Development and validation of a genetic algorithm for flexible docking

Pharmacological Differences between Human and Guinea Pig Histamine H1 Receptors: Asn84 (2.61) as Key Residue within an Additional Binding Pocket in the H1 Receptor

We tested several histamine H(1) receptor (H(1)R) and antagonists for their differences in agonists binding affinities between human and guinea pig H(1)Rs transiently expressed in African green monkey kidney (COS-7) cells. Especially, the bivalent agonist histaprodifen-histamine dimer (HP-HA) shows a higher affinity for guinea pig than for human H(1)Rs. Based on the structure of HP-HA, we have further identified VUF 4669 [7-(3-(4-(hydroxydiphenylmethyl)piperidin-1-yl)propoxy)-4-oxochroman-2-carboxylic acid] as a guinea pig-preferring H(1)R antagonist, demonstrating that the concept of species selectivity is not limited to agonists. To delineate the molecular mechanisms behind the observed species selectivity, we have created mutant human H(1)Rs in which amino acids were individually replaced by their guinea pig H(1)R counterparts. Residue Asn(84) (2.61) in transmembrane domain (TM) 2 seemed to act as a selectivity switch in the H(1)R. Molecular modeling and site-directed mutagenesis studies suggest that Asn(84) interacts with the conserved Tyr(458) (7.43) in TM7. Our data provide the first evidence that for some H(1)R ligands, the binding pocket is not only limited to TMs 3, 4, 5, and 6 but also comprises an additional pocket formed by TMs 2 and 7.

Fragment-Based Lead Discovery Using X-ray Crystallography

Fragment screening offers an alternative to traditional screening for discovering new leads in drug discovery programs. This paper describes a fragment screening methodology based on high throughput X-ray crystallography. The method is illustrated against five proteins (p38 MAP kinase, CDK2, thrombin, ribonuclease A, and PTP1B). The fragments identified have weak potency (>100 microM) but are efficient binders relative to their size and may therefore represent suitable starting points for evolution to good quality lead compounds. The examples illustrate that a range of molecular interactions (i.e., lipophilic, charge-charge, neutral hydrogen bonds) can drive fragment binding and also that fragments can induce protein movement. We believe that the method has great potential for the discovery of novel lead compounds against a range of targets, and the companion paper illustrates how lead compounds have been identified for p38 MAP kinase starting from fragments such as those described in this paper.

Docking and scoring in virtual screening for drug discovery: methods and applications

Computational approaches that 'dock' small molecules into the structures of macromolecular targets and 'score' their potential complementarity to binding sites are widely used in hit identification and lead optimization. Indeed, there are now a number of drugs whose development was heavily influenced by or based on structure-based design and screening strategies, such as HIV protease inhibitors. Nevertheless, there remain significant challenges in the application of these approaches, in particular in relation to current scoring schemes. Here, we review key concepts and specific features of small-molecule-protein docking methods, highlight selected applications and discuss recent advances that aim to address the acknowledged limitations of established approaches.

Design and Synthesis of a Focused Library of Novel Aryl- and Heteroaryl-Ketopiperazides

1-Phenyl-4-piperazinyl-carbonyl-substituted nitrogen-containing heterocycles were discovered at Zentaris as a new class of potent, synthetic, small molecule tubulin inhibitors with strong antiproliferative activity. The lead structure of this class, D-24203, proved to be a potent inhibitor of in vivo tumor growth in different xenograft models including mammary and renal cancers. As part of our efforts in the lead optimization process to expand structural diversity as well as to optimize bioavailability parameters such as solubility and metabolic stability for these compounds, we produced and evaluated a focused library containing 320 compounds. Five new heterocyclic compound classes with comparable activity properties in the cytotoxicity and tubulin polymerization assay could be identified. In silico calculated bioavailability parameters for selected library members provides new compound classes with improved solubility properties. Library design, development of adequate solution phase methodology, and synthesis will be presented, as well as results of lead optimization.

β-Diketo Acid Pharmacophore Hypothesis. 1. Discovery of a Novel Class of HIV-1 Integrase Inhibitors

HIV-1 Integrase (IN) is an essential enzyme for viral replication. The discovery of beta-diketo acids was crucial in the validation of IN as a legitimate target in drug discovery against HIV infection. In this study, we discovered a novel class of IN inhibitors using a 3D pharmacophore guided database search. We used S-1360 (1), the first IN inhibitor to undergo clinical trials, and three other analogues to develop a common feature pharmacophore hypothesis. Testing this four-featured pharmacophore against a multiconformational database of 150,000 structurally diverse small molecules yielded 1,700 compounds that satisfied the 3D query. Subsequently, all 1,700 compounds were docked into the active site of IN. On the basis of docking scores, Lipinski's rule-of-five, and structural novelty, 110 compounds were selected for biological screening. We found that compounds that contain both salicylic acid and a 2-thioxo-4-thiazolidinone (rhodanine) group (e.g. 5-13) showed significant inhibitory potency against IN, while the presence of either salicylic acid or a rhodanine group alone did not. Although some of the compounds containing only a salicylic acid showed inhibitory potency against IN, none of the compounds containing only rhodanine exhibited considerable potency. Of the 52 compounds reported in this study, 11 compounds (5, 6, 8, 10-13, 32-33, 51, and 53) inhibited 3'-processing or strand transfer activities of IN with IC(50) < or = 25 microM. This is the first reported use of S-1360 and its analogues as leads in developing a pharmacophore hypothesis for IN inhibition and for identification of new compounds with potent inhibition of this enzyme.

Potent and Selective Farnesyl Transferase Inhibitors

We recently described a novel series of CA(1)A(2)X peptidomimetics as farnesyl transferase inhibitors (FTIs). These compounds possess an N-(4-piperidinyl)benzamide scaffold mimicking A(1)A(2) residue. Extensive exploration of structure--activity relationships revealed that replacement of cysteine by substituted benzylimidazoles provided nanomolar FTIs with in vitro activities (18e, IC(50) = 4.60 nM on isolated enzyme, EC(50) = 20.0 nM for growth inhibition on a tumor cell line). The molecular docking of 18e and 19e in the active site of the enzyme provided details of key interactions with the protein and showed that the methionine or phenylalanine residue fits into the aryl binding site.

