Recently, Jin et al addressed the generally misunderstood problem of the factors responsible for adenosine diphosphate (ADP)–induced thromboxane A2 (TxA2) production by human platelets.1 They used stirred, washed human platelet suspensions to which fibrinogen but no CaCl2 had been added; as a consequence, the concentration of external Ca++ ([Ca++]o) was much lower than physiological. The main results of their study can be summarized as follows: (i) the ADP receptor antagonists A2P5P (anti-P2Y1) or AR-C67085 (anti-P2Y12) inhibited platelet aggregation and TxA2 production induced by ADP; (ii) the fibrinogen receptor antagonist SC49992 inhibited platelet aggregation and TxA2 production induced by ADP; (iii) the Fab fragment of ligand-induced binding site 6 (LIBS6) antibody, which induces a fibrinogen binding site on αIIbβ3, caused platelet aggregation and TxA2 production when added to platelet suspensions with fibrinogen; (iv) A2P5P (P2Y1 antagonist) or AR-C67085 (P2Y12 antagonist), when added together or alone, inhibited the secondary wave of aggregation and TxA2 production induced by LIBS6 plus fibrinogen; and (v) in the presence of physiological concentrations of Ca++, there was no production of TxA2 by ADP-aggregated platelets. The authors concluded that ADP induces TxA2 generation in human platelets, which requires coordinated signaling through the integrin αIIbβ3 and ADP receptors.
We do not agree with the authors' interpretation of their own results. Rather, we think that their results are in accord with well-established knowledge of the factors involved in ADP-induced TxA2generation: (1) ADP does not stimulate TxA2 production directly; (2) it is the close platelet-to-platelet contact that is brought about by ADP-induced platelet aggregation that triggers the production of TxA2; and (3) this effect is greatly enhanced and can be seen in most healthy individuals when [Ca++]o is decreased to micromolar levels.2-5 Based on this interpretation, it is not surprising that the experiments performed by Jin et al showed that antagonists of P2Y1, P2Y12, or the fibrinogen receptor, which inhibit ADP-induced platelet aggregation, abolished the TxA2production. The dependency of TxA2 production and the ensuing platelet secretion on platelet aggregation is demonstrated by the observation that no TxA2 production or platelet secretion occurs from normal human platelets that are stimulated by ADP, even at high concentration, under conditions in which platelet aggregation does not occur: for instance, if the receptor function for adhesive proteins on αIIbβ3 is inhibited with inhibitory monoclonal antibodies or Arg-Gly-Asp–containing peptides, or, more simply, if the platelet suspension is not stirred. The mechanism by which platelet aggregation triggers TxA2 formation at low [Ca++]o is presently unknown. But the mechanism is by no means selective for ADP-induced platelet aggregation, because similar effects occur when platelets are aggregated by other weak agonists,6 and it can be observed also when close platelet-to-platelet contact is brought about by LIBS6 and fibrinogen,1 or the agglutinating agents polylysine or ristocetin.7,8 The platelet-to-platelet contact caused by weak agonists, agglutinating agents or LIBS6 and fibrinogen causes the formation of trace amounts of TxA2 and the secretion of ADP, which supports the secondary full aggregation responsible for the production of large amounts of TxA2. None of these responses, including TxA2 production, is observed when the same agonists or agglutinating agents are added to the same platelet suspension under nonstirring conditions.6 This observation clearly indicates that a definite distinction should be made betweenplatelet aggregation and fibrinogen receptor occupancy. When platelets are exposed to ADP under nonstirring conditions to prevent their aggregation, ADP elicits the inside-out signaling, which promotes the occupancy of the fibrinogen receptor on αIIbβ3, which in turn elicits the outside-in signaling. According to the model proposed by Jin et al, this should be enough to stimulate the generation of TxA2by platelets; under these conditions, however, TxA2production does not occur: it is only when platelets are stirred and, as a consequence, they are allowed to aggregate that platelet TxA2 production can be observed.
