Studies on the different modes of action of the anticoagulant protease inhibitors DX-9065a and Argatroban. I. Effects on thrombin generation

The present study began with mathematical modeling of how inhibitors of both factor Xa (fXa) and thrombin affect extrinsic pathway-triggered blood coagulation. Numerical simulation demonstrated a stronger inhibition of thrombin generation by a thrombin inhibitor than a fXa inhibitor, but both prolonged clot time to a similar extent when they were given an equal dissociation constant (30 nm) for interaction with their respective target enzymes. These differences were then tested by comparison with the real inhibitors DX-9065a and argatroban, specific competitive inhibitors of fXa and thrombin, respectively, with similar K(i) values. Comparisons were made in extrinsically triggered human citrated plasma, for which endogenous thrombin potential and clot formation were simultaneously measured with a Wallac multilabel counter equipped with both fluorometric and photometric detectors and a fluorogenic reporter substrate. The results demonstrated stronger inhibition of endogenous thrombin potential by argatroban than by DX-9065a, especially when coagulation was initiated at higher tissue factor concentrations, while argatroban appeared to be slightly less potent in its ability to prolong clot time. This study demonstrates differential inhibition of thrombin generation by fXa and thrombin inhibitors and has implications for the pharmacological regulation of blood coagulation by the anticoagulant protease inhibitors.     

Thrombin inhibition by serpins disrupts exosite II

Thrombin uses three principal sites, the active site, exosite I, and exosite II, for recognition of its many cofactors and substrates. It is synthesized in the zymogen form, prothrombin, and its activation at the end of the blood coagulation cascade results in the formation of the active site and exosite I and the exposure of exosite II. The physiological inhibitors of thrombin are all serpins, whose mechanism involves significant conformational change in both serpin and protease. It has been shown that the formation of the thrombin-serpin final complex disorders the active site and exosite I of thrombin, but exosite II is thought to remain functional. It has also been hypothesized that thrombin contains a receptor-binding site that is exposed upon final complex formation. The position of this cryptic site may depend on the regions of thrombin unfolded by serpin complexation. Here we investigate the conformation of thrombin in its final complex with serpins and find that in addition to exosite I, exosite II is also disordered, as reflected by a loss of affinity for the γ'-peptide of fibrinogen and for heparin and by susceptibility to limited proteolysis. This disordering of exosite II occurs for all tested natural thrombin-inhibiting serpins. Our data suggest a novel framework for understanding serpin function, especially with respect to thrombin inhibition, where serpins functionally "rezymogenize" proteases to ensure complete loss of activity and cofactor binding.     

Synthesis,structure, and structure-activity relationships of divalent thrombin inhibitors containing an α-keto-amide transition-state mimetic

A new class of divalent thrombin inhibitors is described that contains an α-keto-amide transition-state mimetic linking an active site binding group and a group that binds to the fibrinogen-binding exosite. The X-ray crystallographic structure of the most potent member of this new class, CVS995, shows many features in common with other divalent thrombin inhibitors and clearly defines the transition-state-like binding of the α-keto-amide group. The structure of the active site part of the inhibitor shows a network of water molecules connecting both the side-chain and backbone atoms of thrombin and the inhibitor. Direct peptide analogues of the new transition-state-containing divalent thrombin inhibitors were compared using in vitro assays of thrombin inhibition. There was no direct correlation between the binding constants of the peptides and their α-keto-amide counterparts. The most potent cv-keto-amide inhibitor, CVS995, with a K_i = 1 pM, did not correspond to the most potent divalent peptide and contained a single amino acid deletion in the exosite binding region with respect to the equivalent region of the natural thrombin inhibitor hirudin. The interaction energies of the active site, transition state, and exosite binding regions of these new divalent thrombin inhibitors are not additive.     

Site-specific interaction of thrombin and inhibitors observed by fluorescence correlation spectroscopy.

We report on the application of fluorescence correlation spectroscopy (FCS) to observe the interaction between thrombin and thrombin inhibitors. Two site-specific fluorescent labels were used to distinguish between inhibitors directed to the active site, the exosite, or both binding sites of thrombin. For several well-known inhibitors of thrombin, the binding sites observed by FCS correspond to previous studies. The interaction of the recently discovered thrombin inhibitor ornithodorin from the tick Ornithodorus moubata with thrombin was investigated. It was found that this inhibitor, like hirudin and rhodniin, binds to both the active site and exosite of thrombin simultaneously. This study shows the feasibility of FCS as a sensitive and selective method for observing protein-ligand interactions. As an additional technique, simultaneous labeling with both fluorescent labels was successfully demonstrated.     

