Search

Being central players in thrombosis and hemostasis, platelets react in manifold and complex ways to extracellular stimuli. Cell-matrix and cell-cell interactions are mandatory for initial adhesion as well as for final development of stable plugs. Primary interfaces for interactions are plasma membrane proteins, of which many have been identified over the past decades in individual studies. However, due to their enucleate structure, platelets are not accessible to large-scale genomic screens and thus a comprehensive inventory of membrane proteins is still missing. For this reason, we here present an advanced proteomic setup for the detailed analysis of enriched platelet plasma membranes and the so far most complete collection of platelet membrane proteins. In summary, 1282 proteins were identified, of which more than half are termed to be of membrane origin. This study provides a brief overview of gene ontology subcellular and functional classification, as well as interaction network analysis. In addition, the mass spectrometric data were used to assemble a first tentative relative quantification of large-scale data on the protein level. We therefore estimate the presented data to be of major interest to the platelet research field and to support rational design of functional studies.

CEACAM1 is a cell adhesion molecule that has been implicated in a number of physiological processes (eg, tumor suppressor in epithelial tissues, potent angiogenic factor in microvessel formation, microbial receptor in human granulocytes and epithelial cells). The mechanism of CEACAM1 action is still largely unresolved but recent findings demonstrated that the cytoplasmic CEACAM1 domain is linked indirectly to the actin-based cytoskeleton. We have isolated integrin beta(3) as an associated protein using CEACAM1 tail affinity purification. This association depends on phosphorylation of Tyr-488 in the CEACAM1 cytoplasmic domain. Confocal laser scanning microscopy confirmed in vivo colocalization of both molecules in human granulocytes and epithelial cells. Furthermore, the concentrated colocalization at the tumor-stroma interface of invading melanoma masses suggests a functional role of CEACAM1-integrin beta(3) interaction in melanoma invasion. Moreover, colocalization of the two adhesion molecules is also found at the apical surface of glandular cells of pregnancy endometrium. Colocalization of CEACAM1 and integrin beta(3) at the transitional zone from proliferative to invasive extravillous trophoblast of the maternal-fetal interface supports a role for CEACAM1/integrin beta(3) complexes in cell invasion.

Signaling across integrins is regulated by interaction of these receptors with cytoskeletal proteins and signaling molecules. To identify molecules interacting with the cytoplasmic domain of the beta3-integrin subunit (glycoprotein IIIa), a placental cDNA library was screened in the yeast two-hybrid system. Two identical clones coding for a 96-amino acid sequence were identified. This sequence was 100% identical to a sequence in skelemin, a protein identified previously in skeletal muscle. Skelemin is a member of a superfamily of cytoskeletal proteins that contain fibronectin-type III-like motifs and immunoglobulin C2-like motifs and that regulate the organization of myosin filaments in muscle. The amino acid residues in the isolated clones encompassed C2 motifs 4 and 5 of skelemin. A recombinant skelemin protein consisting of C2 motifs 3-7 interacted with beta1- and beta3-integrin cytoplasmic domains expressed as glutathione S-transferase (GST) fusion proteins, but not with GST-beta2-integrin cytoplasmic tail or GST alone. The skelemin-binding region was in the membrane proximal cytoplasmic domains of the integrins. Full-length skelemin interacted with integrin in intact cells as demonstrated by the colocalization of hemagglutinin-tagged skelemin in Chinese hamster ovary (CHO) cells containing alphaIIbbeta3-integrin and by the finding that microinjection of C2 motif 4 of skelemin into C2C12 mouse myoblast cells caused spread cells to round up. A skelemin-like protein was detected in CHO cells, endothelial cells, and platelets, and this protein colocalized with beta1- and beta3-integrins in CHO cells. This study suggests the presence of a skelemin-like protein in non-muscle cells and provides evidence that it may be involved in linking integrins to the cytoskeleton.

Enzymatic and immunochemical assays show a phosphatidylinositol 4-kinase in novel and specific complexes with proteins (CD63 and CD81) of the transmembrane 4 superfamily (TM4SF) and an integrin (alpha3beta1). The size (55 kDa) and other properties of the phosphatidylinositol 4-kinase (PI 4-K) (stimulated by nonionic detergent, inhibited by adenosine, inhibited by monoclonal antibody 4CG5) are consistent with PI 4-K type II. Not only was PI 4-K associated with alpha3beta1-CD63 complexes in alpha3-transfected K562 cells, but also it could be co-purified from CD63 in untransfected K562 cells lacking alpha3beta1. Thus, TM4SF proteins may link PI 4-K activity to the alpha3beta1 integrin. The alpha5beta1 integrin, which does not associate with TM4SF proteins, was not associated with PI 4-K. Notably, alpha3beta1-CD63-CD81-PI 4-K complexes are located in focal complexes at the cell periphery rather than in focal adhesions. The novel linkage between integrins, transmembrane 4 proteins, and phosphoinositide signaling at the cell periphery may play a key role in cell motility and provides a signaling pathway distinct from conventional integrin signaling through focal adhesion kinase.

Human natural "anti-Gal" antibodies are specifically directed to Gal alpha 1-3Gal beta 1-4GlcNAc residues expressed on non-primate mammal and new world monkey cells. We investigated the relative involvement of purified IgG and IgM anti-Gal as xenoreactive natural antibodies (XNA). IgG and IgM were isolated from human plasma, and anti-Gal antibodies were purified by affinity chromatography on a Synsorb-14 column (Chembiomed, Edmonton, Alberta, Canada). Anti-Gal of both IgM and IgG classes represent the bulk of human XNA that bind to porcine platelets in enzyme-linked immunosorbent assay (ELISA). On immunoblots, normal human sera, as well as purified IgM and IgG fractions, reacted with 115-, 125-, 135-, 150-, 180-, 210-, and 240-kd) pig platelet proteins, whereas purified anti-Gal antibodies of both IgM and IgG classes mainly bound to 135-, 150-, 180-, and 210-kD glycoproteins. A low reactivity was observed in ELISA with anti-Gal free IgM and IgG, indicating that xenoantibodies are not solely directed to galactosyl epitopes. These antibodies revealed bands of 115, 125, and 240 kD, alpha-Galactosidase treatment of porcine platelet glycoproteins (gps) enriched by affinity chromatography abrogated the reactivity of 135- and 210-kD proteins. N- and O-glycosidase treatments demonstrated that alpha-galactosyl residues are located on the O-glycans of the 135-kD component. Finally, glycoproteins of 90 and 135 kD were identified by amino acid sequencing as the pig analogs of the human glycoproteins IIIa and IIb, respectively, whereas the 240-kD) component was identified as the porcine fibrinogen, using a new murine monoclonal antibody (naM147-7B6; IgG1) specific for its beta-chain.

To define the molecular basis of human chemoattractant receptor regulation, rat basophilic leukemia RBL-2H3 cells, which are thrombin-responsive, were transfected to stably express epitope-tagged receptors for C5a, interleukin-8 (IL-8), formylpeptides (e.g. N-formyl-methionyl-leucyl-phenylalanine (fMLP)), and platelet-activating factor (PAF). Here we demonstrate that both thrombin and a synthetic peptide ligand for the thrombin receptor (sequence SFLLRN) caused phosphorylation and heterologous desensitization of the receptors for C5a, IL-8, and PAF but not that for formylpeptides as measured by agonist-stimulated [35S]guanosine 5'-3-O-(thio)triphosphate binding to membranes. Consistent with the PAF receptor phosphorylation, both thrombin and thrombin receptor peptide inhibited phosphoinositide hydrolysis, Ca2+ mobilization, and degranulation stimulated by PAF. Unexpectedly, despite heterologous desensitization at the level of receptor/G protein activation, there was enhancement ("priming") by thrombin of subsequent activities stimulated by C5a and IL-8 as well as fMLP. The priming effect of thrombin was blocked by its inhibitor, hirudin. However, two other activators of the thrombin receptor, the peptide SFLLRN and trypsin, stimulated Ca2+ mobilization in RBL-2H3 cells but did not cause priming. In addition, SFLLRN and the thrombin receptor antagonist peptide FLLRN both inhibited thrombin-induced Ca2+ mobilization but not priming. Furthermore, the proteolytically active gamma-thrombin, which does not stimulate the tethered ligand thrombin receptor and caused little or no Ca2+ mobilization in RBL-2H3 cells, effectively primed the response to fMLP. These data demonstrate that heterologous receptor phosphorylation and attenuation of G protein activation are not, by themselves, sufficient for the inhibition of biological responses mediated by C5a and IL-8. Moreover, thrombin appears to utilize mechanism(s) independent of its tethered ligand receptor to selectively prime phospholipase C-mediated biological responses of the C5a, IL-8, and formylpeptide receptors but not PAF. Because C5a, IL-8, and formylpeptide activate phospholipase Cbeta2, whereas PAF stimulates a different phospholipase C, the striking selectivity of thrombin's priming may be mediated via its ability to enhance receptor-mediated activation of phospholipase Cbeta2.

