Inhibition of thrombin by antithrombin III in the presence of certain glycosaminoglycans found in the mammalian aorta

Chondroitin-4-sulphate, chondroitin-6-sulphate, dermatan sulphate, heparan sulphate and hyaluronic acid were compared with heparin in their abilities to influence the inactivation of bovine thrombin by rabbit antithrombin III. The effect of the glycosaminoglycans on the enzymic activity of thrombin was examined using the substrate α-N-benzoyl arginine ethyl ester. Heparin, dermatan sulphate, heparan sulphate and, to a smaller extent, chondroitin-6-sulphate increased the esterase activity, whereas chondroitin-4-sulphate and hyaluronic acid had a negligible effect. Heparan sulphate and dermatan sulphate markedly accelerated the inactivation of thrombin by antithrombin III, but chondroitin-4-sulphate, chondroitin-6-sulphate and hyaluronic acid did not significantly affect the reaction.The glycosaminoglycans were adsorbed on Sepharose-lysine or polyacrylamide-ethylenediamine to test their ability to bind thrombin or antithrombin III under conditions where neither protein could directly bind to the conjugates. The binding of thrombin to immobilised heparan sulphate or dermatan sulphate compared well with that to heparin. On immobilised chondroitin-6-sulphate, thrombin was retarded and on chondroitin-4-sulphate no thrombin binding was observed. In contrast to thrombin, antithrombin III only bound to heparin: antithrombin III did not bind to dermatan sulphate, heparan sulphate or the other glycosaminoglycans. Antithrombin III inactivation of ¹²⁵I-thrombin adsorbed to either heparan sulphate or dermatan sulphate produced a thrombin-antithrombin III complex which was spontaneously released from the glycosaminoglycan. However, no complex was recovered from immobilised heparin unless the column was eluted by 1M NaCl. These results support the view that heparan sulphate and dermatan sulphate could play a role similar to that of heparin in the inactivation of thrombin by antithrombin III.     

Role of heparin in the inactivation of thrombin, factor Xa, and plasmin by antithrombin III

To characterize the mode of action of heparin, the kinetic behaviour of the inhibition of thrombin, factor Xa, and plasmin by antithrombin III was studied in the presence and absence of heparin. Following the concentration dependence of inactivation, in both cases a linear dependence of the apparent first-order inactivation rate constant on the antithrombin III concentration was found. These results are explained by enzyme-heparin interactions. Thus, heparin is believed to act as an activator of the enzymes that makes them more susceptible to antithrombin III.

Values of kinetic constants for the inactivation reaction of the several enzymes were determined. From these values it is concluded that heparin influences primarily the thrombin-antithrombin III interaction.     

Thrombin inactivation on surfaces with covalently bonded heparin

About 8000 Daltons porcine mucosa heparin fragments were covalently bonded by end-point attachment to polyethylene. The interaction between the immobilized heparin, added thrombin, and antithrombin III [AT] was investigated.

The heparin surface was adsorbed with either albumin, AT dissolved in albumin or Tyrode, or platelet free plasma. Irrespective of the pre-treatment procedure, exposure of the surface to thrombin resulted in the same substantial decrease of thrombin in solution and the same degree of surface-confined thrombin activity. It was concluded that the heparin surface has a large capacity to bind thrombin and that the thrombin inhibitory capacity of high affinity heparin fragments is limited.

On exposure of the thrombin-loaded surfaces to defibrinogenated plasma or AT, the surface-confined thrombin was inhibited within 30 seconds. Successive dilutions of plasma or AT decreased the inhibition rate but not the inhibition capacity. It is concluded that inhibition of thrombin adsorbed on the heparin surface occurs as follows: Added AT adheres to high affinity heparin fragments on the surface whereupon adsorbed thrombin migrates in the hydrophilic heparin coating towards the reaction site of AT and becomes inhibited. The inactivated thrombin-AT complex leaves then the surface, thus enabling the process to be repeated.     

GAGs-potentiated inhibition of thrombin, factor Xa and plasmin in plasma and in a purified system containing antithrombin III - correlation with total charge density

The ability of heparin and related glycosaminoglycans (GAGs) to accelerate the inhibition of thrombin, factor Xa and plasmin in plasma and in a purified system containing antithrombin III (At III) was studied using chromogenic peptide substrate assays. There was good correlation between the charge density of the mucopolysaccharides and the activities investigated. While the difference between potentiation of the antithrombin activity by GAGs in plasma and in the purified system was slight, the inhibition of factor Xa in plasma was more pronounced than in the presence of purified At III, indicating the mechanisms for GAGs-potentiated inhibition of thrombin and factor Xa are not identical. For the antiplasmin activity, there was a good correlation between the chemical structure and biological activity only in the pure system, confirming that the antithrombin-GAG complex plays a very limited role in the inactivation of plasmin in plasma.     

