Identification of potential glycogen kinase-3 inhibitors by structure based virtual screening

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 [Formula: see text]; test set: [Formula: see text]; whole data set: stability [Formula: see text]). 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.

Structure-Based Discovery of 1H-Indazole-3-carboxamides as a Novel Structural Class of Human GSK-3 Inhibitors

An in silico screening procedure was performed to select new inhibitors of glycogen synthase kinase 3β (GSK-3β), a serine/threonine protein kinase that in the last two decades has emerged as a key target in drug discovery, having been implicated in multiple cellular processes and linked with the pathogenesis of several diseases. GSK-3β inhibitors might prove useful as therapeutic compounds in the treatment of conditions associated with elevated levels of enzyme activity, such as type-2 diabetes and neurological disorders, for example, Alzheimer's disease, bipolar disorder, neuronal cell death, stroke, and depression. In this work, virtual screening studies were applied to proprietary compound libraries, and the functional activities of selected compounds were assayed on human GSK-3β. The in silico screening procedure enabled the identification of eight hit compounds showing pIC50 values ranging from 4.9 to 5.5. X-ray crystallographic studies resulted in a 2.50 Å three-dimensional structure of GSK-3β complexed with one of the selected compounds, confirming that the inhibitor interacts with the enzyme according to the docking hypothesis. Importantly, molecular docking was able to find a new chemical scaffold for GSK-3β inhibition, providing grounds for rational structure-based design aimed at further optimization of the initial hits.

Structural basis for the complete loss of GSK3beta catalytic activity due to R96 mutation investigated by molecular dynamics study

Many Ser/Thr protein kinases, to be fully activated, are obligated to introduce a phospho-Ser/Thr in their activation loop. Presently, the similarity of activation loop between two crystal complexes, i.e. glycogen synthase kinase 3beta (GSK3beta)-AMPNP and GSK3beta-sulfate ion complex, indicates that the activation segment of GSK3beta is preformed requiring neither a phosphorylation event nor conformational changes. GSK3beta, when participated in glycogen synthesis and Wnt signaling pathways, possesses a unique feature with the preference of such substrate with a priming phosphate. Experimental mutagenesis proved that the residue arginine at amino acid 96 mutations to lysine (R96K) or alanine (R96A) selectively abolish activity on the substrates involved in glycogen synthesis signaling pathway. Based on two solved crystal structures, wild type (WT) and two mutants (R96K and R96A) GSK3beta-ATP-phospho-Serine (pSer) complexes were modeled. Molecular dynamics simulations and energy analysis were employed to investigate the effect of pSer involvement on the GSK3beta structure in WT, and the mechanisms of GSK3beta deactivation due to R96K and R96A mutations. The results indicate that the introduction of pSer to WT GSK3beta generates a slight lobe closure on GSK3beta without any remarkable changes, which may illuminate the experimental conclusion, whereas the conformations of GSK3beta and ATP undergo significant changes in two mutants. As to GSK3beta, the affected positions distribute over activation loop, alpha-helix, and glycine-rich loop. Based on coupling among the mentioned positions, the allosteric mechanisms for distorted ATP were proposed. Energy decomposition on the residues of activation loop identified the important residues Arg96 and Arg180 in anchoring the phosphate group.

4-Aminoethylamino-emodin--a novel potent inhibitor of GSK-3beta--acts as an insulin-sensitizer avoiding downstream effects of activated beta-catenin

Glycogen synthase kinase-3β (GSK-3β) is a key target and effector of downstream insulin signalling. Using comparative protein kinase assays and molecular docking studies we characterize the emodin-derivative 4-[N-2-(aminoethyl)-amino]-emodin (L4) as a sensitive and potent inhibitor of GSK-3β with peculiar features. Compound L4 shows a low cytotoxic potential compared to other GSK-3β inhibitors determined by the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide assay and cellular ATP levels. Physiologically, L4 acts as an insulin-sensitizing agent that is able to enhance hepatocellular glycogen and fatty acid biosynthesis. These functions are particularly stimulated in the presence of elevated concentrations of glucose and in synergy with the hormone action at moderate but not high insulin levels. In contrast to other low molecular weight GSK-3β inhibitors (SB216763 and LiCl) or Wnt-3α-conditioned medium, however, L4 does not induce reporter and target genes of activated β-catenin such as TOPflash, Axin2 and glutamine synthetase. Moreover, when present together with SB216763 or LiCl, L4 counteracts expression of TOPflash or induction of glutamine synthetase by these inhibitors. Because L4 slightly activates β-catenin on its own, these results suggest that a downstream molecular step essential for activation of gene transcription by β-catenin is also inhibited by L4. It is concluded that L4 represents a potent insulin-sensitizing agent favouring physiological effects of insulin mediated by GSK-3β inhibition but avoiding hazardous effects such as activation of β-catenin-dependent gene expression which may lead to aberrant induction of cell proliferation and cancer.

Virtual screening to identify lead inhibitors for bacterial NAD synthetase (NADs)

Virtual screening was employed to identify new drug-like inhibitors of NAD synthetase (NADs) as antibacterial agents. Four databases of commercially available compounds were docked against three subsites of the NADs active site using FlexX in conjunction with CScore. Over 200 commercial compounds were purchased and evaluated in enzyme inhibition and antibacterial assays. 18 compounds inhibited NADs at or below 100 microM (7.6% hit rate), and two were selected for future SAR studies.

Differentiation of AmpC beta-lactamase binders vs. decoys using classification kNN QSAR modeling and application of the QSAR classifier to virtual screening

The use of inaccurate scoring functions in docking algorithms may result in the selection of compounds with high predicted binding affinity that nevertheless are known experimentally not to bind to the target receptor. Such falsely predicted binders have been termed 'binding decoys'. We posed a question as to whether true binders and decoys could be distinguished based only on their structural chemical descriptors using approaches commonly used in ligand based drug design. We have applied the k-Nearest Neighbor (kNN) classification QSAR approach to a dataset of compounds characterized as binders or binding decoys of AmpC beta-lactamase. Models were subjected to rigorous internal and external validation as part of our standard workflow and a special QSAR modeling scheme was employed that took into account the imbalanced ratio of inhibitors to non-binders (1:4) in this dataset. 342 predictive models were obtained with correct classification rate (CCR) for both training and test sets as high as 0.90 or higher. The prediction accuracy was as high as 100% (CCR = 1.00) for the external validation set composed of 10 compounds (5 true binders and 5 decoys) selected randomly from the original dataset. For an additional external set of 50 known non-binders, we have achieved the CCR of 0.87 using very conservative model applicability domain threshold. The validated binary kNN QSAR models were further employed for mining the NCGC AmpC screening dataset (69653 compounds). The consensus prediction of 64 compounds identified as screening hits in the AmpC PubChem assay disagreed with their annotation in PubChem but was in agreement with the results of secondary assays. At the same time, 15 compounds were identified as potential binders contrary to their annotation in PubChem. Five of them were tested experimentally and showed inhibitory activities in millimolar range with the highest binding constant K(i) of 135 microM. Our studies suggest that validated QSAR models could complement structure based docking and scoring approaches in identifying promising hits by virtual screening of molecular libraries.