Several Bifidobacterium adolescentis strains were isolated from the gastrointestinal tracts (GIT) of healthy individuals residing in Japan. These isolates are part of the microbial strain collection maintained at the Laboratory of Genome Microbiology, Gifu University, Japan. Additional commercial strains were obtained from the Japan Collection of Microorganisms (JCM).

All strains were cultivated in de Man, Rogosa, and Sharpe (MRS) broth (BD, Cat# MD21152, USA) at 37 °C under anaerobic conditions using a BUG Box anaerobic chamber (Ruskinn Technology Ltd., Bridgend, Wales, UK) and a gas mixture composed of 80% N₂, 10% CO₂, and 10% H₂. To support GABA production, 1% (v/v) monosodium glutamate (MSG; Sigma-Aldrich, Cat# G1626, St. Louis, MO, USA) was added to the MRS broth. For standard cultivation, 30–50 µL of frozen stock (stored at –80 °C) was inoculated into 12 mL of MRS broth and incubated at 37 °C for 24 hours. The culture was subsequently sub-cultured in fresh MRS broth containing 1% MSG and incubated for an additional 48 hours. Cells were then harvested by centrifugation at 6000 $\times$g for 10 minutes, and the supernatant was collected for high-performance liquid chromatography (HPLC) analysis.

Fecal samples were obtained from healthy adult volunteers with no history of systemic or psychiatric disorders and no antibiotic usage for at least three months prior to sample collection. Donors (n = 8) were healthy adults aged 21–36 years, including both males and females, with no known dietary restrictions. All reported following a mixed diet typical of the local population. Samples were collected using BD BBL CultureSwab Plus (BD Co., New Jersey, US, Cat# 212550), immediately stored at 4 °C, and processed within 12 hours of collection. Written informed consent was obtained from all participants. The study protocol was approved by the institutional ethics review board of Gifu University (Certificate No. 2019-283).

Batch fermentation was conducted using a pH-controlled, multi-channel fermentation system (Kobe University Human Intestinal Microbiota Model, KUHIMM; Kobe, Hyogo, Japan) [22]. The simulator consisted of six parallel and independent vessels. The working volume per vessel was 100 mL of Gifu Anaerobic Medium (GAM broth; Nissui Pharmaceutical Co., Ltd., Code: 05422, Tokyo, Japan) with an initial pH adjusted to 6.5 using manual acid/base titration. KUHIMM is equipped with a pH sensor that continuously monitors and records pH values throughout the entire cultivation period. No fixed pH set point was applied, as the system was designed to record the natural pH changes resulting from microbial fermentation. Anaerobic conditions were maintained by continuous flushing with an N₂/CO₂ (80:20) gas mixture at a rate of 10 mL min^-1 through a 0.2 µm polytetrafluoroethylene membrane (Pall Corporation, Port Washington, NY, USA) for 30 minutes at 37 °C prior to fermentation. The pH was continuously monitored for each vessel, and continuous stirring was maintained at 300 rpm to ensure homogeneous microbial distribution.

For inoculation, fecal samples were suspended in 2 mL of physiological saline and introduced through the side port of each vessel. Based on the experimental design, three additives were prepared: mannooligosaccharides (MOS; van wankum ingredients BV, Maarssen, Netherlands; powder form), fructooligosaccharides (FOS; SigmaAldrich, Cat. F8052, St. Louis, MO, USA) and dextrin (Dex; Shandong Bailong Chuangyuan BioTech Co., Ltd., Dezhou, Shandong, China), each dissolved in 10 mL of sterilized water at a final concentration of 0.5%. These additives were then added to the test vessels. Additionally, sterile water was added to the control vessels. Fully grown B. adolescentis 4-2 was added to designated vessels at 0.1% (v/v) of the total volume.