Heparin with low affinity to antithrombin III inhibits the activation of prothrombin in normal plasma

Standard unfractionated heparin is known to have two actions on blood clotting. Unfractionated heparin enhances the rates at which antithrombin III inactivates activated clotting factors, and inhibits the activation of both Factor X and prothrombin by disrupting the calcium and phospholipid dependent assembly of the Factor X and prothrombin activator complexes. This latter inhibitory action of heparin occurs independently of antithrombin III. A heparin fraction with low affinity to antithrombin III was prepared from standard heparin by affinity chromatography on antithrombin-III-Sepharose and its properties compared with unfractionated heparin. The low affinity heparin fraction and the unfractionated heparin had equivalent inhibitory effects on prothrombin activation in antithrombin III depleted plasma. In normal plasma, the low affinity fraction inhibited the activation of prothrombin. Unlike the unfractionated heparin, however, the fraction of heparin with low affinity to antithrombin III did not enhance the inactivation of either Factor Xa or thrombin. This antithrombin III independent inhibition of the activation of prothrombin was also evident when activated platelets were used as the source of the procoagulant phospholipids. The antithrombin III independent effect of heparin is unlikely to be important therapeutically, however, if this property of heparin is shared by other naturally occurring glycosaminoglycans, it could be important in maintaining the fluidity of blood under physiological conditions.     

Inhibition by calcium ions of thrombin

Amidolytic activity of thrombin for Boc-Val-Pro-Arg-MCA was inhibited by CaCl2 at relatively high concentrations. Kinetic analysis showed neither competitive nor noncompetitive inhibition by CaCl2 of thrombin. The nonspecific interaction of thrombin with calcium ions prolonged fibrinogen-clotting and neutralization by antithrombin III (AT III) in the absence of heparin. However, calcium ions had no effects on rate of thrombin inactivation in dilute solution and affinity of the enzyme for Sepharose 6B. The inverse effects of calcium ions on the thrombin activities in vitro and on generation of the enzyme in plasma ex vivo depended mostly on the ion concentration.     

The molecular mechanisms of heparin action: II. Separation of functionally different heparins by affinity chromatography

Heparin was separated into functionally distinct fractions based on an affinity for bovine antithrombin III or bovine Factor X. Fractions with high affinity for antithrombin III (HA-ATIII) had high activity in the enhancement of antithrombin III inhibition of thrombin. These fractions functioned poorly in the anticoagulation of plasma and as inhibitors of prothrombin activation by Factor Xa, calcium and phospholipid. In the absence of phospholipid, HA-ATIII heparin stimulated antithrombin III inhibition of Factor Xa while in the presence of phospholipid HA-ATIII had little effect. Heparin fractions with a high affinity for antithrombin III were poor inhibitors of prothrombin activation and did not inhibit Factor Xa-phospholipid binding.Heparin fractions with high affinity for Factor X(HA-FX) were very effective in the inhibition of prothrombin activation by Factor Xa, calcium and phospholipid, as well as inhibition of Factor Xa binding to phospholipid vesicles. HA-FX fractions of heparin were poor anticoagulants in plasma and did not accelerate inhibition of thrombin by antithrombin III. HA-FX heparin accelerated antithrombin III inhibition of Factor Xa both in the presence and absence of phospholipid.When HA-FX heparin fractions were combined with HA-ATIII heparin fractions, the anticoagulant activity of the mixture was greater than the sum of the activities of the individual fractions. These results indicaté that the interaction of heparin with both Factor Xa and antithrombin III is functionally important in the anticoagulation of blood.     

Inactivation of thrombin by antithrombin III on a heparinized biomaterial

Heparin covalently bonded to polyvinyl alcohol (PVA) is potentially useful as a nonthrombogenic coating in the preparation of small diameter vascular prostheses and blood sampling catheters. PVA-heparin is highly stable: the elution rate of ³⁵S-heparin from the polymer was determined to be negligible (approx. 2 × 10⁻¹¹ g/cm² min) when washed with either buffered saline (pH 7.4) or citrated human plasma. The inactivation of thrombin by antithrombin III was studied on PVA-heparin. Using small columns of PVA-heparin beads eluted by 0.14M NaCl buffered at pH 7.4 both thrombin and antithrombin III bound to the immobilized heparin. If thrombin was loaded before an excess of antithrombin III, significant inactivation of thrombin was observed; however, loading antithrombin III before thrombin did not measurably inactivate thrombin. The results suggest that the covalently-bound heparin effectively participates in the inactivation of thrombin through the formation of surface-bound heparin-thrombin, which then reacts with antithrombin III to yield a surface-bound thrombin-antithrombin III complex. The fate of this surface-bound complex has yet to be clarified.     

