Biosynthesis and processing of platelet GPIIb-IIIa in human megakaryocytes.

Platelet membrane glycoprotein IIb-IIIa forms a calcium-dependent heterodimer and constitutes the fibrinogen receptor on stimulated platelets. GPIIb is a two-chain protein containing disulfide-linked alpha and beta subunits. GPIIIa is a single chain protein. These proteins are synthesized in the bone marrow by megakaryocytes, but the study of their synthesis has been hampered by the difficulty in obtaining enriched population of megakaryocytes in large numbers. To examine the biosynthesis and processing of GPIIb-IIIa, purified human megakaryocytes were isolated from liquid cultures of cryopreserved leukocytes stem cell concentrates from patients with chronic myelogenous leukemia. Immunoprecipitation of [35S]methionine pulse-chase-labeled cell extracts by antibodies specific for the alpha or beta subunits of GPIIb indicated that GPIIb was derived from a precursor of Mr 130,000 that contains the alpha and beta subunits. This precursor was converted to GPIIb with a half-life of 4-5 h. No precursor form of GPIIIa was detected. The glycosylation of GPIIb-IIIa was examined in megakaryocytes by metabolic labeling in the presence of tunicamycin, monensin, or treatment with endoglycosidase H. The polypeptide backbones of the GPIIb and the GPIIIa have molecular masses of 120 and 90 kD, respectively. High-mannose oligosaccharides are added to these polypeptide backbones co-translationally. The GPIIb precursor is then processed with conversion of high-mannose to complex type carbohydrates yielding the mature subunits GPIIb alpha (Mr 116,000) and GPIIb beta (Mr 25,000). No posttranslational processing of GPIIIa was detected.

Differential structural requirements for fibrinogen binding to platelets and to endothelial cells.

The cytoadhesins represent a group of RGD receptors that belongs to the integrin superfamily of adhesion molecules. Members of this cytoadhesin family include the platelet GPIIb-IIIa and the vitronectin receptors. These glycoproteins share the same beta-subunit, which is associated with different alpha subunits to form an alpha/beta heterodimer. In the present study, we have analyzed the fine recognition specificy of the cytoadhesins from platelets and endothelial cells for the adhesive protein, fibrinogen. Two sets of synthetic peptides, RGDX peptides and peptides corresponding to the COOH terminus of the fibrinogen gamma chain, were compared for their structure-function relationships in the two cellular systems. The results indicate that: (a) both RGDX and gamma-chain peptides inhibit the binding of fibrinogen to platelets and endothelial cells; (b) a marked influence of the residue at the COOH- and NH2-terminal positions of each peptide set can be demonstrated on the two types; and (c) RGDX and gamma peptides have differential effects on platelets and endothelial cells with respect to fine structural requirements. These results clearly indicate that while the platelet and endothelial cytoadhesins may interact with similar peptidic sequences, they express a different fine structural recognition.

Interaction between fibrinogen and cultured endothelial cells. Induction of migration and specific binding.

It has been suggested that fibrinogen (fg) or its physiological derivatives influence the motility and growth of endothelial cells (ECs), but direct support for this concept is still lacking. In the present study, the capacity of fg to interact with ECs and induce the migration of ECs was examined. The capacity of fg to induce EC migration was studied by means of a modification of the Boyden chamber technique. fg in the lower compartment of the chamber caused a time- and concentration-dependent migration of ECs across filters. fg present in equal concentrations above and below the filter increased EC migration, but the maximal effect invariably occurred in the presence of a gradient between the lower and the upper compartments. Trypsin or plasmin digestion of fg and preincubation of fg with Fab fragments from specific antibody completely abolished fg-induced EC migration. Dialysis of fg to eliminate small peptides that might contaminate the preparation did not modify fg-induced migration. Plasma obtained from healthy donors induced EC migration, but plasma from an afibrinogenemic patient was completely ineffective. The addition of purified fg to afibrinogenemic plasma restored plasma-induced EC migration. Plasmin degradation fragments D and E, of 100,000 and 50,000 mol wt, respectively, did not induce EC migration. However, fragment E caused dose-related inhibition of fg-induced EC migration Direct interaction of highly purified radioiodinated human fg with cultured human and bovine Ecs was observed. The binding was time dependent and plateaued at 10 min. Nonlabeled fg in a large molar excess inhibited the interaction, but unrelated proteins, including fibronectin, ovalbumin, and myoglobin, did not. Monospecific Fab fragments directed to fg inhibited binding by 38% at a 50 to 1 molar ratio whereas nonimmune Fab caused only 2% inhibition at a similar concentration. The binding of 125I-fg with ECs was saturable, and an apparent dissociation constant of 0.23 x 10(-6) M was estimated from binding isotherms. After 30 min of incubation the interaction between 125I-fg and the cells was completely reversible and displaceable by a large molar excess of unlabeled fg. Autoradiography of the display of EC-bound 125I on polyacrylamide gel showed the constitutive B beta- and gamma-chains of the fg molecule, with a partial loss of the A alpha-chain. Purified fragment E and E were tested for their capacity to inhibit fg binding. At a 1 to 400 125I-fg-to-fragment molar ratio, fragment E, which also inhibited migration, competed for binding by 44%, but fragment D was completely ineffective. These data show that fg may specifically associate with ECs and induce migration of these cells; it also appears that the structural requirement of this activity is located in the N-terminal part of the molecule.

