Production of Indole-3-Lactic Acid by Bifidobacterium Strains Isolated fromHuman Infants

Recent studies have shown that metabolites produced by microbes can be considered as mediators of host-microbial interactions. In this study, we examined the production of tryptophan metabolites by Bifidobacterium strains found in the gastrointestinal tracts of humans and other animals. Indole-3-lactic acid (ILA) was the only tryptophan metabolite produced in bifidobacteria culture supernatants. No others, including indole-3-propionic acid, indole-3-acetic acid, and indole-3-aldehyde, were produced. Strains of bifidobacterial species commonly isolated from the intestines of human infants, such as Bifidobacterium longum subsp. longum, Bifidobacterium longum subsp. infantis, Bifidobacterium breve, and Bifidobacterium bifidum, produced higher levels of ILA than did strains of other species. These results imply that infant-type bifidobacteria might play a specific role in host–microbial cross-talk by producing ILA in human infants.


Introduction
Tryptophan can be metabolized by microbiota organisms.
Tryptophan metabolites, including indole-3-lactic acid (ILA), indole-3-propionic acid (IPA), indole-3-acetic acid (IAA), and indole-3-aldehyde (IAld), play important roles in host homeostasis. These tryptophan metabolites have been reported to act as agonists of the aryl hydrocarbon receptor and farnesoid X receptor (FXR) [1,2]. IAA can suppress inflammatory responses of cytokine-mediated lipogenesis in hepatocytes via the reduction of pro-inflammatory cytokine production in macrophages [3]. IAld stimulates lamina propria lymphocytes to secret IL-22 and restores the barrier function of damaged intestinal mucosa by, in turn, stimulating the proliferation of intestinal epithelial cells [4]. IPA regulates gastrointestinal barrier functions by the downregulation of enterocyte tumor necrosis factor-α (TNF-α) and the upregulation of junctional proteins [5]. It has also been reported that these tryptophan metabolites can inhibit amyloid fibrillation of lysozymes and that they possess neuroprotective properties [6]. ILA scavenges free radical products and inhibits the UVB-induced production of interleukin-6 (IL-6) [7,8]. ILA was reported to reduce TH17 polarization which suppresses inflammatory T cells and gut intraepithelial CD4 + CD8αα + T cells (immunoregulatory T cells) [9,10]. ILA acts as an agonist of human hydroxycarboxylic acid receptor 3 and induces a decrease in cAMP in human monocytes [11]. It has been reported that Bifidobacterium strains produce ILA [12]. However, to the best of our knowledge, information relating to tryptophan metabolite-producing Bifidobacterium strains is scarce [13].
Bifidobacterium strains commonly found to colonize the human gut are designated as human-residential bifidobacteria (HRB), while Bifidobacterium strains that naturally colonize the gut of other animals are referred to as non-HRB. B. breve, B. longum subsp. infantis, B. bifidum, and B. longum subsp. longum are most frequently observed Bifidobacterium species in human infants (infant-type HRB) [14][15][16][17][18]. It is important to note that the distribution of bifidobacterial species changes with host age, which is caused by age-related changes in dietary habits [19]. Dominant HRB in adults are referred to as adult-type HRB.
The present study aimed to evaluate the capacity of Bifidobacterium strains to produce various tryptophan metabolites (ILA, IAA, IAld, and IPA). We first examined 19 typical strains that are available from public culture collection facilities. Then, the ability of 100 newly isolated strains [20] to produce ILA was examined.

Production of Tryptophan Metabolites by Bifidobacterium strains
To begin with, 19 bifidobacterial strains obtained from culture collections were tested by culturing in de Man, Rogosa and Sharpe (MRS) ( Table 1). No obvious differences in growth were observed. MRS did not contain ILA (< 0.005 µg/mL), and explicit production of ILA was observed in culture supernatants (CSs). The average concentration of ILA in CSs of infant-type HRB (B. longum subsp. longum, B. longum subsp. infantis, B. breve, and B. bifidum) was higher compared with other strains (Table 1). Other tryptophan metabolites (IAld, IAA, and IPA) were not produced by any of the strains tested ( Figure S1). To confirm the differences in ILA production among each of the Bifidobacterium species, a total of 100 newly isolated strains were also tested. Figure S2 shows the concentration of ILA in the CSs of these 100 strains, and the data are summarized in Table 2. The average concentration of ILA in CSs of B. longum, B. breve, B. bifidum, and Bifidobacterium kashiwanohense was higher than in CSs of Bifidobacterium pseudocatenulatum, Bifidobacterium adolescentis, and Bifidobacterium dentium.
** Statistically significant difference in ILA production between infant-type HRB and adult-type HRB. ## Statistically significant difference in ILA production between infant-type HRB and non-HRB. The rate of growth (OD600) and concentration of ILA in culture supernatants is shown. Values are expressed as means ± S.D.

