Analysis of inhibitor binding to the mitochondrial cytochrome c reductase by fluorescence quench titration. Evidence for a 'catalytic switch' at the Qo center

The binding characteristics of inhibitors of the mitochondrial cytochrome c reductase were studied by fluorescence quench titration. Based on the standard binding equation, the applied numerical method allowed the online recorded titration curves to be interpreted by fitting the Kd, the number of binding sites, and the specific fluorescence of the free and the bound inhibitor. For the Qi center, 2-n-nonyl-4-hydroxyquinoline N-oxide and for the Qo center (E)-beta-methoxyacrylate-stilbene (MOA-stilbene) were used as fluorescing inhibitors. The experiments could be extended to other, non-fluorescing inhibitors by competition analysis. Using this method we were able to compare the binding behaviour of Qi and Qo center inhibitors under different redox states of the enzyme using the same experimental set up. We studied the competition between inhibitors of the cytochrome c reductase representative for all subgroups and demonstrated that at least three inhibitor binding sites exist, two located in the Qo center, one located in the Qi center. Determination of the dissociation constants of the oxidized, the partially reduced and the fully reduced enzyme showed that inhibitor binding at the Qi center is not redox-dependent. In contrast, the binding of MOA-stilbene to the Qo center is decreased after reduction of the iron-sulfur center and cytochrome c1, whereas this redox change increases the affinity for a Qo center inhibitor of the hydroxynaphthoquinone type, 3-n-undecyl-2-hydroxynaphthoquinone. From these results, aware of the fact that the inhibitory mechanism at the Qo center is a non-competitive one, we made the hypothesis of a 'catalytic switch' to explain both the bifurcation of electron flow and the inhibition at the Qo center. A steric blockage of one of two conformational states could serve as a cogent explanation for the great structural variability of the inhibitors and differential effects on the redox centers exerted by the inhibitors. Moreover, the proposed 'switch' gives some insight into other experimental results which are difficult to explain with the ubiquinone cycle as currently formulated.     

Inhibitory analogs of ubiquinol act anti-cooperatively on the Yeast cytochrome bc1 complex. Evidence for an alternating, half-of-the-sites mechanism of ubiquinol oxidation.

The cytochrome bc(1) complex is a dimeric enzyme that links electron transfer from ubiquinol to cytochrome c by a protonmotive Q cycle mechanism in which ubiquinol is oxidized at one center in the enzyme, referred to as center P, and ubiquinone is re-reduced at a second center, referred to as center N. To understand better the mechanism of ubiquinol oxidation, we have examined the interaction of several inhibitory analogs of ubiquinol with the yeast cytochrome bc(1) complex. Stigmatellin and methoxyacrylate stilbene, two inhibitors that block ubiquinol oxidation at center P, inhibit the yeast enzyme with a stoichiometry of 0.5 per bc(1) complex, indicating that one molecule of inhibitor is sufficient to fully inhibit the dimeric enzyme. This stoichiometry was obtained when the inhibitors were titrated in cytochrome c reductase assays and in reactions of quinol with enzyme in which the inhibitors block pre-steady state reduction of cytochrome b. As an independent measure of inhibitor binding, we titrated the red shift in the optical spectrum of ferrocytochrome b with methoxyacrylate stilbene and thus confirmed the results of the inhibition of activity titrations. The titration curves also indicate that the binding is anti-cooperative, in that a second molecule of inhibitor binds with much lower affinity to a dimer in which an inhibitor molecule is already bound. Because these inhibitors bind to the ubiquinol oxidation site in the bc(1) complex, we propose that the yeast cytochrome bc(1) complex oxidizes ubiquinol by an alternating, half-of-the-sites mechanism.     