Production of anti-P1A monoclonal antibodies.

Two peptides corresponding to the sequence of platelet glycoprotein IIIa between serine 27 and arginine 37 were synthesized and used to produce monoclonal antibodies. These two synthetic peptides were identical except for a single substitution at position 33, where a Pro/Leu polymorphism was shown to occur in human platelets and was predicted to be responsible for the P1A1-P1A2 alloantigen system (Newman et al., J. Clin Invest 1989:83:1778-81). Two monoclonal antibodies named 3C1 for the anti-"P1A1 peptide" and AD3 for the anti-"P1A2 peptide" were characterized. These monoclonal antibodies interacted with the two allelic forms of the reduced glycoprotein IIIa. They were used to type P1A1 and P1A2 homozygote as well as heterozygote platelets. Thus these two immunoprobes confirm that the Pro-Leu substitution is associated with the P1A1-P1A2 alloantigenic system. Although they interact only with reduced glycoprotein IIIa, these antibodies can be used to design simple tests for the typing of the P1A status of patients.

In vitro studies of the effect of buflomedil on platelet responsiveness.

The effect of buflomedil (Fonzylane; Laboratoire Lafon, Maisons-Alfort, France) on platelet function, a drug used clinically for the treatment of peripheral vascular diseases, was investigated in vitro. The compound significantly inhibits epinephrine-induced aggregation at the micromolar level. At higher doses (approximately 1 mM), a weak inhibition of ADP- and collagen-induced aggregation was observed; at these concentrations, buflomedil inhibits granular secretion and the interaction of fibrinogen with its receptor on platelet. Further investigations indicate that the drug affects calcium uptake at the membrane level and inhibits the binding of [3H]-yohimbine to the same extent as observed with phentolamine. The IC50 determined from competition binding assays was 1 +/- 0.5 microM. This value was consistent with the affinity constant approximated for the binding of [3H]-buflomedil to non-stimulated platelets. Taken-together, these results indicate that the vasoactive compound buflomedil is a weak antiaggregating agent which exhibits alpha 2-adrenergic antagonistic properties.

Evaluation of monoclonal antibodies with specificity for the human platelet HPA-1 allotypes.

Two mouse monoclonal antibodies have been characterized. Both antibodies specifically recognized the amino acid sequences which are responsible for the presence of the HPA-1a and HPA-1b alloantigens at the surface of human platelet. The monoclonality allied with an exquisite specificity of these antibodies allow a perfect typing of platelet HPA-1 status. A fast and reliable Elisa assay which fulfills all the needs and requirements of hematology laboratories has been set up and is described here. The Elisa assay has been tested on a population of 689 blood donors. The results agreed with the known phenotypic frequencies of the HPA-1 alloantigens.

Identification of a monoclonal antibody against platelet GPIIb that interacts with a calcium-binding site and induces aggregation.

We have characterized a monoclonal antibody named D33C, specific for platelet glycoprotein (GP) IIb, which induces fibrinogen binding and platelet aggregation. D33C Fab fragments interact with an average of 44,000 +/- 20,000 sites on resting platelet with a Kd value of 0.8 microM. This value decreased to 0.17 microM in the presence of 1 mM EDTA suggesting that Ca2+ chelation increases the antibody affinity. Purified IgGs and Fab fragments exhibit a similar potency and induce binding of fibrinogen and aggregation at levels comparable to those obtained with ADP. D33C-induced platelet aggregation, however, was not inhibited by 1 microM PGE1 and was not associated with a significant [14C]serotonin release, suggesting differences with ADP in the mechanism of activation. Among a large series of synthetic peptides corresponding to potential antigenic sequences within the structure of GPIIb, one peptide with the sequence DIDDNGYPDLIV was found to inhibit D33C activity. This peptide corresponds to a putative calcium-binding site whose sequence is highly homologous to similar sequences present in the alpha subunits of the fibronectin and the vitronectin receptors. Despite this homology, D33C interacts only with platelet GPIIb suggesting that the identified epitope may be differently exposed at the surface of the cells. This antibody may prove to be a valuable tool to study the induction reaction on recombinant GPIIbIIIa expressed in cells that lack the appropriate signal transduction reactions.

