Discovery of novel tropomyosin receptor kinase A inhibitors by virtual screening merging ligand-based and structure-based methods

Tropomyosin receptor kinase (Trk) is a crucial broad-spectrum Anticancer target. Its inhibitors are among the first "tumor agnostic" drugs approved for pan-cancer therapy. TrkA, a subtype of tropomyosin receptor kinase, is one of the most frequently detected in human cancers and has emerged as a key target for Anticancer drug development. Our work employed a virtual screening approach integrating ligand- and structure-based strategies to identify novel inhibitors targeting TrkA. A large-scale virtual screening system was constructed for a screening database of 9.87 million commercial compounds containing a broad chemical space. A funnel-shaped multistage combinatorial screening process was adopted, with the combined time-consuming and precision of the methods in the following order: screening based on pharmacophore and shape, screening based on the QSAR model obtained from the prequel study, screening based on molecular docking, prediction of ADME properties, and patent searching. The above virtual screening system received sixteen potentially active novel inhibitors. Subsequent enzymatic activity assays revealed one hit compound ZA16 with nanomolar-level inhibitory potency, validating the efficacy of the large-scale virtual screening system. This approach can be a robust framework for guiding drug discovery and design across diverse compound libraries. Furthermore, molecular dynamics simulations revealed the interaction between this novel inhibitor and the TrkA protein domain, emphasizing the crucial role of the hydrogen bond at MET592. Through the analysis of MM/GBSA binding free energy, this study offered significant insights for drug discovery and design targeting the same protein.

Large-scale virtual screening; Machine learning; Molecular docking; Nanomolar-level inhibtor; Tropomyosin receptor kinase A (TRKA).