of Platelet-Platelet and Inhibitors
FRANCOIS M. BOOYSE, THOMAS P. HOVEKE, DONNA KISIELESKI AND MAX E. RAFELSON, JR. Department of Biochemistry, Rush Medical College, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, Illinois 60612 Received December 10, 1971 With the technique of shadow-casting of whole platelets, it has been possible to observe both the very early events produced on the platelet surface by aggregating agents and what appears to be the entire process of platelet aggregation. It has been shown that a variety of aggregating agents and conditions such as ADP, thrombin, collagen, polylysine, glass and cold induce a common sequence of reactions leading to aggregation. The data presented here show that the platelet membrane does not directly participate in the early events preceding and leading to platelet aggregation. It is extruded platelet cytoplasm (cytogel) that appears to interact to form interplatelet attachment. Subsequent platelet membrane interactions appear to follow the contraction of the interacted cytogels which brings the platelets into close contact. The action of several inhibitors of aggregation have been localized to specific steps in the sequence of events observed.
INTRODUCTION A wide variety of agents such as ADP (Hellem, 1960), thrombin (Grette, 1962), trypsin (Grette, 1962), catecholamines (Mitchell and Sharp, 1964), cold (Kattlove and Alexander, 1971),polylysine (Schneider et al., 1968),collagen (Hovig, 1963),etc., have the common ability to induce platelet aggregation. A considerable amount of data has been obtained on the effects of these agents on platelet morphology and the releaseof cellular constituents (Hovig, 1968; Mustard and Packham, 1970). However, very little is known about the initial attachment and action of these diverse agents and the subsequent early eventspreceding and leading to aggregation. Our recent studies have shown that the process whereby platelets become activated to aggregate and adhere is the results of very rapid molecular events taking place on or near the platelet surface (membrane and coat) within secondsafter the addition of the aggregating agent (Booyse and Rafelson, 1971; Rafelson and Booyse, 1971). The published papers describing various aspects of platelet aggregation and the theories to explain this phenomenon are too numerous to catalog here. Hence, we shall restrict this discussion to the very early events that occur in the presenceof aggregating agents, and the presenceof macromolecules associatedwith the surface and the participation of thesemacromoleculesin the molecular mechanismsleading to platelet-platelet ’ Supported in part by grants from the National Institutes of Health (HE-07565), the Chicago and Illinois Heart Associations, the Illinois Federation of Women’s Clubs, the Clow Foundation and General Research Support Grant RRO-5477. 0 1972 by Academic Press, Inc. 179
BOOYSE ET AL.
interactions. The contractile protein, thrombosthenin, has been isolated from platelet membranes (Nachman et al., 1967) and it has been suggestedthat the “ecto-ATPase” of platelets has properties similar, if not identical to thrombosthenin (Chambers et al., 1967).We have shown that about 5-10 % of the total platelet thrombosthenin is surface or membrane associated (S-thrombosthenin) (Booyse and Rafelson, 1971). B. M. Jones (1966) and P. C. T. Jones (1966) suggestedthat a membrane localized contractile protein is responsible for cellular adhesion in general. Allosteric transformations of the contractile protein, induced by various aggregating agents can produce changes of the cell surface leading to cellular interaction. We have demonstrated that very early changesdo occur on the platelet surface preceding aggregation. The surface extended structures that are produced are composed, at least in part, of immunohistochemically identifiable thrombosthenin (Booyse and Rafelson, 1971).The purpose of this communication is to describe in more detail, the early eventspreceding plateletplatelet interaction, and the effects of various inhibitors on this sequenceof events. Shadow-casting of whole platelets was usedto study the early eventspreceding plateletplatelet interaction. MATERIALS
Human thrombin was obtained from Ortho Diagnostics and dissolved in 0.15 M NaCl to give a stock solution of 50 U/ml. Polylysine was obtained from Miles Laboratories, Inc., had a molecular weight of 37,200 daltons and was dissolved in deionized water. The collagen preparation used was a generous gift from Dr. J. Caen of Paris, France. ADP was purchased from Sigma and kept as a frozen stock solution (10 mM) in deionized water neutralized to pH 6.8. Analytical grade reagents were used in all other procedures. Preparation of Platelets Platelets were prepared as platelet-rich plasma (PRP) from freshly drawn titrated human blood (ACD, formula A-NIH) by centrifugation at 33OOg,,,for 2 min in a swinging bucket rotor at 37”. All procedures were carried out in cleaned, unused polypropylene plastic tubes at 37” unless stated otherwise. The freshly prepared PRP was kept at 37” and used within 15-30 min after the initial drawing of the whole blood. Only platelet preparations that showed normal responsesto ADP, thrombin and collagen in the aggregometer were used in these experiments. Changes in optical density (OD) at 600 nm (Born, 1962, 1969) were routinely used to obtain the time, in seconds, between the addition of the aggregating agent and the subsequent increase and rapid decreasein OD that followed. Activation of Platelets with Aggregating Agents Theseexperiments were performed by rapidly injecting the required amount of aggregating agent (contained in 2 ml neutralized warm saline) into 2 ml warm, freshly prepared PRP that was continuously being stirred at 1000 rpm. After the appropriate activation time had elapsed, 4 vol warm 4% gluteraldehyde (containing 0.1 M cacodylate buffer, pH 6.7) were very rapidly injected into the PRP with a secondhigh-speed syringe injection system. The activation and aggregation processes appear to be
