Elucidating the inhibiting mode of AHPBA derivatives against HIV-1 protease and building predictive 3D-QSAR models

The recent application of 3D-QSAR and docking studies to novel HIV-protease inhibitor drug discovery

INTRODUCTION

Despite the availability of FDA approved inhibitors of HIV protease, numerous efforts are still ongoing to achieve 'near-perfect' drugs devoid of characteristic adverse side effects, toxicities, and mutational resistance. While experimental methods have been plagued with huge consumption of time and resources, there has been an incessant shift towards the use of computational simulations in HIV protease inhibitor drug discovery.

AREAS COVERED

Herein, the authors review the numerous applications of 3D-QSAR modeling methods over recent years relative to the design of new HIV protease inhibitors from a series of experimentally derived compounds. Also, the augmentative contributions of molecular docking are discussed.

EXPERT OPINION

Efforts to optimize 3D QSAR and molecular docking for HIV-1 drug discovery are ongoing, which could further incorporate inhibitor motions at the active site using molecular dynamics parameters. Also, highly predictive machine learning algorithms such as random forest, K-means, decision trees, linear regression, hierarchical clustering, and Bayesian classifiers could be employed.

Design and prediction of new anticoagulants as a selective Factor IXa inhibitor via three-dimensional quantitative structure-property relationships of amidinobenzothiophene derivatives

Factor IXa (FIXa), a blood coagulation factor, is specifically inhibited at the initiation stage of the coagulation cascade, promising an excellent approach for developing selective and safe anticoagulants. Eighty-four amidinobenzothiophene antithrombotic derivatives targeting FIXa were selected to establish three-dimensional quantitative structure–activity relationship (3D-QSAR) and three-dimensional quantitative structure–selectivity relationship (3D-QSSR) models using comparative molecular field analysis and comparative similarity indices analysis methods. Internal and external cross-validation techniques were investigated as well as region focusing and bootstrapping. The satisfactory q² values of 0.753 and 0.770, and r² values of 0.940 and 0.965 for 3D-QSAR and 3D-QSSR, respectively, indicated that the models are available to predict both the inhibitory activity and selectivity on FIXa against Factor Xa, the activated status of Factor X. This work revealed that the steric, hydrophobic, and H-bond factors should appropriately be taken into account in future rational design, especially the modifications at the 2′-position of the benzene and the 6-position of the benzothiophene in the R group, providing helpful clues to design more active and selective FIXa inhibitors for the treatment of thrombosis. On the basis of the three-dimensional quantitative structure–property relationships, 16 new potent molecules have been designed and are predicted to be more active and selective than Compound 33, which has the best activity as reported in the literature.

Bioinformatics approach to predicting HIV drug resistance

The emergence of drug resistance remains one of the most challenging issues in the treatment of HIV-1 infection. The extreme replication dynamics of HIV facilitates its escape from the selective pressure exerted by the human immune system and by the applied combination drug therapy. This article reviews computational methods whose combined use can support the design of optimal antiretroviral therapies based on viral genotypic and phenotypic data. Genotypic assays are based on the analysis of mutations associated with reduced drug susceptibility, but are difficult to interpret due to the numerous mutations and mutational patterns that confer drug resistance. Phenotypic resistance or susceptibility can be experimentally evaluated by measuring the inhibition of the viral replication in cell culture assays. However, this procedure is expensive and time consuming.

Design, biologic evaluation, and SAR of novel pseudo-peptide incorporating benzheterocycles as HIV-1 protease inhibitors

A series of novel HIV-1 protease inhibitors based on the (hydroxyethylamino)-sulfonamide isostere incorporating substituted phenyls and benzheterocycle derivatives bearing rich hydrogen bonding acceptors as P(2) ligands were synthesized. Prolonged chain linking the benzhereocycle to the carbonyl group resulted in partial loss of binding affinities. Introduction of a small alkyl substituent with appropriate size to the -CH2- of P(1)-P(2) linkage as a side chain resulted in improved inhibitory potency, and in this study, isopropyl was the best side chain. Replacement of the isobutyl substituent at P(1)'group with phenyl substituent decreased the inhibitory potency. One of the most potent inhibitor, compound 23 showing high affinity to HIV-1 protease with an IC(50) value of 5 nM, also exhibited good anti-SIV activity (EC(50) = 0.8 microM) with low toxicity (TC(50) > 100 microM). The flexible docking of inhibitor 23 to HIV-1 protease active site rationalized the interactions with protease.

