Commensal microbiota in the intestinal tract plays an important role in the normal development of the immune system(Reference Wills-Karp, Santeliz and Karp1, Reference Macpherson and Harris2). Therefore, strategies for microbiota manipulation have been explored for the prevention of the onset of immune diseases such as allergy. This idea is supported by epidemiological data demonstrating that differences in the composition of gut microbiota in infancy precede the development of atopic dermatitis(Reference Björkstén3). In addition, clinical trials showed that administration of probiotic bacterial strains, such as Lactobacillus rhamnosus GG, was beneficial in both the prevention(Reference Kalliomaki, Salminen and Arvilommi4, Reference Kalliomaki, Salminen and Poussa5)and treatment(Reference Isolauri, Arvola and Sutas6) of early allergic diseases. However, some studies reported that L. rhamnosus GG and Lactobacillus acidophilus did not have a preventative effect in infancy(Reference Taylor, Dunstan and Prescott7–Reference Folster-Holst, Muller and Schnopp9). Therefore, there is currently insufficient data to recommend probiotics as part of a standard therapy in the treatment or prevention of allergic conditions(Reference Prescott and Björkstén10).
Prebiotics, such as indigestible oligosaccharides, have also been considered to promote immune health by selectively stimulating the growth and/or activity of beneficial bacteria, such as bifidobacteria and lactobacilli, in the intestinal tract(Reference Gibson, Beatty and Wang11–Reference Macfarlane, Macfarlane and Cummings14). Indeed, a mixture of long-chain fructo-oligosaccharide (FOS) and short-chain galacto-oligosaccharide reportedly reduced the incidence of atopic dermatitis in formula-fed high-risk infants(Reference Moro, Arslanoglu and Stahl15, Reference Arslanoglu, Moro and Schmitt16). In terms of experimental evidence, we demonstrated that dietary raffinose and α-linked galacto-oligosaccharide reduced allergic airway inflammation in ovalbumin-sensitised Brown Norway rats(Reference Watanabe, Sonoyama and Watanabe17, Reference Sonoyama, Watanabe and Watanabe18). In addition, Fujitani et al. (Reference Fujitani, Ueno and Kamiya19) showed that dietary FOS reduced infiltration of inflammatory cells and oedema formation in duodenal mucosa using an ovalbumin-induced food allergy model of NC/jic mice. Furthermore, Vos et al. (Reference Vos, van Esch and Stahl20) reported that consumption of a mixture of long-chain FOS and short-chain galacto-oligosaccharide suppressed allergic airway inflammation in an ovalbumin-induced allergic asthma model of BALB/c mice. These findings suggest that administration of indigestible oligosaccharides is effective in the prevention of allergic diseases.
Allergic contact dermatitis is one of the most prevalent human skin diseases, causing moderate to severe inflammatory damage. This pathological condition arises after contact hypersensitivity (CHS)(Reference Olmos, Giner and Recio21). CHS is a T-cell-mediated, antigen-specific type of skin inflammation that is induced by topical skin contact with haptens in a previously sensitised host(Reference Krasteva, Kehren and Ducluzeau22, Reference Saint-Mezard, Berard and Dubois23). When the host is sensitised by application of the hapten, skin dendritic cells capture the hapten and migrate to draining lymph nodes, where they prime specific T cells. These cells differentiate into CHS effector cells, which recirculate through the blood. The second contact with the same hapten leads to a skin inflammatory response that peaks 24–48 h after the challenge. Avoidance of causal allergens is one of the emphasised therapeutic suggestions for CHS; however, avoidance is not practical in the majority of cases.
We previously reported that dietary supplementation with short-chain FOS reduced DNFB-induced CHS response in BALB/c mice(Reference Watanabe, Sasajima and Aramaki24). Molecular biological analyses of gut microbiota in these mice, based on 16S rRNA gene sequences, showed that the CHS response, i.e. ear swelling, was negatively correlated with the numbers of bifidobacteria. Because sequence analysis revealed that Bifidobacterium pseudolongum was the most predominant bifidobacterium in the intestine of FOS-supplemented mice, we postulated that a reduction in the CHS response by feeding FOS was associated with the proliferation of B. pseudolongum in the intestinal tract. In the present study, therefore, we aimed to examine whether oral administration of B. pseudolongum, isolated from FOS-supplemented mice, affects the development of DNFB-induced CHS response in mice.
Materials and methods
Animals and diets
The following study was approved by the Hokkaido University Animal Use Committee, and animals were maintained in accordance with the Hokkaido University guidelines for the care and use of laboratory animals.