ADP is an important platelet agonist which initiates platelet shape change, aggregation, exposure of fibrinogen receptors, and calcium mobilization. Because of the limitations of previously used affinity analogs and photo-labeling studies as well as controversies surrounding the identity of an ADP receptor on platelets, we have used an affinity label capable of alkylating a putative exofacial receptor on platelets. We now report that 8-(4-bromo-2,3-dioxobutylthio)adenosine-5'-diphosphate (8-BDB-TADP), which is an analog of the natural ligand ADP, blocked ADP-induced platelet shape change, aggregation, exposure of fibrinogen-binding sites, secretion, and calcium mobilization. Following modification by 8-BDB-TADP, the rates of aggregation of platelets induced by thrombin, a calcium ionophore (A23187) or a stimulator of protein kinase C (phorbol myristate acetate) were minimally affected. However, the 8-BDB-TADP-modified platelets exhibited decreased rates of aggregation in response to ADP, as well as collagen and a thromboxane mimetic (U46619), both of which partially require ADP. Autoradiograms of the gels obtained by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of solubilized platelets modified by either [beta-32P]8-BDB-TADP, or 8-BDB-TADP and NaB[3H]4 showed the presence of a single radiolabeled protein band at 100 kDa. The intensity of this band was reduced when platelets were preincubated with ADP, ATP, and 8-bromo-ADP prior to labeling by the radioactive 8-BDB-TADP. The results show that 8-BDB-TADP selectively and covalently labeled aggregin (100 kDa), a putative ADP receptor, resulting in a loss of ADP-induced platelet responses.

Transient phosphorylation of histidine characterizes the two-component systems in prokaryotes that control important physiological functions, but analogous events have not been implicated in signal transduction in mammalian cells. To explore histidine phosphorylation during activation of human cells, stimulated platelets were analyzed for the formation of protein phosphohistidine in a model system employing P-selectin. P-selectin, a leukocyte adhesion molecule, undergoes rapid phosphorylation and selective dephosphorylation of tyrosine, serine, and threonine. We now establish that phosphorylation following platelet activation with thrombin or collagen generates phosphohistidine at histidines on the cytoplasmic tail of P-selectin. With thrombin stimulation, the kinetics of phosphohistidine appearance and disappearance of P-selectin are very rapid. Platelets exhibit a novel ligand-induced signaling pathway to generate phosphohistidine. These results provide direct biochemical evidence for the induction of rapid and reversible histidine phosphorylation in mammalian cells upon cell activation and represent a novel paradigm for mammalian cell signaling.

The platelet antigens, PlA1 and PlA2, are responsible for most cases of posttransfusion purpura (PTP) and neonatal alloimmune thrombocytopenia (NAIT) in the caucasian population and are determined by two allelic forms of the platelet glycoprotein GPIIIa gene. To study the interaction between these antigens and their respective antibodies, we inserted the sequence that encodes the signal peptide and the N-terminal 66 amino acids of the PlA1 form of GPIIIa into the expression vector pGEX1. To express the PlA2 antigen, nucleotide 196 of the PlA1 coding sequence was mutated to the PlA2 allelic form. When transformed and induced in Escherichia coli, the two constructs produce glutathione S-transferase (GST)/N-terminal GPIIIa fusion proteins, one containing leucine at position 33 (PlA1), the other proline (PlA2). These proteins are easily purified in milligram quantities using glutathione-Sepharose and react specifically with their respective antibodies by immunoblot and enzyme-linked immunosorbent assay. Antigenicity of the PlA1 fusion protein in reduced glutathione increases with time; moreover, the addition of oxidized glutathione accelerates this process, presumably because of formation of the native disulfide bonds. Neutralization assays indicate that the PlA1 fusion protein competes for all of the anti-PlA1 antibody in the serum of patients with PTP and NAIT that is capable of interacting with the surface of intact platelets. This study shows that the GST/N-terminal GPIIIa fusion proteins contain conformational epitopes that mimic those involved in alloimmunization, and that regions other than the amino terminal 66 amino acids of GPIIIa are not likely to contain or be required for the development of functional PlA1 epitopes. Furthermore, these recombinant proteins can be used for the affinity-purification of clinical anti-PlA1 antibodies and specific antibody identification by western blotting, making them useful in the diagnosis of patients alloimmunized to PlA1 alloantigens.

Platelet adhesion to subendothelial von Willebrand factor involves receptor recognition by the platelet glycoprotein (GP) Ib-IX and initiates activation signals that contribute to primary hemostasis. We show here that GPIb-IX is specifically associated with an intracellular 29-kDa protein. The physicochemical characteristics and amino acid sequence of this protein indicate that it is identical to the human zeta-isoform 14-3-3 protein, previously characterized as a platelet phospholipase A2 (PLA2). As activation of PLA2 is an early event in GPIb-IX-mediated signaling, this result suggests that ligand occupancy of GPIb-IX may directly activate PLA2, leading to platelet activation.

The stimulation of human platelets with physiological agonists results in the incorporation of several proteins into the cytoskeleton, fibrinogen binding, and platelet aggregation. We recently demonstrated that the Ras-related low molecular weight GTP-binding protein Rap2B associates with the cytoskeleton in activated platelets and that this interaction requires platelet aggregation. In the present study we demonstrate that agonist-induced actin polymerization is necessary for the translocation of Rap2B to the cytoskeleton, suggesting that Rap2B interacts with the newly formed actin filaments. Moreover, the association of Rap2B with Triton X-100-insoluble material from platelets was totally blocked by treatment of intact platelets with monoclonal antibodies against the fibrinogen receptor glycoprotein IIb-IIIa. Platelets from patients affected by Glanzmann thrombastenia, a genetic disorder in which platelet plasma membranes lack glycoprotein IIb-IIIa but possess normal levels of Ras-related proteins, failed to incorporate Rap2B into the cytoskeleton upon activation by thrombin. Comparative immunoblotting revealed that the translocation of Rap2B to the cytoskeleton during platelet aggregation was accompanied by the simultaneous translocation of glycoprotein IIb-IIIa. Moreover, the cytoskeleton from aggregated platelets contained Rap2B and glycoprotein IIb-IIIa in comparable amounts. These results demonstrate the association of Rap2B and glycoprotein IIb-IIIa and their translocation to the cytoskeleton in aggregated human platelets.

Glycoprotein (GP) Ib alpha is required for expression of the highest affinity alpha-thrombin-binding site on platelets, possibly contributing to platelet activation through a pathway involving cleavage of a specific receptor. This function may be important for the initiation of hemostasis and may also play a role in the development of pathological vascular occlusion. We have now identified a discrete sequence in the extracytoplasmic domain of GP Ib alpha, including residues 271-284 of the mature protein, which appears to be part of the high affinity alpha-thrombin-binding site. Synthetic peptidyl mimetics of this sequence inhibit alpha-thrombin binding to GP Ib as well as platelet activation and aggregation induced by subnanomolar concentrations of the agonist; they also inhibit alpha-thrombin binding to purified glycocalicin, the isolated extracytoplasmic portion of GP Ib alpha. The inhibitory peptides interfere with the clotting of fibrinogen by alpha-thrombin but not with the amidolytic activity of the enzyme on a small synthetic substrate, a finding compatible with the concept that the identified GP Ib alpha sequence interacts with the anion-binding exosite of alpha-thrombin but not with its active proteolytic site. The crucial structural elements of this sequence necessary for thrombin binding appear to be a cluster of negatively charged residues as well as three tyrosine residues that, in the native protein, may be sulfated. GP Ib alpha has no significant overall sequence homology with the thrombin inhibitor, hirudin, nor with the specific thrombin receptor on platelets; all three molecules, however, possess a distinct region rich in negatively charged residues that appear to be involved in thrombin binding. This may represent a case of convergent evolution of unrelated proteins for high affinity interaction with the same ligand.