stopped immediately upon the addition of gluteraldehyde. Control, unactivated platelets (PRP) were fixed immediately after the addition of 2 ml warm saline. Each aggregating agent was added to the warm PRP in an equal volume of saline (2 ml) to facilitate rapid and complete mixing and also uniform activation of the PRP. Sufficient platelets were present in l-2 ml normal PRP (300,000-800,000platelets/mm3) to complete one shadow-casting experiment. ADP Activation ADP was added to warm PRP and fixative added at various times between 1-6 set after the addition of ADP. In order to observe very rapid changes,the gluteraldehyde and ADP were injected simultaneously. Control platelets were fixed immediately without the prior addition of ADP. The sequenceof events induced by ADP was studied at 1 x 10m7and 1 x 1Om4M ADP, final concentration. Thrombin Activation Thrombin (final concentration of 0.5 U/ml) was added to warm PRP and fixative added at various times between 1-12 set after the addition of thrombin. Collagen Activation Lyophilized collagen was dissolved in 0.1 Macetic acid (pH 3.5) and then exhaustively dialyzed against neutralized saline (pH 7.5). The dialyzed collagen was prepolymerized by heating at 33” for 2 min and added directly to warm PRP. Fixative was added at various times between l-90 set after the addition of the collagen. The final concentration of collagen was 40 pg/ml. Polylysine Activation Polylysine (final concentration of I x low6 M) was added to warm PRP and fixative added at various times between l-10 set after the addition of the activating agent. Cold Activation Freshly prepared PRP was placed in an ice-bath (4”) for 15-30 min and then fixed by the addition of 4 vol cold 4 % gluteraldehyde. Fixation proceeded for 30 min at room temperature. Glass Activation A few drops of warm PRP were placed on a glass slide for 1 min. The platelets were fixed on the slide with 4 % gluteraldehyde for 15 min and then rinsed with saline, deionized water and absolute alcohol. The glass slide was air-dried and shadowed as described below. Activation of Platelets in the Presenceof Various Inhibitors The effect of the following inhibitors on the activation sequence was studied: adenosine, 5 x 10e4M; iodoacetate, 4 x lo-) M; PGE,, 1 x lo+ M; N-ethyl maleimide (NEM), 1 x 10M3M; colchicine, 2 x 10m3M; EDTA, 2 x lo-) M; ATP, 5 x 1O-4 M; salyrgan, 2.5 x lop3 M; ADP, 1 x 10e2 M; p-hydroxymercuribenzoate (PHMB), 1 x lo-’ M and ouabain, 1 x 1O-3M. I*
BOOYSE ET AL.
Platelets (as PRP) were first incubated for 2 min with each of the inhibitors, at the final concentrations shown above, to determine whether the inhibitors by themselves caused any discernable changes in the platelets. A second seriesof experiments were then performed to determine the effectsof these inhibitors on aggregation induced by 10m4M ADP. PRP was preincubated for 2 min with eachof the inhibitors, ADP was then added and the platelets fixed and prepared for shadow-casting at various times after the addition of ADP. Specific Antibody Staining Platelets (as PRP) were activated with lo-’ M ADP for 3-4 set and stained with monospecific anti-human thrombosthenin serum (rabbit) using an unlabeled antibodyenzyme staining technique as described previously (Booyse et al., 1971b; Booyse and Rafelson, 1971). Interplatelet bonds are present at this time period. Preparation of Samplesfor Shadow-Casting Activated platelets (PRP) were fixed for 30 min at room temperature. Fixed cells were centrifuged at 1lOOgfor 10 min in a swinging bucket rotor. Under these conditions of centrifugation the fixed platelets form a thin cell layer on the inside of the plastic centrifuge tube. The platelet layer was carefully rinsed, without resuspension, with 20 ml saline, 20 ml deionized water and 10 ml absolute alcohol. The washed cell layer was resuspendedin 0.5 ml absolute alcohol and spread onto a piece (1Hx 2”) of freshly cleaved p-dioxane coated mica. Mica slides were air-dried for 30 min before shadow-casting. Platelets (on mica) were shadow-castedwith Pt at an angle of 22” and a pressure of 1 x 10-j Torr and the specimensprepared essentially as described by Stewart (1970). Shadowedpreparations were examined on a RCA EMU-3F electron microscopewith an accelerating voltage of 50 kV. Isolation of Thrombosthenin Platelets were isolated from titrated whole human blood by differential centrifugation as described previously (Booyse et al., 1968).The platelet pellet was washed twice with saline and thrombosthenin isolated as described by Bettex-Galland and Luscher (1961). Three times reprecipitated thrombosthenin was used for the determination of ATPase activity and in the superprecipitation experiments. Determination of ATPase Activity ATPase activity of thrombosthenin was determined on 0.5-1.0 mg protein by the method of Marsh asmodified by Grette (1962). Protein was determined by a microbiuret method using bovine serum albumin as the standard (Itzhaki and Gill, 1964).The effect of inhibitors on ATPase activity was determined after preincubating the enzyme and inhibitor for 10 min at room temperature. Final concentrations of inhibitors were the same as described above. Superprecipitation The change in OD was measured at 624 nm in a final volume of 1 ml containing: 1 x 10e3 M ATP; 2.5 x 10m2M imidazole buffer, pH 7.0; 5 x 1O-3 M MgC12; 5 x 10m2A4 KC1 and 0.8-1.0 mg thrombosthenin. Inhibitor concentrations were as shown above. Reactions were measured at room temperature.