Combined 3D-QSAR modeling and molecular docking study on quinoline derivatives as inhibitors of P-selectin

P-selectin is a promising target for developing novel atherosclerosis drugs. To understand the structure-activity correlation of quinolines-based P-selectin inhibitors, we have carried out a combined molecular docking and three-dimensional quantitative structure-activity relationship (3D-QSAR) modeling study. The study has resulted in two types of satisfactory 3D-QSAR models, including the CoMFA model (r(2), 0.863; q(2), 0.589) and CoMSIA model (r(2), 0.866; q(2), 0.636), to predict the biological activity of new compounds. The detailed microscopic structures of P-selectin binding with inhibitors have been studied by molecular docking. We have also developed docking based 3D-QSAR models (CoMFA with r(2), 0.934; q(2), 0.591; CoMSIA with r(2), 0.896; q(2), 0.573). The contour maps obtained from the 3D-QSAR models in combination with the docked binding structures help to better interpret the structure-activity relationship. All of the structural insights obtained from both the 3D-QSAR contour maps and molecular docking are consistent with the available experimental activity data. The satisfactory results strongly suggest that the developed 3D-QSAR models and the obtained P-selectin-inhibitor binding structures are reasonable for the prediction of the activity of new inhibitors and in future drug design.

4D-fingerprints, universal QSAR and QSPR descriptors

An elusive goal in the field of chemoinformatics and molecular modeling has been the generation of a set of descriptors that, once calculated for a molecule, may be used in a wide variety of applications. Since such universal descriptors are generated free from external constraints, they are inherently independent of the data set in which they are employed. The realization of a set of universal descriptors would significantly streamline such chemoinformatics tasks as virtual high-throughout screening (VHTS) and toxicity profiling. The current study reports the derivation and validation of a potential set of universal descriptors, referred to as the 4D-fingerprints. The 4D-fingerprints are derived from the 4D-molecular similarity analysis. To evaluate the applicability of the 4D-fingerprints as universal descriptors, they are used to generate descriptive QSAR models for 5 independent training sets. Each of the training sets has been analyzed previously by several varying QSAR methods, and the results of the models generated using the 4D-fingerprints are compared to the results of the previous QSAR analyses. It was found that the models generated using the 4D-fingerprints are comparable in quality, based on statistical measures of fit and test set prediction, to the previously reported models for the other QSAR methods. This finding is particularly significant considering the 4D-fingerprints are generated independent of external constraints such as alignment, while the QSAR methods used for comparison all require an alignment analysis.

Design, synthesis, and biological evaluation of novel 4-hydroxypyrone derivatives as HIV-1 protease inhibitors

Twenty-four 4-hydroxypyrone derivatives were synthesized with a facile synthetic method to develop novel HIV protease inhibitors. Most of them were shown to display good antiviral activities in SIV-infected CEM cells. The introduction of alpha-naphthylmethyl group to C-6 of 5,6-dihydropyran-2-ones led to an effective antiviral compound that showed an EC(50) value at 1.7 microM with a therapeutic index of 46.

Molecular docking and 3D-QSAR studies on the binding mechanism of statine-based peptidomimetics with β-secretase

beta-Secretase is an important protease in the pathogenesis of Alzheimer's disease. Some statine-based peptidomimetics show inhibitory activities to the beta-secretase. To explore the inhibitory mechanism, molecular docking and three-dimensional quantitative structure-activity relationship (3D-QSAR) studies on these analogues were performed. The Lamarckian Genetic Algorithm (LGA) was applied to locate the binding orientations and conformations of the peptidomimetics with the beta-secretase. A good correlation between the calculated binding free energies and the experimental inhibitory activities suggests that the identified binding conformations of these potential inhibitors are reliable. Based on the binding conformations, highly predictive 3D-QSAR models were developed with q(2) values of 0.582 and 0.622 for CoMFA and CoMSIA, respectively. The predictive abilities of these models were validated by some compounds that were not included in the training set. Furthermore, the 3D-QSAR models were mapped back to the binding site of the beta-secretase, to get a better understanding of vital interactions between the statine-based peptidomimetics and the protease. Both the CoMFA and the CoMSIA field distributions are in well agreement with the structural characteristics of the binding groove of the beta-secretase. Therefore, the final 3D-QSAR models and the information of the inhibitor-enzyme interaction would be useful in developing new drug leads against Alzheimer's disease.

Quantitative structure-activity relationship by CoMFA for cyclic urea and nonpeptide-cyclic cyanoguanidine derivatives on wild type and mutant HIV-1 protease

