Uptake and Inactivation of Thrombin on Rabbit Aortic Endothelium Studied with Two Different Substrates

The endothelium is an important compartment for uptake and inhibition of thrombin. The amount of enzymatically active bound thrombin can be detected with both small synthetic substrates and with aid of fibrinogen as substrate. The present study was designed to investigate the relation between endothelially bound thrombin with amidolytic activity towards a synthetic substrate (S-2238) and thrombin capable of converting fibrinogen by measuring generation of fibrinopeptide A (FPA). The luminal surfaces of rabbit aortae (2 cm2) were exposed in vitro to thrombin (0.625-5.0 NIH units/ml). Thrombin disappeared from the solution and a certain fraction was recovered on the surface. There was a linear relationship between the amount of thrombin on the surface and the concentration of thrombin in the incubation mixture. Approximately one third of the thrombin measured with S-2238 was also able to cleave fibrinogen. After incubation with defibrinogenated plasma almost total inhibition of fibrinogen splitting activity occurred within 30 sec. The inhibition of the amidolytic activity was less complete. When endothelially bound thrombin was exposed to plasma much less FPA was generated than in a fibrinogen solution. A minor fraction of endothelially bound thrombin was inhibited also upon incubation with Tyrode without recovery of any enzymatic activity in the solution. The results indicate that a fraction of thrombin bound to the endothelium has retained enzymatic activity and that a fraction of the enzymatically active thrombin is capable of converting fibrinogen. Inhibition of thrombin enzymatic activity occurs rapidly upon exposure to plasma. The endothelium itself has a minor inhibitory effect also in the absence of plasma. Thrombin with retained fibrinogen splitting activity may contribute to thrombotic events particularly if plasma inhibitors are insufficient. Thrombin with activity towards fibrinogen could be bound to GAG:s whereas thrombin active towards S-2238 could be bound to GAGrs TM or possibly other receptors.

Thrombogenicity of the Injured Vessel Wall – Role of Antithrombin and Heparin

The thrombogenicity of the vessel wall after endothelial denudation is partly explained by an impaired inhibition of thrombin on the subendothelium. We have previously reported that thrombin coagulant activity can be detected on the vessel wall after balloon injury in vivo. The glycosaminoglycans of the subendothelium differ from those of the endothelium and have a lower catalyzing effect on antithrombin III, but inhibition of thrombin can still be augmented by addition of antithrombin III to the injured vessel surface.

In this study the effect of antithrombin III and heparin on thrombin coagulant activity on the vessel wall was studied after in vivo balloon injury of the rabbit aorta using biochemical and immunohistochemical methods and thrombin was analysed after excision of the vessel. Continuous treatment with heparin, lasting until sacrifice of the animal, or treatment with antithrombin III resulted in significant reduction of thrombin coagulant activity on the injured aorta. Heparin given only in conjunction with the injury did not prevent thrombin coagulant activity or deposition of fibrin on the surface.

The capacity of the injured vessel wall to inhibit thrombin in vitro was improved on aortic segments obtained from animals receiving antithrombin III but not from those given heparin. It is concluded that treatment with antithrombin III interferes with thrombin appearance on the vessel wall after injury and thereby reduces the risk for thrombosis.

Effect of Heparin on Thrombin Inhibition in the Microcirculation

Disappearance of thrombin enzymatic activity was measured during recirculation through the microvasculature in a rat Langendorff heart preparation. This resulted in a 50% loss of thrombin from the recirculating solution. No increased loss of thrombin could be demonstrated if a mixture of antithrombin III and thrombin was recirculated, compared to thrombin alone. If, however, a heparin/thrombin mixture was recirculated, a 90% loss of thrombin could be demonstrated. Pretreating the microvasculature with large amounts of heparin resulted in recovery of antithrombin III in the recirculated heparin solution. At a subsequent recirculation with a heparin/thrombin mixture the loss of thrombin was decreased to the control level, as seen when recirculation with thrombin alone was performed. It is concluded that disappearance of thrombin enzymatic activity from a solution when recirculated through the microcirculation can be considerably increased if recirculated together with heparin, which probably reacts with endogenous antithrombin III on the vessel wall. The disappearance of thrombin in the absence of heparin was, however, unaffected by antithrombin III. The latter finding is compatible with the hypothesis that, in the microcirculation, antithrombin III/glycosaminoglycans play only a minor role for inhibition of thrombin coagulant activity and that thrombin binds mainly to thrombomodulin.

