Identification and neuroprotective evaluation of a potential c-Jun N-terminal kinase 3 inhibitor through structure-based virtual screening and in-vitro assay

• Published in: Journal of computer-aided molecular design Vol. 34; no. 6; pp. 671 - 682
• Main Authors: Rajan, Ravi Kumar, Ramanathan, M.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01-06-2020; Springer Nature B.V
• Subjects: Animal Anatomy; Assaying; Binding; Binding energy; Chemistry; Chemistry and Materials Science; Computer Applications in Chemistry; Enzymes; Free energy; Histology; Hydrogen bonds; Ischemia; Kinases; Methionine; Morphology; Pharmacology; Physical Chemistry; Receptors; MM/GBSA (molecular mechanics/ generalized born and surface area continuum solvation); IFD (Induced fit docking); GLIDE (grid-based ligand docking with energetics); RRD (Rigid receptor docking); JNK3I (c-Jun N-terminal kinase 3 inhibitor)
• ISSN: 0920-654X; 1573-4951
• DOI: 10.1007/s10822-020-00297-y

The c-Jun N-terminal kinase 3 (JNK3) signaling cascade is activated during cerebral ischemia leading to neuronal damage. The present study was carried out to identify and evaluate novel JNK3 inhibitors using in-silico and in-vitro approach. A total of 380 JNK3 inhibitors belonging to different organic groups was collected from the previously reported literature. These molecules were used to generate a pharmacophore model. This model was used to screen a chemical database (SPECS) to identify newer molecules with similar chemical features. The top 1000 hits molecules were then docked against the JNK3 enzyme coordinate following GLIDE rigid receptor docking (RRD) protocol. Best posed molecules of RRD were used during induced-fit docking (IFD), allowing receptor flexibility. Other computational predictions such as binding free energy, electronic configuration and ADME/tox were also calculated. Inferences from the best pharmacophore model suggested that, in order to have specific JNK3 inhibitory activity, the molecules must possess one H-bond donor, two hydrophobic and two ring features. Docking studies suggested that the main interaction between lead molecules and JNK3 enzyme consisted of hydrogen bond interaction with methionine 149 of the hinge region. It was also observed that the molecule with better MM-GBSA dG binding free energy, had greater correlation with JNK3 inhibition. Lead molecule (AJ-292-42151532) with the highest binding free energy (dG = 106.8 Kcal/mol) showed better efficacy than the SP600125 (reference JNK3 inhibitor) during cell-free JNK3 kinase assay (IC50 = 58.17 nM) and cell-based neuroprotective assay (EC50 = 7.5 µM).
Graphic Abstract