay et al., 1986). Thirdly, if Gi were responsible for the AVP-evoked inhibition of adenylate cyclase in
platelet membrane, there would be a linear relationship between AVP- and epinephrine-induced adenylate cyclase inhibition, which did not occur in our experiments. Therefore, the G-protein involved in the AVP-induced inhibition of platelet membrane adenylate cyclase appears to be different from G i, which mediate epinephrine or thrombin-induced adenylate cyclase inhibition. The occurrence of distinct forms of cDNA coding for G i (for review, Spiegel, 1987) might result in the expression of different Gi-proteins, sensitive or not to pertussis toxin and which would mediate agonist-specific adenylate cyclase inhibition. The pertussis toxin effect on AVP-induced adenylate cyclase inhibition in comparison to ADP-ribosylation of specific Gi-proteins must be investigated before a conclusion can be reached. The vasopressin receptor subtype present on human platelet membrane is V1 (Thibonnier and Roberts, 1985; Vittet et al., 1986, Launay et al., 1987). V1-Activation mediates the stimulation of inositol phospholipid metabolism in many tissues. In the case of platelets, AVP has been shown to induce increases in the intracellular concentrations of diacylglycerol and inositol trisphosphate (Siess et al., 1986; Vittet et al., unpublished observation), suggesting the involvement of G Pi in the action of AVP. Nevertheless, the KAct value reported by Siess et al. (1986) for diacylglycerol formation, like the KAc t value we measured for inositol trisphosphate accumulation was about 50 nM AVP, which is far higher than both the K d value for AVP-specific binding on platelet membranes (1.8 nM) and the KAe t value now observed for AVP-induced inhibition of adenylate cyclase. Therefore, the low concentrations of AVP sufficient to inhibit adenylate cyclase activity and the higher concentrations required to stimulate phospholipase C suggest that the latter phenomenon may be indirect, or that platelet functional responsiveness is impaired under the conditions used to monitor diacylglycerol or inositol phosphate recovery. Nevertheless, further studies are needed to determine which of these transduction processes occurs initially or whether they occur concomitantly to induce platelet activation. The physiological significance of AVP-induced platelet aggregation nevertheless remains unclear
372 even though other aggregating agents such as epin e p h r i n e or A D P can p o t e n t i a t e it ( L a u n a y et al., 1987). Indeed, i n all these cases, the A V P conc e n t r a t i o n required to lead to the s a t u r a t i o n of receptors is far higher t h a n the h u m a n b l o o d levels of A V P ( H a s l a m a n d Rosson, 1972) a n d n o m a r k e d a m p l i f i c a t i o n was seen with the t r a n s d u c t i o n m e c h a n i s m s studied. T o summarize, the p r e s e n t investigation showed that A V P c a n i n h i b i t adenylate cyclase activity i n platelet m e m b r a n e s from h u m a n subjects i n a c o n c e n t r a t i o n - d e p e n d e n t m a n n e r , with a low KAct. This i n h i b i t i o n appears to b e subject to regulation i n view of the m a r k e d differences b e t w e e n the A V P responses of different donors, a n d given the linear relationship which occurs b e t w e e n A V P - i n duced G T P a s e s t i m u l a t i o n a n d adenylate cyclase i n h i b i t i o n . I n spite of the observed a d e n y l a t e cyclase i n h i b i t i o n , the G - p r o t e i n involved does n o t a p p e a r to be Gi, which mediates the e p i n e p h r i n e i n d u c e d a d e n y l a t e cyclase i n h i b i t i o n .
Acknowledgements This work was supported by the Institut National de la Sant6 et de la Recherche M&licale.The authors are grateful to Colette Bellegarde for expert secretarial assistance.
References Aktories, K. and K.H. Jakobs, 1981, Epinephrine inhibits adenylate cyclase and stimulates a GTPase in human platelet membranes via a-adrenoceptors, FEBS Lett. 130, 235. Cassel, D. and Z. Selinger, 1976, Catecholamine-stimulated GTPase activity in turkey erythrocyte membranes, Biochim. Biophys. Acta 452, 538. Grandt, R., K. Aktories and K.H. Jakobs, 1986, Evidence for two GTPases activated by thrombin in membranes of human platelets, Biochem. J. 237, 669.
Hallam, T.J., N.T. Thompson, M.C. Scrutton and T.J. Rink, 1984, The role of cytoplasmic free calcium in the responses of quin-2 loaded human platelets to vasopressin, Biochem. J. 221, 897. Haslam, R.J. and G.M. Rosson, 1972, Aggregation of human blood platelets by vasopressin, Am. J. Physiol. 223, 958. Houslay, M.D., D. Bojanic, D. Gawler, S. O'Hagan and A. Wilson, 1986, Thrombin, unlike vasopressin appears to stimulate two distinct guanine nucleotide regulatory proteins in human platelets, Biochem. J. 238, 109. Launay, J-M., D. Vittet, M. Vidand, A. Rondot, M-N. Mathieu, C. Lalau-Keraly, B. Cantau and C. Chevillard, 1987, Vlavasopressin specific receptors on human platelets: potentiation by ADP and epinephrine and evidence for homologous down-regulation, Thromb. Res. 45, 323. Lowry, O.H., N.H. Rosebrough, A.L. Fan and R.J. Randall, 1951, Protein measurement with the folin phenol reagent, J. Biol. Chem. 193, 265. Pollock, W.K. and D.E. Mclntyre, 1986. Desensitization and antagonism of vasopressin-induced phosphoinositide metabolism and elevation of cytosolic free calcium concentration in human platelets, Biochem. J. 234, 67. Roos, J., F. Ferracin and A. Pletscher, 1986, Interaction of vasopressin with human blood platelets: dependency on Mg 2÷, Thromb, Haemostas. 56, 260. Salomon, Y., C. Londos, and M. Rodbell, 1974, A highly sensitive adenylate cyclase assay method, Anal. Biochem. 58, 541. Siess, W., M. Stifel, H. Binder and P.C. Weber, 1986, Activation of Vl-receptors by vasopressin stimulates inositol phospholipid hydrolysis and arachidonate metabolism in human platelets, Biochem. J. 233, 83. Spiegel, A.M., 1987, Signal transduction by guanine nucleotide binding proteins, Mol. Cell. Endocrinol. 49, 1. Thibonnier, M. and J.M. Roberts, 1985, Characterization of human platelet vasopressin receptors, J. Clin. Invest. 76, 1987. Vanderwel, M., D.S. Lum and R.J. Haslam, 1983, Vasopressin inhibits the adenylate cyclase activity of human platelet particulate fraction through Vl-receptors, FEBS Lett. 164, 340. Vittet, D., A. Rondot, B. Cantau, J.-M. Launay and C. Chevillard, 1986, Nature and properties of human platelet vasopressin receptors, Biochem. J. 233, 631.