Circulating microparticles and endogenous estrogen in newly menopausal women.

BACKGROUND: Estrogen modulates antithrombotic characteristics of the vascular endothelium and the interaction of blood elements with the vascular surface. A marker of these modulatory activities is formation of cell-specific microparticles. This study examined the relationship between blood-borne microparticles and endogenous estrogen at menopause. METHODS: Platelet activation and plasma microparticles were characterized from women being screened (n = 146) for the Kronos Early Estrogen Prevention Study. Women were grouped according to serum estrogen (< 20 pg/ml; low estrogen, n = 21 or > 40 pg/ml; high estrogen, n = 11). RESULTS: Age, body mass index, blood pressure and blood chemistries were the same in both groups. No woman was hypertensive, diabetic or a current smoker. Platelet counts, basal and activated expression of P-selectin on platelet membranes were the same, but activated expression of glycoprotein IIb/IIIa was greater in the high-estrogen group. Numbers of endothelium-, platelet-, monocyte- and granulocyte-derived microparticles were greater in the low-estrogen group. Of the total numbers of microparticles, those positive for phosphatidylserine and tissue factor were also greater in the low-estrogen group. CONCLUSION: These results suggest that, with declines in endogenous estrogen at menopause, numbers of procoagulant microparticles increase and thus may provide a means to explore mechanisms for cardiovascular risk development in newly menopausal women.

Clearance of thrombin from circulation in rabbits by high-affinity binding sites on endothelium. Possible role in the inactivation of thrombin by antithrombin III.

The clearance of (125)I-thrombin and diisopropylphosphoryl-(125)I-thrombin (DIP-thrombin) from the circulation in rabbits was studied. When given either intraarterially or intravenously, DIP-thrombin, which is active-site blocked, was approximately 90% cleared from the circulation by 1 min, the time of earliest sampling, indicating a large first-pass effect. DIP-thrombin given intravenously is found predominantly in the lungs, whereas DIP-thrombin injected into the aortic arch is distributed diffusely in approximate proportion to the blood supply. Renal artery, femoral artery, ear artery, left atrium, and portal vein infusions demonstrate that kidney, muscle, ear, heart, and liver, respectively, can remove DIP-thrombin from the circulation. These data imply that the clearance of DIP-thrombin is not a function of a specific organ but of the vascular bed per se. The clearance of DIP-thrombin was reversible since injection of 0.5 mg of unlabeled DIP-thrombin 10 min after the injection of a tracer dose of DIP-(125)I-thrombin resulted in the rapid reappearance of the DIP-(125)I-thrombin into the circulation. In addition, the clearance of DIP-thrombin was saturable, i.e., clearance of DIP-(125)I-thrombin was inhibited by unlabeled DIP-thrombin in a dose-dependent fashion. In vivo Scatchard analysis of the saturation of the clearance process demonstrated that DIP-thrombin can be removed by binding to high-affinity binding sites, since dissociation constants (K(D)) of 10 and 13 nM were obtained for human and bovine DIP-thrombin, respectively. In contrast to DIP-thrombin, approximately 75% of the radioactivity associated with active thrombin remained in the circulation at 1 min. By 10 min 55% of (125)I-thrombin had been removed from the circulation, and essentially all of the radioactivity can be accounted for in the liver. Sodium dodecyl sulfate-polyacrylamide gel radioelectrophoresis of plasma samples taken after injection of (125)I-thrombin demonstrated that all of the active thrombin was converted to covalent thrombin-antithrombin III complex by the time of initial sampling (30 s). The in vitro conversion of (125)I-thrombin to thrombin-antithrombin III complex was considerably slower (50+/-5% conversion at 30 s). The simultaneous injection of excess unlabeled DIP-thrombin inhibited the rate of formation of (125)I-thrombin-antithrombin III complex formation in vivo (but not in vitro), which suggests that the binding of active thrombin to the high affinity binding sites is required for the rapid inactivation of thrombin in vivo. We propose that (a) thrombin in the circulation binds to active site-independent high-affinity binding sites on the endothelial cell surface; (b) the inactivation of thrombin by antithrombin III is faster in vivo than in vitro because the high-affinity binding sites, present in a high concentration in the microcirculation, catalyze the reaction; (c) thrombin-antithrombin III complexes are selectively removed by the liver.

