Isolation, purification, and partial characterization of platelet membrane glycoproteins IIb and IIIa

Molecular mechanisms of platelet exocytosis: role of SNAP-23 and syntaxin 2 in dense core granule release

To characterize the molecular mechanisms of platelet secretion, we focused on the calcium-induced exocytosis of dense core granules. Platelets contain several known t-SNAREs (soluble N-ethylmaleimide sensitive factor [NSF] attachment protein receptors) such as syntaxins 2, 4, and 7 and SNAP-23 (synaptosomal associated protein 23). By using an in vitro exocytosis assay, we have been able to assign roles for some of these t-SNAREs in dense core granule release. This calcium-induced secretion relies on the SNARE proteins because it is stimulated by the addition of recombinant -SNAP and inhibited by a dominant negative -SNAP–L294A mutant or by anti–-SNAP and anti-NSF antibodies. SNAP-23 antibodies and an inhibitory C-terminal SNAP-23 peptide both blocked dense core granule release, demonstrating a role for SNAP-23. Unlike other cell types, platelets contain a significant pool of soluble SNAP-23, which does not partition into Triton X-114. Of the anti-syntaxin antibodies tested, only anti–syntaxin 2 antibody inhibited dense core granule release. Immunoprecipitation studies showed that the 2 t-SNAREs syntaxin 2 and SNAP-23 do form a complex in vivo. These data clearly show that SNAPs, NSF, and specific t-SNAREs are used for dense core granule release; these data provide a greater understanding of regulated exocytosis in platelets.

Erythrocytes with covalently bound fibrinogen as a cellular replacement for the treatment of thrombocytopenia

Thrombocytopenia in general, and autoimmune thrombocytopenia in particular, is a disease of high prevalence with a non-satisfactory regime of treatment. The present study aimed to explore the feasibility of an alternative treatment, based on the rationale that autologous erythrocytes modified to bear covalently bound fibrinogen would participate passively in the aggregation of the remaining platelets, thus augmenting the haemostatic needs, while resisting the autoimmune reaction directed towards the platelets. Several procedures for the cross-linking of fibrinogen to red blood cells (RBCs) were tested. Formaldehyde (33 microM) for 10 min at 23 degrees C attached 58 fibrinogen molecules per erythrocyte. These erythrocytes were indistinguishable from untreated erythrocytes in the following properties: osmotic fragility, bound haemoglobin, sedimentation rate, acetylcholinesterase activity, phagocytosis by macrophages, rosette formation with K562 cells. It is shown that RBCs cross-linked with fibrinogen are capable of participating in the in vitro aggregation of platelets and are indeed effective in the in vivo process of arrest of bleeding in an animal model of autoimmune thrombocytopenia.

A large-scale procedure for the isolation of integrin GPIIb/IIIa, the human platelet fibrinogen receptor

The heterodimer GPIIb/IIIa, formed by the Ca(2+)-dependent association of glycoproteins IIb (GPIIb) and IIIa (GPIIIa), is the major integrin at the platelet surface, where it serves as the receptor for fibrinogen and other adhesive proteins and plays a central role in platelet aggregation and in platelet adhesion to the subendothelium. Here we describe a procedure for the isolation of GPIIb/IIIa using as starting material either the whole particulate fraction, obtained by differential centrifugation after hypoosmotic lysis of glycerol-loaded platelets, or any of the fractions obtained by density gradient centrifugation of the whole particulate fraction. The procedure consists simply of differential extraction with Triton X-100 of the starting particulate fraction, anion-exchange chromatography of the 4% Triton X-100 supernatant, and size-exclusion chromatography of the GPIIb/IIIa-rich fraction retained in the ion-exchange column. The use of particulate fractions instead of whole platelets as the starting material for extraction together with differential extraction with Triton X-100 (two steps that are simple and inexpensive to perform) results in the early removal of many unwanted proteins, which otherwise would have to be removed at later stages at the expense of severely impairing the final yield of GPIIb/IIIa. Pure GPIIb/IIIa is obtained with a yield of about 48%, the highest so far reported, calculated with respect to the GPIIb and GPIIIa content in the starting particulate fraction. The final product can be stored in freeze-dried form without apparent changes in its physical and chemical properties.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of an antibody raised against reduced glycoprotein IIIa of human platelets

A polyclonal antibody against reduced and vinylpyridylethylated human glycoprotein IIIa was raised in rabbits. Its reactivity with reduced GPIIIa was about 500 times higher than that of the antibody against native GPIIIa. The lowest amounts of purified reduced and native GPIIIa recognized by the antibody against reduced GPIIIa were 25 and 400 ng, respectively. The antibody did not recognize native GPIIIa (about 1-2 micrograms) in platelet extracts and chymotryptic degradation products of GPIIIa. It inhibited ADP-induced platelet aggregation but it did not inhibit fibrinogen binding to ADP-stimulated platelets. Our experiments suggest that the antigenicity of GPIIIa (beta 3 integrin) depends on the conformation of the molecule determined by numerous S-S bridges between cysteine residues.

