Discovery of a Hidden Trypanosoma cruzi Spermidine Synthase Binding Site and Inhibitors through In Silico, In Vitro, and X-ray Crystallography.

In drug discovery research, the selection of promising binding sites and understanding the binding mode of compounds are crucial fundamental studies. The current understanding of the proteins-ligand binding model extends beyond the simple lock and key model to include the induced-fit model, which alters the conformation to match the shape of the ligand, and the pre-existing equilibrium model, selectively binding structures with high binding affinity from a diverse ensemble of proteins. Although methods for detecting target protein binding sites and virtual screening techniques using docking simulation are well-established, with numerous studies reported, they only consider a very limited number of structures in the diverse ensemble of proteins, as these methods are applied to a single structure. Molecular dynamics (MD) simulation is a method for predicting protein dynamics and can detect potential ensembles of protein binding sites and hidden sites unobservable in a single-point structure. In this study, to demonstrate the utility of virtual screening with protein dynamics, MD simulations were performed on Trypanosoma cruzi spermidine synthase to obtain an ensemble of dominant binding sites with a high probability of existence. The structure of the binding site obtained through MD simulation revealed pockets in addition to the active site that was present in the initial structure. Using the obtained binding site structures, virtual screening of 4.8 million compounds by docking simulation, in vitro assays, and X-ray analysis was conducted, successfully identifying two hit compounds.

In silico, in vitro, X-ray crystallography, and integrated strategies for discovering spermidine synthase inhibitors for Chagas disease.

Chagas disease results from infection by Trypanosoma cruzi and is a neglected tropical disease (NTD). Although some treatment drugs are available, their use is associated with severe problems, including adverse effects and limited effectiveness during the chronic disease phase. To develop a novel anti-Chagas drug, we virtually screened 4.8 million small molecules against spermidine synthase (SpdSyn) as the target protein using our super computer "TSUBAME2.5" and conducted in vitro enzyme assays to determine the half-maximal inhibitory concentration values. We identified four hit compounds that inhibit T. cruzi SpdSyn (TcSpdSyn) by in silico and in vitro screening. We also determined the TcSpdSyn-hit compound complex structure using X-ray crystallography, which shows that the hit compound binds to the putrescine-binding site and interacts with Asp171 through a salt bridge.

Pharmacophore modeling for anti-Chagas drug design using the fragment molecular orbital method.

BACKGROUND: Chagas disease, caused by the parasite Trypanosoma cruzi, is a neglected tropical disease that causes severe human health problems. To develop a new chemotherapeutic agent for the treatment of Chagas disease, we predicted a pharmacophore model for T. cruzi dihydroorotate dehydrogenase (TcDHODH) by fragment molecular orbital (FMO) calculation for orotate, oxonate, and 43 orotate derivatives. METHODOLOGY/PRINCIPAL FINDINGS: Intermolecular interactions in the complexes of TcDHODH with orotate, oxonate, and 43 orotate derivatives were analyzed by FMO calculation at the MP2/6-31G level. The results indicated that the orotate moiety, which is the base fragment of these compounds, interacts with the Lys43, Asn67, and Asn194 residues of TcDHODH and the cofactor flavin mononucleotide (FMN), whereas functional groups introduced at the orotate 5-position strongly interact with the Lys214 residue. CONCLUSIONS/SIGNIFICANCE: FMO-based interaction energy analyses revealed a pharmacophore model for TcDHODH inhibitor. Hydrogen bond acceptor pharmacophores correspond to Lys43 and Lys214, hydrogen bond donor and acceptor pharmacophores correspond to Asn67 and Asn194, and the aromatic ring pharmacophore corresponds to FMN, which shows important characteristics of compounds that inhibit TcDHODH. In addition, the Lys214 residue is not conserved between TcDHODH and human DHODH. Our analysis suggests that these orotate derivatives should preferentially bind to TcDHODH, increasing their selectivity. Our results obtained by pharmacophore modeling provides insight into the structural requirements for the design of TcDHODH inhibitors and their development as new anti-Chagas drugs.

Accelerating quantum chemistry calculations with graphical processing units - toward in high-density (HD) silico drug discovery.

The growing power of central processing units (CPU) has made it possible to use quantum mechanical (QM) calculations for in silico drug discovery. However, limited CPU power makes large-scale in silico screening such as virtual screening with QM calculations a challenge. Recently, general-purpose computing on graphics processing units (GPGPU) has offered an alternative, because of its significantly accelerated computational time over CPU. Here, we review a GPGPU-based supercomputer, TSUBAME2.0, and its promise for next generation in silico drug discovery, in high-density (HD) silico drug discovery.

