An integrated structure- and pharmacophore-based MMP-12 virtual screening

MMP-12 belongs to a large family of proteases called matrix metalloproteinases (MMPs) that degrades elastin. The main pathologic role of MMP-12 overexpression was suggested to be associated with pathogenesis mechanism of inflammatory respiratory diseases and atherosclerosis. An integrated ligand- and structure-based virtual screening was employed in hope of finding inhibitors with new scaffolds and selectivity for MMP-12. Seven compounds among 18 experimentally tested compounds had a measurable effect on the inhibition of MMP-12 enzyme. Our results demonstrated the applicability of the developed pharmacophore model and selected crystal structure (PDB code: 3F17) to discover new MMP-12 inhibitors. The receptor structure was selected based on cross-docking results. Here, we report the discovery of new class of MMP-12 inhibitors that could be used for lead optimization. For the inhibition of MMP-12, the significance of its interactions with the catalytic residues Glu219 and Ala182 was emphasized through the inspection of the docking poses.     

Identification of novel potential HIF-prolyl hydroxylase inhibitors by in silico screening

Suppression of HIF-prolyl hydroxylase (PHD) activity by small-molecule inhibitors leads to the stabilization of hypoxia inducible factor and has been recognized as promising drug target for the treatment of ischemic diseases. In this study, pharmacophore-based virtual screening and molecular docking approaches were concurrently used with suitable modifications to identify target-specific PHD inhibitors with better absorption, distribution, metabolism, and excretion properties and to readily minimize false positives and false negatives. A customized method based on the active site information of the enzyme was used to generate a pharmacophore hypothesis (AAANR). The hypothesis was validated and utilized for chemical database screening and the retrieved hit compounds were subjected to molecular docking for further refinement. AAANR hypothesis comprised three H-bond acceptor, one negative ionizable group and one aromatic ring feature. The hypothesis was validated using decoy set with a goodness of fit score of 2 and was used as a 3D query for database screening. After manual selection, molecular docking and further refinement based on the molecular interactions of inhibitors with the essential amino acid residues, 18 hits with good absorption, distribution, metabolism, and excretion (ADME) properties were selected as excellent PHD inhibitors. The best pharmacophore hypothesis, AAANR, contains chemical features required for the effective inhibition of PHD. Using AAANR, we have identified 18 potential and diverse virtual leads, which can be readily evaluated for their potency as novel inhibitors of PHD. It can be concluded that the combination of pharmacophore, molecular docking, and manual interpretation approaches can be more successful than the traditional approach alone for discovering more effective inhibitors.     

Discovery and identification of PIM-1 kinase inhibitors through a hybrid screening approach

PIM-1 kinase is an important therapeutic target in the treatment of cancer. Discovery and identification of PIM-1 Inhibitors with novel scaffolds are an effective way for developing potent therapeutic agents for the treatment of cancers. Here we proposed a hybrid screening approach which combines an optimal structure-based drug design strategy and a simple pharmacophore model to discover PIM-1 kinase inhibitors. With the proposed hybrid screening approach, the SPECS database containing 204,580 molecules was screened. In total, 89 hits were obtained. Forty three of them were purchased and tested in bioassays. Finally, 5 lead compounds with novel scaffolds were identified to exhibit promising antitumor activities against human leukemia cell line MV4-11, K-562 and human prostate cancer cell line PC-3 and DU145. Their $\hbox {IC}_{50}$ values range from 4.40 to $37.96 \,\upmu \hbox {M}$ . Three hits with 3 different scaffolds were selected from these five hits for binding mode analysis. It was demonstrated that the subtle differences in the interactions of the representatives with PIM-1 kinase contribute to the different inhibitory activities. It was also demonstrated that the suggested hybrid screening approach is an effective method to discover PIM-1 inhibitors possessing different scaffolds. These leads have a strong likelihood to act as further starting points for us in the optimization and development of potent PIM-1 inhibitors.     