Synergistic effect of aptamers that inhibit exosites 1 and 2 on thrombin

Thrombin is a multifunctional protease that plays a key role in hemostasis, thrombosis, and inflammation. Most thrombin inhibitors currently used as antithrombotic agents target thrombin''s active site and inhibit all of its myriad of activities. Exosites 1 and 2 are distinct regions on the surface of thrombin that provide specificity to its proteolytic activity by mediating binding to substrates, receptors, and cofactors. Exosite 1 mediates binding and cleavage of fibrinogen, proteolytically activated receptors, and some coagulation factors, while exosite 2 mediates binding to heparin and to platelet receptor GPIb-IX-V. The crystal structures of two nucleic acid ligands bound to thrombin have been solved. Previously Padmanabhan and colleagues solved the structure of a DNA aptamer bound to exosite 1 and we reported the structure of an RNA aptamer bound to exosite 2 on thrombin. Based upon these structural studies we speculated that the two aptamers would not compete for binding to thrombin. We observe that simultaneously blocking both exosites with the aptamers leads to synergistic inhibition of thrombin-dependent platelet activation and procoagulant activity. This combination of exosite 1 and exosite 2 inhibitors may provide a particularly effective antithrombotic approach.     

Studies on the different modes of action of the anticoagulant protease inhibitors DX-9065a and Argatroban. II. Effects on fibrinolysis

The accompanying paper (Nagashima, H. (2002) J. Biol. Chem. 277, 50439-50444) has demonstrated that argatroban can yield a stronger inhibitory effect on thrombin generation than DX-9065a during extrinsic pathway-stimulated human plasma coagulation, while these anticoagulant compounds have comparable abilities to prolong clot time. Since thrombin generation is known to be an important determinant for fibrinolytic resistance of clots formed during coagulation, the two compounds are compared by tissue plasminogen activator-induced clot lysis assays. The results demonstrated that, in the presence of thrombomodulin, argatroban dose dependently accelerated fibrinolysis of the clots, whereas DX-9065a did not. The activation of thrombin activatable fibrinolysis inhibitor (TAFI) determined in separate assays reflected the differential influence on thrombin generation by these compounds. Moreover, TAFI activation correlated closely with the fibrinolytic resistance observed during tissue plasminogen activator-induced clot lysis. This study demonstrates the differential effects of DX-9065a and argatroban on thrombin generation, which in turn results in a differential acceleration of fibrinolysis as well as TAFI activation in the clots formed under the influence of these compounds. The data implicate a possible difference in the antifibrinolytic properties of clots formed during treatment with these compounds.     

Thermodynamic compensation upon binding to exosite 1 and the active site of thrombin

Several lines of experimental evidence including amide exchange and NMR suggest that ligands binding to thrombin cause reduced backbone dynamics. Binding of the covalent inhibitor dPhe-Pro-Arg chloromethyl ketone to the active site serine, as well as noncovalent binding of a fragment of the regulatory protein, thrombomodulin, to exosite 1 on the back side of the thrombin molecule both cause reduced dynamics. However, the reduced dynamics do not appear to be accompanied by significant conformational changes. In addition, binding of ligands to the active site does not change the affinity of thrombomodulin fragments binding to exosite 1; however, the thermodynamic coupling between exosite 1 and the active site has not been fully explored. We present isothermal titration calorimetry experiments that probe changes in enthalpy and entropy upon formation of binary ligand complexes. The approach relies on stringent thrombin preparation methods and on the use of dansyl-l-arginine-(3-methyl-1,5-pantanediyl)amide and a DNA aptamer as ligands with ideal thermodynamic signatures for binding to the active site and to exosite 1. Using this approach, the binding thermodynamic signatures of each ligand alone as well as the binding signatures of each ligand when the other binding site was occupied were measured. Different exosite 1 ligands with widely varied thermodynamic signatures cause a similar reduction in ΔH and a concomitantly lower entropy cost upon DAPA binding at the active site. The results suggest a general phenomenon of enthalpy-entropy compensation consistent with reduction of dynamics/increased folding of thrombin upon ligand binding to either the active site or exosite 1.     

