A method to determine the affinity of heparin to thrombin and antithrombin III on equilibrium gel permeation chromatography

Equilibrium gel permeation chromatography was employed to determine the ability of heparin to form complexes with thrombin and antithrombin III. In the eluate from a Sephacryl S-200 column, heparin caused a peak and then a trough in the fluorescence of 48 nM antithrombin III or 63 nM thrombin. The peak-heights with known amounts of heparin were used for standard curves to determine the extent of complex formation by test heparin preparations. Only heparin species with high-affinity for antithrombin III specifically formed a complex with antithrombin III under the conditions used. The ability to form a complex of heparin preparations with different anticoagulant activities for thrombin and antithrombin III could be determined satisfactorily. The heparin species with different affinities for antithrombin III did not coincide those with different affinities for thrombin. Of 4 preparations with one low-affinity and three high-affinity subfractions of heparin for antithrombin III, the species with the lowest affinity for antithrombin III had the highest affinity for thrombin. All of these observations showed that the method could be used to determine the ability to form a complex of test heparin preparations.     

Behavior of antithrombin III isoforms on immobilized heparins. Evidence that the isoforms bind to different numbers of low-affinity heparin sites

Antithrombin III exists in plasma as major and minor isoforms differing in affinity for heparin. The nature of the binding of each purified isoform to immobilized heparins was investigated. Unfractionated, mixed-affinity heparin bound each isoform with both high affinity and concentration-dependent low affinity. The isoforms were resolved when filtered through low-affinity heparin (heparin repeatedly passed over immobilized antithrombin III) columns. Following chemical modification of a specific tryptophan residue required for heparin binding, each isoform failed to bind to either low-affinity or mixed-affinity heparin-agarose, but elution of the modified higher-affinity isoform was retarded on both gels. Because the modified lower-affinity isoform eluted with the similarly sized bovine serum albumin in these experiments, the difference in isoform affinity for heparin appears to be the result of a unique, secondary heparin-binding site in the higher-affinity isoform that can bind a heparin site with low affinity for antithrombin III. This interpretation was supported by the chromatographic behavior of the isoforms on mixed-affinity agarose during reverse gradient elution. Two other populations of each of the tryptophan-modified isoforms were identified. Since these isoforms bound tightly to mixed-affinity heparin-agarose but eluted at lower salt concentrations than the corresponding unmodified isoforms, both isoforms may contain additional secondary sites that interact weakly with heparin. A general model of heparin-antithrombin III interaction is proposed in which a high-affinity heparin site initially interacts with a primary site on antithrombin III. The subsequent conformational change leads to a cooperative, entropy-driven association between secondary sites on the protein and low-affinity sites on heparin, stabilizing antithrombin III in its activated form.     

Role of ternary complexes, in which heparin binds both antithrombin and proteinase, in the acceleration of the reactions between antithrombin and thrombin or factor Xa

Oligosaccharides (10-20 monosaccharide units) with high affinity for antithrombin, as well as larger high-affinity heparin fractions (having relative molecular masses between 6,000 and 21,500), all markedly accelerated the inhibition of Factor Xa by antithrombin. Moreover, all high-affinity oligosaccharides and heparins enhanced, to a similar extent, the amount of free proteolytically modified antithrombin cleaved at the reactive bond by Factor Xa. In contrast, a minimum high-affinity heparin size of approximately 18 monosaccharide units was required to significantly accelerate the inactivation of thrombin by antithrombin and to enhance the production of modified antithrombin by this enzyme. All high-affinity fractions studied had similar affinities for antithrombin, as determined by fluorescence titrations. In competition experiments, binary complexes of antithrombin with octadecasaccharide or larger high-affinity heparins, but not with smaller oligosaccharides, displaced inactivated 125I-thrombin from matrix-linked low-affinity heparin. Moreover, similar binary complexes with 3H-labeled octadecasaccharide or larger chains, but not with smaller oligosaccharides, were capable of binding to matrix-linked inactivated thrombin. These results indicate that simultaneous binding of antithrombin and thrombin to high-affinity heparin is a prerequisite to the acceleration of the antithrombin-thrombin reaction and that the minimum heparin sequence capable of binding both proteins comprises approximately 18 monosaccharide units. Similar complex formation apparently is not required for the acceleration of the antithrombin-Factor Xa reaction.     