INDUCERS AND INHIBITORS
RESULTS Early Structural Etlents Preceding Aggregation With the technique of shadow-casting of whole platelets, we have been able to observeboth the very early eventsproduced on the platelet surfaceby aggregating agents and what appears to be the entire process of platelet-platelet interaction. That we are observing very early events preceding aggregation may be ascertained by reference to Fig. 1. Figure 1 shows a typical aggregation pattern obtained by measuring the change in OD at 600 nm after the addition of an aggregating agent to PRP. As may be seen,we are observing events well before any macroscopic aggregation (rapid decreasein OD) has taken place. I
20 set . 1
FIG. 1. Typical aggregation pattern obtained after the addition of aggregating agents to PRP. Arrows indicate times at which aggregating agents were added. The activation process was stopped at various times after the addition of the aggregating agents as described in text.
ADP-induced changes. Representative electron micrographs of shadow-casted (on mica) control and ADP-activated (lo-’ A4 and 10m4M ADP) platelets are shown in Figs. 2, 3 and 5. Unactivated control platelets that have been fixed immediately with gluteraldehyde, have a smooth, flat surface and what appears to be a well-defined edge (Fig. 2a). Approximately l-2 set after activation with ADP (lo-’ M) the edge appears to disappear or “relax” and the equatorial region of the platelet becomes amorphous or fluffy (Figs. 2b, c) ; the surfacehas the sameappearanceaspreviously observedwith both the horseradish peroxidase staining and the specific antibody staining (Rafelson and Booyse, 1971).At 3-4 seethe amorphous material on the surface of different platelets appears to interact and form well-defined interplatelet bonds ranging in length from 2oo(r-4000A (Figs. 2 and 3). At 4-6 set, there appearsto be a contraction of the surface
amorphous material including the interplatelet bonds bringing the platelets closer together, and in some cases,resulting in what appears to be fusion of the cell surfaces (Fig. 3). Up to this point (l-6 set) with 10m7M ADP there doesnot seemto be any significant change in the shape of the platelets. All the cells retain their discoid shape. How-
FIG. 2. Shadow-casted (on mica) platelets activated with lo-’ MADP. (a) control unactivated platelet x16,291 ; (b) 1-2 set after activation ~16,931; (c) 2-3 set after activation x16,291; (d) 4-6 set after activation x 16,291. Arrows indicate interacting material and interplatelet bonds.
ever, at longer time periods (about 7-8 set and longer) the shape of the cells becomes distorted and large aggregates start to form. An artist’s conception of the general sequenceof events obtained with 10e7M ADP is shown in Fig. 4. Although the sequenceof events obtained with 10e4M ADP are basically the same as those obtained with lo-’ M ADP, there are quantitative differences. The changes produced with 10m4M ADP are considerably more extensive. As shown in Fig. 5a, ex-
INDUCERS AND INHIBITORS
tensive globular projections are produced at the equatorial region or edge of the platelets within l-2 set after the addition of the ADP. This is followed at 2-3 set by the extrusion of an extensive amount of amorphous material that appears to be the thrombosthenin-rich, nonparticulate platelet cytogel (Fig. 5b). These cytogels also appear to interact and contract, drawing the platelets together to form large aggregatesas shown in Fig. 5d (3-6 set). At this concentration of ADP ( lop4 M) the initial extrusion reaction (2-4 set) was accompanied by an extensive shapechange as indicated by the increase in shadow height, shown in Figs. 5b-d. Thrombin-induced changes. The morphological changesand initial sequenceof events preceding platelet-platelet interactions produced by thrombin (0.25-l U/ml) are essen-
FIG. 3. Shadow-tasted platelets activated with lo-’ MADP. (a) 46 set after activation, arrows indicate interplatelet bridges x 16,291; (b) platelet aggregate formed after contraction of interplatelet bonds x16.291.
