Structure-based understanding of ligand affinity using human thrombin as a model system

The effect of the P1 side chain on the binding of optimized carboxylate and activated carbonyl inhibitors of the hepatitis C virus NS3 protease

Peptidyl inhibitors of the hepatitis C virus NS3 protease hold much promise as direct-acting antiviral agents against hepatitis C infection. The optimization of N-terminal cleavage products, found to exhibit activity (product inhibition) against the enzyme, has led to potent tripeptide inhibitors that bear free C-terminal carboxylate groups. An analogous activated carbonyl compound (pentafluoroethyl ketone) bearing a P1 norvaline (Nva) was found to possess comparable activity against hepatitis C virus protease. However, an analogue bearing an aminocyclopropylcarboxylic acid (Acca) P1 residue exhibited very poor activity. 19F-NMR studies indicate that the propensity of the Acca-derived activated carbonyl to form hemiketals is only slightly reduced compared with that of a P1 Nva equivalent. These results, as well as molecular modeling studies, argue against steric hindrance of the nucleophilic attack of Ser-139 accounting for the poor mechanism-based inhibition by the former. We hypothesize that the conformational properties of the respective C-termini in the context of an adaptable active site account for the divergent P1 structure–activity relationships.

Proton bridging in the interactions of thrombin with small inhibitors

Thrombin is the pivotal serine protease enzyme in the blood cascade system. Phe-Pro-Arg-chloromethylketone (PPACK), phosphate, and phosphonate ester inhibitors form a covalent bond with the active-site Ser of thrombin. PPACK, a mechanism-based inhibitor, and the phosphate/phosphonate esters form adducts that mimic intermediates formed in reactions catalyzed by thrombin. Therefore, the dependence of the inhibition of human alpha-thrombin on the concentration of these inhibitors, pH, and temperature was investigated. The second-order rate constant (ki/Ki) and the inhibition constant (Ki) for inhibition of human alpha-thrombin by PPACK are (1.1 +/- 0.2) x 10(7) M(-1) s(-1) and (2.4 +/- 1.3) x 10(-8) M, respectively, at pH 7.00 in 0.05 M phosphate buffer and 0.15 M NaCl at 25.0 +/- 0.1 degrees C, in good agreement with previous reports. The activation parameters at pH 7.00 in 0.05 M phosphate buffer and 0.15 M NaCl are as follows: DeltaH = 10.6 +/- 0.7 kcal/mol, and DeltaS = 9 +/- 2 cal mol(-1) degrees C(-1). The pH dependence of the second-order rate constants of inhibition is bell-shaped. Values of pKa1 and pKa2 are 7.3 +/- 0.2 and 8.8 +/- 0.3, respectively, at 25.0 +/- 0.1 degrees C. A phosphate and a phosphonate ester inhibitor gave higher values, 7.8 and 8.0 for pKa1 and 9.3 and 8.6 for pKa2, respectively. They inhibit thrombin more than 6 orders of magnitude less efficiently than PPACK does. The deuterium solvent isotope effect for the second-order rate constant at pH 7.0 and 8.3 at 25.0 +/- 0.1 degrees C is unity within experimental error in all three cases, indicating the absence of proton transfer in the rate-determining step for the association of thrombin with the inhibitors, but in a 600 MHz 1H NMR spectrum of the inhibition adduct at pH 6.7 and 30 degrees C, a peak at 18.10 ppm with respect to TSP appears with PPACK, which is absent in the 1H NMR spectrum of a solution of the enzyme between pH 5.3 and 8.5. The peak at low field is an indication of the presence of a short-strong hydrogen bond (SSHB) at the active site in the adduct. The deuterium isotope effect on this hydrogen bridge is 2.2 +/- 0.2 (phi = 0.45). The presence of an SSHB is also established with a signal at 17.34 ppm for a dealkylated phosphate adduct of thrombin.

