Search

We report the results of an immunogold electron microscopical analysis on microfibrils from the arterial subendothelium showing that thrombospondin (TSP) is present on 40 nm-diameter structures joining 8-10 nm-diameter microfibrils containing fibrillin. They differ from type VI collagen which forms 3-5 nm-diameter microfibrils. TSP containing microfibrils (TSP-MF) extracted from human umbilical arteries did not contain fibrillin or type VI collagen. Blood platelet interactions with TSP-MF were not modified by anti-fibrillin or anti-type VI collagen antibodies. In situ, vWF was bound to cross-linked microfibrils, at the level of their 40 nm junction, and a double-labeling with the anti-thrombospondin and anti-vWF antibodies was observed. In vitro, vWF binding to TSP-MF was not inhibited by anti-fibrillin or anti-type VI collagen antibodies. These results suggest a structural and functional heterogeneity of microfibrils and emphasize the role of TSP-MF in the thrombogenicity of the subendothelium.

We examined the effects of proinflammatory cytokines on the expression of two extracellular matrix proteins, e.g., thrombospondin (TSP) and fibronectin (FN) b cultured human umbilical vein endothelial cells (HUVECs). Treatment of HUVECs with human recombinant interleukin-1 beta (IL-1 beta) or human tumor necrosis factor-alpha (TNF-alpha) caused a time-and dose-dependent decline in TSP production whereas FN production was not modified. At low concentrations, IL-1 beta and TNF-alpha in combination ha a greater effect than either agent alone. Interferon-gamma (IFN-gamma) was without effect. The decline in TSP synthesis resulted in a decreased secretion of this glycoprotein into the extracellular matrix. Endothelial cell monolayers cultured on porous filters were used to study the polarity of TSP secretion. Approximately two thirds of the synthesized protein was secreted to the apical side medium and one third to the basal side medium and both types of secretion were inhibited to a similar extent by cytokine treatment. Immunoprecipitation experiments revealed no apparent degradation of secreted TSP, either in the apical or in the basal compartment. Treatment of HUVECs with lL-1 beta, either alone or in combination with TNF-alpha, had no significant effect on the steady-state TSP mRNA levels, suggesting a posttranscriptional regulation. Our results indicate that IL-1 beta decreasing TSP deposition and suggest different regulatory mechanisms for the expression of various secreted proteins by endothelial cells.

We report herein the characterization of a mouse monoclonal antibody (Mab) raised against the recombinant NH2-terminal heparin-binding domain (rHBD) of human endothelial cell thrombospondin (TSP). The antibody, a IgG1 (kappa), hereafter referred to as V58A4, reacted with two rHBD, TSPN18 and TSPN28 (i.e. 18 kDa and 28 kDa, respectively) with an affinity constant of 1.33 x 10(-8) M. However, V58A4 failed to recognize native or deglycosylated forms of TSP purified from platelets or endothelial cells, as well as a 25-30 kDa HBD fragment produced by limited proteolysis of native TSP. In contrast, Mab V58A4 was shown to react with larger HBD fragments (50-60 kDa) that were present in platelet or endothelial cell extracts and could be retained on a heparin-Sepharose column at low salt concentrations. These fragments also reacted with MA-II, a mouse Mab (IgG1), which recognizes both rHBD and HBD as well as intact TSP. Thus, V58A4 Mab appears to selectively recognize naturally occurring HBD fragments of TSP and may thus prove to be useful for detecting TSP proteolysis in situ under various physiopathological conditions.

We present an easy method in which the adhesion of platelets to collagen or to MFs was measured in a spectrophotometer, after an incubation of hypercitrated platelet rich plasma (PRP), or of a platelet suspension, with an inducer, followed by the filtration of non adhering platelets through translucent Isopore membrane filters (pore diameter = 5 microns). The adhering platelets, which are retained on the filter, were stained by Coomassie blue to quantify the adhesion by the simple reading of the O.D. 580 nm of the stained platelets which appear as a blue spot on the translucent membranes.

