Indole Derivatives Biosynthesis in Bifidobacterium longum subsp. infantis and the Tryptophan Substrate Availability

ABSTRACT

The metabolic processes of Bifidobacterium longum subsp. infantis, an early coloniser of the human gut, are essential for gut health, mainly due to the production of indole derivatives from tryptophan. This study investigates the capacity of B. infantis ATCC 15697 to biosynthesise indole‐3‐lactate (ILA), indole‐3‐acetate (IAA), and indole‐3‐carboxaldehyde (I3CA) and the regulatory effects of substrate availability on these pathways. The tryptophan catabolic profile of B. infantis ATCC 15697 under a non‐growing but metabolically active state was investigated. Through HPLC‐PDA and LC–MS analyses, we confirmed for the first time the production of IAA and I3CA by B. infantis ATCC 15697. The results revealed a dose‐dependent relationship between tryptophan availability and the production of indole derivatives, highlighting the nutrient‐driven effect of these metabolic pathways. By integrating genomic analysis with metabolic profiles, we proposed potential pathways underlying the biosynthesis of IAA and I3CA from tryptophan. These findings enhance our understanding of the role of B. infantis ATCC 15697 in human health, with ILA, IAA, and I3CA contributing to immune modulation and gut health. We also provide a platform for using B. infantis ATCC 15697 as a biocatalyst for the biosynthesis of beneficial indole derivatives through whole‐cell bioconversion, which was further demonstrated in B. infantis ATCC 25962 and ATCC 15702. Future in vivo studies will help clarify the impact of these metabolites on the gut environment and inform dietary and probiotic strategies for enhancing indole derivatives production.

Bifidobacterium longum subsp. infantis converts tryptophan into indole‐3‐lactate, indole‐3‐acetate and indole‐3‐carboxaldehyde. Production is substrate‐dependent, with higher tryptophan levels enhancing yields. This study reveals metabolic modulation in nutrient‐rich and resting states, emphasising B. infantis ' role in producing beneficial indole derivatives for gut health.

1. Introduction

Bifidobacterium spp. is a genus of obligate, anaerobic, Gram‐positive bacteria characterised by a high G+C content and commonly found in the intestines and faeces of humans and animals (Whitman et al. 2012). These bacteria are known for their unique ability to metabolise carbohydrates to produce acetate and lactate, contributing to a balanced gut environment (Reens et al. 2024). They have been studied for their health‐promoting properties, such as producing anti‐inflammatory compounds like indole‐3‐lactate, derived from tryptophan metabolism (Meng et al. 2020). Among the different species, Bifidobacterium longum subsp. infantis ( B. infantis ) stands out as a significant coloniser of the infant gut, aiding in developing a healthy microbiota (Ficara et al. 2020).

B. infantis , a short, rod‐shaped bacillus, was initially isolated from the intestines of infants and is considered an essential part of the early gut microbial community (Whitman et al. 2012). B. infantis catabolises tryptophan and produces indole‐derivatives including indole‐3‐pyruvate, indole‐3‐lactate and indole‐3‐propionate (Smith and Macfarlane 1996) (Laursen et al. 2021). B. infantis is one of the earliest colonisers in newborns, playing a critical role in shaping the gut microbiota and contributing to immune system development (Underwood et al. 2015). Recent studies have established a link between the presence of gut B. infantis and reduced markers of inflammation, highlighting its potential protective effects (Chichlowski et al. 2020). Due to these benefits, B. infantis has been incorporated into dietary supplements and probiotic formulations. Supplementation with synbiotic combinations, such as human milk oligosaccharides and B. infantis , has demonstrated reversible engraftment in the microbiomes of healthy adults, showcasing its adaptability across different age groups (Button et al. 2022).

A pivotal mechanism through which B. infantis exerts its beneficial effects is the production of indole derivatives, such as indole‐3‐lactate, via tryptophan catabolism (Meng et al. 2020). However, its pathways for synthesising other potentially beneficial indole derivatives, such as indole‐3‐acetate and indole‐3‐carboxaldehyde, are less understood. Indole‐3‐acetate and indole‐3‐carboxaldehyde (also known as indole‐3‐aldehyde) were shown to activate immune responses and help prevent infections such as candidiasis (Zelante et al. 2021).

Although current metabolic databases and studies have mapped specific metabolic pathways and enzymes responsible for bacterial tryptophan catabolism, many pathways remain incompletely understood. For B. infantis strains, current research focuses on indole‐3‐lactate, but there is still no direct evidence on whether it can produce indole‐3‐acetate and indole‐3‐carboxaldehyde. Previous studies have shown that Lactobacillus spp. can synthesise indole‐3‐acetate and indole‐3‐carboxaldehyde through an uncharacterised enzyme system (Zelante et al. 2021), but it remains unclear whether B. infantis has a similar metabolic mechanism.

This study hypothesised that B. infantis has the potential to synthesise indole‐3‐acetate and indole‐3‐carboxaldehyde and that their production is regulated by tryptophan substrate availability. We investigated the production of indole derivatives from the tryptophan catabolic pathway of B. infantis ATCC 15697 and assessed how substrate availability influences this process. Our research demonstrates that B. infantis ATCC 15697 can metabolise tryptophan to produce indole‐3‐lactate, indole‐3‐acetate, and indole‐3‐carboxaldehyde, with these metabolic pathways modulated by the availability of tryptophan. We further explored the potential metabolic pathways through genomic analysis to identify specific genes or enzymes associated with indole derivatives production. This exploration offers valuable insights into the metabolic flexibility of B. infantis ATCC 15697 and its potential role in human health.

