A Comparative Molecular Dynamics Study of Food-Derived Compounds as PD-L1 Inhibitors: Insights Across Six Flavonoid Subgroups

In this study, we investigated the inhibitory potential of 60 flavonoids from six distinct subgroups on the programmed cell death ligand 1 (PD-L1) dimer through molecular docking and dynamics simulations. Using AutoDock Vina for docking, the binding poses and affinities were evaluated, revealing an average binding affinity of -8.5 kcal/mol for the flavonoids. Among them, ginkgetin exhibited the highest binding free energy of -46.73 kcal/mol, indicating a strong interaction with PD-L1, while diosmin followed closely, with -44.96 kcal/mol. Molecular dynamics simulations were used to further elucidate the dynamic interactions and stability of the flavonoid-PD-L1 complexes, with the analyses showing minimal root mean square deviation (RMSD) and favorable root mean square fluctuation (RMSF) profiles for several compounds, particularly formononetin, idaein, and neohesperidin. Additionally, contact number and hydrogen bond analyses were performed, which highlighted ginkgetin and diosmin as key flavonoids with significant binding interactions, evidenced by their stable conformations and robust molecular interactions throughout the simulations. Ultimately, a cell-based assay confirmed their ability to inhibit the proliferation of cancer cells. These results, validated through cell-based assays, indicate that the strategy of identifying natural compounds with anticancer activity using computational modeling is highly effective.