Identification in vitro of an endothelial cell surface cofactor for antithrombin III. Parallel studies with isolated perfused rat hearts and microcarrier cultures of bovine endothelium.

Two in vitro systems were used to identify an antithrombin III cofactor activity on vascular endothelium. Langendorff rat heart preparations or columns packed with endothelium cultured on microcarrier beads were perfused with mixtures of purified thrombin and antithrombin III. With each preparation, accelerated inhibition of thrombin by antithrombin III occurred during passage over endothelium. Platelet factor 4, protamine sulfate and diisopropylphosphoryl thrombin, all antagonists of the antithrombin III cofactor activity of heparin, significantly reduced the capacity of the preparation to inhibit thrombin. It is concluded that a substance with the functional properties of a stationary phase cofactor for antithrombin III is present on the microvascular endothelium and there catalyzes the inactivation of circulating free thrombin.

Identification of a congenital dysthrombin, thrombin Quick.

A dysprothrombin designated prothrombin Quick, is isolated from the plasma of an individual with < 2% of normal functional prothrombin activity and 34% of the normal prothrombin level by immunologic assay. With Factor Xa or taipan snake venom as activators, a fragmentation pattern identical to that of normal prothrombin is observed on gel electrophoresis in dodecylsulfate. This evidence combined with the observed barium citrate adsorption of prothrombin Quick and the low activity suggests that the defect in prothrombin Quick is in the thrombin portion of the molecule. Thrombin Quick is isolated and comigrates with thrombin on dodecyl sulfate gel electrophoresis, either reduced or nonreduced. The activity of thrombin Quick on several biological substrates of thrombin is investigated. Relative to normal thrombin, thrombin Quick is 1/200 as active on fibrinogen and 1/20-1/50 as effective in activating Factors V and VIII and aggregating platelets. A complex with antithrombin III is detected by dodecyl sulfate gel electrophoresis. Further investigation with the active site titrant, dansylanginine-N-(3-ethyl-1,5-pentanediyl)amide showed that the thrombin Quick preparation has the same affinity for the titrant as thrombin, but apparently only 40% active sites per mole protein are titrable.

Evidence for the formation of an ester between thrombin and heparin cofactor.

Heparin cofactor, a thrombin inhibitor, is purified from human plasma by affinity chromatography on heparin-agarose. The nature of the binding between thrombin and the inhibitor is studied by treatment of the complex with 6 M guanidinium chloride, hydroxylamine, and dilute alkali. The complex is not dissociated during gel chromatography in 6 M guanidinium chloride. This result supports an earlier proposal that formation of the complex includes the formation of a covalend bond. Treatment of dodecylsulfate-denatured complex with hydroxylamine results in dissociation of the complex to yield free thrombin and heparin cofactor. The complex is also dissociated in dilute NaOH (pH 12) solutions. These results indicate that the covalent bond between thrombin and the inhibitor is a carboxylic ester.

Effect of heparin on the extent of inhibition of thrombin by heparin cofactor.

A heparin preparation obtained by gel chromatography is compared to unfractionated heparin with respect to the effects of heparin on the reaction between thrombin and heparin cofactor. Whereas both preparations enhance the rate of inhibition of thrombin by heparin cofactor, the extent of inhibition is decreased by the unfractionated, but not by the fractionated heparin. The decreased extent of inhibition is accounted for by residua of unreacted and undegraded heparin cofactor and thrombin, as demonstrated by gel electrophoresis in dodecyl sulfate. However both heparin preparations enhance the rate of degradation by thrombin of the thrombin-heparin cofactor complex.

6-aminohexanoate and chloride ion in the activation by urokinase of porcine plasminogens.

The rate of activation by urokinase of porcine plasminogen is accelerated by 6-aminohexanoate, although the maximally enhanced rate is 10-fold less than that of human plasminogen without the amino acid. 6-Aminohexanoate facilitates only activation of native porcine plasminogen (asp-plasminogen), but has no effect on activation of des-kringle1-4-plasminogen. Sodium chloride, on the other hand, inhibits activation by urokinase of both porcine asp-plasminogen and des-kringle1-4-plasminogen. It is concluded that 6-aminohexanoate exerts its effect via kringle1-4 domains of plasminogen, whereas Cl- acts, at least in part, through effects on the kringle5 or proteinase domains.

A catalytic role for heparin. Evidence for a ternary complex of heparin cofactor thrombin and heparin.

