Exploring the Selectivity of Tetrahydropyrido[1,2-a]isoindolone Derivatives to GSK3 beta and CDK5 by Computational Methods

Dong Ke-Ke,Yang Xue-Yu,Zhao Teng-Teng,Zhu Xiao-Lei
DOI: https://doi.org/10.11862/CJIC.2016.263
2016-01-01
Chinese journal of inorganic chemistry
Abstract:Tetrahydropyrido [1,2-a]isoindolone derivatives are potent inhibitors of glycogen synthase kinase 3 beta (GSK3 beta) instead of homologous cyclin-dependent kinase 5 (CDK5). Molecular docking, molecular dynamics simulation, and MM/PBSA energy calculation are utilized to reveal the kinase inhibitors' selective mechanism at the molecular level for improving selectivity. Dynamic cross-correlation map (DCCM) analysis is applied to study the effect of the inhibitor on the interactions between each residue in CDK5 and GSK3 beta. The results of molecular docking indicate that the binding modes of three inhibitors with two kinases are especially similar, and residues in the binding pockets of two kinases are aligned with each other based on the sequence comparing analysis of crystal structures. The analysis of Root Mean Square Deviation (RMSD) with little fluctuation underlies the stability and reliability of systems. Its values of CDK5 (similar to 0.15 nm) are less than GSK3 beta (similar to 0.17 nm), and the inhibitor with higher value holds stronger flexibility and conformational changes of kinases. In terms of energies, the electrostatic and van der Walls energies are the major interactions for differentiating the activity between the same inhibitor and two kinases. And the polar solvation energy plays pivotal role in discriminating the selectivity of kinase inhibitor. The residue decomposition indicates that the residues Glu97 and Thr138 of GSK3 beta are the key residues for differentiating the inhibitor selectivity. On the other hand, in the aspect of inter-residue interaction in one kinase, results indicate that the dynamic correlation of residues is different during the binding process of CDK5 and GSK3 beta with inhibitors. The correlation of Thr138 in the hinge domain of GSK3 beta with that of residues Val135-Gln206 is positive, while the correlation of Gln85 and Cys83 similar to Ala150 in CDK5 is unclear, which is a key factor to distinguish inhibitor selectivity.
What problem does this paper attempt to address?