Neutrophil Elastase-Derived Fibrin Degradation Products Indicate Presence of Abdominal Aortic Aneurysms and Correlate with Intraluminal Thrombus Volume

Background The intraluminal thrombi (ILT) of abdominal aortic aneurysms (AAA) contain neutrophils, which can secrete elastase. We evaluated whether plasma neutrophil elastase-derived cross-linked fibrin degradation products (E-XDP) could reveal the presence, size and mechanical stress of AAAs and its ILTs.

Methods E-XDP and D-dimer were measured in plasma from 37 male patients with AAA and 42 male controls. The ILT volumes of the AAAs and any coexisting aneurysms could be measured in 29 patients and finite element analysis was performed to estimate mechanical stress of the ILT. E-XDP, neutrophil elastase and neutrophil marker CD66acd were evaluated in aortic tissue with immunohistochemistry (IHC). The association between ILT volume and E-XDP was validated in a separate cohort (n = 51).

Results E-XDP levels were elevated in patients with AAA compared with controls (p = 5.8e-13), indicated AAA with 98% sensitivity, 86% specificity and increased with presence of coexisting aneurysms. The association between AAA and increased E-XDP was independent of smoking, comorbidities and medication. E-XDP correlated with volume of all ILTs (r = 0.76, p = 4.5e-06), mean ILT stress (r = 0.46, p = 0.013) and the volume of the AAA-associated ILT (r = 0.64, p = 0.00017). E-XDP correlated stronger with ILT volume compared with D-dimer (r = 0.76 vs. r = 0.64, p = 0.018). The correlation between E-XDP and ILT volume was validated in the separate cohort (r = 0.53, p = 7.6e-05). IHC revealed E-XDP expression in the ILT, spatially related to neutrophil elastase and neutrophils.

Conclusion E-XDP is a marker of the presence of AAA and coexisting aneurysms as well as the volume and mechanical stress of the ILT.

Proteolytically active ADAM10 and ADAM17 carried on membrane microvesicles in human abdominal aortic aneurysms

The intraluminal thrombus (ILT) of human abdominal aortic aneurysm (AAA) has been suggested to damage the underlying aortic wall, but previous work found scant activity of soluble proteases in the abluminal layer of the ILT, adjacent to the aneurysm. We hypothesised that transmembrane proteases carried by membrane microvesicles (MV) from dying cells remain active in the abluminal ILT. ILTs and AAA segments collected from 21 patients during surgical repair were assayed for two major transmembrane proteases, ADAM10 (a disintegrin and metalloprotease-10) and ADAM17. We also exposed cultured cells to tobacco smoke and assessed ADAM10 and ADAM17 expression and release on MVs. Immunohistochemistry showed abundant ADAM10 and ADAM17 protein in the ILT and underlying aneurysmal aorta. Domain-specific antibodies indicated both transmembrane and shed ADAM17. Importantly, ADAM10 and ADAM 17 in the abluminal ILT were enzymatically active. Electron microscopy of abluminal ILT and aortic wall showed MVs with ADAM10 and ADAM17. By flow cytometry, ADAM-positive microvesicles from abluminal ILT carried the neutrophil marker CD66, but not the platelet marker CD61. Cultured HL60 neutrophils exposed to tobacco smoke extract showed increased ADAM10 and ADAM17 content, cleavage of these molecules into active forms, and release of MVs carrying mature ADAM10 and detectable ADAM17. In conclusion, our results implicate persistent, enzymatically active ADAMs on MVs in the abluminal ILT, adjacent to the aneurysmal wall. The production of ADAM10- and ADAM17-positive MVs from smoke-exposed neutrophils provides a novel molecular mechanism for the vastly accelerated risk of AAA in smokers.

Abstract 261: Neutrophil Elastase Derived Fibrin Degradation Products are Increased in the Plasma of Patients with Abdominal Aortic Aneurysms and Correlate to the Intraluminal Thrombus Volume

Introduction: Abdominal Aortic Aneurysms (AAA) often contain an intraluminal thrombus (ILT). AAA diameter and ILT volume are associated with growth of the aneurysm. Neutrophils, present in the ILT, contain elastase (NE). NE activity leads to production of fibrin degradation products (FDPs) with a specific epitope [[Unable to Display Character: –]] XDP. The present study evaluates NE-derived FDPs in aneurysm patients scheduled for elective aortic repair. The purpose of the study is to introduce an additional bio-marker for presence of AAA and possibly risk of rupture by measuring levels of NE derived FDPs in plasma of patients with AAA.

