Determinants in the cytoplasmic domain of P-selectin required for sorting to secretory granules.

P-selectin is a granule membrane protein of platelets and endothelial cells that is expressed at the plasma membrane after cell activation. To determine which residues in its cytoplasmic tail are important for sorting to storage granules during biosynthesis, we expressed P-selectin mutants in AtT-20, a murine cell line with secretory granules that contain the hormone corticotropin ('ACTH'). Immunofluorescence microscopy of permeabilized cells revealed that wild-type P-selectin and mutants with alanine substitutions at 14 different positions in the cytoplasmic tail were concentrated in the tips of the cellular processes, which contain the majority of corticotropin granules. However, targeting to the cell tips was greatly decreased for Tyr777-->Ala, Tyr777-->Phe, Gly778-->Ala, Phe780-->Ala and Leu768/Asn769-->Ala/Ala mutants. The reduced presence of these mutants in corticotropin granules was confirmed by immunoelectron microscopy. Stimulation of AtT-20 transfectants with 8-Br-cAMP resulted in a significant increase in membrane expression of wild-type P-selectin, but in only a marginal increase in the surface expression of the five mutants. Antibody binding studies with intact and permeabilized cells demonstrated that the percentage of P-selectin that is expressed on the surface of the cells was considerably higher for these mutants than for wild-type P-selectin (6%), ranging from approximately 20% for the Gly778 and Phe780 mutants to 63% for the Leu768/Asn769 mutant. Taken together, these results indicate that Tyr777, Gly778 and Phe780 form part of an atypical tyrosine-based motif, which also requires the presence Leu768 and/or Asn769 to mediate sorting of P-selectin to secretory granules.

Identification of a homozygous single base pair deletion in the gene coding for the human platelet glycoprotein Ib alpha causing Bernard-Soulier syndrome.

Bernard-Soulier Syndrome (BSS) is a hereditary bleeding disorder which is caused by the absence or the dysfunction of the platelet glycoprotein Ib/IX/V (GP Ib/IX/V) complex, the major receptor for von Willebrand factor (vWf). BSS is characterized by the presence of giant platelets that show a reduced binding of vWf. Although BSS is a well-characterized disease, and many cases have been described in the literature, the molecular genetic basis of this disorder has been studied in only a few patients. We have studied the genetic basis of the defect in a BSS patient. Flow cytometric analysis of the platelet membrane glycoproteins revealed a significant decrease or absence of GP Ib alpha on the platelet surface, and low levels of GP V and GP IX. In subsequent immunoprecipitation experiments, we confirmed the presence of GP V (although in significantly decreased amounts) on the platelet surface. These results indicated a defect in the GP Ib alpha chain. Genomic DNA coding for GP Ib alpha was amplified, using the polymerase chain reaction (PCR). Subsequent direct sequence analysis demonstrated a homozygous deletion of T317 resulting in a frameshift deletion and predicting a substitution of Arg for Leu76. This deletion causes a shift in the reading frame, predicting a premature stop codon after 19 altered amino-acids, leading to a severily truncated molecule. The molecular genetic defect found in this patient differed from the mutations observed in three other BSS patients described in the literature. This points to a marked hetereogeneity of this disease.(ABSTRACT TRUNCATED AT 250 WORDS)

Tyrosine phosphorylation of P-selectin in intact platelets and in a disulphide-linked complex with immunoprecipitated pp60c-src.