Probing the Kinetic Landscape of Transient Peptide−Protein Interactions by Use of Peptide 15N NMR Relaxation Dispersion Spectroscopy: Binding of an Antithrombin Peptide to Human Prothrombin

Protein-ligand interactions may lead to the formation of multiple molecular complexes in dynamic exchange, affecting the kinetic and thermodynamic characteristics of the binding equilibrium. We followed the dissociation kinetics of the transient and specific complex of an antithrombotic peptide N-acetyl-Asp(55)-Phe-Glu-Glu-Ile-Pro(60)-Glu-Glu-Tyr-Leu-Gln(65) with human prothrombin by use of (15)N NMR relaxation dispersion spectroscopy of the peptide. Every one of the five (15)N-labeled adjacent residues of the peptide exhibited apparently different kinetic exchange and relaxation behaviors, which were especially evident at different concentrations of prothrombin. Binding-induced (15)N relaxation dispersion of residues Phe(56), Glu(57), Glu(58), and Ile(59) can be fitted phenomenologically to a two-site on-and-off exchange mechanism with physically feasible relaxation and kinetic parameters obtained for residues Phe(56), Glu(58), and Ile(59), independent of the prothrombin concentration. The apparent kinetic parameters of Glu(57) show some dependence on the concentration of prothrombin and the extracted transverse relaxation rate for Glu(57) in the bound state was severalfold higher than that expected for a protein-peptide complex with a size of approximately 72 kDa. In addition, the equilibrium population of the bound peptide obtained for Glu(57) was inconsistent with those for Phe(56), Glu(58), and Ile(59) and with the prothrombin/peptide ratios used in the experiments. These discrepancies can be explained by the presence of two conformations for the peptide-protein complex exchanging at a rate of approximately 100 s(-)(1). In all, our study shows that fast dissociation of protein-peptide complexes can be studied quantitatively using peptide (15)N NMR relaxation dispersion measurements without a precise knowledge of the peptide and protein concentrations. In addition, protein titration was found to improve the accuracy of quantitative analysis and may make it possible to determine the rate of conformational changes within the protein-peptide complex.

A personal account of the role of peptide research in drug discovery: the case of hepatitis C

Although peptides themselves are not usually the end products of a drug discovery effort, peptide research often plays a key role in many aspects of this process. This will be illustrated by reviewing the experience of peptide research carried out at IRBM in the course of our study of hepatitis C virus (HCV). The target of our work is the NS3/4A protease, which is essential for maturation of the viral polyprotein. After a thorough examination of its substrate specificity we fine-tuned several substrate-derived peptides for enzymology studies, high-throughput screening and as fluorescent probes for secondary binding assays. In the course of these studies we made the key observation: that the protease is inhibited by its own cleavage products. Single analog and combinatorial optimization then derived potent peptide inhibitors. The crucial role of the NS4A cofactor was also addressed. NS4A is a small transmembrane protein, whose central domain is the minimal region sufficient for enzyme activation. Structural studies were performed with a peptide corresponding to the minimal activation domain, with a series of product inhibitors and with both. We found that NS3/4A is an induced fit enzyme, requiring both the cofactor and the substrate to acquire its bioactive conformation; this explained some puzzling results of 'serine-trap' type inhibitors. A more complete study on NS3 activation, however, requires the availability of the full-length NS4A protein. This was prepared by native chemical ligation, after sequence engineering to enhance its solubility; structural studies are in progress. Current work is focused on the P' region of the substrate, which, at variance with the P region, is not used for ground state binding to the enzyme and might give rise to inhibitors showing novel interactions with the enzyme.

Thrombin inhibitor design: X-ray and solution studies provide a novel P1 determinant

The crystal structures of proflavin and 6-fluorotryptamine thrombin have been completed showing binding of both ligands at the active site S1 pocket. The structure of proflavin:thrombin was confirmatory, while the structure of 6-fluorotryptamine indicated a novel binding mode at the thrombin active site. Furthermore, speculation that the sodium atom identified in an extended solvent channel beneath the S pocket may play a role in binding of these ligands was investigated by direct proflavin titrations as well as chromogenic activity measurements as a function of sodium concentration at constant ionic strength. These results suggested a linkage between the sodium site and the S1 pocket. This observation could be due to a simple ionic interaction between Asp189 and the sodium ion or a more complicated structural rearrangement of the thrombin S1 pocket. Finally, the unique binding mode of 6-fluorotryptamine provides ideas toward the design of a neutrally charged thrombin inhibitor.