Binding of novel inhibitors of electron transfer in photosystem 2, derivatives of perfluoroisopropyldinitrobenzene,with polypeptide D2 of the reaction center

A binding site for novel inhibitors of K15 type (derivatives of perfluoroisopropyldinitrobenzene) with the components of reaction center (RC) of photosystem 2 (PS-2) of higher plants has been investigated. It has been shown that multiple washing the PS-2 submembrane chloroplast fragments (BBY-particles) treated with the K15 inhibitor, including multiple dilution in buffer in the presence of high concentrations of mono- and divalent ions, prolonged (up to 2-5 h) incubation, centrifugation, and subsequent resuspension in buffer deprived of the inhibitor, does not lead to restoration of functional activity of the PS-2. After addition of dithionite, inducing reduction and consequent decomposition of the inhibitor, and subsequent removal of dithionite by washing, the functional activity of PS-2 was completely restored. Incubation in the presence of sodium dodecyl sulfate (SDS), leading to solubilization of the sample to the level of protein components, induced the appearance of a fraction of free K15 retaining the initial inhibitory efficiency. To create a covalent binding of the inhibitor with protein, retained under the conditions of denaturing SDS polyacrylamide gel electrophoresis, the azido-containing analog of K15 (K15-N₃) was used. The need for radioactive label for identification of K15 was avoided by the revealed ability of K15-type inhibitors to emit fluorescence, which retained its features under the experimental conditions. With the technique of photoaffinity binding and denaturing SDS-PAGE in the presence of 6 M urea of submembrane chloroplast fragments enriched in PS-2 the D2-polypeptide, an integral component of the reaction center of PS-2, has been shown to be a binding site for K15-type inhibitors. This conclusion is in agreement with a suggestion (put forward in our earlier publications) that K15-type inhibitors are bound to PS-2 reaction center, replacing Q_A in its binding site. Hence, an agent specifically binding to polypeptide D2 has been found for the first time. The data are compared with information about inhibitory action and binding sites of the known inhibitors of electron transfer in PS-2.     

Interaction of stigmatellin and DNP-INT with the Rieske iron-sulfur center of the chloroplast cytochrome b6-f complex

Stigmatellin and DNP-INT are effective inhibitors of the catalytic activity of the plastoquinol-plastocyanin oxidoreductase complex (cytochrome b6-f complex). Both inhibitors alter the EPR spectrum of the Rieske iron-sulfur center but do not produce band-shifts of cytochrome b-563. The midpoint redox potential of the Rieske center is unaffected by either inhibitor, although both alter the DBMIB-induced g-value shifts of the Rieske center. The results are considered in terms of binding domains for inhibitors in the cytochrome b6-f complex.     

Determinants of the inhibition of a Taiwan habu venom metalloproteinase by its endogenous inhibitors revealed by X-ray crystallography and synthetic inhibitor analogues.

Venoms from crotalid and viperid snakes contain several peptide inhibitors which regulate the proteolytic activities of their snake-venom metalloproteinases (SVMPs) in a reversible manner under physiological conditions. In this report, we describe the high-resolution crystal structures of a SVMP, TM-3, from Taiwan habu (Trimeresurus mucrosquamatus) cocrystallized with the endogenous inhibitors pyroGlu-Asn-Trp (pENW), pyroGlu-Gln-Trp (pEQW) or pyroGlu-Lys-Trp (pEKW). The binding of inhibitors causes some of the residues around the inhibitor-binding environment of TM-3 to slightly move away from the active-site center, and displaces two metal-coordinated water molecules by the C-terminal carboxylic group of the inhibitors. This binding adopts a retro-manner principally stabilized by four possible hydrogen bonds. The Trp indole ring of the inhibitors is stacked against the imidazole of His143 in the S-1 site of the proteinase. Results from the study of synthetic inhibitor analogues showed the primary specificity of Trp residue of the inhibitors at the P-1 site, corroborating the stacking effect observed in our structures. Furthermore, we have made a detailed comparison of our structures with the binding modes of other inhibitors including batimastat, a hydroxamate inhibitor, and a barbiturate derivative. It suggests a close correlation between the inhibitory activity of an inhibitor and its ability to fill the S-1 pocket of the proteinase. Our work may provide insights into the rational design of small molecules that bind to this class of zinc-metalloproteinases.     

Significance of the "Rieske" iron-sulfur protein for formation and function of the ubiquinol-oxidation pocket of mitochondrial cytochrome c reductase (bc1 complex).