The role of thrombospondin on the adhesion of endothelial cells in primary culture was studied using a serum-free defined medium or thrombospondin-depleted fetal bovine serum. Under these conditions, only 6% of the cells adhered to gelatin-coated dishes, whereas cells adhering to gelatin in the presence of normal fetal bovine serum were considered as 100% adhesion. The percentage of cells attached to fibronectin or thrombospondin-coated dishes in thrombospondin-depleted serum was 66 and 32%, respectively. The addition of purified platelet thrombospondin to thrombospondin-depleted serum increased the adhesion of endothelial cells to gelatin and to thrombospondin, up to 32 and 59%, respectively, and restored the attachment to fibronectin to the same extent as that observed in the presence of normal serum. In contrast to the attachment, the spreading of the adhering cells was not further influenced by the addition of soluble thrombospondin. Subcultured cells did not require any protein for adhering to gelatin substrata. These observations indicate that thrombospondin plays a major role in the adhesion of endothelial cells in primary culture.

This short review summarizes the present state of our knowledge concerning the mechanisms of the interaction between subendothelial microfibrils and blood platelets. Microfibrils associated with elastin and with basement membranes have been shown to be able to promote platelet adhesion and subsequent activation and aggregation. A 128 kDa thrombospondin (TSP)-like glycoprotein of microfibrils (GP 128) is involved in their reactivity towards platelets, as established from specific inhibition by antibodies against GP 128 and against TSP of the microfibril-platelet interaction. This interaction only occurs in the presence of plasma von Willebrand factor (vWF). VWF effectively binds to microfibrils; this binding does not implicate GP 128, but a 97 kDa protein of microfibrils not yet characterized. In addition, from data based on inhibition of platelet adhesion and aggregation by monoclonal antibodies against platelet membrane glycoprotein Ib (GP Ib), it appears that GP Ib, a platelet membrane receptor for vWF, acts in the vWF dependent microfibril-platelet interaction as a receptor for VWF bound to protein 97 kDa in microfibrils. A tentative model is presented to illustrate the complex mechanism of the microfibril-platelet interactions: recognition of vWF bound to microfibrils by GP Ib would be the first step, which would then allow subsequent binding of GP 128 to its receptor on the platelet membrane.

We used antisera directed against human platelet thrombospondin (TSP) and microfibril-associated GP 128 to localize the presence of these glycoproteins in fixed sections of human placenta or porcine arteries and skin by immunogold labeling, using electron microscopy. These two antibodies reacted with both human and porcine tissues and always recognized the same structures. In all three tissues the antibodies were associated with the basement membranes and, more precisely, with the microfibrillar structures present at the junction between the basement membrane and the adjacent connective tissue. This localization indicates that GP 128 and TSP are associated with the microfibrils, and suggests their possible role in the attachment of basement membrane to the connective tissue meshwork. Their presence in microfibrils associated with the subendothelial basement membrane in arteries may be important in regard to the thrombogenicity of the subendothelium since, after an endothelial lesion, they may be directly accessible to blood platelets.

We developed an ELISA test to measure the binding of plasma von Willebrand factor (vWF) to arterial microfibrils and compared this interaction to the binding of vWF to collagen under the same conditions. We found that vWF binds to microfibrils in a similar manner as it binds to collagen: the binding was independent of the presence of cations, and temperature of incubation and was displaced by 1 M NaCl. Using purified 125I-vWF we showed that the binding was saturable and could be displaced by cold vWF in excess. Using immunoblotting we showed that vWF binds to a 97 kDa protein present in the microfibrils different from the 128 kDa thrombospondin-like structure (GP 128) which in microfibrils is known to interact with blood platelets. These results indicate that in the subendothelium, microfibrils bind plasma vWF and this reinforces the thrombogenic role of these structures.

