The Thrombomodulin/Protein C/Protein S Anticoagulant Pathway Modulates the Thrombogenic Properties of the Normal Resting and Stimulated Endothelium
Abstract We investigated the role of the thrombomodulin (TM)/protein C/protein S anticoagulant pathway in modulating the thrombogenic properties of the endothelium. Endothelial cells (ECs) were placed in parallel-plate flow chambers and exposed to nonanticoagulated human blood at a venous wall shear rate (50 s−1). Fibrin deposition on resting ECs treated with a control IgG1 was negligible. In contrast, a significant amount of fibrin deposited when TM expression was specifically suppressed by >95% by preincubating ECs with an anti-TM IgG1. Similarly, fibrin deposited on interleukin 1–stimulated ECs, but the fibrin deposition was further increased threefold with anti-TM IgG1. Comparable results were found when ECs were perfused at 650 s−1. When TM surface activity was enhanced by 150% by treating ECs with active phorbol ester (4′-phorbol 12-myristate 13-acetate; PMA), the deposition of fibrin was 30% lower than on ECs not pretreated with PMA. Finally, fibrin deposition on stimulated ECs was significantly higher in 11 untreated patients with well-characterized deficiencies of protein C or S or heterozygous factor V Leiden mutation than in 11 healthy individuals, and it was significantly correlated to basal plasma levels of thrombin-antithrombin complexes. Thus, this study underlines the central role of the TM/protein C/protein S pathway in modulating the thrombogenic status of resting and stimulated ECs and indicates that basal coagulation system markers may be helpful in monitoring patients presenting a disorder of this anticoagulant pathway.
Reprint requests to Y. Cadroy, Laboratoire de Recherche sur l'Hémostase et la Thrombose, Pavillon Ch Lefèbvre, Hôpital Purpan, 31059 Toulouse Cedex, France.
- Received March 18, 1996.
- Accepted June 11, 1996.
The membrane glycoprotein TM is an anticoagulant component of the endothelium of blood and lymphatic vessels and the basis of a major natural anticoagulant system (for review, see Reference 11 ). TM forms a 1:1 stoichiometric complex with thrombin, which becomes less available to act on its substrates, notably platelets and fibrinogen. In addition, TM enhances the rate of protein C activation by thrombin. Activated protein C cleaves and inactivates activated factors V and VIII, thereby exerting a potent negative feedback control on the generation of thrombin. The activity of protein C is substantially potentiated by protein S, which binds to activated protein C and to platelets or EC membranes to form a cell surface–bound complex.1
Experimental and clinical data indicate that the TM/protein C/protein S pathway is important in modulating coagulation. The physiological importance of protein C as an anticoagulant is most dramatically illustrated by the severe thromboembolic disorder that affects individuals with homozygous protein C deficiency in the neonatal period.2 Heterozygous protein C deficiency is also a risk factor for venous thrombosis in adult life.3 Likewise, protein S heterozygous deficiency is strongly associated with venous thrombosis.4 Patients with resistance to activated protein C principally due to an Arg 506→Gln factor V gene mutation, called factor V Leiden, have a risk for developing thrombosis, especially when this defect is homozygous or combined with another risk factor, such as protein C deficiency or estrogen treatment.5 6 Deficiency of TM has not yet been found, but certain lupus anticoagulants that inhibit the function of TM in vitro are associated with thrombosis.7 In addition, a patient with a history of thrombotic disorder and a mutation in the TM gene has been identified recently.8 Conversely, purified and recombinant TM or activated protein C and protein S protect animals against disseminated intravascular coagulation or thrombosis in various experimental models.9 10 11 12 13
One of the main functions of the TM/protein C/protein S pathway is to support the antithrombogenic properties of the endothelium.1 However, besides this system, it is suspected that other components of the endothelium play a role in its antithrombotic properties. These include the expression of heparan sulfate proteoglycans on the EC membrane and the synthesis by the cells of tissue factor pathway inhibitor, tissue plasminogen activator, prostacyclin, and endothelium-derived relaxing factor.14 The relative importance of the TM/protein C/protein S pathway in the antithrombotic properties of the endothelium compared with these other systems is not well known. Likewise, the role of this pathway in protecting the endothelium from promoting thrombogenesis in inflammatory conditions has not been studied. Stimulated by inflammatory mediators, the endothelium changes from an anticoagulant to a procoagulant status15 16 17 and induces thrombus formation when exposed to flowing blood.18 19 20 21 22 This process is primarily dependent on the expression of TF, the major initiator of blood coagulation in vivo.19 22 TF is an integral membrane glycoprotein, which binds to factor VII(a). The TF/VII(a) complex activates factors IX and X and thus promotes the generation of thrombin.1 Stimulation of ECs also results in the downregulation of TM expression,16 but the consequences of this effect on the prothrombogenicity of the cells is not well established.