Ribonucleotide activation by enzyme ribonucleotide reductase: understanding the role of the enzyme

This article focuses on the first step of the catalytic mechanism for the reduction of ribonucleotides catalyzed by the enzyme Ribonucleotide Reductase (RNR). This corresponds to the activation of the substrate. In this work a large model of the active site region involving 130 atoms was used instead of the minimal gas phase models used in previous works. The ONIOM method was employed to deal with such a large system. The results gave additional information, which previous small models could not provide, allowing a much clearer evaluation of the role of the enzyme in this step. Enzyme-substrate interaction energies, specific transition state stabilization, and substrate steric strain energies were obtained. It was concluded that the transition state is stabilized in 4.0 kcal/mol by specific enzyme-substrate interactions. However, this stabilization is cancelled by the cost in conformational energy for the enzyme to adopt the transition state geometry; the overall result is that the enzyme machinery does not lead to a rate enhancement in this step. It was also found that the substrate binds to the active site with almost no steric strain, emphasizing the complementarity and specificity of the RNR active site for nucleotide binding. The main role of the enzyme at the very beginning of the catalytic cycle was concluded to be to impose stereospecifity upon substrate activation and to protect the enzyme radical from the solvent, rather than to be an reaction rate enhancement.

A docking score function for estimating ligand-protein interactions: application to acetylcholinesterase inhibition

Acetylcholinesterase (AChE) inhibition is an important research topic because of its wide range of associated health implications. A receptor-specific scoring function was developed herein for predicting binding affinities for human AChE (huAChE) inhibitors. This method entails a statistically trained weighted sum of electrostatic and van der Waals (VDW) interactions between ligands and the receptor residues. Within the 53 ligand training set, a strong correlation was found (R2 = 0.89) between computed and experimental inhibition constants. Leave-one-out cross-validation indicated high predictive power (Q2 = 0.72), and analysis of a separate 16-compound test set also produced very good correlation with experiment (R2 = 0.69). Scoring function analysis has permitted identification and characterization of important ligand-receptor interactions, producing a list of those residues making the most important electrostatic and VDW contributions within the main active site, gorge area, acyl binding pocket, and periferal site. These analyses are consistent with X-ray crystallographic and site-directed mutagenesis studies.

Studies of Binding Modes of (S)-Mephenytoin to Wild Types and Mutants of Cytochrome P450 2C19 and 2C9 Using Homology Modeling and Computational Docking

PURPOSE

This study investigated the structural features of CYP2C19 complexed with (S)-mephenytoin, using computational methods. In addition to wild-type CYP2C19 proteins (1A and 1B), which have selective 4-hydroxylase activities of (S)-mephenytoin, CYP2C19 mutants were also studied, together with a wild type and artificial mutants of CYP2C19.

METHODS

Three-dimensional structures of wild-type and mutant proteins of CYP2C19 and CYP2C9 were estimated from homology modeling using the crystal structure of rabbit CYP2C5 as a reference. The binding mode of (S)-mephenytoin to CYP2C19 was investigated using computational docking.

RESULTS

The results reproduced the specific bindings between (S)-mephenytoin and the wild types of CYP2C19. Our findings suggest that Asp293 of CYP2C19 plays an important role in the binding of (S)-mephenytoin, which was surrounded by Vall13 and Ala297, and points the phenyl ring at the heme iron. In addition the wild types of CYP2C19, the computational docking studies also accounted for the experimental activities of CYP2C19 mutants, and wild-type and mutant CYP2C19 proteins.

CONCLUSIONS

These results confirm that the predicted three-dimensional structure of the CYP2C19-(S)-mephenytoin complex is reasonable, and that this strategy is useful for investigating complex structures. Virtual screening for drug discovery can also be carried out using these methods.

Scoring functions for protein-ligand interactions: a critical perspective

Scoring functions play an essential role in structure-based virtual screening. They are required to guide the docking of candidate compounds to structures of receptor binding sites, to select probable binding modes, and to discriminate binders from non-binders. Although many scoring functions have successfully been used to identify novel ligands for a wide variety of targets, much work remains to be done to avoid incorrect prediction of binding modes and high numbers of false positives. This review gives an overview of the current state of the field and outlines key issues for the further development of scoring functions.:

Evans Blue is an inhibitor of nuclear factor-kappa B (NF-κB)-DNA binding

Nuclear factor-kappa B (NF-kappaB) is an important transcription factor, involved in many immune and inflammatory responses. It is critical in HIV gene expression as it has kappa B binding sites in the HIV-1 long-terminal repeat. Hence, targeting NF-kappaB to prevent its DNA binding holds a significant therapeutic potential. In this context, we report Evans Blue as a novel inhibitor of NF-kappaB-DNA binding. Evans Blue was found to be inhibiting DNA binding of NF-kappaB at a low concentration of 100 microM. Further, molecular modeling studies using docking and generation of electrostatic potential maps predicted a possible binding mode of EB to the DNA binding region of NF-kappaB, consistent with the experimental activity.

Studies of nonnucleoside HIV-1 reverse transcriptase inhibitors. Part 1: Design and synthesis of thiazolidenebenzenesulfonamides

A random high-throughput screening (HTS) program to discover novel nonnucleoside reverse transcriptase inhibitors (NNRTIs) has been carried out with MT-4 cells against a nevirapine-resistant virus, HIV-1(IIIB-R). The primary hit, a thiazolidenebenzenesulfonamide derivative, possessed good activity. A systematic modification program examining various substituents at the 3-, 4-, and 5-positions on the thiazole ring afforded compounds with enhanced anti-HIV-1 and reverse transcriptase (RT) inhibitory activities. These results confirm the important role of the substituents at these positions and the thiazolidenebenzenesulfonamide motif as a valuable lead series for the next generation NNRTIs.

Model structures of the N-methyl-D-aspartate receptor subunit NR1 explain the molecular recognition of agonist and antagonist ligands

Molecular models of the ligand-binding domain of N-methyl-d-aspartate subunit R1 (NR1) were made using the published crystal structures of rat glutamate receptor B (GluRB), the bacterial glutamate receptor (GluR0), and the glutamine-binding protein (QBP) of Escherichia coli. Separate models of NR1 were built to represent the ligand-binding conformation for agonist (glycine, d- and l-isomers of serine and alanine, and the partial agonist ligand d-cycloserine) and antagonist (5,7-dichloro-4-oxo-1,4-dihydroquinoline-2-carboxylic acid (DCKA) and E-3-(2-phenyl-2-carboxyethenyl)-4,6-dichloro-1-H-indole-2-carboxylic acid (MDL 105,519)) ligands. Side-chain conformations of residues within the NR1 ligand-binding site were selected that optimized the hydrophobic packing and hydrogen bonding among residues, while taking into account published data comparing receptor mutants with wild-type NR1. Ligands docked to the model structures provide a rational explanation for the observed differences in binding affinity and receptor activation among agonist and antagonist ligands. NR1 prefers smaller ligands (glycine, serine, and alanine) in comparison with GluRB and GluR0 that bind l-glutamate: the bulky side chain of W731 in NR1 dramatically reduces the size of the ligand-binding site, functioning to selectively restrict recognition to glycine and the d-isomers of serine and alanine. Nevertheless, many of the interactions seen for ligands bound to GluRB, GluR0, and periplasmic-binding proteins are present for the ligands docked to the model structures of NR1.

