Rational design of novel phenol ether derivatives as non-covalent proteasome inhibitors through 3D-QSAR, molecular docking and ADMET prediction

Abstract

Aim: The purpose of this paper is to use different structures and ligand-based drug design methods properly to provide theoretical guidance for the design of novel non-covalent proteasome inhibitors, and conduct theoretical analysis of the binding interaction mode between receptors and ligands. At the same time, the pharmacokinetic (PK) prediction, drug-likeness, and synthesis prediction were made for the screened novel drugs. Therefore, potentially attractive non-covalent proteasome inhibitors with low toxicity could be found as anticancer drugs. Methods: In this work, computer-aided drug design methods, including quantitative structure-activity relationship (QSAR), molecular docking, absorption, distribution, metabolism, excretion, and toxicology (ADMET) prediction, and drug-likeness prediction methods were performed. Results: In this study, the structure-activity relationship (SAR) of a series of non-covalent proteasome inhibitors were studied and the optimal comparative molecular field analysis (CoMFA; Q2 = 0.574, r2 = 0.999, r2pred = 0.755) and comparative molecular similarity indices analysis (CoMSIA)-SEHA (Q2 = 0.584, r2 = 0.989, r2pred = 0.921) models were obtained. According to the results of the QSAR model, some vital clues were found that would effectively enhance the biological activity of the compound. Based on these clues, 24 novel non-covalent proteasome inhibitors (D01–D24) were finally designed and screened. While the binding models between proteasome [protein data bank (PDB) code: 3MG6] and three representative compounds (15, 20, and D24) were also analyzed by using the molecular docking method. The results suggested that hydrogen bond and hydrophobic interaction played a key role in binding interaction between the receptor and ligand. In addition, the results of ADMET prediction indicated that the new designed compounds had reasonable PK parameters and drug-like properties. Conclusions: These statistical results can provide theoretical guidance for structural optimization, design, and synthesis of more effective non-covalent proteasome inhibitors in the future.