Exposomic Biomonitoring\nof Polyphenols by Non-Targeted\nAnalysis and Suspect Screening

Polyphenols, prevalent\nin plants and fungi, are investigated intensively\nin nutritional and clinical settings because of their beneficial bioactive\nproperties. Due to their complexity, analysis with untargeted approaches\nis favorable, which typically use high-resolution mass spectrometry\n(HRMS) rather than low-resolution mass spectrometry (LRMS). Here,\nthe advantages of HRMS were evaluated by thoroughly testing untargeted\ntechniques and available online resources. By applying data-dependent\nacquisition on real-life urine samples, 27 features were annotated\nwith spectral libraries, 88 with in silico fragmentation,\nand 113 by MS1 matching with PhytoHub, an online database\ncontaining >2000 polyphenols. Moreover, other exogenous and endogenous\nmolecules were screened to measure chemical exposure and potential\nmetabolic effects using the Exposome-Explorer database, further annotating\n144 features. Additional polyphenol-related features were explored\nusing various non-targeted analysis techniques including MassQL for\nglucuronide and sulfate neutral losses, and MetaboAnalyst for statistical\nanalysis. As HRMS typically suffers a sensitivity loss compared to\nstate-of-the-art LRMS used in targeted workflows, the gap between\nthe two instrumental approaches was quantified in three spiked human\nmatrices (urine, serum, plasma) as well as real-life urine samples.\nBoth instruments showed feasible sensitivity, with median limits of\ndetection in the spiked samples being 10–18 ng/mL for HRMS\nand 4.8–5.8 ng/mL for LRMS. The results demonstrate that, despite\nits intrinsic limitations, HRMS can readily be used for comprehensively\ninvestigating human polyphenol exposure. In the future, this work\nis expected to allow for linking human health effects with exposure\npatterns and toxicological mixture effects with other xenobiotics.