Structure-based discovery of nonpeptidic small organic compounds to block the T cell response to myelin basic protein

We have utilized a computational structure-based approach to identify nonpeptidic small organic compounds that bind to a human leukocyte antigen (HLA) DR1301 molecule (HLA-DR1301 or DR1301) and block the presentation of myelin basic protein peptide 152-165 (MBP 152-165) to T cells. A three-dimensional (3D) structure of DR1301 was derived by homology modeling followed by extensive molecular dynamics simulation for structural refinement. Computational structure-based database searching was performed to identify nonpeptidic small-molecule candidates from the National Cancer Institute (NCI) database containing over 150 000 compounds that can effectively interact with the peptide-binding groove of the HLA molecule. By in vitro testing of 106 candidate small molecules, two lead compounds were confirmed to specifically block IL-2 secretion by DR1301-restricted T cells in a dose-dependent and reversible manner. The specificity of blocking DR1301-restricted MBP presentation was further validated in a binding assay using an analogue of the most potent lead compound. Computational docking was performed to predict the three-dimensional binding model of these confirmed small molecule blockers to the DR1301 molecule and to gain structural insight into their interactions. Our results suggest that computational structure-based searching is an effective approach to discover nonpeptidic small organic compounds to block the interaction between DR1301 and T cells. The nonpeptidic small organic compounds identified in this study are useful pharmacological tools to study the interactions between HLA molecules and T cells and a starting point for the development of a novel therapeutic strategy for the treatment of multiple sclerosis (MS) or other immune-related disorders.

Discovery of Bicyclic Thymidine Analogues as Selective and High-Affinity Inhibitors of Mycobacterium tuberculosis Thymidine Monophosphate Kinase

Thymidine monophosphate kinase of Mycobacterium tuberculosis (TMPKmt) represents an attractive target for selectively blocking bacterial DNA synthesis. Hereby, we report on the discovery of a novel class of bicyclic nucleosides (10 and 11) and one dinucleoside (12), belonging to the most selective inhibitors of TMPKmt discovered so far.

Synthesis, Biological Evaluation, and Molecular Modeling Studies of a Novel, Peripherally Selective Inhibitor of Catechol-O-methyltransferase

A novel series of potent, peripherally selective, and long-acting inhibitors of catechol-O-methyltransferase (COMT) has been synthesized. The introduction and nature of heteroatom-containing substituents to the side-chain of the nitrocatechol pharmacophore was found to have a profound effect on both peripheral selectivity and duration of COMT inhibition in the mouse. This approach led to the discovery of 1-(3,4-dihydroxy-5-nitrophenyl)-3-[4-[3-(trifluoromethyl)phenyl]-1-piperazinyl]-1-propanone hydrochloride 35 (BIA 3-335), which was found to possess a superior inhibitory profile in vivo over both the nonselective inhibitor tolcapone 1 and the peripherally selective but short-acting entacapone 2. In this model, 35 retained 75% inhibition of peripheral COMT at 6 h after oral administration, yet significantly, only a minor reduction of central (cerebral) COMT activity was observed. Molecular modeling techniques were applied to review the analysis of the ternary enzyme-inhibitor complex previously determined by X-ray crystallography and to provide a deeper understanding of the structure-activity relationships within this novel series. Furthermore, a computational approach was applied in an effort to elucidate the particular structural factors relevant to the poor blood-brain permeability of 35. In conclusion, the improved biological properties herein reported reveal 35 as a candidate for clinical studies as an adjunct to L-DOPA therapy for Parkinson's disease.

Key Amino Acids Located within the Transmembrane Domains 5 and 7 Account for the Pharmacological Specificity of the Human V1b Vasopressin Receptor

In mammals, the vasopressin V(1b) receptor (V(1b)-R) is known to regulate ACTH secretion and, more recently, stress and anxiety. The characterization of the molecular determinant responsible for its pharmacological selectivity was made possible by the recent discovery of the first V(1b) antagonist, SSR149415. Based upon the structure of the crystallized bovine rhodopsin, we established a three-dimensional molecular model of interaction between the human V(1b)-R (hV(1b)-R) and SSR149415. Four amino acids located in distinct transmembrane helices (fourth, fifth, and seventh) were found potentially responsible for the hV(1b)-R selectivity. To validate these assumptions, we selectively replaced the leucine 181, methionine 220, alanine 334, and serine 338 residues of hV(1a)-R by their corresponding amino acids present in the hV(1b)-R (phenylalanine 164, threonine 203, methionine 324, and asparagine 328, respectively). Four mutants, which all exhibited nanomolar affinities for vasopressin and good coupling to phospholipase C pathway, were generated. hV(1a) receptors mutated at position 220 and 334 exhibited striking increase in affinity for SSR149415 both in binding and phospholipase C assays at variance with the hV(1a)-R modified at position 181 or 338. In conclusion, this study provides the first structural features concerning the hV(1b)-R and highlights the role of few specific residues in its pharmacological selectivity.

Efficient method for high-throughput virtual screening based on flexible docking: discovery of novel acetylcholinesterase inhibitors

A method of easily finding ligands, with a variety of core structures, for a given target macromolecule would greatly contribute to the rapid identification of novel lead compounds for drug development. We have developed an efficient method for discovering ligand candidates from a number of flexible compounds included in databases, when the three-dimensional (3D) structure of the drug target is available. The method, named ADAM&EVE, makes use of our automated docking method ADAM, which has already been reported. Like ADAM, ADAM&EVE takes account of the flexibility of each molecule in databases, by exploring the conformational space fully and continuously. Database screening has been made much faster than with ADAM through the tuning of parameters, so that computational screening of several hundred thousand compounds is possible in a practical time. Promising ligand candidates can be selected according to various criteria based on the docking results and characteristics of compounds. Furthermore, we have developed a new tool, EVE-MAKE, for automatically preparing the additional compound data necessary for flexible docking calculation, prior to 3D database screening. Among several successful cases of lead discovery by ADAM&EVE, the finding of novel acetylcholinesterase (AChE) inhibitors is presented here. We performed a virtual screening of about 160 000 commercially available compounds against the X-ray crystallographic structure of AChE. Among 114 compounds that could be purchased and assayed, 35 molecules with various core structures showed inhibitory activities with IC(50) values less than 100 microM. Thirteen compounds had IC(50) values between 0.5 and 10 microM, and almost all their core structures are very different from those of known inhibitors. The results demonstrate the effectiveness and validity of the ADAM&EVE approach and provide a starting point for development of novel drugs to treat Alzheimer's disease.