Female BALB/c mice (5 weeks old) were purchased from Japan SLC (Hamamatsu, Japan) and housed in standard plastic cages in a temperature-controlled (23 ± 2 °C) room with a dark period from 20.00 to 08.00 hours. They were allowed free access to food and water. Mice were allocated to two groups of eleven mice and fed either a synthetic diet prepared according to AIN-93G guidelines(Reference Reeves, Nielsen and Fahey25) or the same diet supplemented with FOS (Meioligo P, donated by Meiji Foodmateria, Tokyo, Japan). These diets were referred to as FOS( − ) and FOS(+), respectively. The FOS(+) diet was prepared by adding (50 g/kg diet) FOS to the FOS( − ) diet at the expense of α-maize starch. In each dietary group, mice were further divided into two subgroups of five or six mice and intragastrically administered vehicle or 2 × 107 cells of viable B. pseudolongum daily as described below. These treatments were referred to as BP( − ) and BP(+), respectively. Thus, mice were divided into four groups: FOS( − )BP( − ), FOS( − )BP(+), FOS(+)BP( − ) and FOS(+)BP(+) (n 5, 6, 5 and 6, respectively). Mice were fed the test diets and administered B. pseudolongum throughout the experimental period. B. pseudolongum was isolated from the faeces of mice-fed FOS(+) in a separate experiment as described below. Two weeks after starting the test diets and B. pseudolongum administration, fresh faeces were collected, and CHS was induced as described below. At 25 d after the first application of DNFB, mice were anaesthetised by inhalation of diethyl ether and blood was drawn from the carotid artery. Cervical lymph node (CLN) was excised for primary cell culture as described below.
Isolation of Bifidobacterium pseudolongum
Fresh faeces were collected from female BALB/c mice-fed FOS(+) for 3 weeks. Faecal samples were homogenised in 450 μl of 50 mm anaerobic phosphate buffer (pH 6·8) that contained 0·05 % (w/v) l-cystein, 0·05 % (w/v) Tween 80 and 0·1 % (w/v) agar. The samples were diluted, and aliquots were plated on bifidobacteria-selective Bifidobacterium iodoacetate medium (BIM) agar(Reference Silvi, Rumney and Rowland26), which consists of MRS agar supplemented with 20 mg/l nalidixic acid, 8·5 mg/l polymyxin B, 50 mg/l kanamycin sulphate, 35 mg/l iodoacetic acid and 25 mg/l 2,3,5-triphenyltetrazolium chloride, and incubated at 37°C for 5 d under anaerobic condition (N2–CO2–H2, 8:1:1, Coy anaerobic chamber, Coy Laboratory Products, Grass Lake, MI, USA). After incubation, a single colony was picked with a platinum loop and streaked onto an agar plate. Following incubation at 37°C for 5 d under anaerobic conditions, a number of single colonies were picked and suspended in BIM broth. After incubation at 37°C for 3 d under anaerobic conditions, the medium was removed by centrifugation (8000 g 10 min), and the resultant pellet was resuspended in BIM broth that contained 20 % glycerol medium and stored at − 80°C. In addition, aliquots of the pellet were used for isolation of DNA, followed by sequencing analysis as described below. Light microscopic observation following Gram staining was also performed.
Oral administration of Bifidobacterium pseudolongum
A frozen stock sample of isolated B. pseudolongum was thawed, washed twice with PBS and adjusted to 108 cells/ml with PBS before use. Mice were intragastrically administered 200 μl of the bacterial suspension (2 × 107 cells) daily throughout the experimental period. Before administration, viability of bacteria was checked by flow cytometry according to Ben-Amor et al. (Reference Ben-Amor, Heilig and Smidt27). In brief, approximately 106 cells were suspended in 1 ml anaerobic PBS containing 1 mm dithiothreitol, 0·01 % (w/v) Tween 20, 104 particles/ml fluorospheres (Flow-Check fluorospheres, Beckman Coulter, Tokyo Japan), 1 μg/ml propidium iodide (Wako Pure Chemical Industries, Osaka, Japan) and 5 nm SYTO-BC (Molecular Probes, Eugene, OR, USA). After 5 min incubation at room temperature, samples were analysed by flow cytometry (Epics XL, Beckman Coulter). The viability was routinely greater than 80 %.