2. Experimental Procedures

2.1. Bacterial Strains, Growth Media and Culture Conditions

Bifidobacterium longum subsp. infantis ATCC 15697^T ( B. infantis ATCC 15697^T), B. infantis ATCC 15702, and B. infantis ATCC 25962 were purchased from Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan). B. infantis strains were routinely cultured in De Man, Rogosa and Sharpe (MRS) medium and modified reinforced clostridial (MRC) medium. MRS medium is widely recognised as the standard for culturing lactic acid bacteria, whereas MRC medium is recommended by the American Type Culture Collection (ATCC) for cultivating B. infantis . The modified MRC medium (mMRC) was used to evaluate the effect of glucose on indole‐3‐lactate production. The MRS medium was composed of 1% peptone (BD Difco, Franklin Lakes, USA), 1% beef extract (BD Difco, Franklin Lakes, USA), 0.5% yeast extract (BD Difco, Franklin Lakes, USA), 2% dextrose (Sigma‐Aldrich, St. Louis, USA), 0.1% polysorbate 80 (Sigma‐Aldrich, St. Louis, USA), 0.2% ammonium citrate (Sigma‐Aldrich, St. Louis, USA), 0.5% sodium acetate (Sigma‐Aldrich, St. Louis, USA), 0.01% magnesium sulfate (Sigma‐Aldrich, St. Louis, USA), 0.005% manganese sulfate (Sigma‐Aldrich, St. Louis, USA), and 0.2% dipotassium phosphate (J.T. Baker, Phillipsburg, USA) with 0.05% l‐cysteine HCl (TCI, Tokyo, Japan). The MRC medium was composed of 1% tryptone (BD Difco, Franklin Lakes, USA), 1% beef extract (BD Difco, Franklin Lakes, USA), 0.3% yeast extract (BD Difco, Franklin Lakes, USA), 0.5% dextrose (Sigma‐Aldrich, St. Louis, USA), 0.5% sodium chloride (Sigma‐Aldrich, St. Louis, USA), 0.1% soluble starch (BD Difco, Franklin Lakes, USA) and 0.3% sodium acetate (Sigma‐Aldrich, St. Louis, USA) with 0.05% L‐cysteine HCl (TCI, Tokyo, Japan). The mMRC medium had the same composition as the MRC medium, except that the glucose concentration was increased to 2% (w/v). All bacterial cultures were incubated in an anaerobic glove chamber (Forma Scientific Model 1025, Thermo Fisher Scientific, Waltham, USA) under an atmosphere of 85% N₂, 10% CO₂, and 5% H₂. The cultures were maintained at 37°C without shaking.

2.2. Growth and Metabolic Profile of B. infantis ATCC 15697 Grown in MRS and MRC Media

For growth experiments, a 3 mL overnight culture (12 h) of B. infantis ATCC 15697^T grown in MRS was inoculated into 100 mL of MRS medium, while that grown in MRC media was inoculated into 100 mL MRC medium. The bacterial cultures were sampled at 0, 3, 6, 9, 12, 15, 24, 36, and 48 h. Then, the bacterial growth and metabolic activity (carbon source consumption and production of indole‐3‐lactate, acetate and lactate were monitored). The experiments were conducted in three independent biological replicates.

2.2.1. Determination of Bacterial Growth

Bacterial growth was assessed by measuring the culture turbidity and quantifying viable cell counts by plate counting. For turbidity measurement, we determined the optical density at 600 nm (OD₆₀₀) of samples taken from bacterial cultures using a spectrophotometer (UV‐1800, Shimadzu, Kyoto, Japan). For viable cell counts quantification via plate counting, bacterial culture samples were collected at each time point and subjected to 10‐fold serial dilution to achieve an appropriate concentration. A suitable concentration was defined as one in which 100 μL of the diluted suspension yielded 30–300 single colonies per agar plate. Subsequently, a 100 μL aliquot of the diluted suspension was evenly spread onto the corresponding agar plate (MRS agar plate or MRC agar plate, depending on the culture medium) and incubated for enumeration. Anaerobic incubation was conducted in an anaerobic glove chamber (Model 1025, Thermo Fisher Scientific, Waltham, USA) under an atmosphere of 85% N₂, 10% CO₂, and 5% H₂ at 37°C for 48 h. Colony and cellular morphology were examined using light microscopy (Axio Lab A1, ZEISS, Oberkochen, Germany) and compared with standard reference strains to exclude potential contamination. The viable cell counts (expressed as CFU/mL) were determined by multiplying the colony count by the serial dilution factor and a factor of 10, accounting for only 100 μL of suspension being plated.

2.2.2. Quantitative Analysis of Primary Carbon Sources and Metabolites in the Medium

Glucose, indole‐3‐lactate, acetate and lactate contents in the MRS and MRC growing cultures were measured. The glucose content in the medium was measured using a glucose xolorimetric assay kit (Elabscience, Houston, USA) according to the manufacturer's instructions. Indole‐3‐lactate content in the medium was analysed using an HPLC‐PDA system. Lactate and short‐chain fatty acids (SCFAs) were extracted using liquid–liquid extraction. During the SCFA extraction process, the bacterial suspension was centrifuged, and 1000 μL of supernatant was collected. An internal standard of 50 μL of 0.2 M succinic acid (TCI, Tokyo, Japan) was added, followed by thorough mixing. Afterward, 100 μL of concentrated HCl (Sigma‐Aldrich, St. Louis, USA) was added, and diethyl ether (TCI, Tokyo, Japan) was added for the extraction. The aqueous phase was discarded, and the retained organic phase was combined with 1000 μL of 1 M NaOH (TCI, Tokyo, Japan) for a second extraction. The resulting aqueous phase was transferred to a microcentrifuge tube and mixed with 200 μL of concentrated HCl (Sigma‐Aldrich, St. Louis, USA). Finally, the samples were filtered through a 0.22 μm nylon membrane (Dikma, Foothill Ranch, USA) and analysed using an HPLC‐PDA system (LC‐2030, Shimadzu, Kyoto, Japan) to quantify lactate and acetate concentrations.