Distinction of two pathologic antithrombin III molecules: Antithrombin III ‘Aalborg’ and antithrombin III ‘Budapest’

Plasma from two different thrombophilic families with functional inherited antithrombin III deficiency, i.e., with low antithrombin III activity but normal immunoreactive antithrombin III concentration, were investigated simultaneously in the same laboratory. The experiments (thrombin and Factor Xa inactivation, heparin affinity chromatography, modified two dimensional immunoelectrophoresis and gel filtration) showed a distinct difference between the two antithrombin III anomalies. The antithrombin III ‘Aalborg’ had decreased thrombininactivating activity but normal Factor Xa-inactivating activity. The heparin affinity and the molecule weight are normal. The antithrombin III ‘Budapest’ displays a more profound abnormality with pathologic thrombin and Factor Xa inactivation, decreased heparin affinity and abnormal molecular weight.     

Effect of heparin and low molecular weight heparins on thrombin-induced blood platelet activation in the absence of antithrombin III

We have investigated the antithrombin III independent effect of crude heparin, two heparin fractions and a heparinoid on in vitro thrombin-induced platelet activation. Thrombin-induced platelet factor Va generation and thrombin plus collagen-induced platelet prothrombin converting activity were tested. Crude heparin was a more potent inhibitor of these reactions than the fractions or the heparinoid. The inhibitory action of the heparins was found to be the result of a direct effect on thrombin and not of an effect either on platelet activation functions or on the assembly or functioning of the prothrombinase complex. Probably this heparin inhibition is due to the masking of secondary macromolecular substrate binding sites on the thrombin molecule. We found no correlation between IC50 values and the antithrombin III-dependent antithrombin specific activities of the heparins. This supports the notion that heparin properties other than their affinity for antithrombin III may contribute to the action of this drug in blood coagulation.     

Abolition by dextran sulfate of the heparin-accelerated antithrombin III/thrombin reaction

Dextran sulfate did not inhibit the amidolytic activity of thrombin on Boc-Val-Pro-Arg-4-methylcoumaryl-7-amide, but abolished inhibition of the enzyme with antithrombin III (AT III) in the presence of heparin.

Dextran sulfate did not bind to AT III and had less affinity than immobilized heparin for the protein. Dextran sulfate bound strongly to thrombin and had higher affinity than immobilized heparin for the enzyme.

These findings indicate that binding of dextran sulfate to a site other than the active site of thrombin to prevent the approach of AT III in the presence of heparin.     

Catalytic and regulatory functions of N-bromosuccinimide-modified bovine thrombin

At pH 4.1, bovine thrombin reacts rapidly with N-bromo-succinimide to yield modified enzyme containing oxidized tryptophan residue. Both fibrinogen clotting activity and esterase activity are reduced considerably when three moles of tryptophan residues per mole of thrombin are oxidized, but the Michaelis constants for synthetic substrates are not appreciably altered. Reaction of NBS also results in a decrease in the affinity of thrombin for heparin. The dissociation constant for heparin-thrombin complex is increased by 2.6-fold due to the modification of one tryptophan residue. However, the magnitude of the increase in the dissociation constant remains the same for modified enzymes containing approximately two or three oxidized tryptophan residues. The rate constant for the inactivation of thrombin by antithrombin III is increased by 2.5-fold due to the modification of a single tryptophan residue. This increase in rate constant is not further amplified when more than one tryptophan residue is oxidized. In contrast, in the presence of heparin the rate of inactivation of modified and unmodified thrombins by antithrombin III are not significantly different. Thus, the heparin-sensitized inactivation of thrombin by antithrombin III is affected by the modification of one tryptophan residue. Spectrophotometric titrations of the phenolic hydroxyl groups suggest that the structural environments of tyrosyl groups for both unmodified and modified thrombin containing one oxidized tryptophan residue, are similar. The temperature for half loss of catalytic activity of control and NBS-modified thrombin, containing one oxidized tryptophan, are 52 and 51.5 degrees C respectively. It appears that the one tryptophan residue of thrombin is situated at or close to the binding site of heparin.     