Natural products as DNA methyltransferase inhibitors: a computer-aided discovery approach

DNA methyltransferases (DNMTs) represent promising targets for the development of unique anticancer drugs. However, all DNMT inhibitors currently in clinical use are nonselective cytosine analogs with significant cytotoxic side-effects. Several natural products, covering diverse chemical classes, have indicated DNMT inhibitory activity, but these effects have yet to be systematically evaluated. In this study, we provide experimental data suggesting that two of the most prominent natural products associated with DNA methylation inhibition, (−)-epigallocathechin-3-gallate (EGCG) and curcumin, have little or no pharmacologically relevant inhibitory activity. We therefore conducted a virtual screen of a large database of natural products with a validated homology model of the catalytic domain of DNMT1. The virtual screening focused on a lead-like subset of the natural products docked with DNMT1, using three docking programs, following a multistep docking approach. Prior to docking, the lead-like subset was characterized in terms of chemical space coverage and scaffold content. Consensus hits with high predicted docking affinity for DNMT1 by all three docking programs were identified. One hit showed DNMT1 inhibitory activity in a previous study. The virtual screening hits were located within the biological-relevant chemical space of drugs, and represent potential unique DNMT inhibitors of natural origin. Validation of these virtual screening hits is warranted.     

Identification of dual Acetyl-CoA carboxylases 1 and 2 inhibitors by pharmacophore based virtual screening and molecular docking approach

Acetyl-CoA carboxylase (ACC) is a crucial metabolic enzyme that plays a vital role in obesity-induced type 2 diabetes and fatty acid metabolism. To identify dual inhibitors of Acetyl-CoA carboxylase1 and Acetyl-CoA carboxylase2, a pharmacophore modelling approach has been employed. The best HypoGen pharmacophore model for ACC2 inhibitors (Hypo1_ACC2) consists of one hydrogen bond acceptor, one hydrophobic aliphatic and one hydrophobic aromatic feature, whereas the best pharmacophore (Hypo1_ACC1) for ACC1 consists of one additional hydrogen-bond donor (HBD) features. The best pharmacophore hypotheses were validated by various methods such as test set, decoy set and Cat-Scramble methodology. The validated pharmacophore models were used to screen several small-molecule databases, including Specs, NCI, ChemDiv and Natural product databases to identify the potential dual ACC inhibitors. The virtual hits were then subjected to several filters such as estimated $\text{ IC}_{50}$ value, quantitative estimation of drug-likeness and molecular docking analysis. Finally, three novel compounds with diverse scaffolds were selected as potential starting points for the design of novel dual ACC inhibitors.     

Docking of a novel DNA methyltransferase inhibitor identified from high-throughput screening: insights to unveil inhibitors in chemical databases

Inhibitors of DNA methyltransferase (DNMT) are attractive compounds not only as potential therapeutic agents for the treatment of cancer and other diseases, but also as research tools to investigate the role of DNMTs in epigenetic events. Recent advances in high-throughput screening (HTS) for epigenetic targets and the availability of the first crystallographic structure of human DNMT1 encourage the integration of research strategies to uncover and optimize the activity of DNMT inhibitors. Herein, we present a binding model of a novel small-molecule DNMT1 inhibitor obtained by HTS, recently released in a public database. The docking model is in agreement with key interactions previously identified for established inhibitors using extensive computational studies including molecular dynamics and structure-based pharmacophore modeling. Based on the chemical structure of the novel inhibitor, a sequential computational screening of five chemical databases was performed to identify candidate compounds for testing. Similarity searching followed by molecular docking of chemical databases such as approved drugs, natural products, a DNMT-focused library, and a general screening collection, identified at least 108 molecules with promising DNMT inhibitory activity. The chemical structures of all hit compounds are disclosed to encourage the research community working on epigenetics to test experimentally the enzymatic and demethylating activity in vivo. Five candidate hits are drugs approved for other indications and represent potential starting points of a drug repurposing strategy.     