Interaction of heparin with antithrombin III and platelet factor 4: pH dependence demonstrated with a new rocket precipitin electrophoretic procedure

Only 30% of commercial heparin reacts with antithrombin III (ATIII). This study shows that the interaction is pH dependent: 100% of the heparin binds to ATIII at pH 3.0, 30% at physiological pH. Binding of ATIII, platelet factor 4, and protamine to heparin was studied using a new rocket precipitin electrophoresis procedure, adapted from the Laurell rocket immunoelectrophoresis procedure. Protamine is incorporated into agarose gel, and heparin mixtures with protamine, ATIII, or platelet factor 4 electrophoresed into the gel from a series of wells. The residual free heparin is precipitated by the protamine in a rocket-shaped arc, the height of which is proportional to the amount of free heparin. No antibody is employed. This procedure is useful for quantitation of heparin and for studying the binding of heparin to proteins.     

Effect of heparin modification on its activity in enhancing the inhibition of thrombin by antithrombin III

Studies were conducted to determine the effect of modifying specific functional groups of heparin on its antithrombin III-enhancing activity. The derivatives employed were heparin methyl ester, heparinylglycine and N-desulfated heparin. The carboxyl-modified derivatives increase the rate of inhibition of thrombin by antithrombin III, although not to the same extent as heparin. N-Desulfated heparin is devoid of any activity. Heparin methyl ester is more potent than heparinylglycine in activating antithrombin III, as exhibited by its immediate effect on the thrombin-fibrinogen reaction. However, heparinylglycine is the more effective of the two, in increasing the rate of thrombin deactivation by antithrombin III. The results indicate that although free carboxyl groups of heparin are not crucial for its binding to antithrombin III, they are important for the combination of the latter with thromobin. In contrast, N-sulfates are critical for the interaction of heparin with antithrombin III.     

Immobilization of heparin with trichloro-s-triazine: Purification of mouse antithrombin

A method for covalently attaching heparin to agarose in which the bound heparin is resistant to detachment is described. A controlled amount of 4-aminophenethylamine is linked to trichloro-s-triazine-activated agarose via the primary amino group. The resulting immobilized arylamine is then diazo coupled to an appropriately modified heparin. The immobilized heparin was used in a two-step procedure for the purification of mouse antithrombin. Some of the properties of the mouse antithrombin are described.     

Probing the heparin-binding domain of human antithrombin III with V8 protease

From structural analysis on genetically abnormal and chemically modified human antithrombin III [Koide, T., Odani, S., Takahashi, K., Ono, T. and Sakuragawa, N. (1984) Proc. Natl Acad. Sci. USA 81, 289-293; Chang, J.-Y. and Tran, T. H., (1986) J. Biol. Chem. 261, 1174-1176; Blackburn, M. N., Smith, R. L., Carson, J. and Sibley, C. C. (1984) J. Biol. Chem. 259, 939-941], the heparin-binding site of antithrombin III has been suggested to be in the region of Pro-41, Arg-47 and Trp-49. In this study the heparin-binding site was probed by preferential cleavage of V8 protease on heparin-treated and non-treated native antithrombin III. The study has been based on the presumption that the heparin-binding site of antithrombin III is situated at exposed surface domain and may be preferentially attacked during limited proteolytic digestion. Partially digested antithrombin III samples were monitored by quantitative amino-terminal analysis and amino acid sequencing to identify the preferential cleavage sites. 1-h-digested antithrombin III was separated on HPLC and peptide fragments were isolated and characterized both qualitatively and quantitatively. The results reveal that Glu-Gly (residues 34-35), Glu-Ala (residues 42-43) and Glu-Leu (residues 50-51) are three preferential cleavage sites for V8 protease and their cleavage, especially the Glu-Ala and the Glu-Leu sites, was drastically inhibited when antithrombin III was preincubated with heparin. Both high-affinity and low-affinity antithrombin-III-binding heparins were shown to inhibit the V8 protease digestion of native antithrombin III, but the high-affinity sample exhibited a higher inhibition activity than the low-affinity heparin. These findings (a) imply that the segment containing residues 34-51 is among the most exposed region of native antithrombin III and (b) support the previous conclusions that this region may play a pivotal role in the heparin binding.     