Stabilization of a clot is dependent on fibrin cross-linking mediated by the transglutaminase, factor XIIIa (FXIIIa). In addition to fibrin stabilization, FXIIIa acts on a number of platelet-reactive proteins, including fibronectin and vitronectin, as well as the platelet proteins, glycoprotein (GP) IIb-IIIa, myosin, and actin. However, conditions inducing the platelet-activation dependent binding of FXIIIa have not been characterized nor have the sites mediating FXIIIa binding been identified. The generation of FXIIIa and consequent detection of FXIIIa on the platelet surface were compared with other thrombin-induced activation events; the rate at which FXIIIa bound to activated platelets was much slower than platelet degranulation or fibrin(ogen) binding. Whereas platelets could be rapidly induced to express a functional receptor for FXIIIa, the rate of FXIIIa binding to platelets is limited by the rate of conversion of FXIII to FXIIIa. Immunoprecipitation of radiolabeled platelets using polyclonal anti-FXIII A-chain antibody identified two proteins corresponding to GPIIb and GPIIIa. Preincubation of intact platelets with 7E3, a monoclonal antibody that blocks the fibrinogen binding site, or GRGDSP peptide inhibited FXIIIa binding by about 95% when measured by flow cytometry; FXIIIa binding to purified GPIIb-IIIa was also inhibited by 7E3. The binding of FXIIIa to purified GPIIb-IIIa was enhanced by the addition of fibrinogen, but not by that of fibronectin or thrombospondin, suggesting that FXIIIa also binds to fibrinogen associated with the complex. These observations suggest that activated platelets bearing FXIIIa may enhance stabilization of platelet-rich thrombi through surface-localized cross-linking events.

A well-characterized in vitro model system composed of thrombin-stimulated gel-filtered human platelets, fibrin-(ogen), plasminogen, and recombinant tissue plasminogen activator (rt-PA) was used to examine the relationship between platelet-fibrin adhesive interactions and the lytic resistance of a platelet-rich thrombus. Laser light scattering kinetic experiments demonstrated that the ligand-mimetic peptide D-RGDW and an anti-alpha IIb beta 3 monoclonal antibody both inhibited clot retraction, but neither integrin-targeted reagent affected the overall delay in lysis of "bulk" fibrin caused by thrombin-stimulated platelets. However, lysis of the model platelet-rich thrombus did proceed some 30% more quickly when treated with a plasminogen activator inhibitor (PAI)-resistant t-PA variant. Taken together, these results confirm that platelet-released PAI-1 is a major determinant of global lytic resistance. Next events occurring during fibrinolysis in the unique microenvironment near the platelet surface were monitored by scanning electron microscopy and quantitative fluorescence microscopy. Scanning electron micrographs of the partially lysed model thrombus in the presence of 200 mumol/L of D-RGDW showed no platelet aggregates, and fibrin was attached by fewer strands to the platelets. Quantitative fluorescence microscopy, using fluorescein-labeled fibrin, showed that fibrin adherent to the surface of thrombin-stimulated platelets lysed 20% to 50% more slowly than bulk fibrin (monitored in parallel by laser light scattering). Furthermore, this microspectroscopic technique showed that D-RGDW reduced the quantity of platelet-bound fibrin, and accelerated lysis near the platelet surface with both native rt-PA and the PAI-resistant variant. These observations suggest that the dense network of fibrin bound to the platelet surface is protected from fibrinolysis by tissue-type plasminogen activators. Further, uncoupling fibrin from its platelet receptors uniquely hastens fibrinolysis at the cell/fibrin interface.

Cellular receptors for blood proteases regulate chemotaxis, extracellular proteolysis, and growth behavior of normal and malignant cells. Binding of the coagulation protease factor Xa to leukocytes is contributed by a recently identified molecule, denominated Effector cell Protease Receptor-1 (EPR-1). Monoclonal antibodies were generated against EPR-1+ MOLT13 lymphocytes and selected for reactivity with lymphocyte surface proteins by flow cytometry and with affinity-purified EPR-1 in Western blots. Antibody-based functional cloning of the cDNA for EPR-1 reveals the sequence of a novel molecule encoded by a prominent 1.9-kilobase mRNA. The cDNA predicts a glycoprotein of 337 amino acids, characterized by a unique cysteine-rich extracellular module, a single membrane-spanning domain, and a serine-rich (26%) cytoplasmic tail featuring at least 15 potential phosphorylation sites. Genetically engineered EPR-1 transfectants were recognized by monoclonal antibodies to EPR-1 and bound 125I-factor Xa in a specific and saturable manner (Kd approximately 10-15 nM). In the absence of factor V/Va, EPR-1 transfectants promoted prothrombin activation in a factor Xa concentration-dependent reaction, inhibited by a monoclonal antibody to EPR-1. These findings define EPR-1 as a novel cell surface receptor for factor Xa potentially implicated in protease-dependent cellular effector functions.

The platelet plasma membrane is lined with a membrane skeleton composed of short actin filaments, actin-binding protein, spectrin, vinculin, and other unidentified proteins. It is connected to the outside of the cell through association with the cytoplasmic domains of transmembrane receptors. In detergent-lysed platelets, cytoplasmic actin filaments are sedimented by centrifugation at 15,600 x g, but the sedimentation of membrane skeleton fragments requires higher g-forces (100,000 x g). In the present study, we show that the major platelet integrin, glycoprotein (GP) IIb-IIIa, sediments from detergent-lysed platelets at 100,000 x g together with fragments of the membrane skeleton that contain the cytoskeletal proteins spectrin, vinculin, and talin. In addition, this cell fraction contained the tyrosine kinases pp60c-src and pp62c-yes and the p21ras GTPase-activating protein (GAP). After thrombin-induced platelet aggregation mediated by fibrinogen binding to GP IIb-IIIa on adjacent platelets, we detected a redistribution of spectrin, talin, vinculin, pp60c-src, and pp62c-yes to the fraction that sediments at 15,600 x g. The redistribution of these proteins from the high-speed detergent-insoluble fraction to the low-speed fraction correlated with the extent of aggregation and was not detected in aggregation-defective thrombasthenic platelets (which lack the GP IIb-IIIa complex). In addition, many of the proteins phosphorylated on tyrosine in activated platelets were present in detergent-insoluble fractions. These results are consistent with the possibilities that 1) GP IIb-IIIa, pp60c-src, pp62c-yes, and GAP associate with a membrane skeleton fraction that contains spectrin, vinculin, and talin, 2) the association of GP IIb-IIIa with adhesive ligand in a platelet aggregate causes components of the membrane skeleton to undergo altered association with cytoplasmic actin filaments, and 3) many of the proteins phosphorylated on tyrosine residues in activated platelets are components of the cytoskeleton. The results imply that the membrane skeleton may play an important role in binding signaling molecules at sites of integrin-cytoskeleton interactions and in mediating signal transduction events in platelets. Further, GP IIb-IIIa-induced redistribution of components of the membrane skeleton and associated signaling molecules may represent an important step in regulating integrin-induced motile events in platelets.

The transforming properties of v-jun, the viral counterpart of the transcription factor AP1, were investigated in avian hematopoietic cells. Two retroviruses, called JB and JBN, expressing both v-jun and v-erbB, were constructed using an avian erythroblastosis-based vector. We show that the cooperative action of both oncogenes allowed the virus to efficiently transform bone marrow cells. No such transformation was obtained with either oncogene alone. JB-transformed bone marrow cells expressed GATA-1, TAL-1, and histone H5, suggesting that they belong to the erythrocytic/thrombocytic lineage. (Thrombocytes are the avian homologues of mammal megakaryocytes.) Moreover, after induction with phorbol 12-myristate 13-acetate JB-transformed bone marrow cells began to differentiate and synthesized high levels of platelet glycoproteins, indicating that they were of thrombocytic origin. These results were confirmed by c-ets1 analysis since this transcription factor, specifically found in cells with megakaryocytic but not erythrocytic features, was clearly detected in these cells.

P-selectin is an adhesion receptor for leukocytes that is redistributed from secretory granule membranes to the surfaces of activated platelets and endothelial cells. The cytoplasmic domain of P-selectin contains two serines, two threonines, and one tyrosine that could potentially be phosphorylated. We found that P-selectin was phosphorylated in both platelets and endothelial cells and that phosphorylation rapidly increased after cell activation. Approximately 0.02, 0.05, and 0.08 mol of phosphate/mol of P-selectin were incorporated, respectively, into resting, thrombin-activated, and phorbol ester-activated platelets. Phosphorylation was completely inhibited by the protein kinase C inhibitors, staurosporine, H-7, and chelerythrine, and was enhanced by the phosphatase inhibitors, okadaic acid and calyculin-A. Phosphoamino acid analysis of 32P-labeled P-selectin showed that phosphorylation occurred predominantly on serine with lesser amounts on threonine. When expressed in transfected Chinese hamster ovary cells, P-selectin was also phosphorylated. Mutagenesis studies showed that Ser788 was the principal site of phosphorylation, with minor sites on the other serine and threonine residues of the cytoplasmic domain. Phosphorylation may regulate membrane trafficking or other functions of P-selectin.