P-selectin is a 140 kDa membrane glycoprotein found in secretory granules of platelets and endothelial cells where it is rapidly translocated to the plasma membrane upon cell activation. It then functions as a receptor for various types of leucocytes. Metabolic labelling of resting platelets with 32Pi showed that P-selectin is primarily phosphorylated on serine residues, although some tyrosine phosphorylation was observed as well. However, tyrosine phosphorylation of P-selectin was greatly stimulated by treatment with the permeating phosphatase inhibitor, pervanadate. When P-selectin immunoprecipitates were incubated with [gamma-32P]ATP (in vitro kinase assay), a fraction of P-selectin was phosphorylated on its tyrosine residues by a co-precipitated kinase. P-selectin phosphorylated in vitro co-migrated with 140 kDa surface-labelled 125I-P-selectin during SDS/PAGE under reducing conditions. Under non-reducing conditions, however, phosphorylated P-selectin was disulphide-linked to unknown protein(s) in a 205 kDa complex. In vitro kinase assays of the most abundant platelet tyrosine kinase, pp60c-src, demonstrated the presence of similar 140 and 205 kDa phosphorylated proteins in SDS/PAGE under reducing and non-reducing conditions respectively. Extraction and reprecipitation studies with proteins phosphorylated in vitro indicated that P-selectin and pp60c-src form a 205 kDa 1:1 disulphide-linked complex. In the complex, pp60c-src autophosphorylation is inhibited and P-selectin is phosphorylated on tyrosine residues. As protein disulphides in the cytoplasm of intact cells are extremely rare, our results suggest that P-selectin and pp60c-src, which co-localize in platelet dense granules, may be non-covalently associated and spontaneously form disulphide bridges during lysis. In addition, the observed tyrosine phosphorylation of P-selectin in intact platelets suggests that its function might be regulated by phosphorylation by pp60c-src.

P-selectin mediates Ca(2+)-dependent adhesion of activated platelets to many different types of leukocytes: detection by flow cytometry.

Previous studies have shown that thrombin-activated platelets interact through the P-selectin with neutrophils and monocytes. To identify other types of leukocytes capable of such an interaction, eosinophils, basophils, and lymphocytes were isolated from whole blood. Binding of these cells to activated platelets was examined in a double immunofluorescence assay and the results show that activated platelets not only bind to neutrophils and monocytes, but also to eosinophils, basophils, and subpopulations of T lymphocytes. Using monoclonal antibodies (MoAbs) specific for subsets of T cells, we could further demonstrate that the T cells which bind activated platelets are natural killer (NK) cells and an undefined subpopulation of CD4+ and CD8+ cells. All these interactions were dependent on divalent cations and were completely inhibited by an MoAb against P-selectin. Thus, P-selectin mediates the binding of activated platelets to many different types of leukocytes. Studies with leukocytes treated with proteases or neuraminidase have shown that the structures recognized by P-selectin are glycoproteins carrying sialic acid residues. Because the loss of binding of activated platelets to neuraminidase-treated neutrophils was almost complete, but only partial to treated eosinophils, basophils, and monocytes, the latter cell types may have different P-selectin ligands in addition to those present on neutrophils. We found that two previously identified ligands for P-selectin, the oligosaccharides Le(x) and sialyl-Le(x), had little or no inhibitory effect on adhesion of activated platelets to leukocytes and that binding was not inhibited by MoAbs against these oligosaccharides. In addition, there was no correlation between the expression of Le(x) on several cell types and their capacity to bind activated platelets. In contrast, the expression of sialyl-Le(x) on cells was almost perfectly correlated with their ability to bind activated platelets. Thus, while Le(x) cannot be a major ligand for P-selectin, a possible role for sialyl-Le(x) in P-selectin-mediated adhesion processes cannot be dismissed. Finally, activated platelets were found to bind normally to monocytes and neutrophils of patients with paroxysmal nocturnal hemoglobulinuria (PNH) and to neutrophils from which phosphatidyl inositol (PI)-linked proteins had been removed by glycosylphosphatidyl inositol-specific phospholipase C (GPI-PLC) digestion. This suggests that at least part of the P-selectin ligands on these cells are not GPI-anchored.

In vitro effect of plasmin on human platelet function in plasma. Inhibition of aggregation caused by fibrinogenolysis.