Molecular modeling of vasopressin receptor and in silico screening of V1b receptor antagonists.

INTRODUCTION: G protein-coupled receptors (GPCRs) are integral membrane proteins which contain seven-transmembrane-spanning alpha-helices. GPCR-mediated signaling has been associated with various human diseases, positioning GPCRs as attractive targets in the drug discovery field. Recently, through advances in protein engineering and crystallography, the number of resolved GPCR structures has increased dramatically. This growing availability of GPCR structures has greatly accelerated structure-based drug design (SBDD) and in silico screening for GPCR-targeted drug discovery. AREAS COVERED: The authors introduce the current status of X-ray crystallography of GPCRs and what has been revealed from the resolved crystal structures. They also review the recent advances in SBDD and in silico screening for GPCR-targeted drug discovery and discuss a docking study, using homology modeling, with the discovery of potent antagonists of the vasopressin 1b receptor. EXPERT OPINION: Several innovative protein engineering techniques and crystallographic methods have greatly accelerated SBDD, not only for already-resolved GPCRs but also for those structures which remain unclear. These technological advances are expected to enable the determination of GPCR-fragment complexes, making it practical to perform fragment-based drug discovery. This paves the way for a new era of GPCR-targeted drug discovery.

Biological applications of hybrid quantum mechanics/molecular mechanics calculation.

Since in most cases biological macromolecular systems including solvent water molecules are remarkably large, the computational costs of performing ab initio calculations for the entire structures are prohibitive. Accordingly, QM calculations that are jointed with MM calculations are crucial to evaluate the long-range electrostatic interactions, which significantly affect the electronic structures of biological macromolecules. A UNIX-shell-based interface program connecting the quantum mechanics (QMs) and molecular mechanics (MMs) calculation engines, GAMESS and AMBER, was developed in our lab. The system was applied to a metalloenzyme, azurin, and PU.1-DNA complex; thereby, the significance of the environmental effects on the electronic structures of the site of interest was elucidated. Subsequently, hybrid QM/MM molecular dynamics (MD) simulation using the calculation system was employed for investigation of mechanisms of hydrolysis (editing reaction) in leucyl-tRNA synthetase complexed with the misaminoacylated tRNA(Leu), and a novel mechanism of the enzymatic reaction was revealed. Thus, our interface program can play a critical role as a powerful tool for state-of-the-art sophisticated hybrid ab initio QM/MM MD simulations of large systems, such as biological macromolecules.

Unidirectional Mechanistic Valved Mechanisms for Ammonia Transport in GatCAB.

Glutamine amidotransferase CAB (GatCAB), a crucial enzyme involved in translational fidelity, catalyzes three reactions: (i) the glutaminase reaction to yield ammonia (NH3 or NH4(+)) from glutamine, (ii) the phosphorylation of Glu-tRNA(Gln), and (iii) the transamidase reaction to convert the phosphorylated Glu-tRNA(Gln) to Gln-tRNA(Gln). In the crystal structure of GatCAB, the two catalytic centers are far apart, and the presence of a hydrophilic channel to transport the molecules produced by the reaction (i) was proposed. We investigated the transport mechanisms of GatCAB by molecular dynamics (MD) simulations and free energy (PMF) calculations. In the MD simulations (in total ∼1.1 µs), the entrance of the previously proposed channel is closed, as observed in the crystal structure. Instead, a novel hydrophobic channel has been identified in this study: Since the newly identified entrance opened and closed repeatedly in the MD simulations, it may act as a gate. The calculated free energy difference revealed the significant preference of the newly identified gate/channel for NH3 transport (∼10(4)-fold). In contrast, with respect to NH4(+), the free energy barriers are significantly increased for both channels due to tight hydrogen-bonding with hydrophilic residues, which hinders efficient transport. The opening of the newly identified gate is modulated by Phe206, which acts as a "valve". For the backward flow of NH3, our PMF calculation revealed that the opening of the gate is hindered by Ala207, which acts as a mechanistic "stopper" against the motion of the "valve" (Phe206). This is the first report to elucidate the detailed mechanisms of unidirectional mechanistic valved transport inside proteins.

Structural Instability of the Active Site of T1 Lipase Induced by Replacement of Na(+) with Water Complexed with the Phenylalanine Aromatic Ring.