3D-QSAR studies using the Comparative Molecular Field Analysis (CoMFA) methodology were conducted to predict the inhibition constants, K(i), and the inhibitor concentrations, IC90 of 127 symmetrical and unsymmetrical cyclic urea and cyclic cyanoguanidine derivatives containing different substituent groups such as: benzyl, isopropyl, 4-hydroxybenzyl, ketone, oxime, pyrazole, imidazole, triazole and having anti-HIV-1 protease activities. A significant cross-validated correlation coefficient (q2) of 0.63 and a fitted correlation coefficient r2 of 0.70 were obtained, indicating that the models can predict the anti-protease activity from poorly to highly active compounds reliably. The best predictions were obtained for: XV643 (predicted log 1/K(i) = 9.86), a 3,5-dimethoxy-benzyl cyclic urea derivate (molec60, predicted log 1/K(i) = 8.57) and a benzyl cyclic urea derivate (molec 61, predicted log 1/IC90 = 6.87). Using the CoMFA method, we also predicted the biological activity of 14 cyclic urea derivatives that inhibit the HIV-1 protease mutants V82A, V82I and V82F. The predicted biological activities of the: (i) XNO63 (inhibitory activity on the mutant HIV-1 PR V82A), (ii) SB570 (inhibiting the mutant HIV-1 PR V82I) and also (iii) XV652 (during the interaction with the mutant HIV-1 PR V82F) were in good agreement with the experimental values.

Molecular docking and 3D QSAR studies on 1-amino-2-phenyl-4-(piperidin-1-yl)-butanes based on the structural modeling of human CCR5 receptor

In the present study, we have used an approach combining protein structure modeling, molecular dynamics (MD) simulation, automated docking, and 3D QSAR analyses to investigate the detailed interactions of CCR5 with their antagonists. Homology modeling and MD simulation were used to build the 3D model of CCR5 receptor based on the high-resolution X-ray structure of bovine rhodopsin. A series of 64 CCR5 antagonists, 1-amino-2-phenyl-4-(piperidin-1-yl)-butanes, were docked into the putative binding site of the 3D model of CCR5 using the docking method, and the probable interaction model between CCR5 and the antagonists were obtained. The predicted binding affinities of the antagonists to CCR5 correlate well with the antagonist activities, and the interaction model could be used to explain many mutagenesis results. All these indicate that the 3D model of antagonist-CCR5 interaction is reliable. Based on the binding conformations and their alignment inside the binding pocket of CCR5, three-dimensional structure-activity relationship (3D QSAR) analyses were performed on these antagonists using comparative molecular field analysis (CoMFA) and comparative molecular similarity analysis (CoMSIA) methods. Both CoMFA and CoMSIA provide statistically valid models with good correlation and predictive power. The q(2)(r(cross)(2)) values are 0.568 and 0.587 for CoMFA and CoMSIA, respectively. The predictive ability of these models was validated by six compounds that were not included in the training set. Mapping these models back to the topology of the active site of CCR5 leads to a better understanding of antagonist-CCR5 interaction. These results suggest that the 3D model of CCR5 can be used in structure-based drug design and the 3D QSAR models provide clear guidelines and accurate activity predictions for novel antagonist design.

A simple clustering technique to improve QSAR model selection and predictivity: application to a receptor independent 4D-QSAR analysis of cyclic urea derived inhibitors of HIV-1 protease

A training set of 50 tetrahydropyrimidine-2-one based inhibitors of HIV-1 protease, for which the -log K(i) values were measured, was used to construct receptor independent 4D-QSAR models. A novel clustering technique was employed to facilitate and improve model selection as well as test set predictions. Following the manifold model theory, five unique models were chosen by the clustering algorithm (q(2) = 0.81-0.84). The models were used to map the atom type morphology of the inhibitor binding site of HIV-1 protease as well as to predict the potencies (-log K(i)) of 10 test set compounds. The rank-difference correlation coefficient was used to evaluate the quality of the test set predictions, which was improved from 0.39 to 0.68 when the clustering technique was applied. The set of five models, collectively, identify the important binding characteristics of the HIV protease receptor site. This study demonstrates that the selected simple clustering technique provides a discrete algorithm for model selection, as well as improving the quality of test set, or unknown, compound prediction as determined by the rank-difference correlation coefficient.

3D-QSAR illusions

3D-QSAR is typically used to construct models (1) to predict activities, (2) to illustrate significant regions, and (3) to provide insight into possible interactions. To the contrary, examples are described herein which make it clear that the predictivity of such models remains elusive, that so-called significant regions are subject to the vagaries of alignment, and that the nature of possible interactions heavily depends on the eye of the beholder. Although great strides have been made in the imaginative use of 3D-descriptors, 3D-QSAR remains largely a retrospective analytical tool. The arbitrary nature of both the alignment paradigm and atom description lends itself to capricious models, which in turn can lead to distorted conclusions. Despite these illusionary pitfalls, predictions can be enhanced when the test set is bounded by the descriptor space represented in the training set. Interpretation of significant interaction regions becomes more meaningful when alignment is constrained by a binding site. Correlations obtained with a variety of atom descriptors suggest choosing useful ones, in particular, in guiding synthetic effort.