Knowledge-based elastic potentials for docking drugs or proteins with nucleic acids

Elastic ellipsoidal functions defined by the observed hydration patterns around the DNA bases provide a new basis for measuring the recognition of ligands in the grooves of double-helical structures. Here a set of knowledge-based potentials suitable for quantitative description of such behavior is extracted from the observed positions of water molecules and amino acid atoms that form hydrogen bonds with the nitrogenous bases in high resolution crystal structures. Energies based on the displacement of hydrogen-bonding sites on drugs in DNA-crystal complexes relative to the preferred locations of water binding around the heterocyclic bases are low, pointing to the reliability of the potentials and the apparent displacement of water molecules by drug atoms in these structures. The validity of the energy functions has been further examined in a series of sequence substitution studies based on the structures of DNA bound to polyamides that have been designed to recognize the minor-groove edges of Watson-Crick basepairs. The higher energies of binding to incorrect sequences superimposed (without conformational adjustment or displacement of polyamide ligands) on observed high resolution structures confirm the hypothesis that the drug subunits associate with specific DNA bases. The knowledge-based functions also account satisfactorily for the measured free energies of DNA-polyamide association in solution and the observed sites of polyamide binding on nucleosomal DNA. The computations are generally consistent with mechanisms by which minor-groove binding ligands are thought to recognize DNA basepairs. The calculations suggest that the asymmetric distributions of hydrogen-bond-forming atoms on the minor-groove edge of the basepairs may underlie ligand discrimination of G.C from C.G pairs, in addition to the commonly believed role of steric hindrance. The analysis of polyamide-bound nucleosomal structures reveals other discrepancies in the expected chemical design, including unexpected contacts to DNA and modified basepair targets of some ligands. The ellipsoidal potentials thus appear promising as a mathematical tool for the study of drug- and protein-DNA interactions and for gaining new insights into DNA-binding mechanisms.

Phe120 contributes to the regiospecificity of cytochrome P450 2D6: mutation leads to the formation of a novel dextromethorphan metabolite

Although the residues that determine the preference of CYP2D6 (cytochrome P450 2D6) for compounds containing a basic nitrogen are well characterized, the contribution of other active site residues to substrate binding and orientation is less well understood. Our structural model of CYP2D6 identifies the aromatic residue Phe120 as a likely major feature of the active site. To examine the role of Phe120, mutants of CYP2D6 in which this residue has been substituted by alanine, leucine, tyrosine, serine, histidine, tryptophan or methionine residues have been prepared in bacterial membranes co-expressing human cytochrome NADPH cytochrome P450 oxidoreductase. The mutants have been characterized using the prototypical bufuralol 1' hydroxylase and dextromethorphan O- and N-demethylase activities of CYP2D6. Larger effects on K(m) values are observed for dextromethorphan O-demethylation than for bufuralol 1' hydroxylation, indicating that the Phe120 side chain is more important in dextromethorphan than in bufuralol binding. A role for this side chain in determining the regiospecificity of substrate oxidation was indicated by changes in the relative rates of O- and N-demethylation of dextromethorphan and, notably, by the formation of 7-hydroxy dextromethrophan, a novel dextromethorphan metabolite, in mutants in which it had been substituted. Computational studies of dextromethorphan binding to the active site of the Phe120-->Ala mutant were carried out to throw light on the way in which the removal of this side chain leads to different modes of ligand binding.

Design and synthesis of potent, orally active, inhibitors of carboxypeptidase U (TAFIa)

A series of 3-mercapto-propionic acid derivatives that function as reversible inhibitors of carboxypeptidase U have been prepared. We present a successful design strategy using cyclic, low basicity guanidine mimetics resulting in potent, selective and bioavailable inhibitors of carboxypeptidase U (TAFIa).

Structural mining: self-consistent design on flexible protein-peptide docking and transferable binding affinity potential

A flexible protein-peptide docking method has been designed to consider not only ligand flexibility but also the flexibility of the protein. The method is based on a Monte Carlo annealing process. Simulations with a distance root-mean-square (dRMS) virtual energy function revealed that the flexibility of protein side chains was as important as ligand flexibility for successful protein-peptide docking. On the basis of mean field theory, a transferable potential was designed to evaluate distance-dependent protein-ligand interactions and atomic solvation energies. The potential parameters were developed using a self-consistent process based on only 10 known complex structures. The effectiveness of each intermediate potential was judged on the basis of a Z score, approximating the gap between the energy of the native complex and the average energy of a decoy set. The Z score was determined using experimentally determined native structures and decoys generated by docking with the intermediate potentials. Using 6600 generated decoys and the Z score optimization criterion proposed in this work, the developed potential yielded an acceptable correlation of R(2) = 0.77, with binding free energies determined for known MHC I complexes (Class I Major Histocompatibility protein HLA-A(*)0201) which were not present in the training set. Test docking on 25 complexes further revealed a significant correlation between energy and dRMS, important for identifying native-like conformations. The near-native structures always belonged to one of the conformational classes with lower predicted binding energy. The lowest energy docked conformations are generally associated with near-native conformations, less than 3.0 Angstrom dRMS (and in many cases less than 1.0 Angstrom) from the experimentally determined structures.

Improving conformational searches by geometric screening

MOTIVATION

Conformational searches in molecular docking are a time-consuming process with wide range of applications. Favorable conformations of the ligands that successfully bind with receptors are sought to form stable ligand-receptor complexes. Usually a large number of conformations are generated and their binding energies are examined. We propose adding a geometric screening phase before an energy minimization procedure so that only conformations that geometrically fit in the binding site will be prompted for energy calculation.

RESULTS

Geometric screening can drastically reduce the number of conformations to be examined from millions (or higher) to thousands (or lower). The method can also handle cases when there are more variables than geometric constraints. An early-stage implementation is able to finish the geometric filtering of conformations for molecules with up to nine variables in 1 min. To the best of our knowledge, this is the first time such results are reported deterministically.

CONTACT

mzhang@mdanderson.org.

ProPose: a docking engine based on a fully configurable protein–ligand interaction model

Virtual high-throughput screening of molecular databases and in particular high-throughput protein-ligand docking are both common methodologies that identify and enrich hits in the early stages of the drug design process. Current protein-ligand docking algorithms often implement a program-specific model for protein-ligand interaction geometries. However, in order to create a platform for arbitrary queries in molecular databases, a new program is desirable that allows more manual control of the modeling of molecular interactions. For that reason, ProPose, an advanced incremental construction docking engine, is presented here that implements a fast and fully configurable molecular interaction and scoring model. This program uses user-defined, discrete, pharmacophore-like representations of molecular interactions that are transformed on-the-fly into a continuous potential energy surface, allowing for the incorporation of target specific interaction mechanisms into docking protocols in a straightforward manner. A torsion angle library, based on semi-empirical quantum chemistry calculations, is used to provide minimum energy torsion angles for the incremental construction algorithm. Docking results of a diverse set of protein-ligand complexes from the Protein Data Bank demonstrate the feasibility of this new approach. As a result, the seamless integration of pharmacophore-like interaction types into the docking and scoring scheme implemented in ProPose opens new opportunities for efficient, receptor-specific screening protocols. [figure: see text]. ProPose--a fully configurable protein-ligand docking program--transforms pharmacophores into a smooth potential energy surface.