Pharmacophore-based screening and drug repurposing exemplified on glycogen synthase kinase-3 inhibitors

The current study was conducted to elaborate a novel pharmacophore model to accurately map selective glycogen synthase kinase-3 (GSK-3) inhibitors, and perform virtual screening and drug repurposing. Pharmacophore modeling was developed using PHASE on a data set of 203 maleimides. Two benchmarking validation data sets with focus on selectivity were assembled using ChEMBL and PubChem GSK-3 confirmatory assays. A drug repurposing experiment linking pharmacophore matching with drug information originating from multiple data sources was performed. A five-point pharmacophore model was built consisting of a hydrogen bond acceptor (A), hydrogen bond donor (D), hydrophobic (H), and two rings (RR). An atom-based 3D quantitative structure–activity relationship (QSAR) model showed good correlative and satisfactory predictive abilities (training set \({R}^{2}= 0.904\); test set: \({Q}^{2}= 0.676\); whole data set: stability \(s = 0.803\)). Virtual screening experiments revealed that selective GSK-3 inhibitors are ranked preferentially by Hypo-1, but fail to retrieve nonselective compounds. The pharmacophore and 3D QSAR models can provide assistance to design novel, potential GSK-3 inhibitors with high potency and selectivity pattern, with potential application for the treatment of GSK-3-driven diseases. A class of purine nucleoside antileukemic drugs was identified as potential inhibitor of GSK-3, suggesting the reassessment of the target range of these drugs.     

Identification of potential glutaminyl cyclase inhibitors from lead-like libraries by in silico and in vitro fragment-based screening

A glutaminyl cyclase (QC) fragment library was in silico selected by disconnection of the structure of known QC inhibitors and by lead-like 2D virtual screening of the same set. The resulting fragment library (204 compounds) was acquired from commercial suppliers and pre-screened by differential scanning fluorimetry followed by functional in vitro assays. In this way, 10 fragment hits were identified (\(\sim \)5 % hit rate, best inhibitory activity: 16 \(\upmu \hbox {M}\)). The in vitro hits were then docked to the active site of QC, and the best scoring compounds were analyzed for binding interactions. Two fragments bound to different regions in a complementary manner, and thus, linking those fragments offered a rational strategy to generate novel QC inhibitors. Based on the structure of the virtual linked fragment, a 77-membered QC target focused library was selected from vendor databases and docked to the active site of QC. A PubChem search confirmed that the best scoring analogues are novel, potential QC inhibitors.     

A rational approach to selective pharmacophore designing: an innovative strategy for specific recognition of Gsk3β

We propose a novel cheminformatics approach that combines structure and ligand-based design to identify target-specific pharmacophores with well-defined exclusion ability. Our strategy includes the prediction of selective interactions, developing structure, and knowledge-based selective pharmacophore models, followed by database screening and molecular docking. This unique strategy was employed in addressing the off-target toxicity of Gsk3β and CDKs. The connections of Gsk3β in eukaryotic cell apoptosis and the extensive potency of Gsk3β inhibitors to block cell death have made it a potential drug-discovery target for many grievous human disorders. Gsk3β is phylogenetically very closely related to the CDKs, such as CDK1 and CDK2, which are suggested to be the off-target proteins of Gsk3β inhibitors. Here, we have employed novel computational approaches in designing the ligand candidates that are potentially inhibitory against Gsk3β, with well-defined the exclusion ability to CDKs. A structure-ligand -based selective pharmacophore was modeled. This model was used to retrieve molecules from the zinc database. The hits retrieved were further screened by molecular docking and protein-ligand interaction fingerprints. Based on these results, four molecules were predicted as selective Gsk3β antagonists. It is anticipated that this unique approach can be extended to investigate any protein-ligand specificity.     