Secondary signals mediated by GPIIb/IIIa in thrombin-activated platelets

We have previously found that stimulation of aequorin-loaded platelets by thrombin produced a two-peaked increase in intracellular free calcium concentration ([Ca2+]i), and the development of the second peak of [Ca2+]i was closely related with the aggregation. In this report, we studied the interrelationship between the GPIIb/IIIa complex, aggregation, cytoskeletons and [Ca2+]i of platelets. The pretreatment of the platelets with dihydrocytochalasin B (4 microM), an actin polymerization inhibitor, did not inhibit aggregation and TXB2 production, but did inhibit both actin polymerization and the second peak of [Ca2+]i increase induced by thrombin, suggesting that actin polymerization and the second peak of [Ca2+]i are interrelated. GRGDSP (100 microM), a synthetic anti-adhesive peptide, has already been reported to inhibit platelet aggregation and the second peak of [Ca2+]i induced by thrombin. It also inhibited actin polymerization and TXB2 production, suggesting that aggregation was important for not only the generation of the second peak of [Ca2+]i but also for actin polymerization and TXB2 production. PGI2 (5 nM) did not abolish but only delayed aggregation, TXB2 production, actin polymerization and the second peak of [Ca2+]i increase. These findings suggest that the secondary signals are caused by aggregation (fibrinogen-binding to the GPIIb/IIIa) in thrombin-aggregated platelets, which results in the TXA2 production and the secondary peak of [Ca2+]i increase, and the latter was dependent on actin polymerization.

Inhibition of fibrin(ogen) binding to stimulated platelets by a monoclonal antibody specific for a conformational determinant of GPIIIa

We have purified the integrin GPIIb:IIIa from the membrane fraction of human blood platelets by lentil lectin affinity chromatography followed by gel filtration chromatography. With purified GPIIb:IIIa as an antigen, we have produced monoclonal antibody CS-1, which immunoblotting demonstrates to be specific for native GPIIIa; disulfide bond reduction of GPIIIa resulted in loss of immunoreactivity. Radiolabelled ligand binding studies revealed that CS-1 recognized approximately 55,000 sites per platelet and bound with a Kd in the nanomolar range, independent of the state of platelet activation. Binding of CS-1 or its Fab fragment to ADP- and thrombin-stimulated gel filtered platelets caused a 2-3 fold inhibition of binding the soluble ligands fibrinogen and fibrin protofibrils. CS-1 also inhibited aggregation of ADP- and thrombin-stimulated platelets by 2- and 4-fold, respectively. Since CS-1 inhibits fibrin(ogen) interactions with GPIIb:IIIa, we propose that the conformationally dependent epitope on GPIIIa recognized by CS-1 constitutes a region of the receptor which is involved in fibrin(ogen) binding.

Human fibrinogen

The structure and physical properties of human fibrinogen and fibrin are reviewed along with methods for the detection of products of their metabolism. Interactions of human fibrinogen with thrombin, factor XIII, plasminogen, glycoprotein IIb/IIIa, and other proteins are related to their relevance to thrombosis and hemostasis. To the extent information is available, the structural determinants of these interactions are delineated, and kinetic and thermodynamic parameters associated with the interactions are listed. Individual steps in the reaction pathway for the conversion of fibrinogen to cross-linked fibrin are characterized. The altered hemostatic properties of mutational variants of fibrinogen are related to their altered structure. The structures of the genes coding for the polypeptide chains of fibrinogen are discussed along with the current state of knowledge of the control and regulation of fibrinogen synthesis. Fibrinogen catabolism and fibrinolysis are also reviewed.