Homozygous variant of antithrombin III that lacks affinity for heparin, AT III Kumamoto

Abnormal antithrombin III (AT III) was found in the plasma of a 31-year-old female who suffered from recurrent thrombotic episodes. Heparin cofactor activity was 28% of normal and undetectable when measured by inhibition of thrombin and factor Xa (F.Xa), while both progressive antithrombin and antifactor Xa activities were normal. The concentration of plasma AT III antigen was 37 mg/dl. Analysis by crossed-immunoelectrophoresis (CIE) in the presence of heparin and affinity chromatography on heparin-Sepharose revealed that the propositus' AT III did not bind to heparin. When heparin cofactor II (HC II) was removed from propositus' plasma, heparin cofactor activity of AT III was not detected. Thus, HC II seemed to account for the plasma heparin cofactor activity found in the presence of thrombin. The patient's parents and three of her brothers demonstrated qualitative abnormality of AT III; heparin cofactor activity was 30-50% of normal levels in the presence of both thrombin and F.Xa. These findings indicate that the propositus' AT III lacks affinity for heparin and the mode of its inheritance seems to be autosomal dominant and, hence, the propositus would be a homozygote. For this variant, the name of AT III Kumamoto is proposed.     

Effect of calcium phosphate on thrombin and on its sensitivity to heparin and antithrombin-III

Thrombin was inactivated by antithrombin-III and the rate of inactivation was accelerated by heparin either in 0.04 M sodium phosphate buffer or in 0.04 M Tris-HCl buffer, at pH 7.4. Calcium chloride, at a final concentration of 1 mM, influenced slightly thrombin inactivation by both antithrombin and antithrombin plus heparin in Tris-HCl buffer, whereas in phosphate buffer the sensitivity of enzyme to antithrombin and heparin decreased.Thrombin showed also increased stability against the inactivating effect of heat at 54°C in sodium phosphate buffer containing calcium chloride.These findings suggest that thrombin bound to calcium phosphate is extremely stable against inactivation by antithrombin and heparin, while its clotting activity does not change. This nature of enzyme may play a role in predisposition for thrombosis during arteriosclerosis.     

The effects of human thrombomodulin on the inactivation of thrombin by its serum inhibitors

Thrombomodulin is an endothelial cell protein which accelerates thrombin-dependent protein C activation by over 1000 fold. In this study, the effect of thrombomodulin on the inactivation of thrombin by its serum inhibitors was evaluated. 125I-thrombin was incubated at 37 degrees C with serum and the resulting complexes separated by SDS-PAGE. Antithrombin III was the major complex formed with some 125I-thrombin bound to heparin cofactor II and higher molecular weight fractions. Inclusion of thrombomodulin at increasing concentrations inhibited 125I-thrombin binding to antithrombin III and the higher molecular weight fractions but had little effect on thrombin-heparin cofactor II complex formation. Similar results were obtained using a purified antithrombin III/heparin cofactor II system. Kinetic studies, using purified antithrombin III, revealed that thrombomodulin acts as a weak competitive inhibitor towards antithrombin III (Ki = 39 nM). We postulate that in the microcirculation, where the ratio of thrombomodulin to antithrombin III is relatively high, thrombin bound to thrombomodulin may be protected from inactivation by antithrombin III and can thus promote efficient activation of protein C.     

Interaction of the sulfated lactobionic acid amide LW 10082 with thrombin and its endogenous inhibitors

The synthetic polyanion LW 10082 prolonged thrombin clotting times when human or bovine thrombin were used to induce clotting. In the clotting time prolongation LW 10082 was 15 times more potent in the presence of human thrombin than of bovine thrombin. In an amidolytic thrombin inhibition assay, LW 10082 inhibited bovine and human thrombin when plasma was present. Here, too, thrombin inhibition was much more pronounced when human thrombin was used. When purified antithrombin III was used, neither human nor bovine thrombin were inhibited by LW 10082. LW 10082 proved to be a potent inhibitor of human thrombin in the presence of heparin cofactor II in an amidolytic assay. The time course of the inhibition was slow and at least 10 minutes of incubation were required to obtain complete inhibition of thrombin by an excess of HC II and LW 10082. The anticoagulant activity of LW 10082 in the APTT was unchanged in the presence of anti-antithrombin III antibodies while the potency of heparin was clearly reduced. This indicates that the overall anticoagulant effect of LW 10082 was as well independent of antithrombin III.     