Species-specific PCR and DNA sequencing for Bifidobacterium pseudolongum
B. pseudolongum subsp. pseudolongum (JCM1264) was obtained from the Japan Collection of Microorganisms of the Institute of Physical and Chemical Research, and used as a positive control. DNA was isolated from bacterial samples isolated from mouse faeces and B. pseudolongum subsp. pseudolongum (JCM1264) using PrepMan Ultra reagent (Applied Biosystems Japan, Tokyo, Japan), according to the manufacturer's instructions. Bacterial DNA samples were used as a template to amplify the 16S rRNA gene fragments with a primer pair specific to B. pseudolongum (forward: IDB41F, CCC TTT TTC CGG GTC CTG T; reverse: IDBC1R, ATC CGA ACT GAG ACC GGT T)(Reference Kwon, Yang and Lee28). PCR was performed in a reaction volume of 25 μl that contained 500 nm each of primers, 1 × PCR buffer, 0·2 mm dNTP and 1·25 U Taq-HS polymerase (Takara, Otsu, Japan). The reaction conditions were 96°C for 2 min, followed by 35 cycles at 94°C for 30 s, 63°C for 40 s and 72°C for 30 s, and a final extension at 72°C for 5 min. The size of the amplicon (471 bp) was checked by agarose gel electrophoresis. In addition, the nearly complete 16S rRNA gene was sequenced using a BigDye Terminator v3.0 cycle sequencing kit (Applied Biosystems Japan) according to the manufacturer's instructions with universal primers (27F, 5′ AGA GTT TGA TCM TGG CTC AG; 1492R, 5′ TAC GGH TAC CTT GTT ACG ACT T) as the sequencing primers(Reference Gurtler and Stanisich29). The sequence was compared to those in the GenBank database using the BLAST N algorithm.
Induction of contact hypersensitivity
CHS to DNFB was induced according to Nagai et al. (Reference Nagai, Matsuo and Hiyama30). Briefly, 25 μl of 0·15 % (v/v) DNFB (Tokyo Kasei, Tokyo, Japan) in acetone–olive oil (4:1, v/v), or the vehicle alone, were applied to each side of right and left ear auricles of mice, respectively, twice at 7-d intervals. Ear thickness was measured daily with a digital engineer's micrometer (Mitsutoyo, Kawasaki, Japan) from day 0 to day 10 and on day 14 after the first application. Ear thickness measurements were performed by an investigator who was blinded to the treatments of mice. DNFB-specific ear swelling was calculated according to the following equation:
Profile analysis of faecal microbiota by PCR-denaturing gradient gel electrophoresis
DNA was isolated from fresh faeces using a faecal DNA isolation kit (MO Bio Laboratories, Carlsbad, CA, USA) according to the manufacturer's instructions. DNA samples were used as a template to amplify the 16S rRNA gene fragments with universal primers, U968-GC (CGC CCG GGG CGC GCC CCG GGC GGG GCG GGG GCA CGG GGG GAA CGC GAA GAA CCT TAC) and L1401 (GCG TGT GTA CAA GAC CC)(Reference Zoetendal, Akkermans and De Vos31). U968-GC primer contains 40-nucleotide GC-rich sequence (GC clamp) at 5′ end. Denaturing-gradient gel electrophoresis (DGGE) analysis of the amplicon was performed as previously described(Reference Watanabe, Sasajima and Aramaki24). Quantity One software (version 4.6.0; Bio-Rad, Hercules, CA, USA) was used for band identification and normalisation of band patterns from DGGE gels. Subsequently, a dendrogram of the DGGE band profile was constructed using Pearson's curve-based correlation and the unweighted pair-group method using the arithmetic average clustering method in Quantity One software as previously described(Reference Wasaki, Rothe and Kania32).
Quantification of bifidobacteria in faeces
Populations of bifidobacteria in faecal samples were determined by conventional culture method. Briefly, the fresh samples were diluted tenfold with anaerobic phosphate buffer, and then 50 μl of each dilution were inoculated onto bifidobacteria-selective BIM agar(Reference Silvi, Rumney and Rowland26). Anaerobic incubation was performed at 37°C for 72 h using the AnaeroPack system (Mitsubishi gas, Tokyo, Japan). The number of colonies was counted after incubation and is represented as logarithm of colony-forming units.
Additionally, populations of bifidobacteria were quantified by real-time quantitative PCR (RT-qPCR) as previously described(Reference Watanabe, Sasajima and Aramaki24). Amplification and detection of faecal DNA were performed with the Thermal Cycler Dice Real Time System (Takara). Bifidobacterium genus-specific primer pairs (forward: TCG CGT C(C/T)G GTG TGA AAG; reverse: CCA CAT CCA GC(A/G) TCC AC)(Reference Rinttilä, Kassinen and Malinen33) and B. pseudolongum species-specific (IDB41F and IDBC1R as described above)(Reference Kwon, Yang and Lee28) primer pairs were used. RT-qPCR was performed in a reaction volume of 25 μl, containing 12·5 μl SYBR Premix Ex Taq (Takara), 200 nm each of the forward and reverse primers and 1 μl of the faecal DNA samples. The reaction conditions were: 95°C for 10 s, followed by forty cycles at 95°C for 5 s, 64°C for 30 s for the quantification of bifidobacteria; and 95°C for 10 s, followed by forty cycles at 95°C for 5 s, 60°C for 30 s for the quantification of B. pseudolongum. The fluorescent products were detected at the last step of each cycle. A melting curve analysis was performed after amplification to distinguish the targeted PCR product from the non-targeted PCR product. All samples were analysed in duplicate.