P-selectin is an adhesion receptor for leukocytes on thrombin-activated platelets and endothelial cells. It contains a NH2-terminal carbohydrate-recognition domain, an epidermal growth factor motif, nine consensus repeats, a transmembrane domain, and a cytoplasmic tail. We expressed two soluble forms of P-selectin, one truncated after the ninth repeat (tPS) and the other lacking the transmembrane domain due to alternative RNA splicing (asPS). When visualized by electron microscopy, each was a monomeric rod-like structure with a globular domain at one end, whereas membrane P-selectin (mPS) from platelets formed rosettes with the globular domains facing outward. Sedimentation velocity and equilibrium studies confirmed that tPS and asPS were asymmetric monomers, whereas mPS was oligomeric. HL-60 cells adhered to immobilized tPS and asPS, although less efficiently than to mPS. 125I-Labeled tPS and asPS bound to approximately 25,000 sites/neutrophil and approximately 36,000 sites/HL-60 cell with an apparent Kd of 70 nM. Treatment of HL-60 cells with O-sialoglycoprotease eliminated the binding sites for asPS. We conclude that 1) P-selectin is a rigid, asymmetric protein; 2) monomeric soluble P-selectin binds to high affinity ligands with sialylated O-linked oligosaccharides on leukocytes; and 3) oligomerization of mPS enhances its avidity for leukocytes.

Lectin-carbohydrate recognition between the selectins and their ligands are among the earliest events in leukocyte recirculation, leukocyte recruitment into inflamed areas, and abnormal egress of leukocytes in diseases. Previously, we have described a dimeric sialoglycoprotein from myeloid cells with subunits of molecular mass = 120 kDa, which is selectively recognized by P-selectin (Moore, K.L., Stults, N.L., Diaz, S., Smith, D.F., Cummings, R.D., Varki, A., and McEver, R.P. (1992) J. Cell Biol. 188, 445-456). Here, we demonstrate that this P-selectin ligand carries alpha 2-3-linked sialic acids and the sialyl-Lewisx (SLex) tetrasaccharide motif. This glycoprotein contains < 1% of the total membrane-bound sialic acids and a very small fraction of the total SLex on neutrophil membranes. In spite of a relative resistance to sialidase digestion, the predominant form of sialic acid on the ligand is N-acetylneuraminic acid. Selective periodate oxidation of the side chain of sialic acids does not affect P-selectin binding and allows the introduction of tritium label into the truncated sialic acids. beta-Elimination with alkaline borohydride releases labeled O-linked oligosaccharides both from the labeled neutrophil ligand and from the ligand purified from HL-60 cells metabolically labeled with [3H]glucosamine. The ligand from both neutrophils and HL-60 cells is also susceptible to cleavage by the enzyme O-sialoglycoprotease from Pasteurella hemolytica. Analysis of the specificity of this enzyme suggests that the P-selectin ligand carries large numbers of closely spaced sialylated O-linked oligosaccharides. O-Sialoglycoprotease abolishes both direct binding of P-selectin to HL-60 cells and the adhesion of HL-60 cells to immobilized P-selectin, without significantly decreasing overall cell surface SLex expression. This indicates that the 120-kDa ligand may be the major determinant of P-selectin:myeloid cell interaction in vivo. Finally, based on the current and previous data, we hypothesize that the high affinity recognition site(s) of this P-selectin ligand may be derived from a "clustered saccharide patch" of sialylated fucosylated O-linked oligosaccharide sequences.

We have used a series of Escherichia coli-expressed deletion mutants of the glycoprotein (GP) Ib-binding domain of von Willebrand factor (vWF) to study the structural basis of its function. In addition to the prototypic molecule (rvWF441-733), we constructed 11 mutants; seven had deletions of sequence on the amino- and/or carboxyl-terminal side of the Cys509-Cys695 intrachain disulfide loop, and four had limited deletions inside the loop. Other cysteine residues in addition to 509 and 695, when present in the corresponding native sequence, were mutated to glycine; all molecules were purified in the oxidized as well as reduced and alkylated state. The smallest species retaining the ability to interact with GP Ib in the absence of modulators was the oxidized rvWF508-696; the latter, as well as rvWF441-696, became inactive after reduction and alkylation. In contrast, all the other fragments with deletions outside of the loop, but extending at least to residue 700, showed better binding to platelets after reduction and alkylation than when the Cys509-Cys695 disulfide bond was oxidized. Any limited deletion of sequence inside the loop caused complete loss of GP Ib-binding function both in the absence or in the presence of botrocetin, and this persisted even after reduction and alkylation. In contrast, all mutants with intact sequence between residues 509 and 695 bound to GP Ib in the presence of botrocetin, regardless of whether the 2 cysteine residues were oxidized or reduced and alkylated. Ristocetin, unlike botrocetin, appeared to have no effect in modulating the binding of any of the expressed fragments to platelets. Our findings suggest that the GP Ib-binding domain of vWF contains multiple interaction sites, but integrity of the sequence 509-695 is important for function.

The cardioprotective effects of an mAb to P-selectin designated mAb PB1.3 was examined in a feline model of myocardial ischemia (MI) and reperfusion. PB1.3 (1 mg/kg), administered after 80 min of ischemia (i.e., 10 min before reperfusion), significantly attenuated myocardial necrosis compared to a non-blocking mAb (NBP1.6) for P-selectin (15 +/- 3 vs 35 +/- 3% of area at risk, P < 0.01). Moreover, endothelial release of endothelium derived relaxing factor, as assessed by relaxation to acetylcholine, was also significantly preserved in ischemic-reperfused coronary arteries isolated from cats treated with mAb PB1.3 compared to mAb NBP1.6 (67 +/- 6 vs 11 +/- 3, P < 0.01). This endothelial preservation was directly related to reduced endothelial adherence of PMNs in ischemic-reperfused coronary arteries. Immunohistochemical localization of P-selectin was significantly upregulated in the cytoplasm of endothelial cells that lined coronary arteries and veins after 90 min of ischemia and 20 min of reperfusion. The principal site of intracytoplasmic expression was in venous vessels. mAb PB1.3 significantly decreased (P < 0.01) adherence of unstimulated PMNs to thrombin and histamine stimulated endothelial cells in a concentration-dependent manner in vitro. These results demonstrate that PMN adherence to endothelium by P-selectin is an important early consequence of reperfusion injury, and a specific monoclonal antibody to P-selectin exerts significant endothelial preservation and cardioprotection in myocardial ischemia and reperfusion.

We report that the adhesion receptor P-selectin can be metabolically labeled with [3H]palmitic acid in human platelets. Analysis of alkaline methanolysis products from labeled protein demonstrated that the radioactivity associated with P-selectin was covalently bound palmitic acid. [3H]Palmitic acid was cleaved by hydroxylamine treatment at neutral pH and by reducing agents, indicating that acylation occurred through a thioester linkage. Both stearic acid and palmitic acid were detected by gas chromatography-mass spectrometry analysis of alkaline hydrolysates of purified P-selectin. Deletion or mutation of Cys766 eliminated [3H] palmitic acid labeling of P-selectin in transfected COS-7 cells. We conclude that the cytoplasmic domain of P-selectin is acylated at Cys766 through a thioester bond. Fatty acid acylation may regulate intracellular trafficking or other functions of P-selectin.

In addition to thyroglobulin, primary thyrocytes secrete into the culture medium significant quantities of p500, a protein so named because of its M(r) > or = 500,000. Epithelial monolayers cultured on porous filters serve as a useful model system in which to study protein secretion. From these monolayers, thyroglobulin, the precursor in thyroid hormonogenesis, was released with apical predominance, while p500 was found mostly in the basolateral medium. Thyrocyte exposure to thyrotropin augmented selectively thyroglobulin but not p500 production. By contrast, exposure to cycloheximide actually augmented p500 production, a response observed for immediate-early proto-oncogenes. Using thyrocyte conditioned medium, the p500 protein was purified to homogeneity. Peptide sequencing of tryptic fragments of purified p500 showed identity to thrombospondin. Immunoprecipitation of thrombospondin from media bathing primary thyrocytes and the FRTL5 cell line quantitatively recovered p500, confirming its identity and indicating an epithelial origin. Gel filtration of secreted thrombospondin eluted at a high molecular weight, suggesting complexation with components of the extracellular matrix. Further, immunofluorescence showed cellular codistribution of thrombospondin and thyroglobulin, although thrombospondin exhibited predominantly an extracellular, basolateral deposition. It seems likely that thrombospondin production by thyrocytes plays a role in the growth or development of the thyroid epithelium.

The levels and expression of the proteins CD63 and granulophysin in platelets from control and from a Hermansky-Pudlak syndrome subject (a condition characterized by dense granule and lysosomal deficiencies and the accumulation of ceroid-like material in reticuloendothelial cells) were examined. Immunofluorescence studies indicated that anti-CD63 and anti-granulophysin antibodies recognized similar numbers of granules; coapplication of antibodies did not identify more granules than the individual antibodies. Significantly fewer granules were recognized in Hermansky-Pudlak syndrome platelets than in control using either antibody. Immunoblotting studies demonstrated that anti-CD63 and anti-granulophysin antibodies apparently recognize the same protein, which was deficient in Hermansky-Pudlak platelets. Analysis by fluorescence-activated cell sorter (FACS) showed biphasic expression of CD63 and granulophysin after thrombin stimulation of control but not Hermansky-Pudlak platelets. Anti-CD63 effectively blocked detection of the protein by anti-granulophysin using immunofluorescence, ELISA, immunoblotting, and FACS analysis. Amino-terminal sequencing over the first 37 amino acids revealed that granulophysin was homologous to CD63, melanoma antigen ME491, and pltgp40. These results suggest that granulophysin and CD63 are possibly identical proteins. This is the first report of a protein present in platelet dense granules, lysosomes, and melanocytes, but deficient in a patient with Hermansky-Pudlak syndrome.