The aim of the present study was to more specifically determine the role of the TM/protein C/protein S pathway in modulating the respective antithrombotic and prothrombotic properties of resting and stimulated ECs in flowing nonanticoagulated blood at various shear rates. Resting or cytokine-stimulated confluent human umbilical venous ECs were positioned in parallel-plate perfusion chambers and exposed to nonanticoagulated human blood at two wall shear rates, which are characteristic of veins (50 s−1) or medium-sized arteries (650 s−1). In the first part of the study, the role of TM in modulating fibrin and platelet deposition on ECs was investigated by specifically altering its activity on the surface of ECs with the use of appropriate agents. TM activity was either suppressed by treating stimulated ECs with a monoclonal antibody directed against TM or enhanced by pretreating ECs for 48 hours with active phorbol ester.
In a second part, the role of the TM/protein C/protein S pathway in modulating the thrombogenicity of the stimulated endothelium was studied by examining untreated patients with well-characterized deficiencies of protein C or protein S or with resistance to activated protein C due to factor V Leiden mutation. Nonanticoagulated blood from these patients was drawn directly over IL1–stimulated confluent ECs placed in the perfusion chambers for 5 minutes at a venous wall shear rate of 50 s−1. Fibrin deposition on ECs was compared with that obtained in age- and sex-matched healthy individuals. This study was also designed to determine the utility of the ex vivo model in monitoring these patients. Indeed, ≈50% of patients with deficiencies in protein C or protein S or resistance to activated protein C will develop thrombotic events.23 Biological markers of hypercoagulability (eg, F1+2, T-AT, and FPA) may possibly be helpful in monitoring these patients. However, the exact relationship between the levels of activation markers of coagulation and thromboembolic disease is not well known. To clarify this relationship, we compared plasma levels of F1+2, T-AT, and FPA with results obtained with the ex vivo thrombosis model, which may be considered to give a better reflection of the prothrombotic state.
In the first part of the study, perfusion experiments were performed with healthy volunteers. Each set of experiments (treated ECs and control ECs) was performed the same day with the same blood donor and cells obtained from one single culture. In the second part, 11 patients with isolated deficiency of protein C or protein S or with resistance to activated protein C due to factor V Leiden mutation were studied. Seven had a history of thrombosis and 4 were asymptomatic. They were compared with 11 healthy age- and sex-matched volunteers with normal plasma levels of protein C and protein S and no resistance to activated protein C. None of the subjects were taking any drugs known to affect platelet function or coagulation. All subjects gave informed consent to the protocol approved by the Human Subject Committee (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale, Toulouse). The perfusion experiments of patients and of their age- and sex-matched healthy control subjects were performed the same day and with ECs obtained from one single culture. This precaution was taken because the expression of TM and notably of TF of cell cultures obtained from different pools of umbilical cord is known to vary considerably.16
Blood was collected from an antecubital vein by using a 19-gauge butterfly (SFL) in EDTA (reference No. 367652, Becton Dickinson) for DNA studies of factor V gene; in citrate (reference No. 367704, Becton Dickinson) for determination of plasma levels of antithrombin III, protein C, and protein S and of resistance to activated protein C; and in 0°C precooled tubes filled with inhibitors of platelet activation (0.4 mL of sodium citrate, citric acid, theophylline, adenosine, and dipyridamole; Diatube from Stago) and inhibitors of coagulation (0.4 mL of heparin and aprotinin; Assera FPA, Stago) for determination of plasma levels of markers of thrombin formation (F1+2 and T-AT) and thrombin cleavage of fibrinogen (FPA). Except for the EDTA tubes, blood samples were immediately centrifuged at 4°C and 3000g for 30 minutes, and aliquots of plasma were stored at −70°C until assayed. Antithrombin III was measured by a chromogenic substrate method (Coamatic antithrombin, Chromogenix). Protein C and free protein S were measured by a coagulometric method (Staclot protein C and Staclot protein S, Stago). Resistance to activated protein C was measured with purified activated protein C (Enzyme Research Laboratories) as previously described24 25 and the Arg 506→Gln mutation of the factor V gene was searched for by using the method described by Gandrille et al.26 The factor V gene exon 10 was amplified with a modified oligonucleotide (5′-GGTTACTTCAAGGACAAAATACCTGTAAAGCT-3′) to allow the amplified fragment bearing the mutation to be cleaved by the restriction enzyme HindIII. Plasma concentrations of F1+2, T-AT, and FPA were measured by commercially available immunoenzyme assays (Enzygnost F1+2, Enzygnost TAT, Behring, and Assera FPA, Stago, respectively).