2.3. Tryptophan Catabolism by Whole Cell Bioconversion and Indole‐Derivatives Analysis

2.3.1. Resting Cell Preparation

The resting cells of B. infantis ATCC 15697^T, B. infantis ATCC 15702, and B. infantis ATCC 25962 grown in MRC medium were prepared following established protocols (Zhang and Huang 2023). The term resting cells in this study specifically refers to B. infantis strains in a non‐growing but metabolically active state. In brief, bacteria grown to log phase in MRC medium were harvested by centrifugation (4000 g, 10 min, 4°C). Then, bacteria were washed twice with the washing buffer consisting of 75 mM potassium chloride (Sigma‐Aldrich, St. Louis, USA), 75 mM sodium chloride (Sigma‐Aldrich, St. Louis, USA), and 2 mM magnesium chloride (Merck, Darmstadt, Germany) in 2 mM PPB (4°C, pH 7.0). Then, the cell pellet was resuspended in a 0.1 M potassium phosphate buffer (PPB, pH 7.0) containing varying concentrations of tryptophan (5, 25, and 50 mM). The bacterial suspension was adjusted to an optical density (OD) of 1.0 ± 0.05 at 600 nm using the same buffer. Bacterial cultivation, washing, and resuspending were all conducted in anaerobic conditions (85% N₂, 10% CO₂, and 5% H₂).

For the qualitative analysis of indole‐derivatives production of the resting cells of the three B. infantis strains, bacterial suspensions (1 mL) containing 5 mM tryptophan were incubated for 24 h at 37°C under anaerobic conditions (85% N₂, 10% CO₂, and 5% H₂). For the analysis of the substrate availability effect, bacterial suspensions (1 mL) containing different concentrations of tryptophan (5, 25, and 50 mM) were incubated for 24, 48, and 72 h at 37°C under anaerobic conditions (85% N₂, 10% CO₂ and 5% H₂). All experiments were conducted in three independent biological replicates.

2.3.2. Identification of Indole‐Derivatives in the Reaction Supernatant

To identify indole derivatives, B. infantis resting cell suspensions were incubated in 5 mM tryptophan buffer and collected after incubation for 24 h. The samples were centrifuged at 10,000 g for 10 min at 4°C, and the resulting supernatants were analysed by a Waters ACQUITY UPLC system coupled with a Waters Xevo Q‐TOF G2‐XS mass spectrometer. Chromatographic separation was performed on an ACQUITY UPLC BEH C18 column (100 × 2.1 mm, 1.7 μm) under positive ion mode. The column was maintained at 45°C, with mobile phase A comprising water with 0.1% formic acid (FA) and mobile phase B comprising acetonitrile (ACN) with 0.1% FA. The gradient conditions were as follows: an initial flow rate of 0.3 mL/min with 99.0% mobile phase A and 1.0% mobile phase B, held until 0.50 min. The gradient was shifted to 0.0% mobile phase A and 100% mobile phase B at 4.50 min, maintained until 5.00 min, and then returned to 99.0% mobile phase A and 1.0% mobile phase B by 6.00 min, continuing until 9.00 min. Retention times and mass spectra were compared with standards (indole‐3‐lactic acid, indole‐3‐acetic acid, and indole‐3‐carboxaldehyde) to confirm the identification.

2.3.3. Analysis of Substrate Availability Effect

To analyse the substrate availability effect, B. infantis ATCC 15697^T resting cell suspensions containing different concentrations of tryptophan (0, 0.5, 5, 25, and 50 mM) were incubated for 24, 48, and 72 h. After incubation, each sample was centrifuged at 10,000 g for 10 min at 4°C. The supernatant was filtered through a 0.22 μm nylon membrane (Dikma, Foothill Ranch, USA) to analyse tryptophan and its indole derivatives. Detection and quantification were performed using a high‐performance liquid chromatography (HPLC) system (LC‐2030, Shimadzu, Kyoto, Japan) equipped with a YMC‐Triart C18 RP column (250 × 4.6 mm, 5 μm). The mobile phase consisted of a 60:40 mixture of 0.1% acetic acid solution and methanol. The injection volume was set to 10 μL, with an oven temperature of 35°C and a 1 mL/min flow rate.

2.3.4. Resting Cells Viability Analysis

A LIVE/DEAD BacLight Bacterial Viability Kit (Thermo Fisher Scientific, USA) was used to evaluate cell viability according to the manufacturer's instructions. The fluorescent dye working solution consisted of SYTO 9 and propidium iodide. Bacterial suspensions were collected by centrifugation (10,000 g, 3 min, 4°C), washed twice with water, and resuspended in water to an optical density (OD) at 600 nm of 0.25. Positive controls (live bacteria) were prepared from log‐phase cultures grown in MRC medium, while negative controls (dead bacteria) were treated with 75% ethanol. For staining, 100 μL of the cell suspension was combined with 100 μL of the dye solution and incubated in the dark at 25°C for 15 min. The stained bacteria were visualised using an automated fluorescence microscope (Olympus BX63, Tokyo, Japan). Live cells were identified by SYTO 9 fluorescence using a FITC filter, and dead cells were detected by propidium iodide fluorescence using a TRITC filter.

2.4. Bacterial Tryptophan Catabolic Pathway Construction and Related Genomics Analysis

The information on tryptophan to indole‐derivatives pathways was constructed based on a literature review (Tables 2 and 3), MetaCyc Pathway database (Caspi et al. 2014), and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway database (Kanehisa 2002). Both Gram‐negative and Gram‐positive bacteria were included. The genome of B. infantis was analysed to identify potential gene analogs involved in tryptophan metabolism. Protein sequences of known tryptophan catabolic genes from various bacteria were retrieved and used as queries in BLASTp searches against the Bifidobacterium genome database. Functional annotations of identified genes were confirmed using sequence alignments and conserved domain analysis. Homologues of known tryptophan catabolic genes were identified using the following cutoffs: query coverage ≥ 50%, percent identity ≥ 30% and e‐value ≤ 1e−3. Alignment statistics, including query cover and E‐values, were recorded to compare homologues from other bacterial species.