Atomic structures of two nitroxide spin labels complexed with human thrombin: comparison with solution studies

Crystal structures of thrombin complexed with two spin labels called para-V, 4-(2,2,5,5-tetramethylpyrrolidine-1-oxyl)-p-(fluorosulfonyl) benzamidine, and meta-V, 3-(2,2,5,5-tetramethyl-pyrrolidine1-oxyl)-m-(fluorosulfonyl) benzamidine, have been completed at 2.0 and 3.0 A resolution, respectively. Previous electron spin resonance studies with these labels gave rise to a low-resolution "topography map" of thrombin's extended active site. These labels monitor two distinct areas of the thrombin active site: (1) an apolar binding site which manifests itself in an biphasic activation/inhibition effect on thrombin activity and (2) a region sensitive to alpha-thrombin autoproteolytic cleavage(s) to gamma-thrombin (Arg75-Tyr76 and/or Arg77A-Asn78, and Lys149E-Gly150, chymotrypsin numbering). Para-V was found to bind along the substrate binding cleft, while meta-V was found to bind both at the substrate primary specificity pocket and at a site which interacts with the gamma-cleavage loop. These studies reaffirm that accurate information may be gained from solution studies and indicates the complementarity of solid-state studies.

The use of chromogenic reference substrates for the kinetic analysis of penicillin acylases

Determination of kinetic parameters of penicillin acylases for phenylacetylated compounds is complicated due to the low K(m) values for these substrates, the lack of a spectroscopic signal, and the strong product inhibition by phenylacetic acid. To overcome these difficulties, a spectrophotometric method was developed, with which kinetic parameters could be determined by measuring the effects on the hydrolysis of the chromogenic reference substrate 2-nitro-5-[(phenylacetyl)amino]benzoic acid (NIPAB). To that end, spectrophotometric progress curves with NIPAB in the absence and presence of the phenylacetylated substrates and their products were measured and analyzed by numerical fitting to the appropriate equations for competing substrates with product inhibition. This analysis yielded kinetic constants for phenylacetylated substrates such as penicillin G, which are in close agreement with those obtained in independent initial velocity experiments. Using NIPAB analogs with lower k(cat)/K(m) values, kinetic parameters for the hydrolysis of cephalexin and penicillin V were determined. This method was suitable for determining the kinetic constants of penicillin acylases in periplasmic extracts from Escherichia coli, Alcaligenes faecalis, and Kluyvera citrophila. The use of chromogenic reference substrates thus appears to be a rapid and reliable method for determining kinetic constants with various substrates and enzymes.

Structural characterization of the interactions of optimized product inhibitors with the N-terminal proteinase domain of the hepatitis C virus (HCV) NS3 protein by NMR and modelling studies

The interactions of peptide inhibitors, obtained by the optimization of N-terminal cleavage products of natural substrates, with the protease of human hepatitis C virus (HCV) are characterized by NMR and modelling studies. The S-binding region of the enzyme and the bound conformation of the ligands are experimentally determined. The NMR data are then used as the experimental basis for modelling studies of the structure of the complex. The S-binding region involves the loop connecting strands E2 and F2, and appears shallow and solvent-exposed. The ligand binds in an extended conformation, forming an antiparallel beta-sheet with strand E2 of the protein, with the P1 carboxylate group in the oxyanion hole.