Human ECs were isolated from umbilical veins and cultured according to the method of Jaffe et al.27 The cells were grown on gelatin-coated Thermanox coverslips (Miles Laboratories) pretreated with 0.05% glutaraldehyde. The culture medium was composed of RPMI 1640 medium and medium 199 (ATGC Biotechnologie) supplemented with 20% human pooled serum (Institut Jacques Boy). The cells were identified by their typical morphology. Confluence was reached in 5 to 10 days, resulting in 1.25±0.15×105 cells per square centimeter. Primary and secondary cultures were used. Cells were treated according to different protocols. The TM activity at the surface of ECs was depressed by incubating the cells at rest or stimulated for 4 hours with recombinant human IL1 (50 U/mL, Coenzyme) with a mouse monoclonal IgG1 antibody directed against the fifth epidermal growth factor–like domain of TM (3 μg/mL, M617, Dako) for 30 minutes. Control studies were performed with a nonimmune mouse monoclonal IgG1 antibody (3 μg/mL, Dako). In a second set of experiments, TM activity was enhanced by treating the cells for 48 hours with PMA (6 ng/mL, Sigma). The PMA-treated cells were then restimulated for 6 hours with IL1 (50 U/mL) and recombinant human tumor necrosis factor-α (10 ng/mL, Genzyme). Finally, the study of patients with deficiency of protein C or protein S or resistance to activated protein C was performed with ECs stimulated for 4 hours with IL1 (50 U/mL).
Measurement of TF Activity
The TF activity expressed on the apical surface of ECs was measured by a chromogenic assay as previously described.22 The surfaces were washed three times in Tris-buffered saline (50 mmol/L Tris HCl, 120 mmol/L NaCl, 2.7 mmol/L KCl, and 3 mg/mL bovine serum albumin) and incubated with 300 μL of Tris-buffered saline containing purified human factor VII and factor X 5 and 150 nmol/L final concentrations, respectively, from Enzyme Research Laboratories) and CaCl2 (5 mmol/L). After a 30-minute incubation period at 37°C, 250 μL of the supernatant was added to 25 μL (0.2 mmol/L final concentration) of a chromogenic substrate of factor Xa (S2765, Chromogenix), the reaction was stopped after 3 minutes by the addition of 200 μL of 50% (vol/vol) acetic acid, and the amidolytic activity of factor Xa generated was read at 405 nm with a spectrophotometer. The TF activity was expressed in arbitrary units using reference curves determined with a standard human brain TF preparation containing 106 TF-AU/mL (Thromborel S, Behring). The curve was linear up to 2000 TF-AU/mL. The activity was observed to be TF-related, since the Xa generation was fully inhibited in the absence of factor VII and inhibited by 70% to 80% by incubating the ECs with a mixture of two mouse anti-human TF monoclonal antibodies (10 μg/mL, Corvas) for 30 minutes at 37°C.
Measurement of TM Activity
The TM activity expressed on the apical surface of ECs was measured by a chromogenic assay. The EC surface was washed in Tris-buffered saline and incubated with 300 μL of Tris-buffered saline containing purified human protein C (150 nmol/L, Enzyme Research Laboratories), thrombin (1 U/mL, Etablissement de Transfusion Sanguine), and CaCl2 (5 mmol/L). After a 2-hour incubation period at 37°C, 50 μL of the supernatant was added to 450 μL of a chromogenic substrate of activated protein C (S2266, 2 mmol/L final concentration, Chromogenix) and hirudin (25 U/mL, Sigma). The reaction was stopped after 10 minutes by the addition of 200 μL of 50% (vol/vol) acetic acid, and the amidolytic activity of activated protein C generated was read at 405 nm with a spectrophotometer. The TM activity was expressed in arbitrary units using reference curves determined with purified rabbit TM (Stago).