Discussion
Microbiota-derived tryptophan metabolites play important roles in their hosts' homeostasis [21,22]. Some bifidobacterial strains produce ILA, IAA, and IPA [13]. In this investigation, we tested the ability of various bifidobacterial strains to produce tryptophan metabolites (IAld, IAA, IPA, and ILA). We observed only the production of ILA by bifidobacterial strains (Table 1). These tryptophan metabolites are found in plants as auxins or their intermediates, and MRS broth containing a digest of soybean, we suppose that IAld, IAA, and IPA were derived from the ingredients of MRS broth [23,24]. The same results were observed not only in MRS broth CSs but also in Gifu Anaerobic Medium (GAM) broth CSs (Table S2). In addition, we found that the ability to produce ILA reflected strain-specific features. That is B. longum subsp. longum, B. longum subsp. infantis, B. breve, and B. bifidum, which are usually found in the intestines of human infants and designated infant HRBs [16,25], produced relatively higher levels of ILA compared with the other strains (Table 1). We further investigated the production of ILA by 100 newly isolated bifidobacterial strains [20]. The production of ILA by infant HRBs was significantly higher than the production of this compound by B. pseudocatenulatum, B. adolescentis, and B. dentium. We did not examine the type-strain of B. kashiwanohense, which has previously been isolated from the feces of healthy infants [26]. Therefore, although we recognize that B. kashiwanohense can be classified as an infant HRB, there were too few B. kashiwanohense CSs to judge the results.
The mechanism of the production of ILA from bifidobacterial strains was not clarified in this study. However, we suppose two metabolic pathways for the ILA production by infant-type HRB. One possible pathway is through tryptophan deamination by amino acid oxidase (AAO) [27]. Another metabolic pathway is a conversion from tryptophan to indolepyruvic acid by aromatic amino acid aminotransferase (Aat), followed by conversion to ILA by phenyllactate dehydrogenase (fldH) [28], although the related gene was not identified in this study.
Our result suggests that further investigation of ILA biological meaning is needed to fully understand how and why only limited species (infant-type HRB) are allowed to harbor in the human infant gut. As described in the introduction, ILA has been reportedly involved in inducing immunoregulatory T cells [9,10] and suppressing inflammatory T cells [29][30][31][32]. This would be one of the benefits for normal growth, including the immune development in infants. From the bacteria aspect, we speculate that ILA production by infant-type HRB may contribute to the predominance of themselves in the infant's large intestine because ILA was reported to have antimicrobial activity [33] in addition to H 2 O 2 production as a by-product during tryptophan deamination [34].

Bacterial Strains
Bifidobacterial strains were obtained from the Morinaga Culture Collection (Morinaga Milk Industry Co., Ltd., Zama, Japan) or purchased from the American Type Culture Collection (Manassas, VA, USA), the Japan Collection of Microorganisms (Wako, Japan), the German Collection of Microorganisms (DSMZ; Braunschweig, Germany), or the Laboratorium voor Microbiologie (LMG; Belgium). A further 100 newly isolated strains, which were reported in a previous study [25], were also used.

Culture Supernatants (CSs)
Initially, all bifidobacterial strains tested were maintained by culturing at 37 • C for 16 h under anaerobic conditions in MRS-C. The growth-phase bacterial cells were then harvested by centrifugation [high-speed centrifugal refrigerating machine, HIMAC SCR20B (Hitachi Koki Co., Ltd., Tokyo, Japan)] at 5000× g (4 • C for 10 min) and washed twice with phosphate buffered saline (PBS) and Dulbecco's Formula (DS Pharma Biomedical Co., Ltd., Osaka, Japan). Subsequently, whole-cell pellets were suspended in PBS containing 0.05% L-cysteine (PBS-C). The optical density (at 600 nm) of each bacterial cell suspension was adjusted to the same value (OD600 = 0.2) using PBS-C. Cell suspensions (100 µL) were added to MRS-C (3 mL) and cultured at 37 • C for 24 h under anaerobic conditions. The CSs were obtained by centrifuging the culture suspensions at 5000× g (4 • C for 10 min). Following filtration (pore size 0.22 µm; Millipore, MA, USA), the samples were stored at −80 • C until use. All cultures were grown in independent triplicates, and the resulting data were expressed as the mean of these replicates.
Quantitation was performed by comparing metabolite concentrations in CSs with those of the corresponding synthetic compound standards (IAA, IAld, IPA, and ILA) and the internal standard (MOI). The LC-MS/MS spectrum (product ion data) of the positive precursor ion was evaluated to determine their final content (Table S1).

Statistical Analyses
Intergroup differences in ILA production were analyzed using unpaired t-tests. p values < 0.001 were considered statistically significant.

Conclusions
In conclusion, we examined the ability of various bifidobacterial strains to form tryptophan metabolites. We found that typical infant-type HRB produced significantly higher concentrations of ILA compared with adult-type HRB and non-HRB. Future investigations of ILA-producing microbiota will help to further reveal the role of infant-type HRB in the human gut.

Conflicts of Interest:
The authors declare that they have no competing interests.