B. pseudolongum subsp. pseudolongum (JCM1264) was cultured in BIM broth, and the genomic DNA was extracted by PrepMan Ultra reagent according to the manufacturer's instructions. The 16S rRNA gene fragment was amplified by PCR with the B. pseudolongum species-specific primer pair (IDB41F and IDBC1R as described above)(Reference Kwon, Yang and Lee28). The amplicons were purified by QIAquick PCR purification kit (Qiagen, Tokyo, Japan) and cloned in pGEM-Easy T vectors (Promega, Madison, WI, USA). Transformation was performed with competent Escherichia coli XL-1 Blue cells, and the cells were plated onto Luria-Bertani agar plates supplemented with 25 μg/ml ampicillin, 30 μg/ml X-gal and 20 μg/ml IPTG, and incubated overnight at 37°C. White transformants were picked and grown in Luria-Bertani broth. Plasmid DNA was extracted with a QIAprep Spin Miniprep kit (Qiagen) and used as a standard for RT-qPCR.
Antibody ELISA
Serum levels of IgG1 and IgG2a specific to the hapten were determined by ELISA as previously described(Reference Watanabe, Sasajima and Aramaki24). In brief, each well of the ninety-six-well microtitre plates was coated with dinitrophenol-bovine serum albumin (Calbiochem, San Diego, CA, USA). After blocking the wells, serial dilutions of serum were added and then incubated for 2 h at room temperature. Horseradish peroxidase-conjugated rat anti-mouse IgG1 (clone LO-MG1-2, Zymed Laboratories, South San Francisco, CA, USA) or rat anti-mouse IgG2a (clone LO-MG2a-3, Zymed Laboratories) was added and incubated at 37°C for 2 h. Plates were developed at room temperature after the addition of o-phenylenediamine (0·4 mg/ml) and hydrogen peroxide (0·016 %).
Measurement of in vitro cytokine production in cervical lymph node cells
The excised CLN was gently homogenised with RPMI-1640 medium (GIBCO-BRL, Tokyo, Japan) supplemented with 2 % heat-inactivated fetal calf serum (GIBCO-BRL), 100 U/ml penicillin, 100 μg/ml streptomycin and 50 μg/ml gentamycin, and cells were filtered through nylon mesh (Nippon Rikagaku Kikai, Tokyo, Japan) and centrifuged for 5 min at 300 g. To remove erythrocytes, the cell pellet was resuspended in 10 mm 2-amino-2-(hydroxymethyl)propane-1,3-diol–HCl buffer (pH 7·4) supplemented with 0·24 m NH4Cl and incubated for 5 min at room temperature, followed by washing twice with RPMI-1640 medium supplemented with 2 % fetal calf serum. Thereafter, cells were counted by a Z1 coulter particle counter (Yamato Scientific, Tokyo, Japan), and an equal number of cells isolated from each mouse were pooled in each group. The cells (2 × 105 cells in 0·2 ml) were cultured in round-bottomed ninety-six-well plates (Greiner Bio-One, Tokyo, Japan) in RPMI-1640 medium supplemented with 10 % fetal calf serum and 500 μg/ml dinitrobenzene sulphonic acid sodium salt at 37°C in a 5 % CO2 atmosphere. Following 3-d-culture, the conditioned medium was collected and subjected to ELISA for cytokines. ELISA was performed in ninety-six-well microtitre plates. Wells were coated overnight at 4°C with primary antibodies (clones XMG1.2 and JES5-16E3 for IFN-γ and IL-10, respectively, eBioscience, San Diego, CA, USA). After blocking with RPMI-1640 medium supplemented with 10 % fetal calf serum at room temperature for 1 h, serial dilutions of conditioned medium, recombinant mouse interferon (IFN)-γ (eBioscience) and recombinant mouse IL-10 (eBioscience) were added and incubated for 2 h at room temperature. Thereafter, biotin-conjugated secondary antibodies (clones R4-6A2 and JES5-2A5 for IFN-γ and IL-10, respectively, eBioscience) were added and incubated for 1 h at room temperature. Streptavidin-horseradish peroxidase (Zymed Laboratories) was then added and incubated for 30 min at room temperature. Plates were developed at room temperature after the addition of 0·2 m citrate buffer (pH 4·0) containing 0·0123 mg/ml 3,3′,5,5′-tetramethylbenzidine and 0·01 044 % hydrogen peroxide. Finally, 1 m H2SO4 was added, and absorbance was measured at 450 nm with a microplate reader.