Materials and Methods: 42 male patients, undergoing aortic repair for AAA were included. As controls, we collected blood samples from 42 men who attended an AAA screening program but had no AAAs on ultrasound. Computed Tomography (CT) images were available for 34 AAA patients and analyzed using A4 Clinics software (VASCOPS, Austria). Patient demographics, maximum diameter, aortic volume and ILT volume were recorded. Peak wall stress (PWS), peak wall rupture index (PWRI) and mean ILT stress were estimated by Finite Element Analysis using the A4 Clinics software. Plasma levels of elastase digests of cross-linked fibrin (E-XDP) were determined with a sandwich ELISA.

Results: E-XDP levels were higher in AAA patients than in age-matched controls (8.5 vs 1.2 U/ml, p<0.0001). E-XDP levels correlated with ILT volume (r = 0.64, p<0.0001), aortic volume (r = 0.64, p<0.0001) and maximum diameter (r = 0.59, p=0.0003). AAA patients with other concomitant peripheral aneurysms had higher E-XDP levels than those with only an AAA (13.6 vs 6.8 U/ml, p=0.028). PWS, PWRI and bleeding signs in the thrombus did not significantly affect E-XDP levels. Interestingly, the mean ILT stress correlated significantly to E-XDP levels (r= 0.45, p=0.008).

Conclusions: The study shows that it is feasible to measure E-XDP levels in plasma of patients with AAA and that E-XDP correlates with ILT volume and mean ILT stress. These results support the notion that the resident neutrophils in the ILT can actively lyse fibrin in the ILT, which may decrease ILT strength. E-XDP holds potential as a biomarker of the ILT in AAA patients and needs to be further investigated in AAA rupture risk assessment.

Abstract 367: Pcsk6 Is a Key Protease Modulating Smooth Muscle Cell Activation in Vascular Remodeling and Plaque Vulnerability

Proprotein convertases (PCSKs) are conserved among species and involved in processing of MMPs and growth factors, but poorly characterised in atherosclerosis. Previously we demonstrated upregulation of PCSK6 in a large cohort of plaques from symptomatic vs. asymptomatic patients. This protease localized to smooth muscle cells (SMCs) and showed a positive correlation to markers of inflammation, extracellular matrix remodeling and cytokines in plaques. Here we aimed at elucidating its role in vascular development and disease.

In zebrafish embryos PCSK6 localized to heart and vasculature and its ablation caused defective peripheral vascular patterning with cerebral and myocardial hemorrhage. Increased expression of PCSK6 in vascular pathologies was validated by microarrays from carotid plaques vs. undiseased arteries (n=32 patients, p<0.0001), abdominal (AAA, n=14, p<0.0001) and thoracic aortic aneurysms (TAA, n=244, p=0.012). By immunohistochemistry, PCSK6 localized mainly to SMCs in the fibrous cap and neo-vessels in plaques, AAA and TAA tissues. Investigation of mouse carotid ligation, rat artery balloon injury and human restenosis tissues with pronounced intimal hyperplasia confirmed induction of PCSK6 in proliferating SMCs. By microarrays following progression of rat carotid injury, PCSK6 was downregulated in early phases defined by inflammatory pathways, while upregulated in later phases with SMCs activation and localized in SMCs. Expression of PCSK6 in this model was positively correlated to PDGFB and IGF1 (Spearman r>0.7, p<0.0001). PCSK6 overexpression in SMCs in vitro increased their migration in a wound-healing assay, especially upon PDGFBB stimulation. By eQTL analyses several polymorphisms in the PCSK6 gene were found to influence its expression in plaques and aneurysm tissue. Among these, rs6598465 showed association with maximum progression of arterial intima-media thickness in high-risk coronary artery disease subjects (n=3400, p=0.037).

We identified a functional link between elevated expression of PCSK6 and vascular remodeling characterized by SMC activation. The present study establishes PCSK6 as one of the key modulators of pathological processes in relation to plaque vulnerability.

