Factor VIII influences binding of factor IX and factor X to intact human platelets

To investigate the influence of factor VIII on the binding of factors IX and X to the surface of intact human platelets, washed collagen-stimulated platelets were incubated with factor IX and factor X in the presence or absence of factor VIII. Platelets were then lysed and IX:Ag and X:Ag were determined in the platelet lysate. Platelet bound IX:Ag was higher after incubation of platelets with factor IX in the presence of native or activated factor VIII than after incubation of platelets with factor IX alone. Thrombin degraded factor VIII did not support binding of factor IX to stimulated platelets. Preactivation of factor IX enhanced binding to platelets; influence by factor VIII on binding to platelets was greater with native factor IX than with preactivated factor IX. Presence of native or activated factor VIII also increased binding of factor X to platelets. In contrast to factor IX, preactivation of factor X did not influence binding to platelets, but in the presence of factor VIII preactivation of factor X increased binding of factor X to platelets. Our data suggest that mediation of binding of factors IX and X to the surface of platelets is one of the mechanisms by that factor VIII exerts its cofactor function in the activation of factor X.     

Effect of carbohydrate modifications of factor VIII/von Willebrand factor on binding to platelets

This study compares the ability of unmodified and carbohydrate-modified forms of factor VIII/von Willebrand factor (FVIII/vWF) protein to bind to platelets in the presence of ristocetin or thrombin. Treatment of intact FVIII/vWF with alpha-D-neuraminidase results in more than 95% desialylation. Asialo FVIII/vWF retains total activity in ristocetin- and thrombin-mediated binding to platelets as demonstrated by direct and competitive binding assays. Examination of its multimeric pattern by sodium dodecyl sulfate-agarose electrophoresis reveals a normal multimeric structure. Treatment of intact FVIII/vWF with beta-D-galactosidase results in the removal of 20% of galactose (agalacto FVIII/vWF) whereas 55% of galactose is released from asialo FVIII/vWF (asialo agalacto FVIII/vWF). Agalacto and asialo-agalacto FVIII/vWF are both unable to bind to platelets in the presence of ristocetin. In contrast, they still bind to thrombin-stimulated human (except thrombasthenic) platelets. Removal of either ultimate (agalacto FVIII/vWF) or ultimate and penultimate (asialo-agalacto FVIII/vWF) galactose results in the same loss of the larger molecular weight multimers and in an increase of smaller multimers. These results suggest (1) that sialic acid does not play a significant role in ristocetin- or thrombin-mediated FVIII/vWF-platelets interactions and multimeric structure of FVIII/vWF (2) that ultimate beta-linked galactose residues are essential for the maintenance of a normal multimer organization (3) that ristocetin- and thrombin-mediated binding of FVIII/vWF to platelets differ in FVIII/vWF galactose requirement.     

Inhibition by ticlopidine of Paf-acether-induced in vitro aggregation of rabbit and human platelets

Ticlopidine was incubated $\text{in vitro}$ with rabbit or human washed platelets and aggregations were triggered by submaximal concentrations of adenosine-5′-diphosphate (ADP), arachidonic acid (AA) and Paf-acether (platelet-activating factor), the mediators of the three known pathways of platelet activation. Inhibition of Paf-acether-induced rabbit platelet aggregation was proportionnal to the concentrations of Ticlopidine used. The same range of inhibition by Ticlopidine was observed when aggregations were triggered by the two other agonists. Human platelet aggregation induced by Paf-acether was also inhibited by Ticlopidine. Inhibition was increased when platelets were rendered insensitive to ADP and AA. Our results show that Ticlopidine inhibits human and rabbit platelet aggregation triggered by Paf-acether through a mechanism not related to the inhibition of the ADP and prostaglandin pathways.     

Reversibility of fibrinogen fragment D1 binding to human platelets: comparison with native fibrinogen.