Perfusion experiments were performed with a parallel-plate perfusion chamber device as previously described.21 22 After collection of blood samples for the determination of hematological parameters, nonanticoagulated blood was drawn directly by a peristaltic roller pump (Multiperpex LKB, Pharmacia) from the vein through the infusion set and over the ECs in the perfusion chamber for 3 or 5 minutes. The pump was placed distal to the chamber. Flow rates were 5 and 10 mL/min, which given the cross-sectional dimensions of the blood flow channels correspond to wall shear rates of 50 s−1 (5 mL/min) and 650 s−1 (10 mL/min), respectively. At the end of the 3 or 5 minutes of blood perfusion, RPMI/medium 199 was perfused at the same flow rate for 60 seconds (50 s−1) or 20 seconds (650 s−1) to wash out the blood from the flow channel. For the morphometric examinations, the coverslips were subsequently fixed by a perfusion with 2.5% glutaraldehyde in 0.1 mol/L cacodylate, pH 7.4, for 40 seconds. The coverslip with the thrombotic deposits was then removed from the chamber, immersed in freshly prepared fixation solution at 4°C for 90 minutes, and stored in 0.1 mol/L cacodylate/7% sucrose at 4°C until embedded in epoxy resin (Epon). The fixation procedure was not performed when deposited fibrin was measured by immunological determination.
Morphometric Determinations of Fibrin and Platelet Deposition on ECs
Evaluation of thrombotic deposits was performed on Epon-embedded semithin sections (1 μm thick) stained with toluidine blue and basic fuchsin as previously described.21 22 The sections were prepared at an axial position 1 mm downstream of the coverslip flow inlet and perpendicular to the direction of the blood flow. Standard morphometry, performed by light microscopy at 1000× magnification, was used to quantify the percentage of surface coverage (1) with fibrin, recognized as circular spots on ECs associated or not associated with platelets (percent fibrin) and (2) with platelets associated or not associated with fibrin (percent platelets). The surface coverage with cells was determined and expressed as percent endothelial surface coverage. These morphometric evaluations were performed at 10 μm intervals along the surface by moving the section along an eyepiece micrometer positioned in the ocular of the microscope. Approximately 400 evaluations were performed per section.
Immunological Determination of Fibrin on ECs
Fibrin deposition on ECs was quantified by immunological determination of fibrin degradation products of plasmin-digested thrombi.22 At the end of the perfusion period, the coverslip was immersed in 3 mL of a plasmin (Chromogenix) solution (0.7 CU/mL, in Tris-buffered saline, pH 7.4) for 1 hour at 37°C and with gentle shaking. Fibrin degradation products were measured by using an immunoenzymatic assay (Asserachrom D-Di, Stago). The amount of fibrin deposited was directly determined from the levels of fibrin degradation products expressed as fibrin equivalent unit, as indicated by the manufacturer. This unit corresponds to the quantity of clotted fibrinogen that leads to the observed level of fibrin degradation products. No fibrin (<0.05 μg/cm2) was found on nonperfused coverslips coated with ECs.
Determination of the Procoagulant Activity of ECs During Blood Perfusions
The procoagulant activity of ECs was determined by measuring the plasma levels of F1+2, T-AT, and FPA in blood perfused over ECs.21 22 Blood samples (3.2 mL) were collected at the flow outlet of the chamber between 2.2 and 3 minutes of perfusion for the 3-minute perfusion period and between 4.2 and 5 minutes of perfusion at 50 s−1 and between 4.5 and 5 minutes of perfusion at 650 s−1 for the 5-minute perfusion period. The blood samples at the flow outlet were collected by a syringe pump (Harvard Apparatus) withdrawing at the same flow rate, ie, 5 or 10 mL/min, through a puncture made with a 19 G infusion set at the outlet tubing between the chamber and the roller pump. The roller pump was switched off during the sampling procedure to keep the blood flow constant. Blood samples (3.2 mL) were collected in 0°C precooled syringes prefilled with platelet inhibitors and anticoagulants as indicated above.
Statistical analysis was performed using the PCSM program (Deltasoft). Results were expressed as mean±SE and appropriately compared with the Student's t test or the Mann-Whitney U test. Values of P<.05 were considered significant.
Effect of Suppression of TM Surface Activity on Thrombogenic Properties of Resting and Stimulated ECs
The surface of nonstimulated ECs expressed notable levels of TM activity and very low levels of TF activity (Fig 1⇓). When ECs were incubated with a monoclonal antibody directed against TM for 30 minutes, TF activity remained low, but TM activity was depressed by more than 95% (P<.01). In contrast, the mouse control monoclonal antibody had no effect on TM or TF activities (P>.50). Resting ECs treated with the control or the anti-TM monoclonal antibodies were exposed to nonanticoagulated blood for 5 minutes at a venous wall shear rate of 50 s−1. Fibrin deposition on the surface of resting ECs was negligible with the control monoclonal antibody, confirming the antithrombotic nature of the resting endothelium.14 18 21 However, it increased dramatically with the anti-TM monoclonal antibody (P<.05, Table 1⇓). The procoagulant activity of ECs as determined by measuring the plasma levels of T-AT in blood samples collected immediately distal to the perfusion chamber was also increased with anti-TM IgG1 (P<.05, Table 1⇓). Thus, when TM expression of resting ECs was depressed, the normally nonthrombogenic ECs became procoagulant and prothrombogenic, which indicates that TM does support part of the antithrombotic properties of the resting endothelium.
