Characteristics of collagen-induced fibrinogen binding to human platelets.

Polymerized type I calf skin collagen induced a time-dependent specific binding of 125I-fibrinogen to washed human platelets. Binding occurred more rapidly in a shaken rather than in an unstirred system. It was linear in the range 0.05-0.3 microM added fibrinogen and was saturated at higher fibrinogen concentrations (more than 0.8 microM). Scatchard analysis showed a single population of binding sites (16530 +/- 5410 per platelet) with a Kd = 0.53 +/- 0.23 microM. Collagen-induced 125I-fibrinogen binding to platelets was completely inhibited by ADP antagonists such as creatine phosphate/creatine phosphokinase and AMP, and partially inhibited by pretreatment of the platelets with aspirin. With both normal and aspirin-treated platelets a close correlation was observed between the amount of 125I-fibrinogen bound and the extent of dense granule secretion. Our results confirm that fibrinogen becomes bound to platelet surface receptors during collagen-induced platelet aggregation and suggest that secreted ADP is an essential cofactor in this process.

Studies on the interaction of platelet glycoprotein IIb-IIIa and glycoprotein IV with fibrinogen and thrombospondin: a new immunochemical approach.

We have designed a new binding assay based on crossed immunoelectrophoresis that allowed us to test for the relative capacities of platelet membrane glycoprotein IIb-IIIa (GP IIb-IIIa), and glycoprotein IV (GP IV) to bind purified Arg-Gly-Asp (RGD)-containing adhesive proteins. Preformed immune complexes were made by reacting a platelet lysate with murine monoclonal antibodies to GP IV (OKM5 and FA6-152) or to GP IIb-IIIa (AP-2). Upon two-dimensional electrophoretic separation in agarose gels and immunoprecipitation by a polyclonal antibody to mouse IgG, the immobilized complexes containing the desired antigen were further probed with purified 125I-labeled TSP or fibrinogen. Under these conditions, immobilized GP IV was found to specifically bind TSP, whereas it was unreactive with fibrinogen. By contrast, immobilized GP IIb-IIIa demonstrated fibrinogen binding capacity but did not demonstrate any reactivity toward TSP. These observations suggest that the overall structure of the adhesive protein may determine the accessibility of the RGD sequence to its binding site on GP IIb-IIIa.

Identification of the cytoskeletal protein alpha-actinin as a platelet thrombospondin-binding protein.

Binding of the alpha-granular thrombospondin (TSP) to the plasma membrane of activated platelets has long been documented, yet the molecular mechanism involved in its secretion and surface expression have not been elucidated. Using a ligand blot binding assay where electrophoretically separated platelet proteins were incubated with purified 125I-labeled TSP, we observed a strong interaction of [125I]TSP with a 100 kDa single chain protein. On performing a platelet subfractionation, the 100 kDa protein was predominantly localized in the cytosol from which it was purified by preparative electrophoresis and was identified by amino acid sequencing to the cytoskeletal protein, alpha-actinin. We further demonstrated that [125I]TSP interacts with alpha-actinin in a specific manner and with a high affinity (Kd = 6.6 nM) in a solid-phase binding assay.

Correlation between noradrenaline fluxes, metabolism and compartmentalisation in human platelets.

(3H) noradrenaline was taken up by human platelets and partially converted into sulfoconjugated noradrenaline. This uptake was inhibited by drugs which have been previously shown to impair the uptake of 5-HT (ouabain, chlorimipramine) or the storage of 5-HT (tyramine, reserpine) by platelets. In addition, tyramine and reserpine stimulated the formation of sulfoconjugated noradrenaline. The efflux of noradrenaline from platelets was measured in parallel and was found to be directly related to the proportion of non metabolized to metabolized noradrenaline in the cells. Unlike tyramine, which induced a similar release of noradrenaline and 5-HT, reserpine was less effective at inducing noradrenaline release than 5-HT release. This study indicates a preferential localization of noradrenaline in the granular pool of human platelets with the existence of an extragranular sulfoconjugated pool which is increased when the granular storage of noradrenaline is impaired. Studies of noradrenaline fluxes and metabolism may be useful in the understanding of both acquired and inherited platelet storage pool defects.