tially the sameas those produced by ADP. However, the major observable differences in the thrombin-induced sequenceof eventsare the morphological changesand the more extensive cytogel extrusion. It would appear that thrombin (0.25-l U/ml) is more effective than ADP (lo-‘-low4 M) in causing membrane “relaxation” (seeBooyse et al., 1972a),cytogel extrusion and subsequentplatelet-platelet interaction (Figs. 5 and 6). After a brief lag period (1-5 set), thrombin causesan extensive “bulging” (globular projections) at the equatorial region of the cell (Fig. 6b), followed by a rapid and extensive extrusion of the platelet cytogel from the sites of globular projections (5-7 set) (Figs. 6b-d). As soon as cytogel extrusion starts to occur, the platelet shape begins to change (Figs. 6c, d). Several subsequentchangescan be observed in the extruded cytogel : (1) rapid distortion due to mechanical stirring (this distortion does not take place if the cell suspensionis not stirred), (2) a direct interaction of the extruded cytogels (Fig. 6) which does not occur in the absenceof mechanical agitation or stirring, (3) fiber-like structures of varying length start to appear within the interacted cytogels (Figs. 7d-f).
BOOYSE ET AL.
The fiber-like structures range from about 200 to several hundred angstrom in diameter with lengths ranging up to 8-10 p. These structures do not show any discernable periodicity and do not resemble fibrin. These fiber-like structures appear to represent the initial stagesof contraction of the thrombosthenin-rich cytogel, (4) the interacted
FIG. 4. Artist’s impression of sequence of events in platelet aggregation obtained with lo-’ M ADP. (A) control unactivated platelet; (B) l-2 set activation; (C) 2-4 set activation indicating interaction of amorphous material and formation of interplatelet bonds; (D) 5-6 set activation indicating formation of platelet aggregate due to rapid contraction of interplatelet bonds. (Courtesy of Dr. H. H. Sky-Peck.)
cytogels contract, are pulled back or retracted by the platelets and draw the platelets into a tight mass or thrombus. Steps (l)-(4) take place within 7-12 sec. It should be noted that a variety of inhibitors of aggregation have no effect on the initial cytogel extrusion processbut do inhibit cytogel interactions and the contraction of these cytogels (Figs. 10 and 11). Inhibitors such as EDTA, salyrgan and colchicine can be used to “freeze” this sequenceof events at the stageof cytogel extrusion. Direct measurement of ADP in the platelet-poor plasma from PRP activated by thrombin in
INDUCERS AND INHIBITORS
the presence of either EDTA, salyrgan or colchicine (extruded cytogel state) demonstrated no detectable nucleotide at this stage of the reaction sequence.Other investigators have previously shown that these agents do, in fact, inhibit nucleotide releasein
FIG. 5. Shadow-casted platelets activated with 10m4M ADP. (a) l-2 set activation x 19,250; (b) 2-3 set activation ~17,010; (c) 3-5 set activation ~13,670; (d) 67 set activation x 12,480.
thrombin-activated platelets (White, 1969; Mustard and Packham, 1970). From these data, we have concluded that thrombin-induced activation was not mediated by ADP becauseADP could not be detected during the processleading to cytogel extrusion. Collagen-induced changes. Prepolymerized collagen added to platelet-rich plasma at
20-40 pg/ml final concentration, causes the same general sequence of events as described for ADP. The reaction sequencesstudied thus far take place over a period of about l-90 sec. However, the collagen-induced changes occur in two distinct phases.
FIG. 6. Shadow-tastedplateletsactivated with thrombin. (a) control unactivated platelet ~15,970; (b) 4-6 set activation ~16,830;(c) 5-7 set activation ~17,180;(d) 6-8 set activation ~11,080.
The first part of the reaction sequence(I-80 set) involves the direct interaction or attachment of collagen fibers to a very small percentageof the total platelet population. Only those platelets attached to collagen undergo the extrusion-contraction cycle at first. After the contraction has taken place in these cells, a factor is presumably released (ADP?) from these cells that induces the same sequence of events in the remainder of the cells as with ADP or thrombin induced aggregation. All the remaining platelets that
INDUCERS AND INHIBITORS
are not physically associated with collagen fibers are “activated” in this manner (see Fig. 8b). Polylysine-induced changes. Although polylysine-induced aggregation occurs over a
FIG. 7. Shadow-tasted thrombin-activated platelets. (a) interacted cytogels as shown by arrows ~9,260; (b, c) interacted cytogels as shown by arrows ~11,600; (d) early stages of cytogel contraction x 11,110; (e) interaction and contraction of cytogels between two platelets, platelets are visible at both ends of contracting material ~7,070; (f) high magnification of e ~22,300.
