Computer-aided lead optimization: improved small-molecule inhibitor of the zinc endopeptidase of botulinum neurotoxin serotype A

Optimization of a serotype-selective, small-molecule inhibitor of botulinum neurotoxin serotype A (BoNTA) endopeptidase is a formidable challenge because the enzyme-substrate interface is unusually large and the endopeptidase itself is a large, zinc-binding protein with a complex fold that is difficult to simulate computationally. We conducted multiple molecular dynamics simulations of the endopeptidase in complex with a previously described inhibitor (K(i) (app) of 7+/-2.4 microM) using the cationic dummy atom approach. Based on our computational results, we hypothesized that introducing a hydroxyl group to the inhibitor could improve its potency. Synthesis and testing of the hydroxyl-containing analog as a BoNTA endopeptidase inhibitor showed a twofold improvement in inhibitory potency (K(i) (app) of 3.8+/-0.8 microM) with a relatively small increase in molecular weight (16 Da). The results offer an improved template for further optimization of BoNTA endopeptidase inhibitors and demonstrate the effectiveness of the cationic dummy atom approach in the design and optimization of zinc protease inhibitors.     

Successful molecular dynamics simulation of two zinc complexes bridged by a hydroxide in phosphotriesterase using the cationic dummy atom method

I report herein two 2.0 ns (1.0 fs time step) MD simulations of two zinc complexes bridged by a hydroxide in phosphotriesterase (PTE) employing the nonbonded method and the cationic dummy atom method that uses virtual atoms to impose orientational requirement for zinc ligands. The cationic dummy atom method was able to simulate the four-ligand coordination of the two zinc complexes in PTE. The distance (3.39 +/- 0.07A) between two nearby zinc ions in the time-average structure of PTE derived from the MD simulation using the cationic dummy atoms matched that in the X-ray structure (3.31 +/- 0.001A). Unequivocally, the time-average structure of PTE was able to fit into the experimentally determined difference electron density map of the corresponding X-ray structure. The results demonstrate the practicality of the cationic dummy atom method for MD simulations of zinc proteins bound with multiple zinc ions. In contrast, a 2.0 ns (1.0 fs time step) MD simulation using the nonbonded method revealed a striking difference in the active site between the X-ray structure and the time-average structure that was unable to fit into the density map of PTE. The results suggest that caution should be used in the MD simulations using the nonbonded method.     

Magnesium-cationic dummy atom molecules enhance representation of DNA polymerase beta in molecular dynamics simulations: improved accuracy in studies of structural features and mutational effects

Human DNA polymerase beta (pol beta) fills gaps in DNA as part of base excision DNA repair. Due to its small size it is a convenient model enzyme for other DNA polymerases. Its active site contains two Mg(2+) ions, of which one binds an incoming dNTP and one catalyzes its condensation with the DNA primer strand. Simulating such binuclear metalloenzymes accurately but computationally efficiently is a challenging task. Here, we present a magnesium-cationic dummy atom approach that can easily be implemented in molecular mechanical force fields such as the ENZYMIX or the AMBER force fields. All properties investigated here, namely, structure and energetics of both Michaelis complexes and transition state (TS) complexes were represented more accurately using the magnesium-cationic dummy atom model than using the traditional one-atom representation for Mg(2+) ions. The improved agreement between calculated free energies of binding of TS models to different pol beta variants and the experimentally determined activation free energies indicates that this model will be useful in studying mutational effects on catalytic efficiency and fidelity of DNA polymerases. The model should also have broad applicability to the modeling of other magnesium-containing proteins.     

Serotype-selective, small-molecule inhibitors of the zinc endopeptidase of botulinum neurotoxin serotype A

Botulinum neurotoxin serotype A (BoNTA) is one of the most toxic substances known. Currently, there is no antidote to BoNTA. Small molecules identified from high-throughput screening reportedly inhibit the endopeptidase--the zinc-bound, catalytic domain of BoNTA--at a drug concentration of 20 microM. However, optimization of these inhibitors is hampered by challenges including the computational evaluation of the ability of a zinc ligand to compete for coordination with nearby residues in the active site of BoNTA. No improved inhibitor of the endopeptidase has been reported. This article reports the development of a serotype-selective, small-molecule inhibitor of BoNTA with a K(i) of 12 microM. This inhibitor was designed to coordinate the zinc ion embedded in the active site of the enzyme for affinity and to interact with a species-specific residue in the active site for selectivity. It is the most potent small-molecule inhibitor of BoNTA reported to date. The results suggest that multiple molecular dynamics simulations using the cationic dummy atom approach are useful to structure-based design of zinc protease inhibitors.     