BACKGROUND: Plasmin has been reported both to activate platelets and to inhibit platelet functions. The latter effect was thought to be caused by proteolysis of the main membrane glycoproteins. METHODS AND RESULTS: We found that incubation of citrated human platelet-rich plasma with streptokinase (SK) (300 IU/ml) does not produce any detectable activation but leads to a time-dependent inhibition of ADP-induced aggregation accompanied by substantial fibrinogenolysis. These effects were abrogated by previous addition of a plasmin inhibitor, aprotinin. Crossover experiments (SK-treated or control platelets mixed with SK-treated or control plasma) demonstrated that the platelets remained functional and that the aggregation defect was caused by fibrinogenolysis. Further experiments (addition of purified fibrinogen to fibrinogen-depleted plasma with either SK or thrombin) suggested that in addition to the low residual level of fibrinogen, fibrinogen degradation products had an inhibitory effect. Under the same conditions, tissue-type plasminogen activator (t-PA) (3,000 ng/ml) had no effect on platelet aggregation, and plasma fibrinogen was not significantly lowered. The effects on glycoproteins IIb-IIIa of incubation with SK, t-PA, or plasmin were assessed with immunoblots with murine monoclonal antibodies directed against either part of the complex, which is the receptor for fibrinogen. Proteolysis was detected only in the presence of EDTA, a potent chelator of divalent cations. CONCLUSIONS: The incubation of human platelets in citrated plasma with SK concentrations obtained during therapy leads to an aggregation defect that is related to the decrease in fibrinogen, the adhesive protein involved in this function, and to the impeding effect of fibrinogen degradation products on its binding onto platelets but not to an alteration of the corresponding platelet receptor, the heterodimer glycoproteins IIb-IIIa.

Glycoproteins V and Ib-IX form a noncovalent complex in the platelet membrane.

Platelet glycoprotein (GP) V is a Mr 82,000 plasma membrane protein of unknown function that is cleaved by the potent platelet agonist, thrombin, to yield a Mr 69,500 fragment (GPVf1). Platelet GPIb, a disulfide-linked alpha beta heterodimer (Mr 160,000) that forms a noncovalent complex with GPIX (Mr 22,000), functions as the platelet adhesion receptor for surface-bound von Willebrand factor. Association between GPV and GPIb-IX has been suggested by the finding that both proteins are deficient in the Bernard-Soulier syndrome, a bleeding disorder characterized by giant platelets and defective interaction with von Willebrand factor. Here we report that GPV and GPIb-IX are coprecipitated by monoclonal antibodies (mAbs) against GPV, GPIb, or GPIX when platelets are solubilized in the mild detergent, digitonin. Treatment of digitonin immunopreciptates with the nonionic detergent, Nonidet P-40, released GPV from anti-GPIb and anti-GPIX mAb precipitates and GPIb-IX from the anti-GPV mAb precipitate. Removal of the Mr 45,000 amino-terminal part of GPIb alpha by treatment with elastase did not abrogate association of GPV with GPIb-IX, showing that the leucine-rich repeat sequences in GPIb alpha are not required for complex formation. Binding studies with 125I-labeled mAbs showed the presence of 24,370 GPIb-IX complexes and 11,170 molecules of GPV/platelet (n = 5). These data show that the leucine-rich glycoproteins GPV and GPIb-IX form a noncovalent complex in the platelet membrane. GPV may play a role in the interaction of platelets with von Willebrand factor.

Quantification of platelet-bound immunoglobulins of different class and subclass using radiolabelled monoclonal antibodies: assay conditions and clinical application.