Application of Machine Learning To Improve the Results of High-Throughput Docking Against the HIV-1 Protease

We have previously reported that the application of a Laplacian-modified naive Bayesian (NB) classifier may be used to improve the ranking of known inhibitors from a random database of compounds after High-Throughput Docking (HTD). The method relies upon the frequency of substructural features among the active and inactive compounds from 2D fingerprint information of the compounds. Here we present an investigation of the role of extended connectivity fingerprints in training the NB classifier against HTD studies on the HIV-1 protease using three docking programs: Glide, FlexX, and GOLD. The results show that the performance of the NB classifier is due to the presence of a large number of features common to the set of known active compounds rather than a single structural or substructural scaffold. We demonstrate that the Laplacian-modified naive Bayesian classifier trained with data from high-throughput docking is superior at identifying active compounds from a target database in comparison to conventional two-dimensional substructure search methods alone.

Design and Synthesis of Novel Indole β-Diketo Acid Derivatives as HIV-1 Integrase Inhibitors

Diketo acids such as S-1360 (1A) and L-731,988 (2) are potent and selective inhibitors of HIV-1 integrase (IN). A plethora of diketo acid-containing compounds have been claimed in patent literature without disclosing much biological activities and synthetic details (reviewed in Neamati, N. Exp. Opin. Ther. Pat. 2002, 12, 709-724). To establish a coherent structure-activity relationship among the substituted indole nucleus bearing a beta-diketo acid moiety, a series of substituted indole-beta-diketo acids (4a-f and 5a-e) were synthesized. All compounds tested showed anti-IN activity at low micromolar concentrations with varied selectivity against the strand transfer process. Three compounds, the indole-3-beta-diketo acids 5a and 5c, and the parent ester 9c, have shown an antiviral activity in cell-based assays. We further confirmed a keto-enolic structure in the 2,3-position of the diketo acid moiety of a representative compound (4c) using NMR and X-ray crystallographic analysis. Using this structure as a lead for all of our computational studies, we found that the title compounds extensively interact with the essential amino acids on the active site of IN.

A Three-dimensional Model of the Neprilysin 2 Active Site Based on the X-ray Structure of Neprilysin

Neprilysin 2 (NEP2), a recently identified member of the M13 subfamily of metalloproteases, shares the highest degree of homology with the prototypical member of the family neprilysin. Whereas the study of the in vitro enzymatic activity of NEP2 shows that it resembles that of NEP as it cleaves the same substrates often at the same amide bonds and binds the same inhibitory compounds albeit with different potencies, its physiological role remains elusive because of the lack of selective inhibitors. To aid in the design of these novel compounds and better understand the different inhibitory patterns of NEP and NEP2, the x-ray structure of NEP was used as a template to build a model of the NEP2 active site. The results of our modeling suggest that the overall structure of NEP2 closely resembles that of NEP. The model of the active site reveals a 97% sequence identity with that of NEP with differences located within the S'(2) subsite of NEP2 where Ser(133) and Leu(739) replace two glycine residues in NEP. To validate the proposed model, site-directed mutagenesis was performed on a series of residues of NEP2, mutants expressed in AtT20 cells, and their ability to bind various substrates and inhibitory compounds was tested. The results confirm the involvement of the conserved Arg(131) and Asn(567) in substrate binding and catalytic activity of NEP2 and further show that the modifications in its S'(2) pocket, particularly the presence therein of Leu(739), account for a number of differences in inhibitor binding between NEP and NEP2.

Validation of Model of Cytochrome P450 2D6: An in Silico Tool for Predicting Metabolism and Inhibition

There has been much interest in the development of a predictive model of cytochrome P450 2D6 particularly because this enzyme is involved in the oxidation of at least 50 drugs. Previously we have described the combined use of homology modeling and molecular docking to correctly position a range of substrates in the CYP2D6 active site with the known sites of metabolism above the heme. Here, our approach identifies correctly the site of metabolism of the atypical (no basic nitrogen) cytochrome P450 2D6 substrate, spirosulfonamide. The same method is used to screen a small compound database for cytochrome P450 2D6 inhibition. A database containing 33 compounds from the National Cancer Institute database was docked into our cytochrome P450 2D6 homology model using the program GOLDv2.0. Experimental IC50 values for the 33 compounds were determined; comparison with the corresponding docked scores revealed a correlation with a regression coefficient of r2 = 0.61 (q2 = 0.59). The method was able to discriminate between tight and weak binding compounds and correctly identified several novel inhibitors. The results therefore suggest that our approach, which combines homology modeling with molecular docking, has produced a useful predictive in silico tool for cytochrome P450 2D6 inhibition, which is best used as one filter in a multifilter database screen.

A peptide inhibitor of vascular adhesion protein-1 (VAP-1) blocks leukocyte-endothelium interactions under shear stress

Vascular adhesion protein-1 (VAP-1) is an endothelial adhesion molecule mediating leukocyte interactions with blood vessels during leukocyte extravasation. Molecularly VAP-1 is a cell-surface-expressed ecto-enzyme belonging to the group of semicarbazide-sensitive amine oxidases (SSAO; EC 2.4.6.3), which deaminate primary amines. Here we asked whether peptides displaying a suitable free amine group could be a substrate or inhibitor of SSAO and thus regulate VAP-1-mediated leukocyte adhesion. On the basis of a molecular model of VAP-1, we designed synthetic peptides that fit to the substrate channel of VAP-1. One of these lysine-containing peptides effectively inhibits VAP-1-dependent lymphocyte rolling and firm adhesion to primary endothelial cells under physiologically relevant shear conditions. The same peptide inhibits the SSAO activity of endothelial and recombinant VAP-1 in a selective and long-lasting manner. We also show that all enzymatically active VAP-1 is displayed on the cell surface. Our results suggest that, in addition to soluble amines, specific cell-surface-bound molecules containing free NH(2) groups in a suitable position may modulate the enzymatic activity of SSAO. Moreover, the inhibitory peptide diminishes leukocyte interactions with endothelial cells under conditions of shear, and thus it may be useful to treat inflammatory conditions.