longer period of time than ADP, the sequence of events is virtually identical to that of ADP-induced changes. Fig. 8c shows a polylysine activated platelet in the extruded state, 68 set after the addition of the aggregating agent. Cold-induced changes. Platelets that are cooled and maintained at 5” rapidly extrude their cytogels. Extrusion is followed immediately by a change in shape. However, at low
temperature, contraction of extruded cytogels does not occur and the platelets will remain in this extruded state until the temperature is increased to approximately 15” or
FIG. 8. Shadow-casted platelets PRP. (a) platelets applied onto glass slide without prior fixation ~13,540; (b) collagen induced (presumably ADP) extrusion after 85 set incubation ~15,990; (c) polylysine induced extrusion x 13,540; (d) cold induced extrusion x 17,400.
above. Figure 8d shows a typical cold-activated platelet (5” for 15 min). No detectable ADP is releasedby platelets that are maintained in the extruded state. Nucleotides are only released after the contraction cycle is activated and confirms Kattlove and Alexander’s (1971) observations that cold-maintained platelets do not releaseADP. If
INDUCERS AND INHIBITORS OF AGGREGATION
these cold-activated platelets are now stirred and heated simultaneously, normal extruded cytogel interactions occur leading to aggregation as described by Kattlove and Alexander. However, if these cold-activated platelets are heated without stirring, contraction and retraction of individual cytogels takes place and no aggregation results. Such reheated platelets (30-45 min) can then be activated either with ADP or thrombin to follow the normal sequenceof reactions leading to aggregation. Glass-induced changes. Figure 8a shows platelets applied onto a glassslide as plateletrich plasma. It appears that the platelets are activated by the glass surface and that the activated platelets attach to the surfacevia their extruded cytogels. Plasmaproteins have the same tendency to stick to the glass surface and make it difficult to obtain a clean protein-free surface as in the case of mica. Nature of the interplatelet bond. Immunohistochemical staining of the distinct interplatelet bonds that are formed betweenplatelets activated with 1O-7M ADP for 3-4 set shows that these interplatelet bridges are composed, at least in part, of thrombosthenin (Figs. 3 and 12). Requirements for cytogel contraction. Studies with washed platelets, resuspendedin Tyrode’s solution that contains no divalent cations, indicate that platelets can be activated by ADP and thrombin to extrude their cytogels. However, they have very little or no ability to retract their extruded cytogels. The addition of Ca*+ (2.5 x 10m3M) to the synthetic medium restores the ability of the extruded cytogels to contract. Optimal cytogel contraction requires the presenceof Ca2+,a pH of 6.8-7.8, and an ionic strength TABLE EFFECT OF INHIBITORS
Inhibitor plus lo-’ MADP’ Agent Group I Adenosine Iodoacetate PGE, Group II NEM Group III Colchicine EDTA ATP Salyrgan PHMB Ouabain (I .o Y 10-J M)
Shape change Extrusion
No No No
No No No
No No No
No No No
No No No No No
No No No No No
Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes
No No No No No
’ Final concentrations as shown in text. b Incubated with inhibitor for 2 min. ’ Incubated for 2 min with inhibitor followed by 5-30 set treatment with ADP. Duration of ADP treatment did not influence results.
of about 0.15 M (NaCl). The requirements of additional factors such as fibrinogen for cytogel interactions are at present being investigated. Eficts of Inhibitors on Early Events Preceding Aggregation The various inhibitors of aggregation studied in order to determine their site of action in the activation-extrusion-contraction process of platelet aggregation are shown in
FIG. 9. Shadow-casted platelets treated with inhibitors of aggregation. (a) Group I inhibitors only ~16,291; (b) Group I inhibitors +10m4M ADP ~13,823; (c) NEM ~15,797; (d) NEM + ADP ~15,797. See text for inhibitor type and final concentrations.
Table 1. A double-blind study showed that with two exceptions, namely NEM and colchicine, none of the drugs listed in Table 1 produced any significant change in the shape of the platelet within the 2-min incubation period. None of the inhibitors tested by themselvescausedany visible cytogel extrusion (Fig. 9). NEM causeda very distinct
INDUCERS AND INHIBITORS
shape change and distortion of the cell (change in shadow height) without producing extrusion (Fig. 9). Colchicine caused what appeared to be only a slight rounding or swelling of the cell without any cellular distortion or cytogel extrusion (Fig. 10). A further double-blind study was undertaken to determine the effects of these inhi-
FIG. 10. Shadow-casted platelets treated with inhibitors of aggregation. (a) colchicine ~16,291; (b) colchicine + ADP x 17,770; (c) salyrgan + ADP x 18,960; (d) ATP + ADP x 13,160.