Chronic immune thrombocytopenic purpura (ITP) is a syndrome of destructive thrombocytopenia due to autoantibodies against platelet-associated antigens. These antigens are most commonly located on the platelet glycoprotein (GP) IIb/IIIa complex. In the present studies, we show that many platelet-associated anti-GPIIb/IIIa autoantibodies from chronic ITP patients depend on conformationally intact GPIIb/IIIa for maximal binding. We studied anti-GPIIb/IIIa autoantibodies from 19 ITP patients (15 platelet-associated, 8 plasma) and alloantibodies from three patients with posttransfusion purpura (anti-PIA1). Antibodies were preincubated with purified intact GPIIb/IIIa, EDTA-dissociated GPIIb/IIIa, GPIIIa, or GPIIb for 2 hours and then residual antibody was measured in an antigen capture assay. The binding results were compared with those obtained using antibody preincubated in buffer. Of the 15 platelet-associated autoantibodies studied, the intact GPIIb/IIIa complex resulted in greater inhibition of antibody binding than the EDTA-dissociated complex, with a mean inhibition ratio (intact/dissociated) of 7.9 (range, 1.4 to 30.3). Little inhibition was noted using either GPIIb or GPIIIa. Conversely, plasma anti-PIA1 alloantibodies or plasma autoantibodies from ITP patients against the c-terminal region of GPIIIa were more efficiently inhibited by the dissociated complex or purified GPIIIa. We conclude that platelet-associated anti-GPIIb/IIIa autoantibodies in chronic ITP are frequently directed to cation-dependent conformational antigens.

The platelet glycoproteins GPIIb and GPIIIa are integral membrane proteins and form calcium-dependent heterodimers in the endoplasmic reticulum (ER). In the absence of heterodimer formation, GPIIb and GPIIIa are retained in the ER and degraded. To produce soluble forms of these proteins, we truncated each at a site just proximal to its transmembrane anchor and expressed the mutants in COS-1 cells. We found that both truncated GPIIIa (GPIIIatr) and GPIIIatr were secreted by the transfected cells. However, GPIIbtr was retained by the cells and was immunoprecipitated as a doublet with a 115,000 molecular weight protein. Incubation of transfected cells with the calcium ionophore A23187 or the calcium chelator 1,2-bis(2-aminophenoxy)ethane N,N,N',N'-tetraacetic acid tetrakis(acetoxymethyl) ester (BAPTA-AM) failed to induce appreciable GPIIbtr secretion, suggesting that formation of intracellular calcium complexes was not a factor in GPIIbtr retention. Further, immunoblotting of immunoprecipitated GPIIbtr and GPIIIatr revealed that the chaperone binding protein (BiP) was associated with each, arguing that BiP alone was not responsible for GPIIbtr retention. These studies indicate that the intracellular retention of GPIIIa involves sequences located in the transmembrane or cytoplasmic domains of the molecule. GPIIb contains an additional retention signal located in the extracellular portion of the molecule whose effect is abrogated by formation of a GPIIb-IIIa heterodimer. This signal may be involved in the fate of nascent GPIIb monomers and the generation of correctly configured GPIIb-IIIa heterodimers.

Serum platelet-reactive and glycoprotein (GP) IIb-GPIIIa-reactive IgG and F(ab')2 was examined in 39 patients with classic autoimmune thrombocytopenic purpura (ATP), two patients with anti-PLA1 antibody and 25 control subjects in an enzyme-linked immunosorbent assay. IgG was purified by diethyl aminoethyl chromatography and centrifuged at 100,000g before testing of the supernatant. Significant IgG binding (threefold to fourfold control IgG binding) was noted with 8 of 17 ATP patients' IgG, 2 anti-PLA1 IgGs, and 2 ATP patients with multiple platelet transfusions. However, F(ab')2 fragments of nine of nine positive ATP IgGs were nonreactive; F(ab')2 from the two anti-PLA1 and two multiply transfused ATP IgGs were as reactive as their intact IgG. Antiplatelet or anti-GPIIb-GPIIIa reactivity of ATP IgG could be adsorbed to fixed platelets or solid-phase GPIIb-GPIIIa and eluted with 0.1 mol/L glycine, pH 2.5. However, binding of IgG to GPIIb-GPIIIa could not be inhibited with F(ab')2 of ATP IgG or Fc fragments of control subjects. When platelet- or GPIIb-GPIIIa-reactive ATP IgG was applied to a Sephacryl 300 gel filtration column, no reactivity was noted in the 7S region, whereas anti-PLA1 localized to this region. Antiplatelet or anti-GPIIb-GPIIIa reactivity was noted in the void volume and accompanied by a high molecular weight protein region. An immunoblot of the void volume fraction with goat antihuman IgG (gamma chain) antibody showed high molecular weight bands greater than 250 Kd, which after reduction converted to a 55-Kd heavy-chain band. Fresh samples of ATP and control IgG processed within 1 to 2 days of blood withdrawal had no reactivity for GPIIb-GPIIIa. After storage at -20 degrees C for greater than 3 months, 5 of 19 ATP IgG became reactive, whereas 16 of 16 controls were nonreactive. Thus, platelet-reactive IgG of ATP sera appears to be caused by the development of IgG aggregates held together by disulfide bonds that develop on storage, and is not F(ab')2 mediated.

Recombinant hirudin variants have been designed which inhibit alpha-thrombin by the hirudin mechanism and which in addition exhibit disintegrin activity. These proteins, called "hirudisins," have been engineered by replacing the Ser-Asp-Gly-Glu sequence at the tip of hirudin's finger-like structure (residues 32-35) by Arg-Gly-Asp-Ser (RGDS) to yield hirudisin and Lys-Gly-Asp-Ser (KGDS) to obtain hirudisin-1. Comparison of thrombin inhibition activities showed that hirudisin is 2-fold more potent (K(i) = 160 +/- 70 fM) than hirudisin-1 (K(i) = 370 +/- 44 fM) and recombinant (r)-hirudin (K(i) = 270 +/- 50 fM). alpha-Thrombin-stimulated platelet aggregation was effectively inhibited by r-hirudin, hirudisin, and hirudisin-1 with IC50 of 5.7 to 6.8 nM. Unlike r-hirudin, hirudisin inhibits ADP-induced platelet aggregation (IC50 = 65 microM) 3- to 5-fold stronger than the linear GRGDS- and RGDS-peptide. Direct interaction of hirudisin with purified glycoprotein IIb-IIIa demonstrated that antiplatelet aggregation activity is due to the integrin-directed RGD motif. Disintegrin activity of hirudisin relative to that of reduced and carboxymethylated hirudisin suggests that the conformational strain favors binding to integrins. On the basis of these results, hirudisins appear to be interesting molecules for the design of potential antithrombotic agents with antithrombin as well as antiplatelet aggregation activities.

We describe here the alteration of thrombin specificity induced by its interaction with glycocalicin. Glycocalicin is the external part of platelet glycoprotein Ib alpha (GPIb alpha) and contains binding sites for von Willebrand factor and thrombin. Taking advantage of its solubility, we have used glycocalicin in competition assays on various thrombin activities. Glycocalicin did not inhibit chromogenic substrate hydrolysis nor diisopropylfluorophosphate iPr2 (PF) incorporation, indicating that thrombin binding to GPIb does not alter access to or the conformation of the thrombin catalytic site. Glycocalicin competitively inhibited thrombin binding to fibrin (Ki = 0.1 mumol/L) and blocked fibrinogen clotting activity of thrombin. Glycocalicin also inhibited thrombin binding to thrombomodulin in a competitive manner (Ki = 3 to 5 mumol/L), but failed to prevent thrombin interaction with protein C in the absence of thrombomodulin. Previous results have indicated that GPIb binds to thrombin within the anion binding exosite masked by the carboxy-terminal hirudin peptide 54-65. The present results confirm the implication of the anion binding exosite in GPIb recognition, and further indicate that the thrombin binding site for GPIb overlaps with the thrombin binding sites for fibrin and thrombomodulin, whereas it is distinct from the thrombin binding site for protein C. Some of the structural requirements for thrombin binding to GPIb appear to be very similar to those reported for binding to its platelet receptor. However, thrombin-GPIb interaction does not appear to compete with receptor hydrolysis but rather increases the sensitivity and the rate of platelet responses elicited by the receptor.