To gain further insight into the mechanism responsible for rendering fibrinogen bound to stimulated platelets irreversible to dissociation by EDTA or excess unlabeled fibrinogen, the present study compared the reversibility of platelet interactions with fibrinogen and its plasmic degradation product, fragment D1. Like fibrinogen binding, the binding of fragment D1 became progressively less sensitive to dissociation by EDTA, PGE1, or excess unlabeled fibrinogen. Thus in the presence of EDTA, 70 +/- 19% and 55 +/- 24% (mean +/- S.D., n = 9) of bound fragment D1 failed to dissociate from platelets 60 min after stimulation with 0.15 U/ml thrombin or the combination of 5 microM ADP and 5 microM epinephrine, respectively, compared to 75 +/- 8% and 52 +/- 17% of platelet-bound, intact fibrinogen. In contrast, platelet stimulation with chymotrypsin or Zn+2 failed to support the development of irreversible fragment D1 or fibrinogen binding. Only 8 +/- 6% and 9 +/- 3% of bound fragment D1 remained associated with chymotrypsin- or Zn+2-treated platelets, respectively, compared to 7 +/- 11% and 15 +/- 6% (mean +/- S.D., n = 3) of platelet-associated fibrinogen. These observations suggest that irreversible fragment D1 and fibrinogen binding to platelets occurs by a similar mechanism that requires neither fibrinogen alpha chain 95-97 or 572-574 RGD sequences nor multivalent ligand-receptor interactions.     

Studies on the factor VIII/von Willebrand factor antigen on human platelets

A purified protein having both Factor VIII activity (coagulant assay) and von Willebrand factor activity (platelet retention and ristocetin aggregation assays) was used to immunize a goat. The resulting antiserum neutralized Factor VIII coagulant activity, decreased platelet retention of normal blood and blocked ristocetin aggregation of normal platelet rich plasma.

This same antiserum acted as an “anti-platelet” antibody in serotonin release, platelet aggregation and immunofluorescence with normal, hemophilic and von Willebrand platelets. However, after suitable absorption with small numbers of platelets this antiserum recognized Factor VIII/von Willebrand factor antigen only on normal and hemophilic platelets but not on platelets from two patients with severe von Willebrand's disease.

It is possible that antisera produced against Factor VIII purified from plasma not rendered completely free of platelets contains antibodies to platelet membrane material.     

Characterization of calcium-dependent binding of endogenous factor VIII/von Willebrand factor to surface activated platelets

Activation of washed platelets in the presence of EDTA with either 1 U/ml of alpha-thrombin or 2 microM calcium ionophore (A23187) caused the release of one-third to one-half of the platelet factor VIII/von Willebrand factor (FVIII/vWF) into the supernatant. When calcium was present in excess, only 10% of the platelet FVIII/vWF was detected free in the supernatant, regardless of whether calcium was present before stimulation or added to the platelets after thrombin activation. Release of [14C]serotonin and beta-thromboglobulin were not affected by divalent cations indicating that reduced supernatant levels of FVIII/vWF in the presence of calcium were not due to differential release, but were probably due to a calcium-dependent association of released FVIII/vWF with the platelet surface. The presence or absence of intact glycoprotein Ib on the platelet surface made no significant difference to the observed FVIII/vWF partition. Platelets from a patient with Glanzmann's thrombasthenia, however, failed to show a calcium effect with respect to released FVIII/vWF. The combined results suggest that as well as the ristocetin-dependent, divalent cation-independent binding of FVIII/vWF to glycoprotein Ib, there is a divalent cation-dependent binding of FVIII/vWF to the activated platelet surface which is mediated via the glycoprotein IIb/IIIa complex.     

Subcellular distribution of fibrinogen and factor XIII in human blood platelets.

Fresh human platelets were disrupted by nitrogen decompression, and homogenates were fractionated by centrifugation in a linear gradient of sodium diatrizoate (18–33%). Five distinct fractions were obtained. Electron microscopy showed that the membrane fraction was free of contaminating subcellular organelles. The granule fraction contained primarily intact granules, with some mitochondria and some swollen granules. A portion of the granules was also distributed in the fractions above and below the primary granule fraction. Total protein was found to be 1.95 mg/10⁹ platelets, of which 0.18 mg was identified as fibrinogen. The ratio of platelet to plasma fibrinogen in whole blood was determined to be 1:30. Total factor XIII activity was 69 μmole dansylcadaverine incorporation/10⁹ platelets, and factor XIII activity was found to be equally distributed between plasma and platelets. 70% of the platelet fibrinogen was found in association with granules, with the remainder appearing primarily in the soluble fraction. Platelet factor XIII was localized in the cytoplasm, with 94% appearing in the soluble fraction. Platelet fibrinogen was released during the platelet release reaction induced by collagen, whereas factor XIII was not.     

