- Email: [email protected]
Aeguorin-detected calcium changes in stimulated thrombasthenic platelets. Aggregation - dependent calcium movement in response to ADP
THROMBOSIS RESEARCH 58; 561-570,199O 0049-3848/90 $3.00 + .OOPrinted in the USA. Copyright (c) 1990 Pergamon Press plc. All rights reserved.
AEQUORIN - DETECTED CALCIUM CHANGES IN STIMULATED THROMBASTHENIC PLATELETS. AGGREGATION - DEPENDENT CALCIUM MOVEMENT IN RESPONSE TO ADP.
T. Lecompte ?? , F. Potevin ?? , P. Champeix * , M.-C. Morel ?? *, R. Favier ?? , M.-F. Hurtaud ?? **, N. Schlegel ?? **, M. Samama , C. Kaplan ** Laboratoire d’Hematologie - Hotel-Diu : 1, place du Parvis Notre-Dame 75004 * INTS *** Hopital R. Deb& Paris-France ??
??
??
??
(Received 29.7.1988; accepted in revised form 6.3.1990 by Editor M.B. Donati)
Calcium changes in normal and thrombasthenic platelets were recorded using the PICA-apparatus. Aequorin was loaded in the presence of DMSO, EGTA and PGEl. Platelets of three patients with type I thrombasthenia stimulated with A-23,187, thrombin, PMA in the presence 1mM Ca++ and 1mM Mg++ were able to normally raise their calcium concentrations. The maximal values could be found below the normal range with collagen, ADP and PAF-acether. Calcium mobilization from internal stores in response to thrombin was normal. There were two calcium peaks in normal platelets stimulated with ADP. The second one was suppressed by omitting fibrinogen, stirring, or by adding aspirin, and was absent in thrombasthenic platelets. Thus the GP Ilb-llla complex is not a prerequisite for calcium fluxes but is involved, when weak agonists such ADP are used, through an aggregation-dependent reinforcement of platelet activation.
Calcium plays a central role in platelet activation (1). Inositol-trisphosphate is a likely second messenger able to induce the release of calcium from the internal stores, often referred to as mobilization. An influx from the surrounding plasma across the plasma membrane also contributes to the increase in intracellular calcium. The molecular mechanisms of this influx remain to be fully elucidated. It has been suggested that the GP Ilb-llla complex might be involved (2), and so far two competing techniques have been used to adress this issue with intact platelets. One study using the fluorescent probe quin-2 led its authors to conclude that the complex is not required for the rapid influx that results from stimulation, but is -________---______-_--Key-words: thrombasthenia, calcium, aequorin, ADP 561
Ca++ MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
closely adjacent to the putative calcium channel (3). The more recent one used the photophore protein aequorin and brought some lines of evidence in favour of the role of the GP Ilb-llla complex in calcium movements, but there was no experiment with thrombasthenic (T.) platelets (4). With the aim of clarifying this issue, we compared calcium changes in normal and T. platelets, which are profoundly deficient in GP lib-llla (5).
Three patients (referred here to as A.A., F.C. and F.P.) with a lifelong bleeding disorder and a positive familial history were included in the present study. Each of them fulfilled the criteria for type I Glanzmann’s thrombasthenia: prolonged bleeding times with platelet counts above 120 Giga / L, normal routine coagulation parameters and von Willebrand factor plasma levels, lack of platelet aggregation on repeated testing regardless the agonist used, absent clot retraction. Glycoproteins Ilb and Illa, and fibrinogen could not be detected on sodium dodecyl sulfate - polyacrylamide gel electrophoresis (under reducing and non-reducing conditions, PAS and Coomassie blue staining). In addition binding sites for allo-antibodies against GP Ilb-llla and the PlAl antigen, as assessed by an indirect radiolabelled antiglobulin test, and binding sites for two monoclonal antibodies (6,7), PL2-49 (epitope on GP Ilb but complex-dependent binding) and PL2-73 (GP Ilb-llla complex), as assessed by a direct radioligand test, were dramatically decreased. Patients F.C. and F.P. belong to the same kindred, and F.P. was shown to be strongly immunized against the GP complex after multiple transfusions required for surgery (traumatic splenic rupture). This subject (F.P.) could be studied on three different times, his sister (F.C.) and the unrelated subject A.A.. twice. Controls were run throughout the same study period with blood collected from healthy volunteers. Platelets were loaded with aequorin according to Yamaguchi et al. (8) with some modifications. Briefly a platelet button was prepared from acidified titrated platelet rich plasma in the presence of unfractionated heparin, washed with a saline-Hepes buffer in the presence of glucose (5.6 mM), PGEl (0.1 PM), EGTA (5mM) and albumin (O.l%), pH adjusted to 7.3, incubated with aequorin in the same buffer and increasing amounts of dimethylsulfoxide (DMSO) in a stepwise manner up to 8% vol:vol final proportions, washed twice at 10,000 x g during 90 sec., and resuspended in a Tyrode’s buffer with albumin (0.35%), apyrase (0.03 U/mL), 1 mM calcium and 1 mM magnesium. The overall procedure was performed at room temperature, and the LDH release was less than 5 %. Aliquots (1 ml) of the platelet suspension (adjusted to 3 to 4 x 100,000 / uL) were stimulated under stirring at 1,100 rpm in an apparatus (PICA-Aggregometer from Chronolog Corporation - Coultronics France, Margency) able to record changes in light transmission and light emitted by the intracellular aequorin, in parallel. All experiments were performed in the presence of added fibrinogen (1.2 mg/mL), except when thrombin was the agonist or for particular experiments with ADP. Some experiments aimed at investigating calcium mobilization from the internal stores triggered by A-23,187 and thrombin were performed in the presence of EGTA (3 mM). Platelet testing was completed within 60 to 90 minutes after aequorin loading. Calcium levels were calculated assuming an intracellular magnesium concentration of 1 mM, according to Jonhson et al (9). There was no obvious difference between the loading efficiencies with normal and T. platelets. Sensitivities from 0.5 to 0.05 for detection of the luminescent signal generated by aequorin could be used depending on the strength of the used agonist, and allowed a good signal to background ratio (see curves in the figures). The following platelet activators were used at supramaximal concentrations: thrombin from Leo - 1 U/mL; collagen from Horm - 20 ug/mL; A-23.187 from Boehringer-Mannheim PAF-acether (a kind gift from Dr Benveniste) - 1.8 t&l; ADP - 100 f&t ; 1uM; Phorbol-Myristate Acetate (PMA) - 0.1 WM. The two latter agonists were obtained from Sigma. The endoperoxides-thromboxane analogue U-46,61 9 was obtained from Upjohn. The other reagents were obtained as follows: human fibrinogen, grade L: Kabi-Vitrum; bovine serum albumin, apyrase, EGTA and Triton X 100 from Sigma; aequorin was prepared by Dr Blinks and
Vol. 58, No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
563
was a gift from Coultronics-France; PGEl was from Upjohn ; iloprost (16 methyl, 18-19 didehydro-carbacyclin) was a kind gift from Dr Souvignet, Schering (Lys-lez-Lannoy, France); DMSO was obtained from Bruneau (Boulogne-Billancourt, France).