Statistical analysis
Results are presented as means with their standard errors. Tukey–Kramer's test following two-way ANOVA was used to compare mean values. Data analysis was performed with StatView for Macintosh (version 5.0, Statistical Analysis Systems Institute Inc., Spring Valley, CA, USA). P values less than 0·05 were considered statistically significant.
Results
Isolation of Bifidobacterium pseudolongum from mouse faeces
A Gram-positive obligate anaerobe, in the shape of a club or jelly bean, was isolated from the faeces of mice-fed FOS. DNA isolated from this bacterial sample and B. pseudolongum subsp. pseudolongum (JCM1264) were subjected to PCR with B. pseudolongum-specific primers. Fig. 1 shows the PCR amplicons separated on 3 % agarose gel. The size of the amplicons was consistent with the predicted size (471 bp). In addition, the nearly complete 16S rRNA gene sequence (1420 bp) exhibited 99 % similarity to B. pseudolongum subsp. pseudolongum (GenBank accession number, AY174106; Supplementary Fig. 1; the supplementary material for this article can be found at http://www.journals.cambridge.org/bjn). These findings indicate that the isolated bacterium is identical to B. pseudolongum.
Fig. 1 Representative ethidium bromide staining of PCR products separated on 3 % agarose gel. Lane 1, 100 bp ladder; lane 2, amplicon using DNA from isolated bacterium as template; lane 3, amplicon using DNA from Bifidobacterium pseudolongum subsp. pseudolongum (JCM 1246) as template; lane 4, negative control without template DNA.
Effect of fructo-oligosaccharide and Bifidobacterium pseudolongum on faecal microbiota
The diversity of faecal microbiota in mice was compared by PCR–DGGE analysis based on 16S rRNA gene sequences. A representative DGGE band profile is shown in Fig. 2(a). The intensity and position of detected bands were subjected to cluster analysis. The dendrogram shows two large clusters of FOS(+) and FOS( − ) groups (Fig. 2(b)). The results indicate that dietary supplementation with FOS modulated the composition of gut microbiota in mice, being consistent with our previous study(Reference Watanabe, Sasajima and Aramaki24). In contrast, BP( − ) and BP(+) groups showed no clear cluster. Thus, it appears that administration of B. pseudolongum had no influence on the global composition of gut microbiota in mice.
Fig. 2 PCR–denaturing-gradient gel electrophoresis (DGGE) analysis of faecal microbiota based on 16S rRNA gene sequences in BALB/c mice-fed fructo-oligosaccharide (FOS) and Bifidobacterium pseudolongum (BP). DGGE gel image stained with SYBR green is shown in chart (a) Similarities among DGGE band profiles of faecal bacteria of mice were calculated based on the position and intensity of bands, and the dendrogram of DGGE band profiles constructed by the unweighted pair-group method using the arithmetic average clustering method is shown in chart (b) Each lane in the gel and each line in the dendrogram represent an individual mouse. Distances are measured in arbitrary units.
Effect of fructo-oligosaccharide and Bifidobacterium pseudolongum on faecal bifidobacteria
Faecal bifidobacteria were quantified using a conventional cultivation method (Fig. 3(a)). Two-way ANOVA showed that both FOS (P = 0·0281) and B. pseudolongum (P = 0·0281) increased the faecal excretion of bifidobacteria. The faecal bifidobacterial levels were significantly lower in the FOS( − )BP( − ) group than in the other three groups. Only one out of the five mice in the FOS( − )BP( − ) group had a detectable level of bifidobacteria in the faeces. There were no significant differences among the other three groups. In addition, RT-qPCR with Bifidobacterium genus-specific primers showed that both FOS (P = 0·0016) and B. pseudolongum (P = 0·0016) increased the faecal excretion of bifidobacteria (Fig. 3(b)). The bifidobacterial levels were significantly lower in the FOS( − )BP( − ) group than in the other three groups, and the levels were significantly higher in the FOS(+)BP(+) group than in the FOS( − )BP(+) and FOS(+)BP( − ) groups. There was no significant difference between the FOS( − )BP(+) group and the FOS(+)BP( − ) group. Furthermore, faecal B. pseudolongum levels quantified by RT-qPCR with species-specific primers in each group were roughly comparable with the respective levels of bifidobacteria quantified by RT-qPCR with Bifidobacterium genus-specific primers (Fig. 3(c)).