This study analyses early biochemical events in collagen-induced platelet activation. An early metabolic event occurring during the lag phase was the activation of PtdIns(4,5)P2-specific phospholipase C. Phosphatidic acid (PtdOH) formation, phosphorylation of P43 and P20, thromboxane B2 (TXB2) synthesis and platelet secretion began after the lag phase, and were similarly time-dependent, except for TXB2 synthesis, which was delayed. Collagen induced extensive P43 phosphorylation, whereas P20 phosphorylation was weak and always lower than with thrombin. The dose-response curves of P43 phosphorylation and granule secretion were similar, and both reached a peak at 7.5 micrograms of collagen/ml, a dose which induced half-maximal PtdOH and TXB2 formation. Sphingosine, assumed to inhibit protein kinase C, inhibited P43 phosphorylation and secretion in parallel. However, sphingosine was not specific for protein kinase C, since a 15 microM concentration, which did not inhibit P43 phosphorylation, blocked TXB2 synthesis by 50%. Sphingosine did not affect PtdOH formation at all, even at 100 microM, suggesting that collagen itself induced this PtdOH formation, independently of TXB2 generation. The absence of external Ca2+ allowed the cleavage of polyphosphoinositides and the accumulation of InsP3 to occur, but impaired P43 phosphorylation, PtdOH and TXB2 formation, and secretion; these were only restored by adding 0.11 microM-Ca2+. In conclusion, stimulation of platelet membrane receptors for collagen initiates a PtdInsP2-specific phospholipase C activation, which is independent of external Ca2+, and might be the immediate receptor-linked response. A Ca2+ influx is indispensable to the triggering of subsequent platelet responses. This stimulation predominantly involves the protein kinase C pathway associated with secretion, and appears not to be mediated by TXB2, at least during its initial stage.

The culture medium of bovine aortic endothelial cells contains proteins which inhibit the aggregation of platelets induced by aortic microfibrils but not by type III collagen. From this medium, fibronectin, thrombospondin and a glycoprotein with MW of 128 Kd (GP 128), similar to a glycoprotein described in a microfibrillar extract from bovine aorta were separated by affinity and ion exchange chromatography. GP 128 was further purified by molecular sieve chromatography on SW 3000 column. GP 128 inhibited the aggregation of platelets by microfibrils. This suggests a role of GP 128 in the platelet/subendothelium interaction.

Elastase activities were determined near neutral pH on several specific substrates using platelet-derived preparations mixed with decreasing amounts of leukocytes. Activities were extrapolated to zero leukocyte content enabling the estimation of intrinsic platelet elastase activity. In contrast to human leukocyte elastase, metal chelating agents inhibited partly the elastase activity of the platelet extract and soybean trypsin inhibitor did not modify its activity. Serine active site titrants (phenylmethane sulfonyl fluoride) as well as acetyl-di-L-alanyl-L-propyl-L-valine chloromethylketone completely abolished the activity of platelet lysates. The platelet protease was purified from Triton X-100 platelet lysates. No cross-reactivity could be demonstrated by immunoelectrophoresis with either porcine pancreatic elastase or human leukocyte elastase using monospecific antisera. Applying gel electrophoresis, most of the elastase activity of the platelet protease migrated towards the anode, whereas the pancreatic and leukocyte elastases migrated towards the cathode. The anionic character of the platelet enzyme might explain its capacity to degrade better elastin treated with cationic detergents in contradistinction to other elastases which act better on anionic detergent-treated elastins.

We have developed a monospecific antiserum directed against a major glycoprotein in the elastin-associated microfibrils with an apparent molecular mass of 128 kDa (GP 128). When immunoblotting or enzyme-linked immunosorbent microassay was used, its IgGs recognized thrombospondin in a platelet lysate, but did not react with several basement-membrane-derived macromolecules, nor with plasma fibronectin. Similar patterns of immunofluorescence and immunoperoxidase were found after incubation of endothelial cells with either anti-GP 128 or anti-(platelet thrombospondin) IgGs. Both antibodies inhibited the microfibrils- and thrombin-induced platelet aggregation, and were without effect on the aggregation by other inducers. These results confirm that there is an antigenic homology between GP 128 and thrombospondin.

