An ensemble docking-based virtual screening according to different TRPV1 pore states toward identifying phytochemical activators

Abstract

Transient receptor potential vanilloid type 1 (TRPV1) ion channels are involved in the detection and transduction of nociceptive stimuli. The up-regulation of TRPV1 transcription induced by inflammation and nerve damage leads to pain, therefore, modulating TRPV1 is a useful way of relieving pain. An ensemble docking-based virtual screening (VS) was employed to identify phytochemicals with high binding affinity for TRPV1 from a wide range of vanilloid derivatives extracted from natural products. Known activators and non-activators formed the validation dataset while two models of random forest and logistic regression were used as classifiers for machine learning. An ensemble was constructed with multiple TRPV1 structures with different pore channel conformations: open to closed states and sub-states. TRPV1 structures with pore channels in the closed state and sub-states showed high discrimination power for a validation dataset based on classifiers. Three top high-affinity ligands were screened by the docking of phytochemicals on an ensemble established from molecular dynamics (MD) simulation of the best representative structure. Favourable hydrophobic contacts direct the binding of phytochemicals in the TRPV1 vanilloid pocket. Pore conformational transitions that took place during simulation displayed an increase in the distance between two opposing MET644s or ILE679s in the pore channel and are attributed to a probable activation mechanism to the screened phytochemicals. Therefore, the presented results pave the way for developing potent anti-nociceptive agents for TRPV1.