Synergistic effects of exercise and catalase overexpression on gut microbiome

Abstract

Exercise influences metabolic parameters in part by modulating redox stress and as recently suggested, by affecting the gut microbiome. However, whether excess endogenous antioxidant potentiates or interferes with the beneficial effects of exercise on the gut microbiome is not known. A comparison of the gut microbiome of C57Bl6 (C57/WT) mice to the 'stress-less' catalase overexpressing mice models ([Tg(CAT)^± ] and Bob-Cat), that were either exercised or remained sedentary, showed differences in both alpha and beta diversity. The significant variation was explained by genotypes along with exercise, suggesting a synergistic relationship between exercise and genotypic traits. Linear discriminant analysis effect size (LEfSe) analysis also revealed differential taxa within the exercised/genotype cohorts in contrast to those within sedentary/genotype cohorts. Functional pathway predictions from PICRUSt2 showed enrichment for the metabolism of short-chain fatty acids, butanoate and propanoate pathways in exercised groups. Spearman correlations between enriched taxa and metabolic parameters showed correlations with body or fat weight in some of the cohorts. However, there were significant correlations of differential taxa among all cohorts against parameters that predict energy metabolism, such as respiratory exchange ratio and energy expenditure. Overall, our study showed that there was a synergistic beneficial influence of antioxidant overexpression and exercise on the gut microbiome.

Figure 1.. Boxplot of microbiome alpha diversity.

A boxplot of the alpha diversity measured by the Observed Features metric as implemented in QIIME1 for the exercise and sedentary groups of all samples (n≥27). A two-sample t-test and non-parametric Monte Carlo permutations (n = 999) were used for statistical analysis (p≤0.05).

Figure 2.. β-Diversity plots and Firmicutes to Bacteroidetes ratio of sedentary and exercised mice.

PCoA plots of phylogenetic differences in microbial community composition between sequenced sedentary and exercised samples. (A) C57/WT (blue), [Tg(CAT)^±] (orange), Bob-Cat (red), and Ob/Ob (green) sedentary mice groups and (B) of each respective exercised cohort, is displayed in the images showing a defined microbial community composition within each genotype (ANOSIM, p≤ 0.05). (n≥6). (C) Changes in Firmicutes to Bacteroidetes ratio observed in all genotypes.

Figure 3.. β-Diversity plots per genotype.

PCoA plots of phylogenetic differences in microbial community composition between sequenced sedentary and exercised samples per mouse genotype. (A) C57/WT, (B) [Tg(CAT)^±], (C) Bob-Cat, and (D) Ob/Ob mice groups (sedentary depicted as lighter shade and exercise depicted as darker shade of colored dots) as displayed in the images show a defined microbial community composition within each E and S group of each genotype (ANOSIM p< 0.05), (n≥6).

Figure 4.. LEfSe analysis of taxa within all genotypes between exercised and sedentary cohorts.

Linear discriminant analysis (LDA) scores derived from LEfSe analysis showing the biomarker taxa (log LDA score) ≥ 2.0 and significant at p≤0.05) of (A) C57/WT, (B) [Tg(CAT)^±], (C) Bob-Cat, and (D) Ob/Ob mice groups (green- sedentary; red-exercise) determined by Kruskal–Wallis), (n≥7).

Figure 5.. Significant correlations (Spearman, p≤0.05) among metabolic metadata and enriched taxa.

Both taxa enriched in the Exercised and Sedentary cohort were included in the analysis for their respective genotypes. Positive correlations are shown in blue, and negative correlations are shown in red, with the darkness of the squares indicating the strength of the correlations. Correlations were calculated through the hmisc package and visualized using the ggcorrplot package in R. (A) C57/WT Sedentary. (B) C57/WT Exercised. (C) Ob/OB Sedentary. (D) Ob/Ob Exercised mice groups.

Figure 6.. Significant correlations (Spearman, p≤0.05) among metabolic metadata and enriched taxa for the [Tg(CAT)±] and Bob-Cat cohorts.

Both taxa enriched in the Exercised and Sedentary cohort were included in the analysis for their respective genotypes. Positive correlations are shown in blue, and negative correlations are shown in red, with the darkness of the squares indicating the strength of the correlations. Correlations were calculated through the hmisc package and visualized using the ggcorrplot package in R. (A) [Tg(CAT)±] Sedentary. (B) [Tg(CAT)±] Exercised. (C) Bob-Cat Sedentary. (D) Bob-Cat Exercised mice groups.

Figure 7.. Predicted function LEfSe analysis.

Linear discriminant analysis (LDA) scores derived from LEfSe analysis showing differential PICRUSt2 predicted pathways (log LDA score ≥2.0) and significant at p≤0.05) of (A) C57/WT, (B) [Tg(CAT)^±], (C) Bob-Cat, and (D) Ob/Ob mice groups (green-sedentary; red-exercise) determined by Kruskal—Wallis. (n≥7)

Figure 8:. Synergistic effects of exercise and catalase overexpression on microbial function:

Enrichment of beneficial bacteria and predicted microbial functional enrichment in catalase overexpressing mice that were exercised compared to sedentary.