Decoding Structural Fingerprints to Design and Elucidate the Mechanism of Action of Prospective Cholesteryl Ester Transfer Protein Drugs.

Abstract

Cardiovascular diseases (CVDs) have become a leading cause of deaths globally. Recent studies have shown that increasing the level of high-density lipoproteins (HDL) is one of the potential avenues to halt CVD progression. This could be achieved by modulating the neutral lipid transfer activity of cholesteryl ester transfer protein (CETP), a key target in developing effective cardioprotective drugs. This study aims to identify important structural fingerprints and functional moieties as "good" and "bad" contributors toward CETP inhibition, using machine learning (ML) and quantitative structure-activity relationship-based approaches. Results suggest unsaturated heterocyclic rings and trifluoromethyl substitutions as potential promoters and aliphatic carboxylic acid and ester moieties as the detractors in CETP inhibition. Molecular dynamics (MD) simulations of CETP in complexation with recently reported Obicetrapib with "good" fingerprints versus a clinically failed inhibitor, Torcetrapib shows superior inhibitory potential of the former due to stronger binding and better shape complementarity with the CETP hydrophobic tunnel. By leveraging the potentials of ML and MD simulations, this comprehensive study helps judicious pick of the right functional moieties for designing next generation CETP drugs targeting CVD.