Characterizing Binding of Small Molecules. II. Evaluating the Potency of Small Molecules to Combat Resistance Based on Docking Structures

Free energy landscape and thermodynamics properties of novel mutations in PncA of pyrazinamide resistance isolates of Mycobacterium tuberculosis

Pyrazinamide (PZA) is one of the first-line antituberculosis therapy, active against non-replicating Mycobacterium tuberculosis (Mtb). The conversion of PZA into pyrazinoic acid (POA), the active form, required the activity of pncA gene product pyrazinamidase (PZase) activity. Mutations occurred in pncA are the primary cause behind the PZA resistance. However, the resistance mechanism is important to explore using high throughput computational approaches. Here we aimed to explore the mechanism of PZA resistance behind novel P62T, L120R, and V130M mutations in PZase using 200 ns molecular dynamics (MD) simulations. MD simulations were performed to observe the structural changes for these three mutants (MTs) compared to the wild types (WT). Root means square fluctuation, the radius of gyration, free energy landscape, root means square deviation, dynamic cross-correlation motion, and pocket volume were found in variation between WT and MTs, revealing the effects of P62T, L120R, and V130M. The free energy conformational landscape of MTs differs significantly from the WT system, lowering the binding of PZA. The geometric shape complementarity of the drug (PZA) and target protein (PZase) further confirmed that P62T, L120R, and V130M affect the protein structure. These effects on PZase may cause vulnerability to convert PZA into POA.Communicated by Ramaswamy H. Sarma.

Determination of Molecule Category of Ligands Targeting the Ligand-Binding Pocket of Nuclear Receptors with Structural Elucidation and Machine Learning

The mechanism of transcriptional activation/repression of the nuclear receptors (NRs) involves two main conformations of the NR protein, namely, the active (agonistic) and inactive (antagonistic) conformations. Binding of agonists or antagonists to the ligand-binding pocket (LBP) of NRs can regulate the downstream signaling pathways with different physiological effects. However, it is still hard to determine the molecular type of a LBP-bound ligand because both the agonists and antagonists bind to the same position of the protein. Therefore, it is necessary to develop precise and efficient methods to facilitate the discrimination of agonists and antagonists targeting the LBP of NRs. Here, combining structural and energetic analyses with machine-learning (ML) algorithms, we constructed a series of structure-based ML models to determine the molecular category of the LBP-bound ligands. We show that the proposed models work robustly and with high accuracy (ACC > 0.9) for determining the category of molecules derived from docking-based and crystallized poses. Furthermore, the models are also capable of determining the molecular category of ligands with dual opposite functions on different NRs (i.e., working as an agonist in one NR target, whereas functioning as an antagonist in another) with reasonable accuracy. The proposed method is expected to facilitate the determination of the molecular properties of ligands targeting the LBP of NRs with structural interpretation.

Mechanistic analysis of A46V, H57Y, and D129N in pyrazinamidase associated with pyrazinamide resistance

Pyrazinamide (PZA) is a component of first-line drugs, active against latent Mycobacterium tuberculosis (MTB) isolates. The prodrug is activated into the active form, pyrazinoic acid (POA) via pncA gene-encoded pyrazinamidase (PZase). Mutations in pncA have been reported, most commonly responsible for PZA-resistance in more than 70% of the resistant cases. In our previous study, we detected many mutations in PZase among PZA-resistance MTB isolates including A46V, H71Y, and D129N. The current study was aimed to investigate the molecular mechanism of PZA-resistance behind mutants (MTs) A46V, H71Y, and D129N in comparison with the wild type (WT) through molecular dynamic (MD) simulation. MTB positive samples were subjected to PZA drug susceptibility testing (DST) against critical concentration (100ug/ml). The resistant samples were subjected to pncA sequencing. Thirty-six various mutations have been observed in the coding region of pncA of PZA-resistant isolates (GenBank accession No. MH461111) including A46V, H71Y, and D129N. The post-simulation analysis revealed a significant variation in MTs structural dynamics as compared to the WT. Root means square deviations (RMSD) and Root means square fluctuation (RMSF) has been found in variation between WT and MTs. Folding effect and pocket volume were altered in MTs when compared with WT. Geometric matching supports the effect of mutation A46V, H71Y, and D129N on PZase structure that may have an insight effect on PZase dynamics, making them vulnerable to convert pro-PZA into active form, POA. In conclusion, the current analyses will provide useful information behind PZA-resistance for better management of drug-resistant TB.