Platelet prothrombin converting activity was measured in a system using washed human platelets and purified coagulation factors Xa, Va and prothrombin. Exposure of platelet prothrombin converting activity evoked by collagen or the combined action of collagen and thrombin was effectively inhibited when a collagen derived octapeptide was added prior to platelet activation. Half maximal inhibition of prothrombinase activity of platelets stimulated by collagen plus thrombin- or collagen alone was obtained at 0.9 mM and 0.5 mM octapeptide, respectively. This suggests a modifying effect of thrombin on the platelet-collagen interaction. Octapeptide either alone or in combination with thrombin was unable to enhance platelet procoagulant activity. The increased prothrombin converting activity seen upon treatment of platelets with ionophore A23187 was not affected by octapeptide, added either before or after treatment with ionophore. It is concluded that octapeptide specifically interferes with the platelet-collagen interaction required to generate a procoagulant surface which enhances the rate of thrombin formation.

A synthetic octapeptide derived from type III collagen which specifically inhibits the activation and aggregation of platelets by collagen without affecting their adhesion was assayed on the collagen and ADP dependent fibrinogen binding to platelets. With 20 micrograms/ml collagen, the octapeptide (6 mM) inhibited by 68% the fibrinogen binding: this inhibition was correlated (p less than 0.01) to a decrease in the velocity of aggregation, suggesting that the fibrinogen binding might influence this parameter. The octapeptide did not affect the ADP-induced platelet aggregation and fibrinogen binding. This indicates that the octapeptide does not inhibit the binding of fibrinogen to its receptor directly, but interferes with some step(s) preceding the collagen-induced expression of the fibrinogen receptor.

The interaction of platelets with collagen involves short aminoacid sequences which recur along the fibres. Platelet aggregation by collagen and serotonin release is inhibited by a synthetic octapeptide LYS-PRO-GLY-GLU- PRO-GLY-PRO-LYS- derived from type III collagen. In contrast, this octapeptide inhibits only weakly the retention of platelets labelled with 111Indium to collagen, suggesting that it has a limited effect on platelet adhesion. Preincubation of the octapeptide with platelets inhibits the rise of cAMP level caused by activating adenylate cyclase by various concentrations of PGI2. The octapeptide at 5 mM reverses the inhibition by PGI2 of the adhesion of platelets to collagen. These results suggest that the octapeptide affects the intrinsic activity (manifested as platelet aggregation and secretion) more than the recognition of collagen by its receptor (manifested by adhesion).

The effect of carbacyclin, a chemically stable analogue of prostacyclin (PGI2), on the adhesion of platelets to collagen has been examined. The compound was compared to PGI2 which is unstable and rapidly hydrolysed to the inactive derivative, 6-oxo-PGF 1 alpha. The adhesion of 111Indium-labelled human platelets to collagen in the absence of platelet aggregation and secretion was measured. The cAMP level in the platelets was also monitored. Both PGI2 and carbacyclin inhibited platelet-collagen adhesion and caused a rise in the platelet cAMP level. Carbacyclin was approximately 15-fold less effective than PGI2, however, its effect was longer lasting, remaining constant for at least 30 minutes.

Biochemical methods involving chemical and enzymatic cleavage of type III collagen (characteristic of arterial subendothelium) have led to the identification of a nonapeptide common to different overlapping collagen fragments capable of interacting with platelets. This nonapeptide has been synthesized; it specifically inhibits platelet aggregation and secretion of endogenous platelet serotonin by type III collagen from which it originates. Its effect on platelet adhesion is moderate. In view of its specificity in the platelet-collagen interaction, its use as a mean of preventing thrombosis can be envisaged.