The binding of specific inhibitors to the ubiquinol oxidation pocket ("QP center") of cytochrome c reductase was analyzed before and after removal of bound phospholipid and the "Rieske" iron-sulfur protein using optical spectroscopy and fluorescence quench binding assays. The enzyme lacking iron-sulfur protein showed almost unchanged, tight binding of the E-beta-methoxyacrylate inhibitors oudemansin A and MOA-stilbene, whereas binding of the chromone inhibitor stigmatellin was almost completely abolished. The affinity of the weak inhibitor 3-undecyl-2-hydroxy-naphthoquinone was decreased. Oudemansin A binding to the defective pocket of the iron-sulfur protein-depleted enzyme was lowered by added phospholipid. It was deduced from these results that the QP center is a spacious pocket formed by domains of cytochrome b, bearing the E-beta-methoxcyacrylate binding site, and the iron-sulfur protein, bearing the stigmatellin binding site. Moreover, removal of the iron-sulfur protein leaves this pocket defective but essentially unchanged in its remaining binding capability. The affinity of three preparations of cytochrome c reductase, the complete, the delipidated, and the iron-sulfur depleted enzyme for E-beta-methoxyacrylate-stilbene, was analyzed for different redox states of the catalytic centers of cytochrome c reductase. The apparent Kd values for the different redox states were interpreted in terms of two conformational states. It is suggested that these changes reflect the two states of the "catalytic switch" proposed recently for the QP pocket of cytochrome c reductase (Brandt, U., and von Jagow, G. (1991) Eur. J. Biochem. 195, 163-170). According to the refined model presented in this work, changeover to the "b" state is triggered by reduction of the iron-sulfur cluster, and changeover back to the "FeS" state is triggered by electron transfer from the low potential onto the high potential heme b center. Our interpretation implies that the stability of the two states is affected by the redox states of the enzyme, but that additionally changing the redox states of the two centers is required for "switching" on a catalytic time scale.     

Mutational analysis of the mouse mitochondrial cytochrome b gene

The protonmotive cytochrome b protein of the mitochondrial bc1 respiratory chain complex contains two reactions centers, designated Qo and Qi, which can be distinguished by the effects of different inhibitors. The nucleotide sequences have been determined of the mitochondrial cytochrome b genes from a series of mouse cell mutants selected for increased inhibitor resistance. Each mutant contains a single nucleotide change which results in an amino acid substitution. When the proximity of the altered amino acid residues to the histidines involved in heme ligation is considered, the results support a model for cytochrome b folding in which there are eight transmembrane domains rather than the nine of the Widger-Saraste model. Replacement of the Gly38 residue by valine results in resistance to the Qi inhibitors antimycin A and funiculosin but not 2-n-heptyl-hydroxyquinoline-N-oxide. Based upon sequence comparisons of mitochondrial and bacterial cytochrome b and chloroplast b6 proteins, the region of the molecule involved in antimycin binding is as highly conserved as those domains involved in heme ligation. It is suggested that the antimycin binding domain of cytochrome b is involved in forming the Qi reaction center. Alterations of the Gly142 and Thr147 residues result in resistance to myxothiazol and stimatellin, respectively. While both inhibitors block the Qo reaction center, the two mutations do not confer cross-resistance to each other. This region of cytochrome b is the most highly conserved during evolution and these inhibitor binding sites probably occur within the protein domain constituting the Qo reaction center. In addition, there is a less conserved region of the protein, defined by the Leu294 residue, which may function in binding the hydrophobic portions of Qo inhibitors.     

Concanamycin and indolyl pentadieneamide inhibitors of the vacuolar H+-ATPase bind with high affinity to the purified proteolipid subunit of the membrane domain