Typical simultaneous recordings of changes in light transmission and in aequorin-generated luminescence with stimulated normal platelets are showed in HQ. 1 The patterns of light transmission changes fit in with the well-known responses of washed platelets. In particular ADP-induced aggregation was markedly enhanced by the addition of exogenous fibrinogen. Aggregation was not induced by stirring alone, in the presence of fibrinogen.
FIG.1 Typical tracings obtained with normal, aequorin-loaded platelets. Aggregation tracings are shown in the upper part of the figure, and luminescence tracings recorded in parallel in its lower part. Platelets were stimulated with the following agonists: thrombin (T), calcium ionophore A-23,187 (I), collagen (C), PAF-acether (PAF) and phorbol-myristate acetate (PMA). The vertical bar indicates a 10% change in light transmission, the horizontal one one minute. Note that there is no linear relationship between the height of the luminescence peak and the maximal level in intra-platelet calcium, and that there is a time-dependent decrease in the intra-platelet active aequorin.
Thrombin and the calcium ionophore A-23,187 induced immediate and sharp luminescence signals corresponding to fast and maximal aggregation . With collagen there was a similar lag-time for the onset of the decrease in light transmission corresponding to the shape change, and the rise in calcium levels . ADP, PAF-acether and PMA induced calcium changes of lower amplitude. The luminescent signals were almost instantaneous but bell shaped (F~Q.I ), In the table I are the values of the peaks in calcium obtained after stimulation with the different agonists of control and T. platelets. Estimated maximal calcium changes in stimulated T. platelets fall into the normal ranges as far as calcium ionophore, thrombin and PMA are concerned. Signals elicited by collagen, ADP and PAF-acether tend to be lower than with normal platelets.
564
Ca+* MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
TABLE I Aequorin-Detected Calcium Peaka ~__~~_____~~~__~~~__.~~~~~__~~~~___~~~____~~_._~~~___~~~____~ Patient A-23,1 87 thrombin collagen ADP PAF _~___~~~___~~~__~~._~~~~~~__~~~~__~~~~__~~~~~~~~~~____~___~_~ 1-A 12.6 7.9 5.0 1-B 8.9 2.6 2-A 17.8 8.0 2.2 : 2-B 11.2 8.9 2.2 3-A 7.2 3.0 1.2 1 .o 3-B 9.5 5.6 1.4 _ 3-c 14.8 7.0 2.2 ___~__-____--_~_-_~~~~~~-~~~~~-~~~~-~~~~~~~~~~~~~~~~~~~~~~~~Normal 7.9 4.0 5.0 2.0 3.1 range * 31 .O 20.0 11.2 4.0 8.9
PMA
3.9
2.8
2.5 4.5
Maximal levels of intra-platelet calcium are expressed in f&l. Patients with Glanzmann’s thrombasthenia are 1: A.A. and 2: CF., studied on two separate occasions: patient 3: P.F., three separate experiments. ’ Control subjects: at least three. The median values obtained with control platelets were respectively: thrombin: 7.9 (n=18) , ADP: 2.1 (n-5), PAF: 5.0 (n=6). The results with thrombin have not a normal distribution, but most of them fall in a narrow range between 6.0 and 10.0 f&t.
TABLE II Calcium Mobilization Thrombin A-23,1 87 + Ca++ + EGTA + Ca++ + EGTA ___~______________~_._~~~__~~~~_____~_~._~~__~~~~~~~~~~_~~.~. P.F, 7.0 1.4 C.F. 8.9 1.8 ~_~~~~~__~~~~~~~~~~~~~~~-
14.8 6.3 11.a 4.5 _~~_~~___~~___._--_~~~~~~.~~~~~~~~~~
Normal values
4.1 1 .4 10.0 3.8 18.6 5.6 11.2 5.0 7.0 2.0 11.2 4.4 ____________~____~~_~~~~-~~.~~~__~~~~~~~~~~~~~.~~~----------~
Maximal calcium levels are expressed in uM. The addition to the external milieu of EGTA (3.mM) precludes any influx and therefore the remaining signals are the result of calcium mobilization from the internal stores. The experiments the results of which are given here are referred to as 3-C and 2-A in the table I. Normal values are: (i) for thrombin the range and the mean of the results of experiments with platelets of 5 different donors; (ii) for A-23,187 the individual results obtained with three different platelet suspensions. Values obtained with A-23,187 in the presence of EGTA indicate (i) the internal releasable calcium; (ii) the ability of the aequorin loaded in the platelets to detect such a calcium mobilization.