bitors on 10m4M ADP-induced aggregation. Each of the inhibitors acted at specific stepsin the ADP-induced sequenceof events. On the basesof their actions at different stagesin the activationextrusion-contraction process,we have subsequently classified the inhibitors in Table 1 into three specific groups (Groups I, II and III). Group I inhibitors. This group includes adenosine, iodoacetate, and PGEi. These inhibitors appear to exert their action at the level of the membrane. No shapechange or extrusion takes place with the inhibitor alone (Fig. 9a) or after the addition of ADP
(Fig. 9b). The cells maintained their normal shapeand were indistinguishable from control, unactivated platelets. Group II inhibitors. Thus far, we have found only one inhibitor that produced an extensive shape change under our experimental conditions, namely NEM. Although NEM also inhibited ADP-induced cytogel extrusion like the Group I inhibitors, we have classified it separately for the present becauseof the extensive shapechange and cellular distortion produced by this inhibitor alone (Fig. SC).The subsequentaddition of ADP did not produce any additional shape changes (Fig. 9d). Group III inhibitors. This group includes colchicine, EDTA, ATP, salyrgan and PHMB. Only colchicine produced a slight rounding of the platelets when incubated with the inhibitor alone (Fig. 10a). No effect was observed on ADP-induced cytogel extru-
FIG. 11. Shadow-casted platelets treated with inhibitors of aggregation. (a) PHMB + ADP x 16,160; (b) EDTA + ADP ~13,820.
sion. However, contraction of the extruded cytogels was completely inhibited (Figs. 10 and 11). It should also be noted that Group III inhibitors apparently inhibited the interaction of extruded cytogels, because,as shown in Figs. 10 and 11, all of the ADP or thrombin activated cells remained as separate entities (extruded cells). Ouabain had no effect on the ADP-induced sequenceof eventspreceding aggregation or aggregation itself. The various stagesat which Groups I, II and III inhibitors act in the early sequenceof eventsleading to aggregation, is discussedin the relaxation-contraction model for platelet aggregation proposed in a succeedingpaper (Booyse and Rafelson, 1972). Ej2ct of Group III inhibitors on ADP-induced release. The addition of an activating agent such as ADP or thrombin to platelets pretreated with any of the Group III inhibitors resulted in normal cytogel extrusion with very little or no cytogel interaction and no cytogel contraction (Figs. 10 and 11). When ADP or thrombin-induced cytogel
INDUCERS AND INHIBITORS
FIG. 12. Interplatelet bonds formed by treatment of platelets with lo-’ M ADP. (a) shadow-casted interplatelet bonds x 16,291; (b, c)interplatelet bondsstained immunohistochemically withmonospecific antihuman thrombosthenin serum (rabbit) as described in text ~29,620. Arrows indicate interplatelet bonds.
extrusion was maintained by the presenceof a Group III inhibitor, we were unable to demonstrate the presenceof either ADP or serotonin in the platelet-poor plasma of the activated preparation. Only after contraction of the extruded cytogels had taken place could we detect released ADP and serotonin. The same was true for cold-activated, TABLE 2 EFFECTS OF INHIBITORS AND COLD (5”) ON THROMBOSTHENIN ACTIVITY AND SUPERPRECIPITATION
Inhibitor” None (Control) Group 1 PGE, Iodoacetate Group II NEM Group III EDTA Colchicine Salyrgan PHMB Cold (5”)
5.4 16.6 4.4 4.8 3.5
27.1 83.0 22.0 23.7 17.5
No No No No No
’ Final concentration as shown in text. ’ nmoles PJmg proteimmin. ’ Expressed as a percentage of control activity.