Quantification of platelet-bound immunoglobulins (PBIg) with radiolabelled murine monoclonal antibodies (mAbs) has been described only for IgG so far. Here we describe some modifications of this mAb radioimmunoassay (MARIA) and show that by using a panel of radiolabelled specific mAbs it is possible to quantify not only PBIgG but also PBIgG subclasses and PBIgM. Analysis by gel filtration showed that all anti-IgG and anti-IgG-subclass mAbs bound to their respective antigens in a ratio of about 1:1. However, the binding ratio for the anti-IgM Mab could not be established. There was a good correlation between the antibody-density per platelet as determined with the anti-IgG mAb and determined as the sum of the IgG molecules of different subclass per platelet (r = 0.90). Platelet fragments did not interfere in the assay. 89 normal healthy controls had 140 IgG molecules per platelet and bound 269 anti-IgM molecules per platelet (geometric means). In a study on the detection of PBIg in 147 thrombocytopenic patients, it appeared that the MARIA had a sensitivity of 61% and a specificity of 45% for the diagnosis of idiopathic thrombocytopenic purpura (ITP). Both in ITP and in secondary thrombocytopenia (STP), PBIgG1 and PBIgG3 were found more frequently (60% and 61%, respectively) than PBIgG2 and PBIgG4 (13% and 9%, respectively). There was no relation between the amount of total PBIgG or PBIgM and the platelet count in either ITP or STP. Also, if IgG antibodies of only one subclass were found, there was no relation between the severity of the thrombocytopenia and the amount of PBIgG. By applying the MARIA, it is possible to quantify PBIgG, all four PBIgG-subclasses and PBIgM in ITP and STP in a reliable way.

Integrin recognition of different cell-binding fragments of laminin (P1, E3, E8) and evidence that alpha 6 beta 1 but not alpha 6 beta 4 functions as a major receptor for fragment E8.

The involvement of integrins in mediating interaction of cells to well-characterized proteolytic fragments (P1, E3, and E8) of laminin was assessed by antibody blocking studies. Cell adhesion to fragment P1 was affected by mAbs against the integrin beta 1 and beta 3 subunits and furthermore could be prevented completely by a synthetic peptide containing the Arg-Gly-Asp sequence. Because the beta 3 antibody-sensitive cell lines expressed the vitronectin receptor (alpha v beta 3) at high levels, the involvement of this receptor in cell adhesion to P1 is strongly suggested. Integrin-mediated cell adhesion to E3 is of low affinity and was inhibited by antibodies against the integrin beta 1 subunit. In contrast, adhesion of some cell types to E3 was not or only partially sensitive to inhibition by anti-integrin subunit antibodies. Cell adhesion to E8 was blocked completed by integrin alpha 6 or beta 1 antibodies. The alpha 6-specific antibody did not inhibit cell adhesion to E3 or P1. Furthermore, the antibody only blocked adhesion to laminin of those cells that adhered exclusively to the E8 fragment. In addition, expression of alpha 6 beta 1 was closely correlated with the ability of cells to bind to the E8 fragment of laminin. These results indicate that the alpha 6 beta 1 integrin is a specific receptor for the E8 fragment of laminin. Many cell types expressed, instead of or in addition to alpha 6 beta 1 the recently described integrin alpha 6 beta 4. Although the ligand of alpha 6 beta 4 was not identified, it must be different from that of alpha 6 beta 1, because cells that express alpha 6 beta 4, but not alpha 6 beta 1, do not adhere to E8, and cell adhesion to E8 was specifically blocked by beta 1 specific antibodies. In conclusion, the data indicate that distinct integrin receptors belonging to the beta 1 or beta 3 subfamily are involved in adhesion of cells to the various laminin fragments. Adhesion to E3 may also be brought about by other receptor molecules, possibly proteoglycans, not belonging to the integrin family.

Detection of platelet activation with monoclonal antibodies and flow cytometry. Changes during platelet storage.