Cell-free supernatants were obtained by membrane filtration (0.45 µm) from both bacterial and fecal cultures. GABA concentrations were quantified using HPLC (1100 Series, Agilent Technologies, Santa Clara, CA, USA) equipped with a fluorescence detector (excitation at 350 nm and emission at 450 nm) and a COSMOSIL packed column (5C18-MS-II, 3.0 mm ID $\times$ 150 mm; Nacalai Tesque, Kyoto, Japan, cat#34245-31). Prior to analysis, each sample was derivatized with O-phthalaldehyde (OPA; Sigma-Aldrich, St. Louis, MO, USA, cat#P0657) reagent [20, 23]. The mobile phase consisted of A (CH₃CN/CH₃OH/H₂O, 45/40/15, v/v/v) and B (20 mM KH₂PO₄, pH 6.9, H₃PO₄). Compounds were eluted using the following gradient program: 0–9 min, 100% B; 9–12 min, 89% B; and 12–21 min, 78% B. The column temperature was maintained at 35 °C with a flow rate of 0.7 mL/min. GABA concentrations in unknown samples were calculated by comparing peak area and retention time to those of standard curves prepared using known concentrations.

DNA extraction and next-generation sequencing (NGS) were performed as previously described [20]. Raw NGS data were analyzed using QIIME 2 (version 2020.2; QIIME 2 Development Team) [24]. Additionally, bar plots displaying microbial composition were generated at both the phylum and species levels based on identified operational taxonomic units (OTUs).

Apart from the KUHIMM culture, liquid MRS culture (1 mL) was disrupted by sonication. Sonication was performed using a Fisherbrand™ Sonicator with probe (Fisher Scientific, Waltham, MA, USA) for cell disruption prior to enzyme assay. The disrupted cells were then centrifuged at 13,000 rpm for 15 min at 4 °C, and the supernatant (cell-free extract) was used for the $\beta$-mannosidase assay. For the assay, the reaction mixture consisted of 0.4 M sodium acetate buffer (pH 6.0, containing 4 mM CaCl₂; Sigma-Aldrich, Cat# C4901, St. Louis, MO, USA), 10 mM p-nitrophenyl-$\beta$-D-mannopyranoside (PNP$\beta$M; Sigma-Aldrich, Cat# N2127, St. Louis, MO, USA) and distilled water in a 3:1:6 (v/v/v) ratio. A total of 280 µL of this mixture was combined with 46.7 µL of the cell-free extract and incubated at 37 °C for 1 hour. The reaction was terminated by adding 490 µL of 0.2 M Na₂CO₃ (Sigma-Aldrich, Cat# S7899, St. Louis, MO, USA). A 200 µL aliquot of the mixture was then used for spectrophotometric analysis at 450 nm.

Enzymatic activity was quantified based on the concentration of p-nitrophenol released, using a standard curve prepared from known concentrations of p-nitrophenol (Sigma-Aldrich, Cat# N7660, St. Louis, MO, USA), under identical assay conditions.

All statistical analyses were performed using GraphPad Prism (version 9; GraphPad Software, San Diego, CA, USA). Data were presented as mean $\pm$ standard deviation (SD) unless otherwise specified. Comparisons between two groups were assessed using unpaired two-tailed Student’s t-tests. For multiple group comparisons, one-way analysis of variance (ANOVA) was followed by Tukey’s post-hoc test to identify significant differences among groups. Moreover, a p-value of less than 0.05 was considered statistically significant. All experiments were conducted with biological replicates obtained from five to ten individual volunteers, as specified. Additionally, microbial community analysis was performed using QIIME2 (QIIME 2 Development Team, Flagstaff, AZ, USA).

To identify strains with strong GABA-producing capabilities, a collection of Bifidobacterium isolates, comprising both fecal-derived and commercially available strains (Table 1), was screened. Consistent with previous findings [25], GABA production was found to be markedly strain-specific rather than species-dependent. Among the B. adolescentis isolates, strain 4-2 demonstrated the highest GABA yield, achieving 1.4 $\pm$ 0.15 g/L.