The macrolide antibiotic concanamycin is a potent and specific inhibitor of the vacuolar H(+)-ATPase (V-ATPase), binding to the V(0) membrane domain of this eukaryotic acid pump. Although binding is known to involve the 16 kDa proteolipid subunit, contributions from other V(0) subunits are possible that could account for the apparently different inhibitor sensitivities of pump isoforms in vertebrate cells. In this study, we used a fluorescence quenching assay to directly examine the roles of V(0) subunits in inhibitor binding. Pyrene-labeled V(0) domains were affinity purified from Saccharomyces vacuolar membranes, and the 16 kDa proteolipid was subsequently extracted into chloroform and methanol and purified by size exclusion chromatography. Fluorescence from the isolated proteins was strongly quenched by nanomolar concentrations of both concanamycin and an indolyl pentadieneamide compound, indicating high-affinity binding of both natural macrolide and synthetic inhibitors. Competition studies showed that these inhibitors bind to overlapping sites on the proteolipid. Significantly, the 16 kDa proteolipid in isolation was able to bind inhibitors as strongly as V(0) did. In contrast, proteolipids carrying mutations that confer resistance to both inhibitors showed no binding. We conclude that the extracted 16 kDa proteolipid retains sufficient fold to form a high-affinity inhibitor binding site for both natural and synthetic V-ATPase inhibitors and that the proteolipid contains the major proportion of the structural determinants for inhibitor binding. The role of membrane domain subunit a in concanamycin binding and therefore in defining the inhibitor binding properties of tissue-specific V-ATPases is critically re-assessed in light of these data.     

Effect of pH on substrate and inhibitor kinetic constants of human liver alanine aminopeptidase. Evidence for two ionizable active center groups.

The presence of at least two ionizable active center groups has been detected by a study of the effect of pH upon catalysis of hydrolysis of L-alanyl-beta-naphthylamide by human liver alanine aminopeptidase and upon the inhibition of hydrolysis by inhibitors and substrate analogs. Octanoic acid, octylamine, and peptide inhibitors have been found to be competitive inhibitors and are therefore thought to bind the active center. L-Phe was previously shown to bind the active center since it was found to be a competitive inhibitor of the hydrolysis of tripeptide substrates (Garner, C. W., and Behal, F. J. (1975), Biochemistry 14, 3208). A plot of pKm vs. pH for the substrate L-Ala-beta-naphthylamide showed that binding decreased below pH 5.9 and above 7.5, the points at which the theoretical curve undergoes an integral change in slope. These points are interpreted as the pKa either of substrate ionizable groups or binding-dependent enzyme active center groups. Similar plots of pKm vs. pH for L-alanyl-p-nitroanilide (as substrate) and pKi vs. pH for L-Leu-L-Leu-L-Leu and D-Leu-L-Tyr (as inhibitors) gave pairs fo pKa values of 5.8 and 7.4, 6.0 and 7.5, and 5.7 and 7.5, respectively. All the above substrates (and D-Leu-L-Tyr) have pKa values near 7.5; therefore, the binding-dependent group with a pKa value near 7.5 is possibly this substrate group. Similar plots of pKi vs. pH for the inhibitors L-Phe, L-Met, L-Leu, octylamine, and octanoic acid had only one bending point at 7.7, 7.6, 7.4, 6.3, and 5.9, respectively. Amino acid inhibitors, octylamine, and octanoic acid have no groups with pKa values between 5 and 9. These data indicate that there are two active center ionizable groups with pKa values of approximately 6.0 and 7.5 which are involved in substrate binding or inhibitory amino acid binding but not in catalysis since Vmax was constant at all pH values tested.     

Investigating the Qn site of the cytochrome bc1 complex in Saccharomyces cerevisiae with mutants resistant to ilicicolin H, a novel Qn site inhibitor

The cytochrome bc1 complex resides in the inner membrane of mitochondria and transfers electrons from ubiquinol to cytochrome c. This electron transfer is coupled to the translocation of protons across the membrane by the protonmotive Q cycle mechanism. This mechanism topographically separates reduction of quinone and reoxidation of quinol at sites on opposite sites of the membrane, referred to as center N (Qn site) and center P (Qp site), respectively. Both are located on cytochrome b, a transmembrane protein of the bc1 complex that is encoded on the mitochondrial genome. To better understand the parameters that affect ligand binding at the Qn site, we applied the Qn site inhibitor ilicicolin H to select for mutations conferring resistance in Saccharomyces cerevisiae. The screen resulted in seven different single amino acid substitutions in cytochrome b rendering the yeast resistant to the inhibitor. Six of the seven mutations have not been previously linked to inhibitor resistance. Ubiquinol-cytochrome c reductase activities of mitochondrial membranes isolated from the mutants confirmed that the differences in sensitivity toward ilicicolin H originated in the cytochrome bc1 complex. Comparative in vivo studies using the known Qn site inhibitors antimycin and funiculosin showed little cross-resistance, indicating different modes of binding of these inhibitors at center N of the bc1 complex.     

