Identification of polyphenols as novel neuropilin-1 cendR pocket inhibitors to block SARS-CoV-2 entry and enhance variant resistance.

Abstract

Neuropilin-1 (NRP-1) functions as an essential co-receptor for SARS-CoV-2, facilitating viral entry by binding the spike protein's C-end rule (CendR) motif in its b1 domain, yet it has received less attention than ACE2 in therapeutic development. This in silico study evaluates plant-derived polyphenols as potential selective inhibitors of the NRP-1 CendR pocket to disrupt SARS-CoV-2 engagement, addressing limitations of synthetic inhibitors like EG01377, which exhibit modest affinity (-5.83 kcal/mol) and potential off-target risks. High-throughput molecular docking of 10,000 phytochemicals using AutoDock Vina identified 10 polyphenols with binding affinities ranging from -9.87 to -6.63 kcal/mol, led by 6"-O-acetyldaidzin (-9.87 kcal/mol) and phloretin (-8.64 kcal/mol), forming stable hydrogen bonds and π-cation interactions with critical residues (e.g., THR-401, GLU-367 for 6"-O-acetyldaidzin; PRO-311, ILE-400 for phloretin), as visualized in Discovery Studio. Notably, four of the top inhibitors are isoflavonoid derivatives, highlighting a chemical class enrichment. Molecular dynamics simulations over 100 ns using Desmond indicated moderate complex stability (RMSD: 0.6-3.8 Å; RMSF <0.5 Å at binding site). ADMET-Tox profiling via SwissADME and ProTox-II revealed drug-like properties, including high gastrointestinal absorption (>70% for leads) and low toxicity (classes 4-5), though 6"-O-acetyldaidzin shows limited bioavailability due to its high H-bond acceptor count (10) and large size, suggesting need for formulation optimization. The NRP-1 b1 homology model, built with SWISS-MODEL, exhibited high fidelity (GMQE: 0.79; Ramachandran favored regions: 90.3%). This focused computational screening of polyphenols against NRP-1 complements prior studies and identifies candidates for experimental validation as potential SARS-CoV-2 inhibitors. Limitations include the in silico nature, and lack of membrane/sialic acid models, necessitating in vitro and in vivo testing against SARS-CoV-2 variants.