Novel protocol for enhanced sampling of binding pocket dynamics and its integration into the sirtuin 2 inhibitors virtual screening campaign

Abstract

Inhibitors of sirtuin 2 (SIRT2i) represent a promising group of novel therapeutics in treatment of the age-related disorders, including carcinomas, neurodegenerative diseases, metabolic syndrome etc. Despite the promising preclinical results, none of the known SIRT2i reached clinical trials. One of the main obstacles in the structure-based drug design of novel SIRT2i represents intricate conformational dynamics of sirtuin 2 (SIRT2) binding pocket. In order to facilitate the discovery of novel SIRT2i, we have developed the protocol for enhanced sampling of the binding pocket dynamics and validated it by integration into structure-based virtual screening (SBVS) pipeline for discovery of novel SIRT2i. Developed protocol relies on well-tempered metadynamics simulations using the set of pocket-related collective variables derived from time-lagged independent component analysis (tICA). Our protocol outperformed classical molecular dynamics in search for alternative conformational states of the binding pocket. Additionally, the protocol was able to reveal the existence of cryptic subpocket inside SIRT2 binding pocket. To validate our findings, the protocol was implemented into SBVS which resulted in significant expansion of SIRT2i chemical space. To further probe the potential of expanded chemical space in discovery of chemically novel groups of SIRT2i, a few virtual hit molecules were selected and tested in vitro. Compound NDJ18 was shown to be potent and selective SIRT2i with anticancer activity on triple-negative breast cancer cell line MDA-MB-231. Experimental validation supported future generalization of the protocol by application on wider scope of challenging protein targets.