Fig. 3 Populations of bifidobacteria in the faeces of BALB/c mice-fed fructo-oligosaccharide (FOS) and Bifidobacterium pseudolongum (BP). Bifidobacteria (charts (a) and (b)) and B. pseudolongum (chart (c)) were quantified by a conventional culture method (chart (a)) and real-time quantitative PCR (charts (b) and (c)). Each ● represents the value for an individual mouse, and horizontal bars represent mean values. In chart (a), P values estimated by two-way ANOVA were 0·0281, 0·0281 and 0·0618 for FOS, BP and their interaction, respectively. Similarly, P values in chart (b) were 0·0016, 0·0016 and 0·6903 for FOS, BP and their interaction, respectively. P values in chart (c) were 0·0002, 0·0017 and 0·4095 for FOS, BP and their interaction, respectively. Mean values with unlike letters were significantly different (P < 0·05), as estimated by Tukey–Kramer's test.
Effect of fructo-oligosaccharide and Bifidobacterium pseudolongum on contact hypersensitivity
Ear swelling began to be detected 5 d after the first application of DNFB (Fig. 4). At that time, two-way ANOVA showed that both FOS and B. pseudolongum significantly affected ear swelling (Table 1). Thus, ear swelling was significantly higher in the FOS( − )BP( − ) group than in the other three groups, and there were no significant differences among the three groups. From 6 d after the first application of DNFB to 3 d after the second application of DNFB, FOS, but not B. pseudolongum, affected ear swelling. On days 6 and 7 after the first application of DNFB, ear swelling was significantly lower in the FOS(+)BP(+) group than in the FOS( − )BP( − ) group, and the other two groups were intermediate. On days 1, 2 and 3 after the second application of DNFB, ear swelling was significantly lower in the FOS(+)BP(+) and FOS(+)BP( − ) groups than in the FOS( − )BP(+) and FOS( − )BP( − ) groups.
Fig. 4 Temporal changes in ear swelling after starting 2,4-dinitrofluorobenzene (DNFB) application in BALB/c mice-fed fructo-oligosaccharide (FOS) and Bifidobacterium pseudolongum (BP). Two weeks after feeding the indicated diet, DNFB was applied to the ear auricle twice at 7-d intervals (▲). Results are presented as means with their standard errors, n 5–6. Mean values with unlike letters at each time point were significantly different (P < 0·05) as estimated by Tukey–Kramer's test. Two-way ANOVA was summarised in Table 1. ●, FOS( − )BP( − ); ○, FOS( − )BP(+); □, FOS(+)BP( − ); □, FOS(+)BP(+).
Table 1 Summary of two-way ANOVA P values from an analysis of temporal changes in ear swelling, after starting 2,4-dinitrofluorobenzene application, in BALB/c mice-fed fructo-oligosaccharide (FOS), Bifidobacterium pseudolongum (BP), and FOS and BP
Effect of fructo-oligosaccharide and Bifidobacterium pseudolongum on serum antibody levels
There was no detectable hapten-specific antibody in the sera of mice without DNFB treatment (data not shown). At 25 d after the first application of DNFB, all the mice produced detectable levels of hapten-specific IgG1 and IgG2a antibodies. Two-way ANOVA showed that B. pseudolongum (P = 0·0247), but not FOS (P = 0·3871), increased hapten-specific IgG1 levels (Fig. 5(A)). In addition, FOS (P = 0·0002), but not B. pseudolongum (P = 0·7946), decreased hapten-specific IgG2a levels (Fig. 5(B)). The IgG2a levels were significantly lower in the FOS(+)BP( − ) and FOS(+)BP(+) groups than in the FOS( − )BP( − ) and FOS( − )BP(+) groups.
Fig. 5 Serum concentrations of hapten-specific IgG1 (chart A) and IgG2a (chart B) in BALB/c mice-fed fructo-oligosaccharide (FOS) and Bifidobacterium pseudolongum (BP). Each ● represents the value for an individual mouse, and horizontal bars represent mean values. In chart (A), P values estimated by two-way ANOVA were 0·3871, 0·0247 and 0·7194 for FOS, BP and their interaction, respectively. Similarly, P values in chart (B) were 0·0002, 0·7946 and 0·8758 for FOS, BP and their interaction, respectively. a,b Mean values with unlike letters were significantly different (P < 0·05; Tukey-Kramer's test).