Quantitation of platelet fibrinogen and thrombospondin in Glanzmann's thrombasthenia by electroimmunoassay.

Fibrinogen and thrombospondin are major constituents of human platelet alpha-granules and contribute to cell-cell interactions following their release. Glanzmann's thrombasthenia is characterized by the absence of platelet aggregation and reduced levels of GP IIb-IIIa complexes and platelet fibrinogen. The level of thrombospondin is thought to be normal but has not so far been quantified. Using an electroimmunoassay method adapted from Laurell, we have measured fibrinogen and thrombospondin in platelet extracts of four patients with classical Glanzmann's thrombasthenia and two variants with abnormal platelet aggregation associated with subnormal levels of GP IIb-IIIa complexes. Triton X-100 lysates were prepared in the presence of leupeptin or EDTA to avoid endogenous calcium-dependent protease activation during the solubilization procedure. Platelet fibrinogen was not detected in one patient with type I Glanzmann's thrombasthenia; it was reduced to 5-10% of normal values in two other type I patients and to 65% of normal values in one type II patient. It was normal in patient R.P., a variant of Glanzmann's thrombasthenia with 60% of GP IIb-IIIa complexes but decreased in patient A.P. a newly described variant with 35% of GP IIb-IIIa complexes. These findings support a role for GP IIb-IIIa complexes in the packaging of fibrinogen into alpha-granules. Normal or subnormal amounts of thrombospondin were measured in thrombasthenic platelets. Patient A.P., who was investigated on two different occasions, demonstrated variable levels of thrombospondin. This underlines the need for quantifying this protein when evaluating its expression in this disorder.

Characterization of the binding of thrombospondin to human platelets and its association with the platelet cytoskeleton.

To characterize the interaction between thrombospondin and human platelets, thrombospondin was purified from the supernatant of thrombin-activated human platelets, labeled with iodine 125, and allowed to interact with the washed platelets. With concentrations of 10 to 50 micrograms/ml, only minute amounts of 125I-labeled thrombospondin bound to resting platelets or to platelets activated by adenosine diphosphate. In contrast, when platelets were stimulated with thrombin, binding increased fivefold to sixfold in a time-dependent and 125I-labeled thrombospondin concentration-dependent manner. Binding of 125I-labeled thrombospondin to thrombin-activated platelets required the presence of divalent cations, proceeded concomitantly with platelet release, and at a concentration of 1 nmol/L thrombin, reached a maximum of 2200 +/- 260 molecules of 125I-labeled thrombospondin bound per platelet. After its binding to platelets, 125I-labeled thrombospondin was not internalized, because up to 85% of the 125I-labeled thrombospondin was dissociated from the cell surface by adding ethylenediaminetetraacetic acid. Using various experimental approaches, including studies with severe type I thrombasthenic platelets, we further demonstrated that the interaction of 125I-labeled thrombospondin with thrombin-stimulated platelets occurred as a fibrinogen- and fibrin-independent process, and that the glycoprotein IIb-IIIa complex did not function as a physiologic plasma membrane receptor for 125I-labeled thrombospondin. Last, about 60% of the 125I-labeled thrombospondin molecules bound to the platelet surface were found to be associated with the platelet cytoskeleton recovered from platelets solubilized with Triton X-100. On Western blot analysis, this cytoskeletal fraction lacked detectable glycoprotein IV, the putative platelet receptor for thrombospondin. These results suggest that on the surface of thrombin-activated platelets, a fraction of 125I-labeled thrombospondin does not associate with glycoprotein IV but instead with other plasma membrane components that have yet to be identified.

Platelet aggregation: its relation with ADP-induced fibrinogen binding to platelets and ADP-related membrane enzyme activities.