Kinetic mechanism of the Zn-dependent aryl-phosphatase activity of myo-inositol-1-phosphatase

Myo-inositol-1-phosphatase (EC 3.1.3.25) is able to hydrolyze myo-inositol-1-phosphate in the presence of Mg(2+) ions at neutral pH, and also p-nitrophenyl phosphate in the presence of Zn(2+)-ions at acidic pH. This enzyme plays a role in phosphatidylinositol cell signalling and is a putative target of lithium therapy in manic depression. We elucidate here the kinetic mechanism of the Zn-dependent activity of myo-inositol-1-phosphatase. As part of this analysis it was necessary to determine the basicity constants of p-nitrophenyl phosphate and the stability constant of its metal-complex in the presence of zinc chloride. We find that the Zn-dependent reaction may be described either by a rapid-equilibrium random mechanism or an ordered steady-state mechanism in which the substrate binds to the free enzyme prior to the metal ion. In both models the Zn-substrate complex acts as a high affinity inhibitor, yielding a dead-end species through its binding to the enzyme-Zn-substrate in rapid-equilibrium or to the enzyme-phosphate complexes in a steady-state model. Phosphate is a competitive inhibitor of the enzyme with respect to the substrate and an uncompetitive inhibitor with respect to zinc ions.     

Modeling domino effects in enzymes: molecular basis of the substrate specificity of the bacterial metallo-beta-lactamases IMP-1 and IMP-6

Metallo-beta-lactamases can hydrolyze a broad spectrum of beta-lactam antibiotics and thus confer resistance to bacteria. For the Pseudomonas aeruginosa enzyme IMP-1, several variants have been reported. IMP-6 and IMP-1 differ by a single residue (glycine and serine at position 196, respectively), but have significantly different substrate spectra; while the catalytic efficiency toward the two cephalosporins cephalothin and cefotaxime is similar for both variants, IMP-1 is up to 10-fold more efficient than IMP-6 toward cephaloridine and ceftazidime. Interestingly, this biochemical effect is caused by a residue remote from the active site. The substrate-specific impact of residue 196 was studied by molecular dynamics simulations using a cationic dummy atom approach for the zinc ions. Substrates were docked in an intermediate structure near the transition state to the binding site of IMP-1 and IMP-6. At a simulation temperature of 100 K, most complexes were stable during 1 ns of simulation time. However, at higher temperatures, some complexes became unstable and the substrate changed to a nonactive conformation. To model stability, six molecular dynamics simulations at 100 K were carried out for all enzyme-substrate complexes. Stable structures were further heated to 200 and 300 K. By counting stable structures, we derived a stability ranking score which correlated with experimentally determined catalytic efficiency. The use of a stability score as an indicator of catalytic efficiency of metalloenzymes is novel, and the study of substrates in a near-transition state intermediate structure is superior to the modeling of Michaelis complexes. The remote effect of residue 196 can be described by a domino effect: upon replacement of serine with glycine, a hole is created and a stabilizing interaction between Ser196 and Lys33 disappears, rendering the neighboring residues more flexible; this increased flexibility is then transferred to the active site.     

Coordinate repression of cholesterol biosynthesis and cytoplasmic 3-hydroxy-3-methylglutaryl coenzyme A synthase by glucocorticoids in HeLa cells.