Protein classification using comparative molecular interaction profile analysis system

We recently introduced a new molecular description factor, interaction profile Factor (IPF) that is useful for evaluating molecular interactions. IPF is a data set of interaction energies calculated by the Comparative Molecular Interaction Profile Analysis system (CoMIPA). CoMIPA utilizes AutoDock 3.0 docking program, and the system has shown to be a powerful tool in clustering the interacting properties between small molecules and proteins. In this report, we describe the application of CoMIPA for protein clustering. A sample set of 15 proteins that share less than 20% homology and have no common functional motifs in primary structure were chosen. Using CoMIPA, we were able to cluster proteins that bound to the same small molecule. Other structural homology-based clustering programs such as PSI-BLAST or PFAM were unable to achieve the same classification. The results are striking because it is difficult to find any common features in the active sites of these proteins that share the same ligand. CoMIPA adds new dimensions for protein classification and has the potential to be a helpful tool in predicting and analyzing molecular interactions.

On the modeling of snake venom serine proteinase interactions with benzamidine-based thrombin inhibitors

Pit viper venoms contain a number of serine proteinases that exhibit one or more thrombin-like activities on fibrinogen and platelets, this being the case for the kinin-releasing and fibrinogen-clotting KN-BJ from the venom of Bothrops jararaca. A three-dimensional structural model of the KN-BJ2 serine proteinase was built by homology modeling using the snake venom plasminogen activator TSV-PA as a major template and porcine kallikrein as additional structural support. A set of intrinsic buried waters was included in the model and its behavior under dynamic conditions was molecular dynamics simulated, revealing a most interesting similarity pattern to kallikrein. The benzamidine-based thrombin inhibitors alpha-NAPAP, 3-TAPAP, and 4-TAPAP were docked into the refined model, allowing for a more insightful functional characterization of the enzyme and a better understanding of the reported comparatively low affinity of KN-BJ2 toward those inhibitors.

Combining pharmacophore search, automated docking, and molecular dynamics simulations as a novel strategy for flexible docking. Proof of concept: docking of arginine-glycine-aspartic acid-like compounds into the alphavbeta3 binding site

A novel and highly efficient flexible docking approach is presented where the conformations (internal degrees of freedom) and orientations (external degrees of freedom) of the ligands are successively considered. This hybrid method takes advantage of the synergistic effects of structure-based and ligand-based drug design techniques. Preliminary antagonist-derived pharmacophore determination provides the postulated bioactive conformation. Subsequent docking of this pharmacophore to the receptor crystal structure results in a postulated pharmacophore/receptor binding mode. Pharmacophore-oriented docking of antagonists is subsequently achieved by matching ligand interacting groups with pharmacophore points. Molecular dynamics in water refines the proposed complexes. To validate the method, arginine-glycine-aspartic acid (RGD) containing peptides, pseudopeptides, and RGD-like antagonists were docked to the crystal structure of alphavbeta3 holoprotein and apoprotein. The proposed directed docking was found to be more accurate, faster, and less biased with respect to the protein structure (holo and apoprotein) than DOCK, Autodock, and FlexX docking methods. The successful docking of an antagonist recently cocrystallized with the receptor to both apo and holoprotein is particularly appealing. The results summarized in this report illustrated the efficiency of our light CoMFA/rigid body docking hybrid method.

Modeling the E. coli 4-hydroxybenzoic acid oligoprenyltransferase ( ubiA transferase) and characterization of potential active sites

4-hydroxybenzoate oligoprenyltransferase of E. coli, encoded in the gene ubiA, is an important key enzyme in the biosynthetic pathway to ubiquinone. It catalyzes the prenylation of 4-hydroxybenzoic acid in position 3 using an oligoprenyl diphosphate as a second substrate. Up to now, no X-ray structure of this oligoprenyltransferase or any structurally related enzyme is known. Knowledge of the tertiary structure and possible active sites is, however, essential for understanding the catalysis mechanism and the substrate specificity. With homology modeling techniques, secondary structure prediction tools, molecular dynamics simulations, and energy optimizations, a model with two putative active sites could be created and refined. One active site selected to be the most likely one for the docking of oligoprenyl diphosphate and 4-hydroxybenzoic acid is located near the N-terminus of the enzyme. It is widely accepted that residues forming an active site are usually evolutionary conserved within a family of enzymes. Multiple alignments of a multitude of related proteins clearly showed 100% conservation of the amino acid residues that form the first putative active site and therefore strongly support this hypothesis. However, an additional highly conserved region in the amino acid sequence of the ubiA enzyme could be detected, which also can be considered a putative (or rudimentary) active site. This site is characterized by a high sequence similarity to the aforementioned site and may give some hints regarding the evolutionary origin of the ubiA enzyme. Semiempirical quantum mechanical PM3 calculations have been performed to investigate the thermodynamics and kinetics of the catalysis mechanism. These results suggest a near S(N)1 mechanism for the cleavage of the diphosphate ion from the isoprenyl unit. The 4-hydroxybenzoic acid interestingly appears not to be activated as benzoate anion but rather as phenolate anion to allow attack of the isoprenyl cation to the phenolate, which appeared to be the rate limiting step of the whole process according to our quantum chemical calculations. Our models are a basis for developing inhibitors of this enzyme, which is crucial for bacterial aerobic metabolism. [figure: see text]. Structure of the model of ubiA oligoprenyltransferase derived from the photosynthetic reaction center (1PRC). Putative active amino acid residues and substrates are shown as capped sticks to describe their location and geometry in the putative active sites. The violet spheres identify Mg2+.

Identification of 1,5-Naphthyridine Derivatives as a Novel Series of Potent and Selective TGF-β Type I Receptor Inhibitors

Optimization of the screening hit 1 led to the identification of novel 1,5-naphthyridine aminothiazole and pyrazole derivatives, which are potent and selective inhibitors of the transforming growth factor-beta type I receptor, ALK5. Compounds 15 and 19, which inhibited ALK5 autophosphorylation with IC50 = 6 and 4 nM, respectively, showed potent activities in both binding and cellular assays and exhibited selectivity over p38 mitogen-activated protein kinase. The X-ray crystal structure of 19 in complex with human ALK5 is described, confirming the binding mode proposed from docking studies.

Computational approaches to understand alpha-conotoxin interactions at neuronal nicotinic receptors

Recent and increasing use of computational tools in the field of nicotinic receptors has led to the publication of several models of ligand-receptor interactions. These models are all based on the crystal structure at 2.7 A resolution of a protein related to the extracellular N-terminus of nicotinic acetylcholine receptors (nAChRs), the acetylcholine binding protein. In the absence of any X-ray or NMR information on nAChRs, this new structure has provided a reliable alternative to study the nAChR structure. We are now able to build homology models of the binding domain of any nAChR subtype and fit in different ligands using docking programs. This strategy has already been performed successfully for the docking of several nAChR agonists and antagonists. This minireview focuses on the interaction of alpha-conotoxins with neuronal nicotinic receptors in light of our new understanding of the receptor structure. Computational tools are expected to reveal the molecular recognition mechanisms that govern the interaction between alpha-conotoxins and neuronal nAChRs at the molecular level. An accurate determination of their binding modes on the neuronal nAChR may allow the rational design of alpha-conotoxin-based ligands with novel nAChR selectivity.