The receptor for ADP on the platelet membrane, which triggers exposure of fibrinogen-binding sites and platelet aggregation, has not yet been identified. Two enzymes with which ADP interacts on the platelet surface, an ecto-ATPase and nucleosidediphosphate kinase, have been proposed as possible receptors for ADP in ADP-induced platelet aggregation. In the present study, experiments were conducted with washed human platelets to examine if a relationship existed between platelet aggregation, fibrinogen binding and the enzymatic degradation of ADP. With 12 different platelet suspensions, a good correlation (P less than 0.01) was found between the extent of platelet aggregation and the amount of 125I-fibrinogen bound to platelets after ADP stimulation. No correlation was found between these parameters and the rate or extent of transformation of [14C]ADP to [14C]ATP or [14C]AMP. The binding of fibrinogen to platelets was inhibited in parallel with aggregation when ADP stimulation was impaired by the enzymatic degradation of ADP by the system creatine phosphate/creatine phosphokinase, or by the use of specific antagonists, such as ATP and AMP. These antagonists also influenced the enzymatic degradation of ADP. This effect occurred at lower concentrations of ATP or AMP than those required to inhibit ADP-induced platelet aggregation and fibrinogen binding. Our results demonstrate that ATP and AMP may be used as specific antagonists of the ADP-induced fibrinogen binding to platelets. They do not provide evidence to suggest that enzymes which metabolize ADP on the platelet surface are involved in the mechanism of ADP-induced platelet aggregation.

Studies on the mechanism of expression of secreted fibrinogen on the surface of activated human platelets.

Affinity purified anti-fibrinogen (anti-Fg) Fab fragments were used to study the mechanism of expression of alpha-granule fibrinogen on activated platelets. Low amounts of the radiolabeled anti-Fg Fab bound to unstimulated or adenosine diphosphate (ADP)-stimulated cells. They readily bound to platelets stimulated with collagen, alpha-thrombin or gamma-thrombin in the presence of divalent cations. At 1 n mol/L alpha-thrombin or 25 nmol/L gamma-thrombin, platelet fibrinogen was expressed on the surface of the cells notwithstanding the presence of AP-2, a monoclonal antibody to the glycoprotein (GP) IIb-IIIa complex, or the synthetic peptides Arg-Gly-Asp-Ser and gamma 400-411, all substances that prevented the binding of plasma fibrinogen to platelets. These results suggest that platelet fibrinogen may interact with its receptors during its translocation from the alpha-granules to the plasma membrane and, thus, not occupy the same sites as those available for plasma fibrinogen on the surface of the cell. Furthermore, we found that platelet fibrinogen was expressed on the thrombin-stimulated platelets of a Glanzmann's thrombasthenia variant that failed to bind plasma fibrinogen. Normal platelets stimulated with 5 nmol/L alpha-thrombin bound increased amounts of the anti-fg Fab, the additional expression being inhibited by the anti-GP IIb-IIIa monoclonal antibody or by Gly-Pro-Arg-Pro, an inhibitor of fibrin polymer formation. This suggests that rebinding to externally located GP IIb-IIIa complexes becomes important once fibrin is formed.

Use of a monoclonal antibody to measure the surface expression of thrombospondin following platelet activation.

The radiolabelled monoclonal antibody, 5G11, directed against native thrombospondin, has been used to assess the surface expression of secreted thrombospondin on human blood platelets. Emphasis has been placed on studying the role of fibrinogen in this process. Unstimulated platelets bound low amounts of 5G11 (about 2000 molecules/platelet). Binding increased 2-fold and 5-7-fold after stimulation of platelets with ADP or thrombin (or ionophore A23187) respectively. Unstimulated platelets from patients deficient in alpha-granule proteins (gray platelet syndrome) bound baseline levels of 5G11. However, binding was not increased after activation. Thrombospondin expression on thrombin-stimulated normal platelets was for a large part divalent-cation-dependent and was not affected by AP-2, a monoclonal antibody to GPIIb-IIIa complexes. However, binding of 5G11 was some 50% lower when platelets were stimulated in the presence of Fab fragments of a polyclonal rabbit antibody to fibrinogen. This suggested either a direct binding of thrombospondin to surface-bound fibrinogen or a steric inhibition due to a close proximity of the two proteins. The fact that binding of 5G11 was at the lower limit of the normal range to the stimulated platelets of an afibrinogenaemic patient specifically lacking detectable fibrinogen favoured the latter explanation. Thus, a major fibrinogen-independent pathway for thrombospondin expression must exist.