C-reactive protein inhibits binding of platelet-activating factor to human platelets

Serum concentration of C-reactive protein (CRP), a prototypical acute-phase protein rises dramatically in response to tissue injury or inflammation. We report here that CRP (1-20 micrograms/ml) inhibited platelet-activating factor (PAF)-induced aggregation of human platelets in time-, and dose-dependent manner. This inhibitory action of CRP was nearly completely removed by treatment with anti CRP antiserum. At higher concentrations (20-100 micrograms/ml), CRP stabilized platelet membrane against the detergent-like effect of beta-deoxy-lysolecithin. Furthermore, CRP (10 micrograms/ml) diminished specific [3H]PAF binding to platelets and displaced previously bound labeled PAF from platelets. These results suggest that by depressing the bioavailability of PAF, CRP may be an important modulator of platelet activation during acute inflammatory reactions.     

Proaggregatory effect of epinephrine on rabbit platelets inhibited by ticlopidine

Ticlopidine is a potent inhibitor of ADP-induced aggregation of rabbit platelets ex vivo. In vivo, however, multiple agonists play a role in platelet activation. In this study, we examined the effect of epinephrine on the antiplatelet action of ticlopidine in rabbit platelets. Epinephrine reversed the inhibitory effect of drug on ADP-induced platelet aggregation. The potentiating effect of epinephrine was mediated through alpha 2-adrenergic receptors, was reversed by pretreatment with the Na+/H+ exchange inhibitor dimethylamiloride, and was mimicked by agents that increased intracellular sodium or pH. Ticlopidine had no effect on resting intracellular pH, an indication that the effect of epinephrine was not compensating for a drug-induced intracellular acidification. While this potentiation was also found to be inhibited by aspirin, it did not involve enhanced release of thromboxane A2. Our results demonstrate that epinephrine can overcome the inhibitory effect of ticlopidine on ADP-induced aggregation through a mechanism involving activation of Na+/H+ exchange and through an as yet unidentified mechanism sensitive to aspirin.     

Binding of native and thrombin activated factor VIII to platelets

The interaction of human Factor VIII with platelets has been studied using discontinuous albumin gradient centrifugation. When purified Factor VIII complex was mixed with released platelets about 10% of the recovered Factor VIII activity became associated with the platelets. In the corresponding experiment with thrombin-activated Factor VIII about half of the Factor VIII activity was bound to the platelets. This binding was reversible; dilution of the platelet suspension containing bound Factor VIII resulted in dissociation of part of the Factor VIII activity. With non-released platelets very little binding of Factor VIII activity occurred. A mechanism for the in vivo formation of the Factor X activator complex is suggested.     

The fibrinogen gamma chain dodecapeptide inhibits agonist-induced aggregation of rabbit platelets and fibrinogen binding to rabbit glycoprotein IIb-IIIa

The HHLGGAKQAGDV (H12) sequence at the carboxyl termini of the y chains and the RGD sequences in the Aalpha chains of human fibrinogen are potential recognition sites for the binding of soluble fibrinogen to glycoprotein IIb-IIIa (GPIIb-IIIa) on activated human platelets. Thus, addition of either H12 or RGD-containing peptides inhibits aggregation of and fibrinogen binding to human platelets. In contrast, we reported previously that RGDS had relatively little inhibitory effect on these functions of rabbit platelets. In the present study, we found that H12 inhibited ADP- and thrombin-induced aggregation of rabbit platelets in a dose-dependent manner. Specificity was demonstrated by the failure of the variant HHLGGAKQAGEV peptide to inhibit ADP-induced aggregation. Furthermore, flow cytometric analyses demonstrated that H12 inhibited the binding of FITC-fibrinogen to ADP-activated rabbit platelets in a dose-dependent manner. To examine the direct interaction of H12 with rabbit GPIIb-IIIa, we performed affinity chromatography by applying an octylglucoside extract of rabbit platelet proteins onto an affinity matrix containing the fibrinogen gamma chain sequence. Proteins of approximately 135 kDa and approximately 95 kDa were specifically eluted by soluble H12, and the 95 kDa protein band was immunoblotted by anti-LIBS1, a monoclonal antibody against human GPIIIa. In control samples, no detectable protein from rabbit platelet lysates was eluted from an RGD affinity matrix by GRGDSP. Collectively, our results demonstrated that H12 inhibits aggregation of and fibrinogen binding to rabbit platelets by directly interacting with rabbit GPIIb-IIIa. These findings suggest that rabbit platelets would serve as a suitable thrombosis model for testing the efficacy of peptide mimetics derived from H12.     

