SummaryThrombopoietin is produced at a constant rate by the liver and kidney and is removed from the circulation upon binding and subsequent uptake via the Tpo receptor, c-Mpl, expressed by platelets and megakaryocytes. Apart from uptake, this study shows that platelets can also function as a storage pool for Tpo.Upon stimulation with various platelet agonists, full-length biologically active Tpo was released by platelets. Platelet fractionation experiments indicated that this Tpo most likely is contained in the granules. When platelets were preincubated with Tpo-peptide mimetic or truncated Tpo prior to maximal activation, a three- to fivefold increment in Tpo release was seen, whereas, the release of other granule proteins such as vWF-propeptide or serotonin remained unchanged. Therefore, the Mpl agonists might compete with Mpl-bound Tpo, thereby releasing Tpo into the platelet supernatant.Intravascular release of Tpo by platelets might occur in patients with massive platelet activation, as occurs in patients with disseminated intravascular coagulation. The Tpo concentration in these patients is elevated (p <0.01) and correlates with markers for thrombin generation, TAT complexes and Fl+2 (rp =0.8 and 0.9; p <0.01). This suggests that the increment in Tpo concentration was attributed to Tpo release by activated platelets in vivo, which might be instrumental in subsequent stimulation of thrombocytopoiesis.

A monoclonal antibody directed against the von Willebrand factor moiety (vWF) of factor VIII-von Willebrand factor (FVIII-vWF), which blocks ristocetin-induced platelet aggregation as well as the binding of FVIII- vWF to platelets in the presence of ristocetin, inhibited platelet adherence to human artery subendothelium when present in normal flowing blood. This monoclonal antibody, CLB-RAg 35, inhibited platelet adherence as a function of the shear rate. At wall shear rates below 500 s-1, platelet adherence was not affected, but at higher shear rates platelet adherence was gradually inhibited, reaching an average of 11% of the normal value at 2,500 s-1. Indirect immunofluorescence established the reactivity of CLB-RAg 35 with vWF present in artery subendothelium. Pretreatment of normal vessel walls with this antibody inhibited adherence of platelets in blood from a patient with severe homozygous von Willebrand's disease and in blood from normal individuals. The inhibition was shear-rate dependent and significant at high shear rates (2,500 s-1). By adding increasing amounts of purified FVIII-vWF to normal blood, the inhibition was gradually overcome. These data indicate that vWF present in the vessel wall contributes appreciably to platelet adherence. At high wall shear rates, platelet adherence is mediated virtually completely by both plasma FVIII-vWF and vWF in the vessel wall. At low wall shear rates (below 500 s-1), platelet adherence occurs independent of FVIII-vWF in plasma and vWF in the vessel wall.

Pretreatment of endothelial cells with cytokines enhances the adherence of leukocytes, a process that is mediated by surface proteins expressed on both cell types. A three-dimensional model system for the simultaneous determination of leukocyte adherence and migration was used to study the contribution of CD11/CD18, endothelial leukocyte- adhesion molecule-1 (ELAM-1) and VLA-4 in neutrophil and monocyte adherence to and migration through cytokine-activated endothelial cells. Pretreatment of endothelial cells for 4 hours with recombinant interleukin-1 beta (rIL-1 beta) was found to enhance neutrophil adherence and migration to a much greater extent than monocyte adherence and migration. Neutrophil adherence was almost completely prevented by the combined use of monoclonal antibodies (MoAbs) against ELAM-1 and CD18. Although ELAM-1 has been designated an endothelial cell-specific cytokine-inducible receptor for neutrophils, we observed that ENA2, an anti-ELAM-1 MoAb, significantly reduced monocyte adherence about 30%. MoAbs against VLA-4, the ligand of the cytokine- inducible receptor VCAM-1, did not affect monocyte adherence. However, the combined use of the MoAbs against CD18, ELAM-1, and VLA-4 had a very strong and additive inhibitory effect on rIL-1 beta-induced monocyte adherence. The anti-CD18 MoAb reduced both rIL-1 beta-induced neutrophil and monocyte migration far below the level of the unstimulated controls, whereas neither the anti-ELAM-1 nor the anti-VLA- 4 MoAb significantly affected the process of migration. Our results indicate that neutrophils and monocytes initially adhere to cytokine- activated endothelial cells by CD18-independent and (to a lesser extent) by CD18-dependent mechanisms and subsequently change gears to a completely CD18-dependent migratory mechanism.

In this report we show that human endothelial cells (EC) can be detected in circulating blood by means of the EC-specific monoclonal antibody (MoAb) designated as CLB-HEC 19 and expressed quantitatively as number of cells per milliliter of whole blood. We first developed a method that was able to recover cultured human EC added to whole blood by Percoll density gradient centrifugation. The final recovery of the EC was 91.6% (SE = 0.65%). The EC were identified in the gradient subfractions by indirect immunofluorescence with the MoAb CLB-HEC 19. This method was then applied to the separation and characterization of EC or EC remnants from the whole arterial and venous blood taken from two groups of patients subjected to heart catheterization. Firstly, a preliminary blood screening of random samples was performed in a group of eight patients (group I) using a scoring evaluation for the presence of EC and the results were expressed as positivity index. Secondly, the complete blood screening of a group of ten patients (group II) was performed for the detection of immunofluorescent cells and the results were expressed as number of EC per milliliter of whole blood. Our results show in both group I and II a significant presence of EC in the blood after catheterization compared with their basal values. The minimal detectable concentration of EC was 0.06 cells/mL (SE = 0.057) of whole blood. We consider this technique as a suitable clinical test for the detection of EC injury in cardiovascular pathology.