Vol. 58. No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
565
Further experiments were conducted in the presence of EGTA in order to investigate calcium mobilization from the internal stores (table II). The calcium ionophore A-23,187 and thrombin were still able to induce a change in the luminescent signal: the peak values were reduced to 30-40 % and 20-35 % as compared to those obtained in the presence of 1 mM extracellular calcium, respectfvely. Thrombin-induced calcium mobilization in normal and T. platelets was not different. Typical tracings obtained with T. platelets stimulated in the presence of extra-cellular calcium are shown in Figs. 2 and 3. Aggregation of T. platelets was markedly abnormal as indicated by the absence of increase or the very slight increase in light transmission. However the same patterns of luminescence were obtained with T. and control platelets with all agonists but ADP. Control platelets stimulated with ADP showed indeed a peculiar pattern of aequorin-detected calcium changes ( Fig.4 ). With most of the preparations there were two distinct signals, whereas with some others there was a sustained luminesoence following the initial change, with a slow return to the baseline level. With platelets from patients with Glanzmann’s thrombasthenia the calcium signal was always reduced to its first component. One typical example is given in Fjg. 5 . The remaining first peak could be abolished with the addition of the stable prostacyclin analogue iloprost (10 nM), and the shape change too, as shown in Fig. 5 . The responses of control platelets to ADP were further investigated ( Fig. 4 ). In the absence of exogenous fibrinogen, shape change was still detectable but there was only a very small decrease in light transmission thereafter (trace c, left-hand panel). The first peak of luminescence remained unaffected, whereas the second one disappeared. As far as aequorin-generated luminescence is concerned, the same holds true in the presence of aspirin (traces b left- and right- hand panels). Under these experimental conditions however, aggregation still occured, but at a reduced rate and to a lesser maximal extent, and was reversible. In the absence of stirring, aggregation, but not the first calcium peak, was suppressed (not shown). When aspirin-treated platelets were stimulated first with ADP, and thereafter with the endoperoxides-thromboxane mimetic U-46,61 9, aggregation was complete and irreversible and two calcium peaks could be elicited ( Fig. 6).
FIG. 2 Tracings obtained from thrombasthenic agonists as in the previous figures.
platelets (patient
1). Same legends for the
566
Ca++ MOVEMENTS & PLATELET AGGREGATION
T
I
I
Vol. 58, No. 6
PMA
-
I ,
FIG. 3 Tracings obtained with thrombasthenic
platelets (referred to as 3-B in the table I)
b
a
FIG. 4 Experiments performed with two preparations of control platelets, stimulated with ADP. In the presence of fibrinogen ( tracings a): first-peak maximal values of calcium are 2.2 and 4.0 uM respectively. On the left hand, there is a sustained increase in luminescence after the initial discharge. On the right hand, there are at least two distinct peaks. In the presence of fibrinogen and aspirin, 1.4 mM (tracings b): calcium peak values are 2.0 and 4.0 pM respectively (one peak only). In the absence of exogenous fibrinogen (tracing c, left-hand panel only): one single peak at 2.0 PM.
Vol. 58, No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
567
ADP
I
a
cs---I
b -+---, I
FIG. 5 ADP-induced changes in light transmission and luminescence with platelets from a thrombasthenic patient. The stimulation with ADP was performed in the absence (tracings a) or in the presence (tracings b) of the stable prostacyclin analogue iloprost (10 nM). The shape change (upper tracing) is clearly noticeable in a and suppressed in b, and the change in luminescence (lower tracing), indicating intraplatelet calcium levels, as well.
FIG.6 ADP- induced changes in light transmission and luminescence in a typical experiment performed with control platelets, in the presence of exogenous fibrinogen. On the right hand platelets were preincubated with aspirin before the addition of ADP: there is a decrease in maximal aggregation, and the calcium rise is reduced to its first component (peak value: 1.4 PM), very similar to the one induced by ADP in the absence of aspirin (left hand). The subsequent addition of the thromboxane mimetic U-46,619 (0.1 PM) triggers another calcium change (2.2 PM) accompanied by a reinforcement of aggregation. U-46,619 alone at the same concentration induces only a shape change (not shown).
568
Ca++ MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
DISCUSSION. As far as we know we report here the most extensive study with aequorin-loaded platelets according to Yamaguchi’s method. As shown in the figures all the agonists used induced clear-cut increases in luminescence, indicating rises in calcium: thrombin, collagen, A-23,187, ADP, PAF-a&her, and also arachidonic acid and the thromboxane mimetic structurally related to endoperoxides, U-46,619 (data not shown for these two latter agonists). Clearly higher concentrations of ADP were required to obtain a maximal aggregation with changes of light transmission of 30 % or more, as compared to platelets washed with our currently used standard protocol derived from that of Mustard et al (10). But initial rates of aggregation were high and the addition of fibrinogen markedly potentiated aggregation. Furthermore in more recent experiments with platelets prepared from normal volunteers we were able to detect calcium changes in response to 5 f&l ADP. It must be pointed out however that platelets prepared from a few donors became refractory to ADP: low amplitude of maximal aggregation together with slight increase in intracellular calcium and no second component (see below). We then studied platelets collected from patients affected with type I Glanzmann’s thrombasthenia, which were shown to be profoundly deficient in GP Ilb-llla. When strong activators were used, the peaks were sharp, the kinetics and maximal achieved levels were simllar to those obtained with control platelets. This is in agreement with our current knowledge since stimulated T. platelets are known to change their shape and to be able to release the content of their granules and synthetire thromboxane, indicating that the activation processes can occur (11, 12, 13, 14). As far as we know, calcium changes in T. platelets were only evidenced by means of the fluorescent probe quin-2, in response to thrombin and ADP (3). Since quin-2 and aequorin are supposed to investigate different aspects of calcium changes in platelets (9), our data strengthen the fact that these changes may take place almost normally in platelets virtually devoid of GP Ilb-llla. Furthermore we extend this finding to a wide range of agonists, including PMA, and we also studied the mobilization of