To determine the role of TM in modulating thrombus formation in inflammatory conditions, ie, on TF-expressing ECs, we stimulated them with IL1 for 4 hours. The TF activity increased 40 times compared with resting ECs, whereas the TM surface activity was decreased by 40% (P<.05, Fig 1⇑), as indicated previously.16 However, when ECs were further incubated with the anti-TM monoclonal antibody, TM surface activity was abolished, whereas TF activity remained comparable. In contrast, the mouse control monoclonal antibody had no notable effect on TM or TF activities on IL1-stimulated ECs. IL1-stimulated ECs were exposed to nonanticoagulated blood for 5 minutes at a wall shear rate of 50 s−1. In presence of control IgG1 and compared with the respective nonstimulated ECs, the amount of fibrin deposited on these ECs increased 5 times (Tables 1⇑ and 2⇓). However, when the TM surface activity was totally abolished with anti-TM IgG1, the fibrin deposition was further increased three times (P<.05). Likewise, the plasma levels of T-AT collected immediately distal to the chamber were doubled with anti-TM IgG1 (P<.05, Table 2⇓).
To determine whether depression of TM activity resulted in an enhancement of platelet deposition at the surface of ECs in addition to increased fibrin deposition, the nature of thrombotic deposits was evaluated by morphometry on Epon-embedded semithin sections (Fig 2⇓). As shown previously,21 22 ECs, regardless of the treatment, covered more than 99% of the coverslip surface at the end of the 5-minute perfusion period, indicating than blood/extracellular matrix interactions were negligible during the perfusion experiment (data not shown). Deposition of platelet and fibrin on ECs are indicated in Table 3⇓. The anti-TM monoclonal antibody increased the fibrin deposition from 5±2% to 38±11% (P<.05), whereas platelet deposition remained low. Thus, the suppression of TM surface activity resulted in a specific increase in fibrin deposition at a wall shear rate of 50 s−1.
We next examined the effect of TM suppression on the thrombogenicity of stimulated ECs at arterial blood flow rate (wall shear rate of 650 s−1). ECs covered more than 95% of the surface after 5-minute perfusions at 650 s−1 (data not shown). Morphometric evaluation of thrombotic deposits showed that at this wall shear rate, both fibrin and platelet deposition on ECs were low, regardless of their pretreatment with the anti-TM monoclonal antibody (P>.10, Table 3⇑). However, with more sensitive measurements of thrombogenesis, the thrombogenic properties of stimulated ECs treated with the anti-TM monoclonal antibody were seen to be increased. Fibrin deposition, as measured immunologically on the endothelium surface, and plasma levels of T-AT were significantly higher on anti-TM–treated cells than on control cells (P<.05, Table 3⇑). Thus, TM also modulates the prothrombogenic properties of the stimulated endothelium at this arterial wall shear rate.
Effect of Enhancement of TM Surface Activity on Thrombogenic Properties of Stimulated ECs
We next investigated whether enhancement of TM surface activity resulted in a diminution of the prothrombotic properties of stimulated ECs. We treated confluent ECs with PMA (6 ng/mL) for 48 hours. A previous study had shown that the effect of PMA on TM expression by ECs was biphasic, with a reduction occurring between 1 and 8 hours of stimulation followed by an increase at 24 to 48 hours of stimulation.28 TM surface activity was ≈150% of that of control cells after a 48-hour period of stimulation with PMA (Fig 2⇑). To study the effects of different levels of TM expression on the thrombogenicity of ECs that express comparable levels of TF activity, we stimulated PMA-treated ECs and control ECs with IL1 (50 U/mL)+TNF (10 ng/mL) for 6 hours. Preliminary experiments showed that this combination of agonists (IL1+TNF) and this length of stimulation (6 hours) were required so that PMA-treated ECs and control ECs express similar levels of TF activity (Fig 3⇓). Note that PMA-treated ECs that were restimulated with IL1+TNF for 6 hours still expressed significantly higher levels of TM (P<.05).