Double-headed protease inhibitors from black-eyed peas. VI. Singlet-singlet energy transfer and other optical studies on the structure of trypsin and chymotrypsin complexes.

The geometry of the binary and ternary complexes of two black-eyed pea inhibitors with trypsin and chymotrypsin has been established by distance measurements using the technique of singlet-singlet energy transfer. Triangulation of measured distances in the ternary double-headed complex of the trypsin-chymotrypsin inhibitor (BEPCI) with trypsin and chymotrypsin limits the possible structural models for this complex to those in which the center to center distance between trypsin and chymotrypsin is about 64 A, the distance from the center of trypsin to the single fluorescently labeled tyrosyl residue of the BEPCI dimer is about 33 A, and the distance between the chymotrypsin center and the labeled tyrosine of the inhibitor is about 43 A. Energy transfer results for the trypsin inhibitor (BEPTI) complexes show conclusively that the weak trypsin site is structurally analogous to the strong chymotrypsin binding site of BEPCI. The weak chymotrypsin binding site of BEPTI is structurally analogous to the strong trypsin sites of BEPCI and BEPTI. Corresponding distances in binary and ternary complexes are the same, indicating that little or no structural rearrangement occurs when the ternary complexes are formed. Complex formation was shown to involve tryptophan and tryosine residues of both trypsin and chymotrypsin as judged by absorption and circular dichroism difference spectroscopy. In addition, circular dichroism difference spectra revealed some disulfide contributions.     

Insights into the ubiquinol/dioxygen binding and proton relay pathways of the alternative oxidase

The alternative oxidase (AOX) is a monotopic diiron carboxylate protein which catalyzes the four-electron reduction of dioxygen to water by ubiquinol. Although we have recently determined the crystal structure of Trypanosoma brucei AOX (TAO) in the presence and absence of ascofuranone (AF) derivatives (which are potent mixed type inhibitors) the mechanism by which ubiquinol and dioxygen binds to TAO remain inconclusive. In this article, ferulenol was identified as the first competitive inhibitor of AOX which has been used to probe the binding of ubiquinol. Surface plasmon resonance reveals that AF is a quasi-irreversible inhibitor of TAO whilst ferulenol binding is completely reversible. The structure of the TAO-ferulenol complex, determined at 2.7?Å, provided insights into ubiquinol binding and has also identified a potential dioxygen molecule bound in a side-on conformation to the diiron center for the first time.     

Differential efficacy of inhibition of mitochondrial and bacterial cytochrome bc1 complexes by center N inhibitors antimycin, ilicicolin H and funiculosin

We have compared the efficacy of inhibition of the cytochrome bc1 complexes from yeast and bovine heart mitochondria and Paracoccus denitrificans by antimycin, ilicicolin H, and funiculosin, three inhibitors that act at the quinone reduction site at center N of the enzyme. Although the three inhibitors have some structural features in common, they differ significantly in their patterns of inhibition. Also, while the overall folding pattern of cytochrome b around center N is similar in the enzymes from the three species, amino acid sequence differences create sufficient structural differences so that there are striking differences in the inhibitors binding to the three enzymes. Antimycin is the most tightly bound of the three inhibitors, and binds stoichiometrically to the isolated enzymes from all three species under the cytochrome c reductase assay conditions. Ilicicolin H also binds stoichiometrically to the yeast enzyme, but binds approximately 2 orders of magnitude less tightly to the bovine enzyme and is essentially non-inhibitory to the Paracoccus enzyme. Funiculosin on the other hand inhibits the yeast and bovine enzymes similarly, with IC50 approximately 10 nM, while the IC50 for the Paracoccus enzyme is more than 10-fold higher. Similar differences in inhibitor efficacy were noted in bc1 complexes from yeast mutants with single amino acid substitutions at the center N site, although the binding affinity of quinone and quinol substrates were not perturbed to a degree that impaired catalytic function in the variant enzymes. These results reveal a high degree of specificity in the determinants of ligand-binding at center N, accompanied by sufficient structural plasticity for substrate binding as to not compromise center N function. The results also demonstrate that, in principle, it should be possible to design novel inhibitors targeted toward center N of the bc1 complex with appropriate species selectivity to allow their use as drugs against pathogenic fungi and parasites.     