The interaction of human blood platelets with collagenase-treated rabbit subendothelium was studied by histochemical ultrastructural methods and by morphometric semi-quantitative analysis. Aortas were deendothelialized and incubated: 1) with a highly purified bacterial collagenase whose specificity was controlled; and 2) with the same collagenase followed by chymotrypsin. For histochemical studies, tannic acid, ruthenium red, and peroxidase-labeled Ricinus communis and concanavalin A were used. Electron microscopy showed that after digestion of fibrillar collagen by collagenase, adherent and aggregated platelets were observed on Ricinus communis-, concanavalin A-, and ruthenium red-positive glycoprotein microfibrils. After successive incubation with collagenase and chymotrypsin, the microfibrils disappeared. No platelets were observed on the remnant amorphous elastin. Morphometric analysis confirmed the interaction of platelets with collagenase-treated subendothelium. In addition, glycoproteins were extracted from collagenase-treated rabbit aortas using 5 M guanidine. Using an in vitro quantitative test, significant platelet adhesion to these glycoproteins was observed. Our results show an interaction between platelets and noncollagenic glycoprotein microfibrils.

Previous works demonstrated that microfibrils stimulate blood platelets to aggregate. The present study compares the activation of platelets by human placental and bovine aortic microfibrils and by type III collagen. We studied the morphological changes occurring in in platelets during their activation and aggregation, as well as the kinetics of the release reaction and thromboxane B2 formation. As for collagen, the microfibrils-induced platelet aggregation followed a lag phase, during which progressive emission of pseudopodes and centralization of organelles occurred. Aggregation was associated with secretion of beta-thromboglobulin and adenylic adenylic nucleotides, and with formation of thromboxane B2; it was established that the kinetics of secretion and the aggregation curve were parallel. Microfibrils-induced aggregation was also inhibited by ethylenediamine tetraacetic acid, creatine phosphate-creatine phosphokinase, and aspirin, showing that it was calcium-dependent and required a secretion of ADP and formation of endoperoxide and thromboxane. The response to microfibrils was much more rapid than to collagen; placental microfibrils reacted faster than aortic microfibrils. The requirement of plasma in the microfibrils platelets interaction was confirmed: 10 microliters is the minimal amount of plasma necessary for an aggregation of platelets in 400 microliters of buffer. This fact supports the idea of the existence of two different pathways in the interaction between platelet and the subendothelium, depending on the vascular structure (microfibrils or collagen) involved, even though the sequence of events leading to the formation of an aggregate is similar.

Adult bovine aortic tissue was treated with 6 M guanidinium chloride in the presence of proteinase inhibitors to obtain an extract that was essentially devoid of collagenous components and appeared homogeneous by electron microscopy. When this extract was dispersed by sonication it was found to be a very potent inducer of human platelet aggregation. This interaction required the presence of von Willebrand factor and of its receptor (glycoprotein Ib) on platelet membrane. This was demonstrated by the fact that the aggregation of normal blood platelets resuspended in plasmas deficient in von Willebrand factor was significantly diminished as compared to aggregation in control plasma. Moreover, this aggregation was inhibited by a monoclonal antibody, IgG AN51, to platelet glycoprotein Ib. These studies provide direct biochemical evidence for the existence of a thrombogenic constituent of the vessel wall that is noncollagenous and von Willebrand factor-dependent.

A quantitative method (Sepharose test) was devised to measure the adhesion of blood platelets to fibrillar collagen. [14C]5HT-labeled platelets were isolated from plasma, resuspended in EDTA buffer, and incubated with buffer (control) or with fibrillar collagen for 150 sec at 33 degrees C. The mixtures were then filtered through Sepharose 2B columns. In controls the platelets were rapidly eluted, and this was confirmed after 51Cr labeling. [C]5HT was recovered in two stages: 60% with the platelets and 40% retarded, as free 5HT. After incubation with fibrillar collagen (50 micrograms), platelets were retained with the fibrils on the top of the column, and only free [14C]5HT (released from the platelets) was eluted. The percentage of adhesion depended on the number of platelets, the amount of collagen, its degree of polymerization, and the time of incubation at 33 degrees C. [14C]5HT release was markedly diminished when both incubation and filtration were performed at low temperature. ASA, used either in vitro or in vivo in rabbits, did not change the percentage of adhesion but significantly diminished the total amount of [14C]5HT eluted. This method offers a quantitative and reproducible system for the differentiation of adhesion and release, independent of platelet aggregation.