Identification of new novel scaffold for Aurora A inhibition by pharmacophore modeling and virtual screening

Aurora kinases belong to family of highly conserved serine/threonine protein kinases that are involved in diverse cell cycle events and play a major role in regulation of cell division. Abnormal expression of Aurora kinases may lead to cancer; hence, these are considered as a potential target in cancer treatment. In this research article, we identified three novel Aurora A inhibitors using modern computational tools. A four-point common 3D pharmacophore hypothesis of Aurora A (AurA) inhibitors was developed using a diverse set of 55 thienopyrimidine derivatives. A three-dimensional quantitative structure–activity relationship (3D-QSAR) study was carried out using atom-based alignment of diverse set of 55 molecules to evaluate the structure– activity relationships. Docking and 3D-QSAR studies were performed with the 3D structure of AurA to evaluate the generated pharmacophore. The pharmacophore model and 3D-QSAR results complemented the results of our docking study. The pharmacophore hypothesis, which yields the best results, was used to screen the Zinc ‘clean drug-like’ database. Various database filters such as 3D-arrangement of pharmacophoric features, predicted activity and binding interaction score were used to retrieve hits having potential AurA inhibition activity.     

Comparative docking and CoMFA analysis of curcumine derivatives as HIV-1 integrase inhibitors

The docking studies and comparative molecular field analysis (CoMFA) were performed on highly active molecules of curcumine derivatives against 3′ processing activity of HIV-1 integrase (IN) enzyme. The optimum CoMFA model was selected with statistically significant cross-validated r² value of 0.815 and non-cross validated r ² value of 0.99. The common pharmacophore of highly active molecules was used for screening of HIV-1 IN inhibitors. The high contribution of polar interactions in pharmacophore mapping is well supported by docking and CoMFA results. The results of docking, CoMFA, and pharmacophore mapping give structural insights as well as important binding features of curcumine derivatives as HIV-1 IN inhibitors which can provide guidance for the rational design of novel HIV-1 IN inhibitors.     

Prospect of Anterior Gradient 2 homodimer inhibition via repurposing FDA-approved drugs using structure-based virtual screening

Anterior Gradient 2 (AGR2) has recently been reported as a tumor biomarker in various cancers, i.e., breast, prostate and lung cancer. Predominantly, AGR2 exists as a homodimer via a dimerization domain (E60-K64); after it is self-dimerized, it helps FGF2 and VEGF to homo-dimerize and promotes the angiogenesis and the invasion of vascular endothelial cells and fibroblasts. Up till now, no small molecule has been discovered to inhibit the AGR2–AGR2 homodimer. Therefore, the present study was performed to prepare a validated 3D structure of AGR2 by homology modeling and discover a small molecule by screening the FDA-approved drugs library on AGR2 homodimer as a target protein. Thirteen different homology models of AGR2 were generated based on different templates which were narrowed down to 5 quality models sorted by their overall Z-scores. The top homology model based on PDB ID?=?3PH9 was selected having the best Z-score and was further assessed by Verify-3D, ERRAT and RAMPAGE analysis. Structure-based virtual screening narrowed down the large library of FDA-approved drugs to ten potential AGR2–AGR2 homodimer inhibitors having FRED score lower than ? 7.8 kcal/mol in which the top 5 drugs’ binding stability was counter-validated by molecular dynamic simulation. To sum up, the present study prepared a validated 3D structure of AGR2 and, for the first time reported the discovery of 5 FDA-approved drugs to inhibit AGR2–AGR2 homodimer by using structure-based virtual screening. Moreover, the binding of the top 5 hits with AGR2 was also validated by molecular dynamic simulation.

Graphic abstract

A validated 3D structure of Anterior Gradient 2 (AGR2) was prepared by homology modeling, which was used in virtual screening of FDA-approved drugs library for the discovery of prospective inhibitors of AGR2–AGR2 homodimer.