extruded platelets. As long asthe cytogel was maintained in the extruded state by the low temperature, neither ADP nor serotonin was released.Upon heating the cytogels contracted and the releaseof ADP and serotonin was obtained. From these data, we have concluded that contraction of the extruded cytogels is a prerequisite for the release reaction which appears to be a late reaction in the seriesof events we have described. Effect of inhibitors on ATPase activity and superprecipitation. As shown in Table 2, Group I inhibitors enhanced purified ATPase activity by as much as 20-40 % and had very little or no effect on superprecipitation. NEM (Group II) is a powerful inhibitor of ATPase activity and superprecipitation. The Group III inhibitors are all inhibitors of either ATPase activity and/or superprecipitation (Table 2). DISCUSSION We have established that a wide variety of aggregating agents and conditions such as ADP, thrombin, polylysine, collagen, glass and cold induce a common sequenceof reactions the end result of which is aggregation. It would be tempting to speculatethat all of these agents exert their action on the same surface or membrane associated macromolecule such as S-thrombosthenin (Booyse et al., 1971; Rafelson and Booyse, 1971) and that thrombosthenin and its moieties, platelet myosin (Booyse et al., 197la) and platelet actin (Booyse et al., 1972b), are the fundamental reactants and by their properties and ultrastructural localization determine and control the early events leading to platelet-platelet interactions. All of the aggregating agentsdescribedhere seemto exert their action in such a way as to change the membrane permeability sufficiently to allow the nonparticulate, thrombosthenin-rich cytoplasm to be extruded by the platelet. The possible mechanisms by which the diverse agents and conditions such as ADP, thrombin, collagen, polylysine, glassand cold can causechangesin platelet permeability and trigger the early sequenceof events leading to aggregation will be considered in the following two papers in this series(Booyse et al., 1972a; Booyse and Rafelson, 1972).It also should be noted that the actions of both inducers and inhibitors of aggregation are based upon producing changes in the properties of a common molecule, namely the platelet contractile protein, thrombosthenin. The actomyosin-like protein, thrombosthenin, constitutes about 15% of the total platelet protein and would, therefore, be expectedto form a major portion of the cytoplasm of the platelet. Furthermore, the properties of contractile proteins are such that when a viscous contractile protein-rich proteinaceous medium is rapidly exposed to a high Ca2+ and low ionic strength and pH medium such as blood or plasma (0.15 p, pH 6%7.4), rapid mixing and dispersion of protein medium is counteracted by the rapid polymerization and gel formation of the contractile proteins. The extruded thrombosthenin-rich cytoplasm of the platelet does, in fact, appear to undergo a rapid solution-gel transition when it comes into contact with the plasma and remains as such, physically associated with the platelet per se. Although ADP, thrombin and polylysine-induced events are indistinguishable using the shadow-casting technique, there does seemto be some differences between lo-’ M and 10m4M ADP-induced events. At the present time, the data is inconclusive as to whether the amorphous material appearing on the surface of a lo-’ M ADP activated platelet does, in fact, represent cytogel extrusion or merely changes (conformational)
INDUCERS AND INHIBITORS
of surface or membrane associated macromolecules. The interaction of only limited amounts of surface associated material, as indicated by the relatively few interplatelet bridges shown by the shadow-casting technique, versus the very extensive interaction of extruded cytogels in 10m4M ADP activation, could provide a basis for our understanding of the reversibility of lo-’ A4 ADP-induced aggregation. It is difficult to visualize how an entire extruded, interacted and contracted cytogel could be induced to reverse all these reactions and again be induced to aggregate. However, whatever the actual location of the interacting material, be it on the surface or extruded from the cytoplasm of the cell, the sequenceof eventsdoesinvolve the formation of very distinct interplatelet bonds as described by Hovig (1968) and Shirasawa and Chandler (1969). We have presented evidence in this paper that interplatelet bridges formed by the addition of IO-’ M ADP to platelets for 3-4 set contain immunohistochemically identifiable thrombosthenin. We have also shown previously that the interplatelet bridges formed with low concentrations of thrombin also contain thrombosthenin (Booyse and Rafelson, 1971). Thus, there is significant evidence that platelet-platelet or cytogel interactions may well directly involve surface and/or cytogel thrombosthenin interactions. Group III inhibitors appear to exert their specific effectson platelet aggregation and releaseby inhibiting the early processesof cytogel interaction and contraction that precedethe actual processof platelet-platelet interaction. Specific effectsor apparent levels of action described in this paper were obtained with very brief preincubation exposures (l-2 min) of the platelets to the various inhibitors. Longer preincubation exposureswere not studied becauseof the possible deleterious effects of some of these inhibitors (Bull and Zucker, 1965; Behnke, 1970; White, 1968; White, 1969). These agents can inhibit protein-protein interactions such as the formation of actomyosin and polymers of varying length (F-actin, myofibrils, etc.) from monomeric actin, myosin and thrombosthenin (Barany et al., 1962; Bettex-Galland et al., 1962; Ikehara et al., 1961)and also inhibit ATPase activity and superprecipitation (contraction) of purified thrombosthenin. It is conceivable that inhibitors of the Group III type could inhibit cytogel protein-protein interactions and/or the contraction of the interacted cytogels becauseof inhibition of the ATPase activity of the cytogel contractile proteins. Inhibition of cytogel interaction and/or contraction would maintain or “freeze” the activated platelets in the extruded cytogel state as shown by the data. The inhibition of cytogel contraction or retraction due, at least in part, to inhibition of the cytogel contractile protein ATPase activity, would support the energy-dependent nature of releasereactions described by Douglas (1968) and Mtirer (1968). The former author has proposed that the secretory release mechanism may be analogous to the events involved in muscle contraction. On the basis of these data, it would seemthat Grette’s (1962) original suggestion that the platelet releasereaction involved a process analogous to contraction does have some validity. The findings reported here indicate the presence of several distinct events that precedethe actual processof platelet aggregation. It is apparent that thesedynamic changes must be studied while they are occurring, that is within a few secondsafter the addition of an activating (aggregating) agent. There is no way to study these phenomena after the fact. Direct interactions of various aggregating agentswith the platelet membrane appears to be of primary importance in initiating the sequenceof molecular events leading to
BOOYSE ET AL.
