Abstract
Phenolic compounds have been recognized as promising compounds for the prevention of chronic diseases, including neurodegenerative ones. However, phenolics like flavan-3-ols (F3O) are poorly absorbed along the gastrointestinal tract and structurally rearranged by gut microbiota, yielding smaller and more polar metabolites like phenyl-γ-valerolactones, phenylvaleric acids and their conjugates. The present work investigated the ability of F3O-derived metabolites to cross the blood-brain barrier (BBB), by linking five experimental models with increasing realism. First, an in silico study examined the physical-chemical characteristics of F3O metabolites to predict those most likely to cross the BBB. Some of these metabolites were then tested at physiological concentrations to cross the luminal and abluminal membranes of brain microvascular endothelial cells, cultured in vitro. Finally, three different in vivo studies in rats injected with pure 5-(3',4'-dihydroxyphenyl)-γ-valerolactone, and rats and pigs fed grapes or a F3O-rich cocoa extract, respectively, confirmed the presence of 5-(hydroxyphenyl)-γ-valerolactone-sulfate (3',4' isomer) in the brain. This work highlighted, with different experimental models, the BBB permeability of one of the main F3O-derived metabolites. It may support the neuroprotective effects of phenolic-rich foods in the frame of the "gut-brain axis".