     

Novel RIPK3 inhibitors discovered through a structure-based approach exert post-ischemic neuroprotection

Necroptosis or programmed necrosis is evident in various neurological disorders such as ischemic stroke. Receptor interacting serine/threonine protein kinase 3 (RIPK3) is one of the crucial targets of necroptosis and inhibition of this protein exerts neuroprotection. However, knowledge regarding the three-dimensional structure and binding site information of this protein is lacking. In the present study, structure-based in silico methods were implemented to identify the key amino acids in the RIPK3 binding site that might be responsible for ligand interactions. Further, novel RIPK3 inhibitors were identified through a dual ensemble screening strategy. Three inhibitors exhibited binding to RIPK3 in micromolar concentrations and exerted post-ischemic neuroprotection in vitro.     

GSA: a GPU-accelerated structure similarity algorithm and its application in progressive virtual screening

A new algorithm is proposed for accelerating chemical structure similarity search by means of graphic processing unit technology. Experiments demonstrate that the new algorithm is on average more than 120-times faster than the one implemented in conventional central processing unit technology. In order to test the generality of the new algorithm, it has been applied in seven progressive virtual screening experiments on NCI/DTP 60 human cancer cell lines data. The progressive virtual screening results show that the technology can select 10?C20% compounds for screening to get 70?C80% intrinsic hits for a given chemical library and target.     

Integration-mediated prediction enrichment of quantitative model for Hsp90 inhibitors as anti-cancer agents: 3D-QSAR study

The present study describes a systematic 3D-QSAR study consisting of pharmacophore modeling, docking, and integration of ligand-based and structure-based drug design approaches, applied on a dataset of 72 Hsp90 inhibitors as anti-cancer agents. The best pharmacophore model, with one H-bond donor (HBD), one H-bond acceptor (HBA), one hydrophobic_aromatic (Hy_Ar), and two hydrophobic_aliphatic (Hy_Al) features, was developed using the Catalyst/HypoGen algorithm on a training set of 35 compounds. The model was further validated using test set, external set, Fisher’s randomization method, and ability of the pharmacophoric features to complement the active site amino acids. Docking analysis was performed using Hsp90 chaperone (PDB-Id: 1uyf) along with water molecules reported to be crucial for binding and catalysis (Sgobba et al. ChemMedChem 4:1399–1409, 2009). Furthermore, an integration of the ligand-based as well as structure-based drug design approaches was done leading to the integrated model, which was found to be superior over the best pharmacophore model in terms of its predictive ability on internal [integrated model 2: R _(train) = 0.954, R _(test) = 0.888; Hypo-01: R _(train) = 0.912 and R _(test) = 0.819] as well as on external data set [integrated model 2: R _(ext.set) = 0.801; Hypo-01: R _(ext.set) = 0.604].     

Pharmacophore mapping of the crucial mediators of acetylcholinesterase and butyrylcholinesterase dual inhibition in Alzheimer's disease

Optimization and re-optimization of bioactive molecules using in silico methods have found application in the design of more active ones. Herein, we applied a pharmacophore modeling approach to screen potent dual inhibitors of acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) aimed at Alzheimer's disease (AD) treatment. The investigation entails molecular dynamics simulation, docking, pharmacophore modeling, drug-like screening, and binding energy analysis. We prepared a pharmacophore model from approved inhibitors of AChE and BuChE to predict the crucial moieties required for optimum molecular interaction with these proteins. The obtained pharmacophore model, used for database screening via some critical criteria, showed 229 hit molecules. Further analyses showed 42 likely dual inhibitors of AChE/BuChE with drug-like and pharmacokinetics properties the same as the approved cholinesterase inhibitors. Finally, we identified 14 dual molecules with improved potentials over the existing inhibitors and simulated ZINC92385797 bound to human AChE and BuChE structure after noticing that these 14 molecules are similar. The selected compound maintained relative stability at the active sites of both proteins over 120 ns simulation. Our integrated protocols showed the pertinent recipes of anti-AD drug design through the in silico pipeline.

Graphical abstract