Pyrazinamide resistance of novel mutations in pncA and their dynamic behavior

Pyrazinamide (PZA) is one of the essential anti-mycobacterium drugs, active against non-replicating Mycobacterium tuberculosis (MTB) isolates. PZA is converted into its active state, called pyrazinoic acid (POA), by action of pncA encoding pyrazinamidase (PZase). In the majority of PZA-resistance isolates, pncA harbored mutations in the coding region. In our recent report, we detected a number of novel variants in PZA-resistance (PZA^R) MTB isolates, whose resistance mechanisms were yet to be determined. Here we performed several analyses to unveil the PZA^R mechanism of R123P, T76P, G150A, and H71R mutants (MTs) through molecular dynamics (MD) simulations. In brief, culture positive MTB isolates were subjected to PZA susceptibility tests using the WHO recommended concentration of PZA (100 μg ml⁻¹). The PZA^R samples were screened for mutations in pncA along sensitive isolates through polymerase chain reactions and sequencing. A large number of variants (GeneBank accession no. MH461111), including R123P, T76P, G150A, and H71R, have been spotted in more than 70% of isolates. However, the mechanism of PZA^R for mutants (MTs) R123P, T76P, G150A, and H71R was unknown. For the MTs and native PZase structures (WT), thermodynamic properties were compared using molecular dynamics simulations for 100 ns. The MTs structural activity was compared to the WT. Folding effect and pocket volume variations have been detected when comparing between WT and MTs. Geometric matching further confirmed the effect of R123P, T76P, G150A, and H71R mutations on PZase dynamics, making them vulnerable for activating the pro-drug into POA. This study offers a better understanding for management of PZA^R TB. The results may be used as alternative diagnostic tools to infer PZA resistance at a structural dynamics level.

We performed several analyses to unveil the pyrazinamide-resistance mechanism of R123P, T76P, G150A, and H71R mutants through molecular dynamics simulations.

Gibbs Free Energy Calculation of Mutation in PncA and RpsA Associated With Pyrazinamide Resistance

A central approach for better understanding the forces involved in maintaining protein structures is to investigate the protein folding and thermodynamic properties. The effect of the folding process is often disturbed in mutated states. To explore the dynamic properties behind mutations, molecular dynamic (MD) simulations have been widely performed, especially in unveiling the mechanism of drug failure behind mutation. When comparing wild type (WT) and mutants (MTs), the structural changes along with solvation free energy (SFE), and Gibbs free energy (GFE) are calculated after the MD simulation, to measure the effect of mutations on protein structure. Pyrazinamide (PZA) is one of the first-line drugs, effective against latent Mycobacterium tuberculosis isolates, affecting the global TB control program 2030. Resistance to this drug emerges due to mutations in pncA and rpsA genes, encoding pyrazinamidase (PZase) and ribosomal protein S1 (RpsA) respectively. The question of how the GFE may be a measure of PZase and RpsA stabilities, has been addressed in the current review. The GFE and SFE of MTs have been compared with WT, which were already found to be PZA-resistant. WT structures attained a more stable state in comparison with MTs. The physiological effect of a mutation in PZase and RpsA may be due to the difference in energies. This difference between WT and MTs, depicted through GFE plots, might be useful in predicting the stability and PZA-resistance behind mutation. This study provides useful information for better management of drug resistance, to control the global TB problem.