The interaction of heparin with chemically modified thrombin and heparin cofactor is studied. Amidinated heparin cofactor does not bind to heparin-agarose and the reaction rate of the amidinated inhibitor with unmodified thrombin is not affected by heparin. Likewise, thrombin modified with 1,2--cyclohexanedione does not bind to heparin agarose and the reaction rate of the modified enzyme with unmodified inhibitor is not affected by heparin. In the absence of heparin, the modified and unmodified proteins react at the same rate in all possible combinations. Affinity chromatography of diisopropylphosphoryl thrombin on heparin cofactor coupled to Sephadex G--50 is used to study the binding of heparin cofactor and thrombin to heparin. The thrombin for all experiments is tritium-labeled and then inactivated with diispropylfluorophosphate. Thrombin is not bound to heparin cofactor-Sephadex columns. However, after treatment of the columns with a heparin solution, thrombin binds tightly, and is eluted at high ionic strength. Bound thrombin can also be eluted with either excess non-radioactive thrombin or excess free heparin. Heparin-dependent binding of thrombin does not occur if the heparin cofactor-Sephadex is heat-denatured. The ability of heparin to couple solution-phase thrombin to solid-phase heparin cofactor indicates that a ternary complex is formed. Analysis of the binding of the proteins to heparin by a dye displacement method suggests that at least one site on heparin binds to thrombin but not to heparin cofactor. Further support for a catalytic role for heparin derives from the ability of catalytic concentrations of heparin to enhance the rate of hydrolysis of prothrombin by thrombin, another protein pair which bind mutually to heparin.

Dual effect of synthetic plasmin substrates on plasminogen activation.

The effect of plasmin substrates D-valyl-L-leucyl-lysine-p-nitroanilide (S-2251) and H-D-norleucyl-hexahydrotyrosyl-lysine-p-nitro-anilide (Spectrozyme-PL) on the rate of activation of native human plasminogen in physiological salt solution is studied. Plasminogen activation by two-chain urokinase-type plasminogen activator (urokinase), two-chain tissue-type plasminogen activator (tc-tPA) or trypsin, but not by single chain tPA (sc-tPA) is increased 5- to 10-fold by both substrates, as determined by electrophoretic and spectrophotometric kinetic analysis. The amidolytic activity of sc-tPA, on the other hand, is inhibited by the plasmin substrates in a non-competitive manner (K1 of 6.4 . 10(-4) M for S-2251 and 2.9 . 10(-4) M for Spectrozyme-PL), whereas urokinase and tc-tPA activities are not affected. It is concluded that plasmin substrates containing a lysine residue have a general capacity to enhance plasminogen activation presumably by inducing a conformational change in the native zymogen in a manner similar to 6-aminohexanoate, while the same substrates are inhibitory both on the amidolytic activity of sc-tPA and the activation of native and des1-77-plasminogen by sc-tPA.

Prospective randomized study of effects of unopposed estrogen replacement therapy on markers of coagulation and inflammation in postmenopausal women.

Estrogen replacement therapy decreases the risk of arterial disease while at the same time increases the risk for venous thrombosis. Whether a common mechanism(s) of coagulation and inflammation contributes to both responses is unclear. This study determined simultaneous effects of estrogen replacement therapy on regulators of the direct (extrinsic) pathway for activation of coagulation, coagulation, and the acute phase response. Plasma from 26 postmenopausal women without risk factors for cardiovascular disease was collected before (baseline) and after 3 months of treatment with either conjugated equine estrogen (Premarin, 0.625 mg/d) or placebo. Plasma lipids, tissue factor pathway inhibitor antigen and activity, plasminogen, prothrombin, P-selectin, alpha1-protease inhibitor, and C-reactive protein were measured. Estrogen replacement therapy significantly reduced mean concentrations of tissue factor pathway inhibitor (antigen and activity; P < 0.001), which were correlated significantly to decreases in low density lipoprotein (r2 = 0.71). Plasminogen and C-reactive protein increased significantly. Other parameters were unchanged. The results of this prospective study suggest that 3 months of estrogen replacement therapy in healthy postmenopausal women decreases low density lipoprotein with simultaneous decreases in tissue factor pathway inhibitor, a major inhibitor of the extrinsic coagulation pathway, and increases C-reactive protein, a component of the acute phase response. Concomitant changes in these parameters may reduce the risk for arterial disease while altering the threshold for thrombotic events.