Expression of intracellular fibrinogen on the surface of stimulated platelets.

Participation of fibrinogen in platelet aggregation is contingent upon the capacity of various stimuli to induce specific receptors for the molecule on the surface of the cell. The interaction of fibrinogen with this receptor results directly in platelet aggregation, and dissociation of fibrinogen is associated with disaggregation. While the role of exogenous fibrinogen in this process has been fully documented, the mechanisms which control the surface exposure of platelet fibrinogen are less understood. In the present study Fab fragments of antibodies monospecific for fibrinogen have been used to examine the surface expression of intracellular fibrinogen and its involvement in platelet aggregation. Radiolabelled Fab fragments did not interact with non-stimulated platelets but significant binding was observed when the cells were stimulated by ADP, thrombin, collagen and Ca ionophore A23187. Binding was specific for fibrinogen, was not observed with thrombasthenic platelets and was dependent upon the presence of extracellular calcium. With all stimuli tested, the binding of the Fab probe to platelets correlated with platelet secretion. At the following concentrations of stimuli: 30 microM ADP, 4 micrograms/ml collagen, 3 microM A23187 and 0.05 U/ml thrombin, the immune Fab fragments inhibited platelet aggregation. A monoclonal antibody to glycoprotein IIb/IIIa complex and a synthetic peptide gamma 400-411, that inhibited the interaction of plasma fibrinogen with platelets, did not inhibit the binding of 125I-FAB fragments. Taken together these results support the hypothesis that endogenous fibrinogen becomes surface-expressed during stimulation of the cell and can support platelet aggregation, particularly that induced by low concentrations of stimuli. The mechanism for the surface expression of platelet fibrinogen may be distinct from that for the binding of plasma fibrinogen.

Amino acid sequences in fibrinogen mediating its interaction with its platelet receptor, GPIIbIIIa.

Platelet membrane GPIIbIIIa is a member of the family receptors named integrins that recognize RGD sequences in their ligands. GPIIbIIIa interacts with at least three different adhesive ligands: fibrinogen, fibronectin, and von Willebrand factor. These interactions are inhibited by RGD-containing peptides and by peptides corresponding to a sequence unique to fibrinogen in the COOH-terminal domain of its gamma chain (HLGGAKQAGDV). Two RGD sequences are present in fibrinogen A alpha chain: an RGDS sequence at A alpha 572-575, and an RGDF sequence at A alpha 95-98. Polyclonal antibodies raised against the RGDF sequence and the gamma COOH-terminal domain both reacted specifically with fibrinogen in solid phase enzyme-linked immunosorbent assays and immunoprecipitated the protein in solution. The Fab fragments prepared from these antibodies inhibited fibrinogen-platelet interaction and aggregation. These results demonstrate that these two sequences are both accessible within the fibrinogen molecule and are both implicated in ligand binding and cell-cell interaction. In addition, by further examining the interaction of the gamma chain peptide with platelets, it was found that RGDF and the gamma peptide produced a similar dose-dependent inhibition of the binding of the labeled gamma peptide to ADP-stimulated platelets. These results provide evidence that the RGDF sequence present at the A alpha 95-98 constitutes with the gamma 401-411 sequence two recognition sites interacting with the same site or with mutually exclusive sites on GPIIbIIIa.

Fibrinogen-endothelial cell interaction in vitro: a pathway mediated by an Arg-Gly-Asp recognition specificity.

It has been previously shown that fibrinogen (FG) associates specifically with human umbilical vein and bovine aortic endothelial cells (EC) in culture and induces EC migration. In the present study, we have investigated whether the FG-EC interaction is mediated by an Arg-Gly-Asp (RGD) recognition specificity and whether EC membrane proteins related to platelet GPIIb-IIIa are involved. Highly purified radioiodinated human FG, containing no detectable fibronectin, interacted with cultured human and bovine EC in suspension in a time-dependent and specific manner. The binding was inhibited by EDTA. Two polyclonal antibodies to platelet GPIIb-IIIa, which immunoprecipitated a heterodimer molecule from EC, inhibited FG binding to EC. These same antibodies inhibited FG-induced EC migration in a dose-dependent manner as measured in a Boyden chamber. Preabsorption of the antibodies with purified platelet GPIIb-IIIa markedly reduced both inhibitory activities. A series of synthetic RGD-containing peptides inhibited FG binding to EC and FG-induced EC migration. Gly-Arg-Gly-Asp (GRGD) was the most active peptide tested in inhibiting FG binding and EC migration (ID50 of 30 microM), and conservative substitutions in the RGD sequence markedly reduced inhibitory activity (ID50 greater than 1,000 microM). These results indicate that FG binding and EC migration are events mediated by an RGD recognition specificity and that EC surface proteins immunologically related to the GPIIb-IIIa complex on platelets are involved in this recognition.