calcium from the internal stores induced by thrombin (that is luminescence changes in the presence of EGTA, threreby suppressing the influx) and found it to be normal. Taken together our own data and those of Powling and Hardisty (3) indicate that the presence of the GP Ilb-llla complex is a prerequisite neither for calcium influx nor for its mobilization. However part of the calcium movements may also depend on aggregation as evidenced here with ADP. The increase in calcium in response to ADP occurs with most of the studied normal platelets in two distinct phases. This has yet been reported neither with aequorin nor with fluorescent dyes. The first peak of luminescence in response to ADP is still detectable in T. platelets stimulated with ADP. This peak appears to be correlated to the persistent shape change and both events could be suppressed with the prostacyclin analogue iloprost. The second delayed peak depends on aggregation and thromboxane synthesis: in the presence of aspirin under aggregating conditions, or when aggregation cannot occur (no exogenous fibrinogen, no stirring), it is no longer detected. These findings fit in well with the known dependence of ADP-induced thromboxane synthesis on the Occurence of aggregation (14). Thus our data are consistent with an involvement of GP Ilb-llla in platelet activation through thromboxane synthesis (15). The bridging of adjacent platelets, and not only the fibrinogen binding to the glycoproteic complex, seems to be required for such a reinforcement of activation: in the absence of stirring ADP-stimulated platelets are known to bind fibrinogen without optical aggregation. The source of the calcium in this second phase of the increase in its intracellular concentration cannot be determined from our experiments since addition of EGTA prevents aggregation itself. The positive feedback brought by aggregation has been ascribed by Holmsen to close cell contact but still waits for further biochemical explanations (16. 17). Some studies have indicated that ligand binding to GP ilb-llla and aggregation may signal secondary biochemical events in platelet function: activation of calcium-dependent proteoiysis (18);
Vol. 58, No. 6
569
Ca++ MOVEMENTS & PLATELET AGGREGATION
induction of Na+ / H+ exchange (19); association with the cytoskeleton (20). In conclusion ADP-induced activation is reinforced by the occurence of close contacts between adjacent aggregated platelets. This bridging involves the fibrinogen binding to the GP Ilb-llla complex. Since some monoclonal antibodies against GP Ilb-llla have been found to activate platelets (6, 21, 22), the glycoproteic compleic may be directly involved in secondary platelet activation. ACKNOWLEDGEMENTS. We are greatly indebted to Mrs Guilvard and Faudeau from Coultronics France, who allowed us to work with the PICA-aggregometer and gave us the aequorin prepared by Dr Blinks. REFlTRIiNCES
1. Siess W. Molecular mechanisms of platelet activation. wws 1989
69: 58-143,
2. Philipps DR., Charo I.F., Parise L.V. and Fitzgerald L.A. The platelet membrane glycoprotein
Ilb-llla
complex. Blood 71; 831-843,
1988
3. Powling M.J. and Hardisty R.M. Glycoprotein 66: 731-734, 1985 stimulated platelets. mod
Ilb-llla
complex
and calcium
influx
into
4. Yamaguchi A., Yamamoto N., Kitagawa H., Tanoue K. and Yamazaki H. Calcium influx mediated through the GP Ilb/llla complex during platelet activation. FFRS I Ptters 375: 228-232, 1987
5. Nurden A.T. and Caen J.P. An abnormal glycoprotein pattern in three cases of Glanzmann’ s thrombasthenia. B-J.: 253-260, 1974 6. Morel M.C., Lecompte T., Champeix P., Favier R., Potevin F., Samama M., Salmon C. and Kaplan C. PL2-49: a monoclonal antibody against glycoprotein Ilb which activates platelets. BL J. H-1. 71: 57-63, 1989 7. Morel M.C., Favier IX, Champeix P., Lecompte T., Potevin F., Samama M. and Kaplan C. Characterization of an anti- Ilb-llla monoclonal antibody which is a platelet activator. Current od Transfusion 53-63, 1988 8. Yamaguchi A., Suzuki H., Tanoue K. and Yamazaki H. Simple method of aequorin loading into platelets using dimethylsulfoxide. -es. 44: 165-174, 1986 9. Johnson P.C., Ware J.A., Cliveden P.B.C., Smith M., Dvorak A.M. and Salzman E.W. Measurement of ionized calcium in blood platelets with the photoprotein aequorin: comparison with quin 2. J. Riol. Chem.: 2069-2076, 1985 10. Mustard J.F., Perry D.W., Ardlie N.G. and Packham M. Preparations of washed platelets from humans. Dr. J. HaematpL 37: 193-204, 1972 11. Caen J.P., Castaldi P.A., Leclerc J.C., lnceman S., Larrieu M.J., Probst M. and Bernard J. Congenital bleeding disorders with long bleeding time and normal platelet count. I. Glanzmann’s thrombasthenia (report of 15 patients). &I. J. Med. 41: 4-26, 1966 12. Hardisty R.M., Dormandy K.M. and Hutton R.A. Thrombasthenia: Studies on 3 cases U -toI. IQ: 371-387, 1964
570
Ca++ MOVEMENTS & PLATELET AGGREGATION
13. Zucker M.B., Pert J.H. and Hilgartner thrombasthenia. Blood 28: 524 -534, 1966
M.W.
Platelet
function
Vol. 58, No. 6
in a patient
with
14. Malmsten C., Kindahl H., Samuelsson B., Levy-Toledano S., Tobelem G. and Caen J.P. Thromboxane synthesis and the platelet release reaction in Bernard-Soulier syndrome, thrombasthenia Glanzmann and Hermansky-Pudlak syndrome. Br. J. m: 51 I-520, 1977 15. Levy-Toledano S., Maclouf J., Bryon P., Savariau E., Hardisty R.M. and Caen J.P. Human platelet activation in the absence of aggregation: a calcium-dependent phenomenon independent of thromboxane formation. Rlood 59: 1078-1085, 1982 16. Holmsen H. Prostaglandin endoperoxide-thromboxane synthesis and dense granule secretion as positive feedback loops in the propagation of platelet responses during the basic platelet reaction. -0s. s: 1030-1040, 1977 17. Koike K. and Holmsen H. Requirement for thrombin receptor occupancy during platelet secretion under aggregating and non-aggregating conditions. Thrombos.: 1053-I 059, 1987 18. Fox J.E.B., Reynolds C.F. and Phillips D.R. Calcium-dependent platelet aggregation. J. Biol. C-8: 99739981, 1983
proteolysis occurs during
19. Banga H.S., Simons E.R., Brass L.F. and Rittenhouse S.E. Activation of phospholipase A and C in human platelets exposed to epinephrine. Role of glycoprotein Ilb/llla and dual role of epinephrine. Proc. Natl. Sci. USA Q: 9197-9201, 1986 20. Phillips D.R., Jennings L.J. and Edwards H.H. Identification of membrane mediating the interaction of human platelets. J. Cell. R-86: 77-86, 1980
proteins
21. Bai Y., Durbin H. and Hogg N. Monoclonal antibodies specific for platelet glycoproteins react with human monocytes. Rlood 64: 139-146, 1984 22. Modderman P.W., Huisman H.G., van Mourik J.A. and von dem Borne A.E.G.Kr. A monoclonal antibody to the human glycoprotein Ilb/llla complex induces platelet activation. Thrombos. . 601 68-74. 1988
AEQUORIN - DETECTED CALCIUM CHANGES IN STIMULATED THROMBASTHENIC PLATELETS. AGGREGATION - DEPENDENT CALCIUM MOVEMENT IN RESPONSE TO ADP.