Combination of a Naive Bayes Classifier with Consensus Scoring Improves Enrichment of High-Throughput Docking Results

We have previously shown that a machine learning technique can improve the enrichment of high-throughput docking (HTD) results. In the previous cases studied, however, the application of a naive Bayes classifier failed to improve enrichment for instances where HTD alone was unable to generate an acceptable enrichment. We present here a protocol to rescue poor docking results a priori using a combination of rank-by-median consensus scoring and naive Bayesian categorization.

3D-QSAR and docking studies on 4-anilinoquinazoline and 4-anilinoquinoline epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors

The overexpression and/or mutation of the epidermal growth factor receptor (EGFR) tyrosine kinase has been observed in many human solid tumors, and is under intense investigation as a novel anticancer molecular target. Comparative 3D-QSAR analyses using different alignments were undertaken employing comparative molecular field analysis (CoMFA) and comparative molecular similarity analysis (CoMSIA) for 122 anilinoquinazoline and 50 anilinoquinoline inhibitors of EGFR kinase. The SYBYL multifit alignment rule was applied to three different conformational templates, two obtained from a MacroModel Monte Carlo conformational search, and one from the bound conformation of erlotinib in complex with EGFR in the X-ray crystal structure. In addition, a flexible ligand docking alignment obtained with the GOLD docking program, and a novel flexible receptor-guided consensus dynamics alignment obtained with the DISCOVER program in the INSIGHTII modeling package were also investigated. 3D-QSAR models with q2 values up to 0.70 and r2 values up to 0.97 were obtained. Among the 4-anilinoquinazoline set, the q2 values were similar, but the ability of the different conformational models to predict the activities of an external test set varied considerably. In this regard, the model derived using the X-ray crystallographically determined bioactive conformation of erlotinib afforded the best predictive model. Electrostatic, hydrophobic and H-bond donor descriptors contributed the most to the QSAR models of the 4-anilinoquinazolines, whereas electrostatic, hydrophobic and H-bond acceptor descriptors contributed the most to the 4-anilinoquinoline QSAR, particularly the H-bond acceptor descriptor. A novel receptor-guided consensus dynamics alignment has also been introduced for 3D-QSAR studies. This new alignment method may incorporate to some extent ligand-receptor induced fit effects into 3D-QSAR models.

Computational studies of new potential antimalarial compounds--stereoelectronic complementarity with the receptor

One of the most important pharmacological mechanisms of antimalarial action is the inhibition of the aggregation of hematin into hemozoin. We present a group of new potential antimalarial molecules for which we have performed a DFT study of their stereoelectronic properties. Additionally, the same calculations were carried out for the two putative drug receptors involved in the referred activity, i.e., hematin mu-oxo dimer and hemozoin. A complementarity between the structural and electronic profiles of the planned molecules and the receptors can be observed. A docking study of the new compounds in relation to the two putative receptors is also presented, providing a correlation with the defined electrostatic complementarity.

CoMFA and docking study of novel estrogen receptor subtype selective ligands

We present the results from a Comparative Molecular Field Analysis (CoMFA) and docking study of a diverse set of 36 estrogen receptor ligands whose relative binding affinities (RBA) with respect to 17beta-Estradiol were available in both isoforms of the nuclear estrogen receptors (ER alpha, ER beta). Initial CoMFA models exhibited a correlation between the experimental relative binding affinities and the molecular steric and electrostatic fields; ER alpha: r2 = 0.79, q2 = 0.44 ER beta: r2 = 0.93, q2 = 0.63. Addition of the solvation energy of the isolated ligand improved the predictive nature of the ER beta model initially; r2 = 0.96, q2 = 0.70 but upon rescrambling of the data-set and reselecting the training set at random, inclusion of the ligand solvation energy was found to have little effect on the predictive nature of the CoMFA models. The ligands were then docked inside the ligand binding domain (LBD) of both ER alpha and ER beta utilizing the docking program Gold, after-which the program CScore was used to rank the resulting poses. Inclusion of both the Gold and CScore scoring parameters failed to improve the predictive ability of the original CoMFA models. The subtype selectivity expressed as RBA(ER alpha/ER beta) of the test sets was predicted using the most predictive CoMFA models, as illustrated by the cross-validated r2. In each case the most selective ligands were ranked correctly illustrating the utility of this method as a prescreening tool in the development of novel estrogen receptor subtype selective ligands.

Theoretical analysis of peptidyl alpha-ketoheterocyclic inhibitors of human neutrophil elastase: Insight into the mechanism of inhibition and the application of QM/MM calculations in structure-based drug design

It has been suggested from QSAR data (P. D. Edwards, D. J. Wolanin, D.A. Andisik and M. W. Davis, J. Med. Chem., 1995, 38, 76) that the inhibition of elastase by peptidyl alpha-ketoheterocyclic inhibitors can occur in two ways, the less potent inhibitors forming a non-bonded Michaelis complex and the more potent set a covalently bonded enzyme-substrate intermediate. We report QM/MM studies of both binding and reactivity that confirm these findings, showing that the activity of the least potent set of inhibitors correlates with the calculated binding energy, and that of the more potent set correlates with the stability of the intermediate. These calculations show that QM/MM methods can be successfully employed to understand complicated structure-activity relationships and might be employed in the design and assessment of new inhibitors.

Design of EGFR kinase inhibitors: a ligand-based approach and its confirmation with structure-based studies

Three-dimensional quantitative structure-activity relationship (3D-QSAR) models were developed for 100 anilinoquinazolines, inhibiting epidermal growth factor receptor (EGFR) kinase. The studies included molecular field analysis (MFA) and receptor surface analysis (RSA). The cross-validated r2 (r2cv) values are 0.81 and 0.79 for MFA and RSA, respectively. The predictive ability of these models was validated by 28 test set molecules. The results of the best QSAR model were further compared with structure-based investigations using docking studies with the crystal structure of EGFR kinase domain. The results helped to understand the nature of substituents at the 6- and 7-positions, thereby providing new guidelines for the design of novel inhibitors.

