Optimization of Liquid Fermentation of Acanthopanax senticosus Leaves and Its Non-Targeted Metabolomics Analysis

Abstract

To enhance the nutritional value of Acanthopanax senticosus leaves (AL), a fermentation process was conducted using a probiotic Bacillus mixture, and the changes in chemical constituents and biological activities before and after fermentation were compared. A response surface methodology was employed to optimize the liquid fermentation conditions of AL based on their influence on polyphenol content. Non-targeted metabolomics analysis was performed using LC-MS/MS to reveal the differing profiles of compounds before and after fermentation. The results indicated that Bacillus subtilis LK and Bacillus amyloliquefaciens M2 significantly influenced polyphenol content during fermentation. The optimal fermentation conditions were determined to be a fermentation time of 54 h, a temperature of 39.6 °C, and an inoculum size of 2.5% (v/v). In comparison to unfermented AL, the total polyphenol and flavonoid contents, as well as the free radical scavenging capacities measured by DPPH and ABTS assays, and the activities of β-glucosidase and endo-glucanase, were significantly increased. The non-targeted metabolomics analysis identified 1348 metabolites, of which 829 were classified as differential metabolites. A correlation analysis between the differential metabolites of polyphenols, flavonoids, and antioxidant activity revealed that 13 differential metabolites were positively correlated with antioxidant activity. Kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis of the differential metabolites identified 82 pathways, with two of the top 25 metabolic pathways related to flavonoids. This study explores the potential for enhancing the active ingredients and biological effects of AL through probiotic fermentation using Bacillus strains.

Keywords: Acanthopanax senticosus leaves; complex probiotic; fermentation; non-targeted metabolomics.

Figure 1

(A) Effect of single-strain fermentation on TPC. (B) Effect of fermentation with different strain ratios on TPC. (C) Effect of inoculum size on TPC. (D) Effect of fermentation temperature on TPC. (E) Effect of rotational speed on TPC. (F) Effect of fermentation time on TPC. Values are expressed as mean ± standard deviation (n = 3). Note: a, b, c, d, and e represent the significance of the difference within the graph. Containing the same letter means the difference is insignificant (p > 0.05).

Figure 2

Response surface plots. (a,b) show the interaction between fermentation time and fermentation temperature; (c,d) show the interaction between fermentation time and inoculation amount; (e,f) show the interaction between fermentation temperature and inoculation amount. Note: red dot represents above surface; pink dot represents below surface.

Figure 3

Effect of fermentation time on total polyphenol content (A) and total flavonoid content (B). Values are expressed as mean ± standard deviation (n = 3). Note: a, b, c, d, e, and f represent the significance of the difference within the graph. Containing the same letter means the difference is insignificant (p > 0.05). ***: p < 0.001.

Figure 4

Effect of fermentation time on β-glucosidase activity and endo-glucanase activity. Values are expressed as mean ± standard deviation (n = 3).

Figure 5

SEM of AL before and after fermentation. (a) unfermented; (b) fermented using Bacillus subtilis LK and Bacillus amyloliquefaciens M2.

Figure 6

Effect of fermentation on the antioxidant activity of AL. (a) DPPH free radical scavenging ratio, (b) ABTS free radical scavenging ratio, (c) Fe²⁺ chelating activity. Values are expressed as mean ± standard deviation (n = 6). **: p < 0.01.

Figure 7

Schematic representation of the different classifications of FAL metabolites. (a) Metabolites in positive ion mode; (b) metabolites in negative ion mode.

Figure 8

PCA and hierarchical clustering heatmap showing changes in metabolites during AL fermentation. (a) PCA score plot in positive ion mode; (b) PCA score plot in negative ion mode; (c) hierarchical clustering heat map in positive ion mode; (d) hierarchical clustering heat map in negative ion mode.

Figure 8

PCA and hierarchical clustering heatmap showing changes in metabolites during AL fermentation. (a) PCA score plot in positive ion mode; (b) PCA score plot in negative ion mode; (c) hierarchical clustering heat map in positive ion mode; (d) hierarchical clustering heat map in negative ion mode.

Figure 9

OPLS-DA score plots and volcano plots showing changes in metabolites during AL fermentation. (a) OPLS-DA scores plot in positive ion mode; (b) OPLS-DA scores plot in negative ion mode; (c) volcano plot in positive ion mode; (d) volcano plot in negative ion mode.

Figure 10

Correlation plot between differential metabolites in polyphenols and flavonoids, and antioxidant activity (DPPH, ABTS, and Fe²⁺ chelating activity) and enzyme activity (endo-glucanase, β-glucosidase), with the colour of the circles indicating the correlation coefficients and the magnitude of the values within the circles indicating the statistical differences, with the red circles indicating a positive correlation, and the blue circles indicating a negative correlation.

Figure 11

(a–e) Histograms of possible structures and relative peak intensities of five methoxyphenols; (f,g) histograms of possible structures and relative peak intensities of two flavonoid glycosides.

Figure 12

KEGG enrichment pathway based on differential metabolites between AL and FAL. (a) Pathway enrichment analysis. Each bubble in the bubble diagram represents a metabolic pathway, and bubble size is proportional to enrichment. The top 25 items with the highest p-values were selected; (b) network diagram of associations and metabolite heatmap. Orange nodes indicate metabolic pathways and green nodes indicate metabolites.