The effect of ticlopidine on human endothelial cells in culture

The effect of ticlopidine at various concentrations (150, 30, 6 microM), has been studied on cultured endothelial cells derived from human umbilical cord vein. Ticlopidine affects the initial attachment of endothelial cells to artificial substrata and has an inhibitory effect on endothelial cell growth rate which correlates to the concentration of the chemical in the culture medium. These effects are related to a marked reduction of intra- and extracellular fibronectin as evidentiated by immunofluorescence. The drug seems to interfere with the formation of fibronectin filaments from intracellular granules. The reduction of fibronectin availability could affect platelet adhesion to subendothelium as well as endothelial cell repair, and subsequently influence the bleeding time. The inhibition of cell proliferation and its possible effect on the thickness of the vessel wall should be considered as additional mechanisms of action for this substance.     

The identification and functional significance of factor VIII components on normal platelets

Factor VIII-related antigen (VIII R:Ag) has previously been identified on human platelets by fluorescein-labelled heterologous antibodies to factor VIII. However, it is not known whether these antibodies identify only VIII R:Ag or whether antigens associated with factor VIII coagulant activity (VIII:C Ag) are also present. The current studies utilize specific antibodies and their Fab' fragments to evaluate the presence of both VIII R:Ag and VIII:C Ag on human platelets and to study the functional role of the platelet-associated VIII R:Ag on ristocetin-induced platelet agglutination. Immunofluorescent studies demonstrate the presence of VIII R:Ag, but do not identify VIII:C Ag on platelets. The functional studies reveal that after incubation with either of the Fab' fragments, platelets retain their reactivity in the ristocetin cofactor assay. These results indicate that this population of VIII R:Ag is not essential for the ristocetin aggregation reaction and that no antigenic determinants for factor VIII coagulant activity (VIII:C) are detectable on human platelets by immunofluorescence.     

The effect of phospholipase C on platelet factor 3 in human blood platelets

Intact human platelets isolated by gel filtration have been treated with purified phospholipase C. The effect of the enzyme on available and total platelet factor 3 has been tested. The available procoagulant platelet factor 3 was very low. A further small reduction was observed after incubation with phospholipase C when the enzyme was washed away before testing. External attack on platelets by phospholipase C led to a marked inactivation of total platelet factor 3.     

Australian snake venoms and their in vitro effect on human platelets

Thrombocytopenia is generally not associated with cases of envenomation by Australian snakes, however the clinical evidence is conflicting. The in vitro effect of these venoms upon platelets had hitherto not been studied. This study systematically examines the effect on human fresh and fixed platelets of twenty Australian snake venoms, nineteen elapid and an hydrophiid; for comparision four crotalid venoms from the Americas and S.E. Asia were also included. Electron micrographs were taken of platelets after exposure to some of the venoms. Results demonstrated that all venoms except the hydrophiid venom caused fresh platelets to irreversibly aggregate directly, and this was associated with degranulation as evidenced by electron microscopy (EM). Response to all venoms by fixed platelets was less marked and also suggests, that metabolically, active platelets are necessary for the venoms to exert their maximal effect. The hydrophiid venom's action on fresh platelets was unique, as a plasma co-factor was required before aggregation could be induced. Crotalid and hydrophiid venoms were more active against platelets than the elapid venoms. Nevertheless, platelet aggregation and degranulation in the presence of elapid venoms suggests that a platelet response in vitro may be a significant factor in the "defibrination syndrome" induced in humans by Australian snakes.     