Correlation between the predicted and the observed biological activity of the symmetric and nonsymmetric cyclic urea derivatives used as HIV-1 protease inhibitors. A 3D-QSAR-CoMFA method for new antiviral drug design

The predicted inhibition constant (Ki) and the predicted inhibitor concentration (IC90) of the HIV-1 protease (HIV- 1 PR) inhibitors: symmetric and nonsymmetric - benzyl, ketone, oxime, pyrazole, imidazole, and triazole cyclic urea derivatives, were obtained by the 3D-CoMFA (Comparative Molecular Field Analysis) method. The CoMFA statistical parameters: cross-validate correlation coefficient (q2), higher than 0.5, and the fitted correlation coefficient (r2), higher than 0.90 validated the predicted biological activities. The best predictions were found for the trifluoromethyl ketoxime derivative (log 1/Ki predict = 8.42), the m-pyridineCH2 pyrazole derivative (log 1/Ki predict = 9.77) and the 1,2,3 triazole derivative (log 1/Ki predict = 7.03). We attempted to design a new potent HIV-1 protease inhibitor by addition of o-benzyl to the (p-HOPhCH2) pyrazole 12f derivative inhibitor. A favorable steric area surrounded the o-benzyl, suggesting a possible new potent HIV-1 protease inhibitor.

Improved prediction of HIV-1 protease-inhibitor binding energies by molecular dynamics simulations

BACKGROUND

The accurate prediction of enzyme-substrate interaction energies is one of the major challenges in computational biology. This study describes the improvement of protein-ligand binding energy prediction by incorporating protein flexibility through the use of molecular dynamics (MD) simulations.

RESULTS

Docking experiments were undertaken using the program AutoDock for twenty-five HIV-1 protease-inhibitor complexes determined by x-ray crystallography. Protein-rigid docking without any dynamics produced a low correlation of 0.38 between the experimental and calculated binding energies. Correlations improved significantly for all time scales of MD simulations of the receptor-ligand complex. The highest correlation coefficient of 0.87 between the experimental and calculated energies was obtained after 0.1 picoseconds of dynamics simulation.

CONCLUSION

Our results indicate that relaxation of protein complexes by MD simulation is useful and necessary to obtain binding energies that are representative of the experimentally determined values.

Development of biologically active compounds by combining 3D QSAR and structure-based design methods

One of the major challenges in computational approaches to drug design is the accurate prediction of the binding affinity of novel biomolecules. In the present study an automated procedure which combines docking and 3D-QSAR methods was applied to several drug targets. The developed receptor-based 3D-QSAR methodology was tested on several sets of ligands for which the three-dimensional structure of the target protein has been solved--namely estrogen receptor, acetylcholine esterase and protein-tyrosine-phosphatase 1B. The molecular alignments of the studied ligands were determined using the docking program AutoDock and were compared with the X-ray structures of the corresponding protein-ligand complexes. The automatically generated protein-based ligand alignment obtained was subsequently taken as basis for a comparative field analysis applying the GRID/GOLPE approach. Using GRID interaction fields and applying variable selection procedures, highly predictive models were obtained. It is expected that concepts from receptor-based 3D QSAR will be valuable tools for the analysis of high-throughput screening as well as virtual screening data.

Inhibitory mode of 1,5-diarylpyrazole derivatives against cyclooxygenase-2 and cyclooxygenase-1: molecular docking and 3D QSAR analyses

The Lamarckian genetic algorithm of AutoDock 3.0 has been employed to dock 40 1,5-diarylpyrazole class compounds into the active sites of cyclooxygenase-2 (COX-2) and cyclooxygenase-1 (COX-1). The binding models were demonstrated in the aspects of inhibitor's conformation, subsite interaction, and hydrogen bonding. The data of geometrical parameters and RMSD values compared with the known inhibitor, SC-558 (43), show that these inhibitors interact respectively with COX-2 and COX-1 in a very similar way. The r(2) values of 0.648 for COX-2 and 0.752 for COX-1 indicate that the calculated binding free energies correlate well with the inhibitory activities. The structural and energetic differences in inhibitory potencies of 1,5-diarylpyrazoles were reasonably explored, and the COX-2/COX-1 selectivity was demonstrated by the three-dimensional (3D) interaction models of inhibitors complexing with these two enzymes. Using the binding conformations of 1,5-diarylpyrazoles, consistent and highly predictive 3D quantitative structure-activity relationship (QSAR) models were developed by performing comparative molecular field analyses (CoMFA) and comparative molecular similarity analyses (CoMSIA). The q(2) values are 0.635 and 0.641 for CoMFA and CoMSIA models, respectively. The predictive ability of these models was validated by SC-558 (43) and a set of 10 other compounds that were not included in the training set. Mapping these models back to the topology of the active site of COX-2 leads to a better understanding of vital diarylpyrazole compounds and COX-2 interactions. Structure-based investigations and the final 3D QSAR results provided possible guidelines and accurate activity predictions for novel inhibitor design.