Novel Strategy for Three-Dimensional Fragment-Based Lead Discovery

Discovery and development of Factor Xa inhibitors (2015-2022)

As a pathological coagulation process, thrombus can lead to many serious diseases, including ischemic stroke, acute myocardial infarction (AMI), acute coronary syndrome (ACS), and deep venous thrombosis (DVT). And anticoagulant drugs are one of the most effective ways to prevent and treat these diseases. Although macromolecular anticoagulant drugs such as low molecular weight heparins (LMWHs) are widely used in the clinic, their characteristics of requiring injectable use hinder their further promotion in the clinic, and the disadvantages of oral anticoagulant drugs, such as warfarin and dabigatran etexilate, which can easily cause bleeding adverse effects, are also not addressed. Factor Xa (FXa) has gained attention because it lies at the intersection of the coagulation cascade pathways, whereas subsequently introduced Factor Xa inhibitors such as rivaroxaban and apixaban, among others, have gained market popularity because of their high potency for anticoagulation and high specificity for Factor Xa when administered orally. But some of the drawbacks that these Factor Xa inhibitors have simultaneously such as fewer indications and the lack of an effective reversal drug when bleeding occurs are urgently addressed. The development of new Factor Xa inhibitors therefore becomes one means of addressing these questions. This article summarizes the small molecule Factor Xainhibitors developed from 2015 to 2022, classifies them according to their scaffolds, focuses on the analysis of their structure-activity relationships, and provides a brief assessment of them.

Identification of anthranilamide derivatives as potential factor Xa inhibitors: drug design, synthesis and biological evaluation

The coagulation enzyme factor Xa (fXa) plays a crucial role in the blood coagulation cascade. In this study, three-dimensional fragment based drug design (FBDD) combined with structure-based pharmacophore (SBP) model and structural consensus docking were employed to identify novel fXa inhibitors. After a multi-stage virtual screening (VS) workflow, two hit compounds 3780 and 319 having persistent high performance were identified. Then, these two hit compounds and several analogs were synthesized and screened for in-vitro inhibition of fXa. The experimental data showed that most of the designed compounds displayed significant in vitro potency against fXa. Among them, compound 9b displayed the greatest in vitro potency against fXa with the IC50 value of 23 nM and excellent selectivity versus thrombin (IC50 = 40 μM). Moreover, the prolongation of the prothrombin time (PT) was measured for compound 9b to evaluate its in vitro anticoagulant activity. As a result, compound 9b exhibited pronounced anticoagulant activity with the 2 × PT value of 8.7 μM.

DrugOn: a fully integrated pharmacophore modeling and structure optimization toolkit

During the past few years, pharmacophore modeling has become one of the key components in computer-aided drug design and in modern drug discovery. DrugOn is a fully interactive pipeline designed to exploit the advantages of modern programming and overcome the command line barrier with two friendly environments for the user (either novice or experienced in the field of Computer Aided Drug Design) to perform pharmacophore modeling through an efficient combination of the PharmACOphore, Gromacs, Ligbuilder and PDB2PQR suites. Our platform features a novel workflow that guides the user through each logical step of the iterative 3D structural optimization setup and drug design process. For the pharmacophore modeling we are focusing on either the characteristics of the receptor or the full molecular system, including a set of selected ligands. DrugOn can be freely downloaded from our dedicated server system at www.bioacademy.gr/bioinformatics/drugon/.

Molecular modeling of exquisitely selective c-Met inhibitors through 3D-QSAR and molecular dynamics simulations

c-Met has been considered as an attractive target for developing antitumor agents. The highly selective c-Met inhibitors provide invaluable opportunities for the combination with other therapies safely to achieve the optimal efficacy. In this work, a series of triazolopyrazine c-Met inhibitors with exquisitely selectivity were investigated using a combination of molecular docking, three-dimensional quantitative structure-activity relationship (3D-QSAR), and molecular dynamics simulation. Comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA) models were developed to reveal the structural determinants for c-Met inhibition. Both models were validated to have high reliability and predictability, and contour map analysis suggested feature requirements for different substituents on the scaffold. It is worth noting that an important hydrogen bond rich region was identified in the unique narrow channel, which is distinct from other kinases. Molecular dynamics simulations and binding free energy calculations provided further support that suitable groups in this hydrogen bond rich region made great contributions to the binding of ligands. Moreover, hydrogen bonds with residues of the narrow channel were also indicated to be essential to improve the activity and selectivity. This study will facilitate the discovery and optimization of novel c-Met inhibitors with higher activity and selectivity.

