Search

IgG, IgA, IgM, and albumin are primarily known as plasma proteins. Their presence in platelets is poorly understood. The total platelet content of IgG, IgA, and albumin, measured in solubilized platelets by an enzyme-linked immunosorbent-assay (ELISA) technique, was greater than 90% secreted after stimulation by thrombin, consistent with an alpha-granule location. The platelet concentrations of these proteins correlated with their plasma concentrations in normal subjects and over a wide range of abnormalities in patients with IgG or IgA myeloma or liver cirrhosis. IgM was not detectable in normal platelets but was measurable and related to the plasma IgM concentration in patients with macroglobulinemia. In patients with idiopathic thrombocytopenic purpura (ITP), the platelet concentrations of IgG, IgA, and albumin were all twofold to threefold higher than normal despite normal plasma concentrations. Platelet surface IgG, measured by 125I-monoclonal antibody binding, constituted less than 1% of the total platelet IgG, and it appeared to be a pool distinct from the alpha-granule IgG since its concentration in normal subjects and patients did not correlate with either plasma or total platelet IgG concentrations. These observations are consistent with hypotheses that megakaryocytes incorporate plasma proteins into developing alpha-granules by pinocytosis and that the increased ratio of platelet to plasma of IgG, IgA, and albumin in ITP may reflect a younger average age of these platelets.

Methods have been developed to isolate human platelet membrane fragments from plasma and serum. Rabbit antibody produced against the human platelet membrane glycoprotein complex, IIb/IIIa, was utilized in an immunoelectrophoretic assay to evaluate the amount of this antigen in various microparticle preparations. The serum concentration of platelet microparticles was more than tenfold greater than that observed for plasma (65 micrograms/ml versus 4.4 micrograms/ml, respectively). Ultrastructural evaluation of either plasma or serum- derived microparticles disclosed a variety of membrane fragments and membrane-bound vesicles with occasional fragments of red blood cells, white blood cells, and platelets. In contrast, microparticle preparations derived from isolated washed platelets after thrombin stimulation contained a heterogeneous array of membrane fragments, vesicles, and granules but no identifiable red cell, white cell, or platelet fragments. Thus, these studies demonstrate that normal human plasma and serum contain platelet membrane fragments that are produced during cell activation. If a similar loss of platelet membranes occurs in vivo following reversible platelet activation, it is possible that the resulting membrane modifications may be of importance in both the structural and functional changes that develop during platelet senescence.

Thrombin is a physiological agonist that promotes platelet aggregation and secretion. In this study we observed that thrombin can also inhibit a function of platelets related to primary hemostasis. Platelet stimulation by thrombin decreased the binding of von Willebrand factor (vWF) to glycoprotein (GP) Ib and decreased ristocetin-induced agglutination, in vitro reactions that correlate with initial platelet adhesion to the vessel wall. Binding of the monoclonal antibody API to GP Ib was also decreased. Cytoskeletal participation in the change of GP Ib was suggested because pretreatment of platelets with cytochalasin to prevent actin filament formation prevented the thrombin-induced decreases in vWF binding. API binding, and ristocetin-induced agglutination. Measurement of GP Ib in detergent extracts by electroimmunoassay demonstrated no loss after thrombin stimulation. Electroimmunoassay also demonstrated that the API epitope of GP Ib on intact thrombin-treated platelets was accessible for complete digestion by chymotrypsin. Therefore GP Ib was neither released from the platelet surface nor internalized by thrombin treatment. A previously recognized effect of thrombin is its induction of receptor sites on platelet surface GP IIb-IIIa for contact-promoting proteins, including vWF that are involved in the platelet spreading and aggregation that follow adhesion. Therefore the action on GP Ib may combine with the effect on GP IIb-IIIa to shift platelet reactivity from GP Ib-vWF-mediated initial contact with the vessel wall to GP IIb-IIIa-mediated spreading and aggregation.

