The effect of acute moderate-intensity exercise on the serum and fecal metabolomes and the gut microbiota of cross-country endurance athletes

Abstract

Physical exercise can produce changes in the microbiota, conferring health benefits through mechanisms that are not fully understood. We sought to determine the changes driven by exercise on the gut microbiota and on the serum and fecal metabolome using 16S rRNA gene analysis and untargeted metabolomics. A total of 85 serum and 12 fecal metabolites and six bacterial taxa (Romboutsia, Escherichia coli TOP498, Ruminococcaceae UCG-005, Blautia, Ruminiclostridium 9 and Clostridium phoceensis) were modified following a controlled acute exercise session. Among the bacterial taxa, Ruminiclostridium 9 was the most influenced by fecal and serum metabolites, as revealed by linear multivariate regression analysis. Exercise significantly increased the fecal ammonia content. Functional analysis revealed that alanine, aspartate and glutamate metabolism and the arginine and aminoacyl-tRNA biosynthesis pathways were the most relevant modified pathways in serum, whereas the phenylalanine, tyrosine and tryptophan biosynthesis pathway was the most relevant pathway modified in feces. Correlation analysis between fecal and serum metabolites suggested an exchange of metabolites between both compartments. Thus, the performance of a single exercise bout in cross-country non-professional athletes produces significant changes in the microbiota and in the serum and fecal metabolome, which may have health implications.

Figure 1

Partial least square discriminant analysis (PLS-DA) of the metabolites in T1 (before exercise) and T2 (after exercise). The results are presented as principal component score plots, with each point in the plot representing an individual sample. (A,B) PLS-DA score plots obtained from LC-HRMS data in positive mode C18 and negative mode HILIC. (C,D) Statistical validation of the PLS-DA model (A,B), showing R2Y (pink dots) and the Q2 (light-blue dots) values from the permutated analysis (bottom left) lower than the corresponding original values (top right).

Figure 2

Hierarchical clustering heatmaps representing annotated serum metabolites significantly (p < 0.05) different between T1 and T2. (A) Metabolites obtained from HILIC(−) LC-HRMS analysis. (B) Metabolites obtained from C18(+) LC-HRMS analysis.

Figure 3

Variable important in projection (VIP) score from PLS-DA of serum metabolites. (A) VIP score from serum metabolites obtained in positive ionization mode. (B) VIP score of serum metabolites obtained in negative ionization mode.

Figure 4

Summary of metabolic pathway enrichment analysis performed in MetaboAnalyst 4.0 using a panel of 85 metabolites found to be significantly altered by a session of acute exercise. All the matched pathways are displayed as circles. The node color is based on its p-value and the node radius is determined based on its pathway impact values. (A) Serum pathways modified by a session of acute exercise. (B) Fecal pathways modified by a session of acute exercise.

Figure 5

Heatmaps showing significant statistical correlations values (r) between: (A) serum and fecal metabolites, and (B) microbial taxa and fecal metabolites. Significant correlations are marked with a cross (p-adjusted value < 0.05).