Histone acetylation/deacetylation is a key mechanism for transcription regulation which plays an important role in control of gene expression, tissue growth, and development. In particular, histone deacetylase 7 (HDAC7), a member of class IIa HDACs, is crucial to maintain cell homeostasis, and HDAC7 has emerged as a new target for cancer therapy. In this study, molecular docking was applied to screen candidate inhibitors and 21 compounds were found. Following the 50 ns molecular dynamics simulations and binding free energy calculation, ZINC00156160, ZINC01703144, ZINC04293665, and ZINC13900201 were identified as potential HDAC7 inhibitors, which would provide a sound starting point for further studies involving molecular modeling coupled with biochemical experiments. Meanwhile, similarity computation and substructure search were combined, and then we found that compounds sharing common backbone "CC(=O)N[C@@H](CSc1ccccc1)C(=O)O" could be efficient to inhibit the bioactivity of HDAC7. Then comparative molecular similarity indices analysis (CoMSIA) techniques were implemented to investigate the relationship between properties of the substituent group and bioactivities of small molecules. The CoMSIA model exhibited powerful predictivity, with satisfactory statistical parameters such as q2 of 0.659, R2 of 0.952, and F of 268.448. Contour maps of the CoMSIA model gave insight into the feature requirements of the common backbone for the HDAC7 inhibitory activity. Finally, details of designing novel HDAC7 inhibitors were confirmed by a combination of receptor-based docking and ligand-based structure-activity relationship. Communicated by Ramaswamy H. Sarma.
Keywords: CoMSIA; Histone deacetylase 7; molecular docking; molecular dynamics simulations; similarity computation.