Exosites 1 and 2 are essential for protection of fibrin-bound thrombin from heparin-catalyzed inhibition by antithrombin and heparin cofactor II

Assembly of ternary thrombin-heparin-fibrin complexes, formed when fibrin binds to exosite 1 on thrombin and fibrin-bound heparin binds to exosite 2, produces a 58- and 247-fold reduction in the heparin-catalyzed rate of thrombin inhibition by antithrombin and heparin cofactor II, respectively. The greater reduction for heparin cofactor II reflects its requirement for access to exosite 1 during the inhibitory process. Protection from inhibition by antithrombin and heparin cofactor II requires ligation of both exosites 1 and 2 because minimal protection is seen when exosite 1 variants (gamma-thrombin and thrombin Quick 1) or an exosite 2 variant (Arg93 --> Ala, Arg97 --> Ala, and Arg101 --> Ala thrombin) is substituted for thrombin. Likewise, the rate of thrombin inhibition by the heparin-independent inhibitor, alpha1-antitrypsin Met358 --> Arg, is decreased less than 2-fold in the presence of soluble fibrin and heparin. In contrast, thrombin is protected from inhibition by a covalent antithrombin-heparin complex, suggesting that access of heparin to exosite 2 of thrombin is hampered when ternary complex formation occurs. These results reveal the importance of exosites 1 and 2 of thrombin in assembly of the ternary complex and the subsequent protection of thrombin from inhibition by heparin-catalyzed inhibitors.     

Evaluation of Antithrombotic Activity of Thrombin DNA Aptamers by a Murine Thrombosis Model

Aptamers are nucleic acid based molecular recognition elements with a high potential for the theranostics. Some of the aptamers are under development for therapeutic applications as promising antithrombotic agents; and G-quadruplex DNA aptamers, which directly inhibit the thrombin activity, are among them. RA-36, the 31-meric DNA aptamer, consists of two thrombin binding pharmacophores joined with the thymine linker. It has been shown earlier that RA-36 directly inhibits thrombin in the reaction of fibrinogen hydrolysis, and also it inhibits plasma and blood coagulation. Studies of both inhibitory and anticoagulation effects had indicated rather high species specificity of the aptamer. Further R&D of RA-36 requires exploring its efficiency in vivo. Therefore the development of a robust and adequate animal model for effective physiological studies of aptamers is in high current demand. This work is devoted to in vivo study of the antithrombotic effect of RA-36 aptamer. A murine model of thrombosis has been applied to reveal a lag and even prevention of thrombus formation when RA-36 was intravenous bolus injected in high doses of 1.4–7.1 µmol/kg (14–70 mg/kg). A comparative study of RA-36 aptamer and bivalirudin reveals that both direct thrombin inhibitors have similar antithrombotic effects for the murine model of thrombosis; though in vitro bivalirudin has anticoagulation activity several times higher compared to RA-36. The results indicate that both RA-36 aptamer and bivalirudin are direct thrombin inhibitors of different potency, but possible interactions of the thrombin-inhibitor complex with other components of blood coagulation cascade level the physiological effects for both inhibitors.     

The role of thrombin exosites I and II in the activation of human coagulation factor V

Human blood coagulation Factor V (FV) is a plasma protein with little procoagulant activity. Limited proteolysis at Arg(709), Arg(1018), and Arg(1545) by thrombin or Factor Xa (FXa) results in the generation of activated FV, which serves as a cofactor of FXa in prothrombin activation. Both thrombin exosites I and II have been reported to be involved in FV activation, but the relative importance of these regions in the individual cleavages remains unclear. To investigate the role of each exosite in FV activation, we have used recombinant FV molecules with only one of the three activation cleavage sites available, in combination with exosite I- or II-specific aptamers. In addition, structural requirements for exosite interactions located in the B-domain of FV were probed using FV B-domain deletion mutants and comparison with FV activating enzymes from the venom of Russell's viper (RVV-V) and of Levant's viper (LVV-V) known to activate FV by specific cleavage at Arg(1545). Our results indicate that thrombin exosite II is not involved in cleavage at Arg(709) and that both thrombin exosites are important for recognition and cleavage at Arg(1545). Efficient thrombin-catalyzed FV activation requires both the N- and C-terminal regions of the B-domain, whereas only the latter is required by RVV-V and LVV-V. This indicates that proteolysis of FV by thrombin at Arg(709), Arg(1018), and Arg(1545) show different cleavage requirements with respect to interactions mediated by thrombin exosites and areas that surround the respective cleavage sites. In addition, interactions between exosite I of thrombin and FV are primarily responsible for the different cleavage site specificity as compared with activation by RVV-V or LVV-V.     