The crystallographic analysis of T1 lipase suggested an interaction between Na(+) and the aromatic ring of Phe16 in the active site. However, experimental approaches could not dismiss the possible presence of water instead of Na(+). Our previous molecular dynamics (MD) simulations suggested that the significantly large enthalpy gain of the Na(+)-π interaction was required to preserve the catalytic core structure of T1 lipase. In this study, to examine the effects of water, we performed further MD simulations of T1 lipase involving the water-π interaction, instead of the Na(+)-π interaction, exploiting various force fields, such as ff99, ff02, and an accurate potential field to describe the water-π interaction, which was generated using our recently developed scheme (referred to as the grid-based energy representation). The analyses revealed that the water-π complex was unstable in the catalytic core of T1 lipase even when the accurate potential of the water-π complex represented by the grid-based energy function was employed in the MD simulations and led to the disruption of the coordinated structure. In contrast, the catalytic core structure of T1 lipase involving the Na(+)-π complex was significantly stable in the 10 ns MD simulation using the grid-based energy representation of the Na(+)-π interaction. Thus, the possible presence of water may be excluded, and our previous proposal concerning the functional role of the structural element involving the Na(+)-π interaction in the catalytic site of T1 lipase has unambiguously been confirmed. Further, the strong coordination of Na(+) and Nε of His358 was also shown to be substantial to preserve the core structure of the catalytic site.

Editing mechanism of aminoacyl-tRNA synthetases operates by a hybrid ribozyme/protein catalyst.

Aminoacyl-tRNA synthetases (aaRSs) are critical for the translational process, catalyzing the attachment of specific amino acids to their cognate tRNAs. To ensure formation of the correct aminoacyl-tRNA, and thereby enhance the reliability of translation, several aaRSs have an editing capability that hinders formation of misaminoacylated tRNAs. We investigated theoretically the mechanism of the editing reaction for a class I enzyme, leucyl-tRNA synthetase (LeuRS), complexed with a misaminoacylated tRNA(Leu), employing ab initio hybrid quantum mechanical/molecular mechanical potentials in conjunction with molecular dynamics simulations. It is shown that the water molecule that acts as the nucleophile in the editing reaction is activated by a 3'-hydroxyl group at the 3'-end of tRNA(Leu) and that the O2' atom of the leaving group of the substrate is capped by one of the water's hydrogen atoms. Thus, it is shown that editing is a self-cleavage reaction of the tRNA and so it is the tRNA, and not the protein, that drives the reaction. The protein does, however, have an important stabilizing effect on some high-energy intermediates along the reaction path, which is more efficient than the ribozyme would be alone. This indicates that editing is achieved by a novel "hybrid ribozyme/protein catalyst". Analysis of existing experimental data and additional modeling shows that this ribozymal mechanism appears to be widespread, occurring in the ribosome as well as in other aaRSs. It also suggests transitional forms that could have played an important role in the RNA world hypothesis for the origin of life.

Recent advances in jointed quantum mechanics and molecular mechanics calculations of biological macromolecules: schemes and applications coupled to ab initio calculations.

We review the recent research on the functional mechanisms of biological macromolecules using theoretical methodologies coupled to ab initio quantum mechanical (QM) treatments of reaction centers in proteins and nucleic acids. Since in most cases such biological molecules are large, the computational costs of performing ab initio calculations for the entire structures are prohibitive. Instead, simulations that are jointed with molecular mechanics (MM) calculations are crucial to evaluate the long-range electrostatic interactions, which significantly affect the electronic structures of biological macromolecules. Thus, we focus our attention on the methodologies/schemes and applications of jointed QM/MM calculations, and discuss the critical issues to be elucidated in biological macromolecular systems.

Modulation of electronic structures of bases through DNA recognition of protein.

The effects of environmental structures on the electronic states of functional regions in a fully solvated DNA·protein complex were investigated using combined ab initio quantum mechanics/molecular mechanics calculations. A complex of a transcriptional factor, PU.1, and the target DNA was used for the calculations. The effects of solvent on the energies of molecular orbitals (MOs) of some DNA bases strongly correlate with the magnitude of masking of the DNA bases from the solvent by the protein. In the complex, PU.1 causes a variation in the magnitude among DNA bases by means of directly recognizing the DNA bases through hydrogen bonds and inducing structural changes of the DNA structure from the canonical one. Thus, the strong correlation found in this study is the first evidence showing the close quantitative relationship between recognition modes of DNA bases and the energy levels of the corresponding MOs. Thus, it has been revealed that the electronic state of each base is highly regulated and organized by the DNA recognition of the protein. Other biological macromolecular systems can be expected to also possess similar modulation mechanisms, suggesting that this finding provides a novel basis for the understanding for the regulation functions of biological macromolecular systems.

