Development of new approaches for the evaluation of the metabolism of bioactive compounds of nutritional interest

This doctoral thesis developed, optimized, and applied mass spectrometry-based methods to study the in vivo metabolism of dietary (poly)phenols, with a focus on the spatial distribution of phase II and microbial-derived metabolites in animal tissues. Traditional LC-MS methods provide sensitive and selective quantitative data on (poly)phenol metabolites in plasma, urine, and organ extracts. However, this approach requires tissue homogenization, leading to a permanent loss of spatial information about where metabolites accumulate within organs. To overcome this issue, the thesis introduced desorption electrospray ionization mass spectrometry imaging (DESI-MSI) as a complementary technique capable of mapping the in situ localization of endogenous and exogenous compounds at the tissue level. The work is organized into four studies. The systematic review collected and critically assessed all published MSI applications related to (poly)phenol absorption and distribution in animal tissues. The review revealed that, to date, only a few in vivo studies have visualized phenolic compounds in organs such as the liver, kidney, brain, intestine, and eye, and that overall MALDI is the leading MSI technique. It highlights methodological heterogeneity, such as variations in dose levels, administration routes, MSI platforms, matrices, and target tissues. It also discussed significant limitations, including MALDI matrix interference in the low m/z range and the limited use of matrix-free ambient techniques. Despite these challenges, MSI proved to be a valuable approach for revealing tissue-specific distributions of phenolic metabolites that bulk LC-MS analysis cannot capture, underscoring the need for targeted ionization strategies for (poly)phenols and for more physiologically relevant in vivo models. The second study focused on ellagitannin-derived urolithins and aimed to define suitable DESI-MSI conditions to maximize their detection in liver tissue. This study used a mimetic model prepared by homogenizing rabbit liver with urolithin standards (aglycones and glucuronides) to systematically optimize capillary voltage (Cap V), heated transfer line (HTL) temperature, and solvent composition. Under optimized conditions, DESI-MSI reliably detected 3,8-dihydroxy-urolithin and 3-hydroxy-urolithin at concentrations down to 5 µg/g. Glucuronidated conjugates remained undetectable and only became visible when spiked at substantially higher concentrations. Trials on glass slides showed that glucuronides ionize readily in the absence of tissue, suggesting that matrix composition and analyte-matrix interactions limit desorption in biological samples. When the optimized method was applied to rats administered 3-hydroxy-urolithin intraperitoneally, DESI-MSI visualized the distribution of urolithin-3-sulfate in the kidney, whereas urolithin metabolites in the liver and heart, detected by LC-MS/MS, remained below the DESI detection threshold. This study illustrated the feasibility of DESI-MSI for mapping urolithin metabolites in vivo, clarifying concentration- and matrix-dependent sensitivity limits and the need to complement imaging data with quantitative LC-MS. The third study expanded DESI-MSI optimization to flavan-3-ol metabolites and their microbial catabolites using brain, heart, liver, and kidney-enriched mimetic models. For each organ, two distinct mimetic models were prepared, one with (epi)catechins and their phase II conjugates and the other with phenyl-γ-valerolactones, resulting in a total of eight homogenized-enriched models. The sections were analyzed under various combinations of CapV, HTL temperature, and solvent acidification. The results indicated that the optimal instrument parameters differ according to (poly)phenolic class and tissue analyzed. The addition of a small amount of formic acid (0.01% v/v) to the DESI solvent enhanced signal intensity for all compounds, achieving desorption limits of around 5 µg/g for many metabolites, consistent with concentrations found in animal tissues following nutritional dosing. Matrix suppression and analyte structure was confirmed as significant limitations. The last study applied the optimized DESI-MSI conditions in an in vivo mouse model to trace the metabolic fate of deuterated (+)-catechin-d4 after oral gavage. Stability tests revealed a rapid loss of two deuterium atoms under acidic conditions similar to those in the stomach, resulting in (+)-catechin-d2 before absorption. The study combined DESI-MSI analysis with quantitative LC-MS/MS of the stomach, small intestine, colon, liver, and kidney at multiple time points. This approach allowed visualization of the metabolite distribution in situ at time points and organs corresponding to the highest or near-highest concentrations quantified by LC-MS/MS. The traditional LC approach reconstructed organ-specific metabolic and kinetic profiles, highlighting that glucuronidated and methylated forms, not detected by DESI, were present but in lower concentrations and below the imaging detection limit. This study provided the first organ-specific map of catechin-derived metabolites in mice and confirmed that microbial phenyl-γ-valerolactone formation and sulfate conjugation are the major biotransformation pathways following dietary flavan-3-ol intake. Overall, the four studies showed that DESI-MSI, when properly optimized and combined with LC-MS/MS, can investigate dietary (poly)phenol metabolism at the tissue level. The doctoral thesis used mimetic models as effective tools for method development, clarified how DESI-MSI can map low-molecular-weight phenolic metabolites, and identified important limitations in sensitivity and matrix effects. These contributions bridge the gap between traditional nutrikinetic data and spatially resolved information, further advancing the field of spatial metabolomics for (poly)phenols and supporting future applications in more complex dietary interventions and, eventually, in human tissues.