The Tick-Derived Anticoagulant Madanin Is Processed by Thrombin and Factor Xa

The cysteine-less peptidic anticoagulants madanin-1 and madanin-2 from the bush tick Haemaphysalis longicornis are the founding members of the MEROPS inhibitor family I53. It has been previously suggested that madanins exert their functional activity by competing with physiological substrates for binding to the positively charged exosite I (fibrinogen-binding exosite) of α-thrombin. We hereby demonstrate that competitive inhibition of α-thrombin by madanin-1 or madanin-2 involves binding to the enzyme''s active site. Moreover, the blood coagulation factors IIa and Xa are shown to hydrolyze both inhibitors at different, although partially overlapping cleavage sites. Finally, the three-dimensional structure of the complex formed between human α-thrombin and a proteolytic fragment of madanin-1, determined by X-ray crystallography, elucidates the molecular details of madanin-1 recognition and processing by the proteinase. Taken together, the current findings establish the mechanism of action of madanins, natural anticoagulants that behave as cleavable competitive inhibitors of thrombin.     

Multiple pathways of thrombin-induced platelet activation differentiated by desensitization and a thrombin exosite inhibitor.

Recently a thrombin receptor with a unique mechanism of activation was cloned from a megakaryocyte-like cell line (Vu et al., Cell 64:1057-1068, 1991). Thrombin cleaves a portion of this receptor creating a new N-terminus that acts as a "tethered-ligand" to activate the receptor. A thrombin receptor activating peptide (SFLLRNPNDKYEPF) homologous to the new N-terminus was shown to activate platelets. We synthesized this peptide and demonstrated that it desensitized platelets to activation by low concentrations of alpha-thrombin but not gamma-thrombin. We also synthesized a thrombin exosite inhibitor (BMS 180742) that inhibited platelet aggregation induced by low, but not high, concentrations of alpha-thrombin. In contrast, a thrombin active site inhibitor, N alpha-(2-naphthylsulfonyl-glycyl)-D,L-amidinophenylalanylpiperi dide, competitively inhibited thrombin-induced platelet aggregation. We conclude that thrombin-induced platelet activation is mediated by at least two pathways: one activated by low concentrations of alpha-thrombin and blocked by a thrombin exosite inhibitor that appears to be coupled to the "tethered-ligand" thrombin receptor, and another that is stimulated by higher concentrations of alpha-thrombin and by gamma-thrombin and does not require the thrombin exosite for activation. Both pathways are blocked by a thrombin active site inhibitor.     

The reaction of thrombin with platelet-derived nexin requires a secondary recognition site in addition to the catalytic site

A protease nexin released by activated platelets forms stable complexes with alpha-thrombin. Active-site-blocked thrombin does not form the stable complex, but it inhibits formation of the stable complex by active alpha-thrombin. gamma-Thrombin, which has a damaged substrate recognition site (the anion-binding exosite), did not form the complex and did not inhibit formation of the stable complex by alpha-thrombin. Complex formation was inhibited by the C-terminal dodecapeptide of hirudin, which has been shown to bind to the anion-binding exosite. A monoclonal antibody that blocks reactions of thrombin that involve the anion-binding exosite also inhibited formation of a stable complex of alpha-thrombin and the platelet-derived protease nexin. It is concluded that the anion-binding exosite of thrombin, a site that confers a high degree of specificity for substrates with a complementary site, binds to the platelet nexin prior to reaction of the catalytic site with the serpin.     

Thrombin causes neurite retraction in neuronal cells through activation of cell surface receptors.