Control and PMA-treated ECs stimulated with IL1+TNF were exposed for 3 minutes to nonanticoagulated blood at a venous wall shear rate (50 s−1). This time of perfusion, shorter than in the first set of experiments, was chosen to highlight the role of TM, mainly because, in contrast to the previous experiments in which ECs were stimulated only with IL1, IL1+TNF-stimulated ECs showed a threefold higher level of TF expression (Figs 1⇑ and 3⇑). After a 3-minute perfusion period, fibrin deposition on IL1+TNF-stimulated ECs was threefold higher than on ECs stimulated with IL1 and perfused for 5 minutes (Tables 2⇑ and 4⇓). However, the deposition of fibrin on PMA-treated cells that were stimulated with IL1+TNF was moderately but significantly lower than on ECs nonpretreated with PMA (P<.05). Similarly, the plasma concentration of T-AT was significantly lower with PMA-treated cells than with ECs not pretreated with PMA (P<.05, Table 4⇓). When ECs were further incubated in presence of anti-TM IgG1, the deposition of fibrin was comparable (P>.50) regardless of their pretreatment with or without PMA, indicating that the antithrombotic effect induced by the PMA pretreatment of ECs was specifically due to the enhancement of TM surface activity. Thus, enhancement of TM surface activity does result in an enhancement of the antithrombotic status of ECs.
Effect of Deficiency of Protein C or Protein S or of Resistance to Activated Protein C on Thrombogenesis of Stimulated ECs
A total of 11 patients (mean age 43±4 years; 7 females, 4 males) with isolated and heterozygous protein C or protein S deficiency or resistance to activated protein C due to heterozygous factor V Leiden mutation were evaluated (Table 5⇓). Seven had a history of thrombosis. Results obtained with blood from these patients were pooled and compared with those of 11 healthy age- (44±3 years) and sex- (7 females, 4 males) matched individuals with normal protein C and protein S plasma levels and no resistance to activated protein C. In addition, plasma levels of antithrombin III in all these subjects were within the normal limits (data not shown).
The mean plasma levels of F1+2 and T-AT collected before perfusion experiments in these 11 patients were significantly elevated compared with the normal individuals (Fig 4⇓, P<.05). In contrast, the mean plasma FPA values were strictly comparable (P>.50). Fibrin deposition on IL1-stimulated ECs perfused for 5 minutes at 50 s−1 was higher in patients than in control subjects (P<.05, Fig 4⇓). However, due to high interindividual variations, there were no statistical differences in plasma levels of the biological markers of thrombin formation in blood collected immediately distal to the chamber (2.2±0.6 versus 1.3±0.2 pmol/mL, P=.13; 22.8±4.9 versus 16.5±2.0 ng/mL, P=.22; 111±22 versus 81±11 ng/mL, P=.30 for F1+2, T-AT, and FPA in patients and control subjects, respectively). Interestingly, fibrin deposition was significantly correlated with plasma levels of T-AT in blood collected before perfusion experiments (r=.52, P=.01, Fig 5⇓). There was no such relationship between fibrin deposition and plasma levels of F1+2 or FPA (r<.30, P>.15). None of these parameters (ie, F1+2, T-AT, FPA, and fibrin deposition) was correlated with previous history of thrombosis. In addition, due to the low number of patients, these parameters did not statistically differentiate with regard to the type of abnormality (protein C, protein S, or factor V Leiden).
Normal resting endothelium is a nonthrombogenic surface. Many endothelial systems inhibiting platelet function or coagulation activation and stimulating fibrinolysis have been described.14 We show here that TM is one of the main components supporting this function. Using a monoclonal antibody directed against the fifth epidermal growth factor–like repeats of TM, we obtained a total suppression of the ability of TM to activate protein C. The mechanism by which this antibody acts remains to be determined. It is possible that it prevents both thrombin inhibition by TM and protein C activation by the thrombin/TM complex, since TM may be internalized in ECs when bound to antibodies.29 Alternatively, the anti-TM antibody may specifically alter the cofactor protein C function of TM by blocking the interaction between thrombin and TM, since the primary thrombin binding site of TM is thought to be composed of the fifth and sixth epidermal growth factor homology regions.1
The suppression of TM activity was virtually complete (Fig 1⇑). The physiological relevance of this observation is uncertain, since deficiency of TM has not been reported. It is possible that such a deficiency is lethal. Interestingly, the lethality of this deficiency may not be directly related to clotting disorders. Studies by Rosenberg30 showed that TM is necessary for normal embryonic development. A total suppression of TM activity as obtained in the present study may mimic that found in persons with homozygous protein C deficiency, a disorder known to be strongly associated with the development of thrombosis during the neonatal period.2 The thrombogenic properties of ECs that express half the normal level of TM remain to be determined. In the absence of any thrombogenic challenge, heterozygous TM-deficient mice exhibiting a 50% reduction of TM protein are free of thrombotic complications.30 In patients with heterozygous protein C or protein S deficiencies or heterozygous factor V Leiden, the thrombotic events are much less severe than in homozygous patients, and most often they occur when there are other predisposing factors such as surgery.23 In this regard, we found that in such patients, the prothrombotic state, as measured using the ex vivo thrombosis model, was less marked than in studies in which EC TM expression was totally abolished: mean fibrin deposits and plasma levels of T-AT were ≈50% lower (Table 2⇑ and Fig 4⇑).
