A player of many parts: the spotlight falls on thrombin's structure

The Na+ binding site of thrombin

Thrombin is an allosteric serine protease existing in two forms, slow and fast, targeted toward anticoagulant and procoagulant activities. The slow --> fast transition is induced by Na+ binding to a site contained within a cylindrical cavity formed by three antiparallel beta-strands of the B-chain (Met180-Tyr184a, Lys224-Tyr228, and Val213-Gly219) diagonally crossed by the Glu188-Glu192 strand. The site is shaped further by the loop connecting the last two beta-strands and is located more than 15 A away from the catalytic triad. The cavity traverses through thrombin from the active site to the opposite surface and contains Asp189 of the primary specificity site near its midpoint. The bound Na+ is coordinated octahedrally by the carbonyl oxygen atoms of Tyr184a, Arg221a, and Lys224, and by three highly conserved water molecules in the D-Phe-Pro-Arg chloromethylketone thrombin. The sequence in the Na+ binding loop is highly conserved in thrombin from 11 different species and is homologous to that found in other serine proteases involved in blood coagulation. Mutation of two Asp residues flanking Arg221a (D221A/D222K) almost abolishes the allosteric properties of thrombin and shows that the Na+ binding loop is also involved in direct recognition of protein C and antithrombin.

Amide and alpha-keto carbonyl inhibitors of thrombin based on arginine and lysine: synthesis, stability and biological characterization

We report structure-activity investigations in a series of tripeptide amide inhibitors of thrombin, and the development of a series of highly potent active site directed alpha-keto carbonyl inhibitors having the side chain of lysine at P1. Compounds of this class are unstable by virtue of reactivity at the electrophilic carbonyl and racemization at the adjacent carbon (CH). Modifications of prototype alpha-keto-ester 8a have afforded analogs retaining nanomolar Ki. Optimal potency and stability have been realized in alpha-keto-amides 11b (Ki = 2.8 nM) and 11c (Ki = 0.25 nM).

Antithrombotic effects of orally active synthetic antagonist of activated factor X in nonhuman primates

BACKGROUND

Since activated factor X (FXa) has a central role in hemostasis and thrombosis, it is an attractive target for antithrombotic strategies. Accordingly, we evaluated the relative antihemostatic and antithrombotic effects of an orally active amidinoaryl propanoic acid inhibitor of FXa, APAP, in baboons.

METHODS AND RESULTS

With a two-component thrombogenic device that induced the concurrent formation of both arterial-type platelet-rich and venous-type fibrin-rich thrombus when interposed in chronic exteriorized arteriovenous (AV) femoral shunts flowing at 40 mL/min, thrombus formation was compared for oral versus parenteral APAP by measurement of 111In-platelet deposition, 125I-fibrin accumulation, thrombotic obstruction of flow, and circulating levels of blood biochemical markers of thrombosis. The direct infusion of APAP (120 micrograms/min) into AV shunts proximal to thrombogenic devices for 1 hour achieved local drug levels of 4.3 +/- 0.4 mg/L and substantially reduced the accumulation of platelets and fibrin in the formation of venous-type fibrin-rich thrombus (P < .01) but not in the formation of platelet-rich arterial-type thrombus (P > .1). APAP was subsequently removed from plasma with plasma clearance rates of T50 alpha of 6.3 minutes and T50 beta of 99 minutes. The oral administration of APAP (50 mg/kg) produced peak plasma levels of 3.7 +/- 1.4 micrograms/mL at 30 minutes and gradually declining plasma levels over about 6 to 8 hours, with bioavailability estimated to be approximately 5% to 12%. Oral APAP decreased platelet deposition (P < .01) and fibrin accumulation (P < .05) in venous-type thrombus but failed to decrease platelet or fibrin accumulation in arterial-type thrombus (P > .1 in both cases). Oral and infused APAP prolonged the activated partial thromboplastin time and prevented thrombus-dependent elevations in plasma fibrinopeptide A, thrombin-antithrombin III complex, beta-thromboglobulin, and platelet factor 4 levels. Additionally, APAP produced dose-dependent inhibition of FXa bound to thrombus on segments of vascular graft interposed in exteriorized AV shunts for 15 minutes.