The stability constants for the calcium and magnesium complexes of rhodanese are >10⁵m^?1 at both high and low substrate concentrations. The stoichiometry of alkaline earth metal ion binding totals close to 1 per 18,500 molecular weight. The usual assay reagents contain sufficient amounts of these metal ions to maintain added enzyme in its metal-complexed form. When reaction mixtures are treated with oxalate to remove calcium ions, inhibition of rhodanese activity is virtually complete under circumstances such that the contribution of magnesium ion is low.Zinc and a number of transition metal ions are inhibitors of rhodanese activity. Studies of the concentration dependence of these effects with zinc, copper, and nickel showed that: 1) Some cyanide complexes of these metals are competitive with the donor substrate, thiosulfate ion. The binding of the copper and zinc complexes is mutually competitive. 2) Another cyanide species of copper appears to combine with the free enzyme to form a functionally active complex. 3) The zinc cyanide species with a net positive charge is an inhibitor competitive with the acceptor substrate, cyanide ion.All of these observations are consistent with a model in which metal ions serve as the electrophilic site of rhodanese.     

Structure of 3,4-dihydroxy-2-butanone 4-phosphate synthase from Methanococcus jannaschii in complex with divalent metal ions and the substrate ribulose 5-phosphate: implications for the catalytic mechanism

Skeletal rearrangements of carbohydrates are crucial for many biosynthetic pathways. In riboflavin biosynthesis ribulose 5-phosphate is converted into 3,4-dihydroxy-2-butanone 4-phosphate while its C4 atom is released as formate in a sequence of metal-dependent reactions. Here, we present the crystal structure of Methanococcus jannaschii 3,4-dihydroxy-2-butanone 4-phosphate synthase in complex with the substrate ribulose 5-phosphate at a dimetal center presumably consisting of non-catalytic zinc and calcium ions at 1.7-A resolution. The carbonyl group (O2) and two out of three free hydroxyl groups (OH3 and OH4) of the substrate are metal-coordinated. We correlate previous mutational studies on this enzyme with the present structural results. Residues of the first coordination sphere involved in metal binding are indispensable for catalytic activity. Only Glu-185 of the second coordination sphere cannot be replaced without complete loss of activity. It contacts the C3 hydrogen atom directly and probably initiates enediol formation in concert with both metal ions to start the reaction sequence. Mechanistic similarities to Rubisco acting on the similar substrate ribulose 1,5-diphosphate in carbon dioxide fixation as well as other carbohydrate (reducto-) isomerases are discussed.     

Successful molecular dynamics simulation of the zinc-bound farnesyltransferase using the cationic dummy atom approach

Farnesyltransferase (FT) inhibitors can suppress tumor cell proliferation without substantially interfering with normal cell growth, thus holding promise for cancer treatment. A structure-based approach to the design of improved FT inhibitors relies on knowledge of the conformational flexibility of the zinc-containing active site of FT. Although several X-ray structures of FT have been reported, detailed information regarding the active site conformational flexibility of the enzyme is still not available. Molecular dynamics (MD) simulations of FT can offer the requisite information, but have not been applied due to a lack of effective methods for simulating the four-ligand coordination of zinc in proteins. Here, we report in detail the problems that occurred in the conventional MD simulations of the zinc-bound FT and a solution to these problems by employing a simple method that uses cationic dummy atoms to impose orientational requirement for zinc ligands. A successful 1.0 ns (1.0 fs time step) MD simulation of zinc-bound FT suggests that nine conserved residues (Asn127alpha, Gln162alpha, Asn165alpha, Gln195alpha, His248beta, Lys294beta, Leu295beta, Lys353beta, and Ser357beta) in the active site of mammalian FT are relatively mobile. Some of these residues might be involved in the ligand-induced active site conformational rearrangement upon binding and deserve attention in screening and design of improved FT inhibitors for cancer chemotherapy.     

Human glutathione-dependent formaldehyde dehydrogenase. Structures of apo,binary, and inhibitory ternary complexes