TABLE 2.

The known bacterial pathways of tryptophan catabolism to indole‐3‐acetate.

Pathway	Bacteria class	Bacterial specie	Reference
Tryptophan ↓ Indole‐3‐pyruvate ↓ Indole‐3‐acetaldehyde ↓ Indole‐3‐acetate	α‐Proteobacteria	Agrobacterium tumefaciens Azospirillum brasilense Azospirillum lipoferum Bradyrhizobium japonicum Sinorhizobium meliloti Rm2011	Kaper and Veldstra (1958)
	β‐Proteobacteria	Burkholderia pyrrocinia	Lübbe (1983)
	γ‐Proteobacteria	Azotobacter vinelandii Enterobacter cloacae Escherichia coli K4 Pantoea agglomerans Pseudomonas chlororaphis Pseudomonas fluorescens	Koga et al. (1994)
Tryptophan ↓ Indole‐3‐acetamide ↓ Indole‐3‐acetate	α‐Proteobacteria	Agrobacterium tumefaciens	Powell and Morris (1986)
	γ‐Proteobacteria	Pseudomonas savastanoi	Powell and Morris (1986)
Tryptophan ↓ Indole‐3‐acetaldehyde ↓ Indole‐3‐acetate	γ‐Proteobacteria	Pseudomonas fluorescens	Oberhänsli et al. (1991)
Tryptophan ↓ Tryptamine ↓ Indole‐3‐acetaldehyde ↓ Indole‐3‐acetate	α‐Proteobacteria	Azospirillum brasilense Rhizobium phaseoli	Carreno‐Lopez et al. (2000)
	G(+), Low G+C	Bacillus cereus Enterococcus faecalis	Perley and Stowe (1966), Büki et al. (1985)
	G(+), High G+C	Brachybacterium conglomeratum Micrococcus percitreus	Nakazawa et al. (1974)

TABLE 3.

The known bacterial pathways of tryptophan catabolism to indole‐3‐lactate.

Pathway	Bacteria class	Bacterial specie	Reference
Tryptophan ↓ Indole‐3‐pyruvate ↓ Indole‐3‐lactate	G(+), Low G+C	Clostridium botulinum Clostridium sporogenes Lactobacillus casei Lactobacillus helveticus	Elsden et al. (1976), Gummalla and Broadbent (1999a)

2.5. Statistical Analysis

All data were analysed with unpaired t‐tests in GraphPad Prism 8.3, with statistical significance set at p < 0.05. Results are reported as mean ± standard error of the mean (SEM) from three independent experiments.

3. Results

3.1. Growth and Metabolic Profile of B. infantis ATCC 15697 Grown in MRS and MRC Media

B. infantis ATCC 15697 cultures progressed sequentially through the mid‐log, late‐log, early stationary, and stationary growth phases in both MRS and MRC media. B. infantis ATCC 15697 grown in MRS media had higher final culture turbidity than MRC media (Figure 1A,B), while the viable cell counts were comparable. B. infantis ATCC 15697 utilised glucose and produced lactate and acetate in both MRS and MRC media (Figure 1C,D). However, higher net production of lactate and acetate was observed in MRS medium compared to MRC medium (Figure 1C,D). Notably, despite lower overall growth, B. infantis ATCC 15697 produced more indole‐3‐lactate in MRC medium than in MRS medium (Figure 1E,F). The normalised indole‐3‐lactate production of B. infantis ATCC 15697 is consistently higher in MRC medium than in MRS medium across all time points (Table 1). Notably, B. infantis ATCC 15697 started to produce indole‐3‐lactate when entering the early stationary phase in both MRS and MRC media (Figure 1E,F). These findings indicate that the production of indole‐3‐lactate may be influenced by factors such as the availability of substrate, such as tryptophan or carbon source glucose in the medium.

FIGURE 1.

Growth and metabolic profiles of B. infantis ATCC 15697 grown in MRS and MRC media. (A) Growth (OD₆₀₀) and viable cell counts of B. infantis ATCC 15697 grown in MRS and (B) MRC media. (C) Glucose consumption, acetate production, and lactate production of B. infantis ATCC 15697 grown in MRS and (D) MRC media. (E) Indole‐3‐lactate production of B. infantis ATCC 15697 grown in MRS and (F) MRC media. All data are presented as mean ± SEM of three independent biological replicates.

TABLE 1.

Viable cell counts and normalised indole‐3‐lactate (ILA) production at each time point.

Time (h)	CFU/mL		Normalised ILA production (μM/108 cells)
	MRS	MRC	MRS	MRC
0	(1.05 ± 0.46) × 108	(8.17 ± 2.25) × 107	0.00 ± 0.00	0.00 ± 0.00
3	(1.16 ± 0.38) × 108	(9.93 ± 2.37) × 107	0.00 ± 0.00	0.00 ± 0.00
6	(5.98 ± 2.13) × 108	(6.10 ± 2.24) × 108	0.00 ± 0.00	0.10 ± 0.10
9	(2.16 ± 0.92) × 109	(2.25 ± 0.62) × 109	0.00 ± 0.00	0.31 ± 0.10
12	(5.36 ± 1.41) × 109	(4.17 ± 0.67) × 109	0.01 ± 0.01	0.47 ± 0.20
15	(7.57 ± 0.83) × 109	(8.35 ± 0.89) × 109	0.08 ± 0.01	0.29 ± 0.03
24	(1.33 ± 0.46) × 1010	(1.73 ± 0.16) × 1010	0.10 ± 0.02	0.23 ± 0.05
36	(1.70 ± 0.17) × 1010	(2.51 ± 0.30) × 1010	0.07 ± 0.00	0.19 ± 0.03
48	(5.87 ± 0.37) × 109	(2.08 ± 0.49) × 1010	0.21 ± 0.01	0.24 ± 0.06