Under inflammatory circumstances, the endothelium becomes procoagulant and prothrombogenic.14 15 16 17 18 19 20 21 22 Previous authors including ourselves have shown that in these conditions, the prothrombogenicity of the endothelium is fully dependent on TF, since it is fully abolished by preincubating the cells with antibodies directed against TF.19 22 The TM expression diminishes after longer incubation periods (ie, 18 to 24 hours), when TF expression has decreased and returned to baseline levels.15 16 17 Despite this transitory aspect of TF expression, our results indicate that under these conditions in which TM is fully downregulated, the endothelium may remain highly prothrombogenic.
Whereas the abolition of TM expression resulted in an enhancement of the thrombotic properties of the endothelium, we were able to show that the augmentation of TM levels resulted in a moderate but significant diminution of the prothrombogenicity of the stimulated endothelium (Table 4⇑). Enhancement of TM activity by PMA results from an increased de novo synthesis of TM, probably secondary to the downregulation of protein kinase C.28 The smallness of the effect of this surexpression of TM must be compared with the very high levels of TF expressed by ECs in these experiments, which made the cells highly thrombogenic. However, these results suggest that strategies designed to increase TM expression may be a target for decreasing the tendency toward thrombosis. This condition may be produced with pharmacological agents that increase intracellular cAMP levels31 or with retinoic acid,32 which counteracts both the downregulation of TM and the induction of TF.
Previous studies have shown that asymptomatic patients with hereditary deficiencies of protein C or protein S or resistance to activated protein C have a hypercoagulable state characterized by an increased plasma concentration of F1+2.33 34 35 Our results confirm these data and extend them to T-AT complexes (Fig 4⇑). Thus, patients with deficiency of the TM/protein C/protein S pathway have excessive thrombin generation. However, in both studies, the mean plasma level of FPA in patients and control subjects was comparable (Fig 4⇑), indicating that the generated thrombin appears to be efficiently inhibited by other anticoagulant systems such as the heparan sulfate–antithrombin system. This study is the first to utilize the ex vivo perfusion model in the investigation of the prothrombotic state of patients with deficiency of protein C or protein S or resistance to activated protein C. The mean amount of fibrin deposited on IL1-stimulated ECs perfused at a venous wall shear rate was significantly higher in patients with a disorder of the TM/protein C/protein S pathway than in age- and sex-matched control subjects (Fig 4⇑). Interestingly, there were highly significant correlations between basal level of biological markers of a hypercoagulable state (ie, T-AT) and the amount of fibrin deposited on the stimulated endothelium (Fig 5⇑). In contrast and as observed previously,21 the relationship between the fibrin deposition on the cells and the FPA values was not significant (P=.23). This could simply be related to the wide distribution of FPA values but could also signify that not all the fibrin cleaved by thrombin deposited on the procoagulant cells, some remaining soluble. These results may help to validate the utility of biological markers of hypercoagulability, especially T-AT, in determining which patients are destined to develop a clinically relevant prothrombotic state. It remains to be determined with a larger study whether these assays or this thrombosis model can pinpoint patients most likely to benefit from anticoagulant therapy.
Selected Abbreviations and Acronyms
|F1+2||=||prothrombin fragment 1+2|
|PMA||=||4′-phorbol 12-myristate 13-acetate|
|T-AT||=||thrombin-antithrombin III complexes|
|TF-AU||=||TF arbitrary units|
This work was supported in part by a grant from La Délégation Régionale à la Recherche Clinique (Centre Hospitalier Toulouse, No. 95-34-L). The authors thank Dr A. Merle-Beral and the midwives of the Nouvelle Clinique de l'Union and of the Hôpital Joseph Ducuing for providing the umbilical cords.
Esmon CT. Thrombomodulin as model of molecular mechanisms that modulate protease specificity and function at the vessel surface. FASEB J. 1995;9:946-955.