Molecular requirements for the interaction of thrombospondin with thrombin-activated human platelets: modulation of platelet aggregation.

We have investigated the molecular requirements for thrombospondin (TSP) to bind to the platelet surface and to support the subsequent secretion-dependent platelet aggregation. For this, we used two distinct murine monoclonal antibodies (MoAbs), designated MAI and MAII, raised against human platelet TSP, and three polyclonal antibodies, designated R3, R6, and R5, directed against fusion proteins containing the type 1 (Gly 385-Ile 522), type 2 (Pro 559-Ile 669), and type 3 (Asp 784-Val 932) repeating sequences, respectively. Among them, R5 and R6, but not R3, inhibited thrombin-induced aggregation of washed platelets and the concomitant secretion of serotonin. These antibodies, however, did not inhibit the expression of TSP on thrombin-activated platelets, as measured by the binding of a radiolabeled MoAb to TSP, suggesting that they may inhibit platelet aggregation by interfering with a physiologic event subsequent to TSP binding. In contrast, MoAb MAII, which reacts with an epitope located within the heparin-binding domain of TSP, inhibited both TSP surface expression and platelet aggregation/secretion induced by thrombin. In addition, this MoAb inhibited in a dose-dependent manner (IC50 approximately 0.5 mumol/L) the interaction of 125I-TSP with immobilized fibrinogen and platelet glycoprotein IV, both potential physiologic receptors for TSP on thrombin-activated platelets. These results indicate that the interaction of TSP with the surface of activated platelets can be modulated at the level of a specific epitope located within the amino terminal heparin-binding domain of the molecule. Thus, selective inhibition of the platelet/TSP interaction may represent an alternative approach to the inhibition of platelet aggregation.

Evidence for an alpha-granular pool of the cytoskeletal protein alpha-actinin in human platelets that redistributes with the adhesive glycoprotein thrombospondin-1 during the exocytotic process.

In a previous study, we have demonstrated that the platelet adhesive glycoprotein thrombospondin-1 (TSP-1) interacts specifically with the cytoskeletal protein alpha-actinin in a solid-phase binding assay. Stored in the alpha-granules of platelets, TSP-1 is secreted during cell activation and binds to the plasma membrane promoting the platelet macroaggregate formation. However, the molecular mechanism by which TSP-1 reaches and binds to the platelet surface is to date unelucidated. alpha-Actinin is an actin-binding and actinin-cross-linking protein that is present in most cells and may act as a link between the bundles of F-actin and the plasma membrane. In this study, we have investigated a possible interaction of alpha-actinin with TSP-1 in platelets by examining their respective subcellular location during the platelet activation process. By indirect immunofluorescence. alpha-actinin was found to display a granular staining in resting platelets similar to that of TSP-1. Performing postembedding immunogold labeling for electron microscopy, we detected the presence of alpha-actinin throughout the cytoplasm, but the strongest gold staining was found in organelles identified as alpha-granules on the basis of their ultrastructure and TSP-1 content. With the use of double immunogold labeling on platelets at different stages of activation by thrombin, both alpha-actinin and TSP-1 were seen redistributing from the alpha-granules to the platelet surface via the open canalicular system (OCS). At the same time, the cytoplasmic alpha-actinin concentrated toward the plasma membrane, but no colocalization with the F-actin bundles was evidenced. Finally, preembedding immunogold labeling and immunoprecipitation of 125I-surface-labeled, thrombin-activated platelets further demonstrated that alpha-actinin was expressed on the plasma membrane in the absence of any detectable expression of actin and that it could from molecular complexes with TSP-1 on activated platelets. These results suggest that alpha-actinin found to be present on the platelet surface together with TSP-1 originates in the alpha-granules by fusion of the alpha-granules with the plasma membrane during platelet exocytosis.