Structure and ligand-based drug discovery of IL-4 inhibitors via interaction-energy-based learning approaches

Abstract

Interleukin-4 (IL-4), an anti-inflammatory cytokine plays significant in the development of various diseases especially asthmatic allergies. Previous structural and functional studies of IL-4 with its receptor bring forth different types of inhibitors to block their interaction but each of them failed in clinical trials. Since, no synthetic molecules have been identified against IL-4, so far. Therefore, 21 in-house tested IL-4 inhibitors were blindly docked over the entire surface of IL-4 to predict a suitable and druggable binding site as the crystal structure of IL-4 protein in complex with ligand has not been reported yet. After binding site prediction, both ligand-based and structure-based pharmacophore were generated to screen three ZINC libraries (24.5 M) i.e. purchasable, natural product and natural derivative. A total 5,800 top-scored compounds were further subjected towards score-based screening to find the potential leads. Following protein-ligand interaction fingerprints (PLIF) and molecular visualization of selected hits, six top-scored compounds (five from purchasable and one from natural product library) were further moved towards their stability dynamics, followed by their absolute binding free energy and residue-based energy decomposition calculation by MM-GBSA method. These efforts help us to reveal the key factors responsible for ligand binding that might help to improve the binding and stability of these newly discovered hits by structural modifications. Communicated by Freddie R. Salsbury