Endothelial cells express surface molecules that are involved in cell- matrix interaction, including the vitronectin receptor and the fibronectin receptor, both members of a family of cell adhesion receptors (integrins). Here we provide evidence that endothelial cells express a membrane molecule, indistinguishable from the platelet VLA-2 complex, which is a collagen receptor and a member of the integrin family. To identify this endothelial molecule, we have used a monoclonal antibody, CLB-10G11, which recognizes the VLA-2 complex from platelets. The molecule recognized by CLB-10G11 from endothelial cells was characterized as follows. (1) The monoclonal antibody precipitated two proteins from surface-labeled endothelial cells that corresponded to the platelet VLA-2 subunits (glycoprotein Ia and IIa) as judged by one-dimensional sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and two-dimensional nonreduced/reduced SDS- PAGE. (2) Preclearing of endothelial cells with monoclonal antibody A- 1A5, an antibody that is directed against the common VLA beta subunit, removed all the CLB-10G11-binding material. (3) Crossed immunoelectrophoresis revealed that CLB-10G11 recognizes a single precipitation arc from either platelets or endothelial cells. Analysis of these two cell types in one gel again revealed one precipitation arc. The antigen of either cell type, recognized by CLB-10G11 could be precipitated by either polyclonal antiplatelet or polyclonal antiendothelial cell antiserum. Hence, it appears that endothelial cells express at least three different surface molecules (the vitronectin receptor, the fibronectin receptor and a collagen receptor), which may play an important role in controlling the anchorage of endothelial cells to the extracellular matrix.

We have previously demonstrated that endothelial cells synthesize a plasma membrane protein indistinguishable from platelet glycoprotein (GP) IIa. The present study provides evidence for a further analogy between the platelet and the endothelial cell membrane by showing that cultured endothelial cells also synthesize a membrane protein complex immunologically related to the platelet GP IIb/GP IIIa complex. This evidence is based on the following observations: (1) C17, a murine monoclonal antiplatelet GP IIIa antibody, consistently precipitates two proteins, apparent molecular weights, respectively, 115,000 and 125,000 reduced and 95,000 and 135,000 nonreduced, from metabolically (35S- methionine) as well as surface 125I-labeled cultured human endothelial cells; (2) upon crossed immunoelectrophoresis of solubilized endothelial cells against a polyclonal rabbit antiplatelet antiserum and 125I-labeled C17 IgG, a single precipitate of the protein(s) recognized by C17 is observed. As judged by their mobility in 9% polyacrylamide gels, both endothelial proteins appear to have a somewhat larger molecular weight than their platelet counterparts. Patterns obtained by crossed immunoelectrophoresis are also indicative of a difference in electrophoretic behavior of the platelet GP IIb/IIIa complex and the endothelial cell protein complex.

Antiphospholipid antibodies (APA) have thought to be implicated in the pathogenesis of both arterial and venous thrombosis. Because of heterogeneity of APA, direct evidence of their involvement in a thrombotic event is not yet available. Development of thrombosis in the antiphospholipid antibody syndrome (APS) may occur because of the presence of additional risk factors. Here we have analysed 60 patients with APA for the presence of the Arg506-->Gln mutation in factor V. Among them 26 suffered from deep venous thrombosis, 13 from arterial thrombosis and 21 had no history of arterial or venous thrombosis. In the first group four patients were found to be heterozygous and one homozygous for the factor V Arg506-->Gln mutation. None of the patients with the factor V mutation was found in the second and third group. The incidence of factor V mutation was significantly elevated in the group of patients with venous thrombosis. These data suggest that in patients with antiphospholipid antibodies the factor V Arg506-->Gln mutation may play a major role in the occurrence of venous thrombosis.

We have studied the role of factor VIII-von Willebrand factor (FVIII- vWF) in both platelet adherence to subendothelium and ristocetin- induced platelet aggregation using monoclonal antibodies to human FVIII- vWF. Twenty-five monoclonal antibodies were obtained, two of which were directed to the factor VIII moiety of FVIII-vWF; one of these two completely inhibited the procoagulant activity (FVIII:C). The remaining 23 monoclonal antibodies were directed to the von Willebrand factor moiety of FVIII-vWF. The ability of the latter monoclonal antibodies to inhibit platelet adherence to arterial subendothelium was investigated with a perfusion model. According to the number of platelets adhering to the subendothelium, three groups of monoclonal antibodies could be discerned: (A) antibodies not affecting platelet adherence; (B) antibodies that inhibited platelet adherence to the level as observed when von Willebrand's disease plasma was tested; and (C) antibodies that completely inhibited both platelet adherence to subendothelium and ristocetin-induced platelet aggregation. The two antibodies present in group C competed for the same or closely related epitope(s) present on FVIII-vWF. These results demonstrate that a domain is present on the FVIII-vWF molecule that is associated both with ristocetin-induced aggregation and with the ability of FVIII-vWF to support platelet adherence to the subendothelium. Based on these observations, it is concluded that ristocetin-induced binding of FVIII-vWF to platelets reflects, at least in part, a physiologic mechanism regulating the function of FVIII-vWF in primary hemostasis.