To further assess its functional potential, B. adolescentis 4-2 was introduced into the KUHIMM, an in vitro system inoculated with human fecal microbiota from donors exhibiting low baseline GABA levels. Monosodium glutamate (MSG) was provided as the precursor substrate. As depicted in Fig. 1A,B, the presence of B. adolescentis 4-2 resulted in a significant increase in fecal GABA concentration compared with both the control and MSG-only conditions. Concurrently, Fig. 1B illustrates a marked reduction in glutamate levels in the B. adolescentis 4-2 + MSG group, indicating efficient microbial conversion of glutamate to GABA. These results underscore the strain’s capacity to modulate key gut neurotransmitter levels by elevating GABA while reducing glutamate, thereby supporting its application in microbiome-based interventions aimed at the gut-brain axis.

Fig. 1.

Gamma-aminobutyric acid (GABA) and glutamate concentrations across different fecal culture conditions. GABA (A) and glutamate (B) concentrations were measured in fecal culture samples obtained from five individual volunteers under three conditions: Control (green), MSG supplementation (orange), and Bifidobacterium adolescentis 4-2 with MSG (blue). Boxplots depict the data distribution across volunteers, where the horizontal line represents the median, the box represents the interquartile range (IQR), and whiskers extend to 1.5 $\times$ IQR. Diamonds ($\mathrm{◆}$) represent individual statistical outliers falling outside this range. Co-culturing with B. adolescentis 4-2 and MSG resulted in a pronounced increase in GABA levels and a reduction in glutamate levels, indicating microbial conversion during fecal fermentation. Statistical significance was determined using one-way analysis of variance followed by Tukey’s multiple comparison test, with comparisons indicated above the bars (* = p 0.05; ** = p 0.01; and *** = p 0.001). GABA, Gamma-aminobutyric acid; MSG, monosodium glutamate; ANOVA, one-way analysis of variance.

To further investigate the effect of dietary components on microbial activity, prebiotics including MOS, DEX, and FOS were introduced. Supplementation with these oligosaccharides resulted in a significant increase in GABA levels across all treatment groups compared with the control (untreated) group. Among the evaluated prebiotics, MOS and FOS produced the highest GABA concentrations, with statistically significant differences (p 0.01 to p 0.001) observed particularly between the control and MOS/FOS groups (Fig. 2A). In contrast, glutamate concentrations remained relatively stable across all treatments (Fig. 2B), suggesting that the increase in GABA was not merely attributed to elevated substrate availability but rather to enhanced microbial metabolic modulation.

Fig. 2.

Effect of different treatments on gamma amino butyric acid (GABA) and glutamate concentrations in the gut microbiota culture. Bar graphs show the mean concentrations ($\pm$SD) of GABA (A) and glutamate (B) in response to four treatments: Control (Cont.), mannan oligosaccharides (MOS), dextrin (DEX), and fructo-oligosaccharides (FOS). GABA concentrations were significantly higher in the MOS and FOS groups compared with the control. Although MOS and FOS showed higher average GABA levels than DEX, the differences were not statistically significant (p $>$ 0.05). Glutamate levels remained relatively low and stable across treatments. Data are based on pooled measurements from five biological replicates per group. Statistical significance was determined using one-way ANOVA followed by Tukey’s multiple comparison test, with comparisons indicated above the bars (** = p 0.01; *** = p 0.001). Error bars represent standard deviations.

During the 24-hour fermentation, pH values decreased more markedly and remained lower in cultures treated with oligosaccharides, particularly MOS and FOS, compared with the control, which showed only a slight drop followed by partial recovery (Fig. 3A). This sustained acidification indicates enhanced microbial fermentation activity.

Fig. 3.

Impact of different treatments on gut fermentation and microbial composition. (A) pH dynamics over 24 hours of fermentation under different treatments: Control (Cont.), mannooligosaccharides (MOS), dextrin (DEX), and fructooligosaccharides (FOS). A rapid decline in pH occurred in all groups within the first 8 hours, followed by partial recovery in the control and DEX groups, whereas the pH remained low in the MOS and FOS treatments, indicating enhanced fermentation activity. (B) Phylum-level relative abundance of gut microbiota across five biological replicates per treatment. Firmicutes and Proteobacteria were dominant across most treatments. In most donners, MOS and FOS treatments exhibited increased proportions of Actinobacteria and Proteobacteria compared to the control. (C) Relative abundance of Bifidobacterium species. B. adolescentis and other Bifidobacterium species were particularly enriched in MOS and FOS treatments, with noticeable inter-individual variation. B. longum and B. breve were detected in lower but variable proportions.