To investigate this, we measured the initial tryptophan concentrations in both media. The initial tryptophan concentration in MRS medium was 473.8 ± 46.0 μM, which was lower than the 771.1 ± 71.1 μM observed in MRC medium (Figure S1). This difference may account for the reduced production of indole‐3‐lactate in MRS medium compared to MRC medium. We also evaluated the effect of glucose on the indole‐3‐lactate production activity. Our result shows that B. infantis ATCC 15697 produced comparable levels of indole‐3‐lactate when grown in mMRC medium (2% glucose, w/v) and MRC medium (0.5% glucose, w/v), suggesting that glucose is not a major determinant of indole‐3‐lactate production (Table S1). As a result, MRC medium was selected as the primary culture medium for evaluating indole‐3‐lactate production by B. infantis .

3.2. Identification of Indole‐3‐Acetate and Indole‐3‐Carboxaldehyde Production by B. infantis

Given that B. infantis ATCC 15697 produces indole‐3‐lactate during the stationary phase, a physiological state characterised by metabolic shifts despite limited or arrested bacterial growth, we further investigated whether this strain synthesises additional indole derivatives under a non‐growing but metabolically active state, where its metabolic profile diverges from that of the growing phase. We prepared resting cells of B. infantis ATCC 15697 and incubated them in a potassium phosphate buffer containing 5 mM tryptophan for 24 h. Post‐incubation, the supernatants were collected and analysed using an HPLC‐PDA system for initial indole‐derivatives screening.

To facilitate the identification of potential metabolites, we first established chromatography profiles for a set of standard indole derivatives, including indole‐3‐carboxaldehyde, indole‐3‐acetamide, indole‐3‐acetonitrile, tryptamine (indole‐3‐ethanamine), indole‐3‐pyruvate, indole‐3‐butyrate, indole‐3‐lactate, tryptophol and indole‐3‐acetate. These profiles served as reference points for comparison against the chromatograms obtained from the B. infantis ATCC 15697 supernatants after whole‐cell bioconversion.

Initial findings revealed that B. infantis ATCC 15697 metabolised tryptophan into several indole derivatives, notably indole‐3‐lactate, indole‐3‐acetate and indole‐3‐carboxaldehyde. This observation was substantiated through LC–MS analysis, in which UPLC‐ QTOF MS analysis of the supernatants confirmed the presence of signals corresponding to indole‐3‐lactate, indole‐3‐acetate and indole‐3‐carboxaldehyde (Figure 2).

FIGURE 2.

LC–MS identification of indole‐3‐lactic acid, indole‐3‐acetic acid and indole‐3‐carboxaldehyde using LC‐QTOF MS. Chromatography of indole‐3‐lactic acid, indole‐3‐acetic acid (IAA), and indole‐3‐carboxaldehyde (I3CA). Upper and lower chromatographs show the supernatant of resting cells reaction mixtures of B. infantis ATCC 15697 and authentic standards, respectively. For indole‐3‐lactic acid, the sample displayed a retention time of 2.93 min and the standard displayed a retention time of 2.86 min with a detected mass of 206.082 m/z and a mass error within 10.0 ppm. For indole‐3‐acetic acid, the sample and standard both display a peak at 3.14 min with a detected mass of 176.071 m/z and a mass error within 10.0 ppm. For indole‐3‐carboxaldehyde, the sample and standard both show a peak at 3.09 min with a mass of 146.061 m/z and a mass error within 10.0 ppm. The retention time shifts are attributable to batch‐to‐batch variation between analytical runs. Three independent biological experiments were conducted with the representative images presented.

Overall, these results demonstrate that B. infantis ATCC 15697 not only catabolises tryptophan to produce indole‐3‐lactate but also biosynthesises indole‐3‐acetate and indole‐3‐carboxaldehyde.

3.3. The Production of Indole‐Derivatives by B. infantis ATCC 15697 Was Associated With Substrate Availability

Given that indole derivatives are metabolic products of tryptophan catabolism, we examined the relationship between substrate availability and the production of these derivatives. For this investigation, B. infantis cells in a non‐growing but metabolically active state were incubated with varying concentrations of tryptophan (5, 25, and 50 mM) in a potassium phosphate buffer. Following incubation, the supernatants were analysed using an HPLC‐PDA system to assess the production levels of indole derivatives.

Our results demonstrated a clear dose‐dependent relationship between tryptophan concentration and the synthesis of indole derivatives, specifically indole‐3‐lactate, indole‐3‐acetate, and indole‐3‐carboxaldehyde. As the concentration of tryptophan increased from 5 to 50 mM, a corresponding and significant increase in the production of these indole derivatives was identified (Figure 3) The viable cell count of bacteria substantially decreased (~1,000‐fold) by the end of the incubation period, although a small population of viable cells remained detectable (Figure S2). This trend underscores the regulatory effect of substrate availability on the tryptophan catabolic pathway. Notably, the production of indole‐3‐acetate and indole‐3‐carboxaldehyde is approximately 10 times lower than that of indole‐3‐lactate.

FIGURE 3.

Tryptophan concentration‐dependent production of indole derivatives by Bifidobacterium longum subsp. infantis ATCC 15697. The production of (A) indole‐3‐lactate, (B) indole‐3‐acetate, and (C) indole‐3‐carboxaldehyde by B. infantis in a resting state with different concentrations of tryptophan over time. All data are presented as mean ± SEM of three independent biological replicates; an unpaired two‐tailed t‐test was used for statistical analysis. ns, non‐significant; **p < 0.005; ***p < 0.001.