Crystal structure of antithrombin in a heparin-bound intermediate state

Antithrombin is activated as an inhibitor of the coagulation proteases through its specific interaction with a heparin pentasaccharide. The binding of heparin induces a global conformational change in antithrombin which results in the freeing of its reactive center loop for interaction with target proteases and a 1000-fold increase in heparin affinity. The allosteric mechanism by which the properties of antithrombin are altered by its interactions with the specific pentasaccharide sequence of heparin is of great interest to the medical and protein biochemistry communities. Heparin binding has previously been characterized as a two-step, three-state mechanism where, after an initial weak interaction, antithrombin undergoes a conformational change to its high-affinity state. Although the native and heparin-activated states have been determined through protein crystallography, the number and magnitude of conformational changes render problematic the task of determining which account for the improved heparin affinity and how the heparin binding region is linked to the expulsion of the reactive center loop. Here we present the structure of an intermediate pentasaccharide-bound conformation of antithrombin which has undergone all of the conformational changes associated with activation except loop expulsion and helix D elongation. We conclude that the basis of the high-affinity state is not improved interaction with the pentasaccharide but a lowering of the global free energy due to conformational changes elsewhere in antithrombin. We suggest a mechanism in which the role of helix D elongation is to lock antithrombin in the five-stranded fully activated conformation.     

Effect of a heparan sulphate with high affinity for antithrombin III upon inactivation of thrombin and coagulation factor Xa.

The kinetics of inhibition of human alpha-thrombin and coagulation Factor Xa by antithrombin III were examined under pseudo-first-order reaction conditions as a function of the concentration of heparan sulphate with high affinity for antithrombin III. The maximum observed second-order rate constant was, for the antithrombin III-thrombin reaction, 1.2 x 10(9) M-1.min-1 compared with 2.4 x 10(9) M-1.min-1 in the presence of high-affinity heparin. However, the maximum rate was catalysed by much higher concentrations of heparan sulphate (1.3 microM) than of heparin (0.025 microM). Differences were also observed in the maximal acceleration of the antithrombin III-Factor Xa interaction: 1.2 x 10(9) M-1.min-1 at 0.2 microM-heparin sulphate compared with 2.2 x 10(9) M-1.min-1 at 0.04 microM-heparin. The differences in properties of heparan sulphate and heparin were analysed by using the random bi-reactant model of heparin action [Griffith (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 5460-5464]. It was observed that the apparent binding affinity for thrombin was higher for heparan sulphate (180 nM) than for heparin (14 nM). The rate constant for transformation of the antithrombin III-Factor Xa complex into irreversible product differed between heparan sulphate (96 min-1) and heparin (429 min-1). These properties of the high-affinity heparan sulphate may be of importance in consideration of a putative role in the control of intravascular haemostasis.     

Involvement of heparin chain length in the heparin-catalyzed inhibition of thrombin by antithrombin III

The mechanism of the heparin-promoted reaction of thrombin with antithrombin III was investigated by using covalent complexes of antithrombin III with either high-affinity heparin (Mr = 15,000) or heparin fragments having an average of 16 and 12 monosaccharide units (Mr = 4,300 and 3,200). The complexes inhibit thrombin in the manner of active site-directed, irreversible inhibitors: (Formula: see text) That is, the inhibition rate of the enzyme is saturable with respect to concentration of complexes. The values determined for Ki = (k-1 + k2)/k1 are 7 nM, 100 nM, and 6 microM when the Mr of the heparin moieties are 15,000, 4,300, 3,200, respectively, whereas k2 (2 S-1) is independent of the heparin chain length. The bimolecular rate constant k2/Ki for intact heparin is 3 X 10(8) M-1 S-1 and the corresponding second order rate constant k1 is 6.7 X 10(8) M-1 S-1, a value greater than that expected for a diffusion-controlled bimolecular reaction. The bimolecular rate constants for the complexes with heparin of Mr = 4,300 and 3,200 are, respectively, 2 X 10(7) M-1 S-1 and 3 X 10(5) M-1 S-1. Active site-blocked thrombin is an antagonist of covalent antithrombin III-heparin complexes: the effect is monophasic and half-maximum at 4 nM of antagonist against the complex with intact heparin, whereas the effect is weaker against complexes with heparin fragments and not monophasic. We conclude that virtually all of the activity of high affinity, high molecular weight heparin depends on binding both thrombin and antithrombin III to heparin, and that the exceptionally high activity of heparin results in part from the capacity of thrombin bound nonspecifically to heparin to diffuse in the dimension of the heparin chain towards bound antithrombin III. Increasing the chain length of heparin results in an increased reaction rate because of a higher probability of interaction between thrombin and heparin in solution.     