T. Lecompte ?? , F. Potevin ?? , P. Champeix * , M.-C. Morel ?? *, R. Favier ?? , M.-F. Hurtaud ?? **, N. Schlegel ?? **, M. Samama , C. Kaplan ** Laboratoire d’Hematologie - Hotel-Diu : 1, place du Parvis Notre-Dame 75004 * INTS *** Hopital R. Deb& Paris-France ??
??
??
??
(Received 29.7.1988; accepted in revised form 6.3.1990 by Editor M.B. Donati)
Calcium changes in normal and thrombasthenic platelets were recorded using the PICA-apparatus. Aequorin was loaded in the presence of DMSO, EGTA and PGEl. Platelets of three patients with type I thrombasthenia stimulated with A-23,187, thrombin, PMA in the presence 1mM Ca++ and 1mM Mg++ were able to normally raise their calcium concentrations. The maximal values could be found below the normal range with collagen, ADP and PAF-acether. Calcium mobilization from internal stores in response to thrombin was normal. There were two calcium peaks in normal platelets stimulated with ADP. The second one was suppressed by omitting fibrinogen, stirring, or by adding aspirin, and was absent in thrombasthenic platelets. Thus the GP Ilb-llla complex is not a prerequisite for calcium fluxes but is involved, when weak agonists such ADP are used, through an aggregation-dependent reinforcement of platelet activation.
Calcium plays a central role in platelet activation (1). Inositol-trisphosphate is a likely second messenger able to induce the release of calcium from the internal stores, often referred to as mobilization. An influx from the surrounding plasma across the plasma membrane also contributes to the increase in intracellular calcium. The molecular mechanisms of this influx remain to be fully elucidated. It has been suggested that the GP Ilb-llla complex might be involved (2), and so far two competing techniques have been used to adress this issue with intact platelets. One study using the fluorescent probe quin-2 led its authors to conclude that the complex is not required for the rapid influx that results from stimulation, but is -________---______-_--Key-words: thrombasthenia, calcium, aequorin, ADP 561
Ca++ MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
closely adjacent to the putative calcium channel (3). The more recent one used the photophore protein aequorin and brought some lines of evidence in favour of the role of the GP Ilb-llla complex in calcium movements, but there was no experiment with thrombasthenic (T.) platelets (4). With the aim of clarifying this issue, we compared calcium changes in normal and T. platelets, which are profoundly deficient in GP lib-llla (5).
Three patients (referred here to as A.A., F.C. and F.P.) with a lifelong bleeding disorder and a positive familial history were included in the present study. Each of them fulfilled the criteria for type I Glanzmann’s thrombasthenia: prolonged bleeding times with platelet counts above 120 Giga / L, normal routine coagulation parameters and von Willebrand factor plasma levels, lack of platelet aggregation on repeated testing regardless the agonist used, absent clot retraction. Glycoproteins Ilb and Illa, and fibrinogen could not be detected on sodium dodecyl sulfate - polyacrylamide gel electrophoresis (under reducing and non-reducing conditions, PAS and Coomassie blue staining). In addition binding sites for allo-antibodies against GP Ilb-llla and the PlAl antigen, as assessed by an indirect radiolabelled antiglobulin test, and binding sites for two monoclonal antibodies (6,7), PL2-49 (epitope on GP Ilb but complex-dependent binding) and PL2-73 (GP Ilb-llla complex), as assessed by a direct radioligand test, were dramatically decreased. Patients F.C. and F.P. belong to the same kindred, and F.P. was shown to be strongly immunized against the GP complex after multiple transfusions required for surgery (traumatic splenic rupture). This subject (F.P.) could be studied on three different times, his sister (F.C.) and the unrelated subject A.A.. twice. Controls were run throughout the same study period with blood collected from healthy volunteers. Platelets were loaded with aequorin according to Yamaguchi et al. (8) with some modifications. Briefly a platelet button was prepared from acidified titrated platelet rich plasma in the presence of unfractionated heparin, washed with a saline-Hepes buffer in the presence of glucose (5.6 mM), PGEl (0.1 PM), EGTA (5mM) and albumin (O.l%), pH adjusted to 7.3, incubated with aequorin in the same buffer and increasing amounts of dimethylsulfoxide (DMSO) in a stepwise manner up to 8% vol:vol final proportions, washed twice at 10,000 x g during 90 sec., and resuspended in a Tyrode’s buffer with albumin (0.35%), apyrase (0.03 U/mL), 1 mM calcium and 1 mM magnesium. The overall procedure was performed at room temperature, and the LDH release was less than 5 %. Aliquots (1 ml) of the platelet suspension (adjusted to 3 to 4 x 100,000 / uL) were stimulated under stirring at 1,100 rpm in an apparatus (PICA-Aggregometer from Chronolog Corporation - Coultronics France, Margency) able to record changes in light transmission and light emitted by the intracellular aequorin, in parallel. All experiments were performed in the presence of added fibrinogen (1.2 mg/mL), except when thrombin was the agonist or for particular experiments with ADP. Some experiments aimed at investigating calcium mobilization from the internal stores triggered by A-23,187 and thrombin were performed in the presence of EGTA (3 mM). Platelet testing was completed within 60 to 90 minutes after aequorin loading. Calcium levels were calculated assuming an intracellular magnesium concentration of 1 mM, according to Jonhson et al (9). There was no obvious difference between the loading efficiencies with normal and T. platelets. Sensitivities from 0.5 to 0.05 for detection of the luminescent signal generated by aequorin could be used depending on the strength of the used agonist, and allowed a good signal to background ratio (see curves in the figures). The following platelet activators were used at supramaximal concentrations: thrombin from Leo - 1 U/mL; collagen from Horm - 20 ug/mL; A-23.187 from Boehringer-Mannheim PAF-acether (a kind gift from Dr Benveniste) - 1.8 t&l; ADP - 100 f&t ; 1uM; Phorbol-Myristate Acetate (PMA) - 0.1 WM. The two latter agonists were obtained from Sigma. The endoperoxides-thromboxane analogue U-46,61 9 was obtained from Upjohn. The other reagents were obtained as follows: human fibrinogen, grade L: Kabi-Vitrum; bovine serum albumin, apyrase, EGTA and Triton X 100 from Sigma; aequorin was prepared by Dr Blinks and
Vol. 58, No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
563
was a gift from Coultronics-France; PGEl was from Upjohn ; iloprost (16 methyl, 18-19 didehydro-carbacyclin) was a kind gift from Dr Souvignet, Schering (Lys-lez-Lannoy, France); DMSO was obtained from Bruneau (Boulogne-Billancourt, France).
