NMR identification of endogenous metabolites interacting with fatted and non-fatted human serum albumin in blood plasma: Fatty acids influence the HSA-metabolite interaction

Abstract

Metabolites and their concentrations are direct reporters on body biochemistry. Thanks to technical developments metabolic profiling of body fluids, such as blood plasma, by for instance NMR has in the past decade become increasingly accurate enabling successful clinical diagnostics. Human Serum Albumin (HSA) is the main plasma protein (∼60% of all plasma protein) and responsible for the transport of endogenous (e.g. fatty acids) and exogenous metabolites, which it achieves thanks to its multiple binding sites and its flexibility. HSA has been extensively studied with regard to its binding of drugs (exogenous metabolites), but only to a lesser extent with regard to its binding of endogenous (non-fatty acid) metabolites. To obtain correct NMR measured metabolic profiles of blood plasma and/or potentially extract information on HSA and fatty acids content, it is necessary to characterize these endogenous metabolite/plasma protein interactions. Here, we investigate these metabolite-HSA interactions in blood plasma and blood plasma mimics. The latter contain the roughly twenty metabolites routinely detected by NMR (also most abundant) in normal relative concentrations with fatted or non-fatted HSA added or not. First, we find that chemical shift changes are small and seen only for a few of the metabolites. In contrast, a significant number of the metabolites display reduced resonance integrals and reduced free concentrations in the presence of HSA or fatted HSA. For slow-exchange (or strong) interactions, NMR resonance integrals report the free metabolite concentration, while for fast exchange (weak binding) the chemical shift reports on the binding. Hence, these metabolites bind strongly to HSA and/or fatted HSA, but to a limited degree because for most metabolites their concentration is smaller than the HSA concentration. Most interestingly, fatty acids decrease the metabolite-HSA binding quite significantly for most of the interacting metabolites. We further find that competition between the metabolites for binding is absent for most of these metabolites. These mappings in plasma mimics may thus open new opportunities for improved metabolic profiling of blood plasma. For instance, correct metabolite concentrations can be determined for the non-interacting metabolites and/or concentration corrections made for interacting metabolites. Secondly, the interacting metabolites could be used to act as reporters on HSA and fatty acid concentration in plasma, and thus potentially act as biomarker in diagnostic studies of trauma or cardiovascular diseases. Finally, we find in the blood plasma mimics that after ultrafiltration, commonly used to remove the protein from plasma, the measured concentration equals the total metabolite concentration, except for the strongest binding metabolite citrate.