Abstract 150: Identification of SYNPO2, SYNM, LMOD1, PDLIM7 and PLN as Novel Markers of Smooth Muscle Cells in Atherosclerosis

Smooth muscle cell (SMC) activation is a hallmark of vascular remodeling associated with downregulation of SMC specific genes and altered cell function. Our aim was to identify a novel set of genes silenced following SMC activation in vascular disease.

We compared microarrays from carotid plaques (n=177) vs. undiseased arteries and symptomatic vs. asymptomatic patients and found that ‘muscle contraction’, ‘muscle differentiation’ and ‘cytoskeleton based migration’ were the most significantly downregulated pathways (p=6.06e-23; 5.22e-11; 4.62e-11 respectively). Genes clustered in these categories were known SMC markers together with a novel set of genes functionally coupled to actin cytoskeleton: SYNPO2, SYNM, LMOD1, PDLIM7 and PLN. Transcription factors regulating the expression of these genes showed enrichment of SRF, MYOD1 and MYOGENIN, known to control muscle differentiation. Downregulation of these genes was validated in independent microarrays from carotid (n=21) and coronary plaques (n=38). Their expression was positively correlated to typical SMC markers both in human plaques (Pearson r>0.8, p<0.0001) and rat carotid intimal hyperplasia (r>0.9, p<0.0001). In microarrays from rat carotid balloon injury, these genes were downregulated in the early phases of tissue remodeling but gradually reappeared in the mature intima. By immunohistochemistry, the proteins localized to SMCs in normal human vessels, but were mostly absent in plaques and restenosis tissues. In vitro, SYNPO2 and PDLIM7 localized to actin filaments, SYNM to cortical cytoskeleton, PLN to nucleus while LMOD1 was rapidly lost in cultured human and primary rat SMCs. Polymorphisms in some of these genes associated with carotid intima-media thickness in high-risk coronary artery disease subjects (n=3400). By eQTL analyses, rs11746443 influenced PDLIM7 expression in plaques. Interestingly, this polymorphism constitutes the binding site for HEY1 transcription factor implicated in vascular development.

We identified a panel of novel SMC proteins, which are lost in vascular disease and reflect the altered phenotype of SMCs in vascular remodeling. Our results indicate that these could be early and sensitive markers of the SMCs dedifferentiation in atherosclerosis.

Inhibition of neointimal hyperplasia by a specific thrombin inhibitor

OBJECTIVE

Restenosis secondary to neointimal hyperplasia remains the major limiting factor after vascular interventions. Thrombin generated in high concentrations at the site of vascular injury plays a central role in thrombosis and hemostasis. Thrombin has also been implicated as a mitogen for smooth muscle cell proliferation that contributes to restenosis. This study was designed to determine the effects of a specific thrombin inhibitor on neointimal hyperplasia after balloon injury in a rat carotid artery model.

DESIGN

A total of 47 male Sprague-Dawley rats were divided into five groups. All groups underwent balloon injury of the left carotid artery. A specific thrombin inhibitor, inogatran, was given in four different regimens: low and high dose injections, short-term infusion for 3 h, and long-term infusion for 1 week. After 2 weeks the animals were killed and the carotid neointima/media area ratio and the luminal narrowing were calculated.

RESULTS

All treatments significantly reduced the neointimal hyperplasia. Inogatran given as a long-term infusion for 1 week had the lowest neointima/media ratio compared with the other groups. The percentage of lumen narrowing was also significantly lower in all treatment groups compared with the control group.

CONCLUSION

A specific direct thrombin inhibitor, inogatran, reduces neointimal hyperplasia after arterial injury in rats. A more prolonged administration of the thrombin inhibitor gave a further reduction of the neointimal hyperplasia. It seems that inhibition of thrombin activity is not only important early after injury, but also later. This could have clinical implications in the treatment of restenosis and needs to be further evaluated.

The influence of direct and antithrombin-dependent thrombin inhibitors on the procoagulant and anticoagulant effects of thrombin

INTRODUCTION

Clinical trials evaluating direct thrombin inhibitors in unstable coronary artery disease (CAD) have been disappointing. The hypothesis tested in the present study was that these agents may inhibit the anticoagulant effect of thrombin to a further extent than the procoagulant effect of thrombin.