To investigate whether changes in platelet condition during platelet storage correlate with an altered expression of platelet membrane proteins, the binding of monoclonal antibodies (MoAbs) to fresh platelets was compared with MoAbs' binding to thrombin-activated platelets and to platelets stored as platelet concentrates. The MoAbs included antibodies against the platelet glycoprotein (GP) IIb/IIIa complex and against two activation-dependent antigens, one of which was a component of the internal platelet alpha-granule membrane (GMP 140) and the other of which was a 53-kD protein derived from platelet lysosomes. The binding of MoAbs to platelets fixed with 1 percent paraformaldehyde was measured by flow cytometry. In thrombin-activated platelets, a threefold increase was found in the expression of GP IIb/IIIa over that in fresh platelets. The binding of the activation-dependent MoAbs increased from 2 to 3 percent to 70 to 80 percent of the platelets. Storage of platelet concentrates for 5 days resulted in a 60 percent increase in GP IIb/IIIa expression compared to Day 0 and increased binding of the MoAbs directed against GMP-140 from 3 to 16 percent and against the 53-kD protein from 2 to 8 percent of the platelets, respectively. These changes correlated with modifications in platelet morphology (decrease in swirling), leakage of lactate dehydrogenase, and release of beta-thromboglobulin. These data indicate that platelets become activated and are damaged during the storage of platelet concentrates.(ABSTRACT TRUNCATED AT 250 WORDS)

Decreased stability and structural heterogeneity of the residual platelet glycoprotein IIb/IIIa complex in a variant of Glanzmann's thrombasthenia.

A patient is described with a disturbance of platelet function comparable to that in Glanzmann's thrombasthenia. Platelet aggregation and binding of fibrinogen to the patient's platelets were defective and thrombin-induced clot retraction was absent. The platelet fibrinogen content was only moderately reduced. As measured by monoclonal antibody binding in the presence of divalent cations, the platelets contained about 15% of the normal amount of GPIIb and GPIIIa and only 6% of the normal amount of intact GPIIb/IIIa complex. The residual GPIIb/IIIa complex exhibited a decreased stability as shown by the lack of binding of a complex-dependent anti-GPIIb/IIIa antibody to platelets incubated with ethylene diamine tetraacetic acid (EDTA) at 22 degrees C. Crossed immunoelectrophoresis (CIE) in the presence of divalent cations showed partial dissociation of GPIIb/IIIa as well as the presence of two forms of the residual intact GPIIb/IIIa complex. In addition, both CIE in the presence of the EDTA and two-dimensional sodium dodecyl sulphate (SDS) gel electrophoresis showed the presence of two forms of GPIIb. This form of thrombasthenia is characterized by a defective platelet function, a marked reduction of GPIIb and GPIIIa, decreased stability of the residual GPIIb/IIIa complex and structural heterogeneity of GPIIb.

Human vascular endothelial cells express a membrane protein complex immunochemically indistinguishable from the platelet VLA-2 (glycoprotein Ia-IIa) complex.

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

Laminin receptor on platelets is the integrin VLA-6.

Adhesion of platelets to the subendothelial matrix of an injured vessel wall is an essential step in triggering the formation of a haemostatic plug. Fibronectin, collagen and laminin are three major components of the subendothelial matrix which support platelet adhesion. Receptors for fibronectin and collagen have been identified on platelets and are included in the integrin family. Here we report that adhesion of platelets to laminin is inhibited by a rat monoclonal antibody against the integrin family member, VLA-6. This antibody does not affect platelet adhesion to fibrinogen, fibronectin or to type I and III collagen. Binding to laminin does not require platelet activation and is not inhibited by fibronectin and laminin cell-attachment peptides. Platelet adhesion to laminin is supported by Mn2+, Co2+ and Mg2+, but not by Ca2+, Zn2+ and Cu2+. This cation preference is distinct from that characteristic for other platelet-adhesive glycoproteins.

A monoclonal antibody to the human platelet glycoprotein IIb/IIIa complex induces platelet activation.