Combined strategies in structure-based virtual screening

The identification and optimization of lead compounds are inalienable components in drug design and discovery pipelines. As a powerful computational approach for the identification of hits with novel structural scaffolds, structure-based virtual screening (SBVS) has exhibited a remarkably increasing influence in the early stages of drug discovery. During the past decade, a variety of techniques and algorithms have been proposed and tested with different purposes in the scope of SBVS. Although SBVS has been a common and proven technology, it still shows some challenges and problems that are needed to be addressed, where the negative influence regardless of protein flexibility and the inaccurate prediction of binding affinity are the two major challenges. Here, focusing on these difficulties, we summarize a series of combined strategies or workflows developed by our group and others. Furthermore, several representative successful applications from recent publications are also discussed to demonstrate the effectiveness of the combined SBVS strategies in drug discovery campaigns.

Structural Dynamics Behind Clinical Mutants of PncA-Asp12Ala, Pro54Leu, and His57Pro of Mycobacterium tuberculosis Associated With Pyrazinamide Resistance

Pyrazinamide (PZA) is one of the main FDA approved drugs to be used as the first line of defense against Mycobacterium Tuberculosis (MTB). It is activated into pyrazinoic acid (POA) via MTB's pncA gene-encoded pyrazinamidase (PZase). Mutations are most commonly responsible for PZA-resistance in nearly 70% of the resistant samples. In the present work, MTB positive samples were chosen for PZA drug susceptibility testing (DST) against critical concentration (100 ug/ml) of PZA. The resistant samples were subjected to pncA sequencing. As a result, 36 various mutations have been observed in the PZA resistant samples, uploaded to the NCBI (GeneBank accession no. MH461111). Here we report the mechanism of PZA resistance behind the three mutants (MTs), Asp12Ala, Pro54Leu, and His57Pro in comparison with the wild type (WT) through molecular dynamics simulation to unveil how these mutations affect the overall conformational stability. The post-simulation analyses revealed notable deviations as compared to the WT structure. Molecular docking studies of PZA with MTs and WT, pocket volume inspection and overall shape complementarity analysis confirmed the deleterious nature of these mutations and gave an insight into the mechanism behind PZA-resistance. These analyses provide vital information regarding MTB drug resistance and could be extremely useful in therapy management and overcoming its global burden.

Structural dynamics behind variants in pyrazinamidase and pyrazinamide resistance

Pyrazinamide (PZA) is an important component of first-line anti-tuberculosis (anti-TB) drugs. The anti-TB agent is activated into an active form, pyrazinoic acid (POA), by Mycobacterium tuberculosis (MTB) pncA gene encoding pyrazinamidase (PZase). The major cause of PZA-resistance has been associated with mutations in the pncA gene. We have detected several novel mutations including V131F, Q141P, R154T, A170P, and V180F (GeneBank Accession No. MH461111) in the pncA gene of PZA-resistant isolates during PZA drug susceptibility testing followed by pncA gene sequencing. Here, we investigated molecular mechanism of PZA-resistance by comparing the results of experimental and molecular dynamics. The mutants (MTs) and wild type (WT) PZase structures in apo and complex with PZA were subjected to molecular dynamic simulations (MD) at the 40 ns. Multiple factors, including root mean square deviations (RMSD), binding pocket, total energy, dynamic cross correlation, and root mean square fluctuations (RMSF) of MTs and WT were compared. The MTs attained a high deviation and fluctuation compared to WT. Binding pocket volumes of the MTs, were found, lower than the WT, and the docking scores were high than WT while shape complementarity scores were lower than that of the WT. Residual motion in MTs are seemed to be dominant in anti-correlated motion. Mutations at locations, V131F, Q141P, R154T, A170P, and V180F, might be involved in the structural changes, possibly affecting the catalytic property of PZase to convert PZA into POA. Our study provides useful information that will enhance the understanding for better management of TB. AbbreviationsDSTdrug susceptibility testingΔelecelectrostatic energyLJLowenstein-Jensen mediumMGITmycobacterium growth indicator tubesMTsmutantsMDmolecular dynamic simulationsMTBMycobacterium tuberculosisNALC-NaOHN-acetyl-l-cysteine-sodium hydroxideNIHNational Institutes of HealthNPTamount of substance (N), pressure (P) temperature (T)NVTmoles (N), volume (V) temperature (T)PZasepyrazinamidaseΔpspolar solvation energyPTRLProvincial Tuberculosis Reference LaboratoryRMSDroot mean square deviationsRMSFroot mean square fluctuationsΔSASAsolvent accessible surface area energyTBtuberculosisGTotaltotal binding free energyΔvdWVan der Waals energyWTwild typeCommunicated by Ramaswamy H. Sarma.