Crystal structures of cystathionine gamma-synthase inhibitor complexes rationalize the increased affinity of a novel inhibitor

Cystathionine gamma-synthase catalyzes the committed step of methionine biosynthesis. This pathway is unique to microorganisms and plants, rendering the enzyme an attractive target for the development of antimicrobials and herbicides. We solved the crystal structures of complexes of cystathionine gamma-synthase (CGS) from Nicotiana tabacum with inhibitors of different compound classes. The complex with the substrate analog dl-E-2-amino-5-phosphono-3-pentenoic acid verifies the carboxylate-binding function of Arg423 and identifies the phosphate-binding pocket of the active site. The structure shows the function of Lys165 in specificity determination and suggests a role for the flexible side-chain of Tyr163 in catalysis. The importance of hydrophobic interactions for binding to the active-site center is highlighted by the complex with 3-(phosphonomethyl)pyridine-2-carboxylic acid. The low affinity of this compound is due to the non-optimal arrangement of the functional groups binding to the phosphate and carboxylate-recognition site, respectively. The newly identified inhibitor 5-carboxymethylthio-3-(3'-chlorophenyl)-1,2,4-oxadiazol, in contrast, shows the highest affinity to CGS reported so far. This affinity is due to binding to an additional active-site pocket not used by the physiological substrates. The inhibitor binds to the carboxylate-recognition site, and its tightly bent conformation enables it to occupy the novel binding pocket between Arg423 and Ser388. The described structures suggest improvements for known inhibitors and give guidelines for the development of new lead compounds.     

Hydroxyurea is a carbonic anhydrase inhibitor

The interaction of hydroxyurea with the cytosolic isozymes of carbonic anhydrase (CA), hCA I and hCA II has been investigated by means of kinetic and spectroscopic techniques. Hydroxyurea acts as a weak, non-competitive inhibitor of both isozymes, for the 4-nitrophenyl acetate esterase activity, with inhibition constants around 0.1 mM for both isozymes. The spectrum of the adduct of hydroxyurea with Co(II)-hCA II is similar to the spectra of tetrahedral adducts (such as those with sulfamide, acetazolamide or cyanamide), proving a direct interaction of the inhibitor molecule with the metal center of the enzyme, whose geometry remains tetrahedral. Based on the X-ray crystal structure of the adducts of hCA II with ureate and hydroxamate inhibitors, the hypothetical binding of hydroxyurea is proposed to be achieved in deprotonated state, with the nitrogen atom coordinated to Zn(II), and the OH group of the inhibitor making a hydrogen bond with Thr 199. This binding may be exploited for the design of both CA as well as matrix metalloproteinase (MMP) inhibitors, since hydroxyurea is the simplest compound incorporating a hydroxamate functionality in its molecule. Indeed, such inhibitors of the sulfonylated amino acid hydroxamate type have been generated, with potencies in the low nanomolar range for both type of enzymes, CAs and MMPs.     

Probing the location of the substrate binding site of ascorbate oxidase near type 1 copper: an investigation through spectroscopic, inhibition and docking studies

The present investigation addresses the problem of the binding mode of phenolic inhibitors and the substrate ascorbate to the active site of ascorbate oxidase. The results from both types of compounds indicate that the binding site is located in a pocket near the type 1 copper center. This information is of general interest for blue multicopper oxidases. Docking calculations performed on the ascorbate oxidase-ascorbate complex show that binding of the substrate occurs in a pocket near type 1 Cu, and is stabilized by at least five hydrogen bonding interactions with protein residues, one of which involves the His512 Cu ligand. Similar docking studies show that the isomeric fluorophenols, which act as competitive inhibitors toward ascorbate, bind to the enzyme in a manner similar to ascorbate. The docking calculations are supported by 19F NMR relaxation measurements performed on fluorophenols in the presence of the enzyme, which show that the bound inhibitors undergo enhanced relaxation by the paramagnetic effect of a nearby Cu center. Unambiguous support to the location of the inhibitor close to type 1 Cu was obtained by comparative relaxation measurements of the fluorophenols in the presence of the ascorbate oxidase derivative where a Zn atom selectively replaces the paramagnetic type 2 Cu. The latter experiments show that contribution to relaxation of the bound inhibitors by the type 2 Cu site is negligible.     

