|Home | About | Journals | Submit | Contact Us | Français|
Clotting blood contains fibrin-bound thrombin, which is a major source of procoagulant activity leading to clot extension and further activation of coagulation. When bound to fibrin, thrombin is protected from inhibition by antithrombin (AT) + heparin but is neutralized when AT and heparin are covalently linked (ATH). Here, we report the surprising observation that, rather than yielding an inert complex, thrombin-ATH formation converts clots into anticoagulant surfaces that effectively catalyze inhibition of thrombin in the surrounding environment.
Thrombosis results from stimulation of the coagulation pathway, which leads to thrombin cleavage of fibrinogen to form fibrin. During this process, thrombin becomes incorporated into the evolving fibrin clot and becomes a nidus for activation of coagulant factors and fibrin accretion . Historically, thrombosis prevention has been through administration of heparin, which catalyzes the inhibition of thrombin by antithrombin (AT) . Paradoxically, use of heparin can actually lead to protection of fibrin-bound thrombin from reaction with noncovalent AT•heparin by forming a fibrin•heparin•thrombin ternary complex . Increased heparin concentrations reduce circulating prothrombotic activity but clot-associated thrombin activity reappears once heparin treatment is discontinued [3, 4]. We developed a potent nondissociable covalent complex of AT and heparin (ATH)  so that attraction of the heparin moiety to fibrin•thrombin would consequently neutralize thrombin via the attached AT . Investigation of ATH interactions with fibrin monomer and thrombin revealed that the covalent thrombin-ATH complex formed remains adherent to fibrin through the heparin chain . We were intrigued to determine if the heparin in fibrin-bound thrombin-ATH may still be able to interact with exogenous fluid-phase coagulation proteins. Thus, the present work was designed to determine if the heparin chain in the thrombin-ATH complexes bound to fibrin monomer and fibrin clot surfaces is capable of catalyzing the inhibition of free, fluid phase, thrombin by exogenously added AT.
Soluble fibrin monomer was prepared from human fibrinogen (Enzyme Research Laboratories), after removal of contaminating fibronectin by incubation of 15mL of 130μM fibrinogen with 5mL of gelatin agarose (Sigma) for 30minutes. Purified fibrinogen (100μM) was incubated with human thrombin (Enzyme Research, 2nM final) at 37°C for 4hours, followed by centrifugation for 5minutes at 2000g. The fibrin polymer pellet was placed in dialysis tubing with a 12000–14000MW cutoff and dialyzed at 4°C versus H2O, followed by dialysis against 0.02M acetic acid until the fibrin dissolved. The fibrin monomer concentration was determined by absorbance at 280nm, where a concentration of 10mg/mL = an absorbance of 14.0.
Covalent antithrombin-heparin complex (ATH) was prepared by heating 1152mg of antithrombin (AT; Affinity Biologicals Inc.) + 64g of heparin (Sigma) in 900mL of PBS for 14days at 40°C. Following incubation, NaBH3CN was added to a final concentration of 0.05M and the solution was further incubated at 37°C for 5hours. Free AT was removed by making the reaction mixture 2.5M in (NH4)2SO4, loading on a 1000mL column of butyl sepharose (Amersham), washing the beads with 2.5M (NH4)2SO4 and eluting the complex in 0.02 M phosphate pH 7.0. Eluent was dialyzed versus 0.01M Tris-HCl pH 8.0 buffer and bound to 900mL of DEAE sepharose (Amersham), followed by washing with 0.2M NaCl in buffer and elution of purified ATH with 2.0M NaCl in buffer. ATH was pressure-dialyzed versus PBS. The final purified ATH product possessed <5% unconjugated starting materials and contained AT which was >95% active against thrombin. Previous work has established that the AT : heparin molar ratio is essentially 1 : 1 .
High-affinity heparin (HAH) was prepared by loading 4mg of heparin (in 2mL of 0.15M NaCl in 0.01M phosphate pH 7.3 buffer) on a 10mL column of sepharose-AT prepared from AT and CNBr-activated sepharose (Amersham), according to the manufacturer. The column was then washed with 0.5 M NaCl in buffer and eluted with 2 M NaCl in buffer. Recovered HAH was dialyzed against H2O and lyophilized.
In order, 3.24μL of 37μM fibrin monomer, 2.54μL of 59μM GPRP-NH2 (Sigma), 1.3μL of 1M Tris, 4.6μL 0.02 M Tris-HCl 0.15 M NaCl 0.6% polyethylene glycol 8000 pH 7.4 (TSP), 2.06μL of 1mg/mL bovine albumin (globulin free; Sigma) in TSP, and 1.2μL of 1.08μM thrombin were mixed at 23°C in a microfuge tube. After 2minutes, 2.06μL of either 0.125μM AT + 0.125μM heparin or 0.125μM ATH was added. Following a further 2-minute incubation, TSP solutions of excess AT (12.9μM, 1μL) and thrombin (1.08μM, 12μL) were added in rapid succession. After 2minutes of reaction, remaining thrombin activity was determined by mixing 0.8μL of reaction mixture with 79.2μL of 0.62mM S-2238 substrate (diaPharma, West Chester, OH, USA) in TSP, followed by neutralization after 10minutes with 20μL of 50% acetic acid and measurement of absorbance at 405nm. The ability to catalyze the inhibition of thrombin by AT was calculated as the percent decrease in thrombin activity relative to experiments without AT + heparin or ATH.
