Serum metabolomic responses to aerobic exercise in rats subjected to mild chronic unpredictable stress

Behavioral tests

After 21 days of CUMS exposure, the body weight, sucrose preference rate, and number of crosses and ranks of CUMS-treated rats were significantly decreased compared to the normal control group (C), whereas no significant difference was found between the groups before CUMS modeling, indicating that the rats on CUMS had developed depression-like symptoms.

At the end of the 28-day aerobic exercise intervention, the preference rate for sucrose and the number of crossovers and ranks were significantly higher in the DE group compared to the DC group, indicating the beneficial effects of the aerobic exercise on depression-like behavior (Fig. 1). It should be noted that after 28 days of aerobic exercise, the body weight of rats in the DE group was still significantly lower than that in the NC group, suggesting that aerobic exercise did not attenuate the weight loss caused. by CUMS modeling, which could be related to increased energy consumption during exercise. Additionally, the NE group showed a trend of lower body weight and higher sucrose preference rate and number of crosses and ranks compared to the NC group, although the differences were not statistically significant. This suggests that exercise in normal rats did not cause additional stress.

Figure 1

Effects of CUMS and aerobic exercise on behavioral indicators. (a) weight; (b) rate of preference for sucrose; (vs) number of free field test passes; and (D) number of standing rats in an open field test during the 49-day experiment. Data are presented as mean ± SD (n=8 in each group). *p p p p CUMS unpredictable chronic mild stress.

Multivariate statistical analysis of LC–MS/MS data

Data processing in metabolomics typically involves clustering, classification and discrimination of collected data, primary means including unsupervised and supervised methods, and then studying the patterns and mechanisms of change in metabolites after disruption of the organization.18.

To visually distinguish the serum metabolic profiles of each group, the present study first used an unsupervised PCA model to perform a preliminary analysis of all sample data to eliminate outliers. Then, supervised PLS-DA model analysis was performed on each sample data set to obtain PLS-DA score scatter plots and model validation plots (Fig. 2a,b). As shown in the plot of the PLS-DA score, the metabolic profiles of the three groups showed some trend towards clustering, with the DE group clearly separated from the DC group, indicating that the aerobic exercise intervention for depression resulted in significant changes in metabolic profiles. depressed rats. Additionally, PLS-DA QC samples were pooled, indicating that the system and method were stable.

Figure 2
Figure 2

Analysis of multivariate LC–MS/MS data. (a) PLS-DA score plots of NC group, DC group, DE group and QC; (b) PLS-DA model validation scheme; (vs) Graphs of OPLS-DA scores of NC group and DC group; (D) S-plot of NC group and DC group; (and) Graphs of OPLS-DA scores of DC group and DE group; (F) S-plot of DC group and DE group. QCQuality control, PLS-DAdiscriminant analysis of partial least squares, OPLS-DAPartial least squares orthogonal discriminant analysis.

To check the reliability of the model, a permutation test (n=200 times) was performed on the above model data, and the results showed that the parameters R2 and Q2 were close, the slope of the model was large and the intercept with the vertical axis was negative, indicating that the model was robust and had no overfitting phenomenon.

To find differential metabolites associated with depression and exercise to improve depressive symptoms, an OPLS-DA analysis was first performed on data from the NC and DC groups to remove experimentally irrelevant information and to detect the differential variables associated with depressive illness. The OPLS-DA score graph showed a significant separation between the DC and NC groups (Fig. 2c), indicating that depression affected the serum metabolic profile. The variables with the largest contribution to the variance were then examined by combining VIP > 1 in the S-Plot plot (Fig. 2d) and the significant difference (p

Analysis of differential metabolites in serum

A total of 21 differential metabolites linked to depression were identified by OPLS-DA analysis of serum sample data, of which 15 differential metabolites were modulated after exercise (Table 1). The relative peak area levels of these metabolites are shown in the histogram (Fig. 3a). To visualize the overall trend of all differential metabolites in each sample, a hotspot plot of their relative levels was presented (Fig. 3b).