The human glutathione-dependent formaldehyde dehydrogenase is unique among the structurally studied members of the alcohol dehydrogenase family in that it follows a random bi bi kinetic mechanism. The structures of an apo form of the enzyme, a binary complex with substrate 12-hydroxydodecanoic acid, and a ternary complex with NAD+ and the inhibitor dodecanoic acid were determined at 2.0, 2.3, and 2.3 A resolution by X-ray crystallography using the anomalous diffraction signal of zinc. The structures of the enzyme and its binary complex with the primary alcohol substrate, 12-hydroxydodecanoic acid, and the previously reported binary complex with the coenzyme show that the binding of the first substrate (alcohol or coenzyme) causes only minor changes to the overall structure of the enzyme. This is consistent with the random mechanism of the enzyme where either of the substrates binds to the free enzyme. The catalytic-domain position in these structures is intermediate to the "closed" and "open" conformations observed in class I alcohol dehydrogenases. More importantly, two different tetrahedral coordination environments of the active site zinc are observed in these structures. In the apoenzyme, the active site zinc is coordinated to Cys44, His66 and Cys173, and a water molecule. In the inhibitor complex, the coordination environment involves Glu67 instead of the solvent water molecule. The coordination environment involving Glu67 as the fourth ligand likely represents an intermediate step during ligand exchange at the active site zinc. These observations provide new insight into metal-assisted catalysis and substrate binding in glutathione-dependent formaldehyde dehydrogenase.     

Kinetic evidence of the existence of a regulatory phosphoenolpyruvate binding site in maize leaf phosphoenolpyruvate carboxylase

Phenylphosphate, a structural analog of phosphoenolpyruvate (PEP), was found to be an activator of phosphoenolpyruvate carboxylase (PEP carboxylase) purified from maize leaves. This finding suggested the presence in the enzyme of a regulatory site, to which PEP could bind. We carried out kinetic studies on this enzyme using controlled concentrations of free PEP and of Mg-PEP complex and developed a theoretical kinetic model of the reaction. In summary, the main conclusions drawn from our results, and taken as assumptions of the model, were the following: (i) The affinity of the active site for the complex Mg-PEP is much higher than that for free PEP and Mg2+ ions, and therefore it can be considered that the preferential substrate of the PEP-catalyzed reaction is Mg-PEP. (ii) The enzyme has a regulatory site specific for free PEP, to which Mg2+ ions can not bind. (iii) The binding of free PEP, or an analog molecule, to this regulatory site yields a modified enzyme that has much lower apparent Km values and apparent Vmax values than the unmodified enzyme. So, free PEP behaves as an excellent activator of the reaction at subsaturating substrate concentrations, and as an inhibitor at saturating substrate concentrations. These findings may have important physiological implications on the regulation of the PEP carboxylase in vivo activity and, consequently, of the C4 pathway, since increased reaction rates would be obtained when the concentration of PEP rises, even at limiting Mg2+ concentrations.     

Molecular dynamics and free energy analysis of neuraminidase-ligand interactions

We report molecular dynamics calculations of neuraminidase in complex with an inhibitor, 4-amino-2-deoxy-2,3-didehydro-N-acetylneuraminic acid (N-DANA), with subsequent free energy analysis of binding by using a combined molecular mechanics/continuum solvent model approach. A dynamical model of the complex containing an ionized Glu119 amino acid residue is found to be consistent with experimental data. Computational analysis indicates a major van der Waals component to the inhibitor-neuraminidase binding free energy. Based on the N-DANA/neuraminidase molecular dynamics trajectory, a perturbation methodology was used to predict the binding affinity of related neuraminidase inhibitors by using a force field/Poisson-Boltzmann potential. This approach, incorporating conformational search/local minimization schemes with distance-dependent dielectric or generalized Born solvent models, correctly identifies the most potent neuraminidase inhibitor. Mutation of the key ligand four-substituent to a hydrogen atom indicates no favorable binding free energy contribution of a hydroxyl group; conversely, cationic substituents form favorable electrostatic interactions with neuraminidase. Prospects for further development of the method as an analysis and rational design tool are discussed.     

Periplasmic competition for zinc uptake between the metallochaperone ZnuA and Cu,Zn superoxide dismutase

We have investigated the availability of zinc in the periplasmic space of Escherichia coli using a mutant Cu,Zn superoxide dismutase whose dimerization is triggered by zinc binding. This mutant enzyme accumulates in the monomeric form when wild type cells are grown in minimal medium, but assembles in the dimeric form when it is produced in the same medium by a mutant strain lacking the periplasmic zinc metallochaperone ZnuA. These results indicate that periplasmic zinc-containing proteins compete for metal binding when bacteria grow in environments where this element is present in traces. The effective ZnuA ability to sequester the available zinc ions from the periplasm suggests that zinc-containing cytoplasmic proteins are more important for bacterial viability than the periplasmic ones.     