Studies on the C-terminal of hexapeptide inhibitors of the hepatitis C virus serine protease

Replacement of the C-terminal carboxylic acid functionality of peptide inhibitors of hepatitis C virus (HCV) NS3 protease (complexed with NS4A peptide cofactor) by activated carbonyl groups does not produce any substantial increase in potency. These latter inhibitors also inhibit a variety of other serine and cysteine proteases whereas the carboxylic acids are specific. Norvaline was identified as a chemically stable replacement for the P1 residue of Ac-DDIVPC-OH which was also compatible with activated carbonyl functionalities.

Structure-based drug design

Most of the techniques used in structure-based drug design have experienced significant improvements in the past few years, resulting in a marked enhancement of the speed and the efficacy of this approach. At the same time, it was thought that the future of drug design lay in strategies involving solely combinatorial chemistry. It is becoming evident, however, that the development of future drugs will use a combination of methods that will contain a major component of structure-based design.

Structural basis for chemical inhibition of human blood coagulation factor Xa

Factor Xa, the converting enzyme of prothrombin to thrombin, has emerged as an alternative (to thrombin) target for drug discovery for thromboembolic diseases. An inhibitor has been synthesized and the crystal structure of the complex between Des[1-44] factor Xa and the inhibitor has been determined by crystallographic methods in two different crystal forms to 2.3- and 2.4-A resolution. The racemic mixture of inhibitor FX-2212, (2RS)-(3'-amidino-3-biphenylyl)-5-(4-pyridylamino)pentanoic acid, inhibits factor Xa activity by 50% at 272 nM in vitro. The S-isomer of FX-2212 (FX-2212a) was found to bind to the active site of factor Xa in both crystal forms. The biphenylamidine of FX-2212a occupies the S1-pocket, and the pyridine ring makes hydrophobic interactions with the factor Xa aryl-binding site. Several water molecules meditate inhibitor binding to residues in the active site. In contrast to the earlier crystal structures of factor Xa, such as those of apo-Des[1-45] factor Xa and Des[1-44] factor Xa in complex with a naphthyl inhibitor DX-9065a, two epidermal growth factor-like domains of factor Xa are well ordered in both our crystal forms as well as the region between the two domains, which recently was found to be the binding site of the effector cell protease receptor-1. This structure provides a basis for designing next generation inhibitors of factor Xa.

Crystal structure of recombinant human tissue kallikrein at 2.0 A resolution

Human tissue kallikrein, a trypsin-like serine protease involved in blood pressure regulation and inflammation processes, was expressed in a deglycosylated form at high levels in Pichia pastoris, purified, and crystallized. The crystal structure at 2.0 A resolution is described and compared with that of porcine kallikrein and of other trypsin-like proteases. The active and S1 sites (nomenclature of Schechter I, Berger A, 1967, Biochem Biophys Res Commun 27:157-162) are similar to those of porcine kallikrein. Compared to trypsin, the S1 site is enlarged owing to the insertion of an additional residue, cis-Pro 219. The replacement Tyr 228 --> Ala further enlarges the S1 pocket. However, the replacement of Gly 226 in trypsin with Ser in human tissue kallikrein restricts accessibility of substrates and inhibitors to Asp 189 at the base of the S1 pocket; there is a hydrogen bond between O delta1Asp189 and O gammaSer226. These changes in the architecture of the S1 site perturb the binding of inhibitors or substrates from the modes determined or inferred for trypsin. The crystal structure gives insight into the structural differences responsible for changes in specificity in human tissue kallikrein compared with other trypsin-like proteases, and into the structural basis for the unusual specificity of human tissue kallikrein in cleaving both an Arg-Ser and a Met-Lys peptide bond in its natural protein substrate, kininogen. A Zn+2-dependent, small-molecule competitive inhibitor of kallikrein (Ki = 3.3 microM) has been identified and the bound structure modeled to guide drug design.