The platelet glycoprotein (GP) IIb/IIIa complex functions as the receptor for fibrinogen on activated platelets. The effects of two anti-GPIIb/IIIa monoclonal antibodies on platelet function were studied. These antibodies, 6C9 and C17, recognized different epitopes, which were exclusively present on the undissociated GPIIb/IIIa complex. Whereas C17 inhibited the binding of fibrinogen to platelets and platelet aggregation induced by adenosine diphosphate (ADP) or collagen, 6C9 caused irreversible aggregation of platelets, both in the presence and absence of extracellular fibrinogen. When incubated with unstirred (non-aggregating) platelets, 6C9 induced release of alpha and dense granule-constituents as well as binding of 125I-fibrinogen to platelets. The latter was evidently mediated in part by platelet-derived ADP, since it was inhibited to a large extent by apyrase, the ADP-hydrolyzing enzyme. F(ab')2 fragments of 6C9 did not induce platelet-release reactions but caused (slow) aggregation of platelets in the presence of extracellular fibrinogen. These results indicate that binding of an antibody to a specific site on the platelet GPIIb/IIIa complex may cause fibrinogen-mediated aggregation. The Fc part of the platelet-bound antibody appears to be involved in the induction of platelet release.

The platelet alloantigen Zwa or PlA1 is expressed by cultured endothelial cells.

Recently, the synthesis by cultured human endothelial cells of a membrane protein complex immunologically related to platelet glycoprotein (GP) IIb/IIIa complex was demonstrated. Since platelet GP IIIa is known to carry the platelet alloantigen Zwa or PlA1, studies were performed to establish whether this antigen is also expressed on endothelial cells. The present report describes the results of these studies, which provide evidence for the presence of the Zwa or PlA1 antigen on the surface of cultured human endothelial cells. This evidence is based on the following observations: (1) cultured endothelial cells react with anti-Zwa (PlA1) antibodies as shown by indirect immunofluorescence; (2) two proteins are precipitated by anti-Zwa (PlA1) antibodies from lysates of 125I-labelled endothelial cells with an electrophoretic mobility corresponding with that of GP IIb and IIIa; (3) anti-Zwa (PlA1) reacts specifically, as shown by immunoblotting of sodium-dodecylsulphate polyacrylamide gels of solubilized endothelial cells, with a protein with a mobility similar to that of platelet GP IIIa.

Detection of circulating human platelet fragments by using monoclonal antibodies against platelet glycoprotein IIb-IIIa complex.

A radioimmunoassay (RIA) for the detection of platelets and platelet fragments was developed. A sandwich of two monoclonal antibodies directed against the platelet-specific glycoprotein complex IIb-IIIa (GP IIb-IIIa) was used in this assay. A discontinuous 7.5-20% (v/v) albumin gradient was applied to separate platelets and their fragments of various sizes. In platelet suspensions fractionated in this way, we observed that particles smaller than normal platelets still carried the GP IIb-IIIa antigens. This procedure enabled us to detect platelet-derived particles in platelet-rich plasma from thrombocytopenic patients.

The coding site of chloroplast ferredoxin.

Ferredoxins were isolated and purified from leaves of different species of Nicotiana and Petunia and from spinach leaves. Their spectral properties, degree of homogeneity, and molecular weights were determined. The preparations were further analyzed by polyacrylamide gel electrophoresis of tryptic hydrolysates. This allowed us to distinguish between not only ferredoxins of Nicotiana, Petunia, and spinach, but even ferredoxins of various Nicotiana species. We used the differences in tryptic peptide compositions as phenotypic markers to study the mode of inheritance of chloroplast ferredoxin to see whether the coding site is in the chloroplast or in the nucleus. Analysis of the tryptic peptide composition of ferredoxin from different interspecific hybrids of Nicotiana showed that the characteristics of both parental ferredoxins were present. The results indicate that the primary structure of at least the male ferredoxin is coded for in the nucleus. In some of the hybrids the relative contribution of the male parent appeared to be low, suggesting that the female genome (presumably that part located in the plastome) exerted a dominating influence.