The interaction between heparin and antithrombin III: a comparison of two different heparin-dye conjugates

The interactions of antithrombin III with two heparin-dye conjugates have been compared using their fluorescence anisotropy. The first, heparin labelled with 5-isothiocyanatofluorescein, where the dye was mostly bound to unsulphated glucosamine residues, exhibited binding which was characteristic of heparin with a low affinity for antithrombin III. The second, heparin labelled with a reactive naphthalene dye (DENMT), showed similar binding character. However, when the heparin was treated with an amino group blocking agent prior to labelling with DENMT, the resultant heparin-dye conjugate showed binding behaviour, the strength of which was consistent with heparin molecules having both high and low affinity for antithrombin III. Heparin molecules with a high affinity for antithrombin III did not possess free amino groups. The implications of these findings are discussed with regard to the reliability of the data obtained using heparin-fluorescein conjugates.     

Studies on the mechanism of the rate-enhancing effect of heparin on the thrombin-antithrombin III reaction.

The rate of the reaction between thrombin and antithrombin III is greatly increased in the presence of heparin. Several mechanisms for this effect are possible. To study the problems commercial heparin was fractionated into one fraction of high anticogulant activity and one of low anticoagulant activity by affinity chromatography on matrix-bound antithrombin III. The strength of the binding of the two heparin fractions to antithrombin III and thrombin, respectively, was determined by a crossed immunoelectrophoresis technique. As was to be expected, the high activity fraction was strongly bound to antithrombin III while the low activity fraction was weakly bound. In contrast, thrombin showed equal binding affinity for both heparin fractions. The ability of the two heparin fractions to catalyse the inhibition of thrombin by antithrombin III was determined and was found to be much greater for the high activity heparin fraction. A mechanism for the reaction between thrombin and antithrombin III in the presence of small amounts of heparin is suggested, whereby antithrombin III first binds heparin and this complex then inhibits thrombin by interaction with both the bound heparin and the antithrombin III.     

Antithrombin III Padua: a "new" congenital antithrombin III abnormality with normal or near normal activity, normal antigen, abnormal migration and no thrombotic disease

A "new" antithrombin III abnormality is described in four members of a family. The proposita is a 38 years old female who showed no thrombotic disease and the following laboratory pattern: normal routine clotting tests, normal or near normal AT III activity (chromogenic substrates S-2238 and Chromozym Th) both in plasma and in serum and in the presence or absence of heparin, slightly decreased antifactor Xa activity (chromogenic substrate S-2222), normal progressive antithrombin, normal AT III antigen but abnormal migration in the agarose-heparin bidimensional system. In the latter test, one major abnormal peak, less anodal than the normal counterpart, and a smaller, apparently normal peak, were seen. In agarose without heparin the pattern was similar to normal both in plasma and in serum. Heparin tolerance to heparin in vivo and in vitro was slightly increased but still within normal limits. The two sons and a paternal aunt showed the same pattern. The hereditary pattern seems therefore autosomal dominant. The abnormality described appears different from AT III Budapest. The toponym of antithrombin III Padua is proposed to define this peculiar abnormality.     

Effect of heparin on the glia-derived-nexin-thrombin interaction.

In order to determine the specificity of the interaction between thrombin and glia-derived nexin (GdN), the inactivation of proteolytically modified human thrombin species by GdN has been studied. The second-order rate constants for the inactivation of alpha-, beta T-, gamma T- and epsilon-thrombin by GdN were 1.41, 0.63, 0.33 and 1.91 microM-1.s-1 respectively. The kinetic properties of gdN were also investigated in the presence of different types of heparin, fractionated according to antithrombin III-binding affinity. Association rate constants of both gdN and antithrombin III with alpha-thrombin were obtained using unfractionated, low- and high-affinity heparin types. The different heparin types gave optimal rates of inhibition at similar heparin concentrations for both inhibitors. At optimal heparin concentrations, the rate of inactivation of alpha-thrombin by GdN was 0.5-1.2 nM-1.s-1, which suggests that, under these conditions, the interaction is diffusion-controlled.     

The heparin-catalysed inhibition of human factor XIa by antithrombin III is dependent on the heparin type.

