Acute effects of moderate vs. vigorous endurance exercise on urinary metabolites in healthy, young, physically active men-A multi-platform metabolomics approach

Abstract

Introduction: Endurance exercise alters whole-body as well as skeletal muscle metabolism and physiology, leading to improvements in performance and health. However, biological mechanisms underlying the body's adaptations to different endurance exercise protocols are not entirely understood. Methods: We applied a multi-platform metabolomics approach to identify urinary metabolites and associated metabolic pathways that distinguish the acute metabolic response to two endurance exercise interventions at distinct intensities. In our randomized crossover study, 16 healthy, young, and physically active men performed 30 min of continuous moderate exercise (CME) and continuous vigorous exercise (CVE). Urine was collected during three post-exercise sampling phases (U01/U02/U03: until 45/105/195 min post-exercise), providing detailed temporal information on the response of the urinary metabolome to CME and CVE. Also, fasting spot urine samples were collected pre-exercise (U00) and on the following day (U04). While untargeted two-dimensional gas chromatography-mass spectrometry (GC×GC-MS) led to the detection of 608 spectral features, 44 metabolites were identified and quantified by targeted nuclear magnetic resonance (NMR) spectroscopy or liquid chromatography-mass spectrometry (LC-MS). Results: 104 urinary metabolites showed at least one significant difference for selected comparisons of sampling time points within or between exercise trials as well as a relevant median fold change >1.5 or <0. $\overline{6}$ (NMR, LC-MS) or >2.0 or <0.5 (GC×GC-MS), being classified as either exercise-responsive or intensity-dependent. Our findings indicate that CVE induced more profound alterations in the urinary metabolome than CME, especially at U01, returning to baseline within 24 h after U00. Most differences between exercise trials are likely to reflect higher energy requirements during CVE, as demonstrated by greater shifts in metabolites related to glycolysis (e.g., lactate, pyruvate), tricarboxylic acid cycle (e.g., cis-aconitate, malate), purine nucleotide breakdown (e.g., hypoxanthine), and amino acid mobilization (e.g., alanine) or degradation (e.g., 4-hydroxyphenylacetate). Discussion: To conclude, this study provided first evidence of specific urinary metabolites as potential metabolic markers of endurance exercise intensity. Future studies are needed to validate our results and to examine whether acute metabolite changes in urine might also be partly reflective of mechanisms underlying the health- or performance-enhancing effects of endurance exercise, particularly if performed at high intensities.

Keywords: GC×GC-MS; LC-MS; NMR; continuous physical exercise; exercise intensity; exercise metabolomics; metabolic markers; urine metabolome.

FIGURE 1

Study design, urine sample collections, and metabolomics approaches. (A) Participants were randomized to 30 min of CME or CVE at V1 or V2, respectively. (B) Fasting spot urine samples were obtained in the morning of V1 and V2 (U00) and after 24 h (U04). Three urine collection phases covered the (post-)exercise periods (U01: urine produced during exercise trials until 45 min post-exercise; U02: urine produced from 45 to 105 min post-exercise; U03: urine produced from 105 to 195 min post-exercise. (C) 44 metabolites were identified and quantified by targeted NMR or targeted LC-MS. 608 spectral features were detected by untargeted GC×GC-MS.

FIGURE 2

Workflow for metabolomics and statistical analyses.

FIGURE 3

Heatmap of z-transformed median concentrations or signal intensities per urine sample time point in the CME and CVE trial for exercise-responsive and/or intensity-dependent metabolites, sorted by most relevant major metabolic pathways/classes and decreasing median FCs for U01 (CVE)/U00 (CVE) ratio. Metabolites with relevant differences between interventions are indicated with a red asterisk. ^a: GC×GC-MS-detected; ^b: LC-MS-detected; ^c: NMR-detected; *: evaluated based on heuristic/visual approach; U: unknown analyte.

FIGURE 4

Heatmap of z-transformed median concentrations or signal intensities per urine sample time point in the CME and CVE trial for exercise-responsive and/or intensity-dependent metabolites, sorted by most relevant major metabolic pathways/classes and decreasing median FCs for U01 (CVE)/U00 (CVE) ratio (continued). Metabolites with relevant differences between interventions are indicated with a red asterisk. ^a: GC×GC-MS-detected; ^b: LC-MS-detected; ^c: NMR-detected; *: evaluated based on heuristic/visual approach; U: unknown analyte; X: missing values.

FIGURE 5

Venn diagram summarizing the number of exercise-responsive as well as intensity-dependent urinary metabolites. ^a: GC×GC-MS-detected; ^b: LC-MS-detected; ^c: NMR-detected; *: relevant metabolites based on heuristic/visual approach; U: unknown analyte.