Molecular modelling on small molecular CDK2 inhibitors: an integrated approach using a combination of molecular docking, 3D-QSAR and pharmacophore modelling

Cyclin-dependent kinase 2 (CDK2) has been identified as an important target for developing novel anticancer agents. Molecular docking, three-dimensional quantitative structure-activity relationship (3D-QSAR) and pharmacophore modelling were combined with the ultimate goal of studying the structure-activity relationship of CDK2 inhibitors. The comparative molecular similarity indices analysis (CoMSIA) model constructed based on a set of 3-aminopyrazole derivatives as CDK2 inhibitors gave statistically significant results (q (2) = 0.700; r (2) = 0.982). A HypoGen pharmacophore model, constructed using diverse CDK2 inhibitors, also showed significant statistics ([Formula: see text]Cost = 61.483; RMSD = 0.53; Correlation coefficient = 0.98). The small residues and error values between the estimated and experimental activities of the training and test set compounds proved their strong capability of activity prediction. The structural insights obtained from these two models were consistent with each other. The pharmacophore model summarized the important pharmacophoric features required for protein-ligand binding. The 3D contour maps in combination with the comprehensive pharmacophoric features helped to better interpret the structure-activity relationship. The results will be beneficial for the discovery and design of novel CDK2 inhibitors. The simplicity of this approach provides expansion to its applicability in optimizing other classes of small molecular CDK2 inhibitors.

In silico deconstruction of ATP-competitive inhibitors of glycogen synthase kinase-3β

Fragment-based methods have emerged in the last two decades as alternatives to traditional high throughput screenings for the identification of chemical starting points in drug discovery. One arguable yet popular assumption about fragment-based design is that the fragment binding mode remains conserved upon chemical expansion. For instance, the question of the binding conservation upon fragmentation of a molecule is still unclear. A number of papers have challenged this hypothesis by means of experimental techniques, with controversial results, "underlining" the idea that a simple generalization, maybe, is not possible. From a computational standpoint, the issue has been rarely addressed and mostly to test novel protocols on limited data sets. To fill this gap, we here report on a computational retrospective study concerned with the in silico deconstruction of leadlike compounds, active on the pharmaceutically relevant enzyme glycogen synthase kinase-3β.

Identification of novel polo-like kinase 1 inhibitors by a hybrid virtual screening

Polo-like kinase 1 is an important and attractive oncological target that plays a key role in mitosis and cytokinesis. A combined pharmacophore- and docking-based virtual screening was performed to identify novel polo-like kinase 1 inhibitors. A total of 34 hit compounds were selected and tested in vitro, and some compounds showed inhibition of polo-like kinase 1 and human tumor cell growth. The most potent compound (66) inhibited polo-like kinase 1 with an IC(50) value of 6.99 μm. The docked binding models of two hit compounds were discussed in detail. These compounds contained novel chemical scaffolds and may be used as foundations for the development of novel classes of polo-like kinase 1 inhibitors.

Pharmacophore modeling and virtual screening studies to identify new c-Met inhibitors

Mesenchymal epithelial transition factor (c-Met) is an attractive target for cancer therapy. Three-dimensional pharmacophore hypotheses were built based on a set of known structurally diverse c-Met inhibitors. The best pharmacophore model, which identified inhibitors with an associated correlation coefficient of 0.983 between their experimental and estimated IC(50) values, consisted of two hydrogen-bond acceptors, one hydrophobic, and one ring aromatic feature. The highly predictive power of the model was rigorously validated by test set prediction and Fischer's randomization method. The high values of enrichment factor and receiver operating characteristic (ROC) score indicated the model performed fairly well at distinguishing active from inactive compounds. The model was then applied to screen compound database for potential c-Met inhibitors. A filtering protocol, including druggability and molecular docking, were also applied in hits selection. The final 38 molecules, which exhibited good estimated activities, desired binding mode and favorable drug likeness were identified as potential c-Met inhibitors. Their novel backbone structures could be served as scaffolds for further study, which may facilitate the discovery and rational design of potent c-Met kinase inhibitors.

Visualization of molecular fingerprints

A visualization plot of a data set of molecular data is a useful tool for gaining insight into a set of molecules. In chemoinformatics, most visualization plots are of molecular descriptors, and the statistical model most often used to produce a visualization is principal component analysis (PCA). This paper takes PCA, together with four other statistical models (NeuroScale, GTM, LTM, and LTM-LIN), and evaluates their ability to produce clustering in visualizations not of molecular descriptors but of molecular fingerprints. Two different tasks are addressed: understanding structural information (particularly combinatorial libraries) and relating structure to activity. The quality of the visualizations is compared both subjectively (by visual inspection) and objectively (with global distance comparisons and local k-nearest-neighbor predictors). On the data sets used to evaluate clustering by structure, LTM is found to perform significantly better than the other models. In particular, the clusters in LTM visualization space are consistent with the relationships between the core scaffolds that define the combinatorial sublibraries. On the data sets used to evaluate clustering by activity, LTM again gives the best performance but by a smaller margin. The results of this paper demonstrate the value of using both a nonlinear projection map and a Bernoulli noise model for modeling binary data.