A collagen-derived octapeptide KPGEPGPK which specifically inhibits the activation of platelets by collagen has been tested for its ability to affect the collagen-induced phosphoinositide breakdown and protein phosphorylations. Collagen produced a transient decrease followed by a rapid resynthesis of [32P]-phosphatidyl 4-5 bisphosphate (PIP2) and 4-mono phosphate (PIP). Octapeptide, at a concentration preventing aggregation but allowing shape change, did not impair the phosphoinositide breakdown, whereas the P43 phosphorylation was strongly inhibited. Higher concentrations of peptide which did not permit any shape change were needed to hinder the PIP2 and PIP decrease. Therefore, the octapeptide appears to affect early events of the collagen-induced platelet activation involving the P43 phosphorylation, independently of its effect on the receptor-stimulated phosphoinositide hydrolysis.

The effect of 3 prostaglandins (PG's) (I2, D2 and E1) on the adhesion of platelets to purified type III collagen has been investigated. A quantitative method for a specific evaluation of the adhesion has been applied and has revealed an inhibition of adhesion by low concentrations (10(-10)M) of PGs added before collagen; the effect varied as a function of the dose of PGs (maximum at 10(-6)M) which also induced an increase in the level of platelet cAMP. The inhibition of adhesion and the elevation of platelet cAMP followed the same time course and were either of short duration (rapid decrease in the induced effects after 15 and 45 seconds in the case of PGE1) or longer lasting (maximum effect maintained for 5 minutes in the case of PGI2 and D2). These effects were potentiated by a phosphodiesterase inhibitor such as theophylline (10(-3)M). The addition of PGs after collagen resulted in a reduction of the enhancement of cAMP, associated with a decrease in the inhibition of adhesion. Moreover, the addition of exogenous cAMP (dibutyryl N6-02' cAMP) induced a comparable inhibition. A correlation between the adhesion of platelets to collagen and the level of either endogenous or exogenous cAMP has been established. The PGs also inhibited the platelet release reaction from the alpha granules (beta TG) and the dense bodies. (5-HT and ADP). A greater inhibition of release than of adhesion was observed for the same doses of PGs added.

Arterial microfibrils contain a 128 Kd collagenase and pepsin resistant glycoprotein (GP 128) essential for their ability to induce platelet aggregation. A previous report (Fauvel F. et al, (1984) Biochem. Biophys. Res. Comm., 123, 114-120) showed that GP 128 and thrombospondin (TSP) synthetized by endothelial cells each inhibited the aggregation of platelets by microfibrils and not by collagen. We used a monospecific antiplatelet TSP IgG in an immunoblotting assay for the identification of a TSP-like structure in untreated, collagenase-treated and pepsin-treated arterial microfibrils. The only constituent recognized in the three samples of microfibrils was GP 128. Fab fragments of this IgG provoked a dose dependent inhibition of the microfibril induced platelet aggregation (50% inhibition with 0.25 mg, 100% inhibition with 1 mg); in contrast, they did not affect collagen induced aggregation. The results indicate that a glycoprotein constituent with a thrombospondin-like antigenicity is involved in the thrombogenic properties of arterial microfibrils.

Two different subendothelial macromolecules have been identified as being thrombogenic: collagen and the microfibrils associated with elastin. The interaction between platelets and collagen involves the binding of platelet membrane receptors by numerous sites repeatedly staggered along a collagen fiber: this explains why the preservation of ordered structures (quaternary and tertiary structures) is so important in the reactivity of collagen towards platelets. In the case of Type III collagen, a nonapeptide has been identified as possibly being part of these repetitive sites. The microfibrils have not yet been characterized, although the biochemical data presently available show that they are acidic glycoproteins resistant to collagenase. Microfibrils extracted from human placenta or bovine aorta induce the aggregation of platelets in a reaction which involves platelet glycoprotein Ib and FVIII/vWF. A general model proposed for explaining platelet adhesion to subendothelium suggests that two different mechanisms should be envisaged depending on the thrombogenic macromolecules (collagen, microfibrils) involved.