Integrins are membrane receptors which mediate cell-cell or cell-matrix adhesion. Integrin alpha IIb beta 3 (glycoprotein IIb-IIIa) acts as a fibrinogen receptor of platelets and mediates platelet aggregation. Platelet activation is required for alpha IIb beta 3 to shift from noncompetent to competent for binding soluble fibrinogen. The steps involved in this transition are poorly understood. We have studied a variant of Glanzmann thrombasthenia, a congenital bleeding disorder characterized by absence of platelet aggregation and fibrinogen binding. The patient's platelets did not bind fibrinogen after platelet activation by ADP or thrombin, though his platelets contained alpha IIb beta 3. However, isolated alpha IIb beta 3 was able to bind to an Arg-Gly-Asp-Ser affinity column, and binding of soluble fibrinogen to the patient's platelets could be triggered by modulators of alpha IIb beta 3 conformation such as the Arg-Gly-Asp-Ser peptide and alpha-chymotrypsin. These data suggested that a functional Arg-Gly-Asp binding site was present within alpha IIb beta 3 and that the patient's defect was not secondary to a blockade of alpha IIb beta 3 in a noncompetent conformational state. This was evocative of a defect in the coupling between platelet activation and alpha IIb beta 3 up-regulation. We therefore sequenced the cytoplasmic domain of beta 3, following polymerase chain reaction (PCR) on platelet RNA, and found a T-->C mutation at nucleotide 2259, corresponding to a Ser-752-->Pro substitution. This mutation is likely to be responsible for the uncoupling of alpha IIb beta 3 from cellular activation because (i) it is not a polymorphism, (ii) it is the only mutation in the entire alpha IIb beta 3 sequence, and (iii) genetic analysis of the family showed that absence of the Pro-752 beta 3 allele was associated with the normal phenotype. Our data thus identify the C-terminal portion of the cytoplasmic domain of beta 3 as an intrinsic element in the coupling between alpha IIb beta 3 and platelet activation.

The platelet integrin, glycoprotein IIb-IIIa (GPIIb-IIIa), is a calcium-dependent heterodimer that binds fibrinogen, von Willebrand factor, and fibronectin after platelet activation. We examined GPIIb-IIIa alone and bound to these ligands by electron microscopy after rotary shadowing with platinum/tungsten. We found, as observed previously, that in the presence of detergent and 2 mM Ca2+, GPIIb-IIIa consists of an 8 x 12-nm globular head with two 18-nm flexible tails extending from one side. We also found that in the presence of EDTA, GPIIb-IIIa dissociates into two similar comma-shaped subunits, each containing a portion of the globular head and a single tail. Using monoclonal antibodies to GPIIb, GPIIIa, and the GPIIb-IIIa heterodimer, we found that the tails contained the carboxyl termini of each subunit, while the nodular head was composed of amino-terminal segments of both subunits. Electron microscopy of GPIIb-IIIa bound to fibrinogen revealed a highly specific interaction of the nodular head of GPIIb-IIIa with the distal end of the trinodular fibrinogen molecule and with the tails of GPIIb-IIIa extended laterally at an angle of approximately 98 degrees with respect to the long axis of fibrinogen. When a GPIIb-IIIa was bound to each end of a single fibrinogen, the tails were oriented to opposite sides of fibrinogen, enabling fibrinogen to bridge two adjacent platelets. Electron microscopy of GPIIb-IIIa bound to fibronectin revealed GPIIb/IIIa-binding sites approximately two-thirds of the distance from the amino terminus of each end of the fibronectin molecule, while GPIIb-IIIa was found to bind to von Willebrand factor protomers along a rod-like region near the central nodule of the molecule.

Angiogenesis in vitro, the formation of capillary-like structures by cultured endothelial cells, is associated with changes in the expression of several extracellular matrix proteins. The expression of SPARC, a secreted collagen-binding glycoprotein, has been shown to increase significantly during this process. We now show that addition of purified SPARC protein, or an N-terminal synthetic peptide (SPARC4-23), to strains of bovine aortic endothelial cells undergoing angiogenesis in vitro resulted in a dose-dependent decrease in the synthesis of fibronectin and thrombospondin-1 and an increase in the synthesis of type 1-plasminogen activator inhibitor. SPARC decreased fibronectin mRNA by 75% over 48 h, an effect that was inhibited by anti-SPARC immunoglobulins. Levels of thrombospondin-1 mRNA were diminished by 80%. Over a similar time course, both mRNA and protein levels of type 1-plasminogen activator inhibitor (PAI-1) were enhanced by SPARC and the SPARC4-23 peptide. The effects were dose-dependent with concentrations of SPARC between 1 and 30 micrograms/ml. In contrast, no changes were observed in the levels of either type I collagen mRNA or secreted gelatinases. Half-maximal induction of PAI-1 mRNA or inhibition of fibronectin and thrombospondin mRNAs occurred with 2-5 micrograms/ml SPARC and approximately 0.05 mM SPARC4-23. Strains of endothelial cells that did not form cords and tubes in vitro had reduced or undetectable responses to SPARC under identical conditions. These results demonstrate that SPARC modulates the synthesis of a subset of secreted proteins and identify an N-terminal acidic sequence as a region of the protein that provides an active site. SPARC might therefore function, in part, to achieve an optimal ratio among different components of the extracellular matrix. This activity would be consistent with known effects of SPARC on cellular morphology and proliferation that might contribute to the regulation of angiogenesis in vivo.

Soluble fibrinogen binding to the glycoprotein IIb-IIIa complex (integrin alpha IIb beta 3) requires platelet activation. The intracellular mediator(s) that convert glycoprotein IIb-IIIa into an active fibrinogen receptor have not been identified. Because the lipid composition of the platelet plasma membrane undergoes changes during activation, we investigated the effects of lipids on the fibrinogen binding properties of purified glycoprotein IIb-IIIa. Anion exchange chromatography of lipids extracted from platelets exposed to thrombin or other platelet agonists resolved an activity that increased fibrinogen binding to glycoprotein IIb-IIIa. A monoester phosphate was important for activity, and phosphatidic acid coeluted with the peak of activity. Purified phosphatidic acid dose-dependently promoted a specific interaction between glycoprotein IIb-IIIa and fibrinogen which possessed many but not all of the properties of fibrinogen binding to activated platelets. Phosphatidic acid appeared to increase the proportion of fibrinogen binding-competent glycoprotein IIb-IIIa complexes without altering their affinity for fibrinogen. The effects of phosphatidic acid were a result of specific structural properties of the lipid and were not mimicked by other phospholipids. Lysophosphatidic acid, however, was a potent inducer of fibrinogen binding to glycoprotein IIb-IIIa. These results demonstrate that specific lipids can affect fibrinogen binding to purified glycoprotein IIb-IIIa and suggest that the lipid environment has the potential to influence fibrinogen binding to its receptor.

Four polymorphic variants of the platelet receptor for von Willebrand factor, glycoprotein Ib, have been described that differ in molecular weight on sodium dodecyl sulfate-polyacrylamide gels (Moroi, M., Jung, S. M., and Yoshida, N. (1984) Blood 64, 622-629). A recent report localized the polymorphic site to the heavily O-glycosylated region of the glycoprotein Ib alpha-chain known as the macroglycopeptide (Meyer, M., and Schellenberg, I. (1990) Thromb. Res. 58, 233-242). This region contains several tandem repeats of a mucin-like sequence, which appeared to be a likely site for polymorphic variation. We amplified genomic DNA corresponding to the macroglycopeptide from 206 individuals from four ethnic groups and identified three length variants based on the migration of the amplified DNA on denaturing polyacrylamide gels. DNA sequencing revealed that the three variants represented four alleles, two of which varied by only one base pair, a difference that did not result in an amino acid change. The three length variants differed in the number of tandem repeats of a 39-base pair sequence that results in perfect duplication of a 13-amino acid sequence that originated within a region flanked by Glu-396 and Thr-411. The smallest isoform contained one such sequence; the next largest, two repeats; and the largest, three repeats. The DNA sequence containing the tandem repeats was flanked by direct repeats typical of the target site duplications found flanking transposed DNA, suggesting a mechanism for acquisition of this region by the primordial glycoprotein Ib alpha precursor. The amino acid sequence of the repeated element that accounts for the polymorphism contained five sites for potential O-glycosylation, which together with the repeated amino acids would result in incremental differences in molecular weight of approximately 6,000 between the different isoforms. The addition of repeats to the macroglycopeptide is predicted to increase the length of this elongated glycosylated region and extend the distance between the ligand-binding domain of glycoprotein Ib and the platelet plasma membrane, an effect that would project the ligand-binding domain farther into the bloodstream. Such a change may alter the susceptibility of platelets to shear-induced activation, a process that requires an interaction between glycoprotein Ib and von Willebrand factor.