Typical simultaneous recordings of changes in light transmission and in aequorin-generated luminescence with stimulated normal platelets are showed in HQ. 1 The patterns of light transmission changes fit in with the well-known responses of washed platelets. In particular ADP-induced aggregation was markedly enhanced by the addition of exogenous fibrinogen. Aggregation was not induced by stirring alone, in the presence of fibrinogen.
FIG.1 Typical tracings obtained with normal, aequorin-loaded platelets. Aggregation tracings are shown in the upper part of the figure, and luminescence tracings recorded in parallel in its lower part. Platelets were stimulated with the following agonists: thrombin (T), calcium ionophore A-23,187 (I), collagen (C), PAF-acether (PAF) and phorbol-myristate acetate (PMA). The vertical bar indicates a 10% change in light transmission, the horizontal one one minute. Note that there is no linear relationship between the height of the luminescence peak and the maximal level in intra-platelet calcium, and that there is a time-dependent decrease in the intra-platelet active aequorin.
Thrombin and the calcium ionophore A-23,187 induced immediate and sharp luminescence signals corresponding to fast and maximal aggregation . With collagen there was a similar lag-time for the onset of the decrease in light transmission corresponding to the shape change, and the rise in calcium levels . ADP, PAF-acether and PMA induced calcium changes of lower amplitude. The luminescent signals were almost instantaneous but bell shaped (F~Q.I ), In the table I are the values of the peaks in calcium obtained after stimulation with the different agonists of control and T. platelets. Estimated maximal calcium changes in stimulated T. platelets fall into the normal ranges as far as calcium ionophore, thrombin and PMA are concerned. Signals elicited by collagen, ADP and PAF-acether tend to be lower than with normal platelets.
564
Ca+* MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
TABLE I Aequorin-Detected Calcium Peaka ~__~~_____~~~__~~~__.~~~~~__~~~~___~~~____~~_._~~~___~~~____~ Patient A-23,1 87 thrombin collagen ADP PAF _~___~~~___~~~__~~._~~~~~~__~~~~__~~~~__~~~~~~~~~~____~___~_~ 1-A 12.6 7.9 5.0 1-B 8.9 2.6 2-A 17.8 8.0 2.2 : 2-B 11.2 8.9 2.2 3-A 7.2 3.0 1.2 1 .o 3-B 9.5 5.6 1.4 _ 3-c 14.8 7.0 2.2 ___~__-____--_~_-_~~~~~~-~~~~~-~~~~-~~~~~~~~~~~~~~~~~~~~~~~~Normal 7.9 4.0 5.0 2.0 3.1 range * 31 .O 20.0 11.2 4.0 8.9
PMA
3.9
2.8
2.5 4.5
Maximal levels of intra-platelet calcium are expressed in f&l. Patients with Glanzmann’s thrombasthenia are 1: A.A. and 2: CF., studied on two separate occasions: patient 3: P.F., three separate experiments. ’ Control subjects: at least three. The median values obtained with control platelets were respectively: thrombin: 7.9 (n=18) , ADP: 2.1 (n-5), PAF: 5.0 (n=6). The results with thrombin have not a normal distribution, but most of them fall in a narrow range between 6.0 and 10.0 f&t.
TABLE II Calcium Mobilization Thrombin A-23,1 87 + Ca++ + EGTA + Ca++ + EGTA ___~______________~_._~~~__~~~~_____~_~._~~__~~~~~~~~~~_~~.~. P.F, 7.0 1.4 C.F. 8.9 1.8 ~_~~~~~__~~~~~~~~~~~~~~~-
14.8 6.3 11.a 4.5 _~~_~~___~~___._--_~~~~~~.~~~~~~~~~~
Normal values
4.1 1 .4 10.0 3.8 18.6 5.6 11.2 5.0 7.0 2.0 11.2 4.4 ____________~____~~_~~~~-~~.~~~__~~~~~~~~~~~~~.~~~----------~
Maximal calcium levels are expressed in uM. The addition to the external milieu of EGTA (3.mM) precludes any influx and therefore the remaining signals are the result of calcium mobilization from the internal stores. The experiments the results of which are given here are referred to as 3-C and 2-A in the table I. Normal values are: (i) for thrombin the range and the mean of the results of experiments with platelets of 5 different donors; (ii) for A-23,187 the individual results obtained with three different platelet suspensions. Values obtained with A-23,187 in the presence of EGTA indicate (i) the internal releasable calcium; (ii) the ability of the aequorin loaded in the platelets to detect such a calcium mobilization.
