Metabolomic Comparison of Guava (Psidium guajava L.) Leaf Extracts Fermented by Limosilactobacillus fermentum and Lactiplantibacillus plantarum and Their Antioxidant and Antiglycation Activities

Abstract

Probiotic fermentation of plant-based materials can lead to the generation of various bioactive substances via bacterial metabolites and the biotransformation of phenolic compounds. We compared the metabolic differences between fermentation by Limosilactobacillus fermentum KCTC15072BP (LFG) and fermentation by Lactiplantibacillus plantarum KGMB00831 (LPG) in guava leaf extract (0%, 0.5%, and 2% (w/v))-supplemented medium via non-targeted metabolite profiling. By performing multivariate statistical analysis and comparing the different guava leaf extract groups, 21 guava-derived and 30 bacterial metabolites were identified. The contents of guava-derived glucogallin, gallic acid, and sugar alcohols were significantly higher in LFG than they were in LPG. Similarly, significantly higher contents of guava-derived pyrogallol, vanillic acid, naringenin, phloretin, and aromatic amino acid catabolites were obtained with LPG than with LFG. LFG led to significantly higher antioxidant activities than LPG, while LPG led to significantly higher antiglycation activity than LFG. Interestingly, the fermentation-induced increase in the guava-leaf-extract-supplemented group was significantly higher than that in the control group. Thus, the increased bioactivity induced by guava fermentation with the Lactobacillaceae strain may be influenced by the synergistic effects between microbial metabolites and plant-derived compounds. Overall, examining the metabolic changes in plant-based food fermentation by differentiating the origin of metabolites provides a better understanding of food fermentation.

Keywords: antiglycation activity; antioxidant activity; metabolomics; probiotic fermentation.

Figure 1

PCA and PLS-DA score plot of NFG, LFG, and LPG at (A,B) three (0%, 0.5%, and 2%) concentrations of the guava leaf extract and (C,D) 2% guava leaf extract. Multivariate analysis was conducted using datasets derived from (A,C) GC-TOF-MS and (B,D) UHPLC–Orbitrap–MS (NFG, non-fermentation (gray); LFG, guava leaves fermented by L. fermentum (green); LPG, guava leaves fermented by L. plantarum (orange); ●, 0% guava leaf extract group; ▲, 0.5% guava leaf extract group; ■, 2% guava leaf extract group).

Figure 2

Heat map depicting the relative abundance of the (A) bacterial metabolites and (B) guava-derived metabolites of NFG, LFG, and LPG in the 0%, 0.5%, and 2% groups. Bacterial metabolites were selected based on detection in both non-supplemented medium and guava-supplemented medium. Guava-derived metabolites were only detected and selected from the group supplemented with the guava leaf extract (NFG, non-fermented; LFG, fermented by L. fermentum; LPG, fermented by L. plantarum).

Figure 3

Schematic of the metabolic pathway and relative levels of NFG, LFG, and LPG metabolites detected in the different guava leaf extract groups (0%, 0.5%, and 2%). The metabolite pathways were adopted from the KEGG database and modified. The Y-axis of the graph represents peak areas of respective metabolites. Different letters indicate significant difference based on Duncan’s multiple-range test. Metabolites in green and orange font have relatively higher abundance when fermented by L. fermentum and L. plantarum, respectively (0%: 0% guava leaf group; 0.5%: 0.5% guava leaf group; 2%: 2% guava leaf group; : non-fermented (NFG); : fermented by L. fermentum (LFG); : fermented by L. plantarum (LPG)).

Figure 4

Proposed bioconversion pathways of (A) digalloyl glucose and catechin gallate, (B) chlorogenic acid, and (C) prunin and phlorizin. The selected compounds among the guava-derived metabolites are compared the their relative contents in NFG, LFG, and LPG in the groups supplemented with 0%, 0.5%, and 2% of the guava leaf extract. The Y-axis of the graph represents peak areas of respective metabolites. Black text indicates a decrease in metabolites by fermentation with L. fermentum and L. plantarum. Green text indicates an increase in metabolite with LFG alone. Orange text indicates an increase in metabolites with LPG alone. Different letters indicate significant differences based on Duncan’s multiple-range test. (0%: 0% guava leaf group; 0.5%: 0.5% guava leaf group; 2%: 2% guava leaf group; : non-fermented (NFG); : fermented by L. fermentum (LFG); : fermented by L. plantarum (LPG)).

Figure 5

Bioactivities of NFG, LFG, and LPG in the groups supplemented with the guava leaf extract (0%, 0.5%, and 2%). Antioxidant activities based on the (A) DPPH assay and (B) Ferric reducing antioxidant power (FRAP) assay, and (C) antiglycation activity (MGO-AGEs formation inhibition assay). Values represent the averages of triplicate measurements (n = 3). The Y-axis of the (A) and (B) represents Trolox Equivalent Antioxidant Capacity, and in (C), it represents the degree of inhibition of MGO-AGEs formation. Each letter indicates significant differences according to Duncan’s multiple-range test (p < 0.05) (0% group: 0% guava leaf group; 0.5% group: 0.5% guava leaf group; 2% group: 2% guava leaf group; : non-fermented (NFG); : fermented by L. fermentum (LFG); : fermented by L. plantarum (LPG)).

Figure 6

Correlation networks between the metabolites and bioactivity assays (DPPH, FRAP, and MGO-AGEs formation assay) in the 2% guava leaf group. The metabolites were selected based on a Pearson’s correlation coefficient value higher than 0.5. The box and circle symbols indicate bioactivities and the metabolites, respectively (gray color, antioxidant activity using DPPH and FRAP; black color, antiglycation activity, MGO-AGEs formation inhibition assay; , LFG-specific compounds; , LPG-specific compounds; , non-specific compound).