Supramolecular delivery systems for polyphenols: A green approach to predict in vivo permeability through an in vitro setup.

The use of in vitro markers able to reproduce the in vivo permeability and diffusivity of orally administered drugs, could represent an innovative starting point for the formulation of delivery systems, in particular for low soluble and low permeable drugs belonging to BCS class II and IV. Considering the great interest in the green pharmaceutical approaches and the increasing use of natural molecules as novel therapeutic drugs, in this study, rutin, hesperidin and curcumin have been selected as lipophilic model drugs to investigate their possible enhancement of their permeability and bioavailability after oral administration. As the low solubility of the three drugs hinders their application, β-cyclodextrins (CD), amphiphilic natural moieties able to form stable inclusion complexes, have been considered to promote their solubilization. Notably, hydroxypropyl-β-CD (HPBCD) and methyl-β-CD (MBCD), have been selected and the formation of the inclusion complexes with a stoichiometric ratio of 1:1 has been detected through phase-solubility studies and rationalized via docking calculations, revealing a strong complexation and an increased hydrophilicity of the systems. The diffusion experiments performed through the novel UV-Vis localized spectroscopy method confirmed a the extremely high stability of the CD-drugs complexes, especially in the cease of curcumin, which makes this as the predominant chemical specie to diffuse and permeates. The PermeaPad® plate, an in vitro cell-free assay, allowed to investigate the permeability behavior of the drugs, demonstrated that the type of β-cyclodextrins can influence the permeability through the biomimetic membrane, reflecting the effect of the unstirred water layer (UWL). Moreover, in the case of curcumin, the spectroscopic-mathematical approach suggested the formation of nano-supramolecular systems, detected by DLS, supporting the precision of the fitting model.