Microbiota analysis indicated an enrichment of Actinobacteria across all treatment groups, with the greatest increase observed under MOS and FOS conditions (Fig. 3B). Within this phylum, Bifidobacterium species, particularly B. adolescentis, became more predominant, especially in donor samples 1 and 2 (Fig. 3C). This taxonomic shift aligns with previous reports demonstrating B. adolescentis’s preference for oligosaccharides and its ability to synthesize GABA [25, 26].

Mannan hydrolysis requires multiple synergistic glycoside hydrolases. Genomic analysis revealed that B. adolescentis 4-2 possesses two key genes involved in this process: manB, encoding a $\beta$-mannosidase precursor, and manA, encoding an exo-acting $\beta$-mannosidase (Fig. 4A). The $\beta$-mannosidase activity of B. adolescentis 4-2 was evaluated under various carbon sources. MOS supplementation significantly increased enzymatic activity compared with glucose, indicating that MOS serves as a preferred substrate (Fig. 4B). These results confirm the strain’s capacity to assimilate MOS efficiently.

Fig. 4.

$\beta$-mannosidase induction by different carbon sources and proposed manno-oligosaccharide (MOS) utilization genes. (A) Schematic representation of the proposed MOS utilization genes. (B) $\beta$-mannosidase activity in response to different carbon sources. Cultures were grown using glucose, MOS, mannose, or no sugar as the sole carbon source. The $\beta$-mannosidase concentration (mM) was quantified and plotted. Bars represent the mean $\pm$ standard deviation. MOS induced the highest $\beta$-mannosidase production, whereas mannose and no sugar conditions resulted in minimal enzyme activity, indicating that MOS specifically triggers $\beta$-mannosidase induction.

To assess the potential symbiotic effect between B. adolescentis 4-2 and oligosaccharides, a gut microbiota model was employed using fecal samples characterized by low baseline GABA levels. Four experimental setups were tested: control (no additives), 0.5% MOS alone, B. adolescentis 4-2 alone, and a combination of 0.5% MOS with B. adolescentis 4-2. GABA concentrations increased progressively across the groups in the following order: MOS, B. adolescentis 4-2, and MOS + B. adolescentis 4-2, while glutamate levels were similar to those in the control group (Fig. 5A). The combined treatment produced the highest GABA yield, demonstrating a synergistic interaction between the probiotic strain and MOS under gut-simulating conditions. To confirm this symbiosis, $\beta$-mannosidase activity was analyzed across all treatment groups. The highest activity was detected in cultures treated with both B. adolescentis 4-2 and MOS (Fig. 5B), confirming the active utilization of MOS by the strain and reinforcing its role in enhancing GABA biosynthesis through a prebiotic-probiotic synergy. This interaction supports the potential development of symbiotic formulations designed to modulate GABA levels within the gut environment.

Fig. 5.

Effects of treatments on gamma amino butyric acid (GABA) and glutamate concentrations and mannosidase activity. (A) Bar graph displaying the concentrations of GABA (blue bars) and glutamate (red bars) across four treatment groups: Control, MOS (mannan oligosaccharides), B. adolescentis 4-2, and B. adolescentis 4-2 + MOS. GABA concentrations increased significantly in both the B. adolescentis 4-2 and B. adolescentis 4-2 + MOS treatments, accompanied by corresponding reductions in glutamate levels. (B) Bar graph showing mannosidase activity (measured in µM/min/mL) in the same groups. The highest enzymatic activity was recorded in the B. adolescentis 4-2 + MOS group, indicating a synergistic effect. Statistical significance was determined using one-way ANOVA followed by Tukey’s multiple comparison test, with comparisons noted above the bars (ns = not significant; ** = p 0.01; and *** = p 0.001). Error bars represent standard deviations.