Koster T, Rosendaal FR, Briët E, van der Meer FJM, Coly LP, Trienekens PH, Poort SR, Reitsma PH, Vandenbroucke JP. Protein C deficiency in a controlled series of unselected outpatients: an infrequent but clear risk factor for venous thrombosis (Leiden thrombophilia study). Blood. 1995;85:2756-2761.
Dahlbäck B. Inherited thrombophilia: resistance to activated protein C as pathogenic factor of venous thromboembolism. Blood. 1995;85:607-614.
Ohlin AK, Marlar RA. The first mutation identified in the thrombomodulin gene in a 45-year-old man presenting with thromboembolic disease. Blood. 1995;85:330-336.
Gomi K, Zushi M, Honda G, Kawahara S, Matsuzaki O, Kanabayashi T, Yamamoto S, Maruyama I, Suzuki K. Antithrombotic effect of recombinant human thrombomodulin on thrombin-induced thromboembolism in mice. Blood. 1990;75:1396-1399.
Taylor FB, Chang A, Esmon CT, D'Angelo A, Vigano-D'Angelo S, Blick E. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest. 1987;79:918-925.
Gruber A, Griffin JH, Harker LA, Hanson SR. Inhibition of platelet-dependent thrombus formation by human activated protein C in a primate model. Blood. 1989;73:639-642.
Gruber A, Hanson SR, Kelly AB, Yan BS, Bang N, Griffin JH, Harker LA. Inhibition of thrombus formation by activated recombinant protein C in a primate model of arterial thrombosis. Circulation. 1990;82:578-585.
Arnljots B, Dahlbäck B. Protein S as an in vivo cofactor to activated protein C in prevention of microarterial thrombosis in rabbit. J Clin Invest. 1995;95:1987-1993.
Bevilacqua MP, Pober JS, Majeau GR, Fiers W, Contran RS, Gimbrone MA Jr. Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: characterization and comparison with the actions of interleukin-1: Proc Natl Acad Sci U S A. 1986;83:4533-4537.
Moore KL, Andreoli SP, Esmon NL, Esmon CT, Bang NU. Endotoxin enhances tissue factor and suppresses thrombomodulin expression of human vascular endothelium in vitro. J Clin Invest. 1987;79:124-130.
Bevilacqua MP, Pober JS, Majeau GR, Contran RS, Gimbrone MA. Interleukin 1 (IL1) induces biosynthesis and cell surface expression of procoagulant activity in human vascular endothelial cells. J Exp Med. 1984;160:618-623.
Clozel M, Kuhn H, Baumgartner HR. Procoagulant activity of endotoxin-treated human endothelial cells exposed to native human flowing blood. Blood. 1989;73:729-733.
Kirchhofer D, Sakariassen KS, Clozel M, Tschopp TB, Hadvary P, Nemerson Y, Baumgartner HR. Relationship between tissue factor expression and deposition of fibrin, platelets, and leukocytes on cultured endothelial cells under venous blood flow conditions. Blood. 1995;81:2050-2058.
Kirchhofer D, Tschopp TB, Hadvary P, Baumgartner HR. Endothelial cells stimulated with tumor necrosis factor-α express varying amounts of tissue factor resulting in inhomogenous fibrin deposition in a native blood flow system: effects of thrombin inhibitors. J Clin Invest. 1994;93:2073-2083.
Diquélou A, Lemozy S, Dupouy D, Boneu B, Sakariassen KS, Cadroy Y. The effect of blood flow on thrombin generation is dependent on the nature of the thrombogenic surface. Blood. 1994;84: 2206-2213.
Cadroy Y, Sié P, Boneu B. Frequency of a defective response to activated protein C in patients with a history of venous thrombosis. Blood. 1994;83:2008-2009.
Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins: identification by morphologic and immunologic criteria. J Clin Invest. 1973;52:2745-2756.
Horie S, Kizaki K, Ishii H, Kazama M. Retinoic acid stimulates expression of thrombomodulin, a cell surface anticoagulant glycoprotein, on human endothelial cells. Biochem J. 1992;281: 149-154.
Bauer KA, Broekmans AW, Bertina RM, Conard J, Horellou MH, Samama MM, Rosenberg RD. Hemostatic enzyme generation in the blood of patients with hereditary protein C deficiency. Blood. 1988; 71:1418-1426.
Conard J, Bauer KA, Gruber A, Griffin JH, Schwarz HP, Horellou MH, Samama MM, Rosenberg RD. Normalization of markers of coagulation activation with a purified protein C concentrate in adults with homozygous protein C deficiency. Blood. 1993;82: 1159-1164.