MATERIALS AND METHODS

We studied both reversible and irreversible thrombin inhibitors and compared the effects of each inhibitor on activated protein C (APC) generation vs. the effect on fibrinopeptide A (FPA) generation. A mixture of protein C, thrombin inhibitor, fibrinogen, fibrin polymerisation blocker and thrombin was incubated with thrombomodulin (TM)-expressing human saphenous vein endothelial cells (HSVECs). The inhibitors investigated were melagatran, inogatran, hirudin, hirugen, D-Phe-D-Pro-D-arginyl chloromethyl ketone (PPACK), and antithrombin (AT) alone or in combination with unfractionated heparin (UFH) or low-molecular-weight heparin (LMWH).

RESULTS

All agents, except hirugen, inhibited APC and FPA generation in a dose-dependent manner. FPA inhibition/APC inhibition ratios, based on IC50 for inogatran, melagatran, hirudin, PPACK, AT, AT-UFH and AT-LMWH were 1.73, 0.85, 0.55, 2.1, 0.5, 0.65 and 3.1 respectively.

CONCLUSIONS

All agents, except hirugen, inhibited APC and FPA generation approximately to a similar extent. Thus, it can be inferred that the poor efficacy of thrombin inhibitors in recent clinical trials in patients with unstable CAD is unlikely to be a consequence of their effects on the protein C system.

Inhibition of endothelial cell-mediated generation of activated protein C by direct and antithrombin-dependent thrombin inhibitors

The present study investigated the effect of the thrombin inhibitors antithrombin (AT) (with and without unfractionated heparin or low molecular weight heparin), hirudin, inogatran and melagatran on thrombin-thrombomodulin-mediated generation of activated protein C (APC), in solution and on endothelial cells. Sequential incubation with thrombin, thrombin inhibitors and protein C was followed by measurement of APC by an amidolytic assay. The approximate concentrations resulting in 50% inhibition of endothelial cell-mediated APC generation for AT, AT-unfractionated heparin, AT-low molecular weight heparin, hirudin, melagatran and inogatran were 200, 4, 9, 1, 8 and 60 nmol/l, respectively. The normal plasma level of AT is 2800 nmol/l and relevant therapeutic concentrations from clinical trials are 200 nmol/l for hirudin, 500 nmol/l for melagatran and 1000 nmol/l for inogatran. The present study indicates that clinically relevant concentrations of the tested thrombin inhibitors interfere with endothelial-mediated APC generation, which may offer an explanation for the lack of a dose-response effect in clinical trials with thrombin inhibitors.

Thrombin inhibition by antithrombin III on the subendothelium is explained by the isoform AT beta

Balloon injury of the rabbit aorta results in thrombin coagulant activity on the injured vessel wall that causes fibrin formation. The anticoagulant activity of both the intact and injured vessel wall has been partly explained by glycosaminoglycans with heparin-like activity that augment that activity of antithrombin III (AT). AT exists in two isoforms, alpha and beta, AT beta, which constitutes only 5% to 10% of AT in plasma, lacks one carbohydrate side chain, has higher affinity for glycosaminoglycans, and associates more readily with the subendothelium. This study evaluated whether AT can inhibit thrombin on the injured vessel wall and, if so, whether one of the isoforms is more effective then the other. The two isoforms were isolated from human plasma by heparin-Sepharose chromatography, and the purity was investigated by isoelectric focusing and crossed immunoelectrophoresis. Rabbits were subjected to balloon injury of the aorta; 3 hours after injury the aorta was excised. Thrombin coagulant activity on the aorta was measured by exposure to fibrinogen and thereafter by measuring the generation of fibrinopeptide A. Injured animals were treated with AT, AT alpha, or AT beta and were compared with control animals. AT was demonstrated on the injured vessel wall by using an immunohistochemical method. Animals receiving crude AT had significantly lower amounts of thrombin coagulant activity on the injured aortic wall than control animals, but AT alpha at a comparable dose had no effect. AT beta was given in the same dose as crude AT and also at a dose (10%) proportional to its presence in plasma. Animals receiving AT beta had significantly lower values of thrombin on the injured aortic wall than control animals. We conclude that the inhibitory effect of AT on thrombin coagulant activity on the injured vessel wall in explained by its AT beta content.

