Quantifying the human diet in the crosstalk between nutrition and health by multi-targeted metabolomics of food and microbiota-derived metabolites

Abstract

Background: Metabolomics is a powerful tool for investigating the association between nutrition and health status. Although urine is commonly employed for studying the metabolism and transformation of food components, the use of blood samples could be preferable to gain new insights into the bioavailability of diet-derived compounds and their involvement in health. However, the chemical complexity of blood samples hinders the analysis of this biological fluid considerably, which makes the development of novel and comprehensive analytical methods mandatory. Methods: In this work, we optimized a multi-targeted metabolomics platform for the quantitative and simultaneous analysis of 450 food-derived metabolites by ultra-high performance liquid chromatography coupled to tandem mass spectrometry. To handle the chemical complexity of blood samples, three complementary extraction methods were assayed and compared in terms of recovery, sensitivity, precision and matrix effects with the aim of maximizing metabolomics coverage: protein precipitation, reversed solid-phase extraction, and hybrid protein precipitation with solid-phase extraction-mediated phospholipid removal. Results: After careful optimization of the extraction conditions, protein precipitation enabled the most efficient and high-throughput extraction of the food metabolome in plasma, although solid-phase extraction-based protocols provided complementary performance for the analysis of specific polyphenol classes. The developed method yielded accurate recovery rates with negligible matrix effects, and good linearity, as well as high sensitivity and precision for most of the analyzed metabolites. Conclusions: The multi-targeted metabolomics platform optimized in this work enables the simultaneous detection and quantitation of 450 dietary metabolites in short-run times using small volumes of biological sample, which facilitates its application to epidemiological studies.