Japanese cedar pollinosis (JCPsis), an immunoglobulin E (IgE)-mediated type I allergy caused by exposure to Japanese cedar (
Cryptomeria japonica) pollen (JCP), represents a public health issue affecting over 16% of the Japanese population (
4). In clinical studies evaluating the effects of a probiotic strain,
Bifidobacterium longum BB536, on JCPsis, we found that administration of
B. longum BB536 significantly alleviated some subjective symptoms and affected blood markers in individuals with JCPsis (
20-
22). Furthermore, we observed fluctuations in the
Bacteroides fragilis group among individuals with JCPsis in the pollen season, with administration of
B. longum BB536 suppressing these fluctuations (
12,
13). The genus
Bacteroides is known as one of the predominant intestinal bacteria in humans. The
Bacteroides fragilis group has been suggested to be associated with allergic disease in several clinical studies (
5,
8,
16). However, the taxonomy of the genus
Bacteroides has undergone significant changes in the past few years (
15), owing to the redefinition of the genus
Bacteroides and the application of molecular biological techniques leading to the identification of several novel species (
1-
3,
6,
9). Little has been determined regarding the distributions of these bacteria in human fecal microbiota.
In the present study, we designed 14 specific primer pairs to detect species in the Bacteroides fragilis group that have been isolated from and identified in human feces and investigated distributions of each species for individuals with JCPsis and those without JCPsis by real-time PCR, to evaluate possible associations with JCPsis.
Clinical study.
Samples came from a clinical study reported by Xiao et al. (
21) evaluating the effects of
B. longum BB536 on clinical symptoms of JCPsis and blood parameters. Briefly, a total of 44 adults with JCPsis were randomized to ingest either
B. longum BB536 powder (BB536 group; 13 men and 9 women; mean age, 36.0 ±7.3 years) or placebo powder (placebo group; 13 men and 9 women; mean age, 36.5 ±8.1 years), in a randomized, double-blinded design during the pollen season (20 January to 21 April). Fourteen healthy adults who were JCP specific, IgE negative, and without prior history of spring allergic rhinitis (healthy group, 11 men and 3 women; mean age, 33.4 ± 7.6 years) were administered placebo powder during the same intervention period in an identical manner to JCPsis subjects. Participants were instructed to collect specimens in a plastic tube, cool the bag immediately to <10°C, and deliver the sample within 12 h. Collected specimens were stored at −80°C until analysis.
Design and specificity of primer pairs.
DNA extraction from fecal samples was performed as described previously (
13). Fourteen 16S rRNA gene-targeted species-specific primers (Table
1) were designed and checked for specificity according to previous reports (
10,
14). The amplification program consisted of 94°C for 10 s, followed by 35 cycles of 94°C for 5 s and 60°C for 30 s. Melting curves were obtained by heating from 60°C to 95°C in increments of 0.2°C/s, with continuous fluorescence collection. DNA extracts from the type strains listed in Table
1 were used as standards for the determination of the cell number of each species. The specificity of each primer pair was then tested using DNA extracts from all strains listed in Table
1, with the addition of
Parabacteroides distasonis JCM 5825
T,
Parabacteroides merdae JCM 9497
T,
Prevotella intermedia JCM 12248
T, and
Porphyromonas gingivalis JCM 8525, 12257
T. Each specific primer yielded positive PCR results for the corresponding target bacterium and negative PCR results for nontarget microorganisms.
Distribution of each species of the Bacteroides fragilis group in fecal samples.
Analyses were conducted on fecal samples collected from individuals that completed the study before (20 January, before the sample intake) and after (21 April) the pollen season. We observed some different distributions of the
Bacteroides fragilis group for the JCPsis group compared with the healthy group before the pollen season (Table
2). In particular, cell numbers of
Bacteroides fragilis and
Bacteroides intestinalis were significantly higher in the JCPsis group than in the non-JCPsis group (Table
2).
Compared to the pre-pollen season, totals of nine, six, and two species of the
Bacteroides fragilis group were increased significantly after the pollen season in the placebo, BB536 and healthy groups, respectively (Table
3). Among these, when taking notice of those species that increased among the JCPsis subjects, it was found that
Bacteroides caccae, Bacteroides vulgatus, Bacteroides fragilis, and
Bacteroides intestinalis were significantly increased only in the placebo group.
Comparing cell numbers after pollen season, significant intergroup differences were found for Bacteroides fragilis and Bacteroides intestinalis between the placebo and healthy groups and significant intergroup differences were found for Bacteroides fragilis between the placebo and BB536 groups.
Conclusions.
We observed that cell numbers of Bacteroides fragilis and Bacteroides intestinalis were significantly higher in the JCPsis group than in the healthy group before and after the pollen season. Furthermore, significant positive correlations were found between the cell numbers of these two species with composite symptom scores and JCP-specific IgE. Our data suggest that prevalence of Bacteroides fragilis and Bacteroides intestinalis might represent risk factors for JCPsis. In addition, no significant change was observed in cell numbers of Bacteroides fragilis or Bacteroides intestinalis in the BB536 group, suggesting that intake of B. longum BB536 may play a role in stabilizing the microbiota, which might in turn exert suppressive effects on sensitization to pollen and/or symptom development.
Increased prevalence of the
Bacteroides fragilis group has been observed in individuals with allergic diseases or under stress conditions (
5,
7,
8,
16,
17). In vitro studies have demonstrated that
Bacteroides fragilis perturbed host immunity (
11,
12,
18,
19). These lines of evidence implied an exacerbating effect of the
Bacteroides fragilis group on allergic disorders. To the best of our knowledge, this is the first report to outline a possible association between the species of the
Bacteroides fragilis group and allergic diseases, although we cannot deny that there might be some biases in the cell numbers of each species since they have only been determined by the quantitative PCR method. Nevertheless, the results from the intra/intergroup differences should have not been influenced greatly. Further studies are needed to confirm these results, especially for
Bacteroides intestinalis since the prevalence was relatively low.