Table 1 Differential metabolites in serum of CUMS rats with or without aerobic exercise.
picture 3
picture 3

(a) Comparison of relative peak areas of differential metabolites. All data were expressed as mean ± SD (n = 8). *ppppb) The relative heat map content of differential metabolites in rat serum samples under CUMS. CUMSunpredictable chronic mild stress.

After 49 days of CUMS modeling, the levels of 21 different metabolites were changed differently in the NC group and the DC group. Of these, serum levels of 18 different metabolites, including cytosine, L-proline, indole, valine, taurine, L-pipecolic acid, skatole, L-isoleucine, L- norleucine, pyruvate, 6-methylnicotinamide, spermidine, L-lysine, L-methionine, phenylalanine, glucose, citric acid and tryptophan, were significantly lower and levels of 3 different metabolites, choline, hippuric acid and indoleacrylic acid were significantly increased. Aerobic exercise modulated 15 of these metabolites, including cytosine, L-proline, indole, valine, taurine, L-pipecolic acid, skatole, L-isoleucine, L-norleucine, L-lysine, phenylalanine, glucose, indoleacrylic acid, citric acid, and pyruvate, as indicated by that of DE versus DC (Table 1).

Serum differential metabolite pathway analysis

A total of 21 differential metabolites were found to be associated with depression using the LC-MS/MS method and 15 differential metabolites were affected by aerobic exercise. The 15 differential metabolites were imported into the MetaboAnalyst 5.0 database for metabolic pathway analysis to obtain the pathway impact distribution map and pathway enrichment map (Fig. 4). A total of nine metabolic pathways significantly associated with depression were examined with an impact value >0.1 (Fig. 4a), mainly including phenylalanine, tyrosine and tryptophan biosynthesis, taurine metabolism and hypotaurine, phenylalanine metabolism, pyruvate metabolism, tryptophan metabolism, citrate cycle (TCA cycle), arginine and proline metabolism, glycolysis/gluconeogenesis, cysteine ​​metabolism and methionine. Aerobic exercise affected six of these metabolic pathways (Fig. 4b), mainly including phenylalanine, tyrosine and tryptophan biosynthesis, taurine and hypotaurine metabolism, phenylalanine metabolism, pyruvate metabolism, the citrate cycle (TCA cycle), glycolysis/gluconeogenesis. Since animals lack complete pathways for the synthesis of tyrosine, phenylalanine and tryptophan, which are synthesized only by bacteria, other microorganisms and plants19, we have excluded the metabolic pathways of phenylalanine, tyrosine, and tryptophan biosynthesis from further analysis and discussion in this report. Therefore, we have identified 8 metabolic pathways associated with depression, and aerobic exercise may exert antidepressant effects by modulating 5 of them.

Figure 4
number 4

Summary scheme of pathway analysis with Met-PA. (a) Metabolic pathways associated with depression; (b) Regulation of the aerobic exercise metabolic pathway. A: biosynthesis of phenylalanine, tyrosine and tryptophan (×); B: metabolism of taurine and hypotaurine; C: phenylalanine metabolism; D: pyruvate metabolism; E: tryptophan metabolism; F: citrate cycle (TCA cycle); G: arginine and proline metabolism; H: glycolysis/gluconeogenesis; I: metabolism of cysteine ​​and methionine. Each circle represents a metabolic pathway; circle size and color shades are positively correlated with impact on the metabolic pathway.

Correlation analysis of behavioral indicators of depression and serum differential metabolite levels

In the correlation heatmap, the size and color of the circles represent the degree of correlation between the indicators, and blue is positive and red is negative. Correlation coefficients range from 1.0 (maximum positive correlation) to −1.0 (maximum negative correlation), with 0 indicating no correlation. According to the threshold |r|> 0.5 combined with p

Figure 5
number 5

Heatmap of the correlation between differential metabolites and behavioral indicators based on Spearman’s correlation analysis.

Comments are closed.