The mechanism of binding of low-molecular-weight active site inhibitors to human alpha-thrombin

The thrombin inhibitors argatroban, efegatran, NAPAP, CH 1091, CH 248, inogatran and melagatran have been characterised with respect to their mechanism of binding to human alpha-thrombin. Stopped-flow spectrophotometry was used to follow thrombin-catalysed hydrolysis of the chromogenic substrate S-2238 in the presence of inhibitors. The rate of onset or decay of inhibition was evaluated using progress curve analysis. It was possible to obtain apparent association and dissociation rate constants from the dependence of the rates on the inhibitor concentrations. Inhibition constants calculated from the association and dissociation rate constants were in good agreement with those calculated from steady-state rates. The binding of 6 inhibitors was also monitored directly using stopped-flow spectrofluorimetry when two kinetic components were found with all inhibitors. The faster component accounted for the largest part of the change in the intrinsic fluorescence of thrombin induced by inhibitor binding and was dependent on the inhibitor concentration. The slower component was independent of the concentration of the inhibitor. The concentration dependence of the faster component was linear with the compounds argatroban, NAPAP, CH 1091 and melagatran and hyperbolic with the compounds CH 248 and inogatran. The values of the apparent second-order rate constants at pH 7.4 and 37 degrees C range from slow to rapid binding in the interval 16-78 x 10(6) M-1 s-1, which is somewhat higher than 1-34 x 10(6)M-1 s-1 obtained from progress curve analysis of the onset of inhibition. The present results support a mechanism that includes rearrangement of a weak initial thrombin-inhibitor complex towards a tighter complex. Moreover, at least one additional step is required in the mechanism. In this model, the rate-limiting step for the binding of the inhibitor at concentrations in the nanomolar range depends on the primary interaction between the inhibitor and native thrombin.

Empirical scoring functions: I. The development of a fast empirical scoring function to estimate the binding affinity of ligands in receptor complexes

This paper describes the development of a simple empirical scoring function designed to estimate the free energy of binding for a protein-ligand complex when the 3D structure of the complex is known or can be approximated. The function uses simple contact terms to estimate lipophilic and metal-ligand binding contributions, a simple explicit form for hydrogen bonds and a term which penalises flexibility. The coefficients of each term are obtained using a regression based on 82 ligand-receptor complexes for which the binding affinity is known. The function reproduces the binding affinity of the complexes with a cross-validated error of 8.68 kJ/mol. Tests on internal consistency indicate that the coefficients obtained are stable to changes in the composition of the training set. The function is also tested on two test sets containing a further 20 and 10 complexes, respectively. The deficiencies of this type of function are discussed and it is compared to approaches by other workers.

Modulation of human alpha-thrombin activity with phosphonate ester inhibitors

Enantiomers of 4-nitrophenyl 4-X-phenacyl methylphosphonate esters (X = H, PMN; CH3; and CH3O) inactivate human alpha-thrombin with rate constants 4-235 M-1 s-1 in pH 6.5, 0.025 M citrate buffer, and 0.15 M NaCl at 7.0 +/- 0.1 degrees C. Stereoselectivity of the inactivation of thrombin is 2-39 and favors the levorotatory enantiomers. The pH-dependence of inactivation of thrombin by (-)-PMN is sigmoidal and consistent with the participation of a catalytic residue with a pKa of 8.0 +/- 0.1 in 0.15 M NaCl and a pKa of 7.4 +/- 0.2 in 0.15 M choline chloride in the nucleophilic attack of the catalytic Ser at phosphorus. The solvent isotope effect on ki/Ki in the pH-independent region of the reaction is 2.26 +/- 0.17. Thrombin activity returns from the adducts on the 2-7 h time scale at 25.0 +/- 0.1 degrees C via a self-catalyzed intramolecular reaction. The pH dependence of reactivation is significant from the adduct formed with (-)-CH3O-PMN and (-)-CH3-PMN and less so from the adducts formed with the other enantiomers of the inhibitors. Kinetic pKs approximately 7.2, with the exception of the adducts with (-)-PMN and (-)-CH3O-PMN, indicate that a pH-dependent conformational change affects the rate of dephosphonylation. A structural interpretation of the stereoselectivity and other mechanistic features is provided based on the energy-optimized structures of the adducts. Pharmaco-medical use of human alpha-thrombin covalently modified by the PMNs is suggested.