Mechanism and Functional Consequences of Thrombin Binding to Glycoprotein Ib.

Thrombin (Th) is a pleiotropic protease with prothrombotic, anticoagulant and proinflammatory functions. We have characterized the mechanisms through which the binding of Th to glycoprotein (GP) Ibα, a component of the GPIb-IX-V complex on platelets, may regulate these activities. First, we tested the hypothesis, based on crystallographic evidence, that both exosite I and II of α-Th are required for binding to GPIbα. For these studies, in addition to α-Th we used meizothrombin (M-Th), a precursor with functional exosite I but without exosite II, and γ-Th, a degradation product of α-Th with intact exosite II but structurally deranged exosite I. For experiments with purified components, we used a fully sulphated recombinant fragment of human GPIbα (residues -2-290; herein designated GPIbαN) with wild-type sequence (WT) or with the single mutations Y279F (prevents sulphation) or D277N. These residues are predicted to influence predominantly or exclusively the interaction with thrombin exosite I or II, respectively. In experiments measuring the formation of a stable complex in solution, GPIbαN-WT bound α-Th and γ-Th with similar characteristics, while there was no complex formed with M-Th under the same conditions. The mutation Y279F had only a modest effect, while the mutation D277N abolished formation of all complexes, indicating that exosite II is predominantly involved in binding a GPIbα fragment in solution. To obtain more biologically relevant findings, we generated BL6 mice with platelets expressing human GPIbα replacing the corresponding murine component in the GPIb-IX-V complex, either with wild-type sequence (h-WT strain) or with the mutations Y279F (h-279F strain) or D277N (h-277N strain). All had comparable surface expression of GPIbα (∼7,000 molecules/platelet). Platelets from h-WT mice bound α-Th and γ-Th in a saturable manner and with a similar apparent kd (∼50 nM) as human platelets. Because we could not label thrombin directly without affecting binding to platelet GPIbα, we used an indirect approach with biotin-PPACK inserted into the active site of α-Th, γ-Th or M-Th detected by fluorescent streptavidin-PE. With this method, there was no demonstrable M-Th binding to platelet GPIbα. Platelets from h-279F mice showed a markedly decreased binding of α-Th (∼16% of that seen with h-WT platelets and kd of 450 nM) and no binding of γ-Th. Neither α-Th or γ-Th bound to h-277N platelets. Altogether, these results show that α-Th binding to platelet GPIbα concurrently involves exosite I and II, and suggest that the deranged exosite I of γ-Th can contribute to the interaction but is not fully functional, as shown by the greater effect of the mutation Y279F on the binding with γ-Th than α-Th. Functionally, h-279F and h-277N platelets showed a significantly decreased response to stimulation by α-Th and decreased aggregation as compared to h-WT platelets, indicating that GPIbα contributes to this prothrombotic function of α-Th. In contrast, h-WT but not h-279F or h-277N platelets inhibited fibrinogen clotting, likely a consequence of GPIbα competing with fibrinogen for binding to α-Th exosite I, indicating that α-Th binding to GPIbα can also have anticoagulant consequences. Thus, we have shown that GPIbα is a relevant modulator of α-Th activity with potentially opposite effects on thrombogenesis possibly depending on the type of vascular lesion involved. Effects on inflammation remain to be explored.

No relevant conflicts of interest to declare.