Structural Basis for Low-Affinity Binding of Non-R2 Carboxylate-Substituted Tricyclic Quinoline Analogs to CK2α: Comparative Molecular Dynamics Simulation Studies

Machine Learning Models for the Classification of CK2 Natural Products Inhibitors with Molecular Fingerprint Descriptors

Casein kinase 2 (CK2) is considered an important target for anti-cancer drugs. Given the structural diversity and broad spectrum of pharmaceutical activities of natural products, numerous studies have been performed to prove them as valuable sources of drugs. However, there has been little study relevant to identifying structural factors responsible for their inhibitory activity against CK2 with machine learning methods. In this study, classification studies were conducted on 115 natural products as CK2 inhibitors. Seven machine learning methods along with six molecular fingerprints were employed to develop qualitative classification models. The performances of all models were evaluated by cross-validation and test set. By taking predictive accuracy(CA), the area under receiver operating characteristic (AUC), and (MCC)as three performance indicators, the optimal models with high reliability and predictive ability were obtained, including the Extended Fingerprint-Logistic Regression model (CA = 0.859, AUC = 0.826, MCC = 0.520) for training test andPubChem fingerprint along with the artificial neural model (CA = 0.826, AUC = 0.933, MCC = 0.628) for test set. Meanwhile, the privileged substructures responsible for their inhibitory activity against CK2 were also identified through a combination of frequency analysis and information gain. The results are expected to provide useful information for the further utilization of natural products and the discovery of novel CK2 inhibitors.

Structure-based identification of novel CK2 inhibitors with a linear 2-propenone scaffold as anti-cancer agents

Protein kinase CK2 has emerged as an attractive cancer therapeutic target. Previous studies have highlighted the challenge of optimizing CK2 ATP-competitive inhibitors that have low druggability due to their polycyclic ring scaffolds. Therefore the development of novel inhibitors with non-polycyclic scaffolds emerges as a promising strategy for drug discovery targeting CK2. In this current study, based on the similar predicted binding poses of the linear 2-propenone scaffold of isoliquiritigenin with that of the polycyclic inhibitor CX-4945, a series of 2-propenone derivatives containing an amine-substituted five-membered heterocycle and a benzoic acid were designed, synthesized and evaluated for their in vitro CK2 inhibition and anti-cancer activity. Compound 8b was found to be the most potent CK2 inhibitor (IC₅₀ = 0.6 μM) with the anti-proliferative activity on HepG2 cancer cells (IC₅₀ = 14 μM), compared to the activity of isoliquiritigenin (IC₅₀ = 17 μM and 51 μM, respectively). Molecular docking was performed to understand the binding modes of the newly designed 2-propenone derivatives with CK2. Compound 8b formed the most favorable network of hydrogen bonds with both the hinge region and positive area. Our results indicate that CK2 derivatives with a linear 2-propenone scaffold are promising candidates for anti-cancer drug discovery.

Insights into the Impact of Linker Flexibility and Fragment Ionization on the Design of CK2 Allosteric Inhibitors: Comparative Molecular Dynamics Simulation Studies

Protein kinase is a novel therapeutic target for human diseases. The off-target and side effects of ATP-competitive inhibitors preclude them from the clinically relevant drugs. The compounds targeting the druggable allosteric sites outside the highly conversed ATP binding pocket have been identified as promising alternatives to overcome current barriers of ATP-competitive inhibitors. By simultaneously interacting with the αD region (new allosteric site) and sub-ATP binding pocket, the attractive compound CAM4066 was named as allosteric inhibitor of CK2α. It has been demonstrated that the rigid linker and non-ionizable substituted fragment resulted in significant decreased inhibitory activities of compounds. The molecular dynamics simulations and energy analysis revealed that the appropriate coupling between the linker and pharmacophore fragments were essential for binding of CAM4066 with CK2α. The lower flexible linker of compound 21 lost the capability of coupling fragments A and B to αD region and positive area, respectively, whereas the methyl benzoate of fragment B induced the re-orientated Pre-CAM4066 with the inappropriate polar interactions. Most importantly, the match between the optimized linker and pharmacophore fragments is the challenging work of fragment-linking based drug design. These results provide rational clues to further structural modification and development of highly potent allosteric inhibitors of CK2.

Exploring the Pivotal Role of the CK2 Hinge Region Sub-Pocket in Binding with Tricyclic Quinolone Analogues by Computational Analysis

Protein kinase CK2 has been considered as an attractive therapeutic target of cancer therapy. The tricyclic quinoline compound CX-4945 is the first representative of CK2 inhibitors used in human clinical trials. The binding of non-2,6-naphtyridine substituted compounds 27e (IC₅₀ > 500 nM) and 27h (IC₅₀ > 1000 nM) to CK2 is abolished. However, the unbinding mechanisms due to the key pharmacophore group replacement of compounds 27e and 27h are unveiled. In the present work, combined computational analysis was performed to investigate the underlying structural basis of the low-affinity of two systems. As indicated in the results, the loss of hydrogen bonds between the non-2,6-naphtyridine and the hinge region destroyed the proper recognition of the two complexes. Besides, the allosteric mechanisms between the deviated ligands and the changed regions (G-loop, C-loop and β4/β5 loop) are proposed. Furthermore, energetic analysis was evaluated by detailed energy calculation and residue-based energy decomposition. More importantly, the summary of known polar pharmacophore groups elucidates the pivotal roles of hinge region sub-pocket in the binding of CK2 inhibitors. These results provide rational clues to the fragment-based design of more potent CK2 inhibitors.