Human soluble epoxide hydrolase: structural basis of inhibition by 4-(3-cyclohexylureido)-carboxylic acids

X-ray crystal structures of human soluble epoxide hydrolase (sEH) complexed with four different dialkylurea inhibitors bearing pendant carboxylate "tails" of varying length have been determined at 2.3-3.0 A resolution. Similarities among inhibitor binding modes reinforce the proposed roles of Y381 and/or Y465 as general acids that protonate the epoxide ring of the substrate in concert with nucleophilic attack of D333 at the electrophilic epoxide carbon. Additionally, the binding of these inhibitors allows us to model the binding mode of the endogenous substrate 14,15-epoxyeicosatrienoic acid. Contrasts among inhibitor binding modes include opposite orientations of inhibitor binding in the active-site hydrophobic tunnel. Alternative binding orientations observed for this series of inhibitors to human sEH, as well as the binding of certain dialkylurea inhibitors to human sEH and murine sEH, complicate the structure-based design of human sEH inhibitors with potential pharmaceutical applications in the treatment of hypertension. Thus, with regard to the optimization of inhibitor designs targeting human sEH, it is critical that human sEH and not murine sEH be utilized for inhibitor screening, and it is critical that structures of human sEH-inhibitor complexes be determined to verify inhibitor binding orientations that correlate with measured affinities.     

Prediction of inhibitor binding free energies by quantum neural networks. Nucleoside analogues binding to trypanosomal nucleoside hydrolase

A computational method has been developed to predict inhibitor binding energy for untested inhibitor molecules. A neural network is trained from the electrostatic potential surfaces of known inhibitors and their binding energies. The algorithm is then able to predict, with high accuracy, the binding energy of unknown inhibitors. IU-nucleoside hydrolase from Crithidia fasciculata and the inhibitor molecules described previously [Miles, R. W. Tyler, P. C. Evans, G. Furneaux R. H., Parkin, D. W., and Schramm, V. L. (1999) Biochemistry 38, xxxx-xxxx] are used as the test system. Discrete points on the molecular electrostatic potential surface of inhibitor molecules are input to neural networks to identify the quantum mechanical features that contribute to binding. Feed-forward neural networks with back-propagation of error are trained to recognize the quantum mechanical electrostatic potential and geometry at the entire van der Waals surface of a group of training molecules and to predict the strength of interactions between the enzyme and novel inhibitors. The binding energies of unknown inhibitors were predicted, followed by experimental determination of K(i)() values. Predictions of K(i)() values using this theory are compared to other methods and are more robust in estimating inhibitory strength. The average deviation in estimating K(i)() values for 18 unknown inhibitor molecules, with 21 training molecules, is a factor of 5 x K(i)() over a range of 660 000 in K(i)() values for all molecules. The a posteriori accuracy of the predictions suggests the method will be effective as a guide for experimental inhibitor design.     

Crystallographic analysis at 3.0-A resolution of the binding to human thrombin of four active site-directed inhibitors

The mode of binding of four active-site directed inhibitors to human thrombin has been determined by x-ray crystallographic analysis. The inhibitors studied are benzamidine, PPACK, NAPAP, and MD-805, of which the last three are compounds evolved specifically to inhibit thrombin. Crystal structures were determined in the presence of both the inhibitor and the undecapeptide [des-amino Asp55]hirudin(55-65) which binds distant from the active site. Despite having significantly different chemical structures, NAPAP and MD-805 bind to thrombin in a very similar "inhibitor binding mode" which is not that expected by direct analogy with the binding of substrate. Both inhibitors bind to thrombin in a similar way as to trypsin, but thrombin has an extra loop, the "Tyr-Pro-Pro-Trp loop," not present in trypsin, which gives further binding interactions and is seen to move somewhat to accommodate binding of the different inhibitors. The fact that NAPAP and MD-805 require different stereochemistry for potent inhibition is demonstrated, and its structural basis clarified. The wealth of data on analogs and variants of these lead compounds is shown to be compatible with this inhibitor binding mode.