The effect of various well-characterized heparin preparations on the inactivation of human Factor XIa by human antithrombin III was studied. The heparin preparations used were unfractionated heparin and four heparin fractions obtained after anion-exchange chromatography. Inactivation of Factor XIa was monitored with S2366 as chromogenic substrate and followed pseudo-first-order reaction kinetics under all reaction conditions tested. Enhancement of the rate of inhibition of Factor XIa in the presence of unfractionated heparin correlated to the binding of antithrombin III to heparin. From the kinetic data a binding constant of 0.1 microM was inferred. The maximum rate enhancement, achieved at saturating heparin concentrations, was 30-fold. The rate enhancement achieved in the presence of each of the heparin fractions could also be correlated to the binding of antithrombin III to the heparin. The binding constant inferred from the kinetic data varied from 0.10 to 0.28 microM and the number of binding sites for antithrombin III varied from 0.06 to 0.74 site per heparin molecule. The maximum rate enhancements, achieved at saturating heparin concentrations, were strongly dependent on the type of heparin used and varied from 7-fold for fraction A to 41-fold for fraction D. Therefore, although the stimulation of Factor XIa inactivation by antithrombin III could be quantitatively correlated to the binding of antithrombin III to heparin, the heparin-catalysed inhibition of Factor XIa is dependent not only upon the degree of binding of antithrombin III to heparin but also upon the type of heparin to which antithrombin III is bound.     

Endothelial binding sites for heparin. Specificity and role in heparin neutralization.

The specificity of endothelial binding sites for heparin was investigated with heparin fractions and fragments differing in their Mr, charge density and affinity for antithrombin III, as well as with heparinoids and other anionic polyelectrolytes (polystyrene sulphonates). The affinity for endothelial cells was estimated by determining I50 values in competition experiments with 125I-heparin. We found that affinity for endothelial cells increases as a function of Mr and charge density (degree of sulphation). Binding sites are not specific receptors for heparin. Other anionic polyelectrolytes, such as pentosan polysulphates and polystyrene sulphonates, competed with heparin for binding to endothelial cells. Fractions of standard heparin with high affinity for antithrombin III also had greater affinity for endothelium. However, these two properties of heparin (affinity for antithrombin III and affinity for endothelial cells) could be dissociated. Oversulphated heparins and oversulphated low-Mr heparin fragments had lower anticoagulant activity and higher affinity for endothelial cells than did their parent compounds. Synthetic pentasaccharides, bearing the minimal sequence for binding to antithrombin III, did not bind to endothelial cells. Binding to endothelial cells involved partial neutralization of heparin. Bound heparin exhibited only 5% and 7% of antifactor IIa and antifactor Xa specific activity, respectively. In the presence of 200 nM-antithrombin III, and in the absence of free heparin, a limited fraction (approx. 30%) of bound heparin was displaced from endothelial cells during a 1 h incubation period. These data suggested that a fraction of surface-bound heparin could represent a pool of anticoagulant.     

The N-terminal segment of antithrombin acts as a steric gate for the binding of heparin.

The binding of heparin causes a conformational change in antithrombin to give an increased heparin binding affinity and activate the inhibition of thrombin and factor Xa. The areas of antithrombin involved in binding heparin and stabilizing the interaction in the high-affinity form have been partially resolved through the study of both recombinant and natural variants. The role of a section of the N-terminal segment of antithrombin, residues 22-46 (segment 22-46), in heparin binding was investigated using rapid kinetic analysis of the protein cleaved at residues 29-30 by limited proteolysis with thermolysin. The cleaved antithrombin had 5.5-fold lowered affinity for heparin pentasaccharide and 1.8-fold for full-length, high-affinity heparin. It was shown that, although the initial binding of heparin is slightly enhanced by the cleavage, it dissociates much faster from the cleaved form, giving rise to the overall decrease in heparin affinity. This implies that the segment constituting residues 22-46 in the N terminus of antithrombin hinders access to the binding site for heparin, hence the increased initial binding for the cleaved form, whereas, when heparin is bound, segment 22-46 is involved in the stabilization of the binding interaction, as indicated by the increased dissociation constant. When the heparin pentasaccharide is bound to antithrombin prior to incubation with thermolysin, it protects the N-terminal cleavage site, implying that segment 22-46 moves to interact with heparin in the conformational change and thus stabilizes the complex.     

The effect of Ca2+, phospholipid and factor V on the anti-(factor Xa) activity of heparin and its high-affinity oligosaccharides.