Myosin as cofactor and substrate in fibrinolysis.

Myosin accelerates plasminogen activation by tissue-type plasminogen activator (tPA), and is degraded extensively by plasmin. Myosin binds both tPA and plasminogen, and enhances activation of des1-77-plasminogen by tPA but not by urokinase-type plasminogen activator (uPA). Myosin decreases K(M) and increases k(cat) for des1-77-plasminogen activation by tPA, to yield catalytic efficiencies in excess of 8000 M-1 s-1. The effect of myosin is attributed to its C-terminal portion, the myosin rod. With a K(M) of 3 microM, myosin is a high-affinity substrate for plasmin. The findings indicate that myosin is a cofactor for plasminogen activation and a substrate for plasmin.

Chemical and electrical synapses between photoreceptors in the retina of the turtle, Chelydra serpentina.

Electron microscopic observation of the distal retina of the turtle, Chelydra serpentina, revealed that photoreceptors contact each other by means of a variety of junctions. The synaptic terminals of the primary and accessory members of a double cone invariably make punctate contact with each other distal to the basal surfaces. This type of contact was only rarely seen between the synaptic terminals of single photoreceptors. Photoreceptor telodendria which emanate from the basal surface of the synaptic terminal and ramify laterally for up to 40 microns in the outer plexiform layer give off branchlets which contact the terminals of both rods and cones. These contacts resemble superficial basal junctions and can be of the "wide-gap" or "narrow-gap" type. In cones the branchlets of the telodendria sometimes make a "distal junction" with one of the lateral elements of a synaptic dyad. Gap junctions were not found between the fins radiating from the myoids of adjacent receptors but were seen to occur en passant between photoreceptor telodendria in the outer plexiform layer. These junctions, though small, are probably of sufficient size to account for the rod-rod coupling observed physiologically in this retina. A prominent organelle resembling a synaptic lamella was occasionally seen in processes making contact with photoreceptors. These processes were identified as bipolar cell dendrites.

Gap junctions among the perikarya, dendrites, and axon terminals of the luminosity-type horizontal cell of the turtle retina.

Gap junctions of the H1 horizontal cell of the turtle retina (Leeper, '78) were studied in thin-sectioned material and in freeze-fracture replicas. Perikaryal gap junctions were extremely restricted, 0.02-0.07 micron2 in in area, whereas those of axon terminals were much larger, most being 0.1-1.0 micron2. Both varieties, however, had the usual seven-layered appearance in thin section and measured 15 +/- 1 nm in overall width between cytoplasmic faces. Freeze-fractured views of the perikaryal junctions revealed roughly circular patches of P-face 9-nm particles and E-face pits. The axon terminal gap junctions were seen as large areas of P-face particles and E-face pits containing occasional islands of unspecialized membrane. Particle densities varied from 1,455 to 2,448 microns-2. A serial reconstruction was made of a portion of the axon terminal network in order to measure the surface areas of the axons contained therein and the fraction occupied by gap junctions. These data demonstrated that the fractional area occupied by gap junctions was roughly in inverse proportion to the area of the axon region (tuberous core vs. terminal process). It is argued that this constitutes an impedance matching device to ensure adequate current flow through the axon processes. Assuming that each P-face particle represents a connection having a conductance of 10(-10) S and given the P-face particle density and gap junctional areas determined in this report, we calculated that the gap junction distribution is adequate to account for the spatial properties of the horizontal cell axon network (Lamb, '76).

Dopamine D2 receptor-mediated modulation of rod-cone coupling in the Xenopus retina.

We studied the responses of rod photoreceptors that were elicited with light flashes or sinusoidally modulated light by using intracellular recording. Dark-adapted Xenopus rod photoreceptors responded to sinusoidally modulated green lights at temporal frequencies between 1 Hz and 4 Hz. In normal Ringer's solution, 57% of the rods tested could follow red lights that were matched for equal rod absorbance to frequencies >5 Hz, indicating an input from red-sensitive cones. Quinpirole (10 microM), a D2 dopamine agonist, increased rod-cone coupling, whereas spiperone (5 microM), a selective D2 antagonist, completely suppressed it. D1 dopamine ligands were without effect. Neurobiotin that was injected into single rods diffused into neighboring rods and cones in quinpirole-treated retinas but only diffused into rods in spiperone-treated retinas. A subpopulation of rods (ca. 10% total rods) received a very strong cone input, which quickened the kinetics of their responses to red flashes and greatly increased the bandpass of their responses to sinusoidally modulated light. Based on electron microscopic examination, which showed that rod-rod and cone-cone gap junctions are common, whereas rod-cone junctions are relatively rare, we postulate that cone signals enter the rod network through a minority of rods with strong cone connections, from which the cone signal is further distributed in the rod network. A semiquantitative model of coupling, based on measures of gap-junction size and distribution and estimates of their conductance and open times, provides support for this assumption. The same network would permit rod signals to reach cones.

