Binding mechanism of CDK5 with roscovitine derivatives based on molecular dynamics simulations and MM/PBSA methods

Abstract

Roscovitine derivatives are potent inhibitors of cyclin-dependent kinase 5 (CDK5), but they exhibit different activities, which has not been understood clearly up to now. On the other hand, the task of drug design is difficult because of the fuzzy binding mechanism. In this context, the methods of molecular docking, molecular dynamics (MD) simulation, and binding free energy analysis are applied to investigate and reveal the detailed binding mechanism of four roscovitine derivatives with CDK5. The electrostatic and van der Waals interactions of the four inhibitors with CDK5 are analyzed and discussed. The calculated binding free energies in terms of MM-PBSA method are consistent with experimental ranking of inhibitor effectiveness for the four inhibitors. The hydrogen bonds of the inhibitors with Cys83 and Lys33 can stabilize the inhibitors in binding sites. The van der Waals interactions, especially the pivotal contacts with Ile10 and Leu133 have larger contributions to the binding free energy and play critical roles in distinguishing the variant bioactivity of four inhibitors. In terms of binding mechanism of the four inhibitors with CDK5 and energy contribution of fragments of each inhibitor, two new CDK5 inhibitors are designed and have stronger inhibitory potency.

Recent Developments and Applications of the MMPBSA Method

“…MMPBSA was widely applied in the development of anticancer compounds where kinases were identified as the most promising targets. Recent inhibitor design includes the epidermal growth factor receptor kinase domain (Li et al, 2015 ; Moonrin et al, 2015 ; Zhao et al, 2017 ), anaplastic lymphoma kinase (Kong et al, 2015 ), cyclin-dependent kinases (Li X. L. et al, 2014 ; Czelen, 2016 ; Dong et al, 2016a , b , 2017 ; Arba et al, 2017 ), extracellular signal-regulated kinase 2 (Chen, 2017 ), casein kinase 2 (Wang X. W. et al, 2014 ), sphingosine kinases (Fang et al, 2016 ), Src/Abl tyrosine kinases (Fong, 2015 ; Ma et al, 2015 ), RET tyrosine kinase (Gao et al, 2015 ), PRK1 (Slynko et al, 2014 ), Akt kinase (Lu et al, 2015 ), phosphatidylinositol 3 kinase (Bian et al, 2014 ), and Myt1 kinase (Wichapong et al, 2014 ). As an illustration of the method's performance, Slynko et al observed a correlation coefficient of 0.78 between the experimental pIC 50 of 26 PRK1 inhibitors and computational MMPBSA binding affinities.…”

Multitargeting the Action of 5-HT6 Serotonin Receptor Ligands by Additional Modulation of Kinases in the Search for a New Therapy for Alzheimer’s Disease: Can It Work from a Molecular Point of View?

“…Importantly, several studies including X-ray and molecular modeling have indicated the pivotal role of cysteine (Cys83) in ligands binding to CDK5 in the ATP binding pocket [ 139 , 140 ]. This amino acid may be S-nitrosylated and, fascinatingly, the perturbation of such a process leads to the enhancement of dendrite development in cultured hippocampal neurons, which, of course, influences overall neuronal development [ 120 ].…”