Our findings indicate that the biosynthesis of indole derivatives in B. infantis ATCC 15697 is modulated by the concentration of available tryptophan, revealing a substrate‐dependent effect of this metabolic pathway. These findings expand our understanding of how B. infantis ATCC 15697 adjusts its metabolic outputs based on environmental nutrient levels.

3.4. Genomic Analysis of Tryptophan Catabolic Pathway of B. infantis

We have identified novel metabolic capabilities of B. infantis to biosynthesize the indole derivatives, indole‐3‐acetate and indole‐3‐carboxaldehyde, from tryptophan (Figure 2). However, the metabolic pathways underlying the production of these indole derivatives in B. infantis ATCC 15697 remained poorly understood. To address this, we first established a comprehensive overview of the tryptophan catabolic pathways across diverse bacterial species based on existing literature and the current metabolic pathway database. These pathways included Bifidobacterium spp. and eight other groups of bacteria, encompassing more than ten genera from both Gram‐negative and Gram‐positive categories (Tables 2 and 3).

Next, the potential gene analogs of known tryptophan catabolic genes were explored in Bifidobacterium spp. genome using BLASTp. Furthermore, we examined whether intermediates involved in tryptophan catabolic pathways were present in B. infantis ATCC 15697 supernatants following whole‐cell bioconversion using UPLC–QTOF MS. Our results revealed the presence of indole‐3‐pyruvate, indole‐3‐lactate, tryptamine, indole‐3‐acetaldehyde, indole‐3‐acetamide, indole‐3‐acetate, and indole‐3‐carboxaldehyde in B. infantis supernatants after whole‐cell bioconversion (Figures 2 and S3). By integrating genomic and metabolic profile data, we proposed a potential tryptophan catabolic pathway in B. infantis ATCC 15697, leading to the production of indole‐3‐lactate, indole‐3‐acetate, and indole‐3‐carboxaldehyde, as illustrated in Figure 4.

FIGURE 4.

Proposed production pathway of indole‐3‐lactate, indole‐3‐acetate and indole‐3‐carboxaldehyde from tryptophan catabolism in B. infantis . Tryptophan (Trp) is converted to indole‐3‐pyruvate (IPyA) by the amino acid transaminase (aat), which is subsequently reduced to ILA by aromatic lactate dehydrogenase (aldh). Alternative pathways involve the decarboxylation of tryptophan to tryptamine (TA) or indole‐3‐acetamide (IAM) via putative genes. TA is further converted to indole‐3‐acetaldehyde (IAAld) via putative genes and oxidised to IAA by aldehyde dehydrogenase (ald). IAM might also be converted to IAA by amidase (gatA). IAA may be further metabolised into I3CA through downstream reactions. This pathway was deduced from genomic analyses and LC–MS detection of intermediates in B. infantis supernatants. The detection of intermediate indole metabolites is presented in Figure S3. The genes blon_0913 and gatA are shown in grey because their putative functional roles are based on bioinformatic prediction. Trp, tryptophan; IPyA, indole‐3‐pyruvate; ILA, indole‐3‐lactate; TA, tryptamine; IAAld, indole‐3‐acetaldehyde; IAM, indole‐3‐acetamide; IAA, indole‐3‐acetate; I3CA, indole‐3‐carboxaldehyde; aat, amino acid transaminase; aldh, aromatic lactate dehydrogenase; blon_0913, aminotransferase, class V; ald, aldehyde dehydrogenase; gatA, glutamyl‐tRNA amidotransferase, A subunit.

For indole‐3‐lactate production, literature review and genomic analysis revealed that B. infantis ATCC 15697 possesses the gene cluster of aromatic lactate dehydrogenase, which is responsible for the conversion of tryptophan to indole‐3‐lactate (Laursen et al. 2021). This gene cluster includes the aat gene (amino acid transaminase), which catalyses the conversion of tryptophan to indole‐3‐pyruvate, and the aldh gene (aromatic lactate dehydrogenase), which reduces indole‐3‐pyruvate to indole‐3‐lactate (Figure 4). For indole‐3‐acetate production, it is speculated that tryptophan is initially decarboxylated to indole‐3‐acetamide or tryptamine. Although the specific genes responsible for these steps have not been confirmed, the presence of indole‐3‐acetamide and tryptamine in the B. infantis ATCC 15697 supernatant suggests the presence of genes involved in tryptophan decarboxylation (Figure S3). Indole‐3‐acetamide is speculated to be converted to indole‐3‐acetate via gatA (glutamyl‐tRNA amidotransferase, A subunit), which is bioinformatically predicted to function as amidase in other Bifidobacterium species, including B. longum subsp. longum KACC 91563 (Figure 4). Tryptophan is metabolised to tryptamine via tryptophan decarboxylase in other bacterial species. The class V aminotransferase of Bifidobacterium spp. showed strong alignment with l‐tryptophan decarboxylase from several Actinobacteria, including Micrococcus lylae UMB0955 (Query cover: 93%, E value: 4e−28) and Pseudonocardia dioxanivorans CB1190 (Query cover: 83%, E value: 2e−35). The gene blon_0913 of B. infantis ATCC 15697 was annotated as class V aminotransferase. Indole‐3‐acetaldehyde is subsequently oxidised to indole‐3‐acetate via ald (aldehyde dehydrogenase) within B. infantis ATCC 15697 genome (Figure 4).