Functional roles of a structural element involving Na+-pi interactions in the catalytic site of T1 lipase revealed by molecular dynamics simulations.

Interactions between metal ions and pi systems (metal-pi interactions) are known to confer significant stabilization energy. However, in biological systems, few structures with metal-pi coordination have been determined; thus, its roles must still be elucidated. The cation-pi interactions are not correctly described by current molecular mechanics even when using a polarizable force field, and thus they require quantum mechanical calculations for accurate estimation. However, the huge computational costs of the latter methodologies prohibit long-time molecular dynamics (MD) simulations. Accordingly, we developed a novel scheme to obtain an effective potential for calculating the interaction energy with an accuracy comparable to that of advanced ab initio calculations at the CCSD(T) levels, and with computational costs comparable to those of conventional MM calculations. Then, to elucidate the functional roles of the Na(+)-phenylalanine (Phe) complex in the catalytic site of T1 lipase, we performed MD simulations in the presence/absence of the accurate Na(+)-pi interaction energy. A comparison of these MD simulations revealed that a significantly large enthalpy gain in Na(+)-Phe16 substantially stabilizes the catalytic site, whereas a water molecule could not be substituted for Na(+) for sufficient stabilization energy. Thus, the cation-pi interaction in the lipase establishes a remarkably stable core structure by combining a hydrophobic aromatic ring and hydrophilic residues, of which the latter form the catalytic triad, thereby contributing to large structural changes from the complex with ligands to the free form of the lipase. This is the first report to elucidate the detailed functional mechanisms of Na(+)-pi interactions.

Structural modelling of the complex of leucyl-tRNA synthetase and mis-aminoacylated tRNALeu.

To assure fidelity of translation, class Ia aminoacyl-tRNA synthetases (aaRSs) edit mis-aminoacylated tRNAs. Mis-attached amino acids and structural water molecules are not included simultaneously in the current crystal structures of the aaRSctRNA complexes, where the 3'-ends (adenine 76; A76) are bound to the editing sites. A structural model of the completely solvated leucyl-tRNA synthetase complexed with valyl-tRNALeu was constructed by exploiting molecular dynamics simulations modified for the present modelling. The results showed that the ribose conformation of A76 is distinct from those observed in the above-mentioned crystal structures, which could be derived from structural constraints in a sandwiched manner induced by the mis-attached valine and tRNALeu.

Identification of the nucleophilic factors and the productive complex for the editing reaction by leucyl-tRNA synthetase.

To ensure fidelity of translation, several aminoacyl-tRNA synthetases (aaRSs) possess editing capability to hydrolyse mis-aminoacylated tRNAs. In this report, based on our previously-modelled structure of leucyl-tRNA synthetase (LeuRS) complexed with valyl-tRNA(Leu), further structural modelling has been performed along with molecular dynamics simulations. This enabled the identification of the nucleophile, which is different from that suggested by the crystal structure of the LeuRS * Nva2AA complex. Our results revealed that the 3' hydroxyl group of A76 acts as a "gate" to regulate the accessibility of the nucleophile; thus, the opening of the gate leads to the productive complex for the reaction.

A novel computational scheme for accurate and efficient evaluation of π-π and π-σ stacking.

Stacking involving aromatic rings has a significant contribution to the structural stability of biological macromolecules. However, conventional calculations such as density functional theory (DFT) and molecular mechanics (MM) fail to estimate such stabilization energies, most of which are fundamentally derived from van der Waals interactions. For the accurate description, higher level ab initio calculations, such as CCSD(T), should be employed; however, their computational costs are huge. MM calculations provide better estimation of the interactions of the aromatic rings than DFT, but not sufficient. In this report, we propose a novel scheme to calculate the interaction energy at an accuracy compatible to CCSD(T) with the computational costs comparable to MM calculations. In our scheme, the electron density of the aromatic rings is represented by Gaussian-type functions, and the parameters involved in the functions are determined by an optimization scheme to reproduce the CCSD(T) results. Here, we employ model structures involving tryptophan and tyrosine rings, and successfully obtain the optimal parameter set. By using this type of representation of the stacking proposed, the computational time to calculate the interaction energy is dramatically reduced by 10(-10)-fold, compared with CCSD(T).

QM/MM hybrid calculation of biological macromolecules using a new interface program connecting QM and MM engines.