Cultured human endothelial cells seeded on expanded polytetrafluoroethylene support thrombin-mediated activation of protein C

PURPOSE

Reduction of the thrombogenicity of synthetic vascular grafts by endothelialization has been suggested. The purpose of this study was to examine some of the non-thrombogenic properties of cultured adult human great saphenous vein endothelial cells seeded on expanded polytetrafluoroethylene (ePTFE) grafts.

METHODS

Endothelialized grafts, control grafts, and wells were incubated with thrombin. Assays of thrombin loss from solution, thrombin coagulant activity, and protein C activation on the surface were obtained. The presence of thrombomodulin was confirmed with immunohistochemistry.

RESULTS

Significantly more thrombin was found on the ePTFE grafts or wells that underwent endothelialization, and larger amounts were lost from the thrombin solution compared with the control group. Thrombin enzyme activity on the endothelialization group was almost completely represented by activation of protein C and only to a minor extent by activity towards fibrinogen.

CONCLUSIONS

It is concluded that endothelial cells seeded on ePTFE retain the possibility to inhibit thrombin coagulant activity and to activate protein C. These findings provide support for the clinical applicability of cultured autologous endothelium on ePTFE grafts.

Antithrombin III inhibits thrombin-induced proliferation in human arterial smooth muscle cells

Thrombin has attracted increasing attention as a possible mitogen for vascular smooth muscle cells in lesion development both after vascular injury and in atherogenesis. In this study, the ability of antithrombin III to inhibit alpha-thrombin-induced DNA synthesis and cell proliferation in human arterial smooth muscle cells was analyzed. We demonstrate a concentration-dependent initiation of DNA synthesis and cell proliferation by alpha-thrombin. This effect was abolished when complex formation with antithrombin III was allowed before thrombin was added to the cell cultures. Addition of alpha-thrombin and antithrombin III simultaneously at the beginning of the incubation period also resulted in an inhibition of thrombin-induced DNA synthesis, but to a lower degree. The inhibitory activity of antithrombin III was enhanced in the presence of heparin, which on its own had no inhibitory effect on thrombin-induced DNA synthesis. In contrast, the mitogenic activity of alpha-thrombin could be inhibited by heparin in the presence of low concentrations of serum. This inhibition was dependent on the presence of antithrombin III in serum, since heparin lacked effect if antithrombin III was depleted from serum by immunoaffinity chromatography. Analysis of the enzymatic activity of thrombin showed that the influence on catalytic activity of thrombin corresponded to the mitogenic activity of thrombin in the presence of heparin, antithrombin III, and serum. The results suggest that the mitogenic activity of thrombin is regulated by antithrombin III. Therefore, antithrombin III may serve dual functions by inhibiting thrombin in the coagulation cascade and by neutralizing its growth-promoting effects on vascular smooth muscle cells.

Thrombogenicity of the injured vessel wall--role of antithrombin and heparin

The thrombogenicity of the vessel wall after endothelial denudation is partly explained by an impaired inhibition of thrombin on the subendothelium. We have previously reported that thrombin coagulant activity can be detected on the vessel wall after balloon injury in vivo. The glycosaminoglycans of the subendothelium differ from those of the endothelium and have a lower catalyzing effect on antithrombin III, but inhibition of thrombin can still be augmented by addition of antithrombin III to the injured vessel surface. In this study the effect of antithrombin III and heparin on thrombin coagulant activity on the vessel wall was studied after in vivo balloon injury of the rabbit aorta using biochemical and immunohistochemical methods and thrombin was analysed after excision of the vessel. Continuous treatment with heparin, lasting until sacrifice of the animal, or treatment with antithrombin III resulted in significant reduction of thrombin coagulant activity on the injured aorta. Heparin given only in conjunction with the injury did not prevent thrombin coagulant activity or deposition of fibrin on the surface. The capacity of the injured vessel wall to inhibit thrombin in vitro was improved on aortic segments obtained from animals receiving antithrombin III but not from those given heparin. It is concluded that treatment with antithrombin III interferes with thrombin appearance on the vessel wall after injury and thereby reduces the risk for thrombosis.