The mechanism by which thrombin induces neurite retraction was studied in NB2a mouse neuroblastoma cells. The rapid effect of thrombin (completed within minutes) appears to involve an interaction between its anion-binding exosite and the thrombin receptor. Structural alterations of this site increase the EC50 for thrombin-mediated retraction, and a hirudin C-terminal peptide that blocks this site inhibits the response. The thrombin effect was mimicked by a 14 amino acid peptide starting with Ser-42, at the proposed cleavage site of the human thrombin receptor. The protein kinase inhibitors staurosporine and H-7 blocked thrombin-induced retraction. It is therefore proposed that thrombin-mediated neurite retraction is caused by cleavage-induced activation of the thrombin receptor and involves stimulation of a protein kinase(s).     

Inhibitors of factor VIIa affect the interface between the protease domain and tissue factor

Blood coagulation is triggered by the formation of a complex between factor VIIa (FVIIa) and its cofactor, tissue factor (TF). The gamma-carboxyglutamic acid-rich domain of FVIIa docks with the C-terminal domain of TF, the EGF1 domain of FVIIa contacts both domains of TF, and the EGF2 domain and protease domain (PD) form a continuous surface that sits on the N-terminal domain of TF. Our aim was to investigate the conformational changes that occur in the sTF.PD binding region when different types of inhibitors, i.e., one active-site inhibitor (FFR-chloromethyl ketone (FFR)), two different peptide exosite inhibitors (E-76 and A-183), and the natural inhibitor tissue factor pathway inhibitor (TFPI), were allowed to bind to FVIIa. For this purpose, we constructed two sTF mutants (Q37C and E91C). By the aid of site-directed labeling technique, a fluorescent label was attached to the free cysteine. The sTF.PD interface was affected in position 37 by the binding of FFR, TFPI, and E-76, i.e., a more compact structure was sensed by the probe, while for position 91 located in the same region no change in the surrounding structure was observed. Thus, the active site inhibitors FFR and TFPI, and the exosite inhibitor E-76 have similar effects on the probe in position 37 of sTF, despite their differences in size and inhibition mechanism. The allosteric changes at the active site caused by binding of the exosite inhibitor E-76 in turn induce similar conformational changes in the sTF.PD interface as does the binding of the active site inhibitors. A-183, on the other hand, did not affect position 37 in sTF, indicating that the A-183 inhibition mechanism is different from that of E-76.     

Boophilus microplus: its saliva contains microphilin, a small thrombin inhibitor

Saliva of the cattle tick Boophilus microplus contains two thrombin inhibitors, BmAP and microphilin. This work presents the purification and characterization of microphilin. It was purified from the saliva by gel filtration, ultrafiltration through a 3 kDa cut-off membrane and affinity chromatography in a thrombin-Sepharose column. Analysis by mass spectrometry showed a molecular mass of 1770 Da. Microphilin is the smallest salivary thrombin inhibitor peptide known to date. It inhibits fibrinocoagulation and thrombin-induced platelet aggregation with an IC(50) of 5.5 microM, is temperature resistant and its inhibitory activity was abolished by protease K treatment. Microphilin did not inhibit the amidolytic activity of the enzyme upon a small chromogenic substrate, but inhibited the hydrolysis of a substrate that binds both catalytic site and exosite I. Therefore, we propose that microphilin blocks thrombin at exosite I.     

Role of regulatory exosite I in binding of thrombin to human factor V, factor Va, factor Va subunits, and activation fragments.

The blood coagulation proteinase, thrombin, converts factor V into factor Va through a multistep activation pathway that is regulated by interactions with thrombin exosites. Thrombin exosite interactions with human factor V and its activation products were quantitatively characterized in equilibrium binding studies based on fluorescence changes of thrombin covalently labeled with 2-anilinonaphthalene-6-sulfonic acid (ANS) linked to the catalytic site histidine residue by Nalpha-[(acetylthio)acetyl]-D-Phe-Pro-Arg-CH2Cl ([ANS]FPR-thrombin). Exosite I was shown to play a predominant role in the binding of factor V and factor Va from the effect of the exosite I-specific ligand, hirudin54-65, on the interactions. Factor V and factor Va bound to exosite I of [ANS]FPR-thrombin with similar dissociation constants of 3.4 +/- 1.3 and 1.1 +/- 0.4 microM and fluorescence enhancements of 182 +/- 41 and 127 +/- 17%, respectively. Native thrombin and labeled thrombin bound with similar affinity to factor Va. Among factor V activation products, the factor Va heavy chain was shown to contain the site of exosite I binding, whereas exosite I-independent, lower affinity interactions were observed for activation fragments E and C1, and no detectable binding was observed for the factor Va light chain. The results support the conclusion that the factor V activation pathway is initiated by exosite I-mediated binding of thrombin to a site in the heavy chain region of factor V that facilitates the initial cleavage at Arg709 to generate the heavy chain of factor Va. The results further suggest that binding of thrombin through exosite I to factor V activation intermediates may regulate their conversion to factor Va and that similar binding of thrombin to the factor Va produced may reflect a mode of interaction involved in the regulation of prothrombin activation.     