Platelet characteristics associated with coronary artery disease.

BACKGROUND/OBJECTIVE: To test the hypothesis that circulating platelets display evidence of interactions with atherogenesis, platelet capacity to express P-selectin and propensity for spontaneous microaggregation in vitro were measured in samples from normal donors (N), patients with asymptomatic advanced coronary calcification (CC) or acute coronary syndromes (AC). To measure the effect of angioplasty on platelet function, samples obtained before, 30 min after and 24 h after angioplasty were compared. PATIENTS/METHODS: Platelet P-selectin was measured after maximal stimulation with thrombin. Microaggregation was measured as a platelet count deficit in citrate-anticoagulated platelet-rich plasma (PRP) relative to EDTA-anticoagulated blood. RESULTS: P-selectin expression was significantly lower for platelets from patients with either AC or CC compared to normals. In addition, platelets from AC and CC patients have a significantly greater propensity to form microaggregates in citrate anticoagulant. After angioplasty, the PRP-platelet count decreased transiently. CONCLUSION: Both acute unstable and chronic stable coronary disease are associated with an increased share of platelets unable to express P-selectin and an increased share of platelets that microaggregate in citrate anticoagulant. The genesis of these platelet characteristics is not fully explained by focal acute arterial injury and may reflect exposure to systemic atherosclerosis or the atherogenic process.

Inhibition and reversal of platelet-rich arterial thrombus in vivo: direct vs. indirect factor Xa inhibition.

BACKGROUND/OBJECTIVE: The efficacy of a direct factor (F)Xa inhibitor, ZK-807834, was compared with indirect inhibition by enoxaparin for inhibition and deaggregation of acute platelet-rich thrombi in a well-characterized porcine carotid injury model. METHODS: A crush injury was performed on a randomly chosen carotid artery and the thrombus allowed to propagate for 30 min. Pigs then received intravenous drug for 35 min: ZK-807834-Dose 1 (40 microg kg(-1) bolus + 1.5 microg kg(-1) min(-1) infusion, n=6); ZK-807834-Dose 2 (20 microg kg(-1) bolus + 0.75 microg kg(-1) min(-1) infusion; n=6); enoxaparin (1 mg kg(-1) bolus; n=6); or saline (n=6). Five minutes after drug initiation, the contralateral artery was injured. Thrombus size was monitored by scintillation detection of autologous 111In-platelets. RESULTS: The prothrombin time ratio was 2.2 +/- 0.1; 1.4 +/- 0.3; 1.2 +/- 0.9 and 1.1 +/- 0.2, respectively. ZK-807834-Dose 1 significantly inhibited carotid platelet deposition (525 +/- 226 x 10(6) cm(-2); P = 0.008), whereas ZK-807834-Dose 2 (2325 +/- 768) and enoxaparin (1236 +/- 383) were not different from saline (2776 +/- 642). Thrombus deaggregation was greatest for animals receiving ZK-807834-Dose 1 (473 +/- 185). Neither ZK-807834-Dose 2 (1588 +/- 480) nor enoxaparin (1618 +/- 686) was different from saline control (2222 +/- 598). CONCLUSIONS: Direct FXa inhibition with ZK-807834, at a prothrombin time ratio of 2.2, effectively inhibits thrombosis and promptly deaggregates thrombi induced by arterial injury. In contrast, indirect FXa inhibition with enoxaparin was ineffective.

Atrial fibrillation and thrombosis: immunohistochemical differences between in situ and embolized thrombi.