P-selectin (CD62), formerly called GMP-140 or PADGEM, is a membrane protein located in secretory storage granules of platelets and endothelial cells. To study the mechanisms responsible for the targeting of P-selectin to storage granules, we transfected its cDNA into COS-7 and CHO-K1 cells, which lack a regulated exocytic pathway, or into AtT20 cells, which are capable of regulated secretion. P-selectin was expressed on the plasma membrane of COS-7 and CHO-K1 cells but was concentrated in storage granules of AtT20 cells. Immunogold electron microscopy indicated that the electron-dense granules containing P-selectin in AtT20 cells also stored the endogenous soluble hormone ACTH. Activation of AtT20 cells with 8-Br-cAMP increased the surface expression of P-selectin, consistent with agonist-induced fusion of granule membranes with the plasma membrane. Deletion of the last 23 amino acids of the 35-residue cytoplasmic domain resulted in delivery of P-selectin to the plasma membrane of AtT20 cells. Replacement of the cytoplasmic tail of tissue factor, a plasma membrane protein, with the cytoplasmic domain of P-selectin redirected the chimeric molecule to granules. We conclude that the cytoplasmic domain of P-selectin is both necessary and sufficient for sorting of membrane proteins into the regulated pathway of secretion.

Thrombospondin (TSP) was demonstrated to inhibit the growth of bovine aortic endothelial cells, an activity that was not neutralized by antibodies to TSP or by other agents that block TSP-cell interactions but that partially was reversed by a neutralizing antibody to transforming growth factor-beta (TGF-beta). Similar to TGF-beta, TSP supported the growth of NRK-49F colonies in soft agar in a dose-dependent manner, which required epidermal growth factor and was neutralized by anti-TGF-beta antibody. Chromatography of a TSP preparation did not separate the TGF-beta-like NRK colony-forming activity from high molecular weight protein. However, when chromatography was performed at pH 11, this activity was dissociated from TSP. These results suggest that at least some growth modulating activities of TSP are due to TGF-beta associated with TSP by strong non-covalent forces. Most of the active TGF-beta released from platelets after degranulation was associated with TSP, as demonstrated by anti-TSP immunoaffinity and gel permeation chromatography. 125I-TGF-beta binds to purified TSP in an interaction that is specific in the sense that bound TGF-beta could be displaced by TGF-depleted TSP but not significantly by native TSP, heparin, decorin, alpha 2-macroglobulin, fibronectin, or albumin. Hence, TGF-beta can bind to TSP, and the complex forms under physiological conditions. Furthermore, TSP-associated TGF-beta is biologically active, and the binding of TGF-beta to TSP may protect TGF-beta from extracellular inactivators.

GMP-140, a receptor for myeloid cells that is expressed on surfaces of thrombin-activated platelets and endothelial cells, is a member of the selectin family of adhesion molecules that regulate leukocyte interactions with the blood vessel wall. Each selectin contains an N-terminal domain homologous to Ca(2+)-dependent lectins and mediates cell-cell contact by binding to oligosaccharide ligands in a Ca(2+)-dependent manner. The mechanisms by which Ca2+ promotes selectin-dependent cellular interactions have not been defined. We demonstrate that purified GMP-140 contains two high affinity binding sites for Ca2+ as measured by equilibrium dialysis (Kd = 22 +/- 2 microM). Occupancy of these sites by Ca2+ alters the conformation of the protein as detected by a reduction in intrinsic fluorescence emission intensity (Kd = 4.8 +/- 0.2 microM). This Ca(2+)-dependent conformational change exposes an epitope spanning residues 19-34 of the lectin domain that is recognized by a monoclonal antibody capable of blocking neutrophil adhesion to GMP-140 (half-maximal antibody binding at approximately 20 microM Ca2+). Furthermore, a synthetic peptide encoding this epitope, CQNRYTDLVAIQNKNE, inhibits neutrophil binding to GMP-140. Mg2+ also alters the conformation of the protein, but not in a manner that will support leukocyte recognition in the absence of Ca2+. There is a strong correlation between the Ca2+ levels required for neutrophil adhesion to GMP-140, for occupancy of the two Ca(2+)-binding sites, for the fluorescence-detected conformational change, and for exposure of the antibody epitope in the lectin domain. We conclude that binding of Ca2+ to high affinity sites on GMP-140 modulates the conformation of the lectin domain in a manner that is essential for leukocyte recognition.

To assess the individual contributions of the platelet glycoprotein (GP) IIb/IIIa receptor and the alpha v beta 3 vitronectin receptor to platelet levels of fibrinogen and vitronectin, we analyzed the platelets from two groups of Glanzmann thrombasthenic patients: Iraqi-Jews, whose platelets lack both receptors, and Arab patients in Israel, whose platelets lack GPIIb/IIIa, but have normal or increased numbers of alpha v beta 3 vitronectin receptors. The platelets from both thrombasthenic groups had profound deficiencies of fibrinogen, but the defect in the Iraqi-Jewish patients' platelets appeared to be slightly more severe. This finding indicates that GPIIb/IIIa is the major determinant of platelet fibrinogen, presumably acting by receptor-mediated uptake, and that the alpha v beta 3 vitronectin receptor plays little or no role. Arab patients' platelets have normal amounts of platelet vitronectin, whereas Iraqi-Jewish patients' platelets have nearly five times as much vitronectin as control or Arab patients' platelets. To account for these data, we propose a working hypothesis in which vitronectin is synthesized in megakaryocytes and the alpha v beta 3 vitronectin receptor is involved in transport of the protein out of megakaryocytes and/or platelets. Collectively, these observations suggest that in addition to their recognized roles in cell adhesion and in the interaction of cells with extracellular proteins, integrin receptors may be important in protein trafficking into, and perhaps out of, platelets.

Activation of platelets with thrombin and other agonists causes a rapid increase in the phosphorylation of multiple proteins on tyrosine. To identify candidate protein-tyrosine kinases (PTKs; EC 2.7.1.112) that may be responsible for these phosphorylation events, we analyzed the expression of seven Src-family PTKs and examined the association of these kinases with known platelet membrane glycoproteins. Five Src-related PTKs were detected in platelets: pp60SRC, pp60FYN, pp62YES, pp61HCK, and two LYN products of Mr 54,000 and 58,000. The Fgr and Lck PTKs were not detected. Although strict comparative quantification of protein levels was not possible, pp60SRC was detected at higher levels than any of the other kinases. In addition, glycoprotein IV (GPIV, CD36), one of the major platelet membrane glycoproteins, was associated in a complex with the Fyn, Yes, and Lyn proteins in platelet lysates. Similar complexes were also found in two GPIV-expressing cell lines, C32 melanoma cells and HEL cells. Since PTKs appear to be involved in stimulus-response coupling at the plasma membrane, these results suggest that ligand interaction with GPIV may activate signaling pathways that are triggered by tyrosine phosphorylation.

The platelet membrane glycoprotein (GP) IIb-IIIa complex is the receptor for adhesive proteins on activated platelets that mediates platelet aggregation. In the present study, factors affecting the structural stability of the purified GP IIb-IIIa complex and the dissociated subunits were investigated. Purified GP IIb-IIIa was incubated in various Ca2+ concentrations, and the percentage of dissociated subunits was quantitated by sucrose gradient sedimentation. Two Ca(2+)-dependent transitions were observed, one at about 60 microM Ca2+, where half of the complexes became dissociated, and the other at 0.1 microM Ca2+, where half of the dissociated subunits became incapable of reforming heterodimer complexes when higher Ca2+ concentrations were readded. This loss in ability to reform heterodimer complexes was caused primarily by a Ca(2+)-dependent transition in GP IIIa, leading to an apparent unfolding of this subunit, followed by the formation of high molecular weight aggregates. The formation of these aggregates was time- and temperature-dependent and could not be reversed by added Ca2+. Although Mg2+ prevented dissociation of GP IIb-IIIa, it failed to promote reassociation of the dissociated subunits. Based on these findings, conditions were developed for the preparation of dissociated GP IIb and GP IIIa such that 70% of the subunits remained functional in that they retained the ability to reform heterodimer complexes.