CONCLUSIONS

An oral synthetic antagonist of FXa, APAP, inhibits the formation of venous-type fibrin-rich thrombus by inactivating bound and soluble FXa without impairing platelet hemostatic function.

Human thrombin variable region 1, including E39, is involved in interactions with alpha 1-antitrypsin M358R and protein C

We used an antithrombin autoantibody (IgG D), the epitope of which encompasses ABE1 and amino acids located within variable region 1, to study thrombin interactions with R358 alpha 1-AT and protein C. IgG D inhibited the thrombin interaction with R358 alpha 1-AT, while hirugen had no effect, indicating that the interaction of R358 alpha 1-AT with thrombin may involve the VR1 subsite. We also obtained evidence that VR1 may be involved in the activation of protein C by thrombin in the absence of thrombomodulin. Moreover, IgG D attenuated the inhibitory effect of calcium ions during protein C activation by thrombin, probably by masking E39 within the VR1 site.

Identification of the proteolytic thrombin fragments formed after cleavage with rat mast cell protease 1

We have previously identified rat mast cell protease 1 (RMCP-1), a chymotrypsin-like secretory granule serine protease, as a potent inactivator of thrombin. The present study outlines the cleavage pattern obtained after degradation of thrombin by RMCP-1. The cleavage sites in thrombin were identified by N-terminal amino acid sequence analysis of recovered thrombin fragments. Incubation of thrombin with RMCP-1 resulted in the rapid formation of a 37-kDa fragment, due to cleavage of the Phe1G-Gly1F bond in the thrombin A chain (numbering of amino acid residues according to topological equivalencies with chymotrypsinogen). Further incubation resulted in cleavage of the Trp148-Thr149 bond in the B chain, along with the formation of fragments of 27 kDa and 15 kDa. When the RMCP-1/thrombin mixtures were incubated further, successive degradation of the 37-kDa, 27-kDa and 15-kDa fragments was observed, along with cleavage of the Tyr117-Ile118 bond in the B chain and the formation of fragments of 12, 9 and 6 kDa. No residual thrombin activity was detected after the degradation process had proceeded to this stage. Heparin was shown to markedly enhance the rate of thrombin degradation by RMCP-1.

Thrombin, phorbol ester, and cAMP regulate thrombin receptor protein and mRNA expression by different pathways

Human mesangial cells have been used to study the regulation of thrombin receptor protein and mRNA expression during cross-talk between different signal transduction pathways. Persistent activation of thrombin receptor by thrombin led to homologous down-regulation of thrombin receptor protein. However, thrombin receptor mRNA expression was not affected, suggesting that increased receptor degradation is responsible for homologous down-regulation. Chronic activation of protein kinase C by phorbol 12-myristate 13-acetate (PMA) and of adenylylcyclase by prostaglandin E1 (PGE1) resulted in heterologous down-regulation of thrombin receptor protein. In contrast to thrombin, PMA and PGE1 reduced in parallel thrombin receptor mRNA levels to 51% and 24% of control, respectively, indicating that heterologous down-regulation of thrombin receptor protein is, at least in part, due to inhibition of receptor mRNA expression. The mechanisms of heterologous down-regulation of thrombin receptor protein have been studied in detail and compared to homologous down-regulation. PMA-induced down-regulation was completely blocked by GF 109 203 X, an inhibitor of protein kinase C. However, the loss of thrombin receptor induced by thrombin was not prevented by GF 109 203 X, indicating that homologous regulation is not dependent on protein kinase C activation. The heterologous effect of PGE1 was mimicked by 8-bromo-cAMP, isobutylmethylxanthine, and forskolin, suggesting that an increase in intracellular cAMP level is involved in heterologous regulation. Interestingly, heterologous down-regulation induced by PGE1 seems not to require previous internalization of thrombin receptor. These data indicate that thrombin receptor protein and mRNA expression can be regulated in homologous and heterologous ways by different mechanisms.