Characteristics of the Plasmodium falciparum PK5 ATP-binding site: implications for the design of novel antimalarial agents

Increasing worldwide resistance of Plasmodium falciparum (P. falciparum) to traditional chemotherapy strategies such as chloroquine and mefloquine demonstrates the urgent need for the discovery of novel chemotherapeutic agents in the fight against malaria. The recent discovery of P. falciparum Protein Kinase 5 (PfPK5) invites the possibility of selectively targeting the life cycle of P. falciparum in order to prevent cerebral malaria. PfPK5 bears a high degree of sequence identity (>58%) to a structurally conserved family of mammalian kinases known as the cyclin-dependent kinases (CDKs). The CDKs are the key regulatory elements governing the ordered progression of the mammalian cell cycle. With numerous X-ray crystal structures of CDK2 to provide a structural template, here we present a three-dimensional structural model of PfPK5 constructed using computer-based homology modeling techniques. Our model was used to compare the ATP binding site of PfPK5 with that of the mammalian kinase CDK2. Furthermore, kinase-ligand interactions of PfPK5 with known inhibitors were investigated and compared to available crystal structures of CDK2 with inhibitors bound. The focus of the study is to identify similarities and differences between the ATP binding sites of the two kinases that can be exploited for future rational drug design.

Relative free energy of binding and binding mode calculations of HIV-1 RT inhibitors based on dock-MM-PB/GS

Tetrahydroimidazo-[4,5,l-jk][1,4]-benzodiazepin-2-(1H)-one (TIBO) derivatives are important nonnucleoside human immunodeficiency virus-1 reverse transcriptase inhibitors (NNRTI). Several TIBO derivatives have shown high potency to inhibit reverse transcriptase (RT) and one (Tivirapine) has entered into clinical trials. The free energy of binding (FEB) is a numerical way to express the binding affinity of a ligand to its receptor and has been applied in screening candidates in rational drug design. In this work, the FEB of 42 TIBOs in RT was studied. Relative FEB is expressed in the form of a linear combination of vdW, electrostatic, solvation, and nonpolar solvation energy terms. The predicted FEB activity of the TIBOs studied has a good correlation (r(2) = 0.8680, q(2) = 0.8298) with respect to the experimental activity (pIC(50)). Based on the data reported here, the Finite Difference Poisson Boltzmann with a Gaussian Smooth Dielectric Constant Function method (PB/GS) solvation energy term is very important in predicting the binding affinity of TIBOs in RT. In summary, the Dock-Molecular Mechanics (MM)-PB/GS method is a promising technique in predicting ligand/receptor binding affinity and it can be used to screen relatively large sets of molecules in a reasonable amount of computer time.

Assessing scoring functions for protein-ligand interactions

An assessment of nine scoring functions commonly applied in docking using a set of 189 protein-ligand complexes is presented. The scoring functions include the CHARMm potential, the scoring function DrugScore, the scoring function used in AutoDock, the three scoring functions implemented in DOCK, as well as three scoring functions implemented in the CScore module in SYBYL (PMF, Gold, ChemScore). We evaluated the abilities of these scoring functions to recognize near-native configurations among a set of decoys and to rank binding affinities. Binding site decoys were generated by molecular dynamics with restraints. To investigate whether the scoring functions can also be applied for binding site detection, decoys on the protein surface were generated. The influence of the assignment of protonation states was probed by either assigning "standard" protonation states to binding site residues or adjusting protonation states according to experimental evidence. The role of solvation models in conjunction with CHARMm was explored in detail. These include a distance-dependent dielectric function, a generalized Born model, and the Poisson equation. We evaluated the effect of using a rigid receptor on the outcome of docking by generating all-pairs decoys ("cross-decoys") for six trypsin and seven HIV-1 protease complexes. The scoring functions perform well to discriminate near-native from misdocked conformations, with CHARMm, DOCK-energy, DrugScore, ChemScore, and AutoDock yielding recognition rates of around 80%. Significant degradation in performance is observed in going from decoy to cross-decoy recognition for CHARMm in the case of HIV-1 protease, whereas DrugScore and ChemScore, as well as CHARMm in the case of trypsin, show only small deterioration. In contrast, the prediction of binding affinities remains problematic for all of the scoring functions. ChemScore gives the highest correlation value with R(2) = 0.51 for the set of 189 complexes and R(2) = 0.43 for the set of 116 complexes that does not contain any of the complexes used to calibrate this scoring function. Neither a more accurate treatment of solvation nor a more sophisticated charge model for zinc improves the quality of the results. Improved modeling of the protonation states, however, leads to a better prediction of binding affinities in the case of the generalized Born and the Poisson continuum models used in conjunction with the CHARMm force field.

Sequence divergence of Entamoeba histolytica tubulin is responsible for its altered tertiary structure

Atypical microtubular structures of the protozoan parasite Entamoeba histolytica (Eh) have been attributed to amino acid sequence divergence of Eh tubulin. To investigate if this sequence divergence leads to significant differences in the tertiary structure of the Eh alphabeta-tubulin heterodimer, we have modeled alphabeta-tubulin heterodimer of Eh based on the crystal structure of mammalian tubulin. The predicted 3D homology model exhibits an overall resemblance with the known crystal structure of mammalian tubulin except for the 16 residue long carboxy terminal region of Eh beta-tubulin. We propose that this C-terminal region may provide steric hindrance in the polymerization of Eh alphabeta-tubulin for microtubule formation. Using docking studies, we have identified the binding sites for different microtubule specific drugs on Eh beta-tubulin. Our model provides a rational framework, both for understanding the contribution of Eh beta-tubulin C-terminal region to alphabeta-tubulin polymerization and design of new anti-protozoan drugs in order to control amoebiasis.

Creating Artificial Binding Pocket Boundaries To Improve the Efficiency of Flexible Ligand Docking

Traditionally, algorithms for binding site characterization or identification focus on the problem of identifying atoms within a macromolecule that might be responsible for ligand binding. In this manuscript, we focus on the binding pocket problem from a different perspective as a challenge of calculating an artificial binding pocket boundary that is sufficient to isolate binding pocket volume. The approach involves the calculation of a macromolecule encapsulating surface (MES) that separates binding pocket volume from outside space. We show that the MES can be used to increase the efficiency of flexible docking as implemented in AutoDock 3.0. The most significant improvement in docking efficiency is seen when the entire protein is searched and results show additional support for the use of AutoDock, in and of itself, as a feasible tool for binding-site identification for cases in which a protein ligand is known.

Automated protein structure homology modeling: a progress report

Understanding the molecular function of proteins is greatly enhanced by insights gained from their three-dimensional structures. Since experimental structures are only available for a small fraction of proteins, computational methods for protein structure modeling play an increasingly important role. Comparative protein structure modeling is currently the most accurate method, yielding models suitable for a wide spectrum of applications, such as structure-guided drug development or virtual screening. Stable and reliable automated prediction pipelines have been developed to apply large-scale comparative modeling to whole genomes or entire sequence databases. Model repositories give access to these annotated and evaluated models. In this review, we will discuss recent developments in automated comparative modeling and provide selected examples illustrating the use of homology models.