Blood elements at surfaces: platelets

This brief review is designed to highlight the limitations of our knowledge given the difficulties in simulating the in vivo condition in models established in vitro and ex vivo. Despite this uncertainty, mounting evidence indicates that the GPIIb-IIIa receptor mediates crucial platelet functions and that control over its interactions may offer the potential for altering platelet behavior in vivo. Since this receptor appears to be part of a generalized family of adhesive protein receptors that utilize the tripeptide RGD as a recognition signal, insights derived from platelet function may have broad biological implications.

A method for purifying the platelet membrane glycoprotein IIb-IIIa complex

A method has been developed for the rapid isolation of platelet membrane glycoproteins (GP) IIb and IIIa. This method produces an excellent yield and does not require the prior isolation of platelet membranes. Outdated platelets were washed and solubilized in Triton X-100. Concanavalin A affinity chromatography was used to purify a platelet glycoprotein fraction. The concanavalin A-retained glycoproteins were eluted and adsorbed with a heparin-Sepharose column to remove a major contaminant, thrombospondin. Sephacryl S-300 gel filtration was used as the final purification step to remove most fibrinogen and low-molecular-weight contaminants. Wheat germ agglutinin affinity chromatography was used to completely remove trace amounts of fibrinogen. The purified GP IIb and GP IIIa were analyzed by sucrose gradient sedimentation and found to consist of heterodimer complexes.

Interaction of fibronectin with its receptor on platelets

We report that the 12,000 dalton domain of fibronectin that interacts with fibroblast cell surfaces also binds specifically to thrombin-inducible, saturable receptors on platelets. Furthermore, we have used chemical cross-linking and monoclonal antibodies to show that the 12,000 dalton domain of fibronectin interacts directly with glycoprotein IIIa at the platelet cell surface. Both binding and cross-linking of this domain to platelets are competed by a hexapeptide previously shown to block fibroblast adhesion to fibronectin. Finally, we show that a complex of the platelet glycoproteins IIIa and IIb binds to affinity columns of a cell-attachment fragment of fibronectin. These results localize a major fibronectin-platelet interaction to a specific domain of fibronectin and to a specific platelet glycoprotein.

Cleavage of a 135 kD cell surface glycoprotein correlates with loss of fibroblast adhesion to fibronectin

We have previously described a group of three plasma membrane glycoproteins that are recognized by an adhesion-disrupting antiserum and that are involved in fibronectin-mediated BHK cell adhesion. A peculiar property of these molecules is their resistance to tryptic digestion. We have now extended this study in the attempt to identify the active component within this group of molecules. SR/BALB mouse fibroblasts, used in this work, expose at their surface only two trypsin-resistant glycoproteins, gp1 (150 K) and gp2 (135 K), that are recognized by the adhesion-disrupting anti-BHK serum. Controlled proteolysis of the cell surface in the presence of a reducing agent results in the loss of cell adhesion to fibronectin-coated substratum. gp2 is selectively cleaved under these conditions. Moreover, cells treated with trypsin and reducing agent can no longer adsorb the adhesion-relevant antibodies from the anti-BHK serum. These data indicate that gp2 plays a critical role in the adhesion of SR/BALB fibroblasts to fibronectin-coated substratum, and that disulfide bonds are important in the conformation and function of this molecule.

A monoclonal antibody binding to human medulloblastoma cells and to the platelet glycoprotein IIB-IIIA complex

A monoclonal antibody, designated M148, produced by the hybridoma technique from spleen cells of mice immunized with human medulloblastoma, was found by indirect immunofluorescence to bind to normal human platelets (both PlA1 positive and PlA1 negative) and megakaryocytes, as well as to some medulloblastoma and neuroblastoma cells and cell lines and certain other solid tumours. No binding was observed to other marrow constituents, nor to any other normal tissue examined. The antibody bound to platelets from a patient with the Bernard-Soulier syndrome but not to thrombasthenic platelets. It immunoprecipitated glycoproteins IIb and IIIa from 125I-labelled normal platelet membranes, and completely inhibited ADP-induced fibrinogen binding and aggregation of platelets. Aggregation was also inhibited in response to adrenaline, collagen, thrombin, sodium arachidonate and the ionophore A23187; clot retraction was partially inhibited. The antibody was without effect on thromboxane formation or 5-hydroxytryptamine (5HT) secretion in response to thrombin, but inhibited 5HT secretion in response to arachidonate. It did not inhibit factor VIII binding or agglutination in response to ristocetin, but completely inhibited factor VIII binding in response to thrombin. These findings suggest that the epitopes are close to the fibrinogen and factor VIII binding sites on glycoproteins IIb/IIIa, and that the lack of these glycoproteins is sufficient explanation for the pattern of dysfunction observed in thrombasthenic platelets, without invoking any other membrane abnormality.