The clot thickens: clues provided by thrombin structure

Thrombin is the primary promoter of blood clotting; it also plays an important role in the regulation of the coagulation cascade and has been implicated in a number of other cellular processes. How can one molecule catalyse such a variety of events? The recent X-ray structure determination of human a-thrombin and related structures shows that the molecule can be divided into several functional regions that recognize different chemical moieties. By using different combinations of these elements, thrombin can interact with a variety of macromolecules with high specificity.

The X-ray crystal structure of thrombin in complex with N alpha-2-naphthylsulfonyl-L-3-amidino-phenylalanyl-4-methylpiperidide: the beneficial effect of filling out an empty cavity

The 2.5 Angstrum structure of bovine epsilon-thrombin in complex with N alpha-2-naphthyl-sulfonyl-L-3-amidinophenylalanyl-4-methylpiper idide (L-NAPAMP) was solved and crystallographically refined to an R-value of 0.19. The L-NAPAMP moiety is completely and unambiguosly defined in the electron density. NAPAMP binds almost identical to the related 4-methyl deficient 3-amidino-phenylalanyl derivative TAPAP. The overall binding geometry appears dominated by the fixation of the 3-amidinophenyl ring in thrombin's S1-pocket and the hydrogen bonds to Gly 216, irrespective of the presence or absence of a substituent in the 4-position of the piperidine ring. The additional 4-methyl group gives rise to a 17-fold better binding. The more complete spatial occupancy of the hydrophobic S2-cavity therefore accounts for a decrease in free energy of binding of 15 kcal/mol, a value comparable with that anticipated for filling up a stable empty cavity of similar size by a methyl group.

Thrombin inhibitors as antithrombotic agents: the importance of rapid inhibition

For use as an antithrombotic agent, a thrombin inhibitor must be potent and specific, i.e., it should not significantly inhibit the proteases of the anticoagulation (activated protein C) and fibrinolytic systems (plasminogen activator and plasmin). Previous evaluation of potency and specificity has been based on inhibition constants (Ki values). However, consideration of the kinetic parameters for natural plasma serine protease inhibitors indicates that a low Ki value with thrombin is not sufficient; the inhibited complex must also form rapidly. Moreover, potent inhibition of activated protein C and plasmin could be tolerated providing the inhibited complex only forms slowly. An ideal profile of kinetic parameters with thrombin, activated protein C and plasmin is formulated and discussed in relation to various classes of thrombin inhibitors. Examination of kinetic data for thrombin inhibitors currently in clinical trials (hirudin and hirulog) indicates that they possess this ideal profile of kinetic parameters.

Structure/function aspects of neutral P1 residue peptide inhibitors of thrombin

Control of thrombin by its inhibition in indications such as myocardial infarction, unstable angina or stroke has been demonstrated to be therapeutically valuable. However restoration of hemostasis by targeting thrombin while avoiding its fellow serine proteinases, (e.g. plasmin, trypsin), remains a challenge of medicinal chemistry. Tripeptide-boronates and -phosphonates with neutral P1 side chains meet these criteria. Development of novel, high yielding chemical routes furnishes a wide range of un-natural P1 functionalities, demonstrating that this indeed is a class effect with selectivity conferred by the uncharged P1 residue. For example N-benzyloxycarbonyl-D-phenylalanylprolyl-1- (3-methoxypropyl) boroglycine ester (1) has a Ki value for thrombin of 7 nM and greater than two order of magnitude higher with all other serine proteinases tested. The ester group determines the kinetics of inhibition by tripeptide phosphonates, with diphenylphosphonates being slow tight binding inhibitors, showing 50% reversibility of inhibition. Therefore this design of inhibitors offers a facile strategic approach to development as thrombin specific pharmaceutical agents.

Coagulation factors and their inhibitors

A comprehensive three-dimensional picture of the coagulation process is beginning to emerge. Crystallographic structure determinations of prothrombin, factor Xa, factor IXa, tissue factor and factor XIII represent important advances in our understanding of the coagulation cascade. Similarly, structures of antithrombin, tissue factor pathway inhibitor and thrombomodulin provide details of endogenous anticoagulatory mechanisms. NMR spectroscopy of multiple domains of coagulation proteins represents an important contribution to the analysis of flexibility and rigidity of modular proteins. Thrombin, as the prime candidate for antithrombotic drug design, continues to be an object of intense efforts in applied crystallography.

Identification of a region in protein C involved in thrombomodulin-stimulated activation by thrombin: potential repulsion at anion-binding site I in thrombin

During coagulation human protein C is activated by thrombin; however, this cleavage reaction is slow unless thrombin is complexed with a cofactor, thrombomodulin. Near the thrombin cleavage site in protein C is a cluster of basic residues, at positions P5' (Lys-174), P8' (Arg-177) and P9' (Arg-178). We have explored the role of this basic cluster in the activation of protein C by thrombin, and by thrombin-thrombomodulin complex, by substitution of glutamic acid at each position to generate the acidic protein C derivative P'-EEE. The activation rate of P'-EEE by free alpha-thrombin was approx. 12-fold faster than that observed for wild-type (wt) human protein C zymogen (HPC) in the presence of calcium, but unchanged in the absence of calcium. While the thrombin-catalysed activation of wt-HPC was stimulated approx. 300-fold by thrombomodulin, we observed no effect of thrombomodulin on thrombin-catalysed activation of the P'-EEE derivative. Using synthetic peptides that bind to anion-binding site I of thrombin (thrombin-receptor sequence 52-66 and hirudin sequence 54-65 SO4 Tyr), we found that the rate of thrombin-catalysed activation of wt-HPC in the presence of calcium could be increased severalfold in a dose-dependent manner. However, the enhanced rate of thrombin-catalysed activation of P'-EEE could be progressively reduced to wt-HPC levels with increasing concentrations of both synthetic peptides. Our data suggest that the P' basic cluster in protein C reduces interaction with free alpha-thrombin through electrostatic repulsion with anion-binding site I, a site that is masked when thrombomodulin binds thrombin. Further, the lack of thrombomodulin cofactor activity with thrombin-catalysed activation of P'-EEE suggests that the basic cluster in protein C forms a contact site with thrombomodulin.

Thrombin binding to platelets and their activation in plasma

The interactions of alpha-thrombin with platelets are critical in haemostasis and arterial thrombosis. This study established methods for characterizing the binding of alpha-thrombin to platelets and some of its consequences in platelet-rich plasma. The binding of alpha-thrombin to platelets and the subsequent platelet activation were quantified by flow cytometry, using affinity purified polyclonal antibodies to human alpha-thrombin and a monoclonal antibody to GMP-140, respectively. Dose-dependent binding of alpha-thrombin to platelets and their activation occurred in parallel, both reaching the maxima for each enzyme concentration within 10s after > or = 1.0 nM alpha-thrombin was added to recalcified PRP containing 1 microM recombinant tick anticoagulant peptide. The tick anticoagulant peptide abrogated prothrombin activation in the platelet-rich plasma. alpha-Thrombin binding to platelets, and their activation, were abrogated by a monoclonal antibody to the hirudin tail-like domain of the seven transmembrane thrombin receptor on platelets. Therefore this receptor represents an important site for alpha-thrombin binding to platelets suspended in plasma. D-Phe-Pro-ArgCH2-alpha-thrombin only bound to platelets when its concentration was > or = 100 nM, and it did so without inhibiting platelet activation by alpha-thrombin. Whereas concentrations of hirudin equimolar to those of alpha-thrombin failed to abrogate alpha-thrombin-mediated activation of platelets, a 10-fold molar excesses of hirudin over alpha-thrombin abrogated alpha-thrombin binding to platelets. The demonstration that > or = 1.0 nM alpha-thrombin can bind to platelets and initiate their activation raises the possibility that the levels of thrombin generated in venous and arterial thrombosis contribute to platelet activation in vivo.

Direct thrombin inhibitors in cardiovascular medicine

Currently used antithrombotics such as heparin have a number of potential limitations that may be overcome by the new class of agents that directly inhibit thrombin. These agents variously block the active catalytic and/or the anion binding exosites of the thrombin molecule and are potent and specific inhibitors of thrombin's many biological actions, as demonstrated by in vitro and animal models of thrombosis. Preliminary data indicate that the direct antithrombins are safe and efficacious in humans, and their use in acute coronary syndromes and coronary angioplasty in place of heparin has yielded promising early results. Phase III trials in these clinical settings are currently under way. Newer antithrombotics that inhibit thrombin generation and thrombin activity at various strategic points within the coagulation cascade are also in the early stages of development.

The molecular genetics of familial venous thrombosis

This study of naturally occurring mutations predisposing to venous thrombosis has led to a number of important advances in our understanding of protein structure and function relationships and the molecular basis of gene mutation. It has also potentiated the accurate and reliable presymptomatic and antenatal detection of predisposing gene lesions. Perhaps the major challenge facing us is the probabilistic nature of thromboembolism; only a certain proportion of patients with recognized gene defects predisposing to thrombosis will actually suffer from thrombotic episodes. Environmental insults of various kinds, and perhaps epistatic effects resulting from the influence of other loci, are likely to be contributory factors and will help to determine whether a thrombotic event occurs in individuals already compromised by a defect in a gene whose malfunction is known to predispose to thrombosis. Since molecular genetic techniques allow us to dissect the allelioheterogeneity of the different deficiency states by characterizing the wide spectrum of gene mutations giving rise to thrombosis, it may eventually prove possible to relate specific gene lesions to the probability of thromboembolism as well as to the severity and frequency of thrombotic episodes. The multifactorial nature of thrombosis demands a multidisciplinary approach to the analysis of its causation, early detection, treatment and prevention. The application of the new and powerful techniques of molecular genetics promises to make a substantial contribution to all aspects of thrombosis research.

Structure around the cleavage site in the thrombin receptor determined by NMR spectroscopy

NMR spectroscopic experiments were performed to study the structure of synthetic peptides identical to two extracellular regions of the human thrombin receptor. The smaller molecule, comprising 14 amino acids, was biologically active and was equivalent to the "tethered ligand" exposed after cleavage of the receptor by thrombin. The principal structural elements were two overlapping turns (amino acids 5-8 and 6-9), the second of which was stabilized by a hydrogen bond, 6CO-9NH. The five N-terminal residues, considered to be responsible for biological activity, were essentially unstructured. A second version of this peptide, biologically inactive due to the reversal of the two N-terminal amino acids, had a very similar structure. A longer peptide (23 amino acids) covering the proposed thrombin cleavage site was found to be more highly structured. The seven residues from Pro-2 to Arg5 (the N-terminal amino acid exposed after cleavage is taken as residue 1) formed a 3(10) helix which is not present in the shorter tethered ligand peptide. The structure is partially stabilized by a charged hydrogen bond between the side chains of Arg-1 and Asp-3. The overlapping turns observed in the shorter peptides could also be distinguished in the longer molecule. On the basis of the structure determined for the peptide which encompasses the cleavage site and the determined structure of thrombin, a model is postulated for the interaction of the thrombin receptor and the protease during activation.

Structural basis for type I and type II deficiencies of antithrombotic plasma protein C: patterns revealed by three-dimensional molecular modelling of mutations of the protease domain

Familial deficiency of protein C is associated with inherited thrombophilia. To explore how specific missense mutations might cause observed clinical phenotypes, know protein C missense mutations were mapped onto three-dimensional homology models of the protein C protease domain, and the implications for domain folding and structure were evaluated. Most Type I missense mutations either replaced internal hydrophobic residues (I201T, L223F, A259V, A267T, A346T, A346V, G376D) or nearby interacting residues (I403M, T298M, Q184H), thus disrupting the packing of internal hydrophobic side chains, or changed hydrophilic residues, thus disrupting ion pairs (N256D, R178W). Mutations (P168L, R169W) at the activation site destabilized the region containing the activation peptide structure. Most Type II mutations involved solvent-exposed residues and were clustered either in a positively charged region (R147W, R157Q, R229Q, R352W) or were located in or near the active site region (S252N, D359N, G381S, G391S, H211Q). The cluster of arginines 147, 157, 229, and 352 may identify a functionally important exosite. Identification of the spatial relationships of natural mutations in the protein C model is helpful for understanding manifestations of protein C deficiency and for identification of novel, functionally important molecular features and exosites.