Vol. 58. No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
565
Further experiments were conducted in the presence of EGTA in order to investigate calcium mobilization from the internal stores (table II). The calcium ionophore A-23,187 and thrombin were still able to induce a change in the luminescent signal: the peak values were reduced to 30-40 % and 20-35 % as compared to those obtained in the presence of 1 mM extracellular calcium, respectfvely. Thrombin-induced calcium mobilization in normal and T. platelets was not different. Typical tracings obtained with T. platelets stimulated in the presence of extra-cellular calcium are shown in Figs. 2 and 3. Aggregation of T. platelets was markedly abnormal as indicated by the absence of increase or the very slight increase in light transmission. However the same patterns of luminescence were obtained with T. and control platelets with all agonists but ADP. Control platelets stimulated with ADP showed indeed a peculiar pattern of aequorin-detected calcium changes ( Fig.4 ). With most of the preparations there were two distinct signals, whereas with some others there was a sustained luminesoence following the initial change, with a slow return to the baseline level. With platelets from patients with Glanzmann’s thrombasthenia the calcium signal was always reduced to its first component. One typical example is given in Fjg. 5 . The remaining first peak could be abolished with the addition of the stable prostacyclin analogue iloprost (10 nM), and the shape change too, as shown in Fig. 5 . The responses of control platelets to ADP were further investigated ( Fig. 4 ). In the absence of exogenous fibrinogen, shape change was still detectable but there was only a very small decrease in light transmission thereafter (trace c, left-hand panel). The first peak of luminescence remained unaffected, whereas the second one disappeared. As far as aequorin-generated luminescence is concerned, the same holds true in the presence of aspirin (traces b left- and right- hand panels). Under these experimental conditions however, aggregation still occured, but at a reduced rate and to a lesser maximal extent, and was reversible. In the absence of stirring, aggregation, but not the first calcium peak, was suppressed (not shown). When aspirin-treated platelets were stimulated first with ADP, and thereafter with the endoperoxides-thromboxane mimetic U-46,61 9, aggregation was complete and irreversible and two calcium peaks could be elicited ( Fig. 6).
FIG. 2 Tracings obtained from thrombasthenic agonists as in the previous figures.
platelets (patient
1). Same legends for the
566
Ca++ MOVEMENTS & PLATELET AGGREGATION
T
I
I
Vol. 58, No. 6
PMA
-
I ,
FIG. 3 Tracings obtained with thrombasthenic
platelets (referred to as 3-B in the table I)
b
a
FIG. 4 Experiments performed with two preparations of control platelets, stimulated with ADP. In the presence of fibrinogen ( tracings a): first-peak maximal values of calcium are 2.2 and 4.0 uM respectively. On the left hand, there is a sustained increase in luminescence after the initial discharge. On the right hand, there are at least two distinct peaks. In the presence of fibrinogen and aspirin, 1.4 mM (tracings b): calcium peak values are 2.0 and 4.0 pM respectively (one peak only). In the absence of exogenous fibrinogen (tracing c, left-hand panel only): one single peak at 2.0 PM.
Vol. 58, No. 6
Ca++ MOVEMENTS & PLATELET AGGREGATION
567
ADP
I
a
cs---I
b -+---, I
FIG. 5 ADP-induced changes in light transmission and luminescence with platelets from a thrombasthenic patient. The stimulation with ADP was performed in the absence (tracings a) or in the presence (tracings b) of the stable prostacyclin analogue iloprost (10 nM). The shape change (upper tracing) is clearly noticeable in a and suppressed in b, and the change in luminescence (lower tracing), indicating intraplatelet calcium levels, as well.
FIG.6 ADP- induced changes in light transmission and luminescence in a typical experiment performed with control platelets, in the presence of exogenous fibrinogen. On the right hand platelets were preincubated with aspirin before the addition of ADP: there is a decrease in maximal aggregation, and the calcium rise is reduced to its first component (peak value: 1.4 PM), very similar to the one induced by ADP in the absence of aspirin (left hand). The subsequent addition of the thromboxane mimetic U-46,619 (0.1 PM) triggers another calcium change (2.2 PM) accompanied by a reinforcement of aggregation. U-46,619 alone at the same concentration induces only a shape change (not shown).
568
Ca++ MOVEMENTS & PLATELET AGGREGATION
Vol. 58, No. 6
DISCUSSION. As far as we know we report here the most extensive study with aequorin-loaded platelets according to Yamaguchi’s method. As shown in the figures all the agonists used induced clear-cut increases in luminescence, indicating rises in calcium: thrombin, collagen, A-23,187, ADP, PAF-a&her, and also arachidonic acid and the thromboxane mimetic structurally related to endoperoxides, U-46,619 (data not shown for these two latter agonists). Clearly higher concentrations of ADP were required to obtain a maximal aggregation with changes of light transmission of 30 % or more, as compared to platelets washed with our currently used standard protocol derived from that of Mustard et al (10). But initial rates of aggregation were high and the addition of fibrinogen markedly potentiated aggregation. Furthermore in more recent experiments with platelets prepared from normal volunteers we were able to detect calcium changes in response to 5 f&l ADP. It must be pointed out however that platelets prepared from a few donors became refractory to ADP: low amplitude of maximal aggregation together with slight increase in intracellular calcium and no second component (see below). We then studied platelets collected from patients affected with type I Glanzmann’s thrombasthenia, which were shown to be profoundly deficient in GP Ilb-llla. When strong activators were used, the peaks were sharp, the kinetics and maximal achieved levels were simllar to those obtained with control platelets. This is in agreement with our current knowledge since stimulated T. platelets are known to change their shape and to be able to release the content of their granules and synthetire thromboxane, indicating that the activation processes can occur (11, 12, 13, 14). As far as we know, calcium changes in T. platelets were only evidenced by means of the fluorescent probe quin-2, in response to thrombin and ADP (3). Since quin-2 and aequorin are supposed to investigate different aspects of calcium changes in platelets (9), our data strengthen the fact that these changes may take place almost normally in platelets virtually devoid of GP Ilb-llla. Furthermore we extend this finding to a wide range of