Effect of fructo-oligosaccharide and Bifidobacterium pseudolongum on in vitro cytokine production in cervical lymph node cells
CLN cells isolated from each mouse were pooled in each group and restimulated in vitro with hapten for 3 d (Fig. 6). The concentration of IFN-γ in the conditioned medium tended to be higher in the FOS( − )BP( − ) group than in the other three groups. In contrast, the FOS( − )BP( − ) group showed an undetectable level of IL-10. The concentration of IL-10 tended to be increased by B. pseudolongum and was the highest in the FOS(+)BP(+) group.
Fig. 6 Cytokine production in cervical lymph node cells isolated from BALB/c mice-fed fructo-oligosaccharide (FOS) and Bifidobacterium pseudolongum (BP). Cells isolated from individual mice were pooled in each group and restimulated with hapten for 3 d in vitro. Concentrations of interferon (IFN)-γ (chart (a)) and IL-10 (chart (b)) in the conditioned medium were measured by ELISA.
Discussion
Our previous study demonstrated that dietary supplementation with FOS reduced the DNFB-induced CHS response in BALB/c mice(Reference Watanabe, Sasajima and Aramaki24). Additionally, analyses of gut microbiota raised the possibility that B. pseudolongum proliferation in the gastrointestinal tract was responsible for the inhibitory effect of dietary FOS on the DNFB-induced CHS response. These findings suggest that FOS is a prebiotic oligosaccharide that could prevent allergic contact dermatitis. Because we successfully isolated viable B. pseudolongum from the faeces of mice-fed FOS, the present study addressed whether this bacterium is responsible for the inhibitory effect of dietary FOS on DNFB-induced CHS response in mice.
In the present study, conventional culture-dependent analysis showed that dietary supplementation with FOS and intragastric administration of B. pseudolongum resulted in higher excretion of viable bifidobacteria in the faeces of mice. Likewise, molecular biological culture-independent analysis, i.e. RT-qPCR, showed that FOS and B. pseudolongum increased the faecal excretion of bifidobacteria, mainly B. pseudolongum. The data suggest that the mouse intestine was exposed to a higher population of viable B. pseudolongum by dietary supplementation with FOS and by oral administration of B. pseudolongum.
FOS(+)BP(+) mice showed reduced ear swelling in response to topical application of DNFB when compared with vehicle-administered mice-fed FOS( − ) diet (i.e. FOS( − )BP( − )). Although FOS(+)BP( − ) mice also showed a similarly reduced extent of ear swelling as FOS(+)BP(+) mice, the inhibitory action of B. pseudolongum administration without FOS feeding (i.e. FOS( − )BP(+)) was limited largely to the initial phase of the CHS response (i.e. on day 5 after the first application of DNFB). Because faecal excretion levels of bifidobacteria, mainly B. pseudolongum, in FOS( − )BP(+) group were comparable with those in FOS(+)BP( − ) group as described above, we suppose that B. pseudolongum proliferation in the gastrointestinal tract might be partially responsible for the inhibitory effect of dietary FOS on DNFB-induced CHS response in mice. In other words, the proliferation of other bacterial species than bifidobacteria in the intestinal tract of mice-fed FOS might also contribute to the reduction in DNFB-induced CHS response. Indeed, PCR–DGGE analysis of faecal 16S rRNA gene indicated that dietary supplementation with FOS, but not administration of B. pseudolongum, influenced the global composition of gut microbiota in mice. Therefore, it is of interest to search for the bacterial species or strains that prevent the CHS response. Chapat et al. (Reference Chapat, Chemin and Dubois34) demonstrated that oral administration of Lactobacillus casei DN-114 001, a strain used for the preparation of fermented milk, reduced the DNFB-induced CHS response in C57BL/6 mice.