An interface program connecting a quantum mechanics (QM) calculation engine, GAMESS, and a molecular mechanics (MM) calculation engine, AMBER, has been developed for QM/MM hybrid calculations. A protein-DNA complex is used as a test system to investigate the following two types of QM/MM schemes. In a 'subtractive' scheme, electrostatic interactions between QM/MM regions are truncated in QM calculations; in an 'additive' scheme, long-range electrostatic interactions within a cut-off distance from QM regions are introduced into one-electron integration terms of a QM Hamiltonian. In these calculations, 338 atoms are assigned as QM atoms using Hartree-Fock (HF)/density functional theory (DFT) hybrid all-electron calculations. By comparing the results of the additive and subtractive schemes, it is found that electronic structures are perturbed significantly by the introduction of MM partial charges surrounding QM regions, suggesting that biological processes occurring in functional sites are modulated by the surrounding structures. This also indicates that the effects of long-range electrostatic interactions involved in the QM Hamiltonian are crucial for accurate descriptions of electronic structures of biological macromolecules.

Electronic and geometric structures of the blue copper site of azurin investigated by QM/MM hybrid calculations.

The electronic and geometric structures of the copper-binding site in a fully solvated azurin were investigated using quantum mechanics (QM) and molecular mechanics (MM) hybrid calculations. Two types of computational models were applied to evaluate the effects of the environment surrounding the active site. In model I, long-distance electrostatic interactions between QM region atoms and partial point charges of the surrounding protein moieties and solvent water were calculated in a QM Hamiltonian, for which the spin-unrestricted Hartree-Fock (UHF)/density functional theory (DFT) hybrid all-electron calculation with the B3LYP functional was adopted. In model II, the QM Hamiltonian was not allowed to be polarized by those partial point charges. Models I and II provided different descriptions of the copper coordination structure, particularly for the coordinative bonds including a large dipole. In fact, the Cu-O(Gly45) and Cu-S(Cys112) bonds are sensitive to the treatment of long-distance electrostatic interactions in the QM Hamiltonian. This suggests that biological processes occurring in the active site are regulated by the surrounding structures of protein and solvent, and therefore the effects of long-range electrostatic interactions involved in the QM Hamiltonian are crucial for accurate descriptions of electronic structures of the copper active site of metalloenzymes.

Evaluation of stabilization energies in π-π and cation-π interactions involved in biological macromolecules by ab initio calculations.

Non-covalent interactions involving aromatic rings contribute significantly to the stability of three-dimensional structures of biological macromolecules. Therefore, accurate descriptions of such interactions are crucial in understanding the functional mechanisms of biological molecules. However, it is also well known that, for some cases where van der Waals interactions make a dominant contribution, conventional ab initio electronic structure calculations, such as density functional theory, do not produce accurate interaction energies. In this study, we evaluated molecular mechanics (MM) calculations for two types of interactions involving aromatic rings, π-π interactions and cation-π interactions, by comparing our results with those obtained by advanced ab initio calculations at the coupled-cluster with singles, doubles and perturbative triples level. In structures with stacked aromatic rings, interaction energies obtained by MM calculations are overestimated. On the other hand, for cation-π interactions, the energies in MM calculations are significantly underestimated. In both cases, addition of an induction energy based on polarization effects also fails to improve the estimate given by MM calculations. The results indicate that current effective pairwise potentials are inappropriate to represent π-π and cation-π interactions.

First principles molecular dynamics study of catalytic reactions of biological macromolecular systems: toward analyses with QM/MM hybrid molecular simulations.

First principles molecular dynamics simulations performed on a fully solvated RNA model structure allowed us to investigate the mechanism for enzymatic cleavage reactions, in vitro, of RNA enzymes (ribozymes). The concerted action of two metal catalysts turns out to be the most efficient way to promote, on the one hand, the proton abstraction from 2(')-OH that triggers the nucleophilic attack and, on the other hand, the cleavage of the P-O(5(')) bond. In fact, the elimination of one of the two metal cations leads to an increase in the activation energy of the reaction. The simulated pathway shows that an OH(-) in the coordination shell of the Mg(2+) close to O(2(')) promotes the initial proton abstraction and prevents its transfer to the ribozyme. This suggests that, in a real ribozyme, the double-metal-ion reaction mechanism is preferred with respect to single-metal-ion mechanisms either in the presence or in absence of the OH(-) anion. Finally, an insight into the importance of hybrid quantum mechanics/molecular mechanics (QM/MM) schemes is discussed in view of the modelling of a realistic system carrying all the features of a true ribozyme.