Influence of mathematical models and correction factors on binding results of polyphenols and retinol with β-lactoglobulin measured with fluorescence quenching

beta-Lactoglobulin is frequently researched as food grade nanotransporter for hydrophobic compounds like polyphenols and fatty soluble vitamins. The effects of self-fluorescence and inner filtering caused by ligands on the analysis of ligand-beta-lactoglobulin-binding were evaluated using fluorescence quenching data. Without correction for self-fluorescence or inner filtering the binding constants are either overestimated or binding is erroneously assumed. Additionally, the variance between the derived numbers of binding sites resulting by using different mathematical models with the same data set is about 50 %. Thus here we suggest the use of at least two very different mathematical approaches when deriving fluorescence titration data and evaluating the underlying restrictions of the respective model thoroughly. The greatest differences were found to concern K'a or K'd, whereas n seems to be relatively stable. The Cogan plot seems to be the best choice for FQ calculations, given that it does not require an estimation of added ligand concentration and that it is insensitive to residual fluorescence.