Preparation of monoclonal antibodies against glycoprotein IIIa of human platelets. Their effect on platelet aggregation

Monoclonal antibodies against purified glycoprotein IIIa (GPIIIa) of human platelet membranes have been obtained. These antibodies, except one, are able to bind to intact platelets; the exception is M108/p98 antibody which recognizes a new epitope, unmasked after proteolysis of GPIIIa in vitro. Several antigenic areas can be delineated on the molecule, by testing the ability of different antibodies to compete in their simultaneous binding to GPIIIa. One of the monoclonal antibodies inhibits ADP-induced platelet aggregation while others do not have an effect or induce agglutination of platelets independent of ADP. Conventional antiserum raised against purified GPIIIa also blocks the aggregation induced by ADP. These results favour the hypothesis that GPIIIa plays a direct role in the mechanism of platelet aggregation.

Comparison of platelet fibrinogen receptors on intact and proteolytically-treated platelets by use of an anti-glycoprotein IIIa monoclonal antibody (MA 123)

A murine monoclonal antibody (MA 123) was selected by screening 153 supernatants of hybridoma cells secreting anti-human platelet antibodies for their ability to inhibit the fibrinogen-induced aggregation of chymotrypsin-treated platelets. MA 123 inhibited the binding of 125I-fibrinogen to ADP-stimulated intact human platelets and to platelets treated with chymotrypsin or pronase. Moreover, it inhibited the fibrinogen-induced aggregation of these platelet suspensions. The degree of inhibition was similar in each of the three types of platelets tested. The interactions of MA 123 with the 125I-labeled surface components of intact and chymotrypsin-treated platelets were studied by immunoprecipitation using Staphylococcus aureus coated with goat anti-mouse IgG, followed by SDS-polyacrylamide gel electrophoresis and autoradiography. MA 123 precipitated the glycoprotein IIb-glycoprotein IIIa (GPIIb-GPIIIa) complex from the surface of detergent solubilized intact human platelets; and it precipitated GPIIIa from the surface of chymotrypsin-treated platelets. Partially purified GPIIIa was also immunoprecipitated by MA 123. Our data suggest that the exposure of fibrinogen receptors by ADP, chymotrypsin or pronase, is associated with alterations of GPIIIa on the platelet surface.

Action mechanism of the potent platelet aggregation inhibitor from Trimeresurus gramineus snake venom

The platelet aggregation inhibitor (PAI) purified from T. gramineus snake venom inhibited platelet aggregation even when it was added after platelet shape change had occurred. It inhibited the initial platelet aggregation phase in rabbit platelet-rich plasma (PRP), but not the second lytic phase induced by thimerosal (0.5 mM), a -SH reagent. In human platelet-rich plasma, the venom inhibitor also inhibited the platelet aggregation induced by ADP, collagen, epinephrine or ristocetin. The epinephrine-induced platelet aggregation was more susceptible to the venom inhibitor. The venom inhibitor (10 micrograms/ml) as well as N-ethylmaleimide (1.2 mM), a -SH reagent, completely inhibited the thrombin-induced clot retraction of PRP. It is concluded that PAI acts on a common step of platelet aggregation. It acts on membrane -SH containing macromolecule which directly mediates the platelet aggregation subsequent to platelet shape change.

Comparative study of the major glycoproteins of the plasma membrane and secretory granule membranes of porcine platelets

The major glycoproteins of porcine platelet plasma membranes and alpha-granule membranes were compared. Significant cross-contamination of the two preparations was ruled out by surface labeling and proteolysis experiments. At least four of the major glycoproteins of the alpha-granule membrane had counterparts in the plasma membranes with identical molecular weight and lectin-binding properties. Two of these (mol. wts 110,000 and 125,000) were further analysed by one-dimensional peptide mapping. The results confirmed that there are two distinct pools of identical glycoproteins: one on the surface membrane and the other on the alpha-granular membrane. The 110,000 and 125,000 mol. wt glycoproteins are probably equivalent to glycoproteins IIb and IIIa of the human platelet and may therefore be involved in fibrinogen binding.