A volume of 0.5mL of plasma was mixed with 9μL of 1M CaCl2 in microfuge tubes containing 6mm plastic loops (Fisher). After 1hour at 37°C, clots (on loops) were removed and washed 6 times by dipping in 1mL solutions of 0.02M Tris-HCl 0.15M NaCl pH 7.4 (TBS). After storage overnight at 4°C in TBS, clots were placed in TBS for 20minutes at 23°C and 1hour at 37°C, followed by 6 TBS washes and storage overnight at 4°C in TBS. Clots were placed in fresh TBS for 20minutes and then incubated for 1hour at 37°C in 1mL of either TSP, 0.1μM AT + 0.1μM heparin; TSP, 0.1μM AT + 0.1μM HAH; or TSP, 0.1μM ATH. After washing briefly 6 times in 1mL of TSP and once for 20minutes in TSP at 23°C, catalytic activity of the clots was determined at 37°C in a solution of 0.5mL 0.043μM thrombin + 0.5mL 0.2μM AT, mixed just prior to use. Clots were dipped up and down (once per second) within the thrombin + AT solution for 1 minute. After removal of the clot, 12μL of the thrombin + AT solution was reacted with 67.2μL of 0.71mM S-2238 in TSPfor 10minutes at 23°C, neutralized with 20μL of 50% acetic acid and the absorbance was taken at 405nm to determine activity. Results were calculated as a percent of thrombin activity in thrombin solutions without AT. Inhibition was determined relative to reactions without clots. Subtraction of percent thrombin activity lost in clots incubated without heparin compounds gave the net catalytic activity observed.
Data were judged to be normally distributed since variance about the mean was not skewed and application of the Anderson-Darling normality test did not reveal any nonGaussian distributions. t-tests (2 tailed, nonpaired) were used for the fibrin monomer studies (n ≥6), as there were two main experimental groups. Since there were multiple groups in the fibrin clot experiments, a one-way ANOVA was performed (n ≥4, P = .004), followed by the Tukey posttest for comparisons between ATH and the other heparin experimental groups. An alpha level of 0.05 was used throughout the analyses. Data are expressed as mean ± standard error of the mean (SEM).
Initial experiments were conducted to ensure that thrombin on fibrin was neutralized by ATH and that the inhibited surface did not protect additional free thrombin from attack by fluid-phase AT. Results from chromogenic assays showed that fibrin•thrombin-ATH had no effect on soluble thrombin functional activity. While all thrombin on fibrin was capable of reaction with ATH, a rapid inactivation of supplemental thrombin ensued in the presence of exogenous AT. Careful studies confirmed that, compared to fibrin alone, thrombin-ATH on fibrin significantly catalyzed inhibition of excess thrombin by AT (P = .007) while similar tests with noncovalent AT + heparin mixtures demonstrated no effect (Figure 1).
We next examined the capacity for surface anticoagulation of ATH in a physiological system. Clots from recalcified plasma were treated with buffer or small amounts of ATH. Incubation of thrombin + AT solutions with ATH-neutralized clots showed strong inhibition of enzyme activity relative to AT + heparin or AT + heparin with high affinity for AT (HAH), with P values of 0.026 and 0.023, respectively (Figure 2). Moreover, insignificant effects on soluble thrombin + AT were observed for clots treated with either AT + heparin or AT + HAH (95% confidence intervals of −17.2–19.1 and −2.2–10.2, resp.). Thus, we have demonstrated that not only does ATH inactivate clot prothrombotic activity, but the surface produced on the clot also contains heparin which maintains anticoagulant activity in the surrounding milieu. This finding is in contrast to that with heparin, where heparin absorbed on procoagulant clots was ineffective at influencing thrombin/AT reactions (Figure 2). Furthermore, the lack of enhancement by HAH of thrombin reaction with AT (Figure 2) confirmed that enrichment of high-affinity AT binding sites in the heparin chains of clot-bound ATH  was not responsible for ATH's superior catalytic properties.
A model (Figure 3) shows the thrombin-ATH complex, noncovalently tethered through heparin to fibrin, which is catalyzing reaction of incoming thrombin + AT via conventional heparin template bridging . This mechanism is consistent with data showing that ATH causes reduction in clot size within injured vessels, compared to clot growth which occurs with AT + heparin . Our findings have implications that could alter treatment paradigms. By “coating” fibrin-thrombi with ATH, clinicians may use clots as platforms to pacify hyperthrombotic activity in damaged vasculature.
Covalent linkage of AT to heparin facilitates formation of fibrin clot-bound structures with heparin moieties possessing potent thrombin-inhibitory catalytic activity. Conversion of clot surfaces into functional anticoagulants may be done selectively since ATH heparin chains are electrostatically attracted to fibrin•thrombin . In vivo experiments will confirm the utility of anticoagulated thrombi in treatment models.
We gratefully acknowledge the work of Nethnapha Paredes in preparation of this manuscript. This work is supported by a grant-in-aid from the Heart and Stroke Foundation of Ontario (Grant no. T5414). Anthony Chan is a career investigator of the Heart and Stroke Foundation of Canada.