4. Discussion

This study presents significant findings on the metabolic capabilities of Bifidobacterium longum subsp. infantis ATCC 15697 in converting tryptophan into indole‐3‐lactate, indole‐3‐acetate and indole‐3‐carboxaldehyde. To our knowledge, this is the first report of B. infantis ATCC 15697 producing indole‐3‐acetate and indole‐3‐carboxaldehyde. These findings enhance our understanding of tryptophan metabolism in gut‐associated bacteria and highlight the potential role of B. infantis ATCC 15697 in promoting gut health through the production of diverse indole derivatives. It is noteworthy that the production of indole‐3‐acetate and indole‐3‐carboxaldehyde by B. infantis ATCC 15697 was observed exclusively under a non‐growing but metabolically active state. These compounds were not detected in actively growing cultures or stationary phase cultures (Figure 1). Although both stationary phase and resting cells are non‐growing, their metabolic states differ. Stationary phase cells are subjected to nutrient stress and global downregulation of metabolism, whereas resting cells (prepared from log phase cultures and resuspended in defined buffer with tryptophan) retain enzymatic activity under nutrient‐limited conditions (Figure 3), which may favour specific bioconversions such as indole‐3‐acetate and indole‐3 carboxaldehyde production. Our findings align with a previous study that reported no indole‐3‐acetate production by B. infantis ATCC 15697 in growing cultures (Sakurai et al. 2019). These results highlight the significance of utilising cells in a non‐growing but metabolically active state to comprehensively assess their metabolic potential, particularly since the majority of microbes in natural environments exist in a non‐growing state (Bergkessel et al. 2016). From an ecological perspective, this observation may better reflect microbial metabolism in natural environments, where microbes usually do not undergo classical batch‐culture transitions from rapid growth to stationary phase. Instead, they persist in nutrient‐limited, non‐growing but metabolically active states.

Although the complete tryptophan metabolism pathway for B. infantis ATCC 15697 and its associated enzymes have not yet been fully described, the pathways and enzymes identified in Lacticaseibacillus casei and Lactobacillus helveticus provide valuable references for identifying similar gene functions within the B. infantis ATCC 15697 genome (Gummalla and Broadbent 1999a). Phylogenetic analysis confirms that B. infantis ATCC 15697 has two l‐lactate dehydrogenase (LDH) genes: Blon_0840, involved in pyruvate metabolism, and Blon_1090, a type 4 l‐LDH that metabolises indole‐3‐pyruvate to indole‐3‐lactate. Type 4 l‐LDH, with a high affinity for aromatic pyruvate, is classified as an aromatic LDH, supporting a pathway for B. infantis ATCC 15697 to convert indole‐3‐pyruvate to indole‐3‐lactate (Laursen et al. 2021).

Previous studies have established that B. infantis plays an essential role in the gut, producing beneficial metabolites such as indole‐3‐lactate with anti‐inflammatory properties (Ventura et al. 2018). This study expands on these findings by demonstrating that B. infantis ATCC 15697 can also synthesise indole‐3‐acetate and indole‐3‐carboxaldehyde, compounds that have not been previously associated with this species. Indole‐3‐carboxaldehyde is linked to immune modulation in other bacteria like Lactobacillus spp., where it helps prevent conditions such as candidiasis and atherosclerosis (Zelante et al. 2021) (Lu et al. 2023). Indole‐3‐acetate supplementation reduces diet‐induced steatosis and inflammation in mice (Ding et al. 2024). However, it is important to note that the production of indole‐3‐acetate and indole‐3‐carboxaldehyde is approximately 10 times lower than that of indole‐3‐lactate in the resting cell experiment with B. infantis ATCC 15697, which likely reflects differences in metabolic flux between these pathways. Indole‐3‐lactate is synthesised from indole‐3‐pyruvate via aromatic lactate dehydrogenase (aldh), a highly conserved pathway in lactic acid bacteria and Bifidobacteria (Gummalla and Broadbent 1999b; Laursen et al. 2021), suggesting the indole‐3‐lactate synthesis pathway to be the primary tryptophan catabolic route in B. infantis ATCC 15697. In contrast, indole‐3‐acetate biosynthesis likely occurs through either the indole‐3‐acetamide or indole‐3‐acetaldehyde pathway (Figure 4). These pathways are less frequently reported in Bifidobacterium spp., suggesting that the indole‐3‐acetate synthesis pathway may be a minor metabolic route. Furthermore, the multiple enzymatic steps required for indole‐3‐acetate synthesis could be the rate‐limiting factors, leading to the lower production of indole‐3‐acetate compared to indole‐3‐lactate. Notably, the serum concentrations of indole‐3‐acetate and indole‐3‐carboxaldehyde in mice were shown to be much lower compared to indole‐3‐lactate (Lu et al. 2023).

The regulation of indole production by substrate availability aligns with general metabolic control mechanisms seen in lactic acid bacteria and supports the idea that nutrient conditions in the gut can significantly influence microbial activity (Sinha et al. 2024). The dose‐dependent relationship observed in this study, where higher tryptophan concentrations led to increased production of indole derivatives, reinforces the importance of dietary components in influencing microbiome function. The tryptophan catabolic pathway of Clostridium sporogenes is also under the regulation of substrate availability (Sinha et al. 2024).

We also examine two other B. infantis strains for genome analysis and two B. infantis strains for indole‐derivatives analysis to investigate if the tryptophan‐derived indole‐derivatives biosynthetic activity and related genes are conserved across other B. infantis strains. Our findings indicate that these genes are highly conserved across the examined B. infantis strains (Table 4). For the indole derivatives biosynthesis, B. infantis ATCC 25962 and B. infantis 15702 also produced indole‐3‐lactate, indole‐3‐acetate, and indole‐3‐carboxaldehyde under non‐growing but metabolically active states (Figure 5). These results are consistent with the finding derived from B. infantis ATCC 15697, indicating that the production of indole derivatives may be a general property of this species rather than a specific phenomenon of an individual strain.

TABLE 4.

The genes potentially involved in producing ILA, IAA and I3CA in B. infantis ATCC 15697, JCM 7007 and JCM 7009.