Effect of heparin on thrombin inhibition in the microcirculation

Disappearance of thrombin enzymatic activity was measured during recirculation through the microvasculature in a rat Langendorff heart preparation. This resulted in a 50% loss of thrombin from the recirculating solution. No increased loss of thrombin could be demonstrated if a mixture of antithrombin III and thrombin was recirculated, compared to thrombin alone. If, however, a heparin/thrombin mixture was recirculated, a 90% loss of thrombin could be demonstrated. Pretreating the microvasculature with large amounts of heparin resulted in recovery of antithrombin III in the recirculated heparin solution. At a subsequent recirculation with a heparin/thrombin mixture the loss of thrombin was decreased to the control level, as seen when recirculation with thrombin alone was performed. It is concluded that disappearance of thrombin enzymatic activity from a solution when recirculated through the microcirculation can be considerably increased if recirculated together with heparin, which probably reacts with endogenous antithrombin III on the vessel wall. The disappearance of thrombin in the absence of heparin was, however, unaffected by antithrombin III. The latter finding is compatible with the hypothesis that, in the microcirculation, antithrombin III/glycosaminoglycans play only a minor role for inhibition of thrombin coagulant activity and that thrombin binds mainly to thrombomodulin.

Thrombin inactivation and the effects of antithrombin and heparin in a recirculating Langendorff preparation

A recirculating Langendorff preparation was developed in order to study the effects of antithrombin III (AT) and heparin on thrombin inactivation in the capillary bed. Using this technique, rat hearts were shown to inhibit approximately 50% of the recirculated thrombin, while surface-bound thrombin displayed no enzymatic activity towards recirculated fibrinogen. By displacing thrombin bound to the endothelial surface with Polybrene, thrombin enzymatic activity was found to be equally active against a synthetic chromogenic substrate (S-2238) and fibrinogen. This indicates that thrombin bound to the capillary endothelial surface is not coagulantly active, but, when displaced from the endothelium, its coagulant activity is retained. Hearts pretreated with AT or heparin had no effect on thrombin inhibition or surface-bound thrombin, although there was an uptake of both substances to the heart. These findings are in contrast with the results obtained from large vessel preparations, where pretreatment with AT increased thrombin inhibition and also demonstrated lower levels of surface-bound thrombin, and where heparin eluted endogenous AT from the endothelium. It is concluded that AT plays a more important role in the inhibition of thrombin on large vessels than in the microcapillary system where thrombomodulin is the major inactivator of thrombin.

Heparin cofactor II significance for the inhibition of thrombin at the injured vessel wall

The thrombin inhibitory role of antithrombin III (ATIII) and heparin cofactor II (HCII) was studied in vitro using intact and injured rabbit aortae. When intact vessels were loaded with thrombin and then exposed to either heat defibrinogenated human plasma (HDHP) or ATIII the same degree of thrombin inhibition was achieved demonstrating that ATIII was the only plasma component involved in thrombin inhibition on the intact vessel wall. When the media of the vessel wall was loaded with thrombin and then exposed to ATIII or HCII a significantly higher thrombin activity remained on the surface than when it was exposed to defibrinogenated plasma. A mixture of ATIII and HCII resulted in a greater inhibition of thrombin than ATIII or HCII alone. It is concluded that, contrary to what happens on the endothelium, HCII and ATIII inhibit additively thrombin on the injured vessel wall. HCII thus plays an essential role for the inhibition of thrombin at the injured vessel wall. It is also concluded that an additional plasma component participates in thrombin inhibition on the media but its contribution is negligible as compared with ATIII or HCII.

Coagulant and noncoagulant thrombin enzymatic activity on the endothelium

The present study examines the role of antithrombin III (ATIII) and heparin in the inactivation of thrombin on rabbit aortae in vitro. Thrombin on the endothelium was functionally assayed with a synthetic chromogenic substrate (thrombins) or with fibrinogen (thrombinf), measured as the liberation of fibrinopeptide A. After incubation with thrombin, residual amounts were found on the endothelium. Surface-bound thrombinf constituted approximately one-half of surface-bound thrombins. Displacement of thrombin activity from the thrombin loaded endothelial surface was achieved with Polybrene. The thrombin could be quantitatively recovered in the Polybrene solution and equal amounts of thrombins and thrombinf were found. Incubation of endothelial segments with ATIII resulted in only a minimal uptake of ATIII on the surface but in a significantly increased capacity to inactivate thrombin, whereas incubation of the segments with heparin alone had the opposite effect. The effect of ATIII could be inhibited by prior incubation with Polybrene or by a subsequent incubation with heparin. It is concluded that the vessel wall content of ATIII is one determinant for the capacity to inhibit thrombin and that ATIII is bound to glycosaminoglycans on the vessel wall since it could be displaced by heparin.