Finally, the fact that platelet secretion and the production of large amounts of TxA2 occur with platelets from most healthy individuals only when [Ca++]o is at micromolar levels demonstrates that the whole process that can be observed in the turbidometric aggregometer is mostly an in vitro artifact.2-5 (The widespread use of citrate as an anticoagulant in the preparation of platelet-rich plasma introduces the same artifact as suspending media without added CaCl2.) Despite this, Jin et al suggest that ADP-induced TxA2production may have physiological relevance in vivo, because [Ca++]o might be “drastically lower” 1(p196) at sites of thrombus formation. We do not know of any support for this hypothesis, and we believe that, due to the secretion of Ca++ from platelet granules, [Ca++]o might even increase, rather than decrease. Needless to say, we agree with the authors' statement that the results of clinical trials with aspirin indicate that TxA2 plays an important role in thrombus formation in vivo.9But this is not a demonstration that, in vivo, ADP causes platelet TxA2 formation. Platelet thrombus formation in vivo depends on the synergistic interaction of several agonists, some of which, such as collagen and thrombin, directly cause the formation of TxA2. The recent trials combining drugs such as clopidogrel with aspirin10 add evidence to the fact that separate pathways contribute to thrombosis and that combined therapies are more effective than aspirin alone or clopidogrel alone.
Signaling events regulating thromboxane A2generation in platelets
Cattaneo et al argue that the results of our recently published paper are in accord with well-established knowledge of the factors involved in adenosine diphosphate (ADP)–induced thromboxane A2 generation, but they do not agree with the interpretation of our results. They state that (1) ADP does not stimulate thromboxane A2 production directly; (2) it is the close platelet-to-platelet contact that is brought about by ADP-induced platelet aggregation that triggers the production of thromboxane A2; and (3) this effect is greatly enhanced and can be seen in most healthy individuals when extracellular calcium is decreased to micromolar levels.
Our results have demonstrated that blockade of the P2Y1 receptor by A2P5P, of the P2Y12 receptor by AR-C67085, or of fibrinogen binding to its receptor by SC49992 inhibits both platelet aggregation and thromboxane A2 generation.1-1 These results are consistent with the argument of Cattaneo et al that ADP-induced thromboxane A2 generation is dependent on ADP-induced platelet aggregation. We addressed this issue and went further to investigate whether A2P5P and AR-C67085 block thromboxane A2 generation simply by inhibiting ADP-induced platelet aggregation, or whether ADP contributes beyond that to signaling events in thromboxane A2 generation. ADP under stirring conditions causes platelet aggregation, which results in close cell-to-cell contact. Signaling under these conditions is termed as “outside-in” signaling, which Shattil et al defined as “initiated at localized regions of cell matrix and cell-cell contact.” 2(p2651)This outside-in signaling translates into several intracellular events including tyrosine phosphorylation of platelet proteins.1-2 Blockade of fibrinogen binding to activated integrin αIIbβ3 by Arg-Gly-Asp-Ser, EDTA (ethylenediaminetetraacetic acid), or mAb A2A9 abrogated tyrosine phosphorylations of platelet proteins, indicating the important role of fibrinogen binding.1-3 Our results1-1 with SC49992 are consistent with the importance of fibrinogen binding in outside-in signaling events. When LIBS6 Fab fragments and fibrinogen were used to induce platelet aggregation without adding any other agonist, thromboxane A2was generated.1-1 But under these conditions, the ADP-receptor antagonists A2P5P or AR-C67085 inhibited thromboxane A2 generation.1-1 Furthermore, a combination of both A2P5P and AR-C67085 nearly abolished thromboxane A2generation induced by LIBS6 and fibrinogen, without affecting primary aggregation, although full irreversible aggregation was impaired.1-1 But inhibition of full, irreversible aggregation by wortmannin did not impair thromboxane A2accumulation in response to LIBS6 plus fibrinogen. These data suggest that the close platelet-to-platelet contact that is brought about by platelet aggregation (primary or full) is not sufficient to generate thromboxane A2 and depends on signaling from the ADP receptors. Hence we disagree with the interpretation of Cattaneo et al of the previously published results from several laboratories. We stand by our interpretation of our results that outside-in signaling from both the integrin αIIbβ3 and the P2 receptors is required for ADP-induced thromboxane A2 generation in platelets. We agree with Cattaneo et al that ADP-induced thromboxane generation is greatly enhanced and can be seen in most healthy individuals when extracellular calcium is decreased to micromolar levels.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal