Importance of the Arg-Gly-Asp triplet in human thrombin for maintenance of structure and function

Site-directed mutagenesis was employed to assess the importance of the Arg-Gly-Asp triplet that comprises residues 197 to 199 in the B-chain of thrombin. Properties of the R197E and the D199E variants were compared with those of zeta-thrombin and the inactive S205A variant wherein the active site Ser is replaced by Ala. Relative to zeta-thrombin, the R197E thrombin variant under the assay conditions used exhibits 26% activity toward a small chromogenic substrate, 13% activity in the activation of protein C in the presence of thrombomodulin, < 3% activity in processing fibrinogen, and 1% activity in inducing platelet activation. Thus, the substrate specificity of thrombin was altered by the R197-->E replacement. The D199E variant was essentially inactive. It exhibited only 0.02% of the activity of thrombin toward the chromogenic substrate and its reactivity toward the active site-directed alkylating agent D-Phe-Pro-Arg-CH2Cl was 10,000-fold lower than that of thrombin. Like the inactive S205A thrombin variant, the D199E variant antagonized the interactions of thrombin with hirudin and thrombomodulin, but was a less effective antagonist. The dependence of the antagonism of the thrombin-thrombomodulin interaction on the concentration of D199E thrombin variant provided evidence suggesting the presence of two or more domains in thrombin that independently interact with their counterparts in thrombomodulin. Although the S205A thrombin variant antagonized the action of thrombin on platelets no such activity could be demonstrated for the D199E variant in the concentration range studied (< 800 nm). Comparison of the circular dichroism spectra of zeta-thrombin, the D199E, R197E, and S205A variants indicated that subtle differences in conformation exist between the D199E variant and the other thrombins. These differences in conformation might well account for the altered behavior of the D199E variant with respect to its interactions toward thrombomodulin, hirudin, and platelets.

Thrombin receptor agonist peptide induction of mitogenesis in CCL39 cells

The role of the thrombin receptor tethered ligand hypothesis in mediating the mitogenic responses of cells to thrombin was explored. We have found that small (5-14 amino acid) peptides corresponding to the proposed amino terminus of thrombin activated human and hamster thrombin receptors are mitogenic for the Chinese hamster fibroblast cell line, CCL39. Hirudin and hirugen block the mitogenic effects of thrombin but not the activity of the agonist peptides. Pertussis toxin treated cells do not respond to either alpha-thrombin or the agonist peptides. The data support the idea that the thrombin receptor on CCL39 cells, which is homologous to the thrombin receptor on human platelets, is capable of transmitting mitogenic signals by a mechanism consistent with the tethered ligand hypothesis.

Inhibition of dysthrombins Quick I and II by heparin cofactor II and antithrombin

Heparin cofactor II and antithrombin are plasma serine proteinase inhibitors whose ability to inhibit alpha-thrombin is accelerated by glycosaminoglycans. Dysfunctional thrombin mutants Quick I (Arg67-->Cys) and Quick II (Gly226-->Val) were used to further compare heparin cofactor II and antithrombin interactions. Quick I, Quick II, and alpha-thrombin were eluted at the same salt concentration from heparin-Sepharose suggesting that the putative heparin-binding site (also termed anion binding exosite-II) is functional. Antithrombin yielded similar inhibition rates for Quick I and alpha-thrombin in the absence or presence of various amounts of heparin. Also, Quick I was inhibited similarly to alpha-thrombin by heparin cofactor II in the absence of glycosaminoglycan. In contrast, glycosaminoglycan-accelerated Quick I inhibition by heparin cofactor II was greatly reduced indicating that anion binding exosite-I (where the mutation occurs in Quick I) is critical for increased inhibition by heparin cofactor II. We also found that heparin cofactor II formed a SDS-resistant bimolecular complex with Quick II and alpha-thrombin at similar rates and the rate of complex formation was accelerated in the presence of glycosaminoglycans. A three-dimensional molecular model of the Quick II active site compared to alpha-thrombin suggested that the heparin cofactor II Leu-Ser-reactive site sequence (P1-P1') is a compatible "pseudosubstrate" in contrast to the Arg-Ser sequence found in antithrombin. The importance of heparin cofactor II as a thrombin regulator will depend upon its ability to interact with glycosaminoglycans and the functional availability of thrombin exosites.

The linkage between binding of the C-terminal domain of hirudin and amidase activity in human alpha-thrombin

A method derived from the analysis of viscosity effects on the hydrolysis of the amide substrates D-phenylalanylpipecolyl-arginine-p-nitroaniline, tosylglycylprolylarginine-p-nitroanaline and cyclohexylglycylalanylarginine-p-nitroalanine by human alpha-thrombin was developed to dissect the Michaelis-Menten parameters Km and kcat into the individual rate constants of the binding, acylation and deacylation reactions. This method was used to analyse the effect of the C-terminal hirudin (residues 54-65) [hir-(54-65)] domain on the binding and hydrolysis of the three substrates. The results showed that the C-terminal hir-(54-65) fragment affects only the acylation rate, which is increased approx. 1.2-fold for all the substrates. Analysis of the dependence of acylation rate constants on hirudin-fragment concentration, allowed the determination of the equilibrium binding constant of C-terminal hir-(54-65) (Kd approximately 0.7 microM). In addition this peptide was found to competitively inhibit thrombin-fibrinogen interaction with a Ki which is in excellent agreement with the equilibrium constant derived from viscosity experiments. These results demonstrate that binding of hir-(54-65) to the fibrinogen recognition site of thrombin does not affect the equilibrium binding of amide substrates, but induces only a small increase in the acylation rate of the hydrolysis reaction.

Thrombin and albumin adsorption to PVA and heparin-PVA hydrogels. 2: Competition and displacement

Thrombin adsorption to polyvinyl alcohol (PVA) was different from its adsorption to polyethylene (PE)--not so much in amount, but in its affinity. Thrombin was more easily displaced from polyethylene and its adsorption was more readily prevented by prior or simultaneous exposure to albumin. From PVA (or heparin-PVA), only approximately 30% of the adsorbed protein could be removed by a series of eluents, including even harsh ones such as 2.5M NaOH and 6M guanidine; > 85% could be removed from PE. Thrombin adsorption to PVA was not affected by the presence of BSA in solution or at the surface, but was virtually prevented on PE by preexposure to or adsorption with BSA. Heparin-PVA was not much different than PVA in most of these experiments, but did exhibit a "Vroman effect". In the absence of fibrinogen or antithrombin III, there was a maximum in thrombin adsorption from plasma at a plasma concentration of 1%. The behavior on this surface was dependent on both exposure time and protein concentration. These studies highlight the complexity of the interaction between plasma proteins and polymer surfaces (particularly hydrogel surfaces) and the difficulty of obtaining a clear picture of what happens when a single protein interacts with a polymer in the presence of other proteins.

Differential activation of p44mapk (ERK1) by alpha-thrombin and thrombin-receptor peptide agonist

alpha-Thrombin (thrombin), a potent mitogen for CCL39 hamster lung fibroblasts, stimulates phosphoinositide-specific phospholipase C (PI-PLC) and inhibits adenylate cyclase via cleavage of a specific G-protein-coupled receptor (TH-R), recently cloned from human and hamster cells. This action can be entirely mimicked by the synthetic peptide SFFLRNP, referred to here as TMP (thrombin-mimicking peptide). TMP corresponds to the first seven amino acids of the new N-terminus generated by thrombin cleavage of the hamster TH-R. Although thrombin and TMP apparently generate identical early transmembrane signals, only thrombin is mitogenic on its own. TMP needs to be associated with fibroblast growth factor (FGF), a tyrosine kinase-activating growth factor, to induce cell-cycle re-entry. Here, we have examined the early and late phase of p44 MAP kinase (p44mapk) activation in G0-arrested CCL39 cells after stimulation by thrombin, TMP, FGF or TMP+FGF. We found that: (i) both thrombin and TMP rapidly activate p44mapk in a dose-dependent manner with maximum activation at around 5 min, (ii) after the initial burst of activation, a second and long-lasting wave of activation is observed in response to thrombin (10-100 nM) but not to TMP (up to 300 microM), (iii) FGF alone (25 ng/ml), like thrombin, rapidly and persistently activates p44mapk (20-fold at 5 min and about 3-fold after 2 h), (iv) TMP added together with FGF strongly potentiates the second and sustained phase of p44mapk activation. From these results we propose that: (1) thrombin-induced mitogenesis is mediated only in part by the TH-R recently cloned and (2) activation of p44mapk, in particular the long-lasting phase that correlates with DNA synthesis, is an obligatory event for cell-cycle re-entry.

Cellular effects of thrombin: pharmacology of the receptor(s) in various cell types and possible development of receptor antagonists

Thrombin is a key enzyme in the coagulation cascade being the final effector responsible for the conversion of fibrinogen to fibrin. Thrombin as an enzyme also acts on other substrates of the coagulation system, thus resulting in procoagulant tendencies, potentiation of ongoing reactions, feedback inhibition and/or limitation of clotting. Thrombin also has, however, a number of 'cellular' effects probably not only dependent on its enzymatic activity. This review focuses mostly on present knowledge about such mechanisms aimed towards the possible identification and development of molecules able to inhibit the thrombin cellular receptor.

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

The wealth of structural information now available for thrombin, its precursors, its substrates, and its inhibitors allows a rationalization of its many roles. alpha-thrombin is a rather rigid molecule, binding to its target molecules with little conformational change. Comparison of alpha-thrombin with related trypsin-like serine proteinases reveals an unusually deep and narrow active site cleft, formed by loop insertions characteristic of thrombin. This canyon structure is one of the prime causes for the narrow specificity of thrombin. The observed modularity of thrombin allows a diversity in this specificity; its "mix-and-match" nature is exemplified by its interactions with macromolecules (Fig. 20). The apposition of the active site to a hydrophobic pocket (the apolar binding site) on one side and a basic patch (the fibrinogen recognition exosite) on the other allows for a fine tuning of enzymatic activity, as seen for fibrinogen. Thrombin receptor appears to use the same sites, but in a different way. Protein C seems only able to interact with thrombin if the recognition exosite is occupied by thrombomodulin. These two sites are also optimally used by hirudin, allowing the very tight binding observed; thrombin inhibition is effected by blocking access to the active site. On the other hand, antithrombin III makes little use of the recognition exosite; instead, its interactions are tightened with the help of heparin, which binds to a second basic site (the heparin binding site). Thrombin's modularity is a result of the conjunction of amino acid residues of like properties, such as charge or hydrophobicity. The charge distribution plays a role, not only in the binding of oppositely charged moieties of interacting molecules, but also in selection and preorientation of them. Nonproteolytic cellular properties are attributed to 1) the rigid insertion loop at Tyr60A, and 2) a partially inaccessible RGD sequence. The former can interact with cells in the native form; the latter would appear to be presented only in an (at least partially) unfolded state. The membrane binding properties of prothrombin can be understood from the ordered arrangement of calcium ions on binding to the Gla domain. Kringle F2 binds to thrombin at the heparin binding site through charge complementarity; a conformational change appears to occur on binding. The observed rigidity of the thrombin molecule in its complexes makes thrombin ideal for structure based drug design. Thrombin can be inhibited either at the active site or at the fibrinogen recognition exosite, or both.(ABSTRACT TRUNCATED AT 400 WORDS)

Alteration of pharmacokinetics and pharmacodynamics of recombinant hirudin (rHV2-Lys 47) after repeated intravenous administration in dogs

Native hirudin is a heterogenous polypeptide obtained from the medicinal leech, Hirudo medicinalis. Recent advances in molecular biological techniques have led to the availability of large amounts of hirudin in the recombinant form. Recombinant hirudins (rH) are currently being investigated for potential use in the prophylaxis and treatment of deep venous thrombosis (DVT), in disseminated intravascular coagulation (DIC) and during cardiovascular bypass surgery. In this study, one specific variant of rH with a lysine residue in position 47 (rHV2-Lys 47) was administered in dogs in a multiple dose regimen of 2 mg/kg (i.v. bolus) for three weeks with a dosing interval of one week. After each dose, blood samples were collected at regular time intervals, plasma separated and stored at -4 degrees C. Concentrations of rHV2-Lys 47 in each sample were determined using an enzyme-linked immunosorbent assay (ELISA). Ex vivo antithrombin responses measured included activated partial thromboplastin time (APTT), calcium-thrombin time (Ca++TT-10 NIH units/ml) and a chromogenic anti-IIa assay. It was the purpose of this study to detect any sensitization or desensitization of antithrombin responses when rHV2-Lys 47 is used in a repeated fashion such as would be expected in the prophylaxis of DVT. The results indicated that there was no attenuation in the responses; however, there was a sensitization of response as measured by the Ca++TT (10 NIH units/ml). These findings could have major implications in the clinical use of rH where this drug is expected to be used in a multiple dose regimen.

Buried water in homologous serine proteases

Buried water molecules in the structurally homologous family of eukaryotic serine proteases were examined to determine whether buried waters and their protein environments are conserved in these proteins. We found 16 equivalent water sites conserved in trypsin/ogen, chymotrypsin/ogen, elastase, kallikrein, thrombin, rat tonin and rat mast cell protease, and 5 additional water sites in enzymes which share the primary specificity of trypsin. Based on an alignment of 30 serine protease sequences, it appears that the protein environments of these 21 conserved buried waters are highly conserved. The protein environments of buried waters are comprised primarily of atoms from highly conserved residues or main chain atoms from nonconserved residues. In one instance, the protein environment of a water is conserved even in the presence of an unlikely Pro/Ala substitution. We also note 3 instances in which a histidine side chain substitutes for water, suggesting that the structural role of water at these sites is satisfied by the presence of an alternative hydrogen bonding partner. Buried waters appear to be integral structural components of these proteins and should be incorporated into protein structures predicted on the basis of sequence homology to this family, including the catalytic domains of coagulation proteases.

Nuclear magnetic resonance solution structure of hirudin(1-51) and comparison with corresponding three-dimensional structures determined using the complete 65-residue hirudin polypeptide chain

The three-dimensional structure of the N-terminal 51-residue domain of recombinant hirudin in aqueous solution was determined by 1H nuclear magnetic resonance (NMR) spectroscopy, and the resulting high-quality solution structure was compared with corresponding structures obtained from studies with the intact, 65-residue polypeptide chain of hirudin. On the basis of 580 distance constraints derived from nuclear Overhauser effects and 109 dihedral angle constraints, a group of 20 conformers representing the solution structure of hirudin(1-51) was computed with the program DIANA and energy-minimized with a modified version of the program AMBER. Residues 3 to 30 and 37 to 48 form a well-defined molecular core with two antiparallel beta-sheets composed of residues 14 to 16 and 20 to 22, and 27 to 31 and 36 to 40, and three reverse turns at residues 8 to 11 (type II), 17 to 20 (type II') and 23 to 26 (type II). The average root-mean-square deviation of the individual NMR conformers relative to their mean co-ordinates is 0.38 A for the backbone atoms and 0.77 A for all heavy atoms of these residues. Increased structural disorder was found for the N-terminal dipeptide segment, the loop at residues 31 to 36, and the C-terminal tripeptide segment. The solution structure of hirudin(1-51) has the same molecular architecture as the corresponding polypeptide segment in natural hirudin and recombinant desulfatohirudin. It is also closely similar to the crystal structure of the N-terminal 51-residue segment of hirudin in a hirudin-thrombin complex, with root-mean-square deviations of the crystal structure relative to the mean solution structure of 0.61 A for the backbone atoms and 0.91 A for all heavy atoms of residues 3 to 30 and 37 to 48. Further coincidence is found for the loop formed by residues 31 to 36, which shows increased structural disorder in all available solution structures of hirudin, and of which residues 32 to 35 are not observable in the electron density map of the thrombin complex. Significant local structural differences between hirudin(1-51) in solution and hirudin in the crystalline thrombin complex were identified mainly for the N-terminal tripeptide segment and residues 17 to 21. These are further analyzed in an accompanying paper.

Impact of protein-protein contacts on the conformation of thrombin-bound hirudin studied by comparison with the nuclear magnetic resonance solution structure of hirudin(1-51)

The impact of protein-protein interactions on the conformation of the N-terminal hirudin domain consisting of residues 1 to 51 in the X-ray crystal structure of a hirudin-thrombin complex was investigated through comparisons with the nuclear magnetic resonance solution structure of hirudin(1-51). The close overall similarity observed between these two structures contrasts with the behavior of the C-terminal 17-residue polypeptide segment of hirudin, which is flexibly disordered in solution but exhibits a defined conformation in the complex with thrombin. Localized structural differences in the N-terminal domain include that residues 1 to 3 of hirudin in the crystalline complex form a hydrogen-bonding network with thrombin that is reminiscent of a parallel beta-sheet. Moreover, the backbone conformation of residues 17 to 20 in the complex does not contain the characteristic hydrogen bond observed for the type II' reverse turn in the solution structure, and the side-chains of Ser19 and Val21 have significantly different orientations in the two structures. Most of these structural changes can be related directly to thrombin-hirudin contacts, which may also be an important factor in the mechanism of hirudin action. In this context, it is of special interest that other residues that also make numerous contacts with thrombin, e.g. Thr4, Asp5 and Asn20, have identical conformations in free hirudin and in the complex.

Interactions of thermally denatured fibrinogen on polyethylene with plasma proteins and platelets

During the investigation of fibrin deposition onto hydrophobic polymers in contact with human blood, a model was developed in which fibrinogen was denatured and irreversibly coated onto a polyethylene surface by heating to 70 degrees C for 10 min. The denatured fibrinogen-polyethylene surface is resistant to fluid wall shear rates of up to 550 s-1 and the fibrinogen does not desorb in the presence of plasma proteins. Compared to uncoated polyethylene, little albumin or fibrinogen adsorbs to heat-denatured fibrinogen. Thrombin binds to the denatured fibrinogen-coated polyethylene with low affinity and also acts on the surface-bound denatured fibrinogen and cleaves fibrinopeptide A (FPA) quantitatively. Washed, 51Cr-labeled platelets do not adhere to the thermally denatured fibrinogen at either low or high shear rates compared to surfaces coated with undenatured fibrinogen (p < 0.01). These observations support the role of the D domain of fibrinogen in platelet adhesion because this is the region that is denatured by heating. In contrast, the E domain of fibrinogen is not altered by heating to 70 degrees C and hence remains susceptible to thrombin and/or plasmin cleavage. The characteristics of this surface are such that it can be used to develop fibrin-coated surfaces for use in studies of thrombus formation on artificial surfaces.

Structure of the hirulog 3-thrombin complex and nature of the S' subsites of substrates and inhibitors

The X-ray crystallographic structure of the human alpha-thrombin complex with hirulog 3 (a potent, noncleavable hirudin-based peptide of the "hirulog" class containing a beta-homoarginine at the scissile bond), which is isomorphous with that of the hirugen-thrombin crystal structure, was solved at 2.3-A resolution by starting with a model for thrombin derived from the hirugen-thrombin complex and was refined by restrained least squares methods (R = 0.132). Residues of hirulog 3 were well-defined in the electron density, which included most of the pentaglycine linker and the C-terminal helical turn that was disordered in a related structure of thrombin with hirulog 1. The interactions of D-Phe1'-Pro2'-beta-homoArg3' with the active site of thrombin were essentially identical to those of related structures of PPACK- (D-Phe-Pro-Arg chloromethyl ketone) and hirulog 1-thrombin, with the guanidinium function of the arginyl P1 residue forming a hydrogen-bonding ion pair with Asp189 of the S1 site. A noticeable shift in the CA atom of beta-homoArg3' due to the methylene insertion displaces the scissile bond from attack by Ser195, thus imparting proteolytic stability to the beta-homoArg hirulog derivative. Resolution of the pentaglycine spacer, linking N- and C-terminal functional domains into a single oligopeptide bivalent inhibitor, permitted delineation of corresponding S' subsites of thrombin. The position of Gly4' (P1') is stabilized by three hydrogen bonds with His57, Lys60F, and Ser195, while the conformational angles maintained in a strained, nonallowed configuration for non-glycyl amino acids.(ABSTRACT TRUNCATED AT 250 WORDS)

Energetics of thrombin-fibrinogen interaction

The kinetic mechanism of thrombin-fibrinogen interaction has been elucidated by steady-state measurements of synthetic substrate hydrolysis by human alpha-thrombin in the presence of human fibrinogen used as a competitive inhibitor and sucrose used as a viscogenic agent. Sucrose greatly affects the FKm for thrombin-fibrinogen interaction, without altering the intrinsic properties of the system. Under conditions of pH 7.5 and 0.1 M NaCl, fibrinogen behaves like a sticky substrate for thrombin, with acylation being comparable to dissociation in the temperature range 20-37 degrees C. In the same temperature range, deacylation is much faster than acylation. The van't Hoff enthalpy of binding for thrombin-fibrinogen interaction is -24 +/- 3 kcal/mol and the entropy is -55 +/- 11 cal mol-1 deg-1. A chemical compensation effect is present in the binding of fibrinogen and synthetic amide substrates to thrombin, with the delta H and delta G values being linked through a linear relationship.

Structural differences in active site-labeled thrombin complexes with hirudin isoinhibitors

Hirudin, a 65 amino acid polypeptide from the medicinal leech, is the most potent thrombin inhibitor known to date. Recently, recombinant forms have been reported, which are as effective as the isolated forms. The studies presented here demonstrate sensitive spectroscopic methods for distinguishing binding of two recombinant hirudins, HV1 and HV2-Lys 47, with active site-labeled human alpha-, epsilon- and zeta-thrombins. Specifically, fluorosulfonylphenyl nitroxide spin labels, dansyl fluoride and p-nitrophenylanthranilate, were employed as active site-directed covalent reporter groups. In general, the nitroxide immobilization was greater for spin-labeled thrombin complexes with HV2-Lys 47 vs. HV1. The two fluorophore moieties, dansyl and anthraniloyl, were also sensitive to differences in HV1 and HV2-Lys 47 binding, including interactions with loop 145-150 of the thrombin structure where the epsilon- and zeta-thrombin cleavages exist. Speculation over sequence differences between the two isoinhibitors centers on residues 24 and 47, both of which involve either a loss or gain of charge on the side chain.

Kinetic characterization of the substrate reaction between a complex of antithrombin with a synthetic reactive-bond loop tetradecapeptide and four target proteinases of the inhibitor

A tetradecapeptide corresponding to the P1 to P14 region of the reactive-bond loop of antithrombin (AT) binds to the inhibitor, presumably as a middle strand of the A beta-sheet, thereby converting AT from an inhibitor to a substrate of thrombin (Björk, I., Ylinenjärvi, K., Olson, S.T., and Bock, P. E. (1992) J. Biol. Chem. 267, 1976-1982). The kinetics of cleavage of the AT reactive bond in the AT-peptide complex by four target proteinases were quantified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and densitometry. The kcat/Km values for thrombin and factor IXa were indistinguishable from the second-order rate constants for AT inhibition of these enzymes, whereas the values for factor Xa and plasmin were 10-17-fold higher than the inhibition rate constants. Heparin with high affinity for AT accelerated the substrate reaction with thrombin to an extent consistent with the reduced heparin affinity of the AT-peptide complex. These data show that blocking by the peptide of the putative intramolecular association of the P1 to P14 region of the AT reactive-bond loop with the A beta-sheet leads to AT functioning as a substrate of its target enzymes with an efficiency that equals or exceeds the action of uncomplexed AT as an inhibitor and with the expected heparin activation. The results thus suggest that a substrate-like attack of the proteinase on the inhibitor reactive bond in an exposed loop initiates the inhibition reaction. This attack presumably induces the subsequent trapping of the enzyme by the insertion of the reactive-bond loop into the A beta-sheet.