Concepts of Artificial Intelligence for Computer-Assisted Drug Discovery

Artificial intelligence (AI), and, in particular, deep learning as a subcategory of AI, provides opportunities for the discovery and development of innovative drugs. Various machine learning approaches have recently (re)emerged, some of which may be considered instances of domain-specific AI which have been successfully employed for drug discovery and design. This review provides a comprehensive portrayal of these machine learning techniques and of their applications in medicinal chemistry. After introducing the basic principles, alongside some application notes, of the various machine learning algorithms, the current state-of-the art of AI-assisted pharmaceutical discovery is discussed, including applications in structure- and ligand-based virtual screening, de novo drug design, physicochemical and pharmacokinetic property prediction, drug repurposing, and related aspects. Finally, several challenges and limitations of the current methods are summarized, with a view to potential future directions for AI-assisted drug discovery and design.

Structural and free energy landscape of novel mutations in ribosomal protein S1 (rpsA) associated with pyrazinamide resistance

Resistance to key first-line drugs is a major hurdle to achieve the global end tuberculosis (TB) targets. A prodrug, pyrazinamide (PZA) is the only drug, effective in latent TB, recommended in drug resistance and susceptible Mycobacterium tuberculosis (MTB) isolates. The prodrug conversion into active form, pyrazinoic acid (POA), required the activity of pncA gene encoded pyrazinamidase (PZase). Although pncA mutations have been commonly associated with PZA resistance but a small number of resistance cases have been associated with mutationss in RpsA protein. Here in this study a total of 69 PZA resistance isolates have been sequenced for pncA mutations. However, samples that were found PZA resistant but pncA wild type (pncAWT), have been sequenced for rpsA and panD genes mutation. We repeated a drug susceptibility testing according to the WHO guidelines on 18 pncAWT MTB isolates. The rpsA and panD genes were sequenced. Out of total 69 PZA resistant isolates, 51 harbored 36 mutations in pncA gene (GeneBank Accession No. MH46111) while, fifteen different mutations including seven novel, were detected in the fourth S1 domain of RpsA known as C-terminal (MtRpsACTD) end. We did not detect any mutations in panD gene. Among the rpsA mutations, we investigated the molecular mechanism of resistance behind mutations, D342N, D343N, A344P, and I351F, present in the MtRpsACTD through molecular dynamic simulations (MD). WT showed a good drug binding affinity as compared to mutants (MTs), D342N, D343N, A344P, and I351F. Binding pocket volume, stability, and fluctuations have been altered whereas the total energy, protein folding, and geometric shape analysis further explored a significant variation between WT and MTs. In conclusion, mutations in MtRpsACTD might be involved to alter the RpsA activity, resulting in drug resistance. Such molecular mechanism behind resistance may provide a better insight into the resistance mechanism to achieve the global TB control targets.

An Overview of Scoring Functions Used for Protein-Ligand Interactions in Molecular Docking

Currently, molecular docking is becoming a key tool in drug discovery and molecular modeling applications. The reliability of molecular docking depends on the accuracy of the adopted scoring function, which can guide and determine the ligand poses when thousands of possible poses of ligand are generated. The scoring function can be used to determine the binding mode and site of a ligand, predict binding affinity and identify the potential drug leads for a given protein target. Despite intensive research over the years, accurate and rapid prediction of protein-ligand interactions is still a challenge in molecular docking. For this reason, this study reviews four basic types of scoring functions, physics-based, empirical, knowledge-based, and machine learning-based scoring functions, based on an up-to-date classification scheme. We not only discuss the foundations of the four types scoring functions, suitable application areas and shortcomings, but also discuss challenges and potential future study directions.