aggregation. However, contrary to previous ideas, the data presentedhere show that the platelet membrane per se does not directly participate in the early events of plateletplatelet interaction. It is, in fact, the extruded cytogels that interact to form interplatelet attachment. Subsequent platelet membrane interactions appear to follow the contraction event which brings the platelets into close apposition. REFERENCES BhRhw, M., FINKELMAN, F., AND THERATTIL-ANTHONY, T. (1962). Arch. Biochem. Biophys. 98,28. BEHNKE, 0. (1970). Stand. J. Haematol. 7, 123. BE~x-GALLAND, M., AND L~~SCHER,E. F. (1961). Biochim. Biophys. Acta 49, 536. BE~XX-GALLAND, M., PORTZEHL, H., AND L~~SCHER,E. F. (1962). Nature London 193, 777. BOOYSE,F. M., HOVEKE, T. P., AND RAFEJ..SON,JR., M. E. (1968). Biochim. Biophys. Acta 157,660. BOOYSE, F. M., AND RAFEISON, JR., M. E. (1971) Ser. Haemutol. 4 (No. l), 152. BOOYSE, F. M., HOVEKE, T. P., ZSCHOCKE,D., AND RAFELSON, JR., M. E. (1971a). J. Biol. Chem. (in press). BWYSE, F. M., STERNBERGER,L. A., ZSCHOCKE, D., AND RAFELSON, JR., M. E. (1971b). J. Histochem.
Cytochem. 19, 540. BOOYSE, F. M., HOVEKE,T. P., KISIELEXI, D., AND RAFELSON, JR., M. E. (1972a). Microvas. Res. (in
press). BOOYSE, F. M., HOVEKE, T. P., AND RAFELSON, JR., M. E. (1972b). J. Biol. Chem. (in press). BOOYSE,F. M., AND RAFELSON, JR., M. E. (1972). Microvas. Res. (in press). BORN, G. V. R. (1962). Nature London 194, 927.
BORN, G. V. R. (1969). J. Physiol. 202, 93P. BULL, B. S., AND ZUCKER,M. B. (1965). Proc. Sot. Exp. Biol. Med. 120,296. CHAMBERS, D. A., SALZMAN, E. W., AND NERI, L. L. (1967). Arch. Biochem. Biophys. 119, 173. DOUGLAS, W. W. (1968). Brit. J. Pharmacol. 34, 451. GRETTE, K. (1962). Acta Physiol. &and. 56 (Suppl. 195), 5. HELLEM, A. J. (1960). Stand. J. Clin. Invest. 12 (Suppl. 51), 1. HO~IG, T. (1963). Thromb. Diath. Haemorrh. 9,248-263. HOVIG, T. (1968). Ser. Huematol. 1 (No. 2), 3. IKEHARA, M., OLTSUKA, E., KITAGAWA, S., YOOI, K., AND TONOMURA, Y. (1961). J. Amer. Chem. Sot. 83,2679. ITZIUKI, R. F., AND GILL, D. M. (1964). Anal. Biochem. 9, 401. JONES,B. M. (1966). Nature London 212, 362. JONES,P. C. T. (1966). Nature London 212, 365. KATILO~E, H. E., AND ALEXANDER, B. (1971). Blood 38, 39. MITCHELL, J. R. A., AND SHARP,A. A. (1964). Brit. J. Haematol. 10, 78. MOORER,E. H. (1968). Biochim. Biophys. Actu 162,320. MUSTARD, J. F., AND PACKHAM, M. A. (1970). Pharmacol. Rev. 22, 97. NACHMAN, R. L., MARCUS, A. J., AND SAFIER, L. B. (1967). J. Clin. Invest. 46, 1380. RAFELSON,JR., M. E., AND BOOYSE,F. M. (1971). In “Platelet Aggregation” (J. Caen, ed.), pp. 95-106. Masson et Cie, Paris, France. SCHNEIDER, W., K~BLER, W., AND GROSS,R. (1968). Thromb. Diath. Huemorrh. 20, 314. SHIRASAWA, K., AND CHANDLER, A. B. (1969). Amer. J. Puthol. 57, 127. STEWART,G. J. (1970). Thromb. Diath. Haemorrh. 23, 228. WHITE, J. G. (1968). Amer. J. Pathol. 53, 281. WHITE, J. G. (1968). Stand. J. Haemat. 5,241. WHITE, J. G. (1969). J. Puthol. 54, 467.