Inhibition of thrombin on subendothelium. Studies on rabbit aorta, ex vivo

The endothelium plays an active role in inhibiting and promoting the activation of coagulation. Loss of the endothelial lining results in a thrombogenic state. The inactivation of thrombin on rabbit aorta after a superficial and deep injury to the vessel wall was studied. The thrombin enzymatic activity on the subendothelium could be modulated and decreased activity was found on both types of injured vessel segments when exposed to antithrombin III (AT). An increased activity by exposing the superficially injured vessels. Most probably, heparin elutes AT from the surface of the superficially injured vessels, but on the deeply injured surface no AT is present and, consequently, heparin has no effect in this respect. These results speak in favour of treatment with AT instead of heparin in cases of arterial injury.

Long-term stability in vivo of a thromboresistant heparinized surface

A surface with covalently bonded heparin was tested in an in vivo model. Heparinized and non-heparinized tubings were inserted in the pig aorta. They were left in place for up to 16 wk. After removal they were analysed for their capacity to take up and inhibit thrombin. A comparison was also made with tubings that had not been inserted in animals. The heparin surface showed some specific alterations after implantation but its capacity to inhibit thrombin was not impaired. The heparinized tubings caused less intimal hyperplasia and thrombi than non-heparinized tubings.

Uptake and inactivation of thrombin on the subendothelium: comparisons with endothelium

One aspect of the non-thrombogenicity of endothelium is the property to bind and inhibit thrombin. Loss of the endothelial layer gives rise to a thrombogenic surface. The present study concerns uptake and inactivation of thrombin on endothelium as well as subendothelium. Both the immediate subendothelial layer and the media were investigated. Binding and inhibition of thrombin was assayed using both a synthetic chromogenic substrate and fibrinogen. In the latter case thrombin enzymatic activity was assessed by measuring liberation of fibrinopeptide A. There was a linear relationship between thrombin in the solution and thrombin enzymatic activity recovered on the surface on both endothelium and media. 30-50 per cent of the surface bound thrombin measured with the chromogenic substrate was also active towards fibrinogen. Preincubation of endothelium with plasma gave rise to an increased loss of thrombin from the solution at the same time as less thrombin was recovered on the surface. This indicates increased inactivation. Increased inactivation could also be obtained by pretreatment with antithrombin (AT) but not with AT-depleted plasma. The same general pattern was seen on the subendothelium but not on the media. It is concluded that inhibition of thrombin occurs on the endothelium and can be augmented by prior binding of AT to the surface. This effect although less pronounced can also be obtained on the subendothelium. The deeply injured vessel is much less capable of inhibiting thrombin.

Uptake and inactivation of thrombin on rabbit aortic endothelium studied with two different substrates

The endothelium is an important compartment for uptake and inhibition of thrombin. The amount of enzymatically active bound thrombin can be detected with both small synthetic substrates and with aid of fibrinogen as substrate. The present study was designed to investigate the relation between endothelially bound thrombin with amidolytic activity towards a synthetic substrate (S-2238) and thrombin capable of converting fibrinogen by measuring generation of fibrinopeptide A (FPA). The luminal surfaces of rabbit aortae (2 cm2) were exposed in vitro to thrombin (0.625-5.0 NIH units/ml). Thrombin disappeared from the solution and a certain fraction was recovered on the surface. There was a linear relationship between the amount of thrombin on the surface and the concentration of thrombin in the incubation mixture. Approximately one third of the thrombin measured with S-2238 was also able to cleave fibrinogen. After incubation with defibrinogenated plasma almost total inhibition of fibrinogen splitting activity occurred within 30 sec. The inhibition of the amidolytic activity was less complete. When endothelially bound thrombin was exposed to plasma much less FPA was generated than in a fibrinogen solution. A minor fraction of endothelially bound thrombin was inhibited also upon incubation with Tyrode without recovery of any enzymatic activity in the solution. The results indicate that a fraction of thrombin bound to the endothelium has retained enzymatic activity and that a fraction of the enzymatically active thrombin is capable of converting fibrinogen. Inhibition of thrombin enzymatic activity occurs rapidly upon exposure to plasma. The endothelium itself has a minor inhibitory effect also in the absence of plasma.(ABSTRACT TRUNCATED AT 250 WORDS)