Gene	Gene product	B. infantis ATCC 15697	B. infantis JCM 7007	B. infantis JCM 7009
aat	Amino acid transaminase	Blon_0097	BLI007_12160 100% (106/106) a	BLI009_02915 98.1% (104/106)
aldh	Aromatic lactate dehydrogenase	Blon_0099	BLI007_12170 100% (79/79)	NA
Aminotransferase, class V	Aminotransferase, class V	Blon_0913	BLI007_RS03325 100% (415/415)	BLI009_07105 99.3% (412/415)
ald	Aldehyde dehydrogenase	Blon_0259	BLI007_RS13015 100% (511/513)	BLI009_03815 99.6% (511/513)
gatA	Glutamyl‐tRNA amidotransferase, A subunit	Blon_1961	BLI007_RS08900 100% (549/549)	BLI009_10970 97.8% (537/549)

Abbreviation: NA: not applicable.

^a The sequence identity of each gene in B. infantis JCM 7007 and JCM 7009 compared to their counterpart in B. infantis ATCC 15697.

FIGURE 5.

Chromatographic analysis of indole derivatives from B. infantis ATCC 25962 and B. infantis ATCC 15702. Chromatographic analysis of indole‐3‐lactic acid, indole‐3‐acetic acid, and indole‐3‐carboxaldehyde was performed using the supernatants of bioconversion reaction mixtures from ATCC 25962 and ATCC 15702, with authentic standards for comparison. The retention time for indole‐3‐lactic acid was 2.86 min for both ATCC strains, while the standard showed a retention time of 2.86 min, with a detected m/z value of 206.082 and a mass error within 10.0 ppm. For indole‐3‐acetic acid, the retention time was 3.07 min for both strains and 3.14 min for the standard, with an m/z value of 176.071 and a mass error within 10.0 ppm. Indole‐3‐carboxaldehyde exhibited retention times of 3.02 min for both ATCC strains, while the standard showed a peak at 3.09 min, with a detected m/z value of 146.061 and a mass error within 10.0 ppm. The retention time shifts are attributable to batch‐to‐batch variation between analytical runs. Three independent biological replicates were performed, and representative chromatograms are shown.

The discovery that B. infantis can produce indole‐3‐acetate and indole‐3‐carboxaldehyde opens new avenues for understanding how this microbe may contribute to gut health and immune regulation. These indole derivatives have potential roles in modulating immune responses through pathways such as the aryl hydrocarbon receptor (AhR), which is known to influence immune and gut barrier functions (Salminen 2023). Enhanced production of these metabolites could be beneficial in preventing or mitigating inflammatory and autoimmune disorders in the gut (Ji et al. 2020; Li et al. 2024). Furthermore, this study suggests that targeted dietary supplementation with tryptophan or tryptophan‐rich foods may be a future direction to optimise the production of these beneficial indole metabolites. This is supported by the evidence showing that a high‐protein diet is associated with increased tryptophan concentration in the colon of rats (Chung et al. 1975). This insight could guide future probiotic formulation strategies to boost specific metabolic outputs.

While this study provides compelling evidence for the metabolic capabilities of indole derivatives biosynthesis by B. infantis strains, it is essential to acknowledge the limitations. The experiments were conducted under in vitro conditions, which cannot fully replicate the complex dynamics of the gut environment. Additionally, while the study identified potential pathways for indole‐3‐acetate and indole‐3‐carboxaldehyde production, we did not validate the specific genes or enzymes responsible for these metabolic processes. Notably, the identified genes are conserved across different B. infantis strains (Table 4). Future work could involve genetic knockout or overexpression studies to confirm the roles of particular genes.

Building on these findings, future research should explore the in vivo implications of indole‐3‐acetate and indole‐3‐carboxaldehyde production by B. infantis and their interactions with host physiology. Understanding how these metabolites influence the overall gut ecosystem, particularly in conjunction with other commensal microbes like lactate‐utilising bacteria (Zhang et al. 2024), is crucial. Future investigation of the regulatory mechanisms that govern gene expression in the tryptophan catabolic pathway under varying nutritional and environmental conditions could provide deeper insight into how dietary factors and gut microbiota composition impact the metabolic activity of B. infantis .

Overall, this study advances the understanding of tryptophan metabolism by B. infantis , illustrating its capacity to produce multiple indole derivatives and the critical role of substrate availability in this process. These findings pave the way for further studies into how dietary and microbial factors shape gut health and immune regulation through microbial metabolism.

5. Conclusion

This study reveals that Bifidobacterium longum subsp. infantis ATCC 15697 can produce indole‐3‐lactate, indole‐3‐acetate, and indole‐3‐carboxaldehyde from tryptophan, with production influenced by substrate availability. These findings expand the current understanding of B. infantis tryptophan catabolic pathway. This research has practical implications for the food and probiotic industries. By demonstrating the substrate‐dependent production of beneficial indole derivatives, this study suggests that probiotics and functional foods enriched with tryptophan or B. infantis may promote gut health and human well‐being. This opens pathways for future innovations in dietary products aimed at boosting health through microbial metabolism.

Author Contributions

Shi‐Min Zhang: writing – original draft, investigation, methodology, writing – review and editing, conceptualization. Hui‐Chu Wu: conceptualization, formal analysis, investigation, methodology. Jia‐He Hung: writing – original draft, investigation, methodology. Shir‐Ly Huang: conceptualization, writing – review and editing, funding acquisition, resources.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1.

Zhang, S.-M. , Wu H.-C., Hung J.-H., and Huang S.-L.. 2025. “Indole Derivatives Biosynthesis in Bifidobacterium longum subsp. infantis and the Tryptophan Substrate Availability.” Microbial Biotechnology 18, no. 6: e70167. 10.1111/1751-7915.70167.

Data Availability Statement

Data are available on reasonable request.