Washed human platelets in buffers containing either 2 mM Ca++ or 4 mM EDTA were stimulated by human alpha-thrombin to induce secretion. The binding of two endogenous secreted proteins, factor-VIII-related protein (VIII-R) (von Willebrand factor) and platelet factor 4, was measured by reacting thrombin-treated and control platelets with specific antibodies to these proteins, then quantifying antibody binding with 125I-staphylococcal protein A. Both of these granule proteins were associated with the platelet membrane surface by a calcium-dependent mechanism after thrombin-induced secretion. This ability to bind endogenous secreted proteins to the plasma membrane surface may provide a mechanism by which the platelet can concentrate and organize its secreted proteins for subsequent physiologic reactions.

To determine if interleukin-6 (IL-6) administration influences platelet function, platelet activation was analyzed sequentially in IL-6-treated (80 micrograms/kg/d) and control dogs. Platelet activation was determined in whole blood by flow cytometry by quantitating the binding of a monoclonal antibody to platelet surface P-selectin after stimulation with graded doses of thrombin. Administration of IL-6 resulted in a twofold decrease in the thrombin concentration required for induction of half-maximal P-selectin expression (ED50) compared with control animals. The ED50 returned to normal after cessation of IL- 6 administration. As measured by P-selectin expression, enhanced responsiveness to the strong agonist platelet activating factor (PAF) was also observed in the IL-6-treated dogs. IL-6 had no effect on the susceptibility of platelets to thrombin activation when incubated with anticoagulated dog blood. The data show that, in addition to augmenting the platelet count in normal dogs, IL-6 enhances the sensitivity of platelets to activation in response to thrombin and PAF.

After the intravenous infusion of N-hydroxysuccinimido biotin into dogs, 80.6% +/- 9.7% (n = 5) of platelets were covalently labeled with biotin. The in vivo survival of the biotinylated platelets was monitored by flow cytometry and was normal as compared with previous reports for dog platelets. The ability of the biotinylated platelets to be activated was analyzed by measuring the expression of cell-surface P- selectin after incubation with graded concentrations of thrombin. When P-selectin expression was examined 3 hours after labeling, biotinylated platelets were indistinguishable from the nonlabeled population of platelets, indicating that biotinylation did not adversely affect the cells. On consecutive days after biotinylation, the thrombin dose- response curves for biotinylated and nonbiotinylated platelets were repeated, and as the biotinylated-platelets aged, they became less responsive to thrombin. On days 3, 4, and 5, the thrombin EC50 for the aged, biotinylated platelets as compared with the total population of platelets was 136%, 150%, and 178%, respectively. Increasing age clearly impairs the reactivity of platelets towards thrombin as quantitated by the expression of cell-surface P-selectin.

Cryoprecipitate has been demonstrated to correct the bleeding abnormality of patients with some congenital (storage pool disease) and acquired (uremia) platelet abnormalities, but the reason for this effect is unknown. We found significant platelet contamination in plasma harvested to prepare fresh frozen plasma and cryoprecipitate. The platelet membrane microparticles produced by freezing and thawing of the plasma were highly concentrated in cryoprecipitate and may contribute to its therapeutic effect.

Platelet membrane changes that accompany in vivo activation may be difficult to detect if only a small fraction of circulating platelets has undergone secretion. This study describes an approach to that problem by using a method to measure the number of molecules of fluorescein-labeled antibody bound to individual platelets by flow cytometry. The platelet response to different concentrations of thrombin was determined by measuring the binding of a monoclonal antibody (S12) to GMP-140, an alpha-granule membrane protein that becomes exposed on the platelet surface during alpha-granule secretion. Unstimulated platelets bound a mean of 1,120 molecules of S12 per cell, and 93% of platelets bound less than 2,000 molecules. Platelet stimulation by 0.25 U/mL thrombin caused maximum S12 binding with a mean of 7,529 molecules per cell. Even at low concentrations of thrombin (0.025 U/mL), 5% of platelets were maximally activated, binding over 7,000 molecules of S12 per cell. Conversely, at 0.25 U/mL thrombin, 13% of platelets continued to bind less than 2,000 molecules of S12 per cell. A mixture of as little as 5% thrombin-activated platelets with unstimulated platelets could be detected by this method. Therefore flow cytometry offers an important tool for investigating patients who may have circulating activated platelets as part of a disorder predisposing to thrombosis or hemorrhage.