Aberrant regulation between IFN-γ-producing Th1 cells and IL-4-producing Th2 cells is considered important in the development of immune diseases(Reference Abbas, Murphy and Sher35). For example, CHS responses are coupled with excessive Th1 cells and also IFN-γ-producing CD8+ (Tc1) cells(Reference Wang, Fujisawa and Zhuang36, Reference Kimber and Dearman37). In CHS responses, allergen re-exposure in the sensitised tissue results in the production of inflammatory cytokines by epidermal cells, which recruit hapten-specific T cells. Infiltrating T cells releases inflammatory cytokines such as IFN-γ that stimulate keratinocytes to produce chemokines. This in turn results in further recruitment of inflammatory cells, including monocytes and non-antigen-specific T cells. Ultimately, inflammation reaches its maximum, characterised by (epi)dermal infiltrates, oedema and spongiosis(Reference Toebak, Gibbs and Bruynzeel38). In order to elucidate whether CHS-reducing action of dietary FOS is mediated by altered response of T cells that produce inflammatory cytokines, the present study examined the IFN-γ production in CLN cells in vitro. Dietary FOS reduced the IFN-γ production in CLN cells in response to hapten restimulation. Because oral administration of B. pseudolongum also reduced the IFN-γ production, these results suggest that CHS-reducing action of dietary FOS is associated with reduced response of IFN-γ-producing T cells, which might be mediated, at least in part, by increased proliferation of B. pseudolongum in the gastrointestinal tract. In addition, serum levels of IgG2a (i.e. the prototypic antibody of Th1-driven B-cell responses in mice) and IgG1 (i.e. the prototypic antibody of Th2-driven B-cell responses in mice) were also measured as indicative parameters for Th1/Th2 balance. Dietary supplementation with FOS reduced Th1 response as indicated by reduced IgG2a levels, while administration of B. pseudolongum increased Th2 response as indicated by increased IgG1 levels. The data suggest that dietary FOS and administration of B. pseudolongum differently influence the Th1/Th2 balance. In other words, other bacterial species than bifidobacteria might contribute to the modulation of cytokine production. Furthermore, the present results are consistent with Chapat et al. (Reference Chapat, Chemin and Dubois34) who reported that hapten-specific IgG2a levels in sera and IFN-γ production in splenic CD8+T cells in vitro were lower in mice-fed L. casei DN-114 001. These findings suggest that probiotics and prebiotics exert immunoregulatory actions influencing hapten-specific T cells in DNFB-induced CHS response in mice.
de Waard et al. (Reference de Waard, Garssen and Snel39, Reference de Waard, Claassen and Bokken40) reported that administration of L. casei Shirota YIT9029 up-regulated Th1-type immune responses including IgG2b (i.e. the prototypic antibody of Th1-driven B-cell responses in rats) production and delayed-type hypersensitivity reaction in response to infection with Trichinella spiralis and Listeria monocytogenes in rats and mice. Additionally, Vos et al. (Reference Vos, Haarman and Buco41) reported that dietary supplementation with a mixture of galacto-oligosaccharide and FOS promoted Th1-type immune responses in a murine vaccination model. In terms of Th1/Th2 balance, these studies are contradictory to Chapat et al. (Reference Chapat, Chemin and Dubois34) and the present results as described above. Thus, probiotics and prebiotics may have both adjuvant properties on immune defence, via up-regulating Th1 responses, and anti-inflammatory actions that prevent CHS responses, via down-regulating Th1 responses. Furthermore, the direction of the immune responses may depend on the bacterial strains relevant to the actions.
However, more recent studies have proposed novel Th cell subsets, i.e. IL-17-producing Th cells (Th17 cells) and regulatory T cells, as key players in the regulation of immune responses and the development of immune diseases(Reference Oboki, Ohno and Saito42). In particular, Nakae et al. (Reference Nakae, Komiyama and Nambu43) demonstrated that IL-17-producing T cells rather than IFN-γ-producing Th1 cells are important for the development of CHS in mice. Additionally, He et al. (Reference He, Wu and Kim44) showed that IL-17-producing CD8+ (Tc17) cells, but not IFN-γ-producing Tc1 cells, are a central player in effector functions at the elicitation of CHS responses in mice. In the present study, however, CLN cells produced undetectable levels of IL-17 in response to hapten stimulation in vitro (data not shown). In contrast, IL-10 is a cytokine produced by regulatory T cells and reportedly associated with termination of CHS responses by suppressing the response of inflammatory cytokines such as IFN-γ(Reference Kimber and Dearman37). The present study showed a higher production of IL-10 in CLN cells isolated from FOS( − )BP(+) and FOS(+)BP(+) mice, suggesting that CHS-reducing action of dietary FOS is associated with increased production of IL-10, which might be mediated, at least in part, by increased proliferation of B. pseudolongum in the gastrointestinal tract. Thus, modulation of antigen-induced production of inflammatory and regulatory cytokines might be involved in the inhibitory effect of gut microbiota on CHS response, although further studies remain to be performed, especially regarding IL-17, a possible key player in CHS responses.
In conclusion, we propose that B. pseudolongum proliferation in the gastrointestinal tract is partially responsible for the reduction in DNFB-induced CHS response by dietary supplementation with FOS in mice.
Acknowledgements
The present study was partly supported by Special Coordination Funds for Promoting Science and Technology and a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan (no. 19380070). N. S. performed the isolation of B. pseudolongum, animal experiments, cell culture experiments and ELISA. R. F., T. O., N. T. and J. W. were involved in the molecular biological analyses of gut microbiota. N. S. and K. S. were involved in designing the study and writing the manuscript.
None of the authors has a conflict of interest.