Crosslinking of platelet glycoprotein Ib by N-succinimidyl(4-azidophenyldithio)propionate and 3,3'-dithiobis(sulfosuccinimidyl propionate)

To examine the relationship between glycoprotein Ib and other proteins in the platelet membrane and the interaction of this protein with thrombin, platelets were crosslinked by two cleavable reagents, SADP (N-succinimidyl(4-azidophenyldithio)propionate) and DTSSP (3,3'-dithiobis(sulfosuccinimidyl propionate]. Two-dimensional, unreduced-reduced sodium dodecyl sulphate (SDS)-polyacrylamide electrophoresis and staining by silver or wheat germ agglutinin-conjugated peroxidase, after protein transfer to nitrocellulose, demonstrated that SADP intramolecularly crosslinked glycoprotein Ib and formed intermolecular complexes of glycorprotein IIb and some high molecular weight proteins. DTSSP intermolecularly crosslinked glycoprotein Ib, glycoprotein IIb, and other high molecular weight proteins. With a low concentration of 125I-labeled TLCK-thrombin (6 nM), crosslinking with SADP yielded a 200 000 Da complex containing radioactive-labeled thrombin, and high TLCK-thrombin concentration (0.1 microM) gave the complex and a 167 000 band. alpha- and TLCK-thrombin crosslinking with DTSSP also yielded the 200 000 complex formed by reaction with SADP or DTSSP was markedly reduced by preincubation of platelets with excess unlabeled TLCK-thrombin and had a pI similar to glycoprotein Ib. These results suggest that glycoprotein Ib is one of the proteins composing the high affinity receptor for thrombin.

Monoclonal antibodies against human platelet glycoprotein IIIa

Two murine monoclonal antibodies specific for human platelets were prepared and characterized by immunofluorescence, immunoprecipitation and by studying their effect on platelet function. Immunoprecipitation with lysates of normal platelets and platelets from a patient with Glanzmann's thrombasthenia revealed that the monoclonal antibodies were directed against glycoprotein GP IIIa. One of these anti-GP-IIIa antibodies (C17) inhibited both ADP- and collagen-induced platelet aggregation as well as ADP-induced fibrinogen binding to platelets. The other anti-GP-IIIa antibody (C15) also caused a complete inhibition of aggregation with collagen. However, a small, and fully reversible, 'primary wave' was observed if the platelets were stimulated with ADP when platelet-rich plasma was used. The ability to bind fibrinogen was unimpaired for platelets incubated with C15. These findings show that C15 and C17 probably recognize different determinants on GP IIIa. Neither of the monoclonal anti-GP-IIIa antibodies blocked the binding to Zwa-positive platelets of human polyclonal anti-Zwa antibodies. This implies that Zwa is different from the epitopes recognized by C15 and C17.

Purification of human platelet membrane glycoproteins IIb and IIIa using high-performance liquid chromatography gel filtration

A method for purifying platelet membrane glycoproteins IIb and IIIa to homogeneity has been developed. The procedure involves high-pressure gel filtration chromatography using a TSK-4000SW column in the presence of sodium dodecyl sulfate. This technique is capable of rapidly preparing milligram quantities of each glycoprotein with greater than 90% recovery. The use of this technique should aid in defining the structural and functional properties of GPIIb and GPIIIa.

Fibrinogen interaction with platelet receptors

In summary: Incubation of platelets with ADP or proteolytic enzymes (chymotrypsin or pronase) results in an exposure of two classes of specific binding sites on platelet surface: low and high affinity fibrinogen receptors. Fibrinogen interaction with these receptors results in platelet aggregation. High affinity fibrinogen receptors are not exposed on thrombasthenic platelets stimulated by ADP but are rendered available on chymotrypsin-treated thrombasthenic platelets; low affinity receptors cannot be exposed by ADP or chymotrypsin on these platelets. Availability of high affinity fibrinogen receptors on thrombasthenic platelets may depend on the residual glycoprotein IIIa. Fibrinogen receptors appear to be associated with glycoproteins IIb, IIIa and a 66,000 Mr platelet membrane component that is exposed during proteolysis of platelet membranes. Some of the platelet-binding sites on the fibrinogen molecule appear to be associated with the COOH-terminal portion of the gamma chain (gamma 374-411). Additional binding sites may also be located in the COOH-terminal portion of the A alpha chain. The conformation of the fibrinogen molecule may be important in its interaction with platelets. Platelet aggregation may result from bridging platelets by fibrinogen molecule in the presence of bivalent cations. In conclusion, platelet interaction with fibrinogen is a complex process involving different binding sites of the fibrinogen molecule. Our own data and review of literature suggest that platelet-interaction with fibrinogen is of major significance in hemostasis.