agonists, including PMA, and we also studied the mobilization of calcium from the internal stores induced by thrombin (that is luminescence changes in the presence of EGTA, threreby suppressing the influx) and found it to be normal. Taken together our own data and those of Powling and Hardisty (3) indicate that the presence of the GP Ilb-llla complex is a prerequisite neither for calcium influx nor for its mobilization. However part of the calcium movements may also depend on aggregation as evidenced here with ADP. The increase in calcium in response to ADP occurs with most of the studied normal platelets in two distinct phases. This has yet been reported neither with aequorin nor with fluorescent dyes. The first peak of luminescence in response to ADP is still detectable in T. platelets stimulated with ADP. This peak appears to be correlated to the persistent shape change and both events could be suppressed with the prostacyclin analogue iloprost. The second delayed peak depends on aggregation and thromboxane synthesis: in the presence of aspirin under aggregating conditions, or when aggregation cannot occur (no exogenous fibrinogen, no stirring), it is no longer detected. These findings fit in well with the known dependence of ADP-induced thromboxane synthesis on the Occurence of aggregation (14). Thus our data are consistent with an involvement of GP Ilb-llla in platelet activation through thromboxane synthesis (15). The bridging of adjacent platelets, and not only the fibrinogen binding to the glycoproteic complex, seems to be required for such a reinforcement of activation: in the absence of stirring ADP-stimulated platelets are known to bind fibrinogen without optical aggregation. The source of the calcium in this second phase of the increase in its intracellular concentration cannot be determined from our experiments since addition of EGTA prevents aggregation itself. The positive feedback brought by aggregation has been ascribed by Holmsen to close cell contact but still waits for further biochemical explanations (16. 17). Some studies have indicated that ligand binding to GP ilb-llla and aggregation may signal secondary biochemical events in platelet function: activation of calcium-dependent proteoiysis (18);
Vol. 58, No. 6
569
Ca++ MOVEMENTS & PLATELET AGGREGATION
induction of Na+ / H+ exchange (19); association with the cytoskeleton (20). In conclusion ADP-induced activation is reinforced by the occurence of close contacts between adjacent aggregated platelets. This bridging involves the fibrinogen binding to the GP Ilb-llla complex. Since some monoclonal antibodies against GP Ilb-llla have been found to activate platelets (6, 21, 22), the glycoproteic compleic may be directly involved in secondary platelet activation. ACKNOWLEDGEMENTS. We are greatly indebted to Mrs Guilvard and Faudeau from Coultronics France, who allowed us to work with the PICA-aggregometer and gave us the aequorin prepared by Dr Blinks. REFlTRIiNCES
1. Siess W. Molecular mechanisms of platelet activation. wws 1989
69: 58-143,
2. Philipps DR., Charo I.F., Parise L.V. and Fitzgerald L.A. The platelet membrane glycoprotein
Ilb-llla
complex. Blood 71; 831-843,
1988
3. Powling M.J. and Hardisty R.M. Glycoprotein 66: 731-734, 1985 stimulated platelets. mod
Ilb-llla
complex
and calcium
influx
into
4. Yamaguchi A., Yamamoto N., Kitagawa H., Tanoue K. and Yamazaki H. Calcium influx mediated through the GP Ilb/llla complex during platelet activation. FFRS I Ptters 375: 228-232, 1987
5. Nurden A.T. and Caen J.P. An abnormal glycoprotein pattern in three cases of Glanzmann’ s thrombasthenia. B-J.: 253-260, 1974 6. Morel M.C., Lecompte T., Champeix P., Favier R., Potevin F., Samama M., Salmon C. and Kaplan C. PL2-49: a monoclonal antibody against glycoprotein Ilb which activates platelets. BL J. H-1. 71: 57-63, 1989 7. Morel M.C., Favier IX, Champeix P., Lecompte T., Potevin F., Samama M. and Kaplan C. Characterization of an anti- Ilb-llla monoclonal antibody which is a platelet activator. Current od Transfusion 53-63, 1988 8. Yamaguchi A., Suzuki H., Tanoue K. and Yamazaki H. Simple method of aequorin loading into platelets using dimethylsulfoxide. -es. 44: 165-174, 1986 9. Johnson P.C., Ware J.A., Cliveden P.B.C., Smith M., Dvorak A.M. and Salzman E.W. Measurement of ionized calcium in blood platelets with the photoprotein aequorin: comparison with quin 2. J. Riol. Chem.: 2069-2076, 1985 10. Mustard J.F., Perry D.W., Ardlie N.G. and Packham M. Preparations of washed platelets from humans. Dr. J. HaematpL 37: 193-204, 1972 11. Caen J.P., Castaldi P.A., Leclerc J.C., lnceman S., Larrieu M.J., Probst M. and Bernard J. Congenital bleeding disorders with long bleeding time and normal platelet count. I. Glanzmann’s thrombasthenia (report of 15 patients). &I. J. Med. 41: 4-26, 1966 12. Hardisty R.M., Dormandy K.M. and Hutton R.A. Thrombasthenia: Studies on 3 cases U -toI. IQ: 371-387, 1964
570
Ca++ MOVEMENTS & PLATELET AGGREGATION
13. Zucker M.B., Pert J.H. and Hilgartner thrombasthenia. Blood 28: 524 -534, 1966
M.W.
Platelet
function
Vol. 58, No. 6
in a patient
with
14. Malmsten C., Kindahl H., Samuelsson B., Levy-Toledano S., Tobelem G. and Caen J.P. Thromboxane synthesis and the platelet release reaction in Bernard-Soulier syndrome, thrombasthenia Glanzmann and Hermansky-Pudlak syndrome. Br. J. m: 51 I-520, 1977 15. Levy-Toledano S., Maclouf J., Bryon P., Savariau E., Hardisty R.M. and Caen J.P. Human platelet activation in the absence of aggregation: a calcium-dependent phenomenon independent of thromboxane formation. Rlood 59: 1078-1085, 1982 16. Holmsen H. Prostaglandin endoperoxide-thromboxane synthesis and dense granule secretion as positive feedback loops in the propagation of platelet responses during the basic platelet reaction. -0s. s: 1030-1040, 1977 17. Koike K. and Holmsen H. Requirement for thrombin receptor occupancy during platelet secretion under aggregating and non-aggregating conditions. Thrombos.: 1053-I 059, 1987 18. Fox J.E.B., Reynolds C.F. and Phillips D.R. Calcium-dependent platelet aggregation. J. Biol. C-8: 99739981, 1983
proteolysis occurs during
19. Banga H.S., Simons E.R., Brass L.F. and Rittenhouse S.E. Activation of phospholipase A and C in human platelets exposed to epinephrine. Role of glycoprotein Ilb/llla and dual role of epinephrine. Proc. Natl. Sci. USA Q: 9197-9201, 1986 20. Phillips D.R., Jennings L.J. and Edwards H.H. Identification of membrane mediating the interaction of human platelets. J. Cell. R-86: 77-86, 1980
proteins
21. Bai Y., Durbin H. and Hogg N. Monoclonal antibodies specific for platelet glycoproteins react with human monocytes. Rlood 64: 139-146, 1984 22. Modderman P.W., Huisman H.G., van Mourik J.A. and von dem Borne A.E.G.Kr. A monoclonal antibody to the human glycoprotein Ilb/llla complex induces platelet activation. Thrombos. . 601 68-74. 1988