Effects of proteolytic degradation products of human fibrinogen and of human factor VIII on platelet aggregation and vascular permeability

Products of proteolysis of highly purified preparations of human fibrinogen (FDP) and of human factor VIII (VIII-DP) by plasmin were compared with respect to their ability to inhibit platelet aggregation and to enhance the permeability of rat skin microvasculature. Low molecular weight dialysable FDP (LMW-FDP, m. w. under 2754) induced complete inhibition of platelet aggregation in PRP at concentrations of and above 2. 5 mg/ml. Slight or very pronounced increase in skin microvasculature permeability was observed at LMW-FDP doses of 10 μg and 78 μg, respectively. In contrast, the whole digest of factor VIII as well as LMW-VIII-DP appeared inactive. Unfractionated digests of fibrinogen and fibrin formed either by thrombin of Defibrase exhibited the same antiaggregating potency.     

Inhibition of platelet-activating factor binding to human platelets by calcium channel blockers

Calcium channel blockers may impair cell activation either by inhibiting calcium influx or by inhibiting agonist binding. Because of this dual action of calcium channel blockers and because of the close relationship between calcium influx and platelet-activating factor (PAF) binding to platelets the current studies examined the effect of calcium channel blockers on PAF binding to washed human platelets. Diltiazem and verapamil inhibited aggregation by PAF in a dose-dependent manner with 50% inhibition at 2.8 +/- 1.4 X 10(-5) M diltiazem (mean +/- SD, n = 5) and 4.2 +/- 2.0 X 10(-5) M verapamil. Both channel blockers also inhibited PAF binding in a dose-dependent manner with 50% inhibition at 4.7 +/- 2.5 X 10(-5) M diltiazem and 6.3 +/- 1.2 X 10(-5) M verapamil. Analysis of the mechanisms of inhibition of binding indicate both competitive and non-competitive effects of the channel blockers. Scatchard analysis of PAF binding in the presence of different fixed concentrations of either diltiazem or verapamil revealed that these agents both increased PAF receptor number and decreased the receptor binding affinity. Lineweaver-Burke analysis of the same data revealed a family of lines which intersect to the right of the ordinate. The channel blockers also dissociated previously-bound PAF from platelets. The current studies indicate that calcium channel blockers inhibit platelet activation by PAF by more than one mechanism and suggest that the PAF receptor may be closely associated with calcium channels.     

The action of immobilized thrombin on factor VIII, fibrinogen and a synthetic tripeptide

Bovine thrombin was insolubilized by attachment to cyanogen bromide-activated Sepharose (Sepharose-thrombin) or to activated (Affi-Gel 10) agarose containing a 10 A long arm (Affi-Gel-thrombin). Coupling in both instances approximated 7,000 units of thrombin per ml packed gel as determined by 125I-thrombin incorporation. The thrombin beads hydrolyzed the synthetic tripeptide Bz-Phe-Val-Arg-pNA (S-2160) at different rates, with the Sepharose-thrombin more active (220 esterase units per ml) than Affi-Gel thrombin (20.4 units per ml). The Km was significantly higher for the insolubilized thrombins (2 X 10(-3) M) than uncoupled thrombin (Km = 8 X 10(-5) M). The Sepharose-thrombin activated factor VIII significantly more rapidly than Affi-Gel-thrombin. Neither matrix-bound thrombin clotted a fibrinogen solution or liberated significant amounts of fibrinopeptides over 48 hr. This data indicates that a proteolysis of factor VIII, rather than a complex with thrombin, is the method of activation of factor VIII and that factor VIII is more accessible to the action of immobilized thrombin than is fibrinogen.     

Monoclonal antibody binding to factor VIII:c

The interactions of human blood clotting Factor VIII:c with 2 monoclonal antibodies, C7F7 and BD10, were examined with an enzyme-linked immunosorbent assay (ELISA) method. A Scatchard-Sips plot approach to data analysis was used to calculate an average affinity (Ko), valence (n), and heterogeneity index (a) for each interaction. Using a Factor VIII:c coating ELISA method, the following constants were determined: BD10: Ko = 5.48 X 10(8) M-1; n = 0.31; a = 0.97; C7F7: Ko = 2.3 X 10(11) M-1; n = 0.018; a = 0.77.