The Development of CK2 Inhibitors: From Traditional Pharmacology to in Silico Rational Drug Design

Casein kinase II (CK2) is an ubiquitous and pleiotropic serine/threonine protein kinase able to phosphorylate hundreds of substrates. Being implicated in several human diseases, from neurodegeneration to cancer, the biological roles of CK2 have been intensively studied. Upregulation of CK2 has been shown to be critical to tumor progression, making this kinase an attractive target for cancer therapy. Several CK2 inhibitors have been developed so far, the first being discovered by "trial and error testing". In the last decade, the development of in silico rational drug design has prompted the discovery, de novo design and optimization of several CK2 inhibitors, active in the low nanomolar range. The screening of big chemical libraries and the optimization of hit compounds by Structure Based Drug Design (SBDD) provide telling examples of a fruitful application of rational drug design to the development of CK2 inhibitors. Ligand Based Drug Design (LBDD) models have been also applied to CK2 drug discovery, however they were mainly focused on methodology improvements rather than being critical for de novo design and optimization. This manuscript provides detailed description of in silico methodologies whose applications to the design and development of CK2 inhibitors proved successful and promising.

Underlying mechanisms of cyclic peptide inhibitors interrupting the interaction of CK2α/CK2β: comparative molecular dynamics simulation studies

Protein-protein interactions (PPIs) are fundamental to all biological processes. Recently, the CK2β-derived cyclic peptide Pc has been demonstrated to efficiently antagonize the CK2α/CK2β interaction and strongly affect the phosphorylation of CK2β-dependent CK2 substrate specificity. The binding affinity of Pc to CK2α is destroyed to different extents by two single-point mutations of Tyr188 to Ala (Y188A) and Phe190 to Ala (F190A), which exert negative effects on the inhibitory activity (IC50) of Pc against the CK2α/CK2β interaction from 3.0 μM to 54.0 μM and ≫100 μM, respectively. However, the structural influences of Y188A and F190A mutations on the CK2α-Pc complex remain unclear. In this study, comparative molecular dynamics (MD) simulations, principal component analysis (PCA), domain cross-correlation map (DCCM) analysis and energy calculations were performed on wild type (WT), Y188A mutant, and F190A mutant systems. The results revealed that ordered communications between hydrophobic and polar interactions were essential for CK2α-Pc binding in the WT system. In addition to the loss of the hydrogen bond between Gln36 of CK2α and Gly189 of Pc in the two mutants, the improper recognition mechanisms occurred through different pathways. These pathways included the weakened hydrophobic interactions in the Y188A mutant as well as decreased polar and hydrophobic interactions in the F190A mutant. The energy analysis results qualitatively elucidated the instability of the two mutants and energetic contributions of the key residues. This study not only revealed the structural mechanisms for the decreased binding affinity of Y188A and F190A mutant CK2α-Pc complexes, but also provided valuable clues for the rational design of CK2α/CK2β subunit interaction inhibitors with high affinity and specificity.

Microsecond molecular dynamics simulations provide insight into the ATP-competitive inhibitor-induced allosteric protection of Akt kinase phosphorylation

Akt is a serine/threonine protein kinase, a critical mediator of growth factor-induced survival in key cellular pathways. Allosteric signaling between protein intramolecular domains requires long-range communication mediated by hotspot residues, often triggered by ligand binding. Here, based on extensive 3 μs explicit solvent molecular dynamics (MD) simulations of Akt1 kinase domain in the unbound (apo) and ATP-competitive inhibitor, GDC-0068-bound states, we propose a molecular mechanism for allosteric regulation of Akt1 kinase phosphorylation by GDC-0068 binding to the ATP-binding site. MD simulations revealed that the apo Akt1 is flexible with two disengaged N- and C-lobes, equilibrated between the open and closed conformations. GDC-0068 occupancy of the ATP-binding site shifts the conformational equilibrium of Akt1 from the open conformation toward the closed conformation and stabilizes the closed state. This effect enables allosteric signal propagation from the GDC-0068 to the phosphorylated T308 (pT308) in the activation loop and restrains phosphatase access to pT308, thereby protecting the pT308 in the GDC-0068-bound Akt1. Importantly, functional hotspots involved in the allosteric communication from the GDC-0068 to the pT308 are identified. Our analysis of GDC-0068-induced allosteric protection of Akt kinase phosphorylation yields important new insights into the molecular mechanism of allosteric regulation of Akt kinase activity.