The 3-dehydroquinate synthase activity of the pentafunctional arom enzyme complex of Neurospora crassa is Zn2+-dependent.

We have demonstrated the co-purification in constant ratio of all five activities of the pentafunctional arom enzyme complex from Neurospora crassa. Progressive inactivation of the 3-dehydroquinate synthase component of the purified enzyme complex by chelating agents was blocked by the substrate, 3-deoxy-D-arabino-heptulosonate 7-phosphate, but not by the cofactor NAD+. Full activity was restored at Zn2+ concentrations as low as 0.05 nM. Atomic absorption data indicated that the intact enzyme complex contained 1 atom per subunit of tightly bound zinc. The arom 3-dehydroquinate synthase had a calculated turnover number of 19s-1, this being within the narrow range of values obtained for the other four activities of the intact multifunctional enzyme. The Km for 3-deoxy-D-arabino-heptulosonate 7-phosphate was 1.4 microM in a phosphate-free buffer; inorganic phosphate was a competitive inhibitor with respect to 3-deoxy-D-arabino-heptulosonate 7-phosphate.     

Formaldehyde dehydrogenase from Pseudomonas putida: a zinc metalloenzyme

The NAD+-dependent formaldehyde dehydrogenase from Pseudomonas putida C-83 was found to contain 4 gram atoms of zinc per mol, corresponding to 2 gram atoms of zinc per subunit monomer. Treatment of the enzyme with o-phenanthroline resulted in removal of 1 gram atom of zinc per subunit and caused a complete inactivation of the enzyme. The activity lost was restored by the addition of zinc ions, by which the zinc content was also reversed to almost the same level as that of the native enzyme. Another zinc atom that was resistant to metal chelator-treatment was liberated from the enzyme only after the irreversible denaturation of the enzyme. These results indicate that the formaldehyde dehydrogenase of P. putida is a zinc metalloenzyme and one of two zinc atoms per subunit participates in the catalytic activity of the enzyme, another zinc being presumably involved in maintaining the native conformation of the enzyme. Treatment of the enzyme with bipyridine also caused a reversible inactivation of the enzyme, but the zinc content remained unchanged. The spectrophotometric analysis indicated that the formation of a enzyme-Zn-bipyridine complex took place. Incubation of the enzyme with p-chloromercuribenzoate also resulted in a complete loss of the activity. These results suggest that an intrinsic zinc and sulfhydryl group together with NAD+ participate in the dehydrogenation reaction of substrate by the enzyme.     

Activity of yeast orotidine-5'-phosphate decarboxylase in the absence of metals.

Yeast orotidine-5'-phosphate decarboxylase was recently shown to contain zinc and to be inhibited by zinc-complexing agents. When the gene for the yeast enzyme was expressed in Escherichia coli, the gene product was devoid of metal atoms but exhibited a specific activity and molecular mass similar to those of the enzyme obtained directly from yeast. This invalidates the hypothesis that zinc is involved in substrate decarboxylation. The zinc-free enzyme undergoes thermal inactivation at a somewhat lower temperature than does the zinc-containing enzyme isolated from yeast.     

The binding of substrate analogs to phosphotriesterase

Phosphotriesterase (PTE) from Pseudomonas diminuta catalyzes the detoxification of organophosphates such as the widely utilized insecticide paraoxon and the chemical warfare agent sarin. The three-dimensional structure of the enzyme is known from high resolution x-ray crystallographic analyses. Each subunit of the homodimer folds into a so-called TIM barrel, with eight strands of parallel beta-sheet. The two zinc ions required for activity are positioned at the C-terminal portion of the beta-barrel. Here, we describe the three-dimensional structure of PTE complexed with the inhibitor diisopropyl methyl phosphonate, which serves as a mimic for sarin. Additionally, the structure of the enzyme complexed with triethyl phosphate is also presented. In the case of the PTE-diisopropyl methyl phosphonate complex, the phosphoryl oxygen of the inhibitor coordinates to the more solvent-exposed zinc ion (2.5 A), thereby lending support to the presumed catalytic mechanism involving metal coordination of the substrate. In the PTE-triethyl phosphate complex, the phosphoryl oxygen of the inhibitor is positioned at 3.4 A from the more solvent-exposed zinc ion. The two structures described in this report provide additional molecular understanding for the ability of this remarkable enzyme to hydrolyze such a wide range of organophosphorus substrates.     