The influence of Ca2+, phospholipid and Factor V was determined on the rate of inactivation of Factor Xa by antithrombin III, in the absence and in the presence of unfractionated heparin and of three high-affinity heparin oligosaccharides in the Mr range 1500-6000. In the absence of heparin the addition of Ca2+, phospholipid and Factor V caused a 4-fold decrease in rate of inactivation of Factor Xa. As concentrations of unfractionated heparin were increased the protective effect of Ca2+/phospholipid/Factor V was gradually abolished, and at a concentration of 2.4 nM there were no differences in rates of neutralization of Factor Xa in the presence or absence of Ca2+, phospholipid and Factor V. In contrast, heparin decasaccharide (Mr 3000) and pentasaccharide (Mr 1500) fragments were unable to overcome the protective effect of Ca2+/phospholipid/Factor V; in the presence of these components their catalytic efficiencies were 16-fold and 40-fold less respectively than that of unfractionated heparin. A heparin 20-22-saccharide fragment (Mr approx. 6000) gave similar inactivation rates in the presence and in the absence of Ca2+/phospholipid/Factor V. Human and bovine Factor Xa gave similar results. These results indicate that in the presence of Ca2+/phospholipid/Factor V optimum inhibition of Factor Xa requires a saccharide sequence of heparin additional to that involved in binding to antithrombin III. The use of free enzyme for the assessment of anti-(Factor Xa) activity of low-Mr heparin fractions could give misleading results.     

Mechanism of inhibition of mammalian DNA topoisomerase I by heparin.

We have previously shown that heparin is a potent inhibitor of a mammalian DNA topoisomerase I. We have now investigated the mechanism of its inhibition. This was carried out first by scrutinizing the structural features of heparin molecules responsible for the inhibition. Commercial heparin preparation was fractionated by antithrombin III-Sepharose into non-adsorbed, low-affinity and high-affinity fractions, of which only the high-affinity fraction of heparin is known to contain a specific oligosaccharide sequence responsible for the binding to antithrombin III. These fractions all exhibited essentially similar inhibitory activities. Furthermore, when chemically sulphated to an extent comparable with or higher than heparin, otherwise inactive glycosaminoglycans such as heparan sulphate, chondroitin 4-sulphate, dermatan sulphate and neutral polysaccharides such as dextran and amylose were converted into potent inhibitors. Sulphated dermatan sulphate, one of the model compounds, was further shown to bind competitively to the same sites on the enzyme as heparin. These observations strongly suggested that topoisomerase inhibition by heparin is attributable primarily, if not entirely, to the highly sulphated polyanionic nature of the molecules. In a second series of experiments we examined whether heparin inhibits only one or both of the topoisomerase reactions, i.e. nicking and re-joining. It was demonstrated that both reactions were inhibited by heparin, but the nicking reaction was more severely affected than was the re-joining reaction.     

[Identification and characterization of two phospholipase A2 activities in resident mouse peritoneal macrophages.]

The interaction between bovine antithrombin, a plasma proteinase inhibitor, and heparin species of different molecular weights was studied. A commercial heparin preparation was divided by gel chromatography into a number of fractions with average molecular weights ranging from 6000 to 34700. Each of these fractions was further fractionated by affinity chromatography on matrix-bound antithrombin. In the latter procedure, those heparin fractions that had molecular weights lower than about 14000 were separated into three peaks. The material in the first of these was not adsorbed on the column, and the other two peaks corresponded to the low-affinity and high-affinity peaks described previously. In contrast, high-molecular-weight heparin samples gave only the low-affinity and high-affinity fractions. U.v. difference absorption studies showed that the non-adsorbed heparin fraction bound to antithrombin in solution with a binding constant at physiological ionic strength only slightly lower than that of low-affinity heparin. The division between the two fractions thus is arbitrary and only dependent on the conditions selected for the affinity-chromatography experiment. Stoicheiometries and binding constants for the binding of several high-affinity heparin species to antithrombin were determined by fluorescence titrations. High-affinity heparin fractions of equal elution positions in the beginning of the peaks of the affinity chromatographies, but with different molecular weights, showed stoicheiometries that were not experimentally distinguishable from 1:1 and also had no appreciable differences in binding constants. However, the anticoagulant activities, calculated on a molar basis, of these fractions increased markedly with molecular weight, a behaviour that thus cannot be explained by differences in the binding of the fractions to antithrombin. In contrast, high-affinity samples of similar molecular weights, which were eluted at increasing ionic strengths from matrix-linked antithrombin, were found to have an increasing proportion of chains with two binding sites for antithrombin and also to have progressively higher binding constants. These binding properties at least partly explain the increasing anticoagulant activities that were observed for these fractions.