Charge reversal at the P3' position in protein C optimally enhances thrombin affinity and activation rate

We have examined the properties of several human protein C (HPC) derivatives with substitutions for acidic residues near the thrombin cleavage site, including changing the P3' Asp to Asn (D172N), Gly (D172G), Ala (D172A), or Lys (D172K). The rate of thrombin-catalyzed activation of D172N, D172G, and D172A was increased 4-9-fold compared to wild-type HPC, primarily due to a reduction in the inhibitory effect of calcium and a resulting increase in affinity for free alpha-thrombin. There was no significant increase in activation rate or affinity with these 3 derivatives in the absence of calcium, confirming that P3' Asp affects calcium dependency in the native protein C molecule. With charge reversal at P3' (D172K), there was a 30-fold increase in activation rate in the presence of calcium, but unlike the other derivatives, there was a substantial effect (5-fold) on the activation rate and affinity for free alpha-thrombin in the absence of calcium. Thus, protein C affinity for thrombin appears to be influenced by a combination of calcium-dependent and -independent effects of the acidic P3' residue.

Binding of alpha-thrombin to fibrin depends on the quality of the fibrin network

Binding of human alpha-thrombin to fibrin was studied in a purified system at pH 7.35, I 0.08 and 37 degrees C. Binding experiments with active thrombin resulted in fibrin clots of variable quality, depending on the thrombin concentration: opaque gels composed of 'coarse' network were produced at low thrombin concentrations, while increasing concentrations of thrombin led to more translucent 'fine' gels. Scatchard analysis showed a non-linear dependence of thrombin binding to fibrin, suggesting the existence in fibrin(ogen) of multiple classes of binding sites for thrombin. Binding of catalytic-site-inhibited thrombin was investigated in clots of defined quality produced with three different concentrations of a thrombin-like enzyme, batroxobin (EC 3.4.21.29). Straight lines of different slopes were established by Scatchard analysis of binding data at each fixed batroxobin concentration. These results favour a model according to which binding affinity for thrombin depends on the thickness of fibrin bundles. Labelled active-site-inactivated thrombin incorporated in batroxobin-induced clots was only sparingly released during incubation for 24 h in the presence of a 200-fold excess of unlabelled thrombin, indicating that thrombin binding to fibrin is not reversible and that Scatchard analysis is not appropriate for quantification of binding parameters. Irreversible binding of thrombin appears to reflect trapping of thrombin molecules within fibrin fibres. The amount of trapped thrombin depends on the quality of the fibrin fibres, which in turn is determined by the concentration of the clotting enzyme.

Long-term effects of thrombin require sustained activation of the functional thrombin receptor

Thrombin is a potent activator of human glomerular epithelial cells (HGEC). Here we compare short-term and long-term effects of thrombin and thrombin receptor agonist peptide (TRAP) which selectively activates the functional thrombin receptor. TRAP, as thrombin, increases intracellular free Ca2+ concentration and acts synergistically with growth factors possessing tyrosine kinase receptors on DNA synthesis. Thrombin induces synthesis of proteins of the fibrinolytic system and cell proliferation if it is present for at least 8 h. TRAP alone does not stimulate protein synthesis and is not mitogenic. However, in the presence of the aminopeptidase inhibitor amastatin all long-term effects of thrombin can be fully mimicked by TRAP. In conclusion, different effects of thrombin and TRAP may be related to the degradation of TRAP by cellular ectoenzymes. The recently cloned thrombin receptor accounts for early intracellular signals and long-term cellular effects that require sustained activation of this receptor.