Interaction of antithrombin III and thrombin-antithrombin III complex with cultured aortic endothelial cells

The binding of antithrombin III, thrombin, thrombin-antithrombin III complex to endothelial cells was investigated. While the rate of the binding of thrombin to these cells was very rapid, that of antithrombin III was relatively slow and the thrombin-antithrombin III complex was intermediate. Binding kinetics indicated that antithrombin III, like thrombin, showed high affinity to endothelial cells; with a Kd of 3 X 10(-8) M and with 5 X 10(4) binding sites per cell. The dissociation of the inhibitor molecule was also rapid, i.e., approximately 70% bound antithrombin III was released in 2 minutes. Heparin, in a 100-fold molar excess to antithrombin III, or the modification of lysine residues of the inhibitor involved in the interaction with heparin, did not influence the association of antithrombin III with endothelial cells. In addition, antithrombin III did not compete with thrombin blocked in its active center for binding to endothelial cells. It is suggested that the binding sites of endothelial cells are different for thrombin and antithrombin III, and antithrombin III does not bind to these cells through its heparin binding domain.     

Uptake and inactivation of thrombin by the fresh, glutardialdehyde or heparin treated human umbilical cord vein endothelium

With the aim to elucidate the non-thrombogenic properties of the vascular endothelium and the lining of a certain arterial substitute, the uptake and inactivation of thrombin by the native and modified human umbilical cord veins were investigated. The modification procedures included variations with glutardialdehyde treatment, heparin exposure and storage in ethyl alcohol. The native and all the modified veins adsorbed substantial amounts of thrombin, which was immobilized in an active state. Following exposure to antithrombin III solution, the adsorbed thrombin was inhibited only on the native veins and those treated with glutardialdehyde, stored in ethyl alcohol, exposed to heparin and again stored in ethyl alcohol.     

Regulation of the thrombin time in cirrhosis.

Standard coagulation profiles, plasmin and plasminogen activities, and measurements of antithrombin III and antithrombin II (heparin co-factor) activities were performed on 9 patients with biopsy-proven cirrhosis and on a 10th patient studied over a 30 month period for biopsy-proven cirrhosis associated with persistent hypofibrinogenemia. Prolongation of the thrombin time was observed in all 10 patients and was not correlated with fibrinogen concentrations, plasmin nor plasminogen activity, nor the appearance nor magnitude of fibrin-fibrinogen split product titers. There was a negative correlation between the presence of free plasmin in these patients and the presence or magnitude of fibrin-fibrinogen split product titers. The one patient with persistent hypo-fibrinogenemia demonstrated consistently shorter prolongation of the thrombin time on all occasions despite lower fibrinogen levels, longer prothrombin and partial thromboplastin times, and higher plasmin and fibrin-fibrinogen split product concentrations than the other 9 patients studied. This patient's relative normalization of the thrombin time was most likely related to negligible antithrombin III and heparin co-factor activities, and suggests the significance of these antithrombins as possible regulators of the thrombin time in cirrhosis.     

Fate of thrombin and thrombin-antithrombin-III complex adsorbed to a heparinized biomaterial: analysis of the enzyme-inhibitor complexes displaced by plasma

Heparin covalently-linked to polyvinyl alcohol (PVA) is a biomaterial which is of potential value as a non-thrombogenic coating. 125I-labelled thrombin adsorbed to heparin-PVA beads was not dislodged by phosphate-buffered saline, pH 7.4, although radioactivity was progressively displaced from the adsorbent by fibrinogen-free human plasma. Analysis by gel filtration and affinity chromatography showed that the released radioactivity was distributed between (thrombin-antithrombin-III) complex (approx. 70%) and, probably, (thrombin-alpha-2-macroglobulin) complex (approx. 30%). Less efficient thrombin displacement was obtained by either bovine serum albumin (5% w/v) or antithrombin-III-free human plasma: in the latter case, the dislodged enzyme was presumably associated with alpha-2-macroglobulin. Purified alpha-2-macroglobulin did not displace thrombin from heparin-PVA. The quantity of thrombin displaced by an alpha-2-macroglobulin-free plasma fraction compared well with fibrinogen-free plasma: The eluted enzyme was largely associated with antithrombin-III. Purified antithrombin-III did not displace thrombin from heparin-PVA despite causing greater than 70% inactivation of the bound enzyme. Subsequent treatment with fibrinogen-free plasma dislodged (thrombin-antithrombin-III) at a similar rate to that of bound thrombin. We conclude that plasma contains a component(s) which displaces (thrombin-antithrombin-III) complex from immobilised heparin: presumably this leaves the heparin sites free for further use in enzyme inactivation.