One of the responses of platelets to stimulation is activation of intracellular calpain (the Ca(2+)-dependent protease). Previously, we have shown that activation of calpain in platelets is involved in the generation of platelet procoagulant activity. Because procoagulant activity is present on the microvesicles that are shed from activated platelets, in this study we examined whether calpain is involved in the shedding of microvesicles. Platelets were incubated with the physiological agonists collagen or thrombin. The extent of activation of calpain correlated positively with the amount of procoagulant-containing microvesicles that formed, and the shedding of procoagulant-containing microvesicles was inhibited by calpeptin, MDL, and EST (E-64-d), three membrane-penetrating inhibitors of calpain. The protein composition of the microvesicles shed from aggregating platelets was similar to that of microvesicles shed by platelets in which the association of the membrane skeleton with the plasma membrane had been disrupted by incubation of platelets with dibucaine or ionophore A23187. Furthermore, like microvesicles shed from dibucaine- or ionophore A23187-treated platelets, those shed from the aggregating platelets possessed procoagulant activity. These results are consistent with the possibility that activation of calpain in aggregating platelets causes the shedding of procoagulant-containing microvesicles. We suggest that the shedding of microvesicles results from the calpain-induced hydrolysis of the platelet membrane skeleton.

The glycoprotein Ib/IX complex on platelets is responsible for the first stage of haemostasis as an essential component in the primary adhesion of platelets to damaged vessel walls. Glycocalicin is the extracellular part of platelet glycoprotein Ib alpha and contains the von Willebrand factor and thrombin binding sites. Disulphide bonds are implicated in the von Willebrand binding site and studies with peptides point towards a region of glycocalicin with four cysteines as containing the binding sites for both von Willebrand factor and thrombin. The position and linkage of these two disulphide bonds are now determined to be 209-248 and 211-264 and the relevance of this double-loop structure for glycoprotein Ib/IX function is discussed.

Sixty-two snake venoms were screened to identify those which specifically inhibit the adhesive protein binding function of the glycoprotein (GP) IIb-IIIa complex, the receptor-mediating platelet aggregation. Although 52 of these venoms inhibited GPIIb-IIIa, only one of these, from the southeastern pigmy rattlesnake, Sistrurus m. barbouri, was specific for GPIIb-IIIa versus other integrins. The peptide responsible for this activity, termed barbourin, was sequenced and found to be highly homologous to other peptides of the viper venom GPIIb-IIIa antagonist family but was the first member which did not contain the Arg-Gly-Asp (RGD) amino acid sequence, believed to be required for inhibition of receptor function. Instead, barbourin contains the sequence, Lys-Gly-Asp (KGD). The conservative Lys for Arg substitution appears to be the sole structural feature which imparts integrin specificity to barbourin, since venom peptide analogs with Lys substitutions were also specific for GPIIb-IIIa. Thus, barbourin represents a new structural model useful for designing potent and GPIIb-IIIa-specific compounds that may have therapeutic value as platelet aggregation inhibitors.

Platelets respond to a large number of stimuli by undergoing complex biochemical and morphological changes. These changes are involved in physiological processes including adhesion, aggregation, and coagulation. Platelet activation produces membrane alterations that can be recognized by monoclonal antibodies. In this report we describe a novel activation-dependent protein recognized by a monoclonal antibody, JS-1. The platelet glycoprotein was designated p-155 according to its apparent reduced molecular weight, p-155 exists in the native state as varying sized, large multimers held together by disulfide bonds. p-155 is released upon platelet activation and binds to the activated platelet surface. Although p-155 and platelet glycoprotein Ia migrate similarly on reduced sodium dodecyl sulfate-polyacrylamide gel electrophoresis, immunodepletion and isoelectric focusing distinguished p-155 from glycoprotein Ia. p-155 differed from von Willebrand factor and from thrombospondin in its reduced molecular weight. Additionally, immunoblotting of immunoprecipitated p-155 with antisera to von Willebrand factor and to thrombospondin confirmed the unique identity of p-155. Evidence for a soluble, nonintegral membrane-associated protein was obtained by Triton X-114 phase separation studies, membrane elution studies, and by the demonstration of the protein in the aqueous phase of platelet releasate. Both radioimmunoprecipitation and direct binding techniques demonstrated the activation-dependent nature of p-155. The protein could not be detected in other blood cells, endothelial cells, HEL cells, liver, or in plasma. The functional role of p-155 in platelets is not yet known.

In this study, the question of whether glycoprotein Ib (GPIb) mediates both high and moderate affinity pathways of alpha-thrombin-induced platelet activation was examined. Flow cytometric studies, using a panel of monoclonal antibodies (MoAbs), showed that Serratia marcescens protease treatment removed greater than 97% of the glycocalicin portion of GPIb but did not affect the changes in the expression of GPIX or GMP-140 that were induced by high concentrations of alpha-thrombin (10 nmol/L). However, Serratia treatment almost completely abolished the increase in platelet surface GMP-140 induced by low concentrations of alpha-thrombin (0.5 nmol/L) and diminished the downregulation of platelet surface GPIX by 60.9% +/- 5.6% (mean +/- SEM, n = 3). When present in 20-fold molar excess, an MoAb directed against the alpha-thrombin/von Willebrand factor (vWf) binding domains of GPIb completely blocked the ristocetin-dependent binding of vWf to platelets but inhibited only to about 50% the binding of alpha-thrombin and the activation-dependent binding of vWf. In platelets treated with Serratia marcescens protease to remove GPIb, a concentration of this MoAb 16,000-fold in excess of the maximum possible remaining copies of GPIb failed to inhibit platelet activation by alpha-thrombin. These studies demonstrate that activation of intact platelets by alpha-thrombin proceeds by both GPIb-dependent and GPIb-independent mechanisms.

Granule membrane protein (GMP-140), also known as platelet activation-dependent granule-external membrane (PAD-GEM) is an integral membrane glycoprotein that is expressed on the platelet surface following degranulation. GMP-140, also expressed by endothelial cells, is part of a new family of cell adhesion molecules (selectins) related to the endothelial leukocyte adhesion molecule (ELAM-1) and to the lymphocyte homing receptors in humans (Leu-8/TQ1) and in mouse (gp90MEL-14). The role of GMP-140 in platelet functions remains to be elucidated. In this study, a monoclonal antibody, LYP20, was raised against GMP-140. LYP20, directed against a disulphide bridge-dependent epitope, significantly binds to thrombin-stimulated platelets (12,200 +/- 1,184 bound molecules/platelet, kd = 5.0 +/- 0.61 nmol/L) compared with controls (2,400 +/- 266 molecules/platelet, kd = 2.3 +/- 0.54 nmol/L) and inhibits collagen or thrombin-induced aggregation of washed platelets or platelets in platelet-rich plasma. In addition, LYP20 inhibits rosetting of thrombin-activated platelets to U937 cells. These results strongly suggest that GMP-140 plays an important role in platelet aggregation and platelet interaction with other blood cells.

The interaction of platelet membrane glycoprotein (GP) Ib-IX complex with the cytoplasmic membrane skeleton is potentially of major importance in regulating platelet function. Indirect evidence suggested that this interaction is mediated by actin-binding protein, but it is not known whether GP Ib-IX and actin-binding protein associate directly. To examine more closely the nature of this association, purified GP Ib-IX complex was specifically bound and oriented on the surface of impermeable polymer beads via a monoclonal antibody, AK 2, directed against the extracytoplasmic domain of GP Ib alpha (glycocalicin). Binding was specific since 1) it was abolished by excess unlabeled actin-binding protein; 2) there was no detectable specific binding of radiolabeled actin-binding protein to beads coated with glycocalicin, the major extracytoplasmic proteolytic fragment of GP Ib alpha; and 3) unlike actin-binding protein, there was no specific binding of bovine serum albumin or human platelet vinculin to the GP Ib-IX complex-coated beads. Binding of actin-binding protein to the GP Ib-IX complex-coated beads, but not to the glycocalicin-coated beads, was saturable and reversible (apparent Kd = 1 x 10(-7) M). These experiments provide direct evidence that actin-binding protein can bind to the cytoplasmic domain of a membrane glycoprotein. Because actin-binding protein is found submembranously in cells other than the platelet, it is possible that this protein may link actin filaments to the plasma membrane in those cells.

The respiratory burst of blood neutrophils has a critical role in the destruction of microorganisms and tissue damage in inflammation. Neutrophils adhere in a dose-dependent fashion to granule membrane protein 140 (GMP140), a member of the LEC-CAM (lectin/epidermal growth factor/complement-binding domain cell adhesion molecule) family of adhesion proteins when it is immobilized onto plastic surfaces. Adherence to GMP140 was associated with less superoxide anion generation than adherence to other surfaces, an effect that is especially remarkable after activation of neutrophils with tumor necrosis factor alpha, an agent that on other surfaces promotes adhesion and spreading. However, on GMP140 the cells fail to spread and instead remain rounded and refractile. Neutrophils adhering to GMP140 were also deficient in superoxide anion generation to formylmethionylleucylphenylalanine. Furthermore, fluid-phase GMP140 also inhibited the superoxide generation by neutrophils stimulated by tumor necrosis factor alpha. The effect of GMP140 was reversible by washing and was inhibited by anti-GMP140 Fab antibody. GMP140 appears to be a natural antiinflammatory molecule that may prevent the inappropriate activation of neutrophils in the circulation.