A nuclear-magnetic-resonance study of the binding of novel N-hydroxybenzenesulphonamide carbonic anhydrase inhibitors to native and cadmium-111-substituted carbonic anhydrase

Various ring- and nitrogen-substituted benzenesulphonamides have been prepared and tested as potential inhibitors of carbonic anhydrase. N-Methoxysulphonamides showed no inhibitory activity, as predicted by the classic work of Krebs on N-substituted inhibitors. By contrast, N-hydroxysulphonamides proved to be very effective inhibitors of carbonic anhydrase. Using 111Cd-NMR it has been possible to analyse the molecular interaction of 4-fluoro-N-hydroxybenzenesulphon[15N]amide, with 111Cd-substituted bovine carbonic anhydrase. A large cadmium-111:nitrogen-15 spin-coupling shows that this inhibitor is directly bound to the metal via its nitrogen rather than through an oxygen atom. The mode of this binding is similar to that for the unsubstituted sulphonamide inhibitor, 4-fluorobenzenesulphon[15N]amide. The 111Cd-chemical shift of the signal for the inhibited enzyme shows that the N-hydroxysulphonamide is bound as its anion. From the relative intensities of free and complexed enzyme signals it can be deduced that the cadmium enzyme complex with the N-hydroxysulphonamide has a longer life-time than that formed with the unsubstituted sulphonamide. By contrast, native zinc-containing bovine carbonic anhydrase shows similar I50 values with both of these sulphonamides. Attempts to monitor the binding using 15N-NMR were unsuccessful, possibly due to a very long relaxation time for the nitrogen nucleus in the N-hydroxysulphonamide when bound to the enzyme leading to loss of the 15N signal.     

Inhibition by excess of free ATP, and free Mg2+ ions of the mitochondrial F1-ATPase moiety from Phycomyces blakesleeanus.

High concentrations of either Mg-ATP complex, free ATP, or free Mg2+ ions were inhibitors of the mitochondrial F1-ATPase moiety from Phycomyces blakesleeanus. Free Mg2+ acts as a linear competitive inhibitor with regard to Mg-ATP hydrolysis with a Ki value of 2.8 mM. The inhibition by free ATP was markedly biphasic and thus simple competitive inhibition alone is not sufficient to explain the inhibitory effect. From these results conclusions were drawn about the binding of the substrate, Mg-ATP complex, to the enzyme.     

Crystal structures of the cadmium- and mercury-substituted metallo-beta-lactamase from Bacteroides fragilis.

The metallo-beta-lactamases require zinc or cadmium for hydrolyzing beta-lactam antibiotics and are inhibited by mercurial compounds. To data, there are no clinically useful inhibitors of this class of enzymes. The crystal structure of the Zn(2+)-bound enzyme from Bacteroides fragilis contains a binuclear zinc center in the active site. A hydroxide, coordinated to both zinc atoms, is proposed as the moiety that mounts the nucleophilic attack on the carbonyl carbon atom of the beta-lactam ring. To study the metal coordination further, the crystal structures of a Cd(2+)-bound enzyme and of an Hg(2+)-soaked zinc-containing enzyme have been determined at 2.1 A and 2.7 A, respectively. Given the diffraction resolution, the Cd(2+)-bound enzyme exhibits the same active-site architecture as that of the Zn(2+)-bound enzyme, consistent with the fact that both forms are enzymatically active. The 10-fold reduction in activity of the Cd(2+)-bound molecule compared with the Zn(2+)-bound enzyme is attributed to fine differences in the charge distribution due to the difference in the ionic radii of the two metals. In contrast, in the Hg(2+)-bound structure, one of the zinc ions, Zn2, was ejected, and the other zinc ion, Zn1, remained in the same site as in the 2-Zn(2+)-bound structure. Instead of the ejected zinc, a mercury ion binds between Cys 104 and Cys 181, 4.8 A away from Zn1 and 3.9 A away from the site where Zn2 is located in the 2-Zn(2+)-bound molecule. The perturbed binuclear metal cluster explains the inactivation of the enzyme by mercury compounds.