Constructing and Validating High-Performance MIEC-SVM Models in Virtual Screening for Kinases: A Better Way for Actives Discovery

The MIEC-SVM approach, which combines molecular interaction energy components (MIEC) derived from free energy decomposition and support vector machine (SVM), has been found effective in capturing the energetic patterns of protein-peptide recognition. However, the performance of this approach in identifying small molecule inhibitors of drug targets has not been well assessed and validated by experiments. Thereafter, by combining different model construction protocols, the issues related to developing best MIEC-SVM models were firstly discussed upon three kinase targets (ABL, ALK, and BRAF). As for the investigated targets, the optimized MIEC-SVM models performed much better than the models based on the default SVM parameters and Autodock for the tested datasets. Then, the proposed strategy was utilized to screen the Specs database for discovering potential inhibitors of the ALK kinase. The experimental results showed that the optimized MIEC-SVM model, which identified 7 actives with IC50 < 10 μM from 50 purchased compounds (namely hit rate of 14%, and 4 in nM level) and performed much better than Autodock (3 actives with IC50 < 10 μM from 50 purchased compounds, namely hit rate of 6%, and 2 in nM level), suggesting that the proposed strategy is a powerful tool in structure-based virtual screening.

MIEC-SVM: automated pipeline for protein peptide/ligand interaction prediction

MOTIVATION

MIEC-SVM is a structure-based method for predicting protein recognition specificity. Here, we present an automated MIEC-SVM pipeline providing an integrated and user-friendly workflow for construction and application of the MIEC-SVM models. This pipeline can handle standard amino acids and those with post-translational modifications (PTMs) or small molecules. Moreover, multi-threading and support to Sun Grid Engine (SGE) are implemented to significantly boost the computational efficiency.

AVAILABILITY AND IMPLEMENTATION

The program is available at http://wanglab.ucsd.edu/MIEC-SVM CONTACT: : wei-wang@ucsd.edu

SUPPLEMENTARY INFORMATION

Supplementary data available at Bioinformatics online.

Improvements, trends, and new ideas in molecular docking: 2012-2013 in review

Molecular docking is a computational method for predicting the placement of ligands in the binding sites of their receptor(s). In this review, we discuss the methodological developments that occurred in the docking field in 2012 and 2013, with a particular focus on the more difficult aspects of this computational discipline. The main challenges and therefore focal points for developments in docking, covered in this review, are receptor flexibility, solvation, scoring, and virtual screening. We specifically deal with such aspects of molecular docking and its applications as selection criteria for constructing receptor ensembles, target dependence of scoring functions, integration of higher-level theory into scoring, implicit and explicit handling of solvation in the binding process, and comparison and evaluation of docking and scoring methods.

Drug/nondrug classification using Support Vector Machines with various feature selection strategies

In conjunction with the advance in computer technology, virtual screening of small molecules has been started to use in drug discovery. Since there are thousands of compounds in early-phase of drug discovery, a fast classification method, which can distinguish between active and inactive molecules, can be used for screening large compound collections. In this study, we used Support Vector Machines (SVM) for this type of classification task. SVM is a powerful classification tool that is becoming increasingly popular in various machine-learning applications. The data sets consist of 631 compounds for training set and 216 compounds for a separate test set. In data pre-processing step, the Pearson's correlation coefficient used as a filter to eliminate redundant features. After application of the correlation filter, a single SVM has been applied to this reduced data set. Moreover, we have investigated the performance of SVM with different feature selection strategies, including SVM-Recursive Feature Elimination, Wrapper Method and Subset Selection. All feature selection methods generally represent better performance than a single SVM while Subset Selection outperforms other feature selection methods. We have tested SVM as a classification tool in a real-life drug discovery problem and our results revealed that it could be a useful method for classification task in early-phase of drug discovery.