BACKGROUND/OBJECTIVE: Thromboembolism secondary to atrial fibrillation accounts for approximately one-fourth of all strokes. Although considerable resources have been targeted to pharmacologic prophylaxis, neither the cellular nor the biochemical composition of atrial thrombi is known. Quantitative immunohistochemistry was undertaken to define the composition of atrial thrombi and to explore morphological differences between atrial appendage thrombi and those that embolize. PATIENTS/METHODS: Serial sections of thrombi obtained during valve replacement surgery or embolectomy from 22 patients with atrial fibrillation were stained with antibodies against fibrin, integrin beta3, or tissue factor and analyzed with NIH-image. RESULTS: Thrombi showed distinct regions staining for either fibrin or platelets and on average, the fibrin-rich regions predominated (P < 0.0001). The platelet content of embolized thrombi was nearly twice that of atrial thrombi (P = 0.02). Non-staining amorphous material comprised nearly half of atrial thrombi in situ, but was rare in embolized thrombi (P < 0.001). Tissue factor colocalized to areas rich in platelets and granulocytes. CONCLUSIONS: The abundance of fibrin relative to platelets underscores the enhanced efficacy of warfarin prophylaxis in clinical trials. The finding of tissue factor localized to platelet-leukocyte clusters suggests its blood-borne origin. Compositional differences between in situ and embolized thrombi suggest directions for investigating propensity for embolization.

Facilitation of plasminogen activation by a plasmin substrate containing a lysyl residue.

The plasmin substrate, H-D-norleucyl-hexahydrotyrosyl-lysine-p-nitroanilide (Spectrozyme-PL), was found to be equivalent to 6-aminohexanoate as an enhancer of porcine and human plasminogen activation by urokinase and of removal of the 1-77 peptide of plasminogen by plasmin. Activation of plasminogen lacking kringles 1-4, on the other hand, was not influenced by Spectrozyme PL. Although the rate of activation of human plasminogen and the modification of human plasminogen by plasmin are faster by an order of magnitude than that of the activation and modification of porcine plasminogen, both reactions in the human zymogen, the hydrolysis at arg561-val562 and at lys77-lys78, are accelerated by Spectrozyme PL. The findings indicate that kinetic interpretation of plasminogen activation in solutions containing substrates, where the substrate has been incorporated to inhibit feedback proteolysis by plasmin, must account for the cofactor activity as well as the inhibitory activity of the substrate.

Reversibility of platelet thrombosis in vivo. Quantitative analysis in porcine carotid arteries.

Reversibility in vivo of acute platelet thrombosis in response to specific anticoagulants is analyzed with thrombi that develop in segments (1 cm) of porcine carotid arteries externally crushed with a hemostat. Most thrombi fill the lumens of the injured segments (ca. 1 cm x 3 mm, 1 x w) within 30 min and comprise masses of platelets interpenetrated with neutrophil-lined seepage channels of blood. Continuous quantitative assay of thrombus mass is provided by a gamma detector placed over the injured segments to collect counts from 111In-labeled platelets. Thrombi established 30 min after injury, otherwise stable for 6 h, clear during 30-60 min of continuous infusion of either hirudin, tick anticoagulant or activated porcine protein C, or intermittent activation of endogenous protein C with a latent thrombin reagent. Anticoagulant dose-dependence of thrombus clearance is established for hirudin between 0.01 and 1.0 mg/kg/min. Thrombi become progressively refractory to hirudin between 0.5 and 6 h after injury. Neither heparin no low-molecular-weight heparin in full (clinical) anticoagulant doses yield significant dethrombosis. It is concluded that, within time limits, controlled thrombin generation in platelet thrombi maintains platelet cohesion without catalyzing irreversible platelet aggregation or clotting of fibrinogen.

Evidence for an ester bond between thrombin and heparin cofactor.

Heparin cofactor, a thrombin inhibitor, is purified from human plasma by affinity chromatography on heparin-agarose. The nature of the binding between thrombin and the inhibitor is studied by treatment of the complex with 6 M guanidinium chloride, hydroxylamine, and dilute alkali. The complex is not dissociated during gel chromatography in 6 M guanidinium chloride. This result supports an earlier proposal that formation of the complex includes the formation of a covalent bond. Treatment of dodecyl sulfate-denatured complex with hydroxylamine results in dissociation of the complex to yield free thrombin and heparin cofactor. Hydroxylamine does not dissociate the complex unless it is denatured. The complex is also dissociated in dilute sodium hydroxide (pH 12) solutions. These results indicate that the covalent bond between thrombin and the inhibitor is a carboxylic ester.