Design, synthesis, and biological activity of novel factor Xa inhibitors: 4-aryloxy substituents of 2,6-diphenoxypyridines

A novel series of triaryloxypyridines have been designed to inhibit factor Xa, a serine protease strategically located in the coagulation cascade. Inhibitor 5e has a K(I) against factor Xa of 0.12nM and is greater than 8000- and 2000-fold selective over two related serine proteases, thrombin and trypsin, respectively. The 4-position of the central pyridine has been identified as a site that tolerates various substitutions without deleterious effects on potency and selectivity. This suggests that the 4-position of the pyridine ring is an ideal site for chemical modifications to identify inhibitors with improved pharmacokinetic characteristics. This investigation has resulted in inhibitor 5d, which has an oral availability of 6% in dogs. The synthesis, in vitro activity, and in vivo profile of this class of inhibitors is outlined.     

A new blood coagulation inhibitor from the snake Bothrops jararaca plasma: isolation and characterization

A novel thrombin inhibitor, Bothrops jararaca inhibitor (BjI), has been identified and purified from B. jararaca snake blood by two anionic chromatographic steps. Purified BjI showed two polypeptide chains with molecular masses of 109 and 138 kDa, by SDS-PAGE in reducing conditions. On the other hand, in nonreducing conditions the molecular masses were 150 and 219 kDa, suggesting that the polypeptide chain 109 kDa can be a dimer form linked by disulfide bond. However, the native BjI shows a molecular mass higher than 1000 kDa by gel filtration chromatography, indicating the need of a quaternary structure formation for the blood coagulation inhibition. BjI is a specific thrombin coagulant activity inhibitor that does not affect other thrombin functions, such as: amidolytic and platelet aggregation activities. BjI is not an antithrombin-like inhibitor. Fibrinogen and heparin competition ELISA assays with BjI and thrombin showed that fibrinogen does not interfere in the BjI and thrombin binding, however, heparin interferes in BjI and thrombin interaction, suggesting that BjI binds to heparin site or other sites close to it. Our findings indicate that BjI is an exosite binding thrombin inhibitor, specific upon coagulant activity thrombin inhibitor, without any anti-platelet aggregation activity.     

Biflavones from Ginkgo biloba as inhibitors of human thrombin

Ginkgo Biloba leaf extract has been widely used for the prevention and treatment of thrombosis and cardiovascular disease in both eastern and western countries, but the bioactive constituents and the underlying mechanism of anti-thrombosis have not been fully characterized. The purpose of this study was to investigate the inhibitory effects of major constituents in Ginkgo biloba on human thrombin, a key serine protease regulating the blood coagulation cascade and the processes of thrombosis. To this end, a fluorescence-based biochemical assay was used to assay the inhibitory effects of sixteen major constituents from Ginkgo biloba on human thrombin. Among all tested natural compounds, four biflavones (ginkgetin, isoginkgetin, bilobetin and amentoflavone), and five flavonoids (luteolin, apigenin, quercetin, kaempferol and isorhamnetin) were found with thrombin inhibition activity, with the IC₅₀ values ranging from 8.05 μM to 82.08 μM. Inhibition kinetic analyses demonstrated that four biflavones were mixed inhibitors against thrombin-mediated Z-GGRAMC acetate hydrolysis, with the K_i values ranging from 4.12 μM to 11.01 μM. Molecular docking method showed that the four biflavones could occupy the active cavity with strong interactions of salt bridges and hydrogen bonds. In addition, mass spectrometry-based lysine labeling reactivity assay suggested that the biflavones could bind on human thrombin at exosite I rather than exosite II. All these findings suggested that the biflavones in Ginkgo biloba were naturally occurring inhibitors of human thrombin, and these compounds could be used as lead compounds for the development of novel thrombin inhibitors with improved efficacy and high safety profiles.