OUP user menu

Rapid identification of human intestinal bifidobacteria by 16S rRNA-targeted species- and group-specific primers

Takahiro Matsuki, Koichi Watanabe, Ryuichiro Tanaka, Hiroshi Oyaizu
DOI: http://dx.doi.org/10.1111/j.1574-6968.1998.tb13216.x 113-121 First published online: 1 October 1998


On the basis of 16S rRNA sequences, species- and group-specific primers for Bifidobacterium adolescentis, B. angulatum, B. bifidum, B. breve, the B. catenulatum group (B. catenulatum and B. pseudocatenulatum), and the B. longum group (B. longum and B. infantis), which are species commonly found in human intestinal tracts, were developed. The specificity of these primers was confirmed through the use of DNA extracted from 46 strains of 31 Bifidobacterium species, as well as 14 non-bifidobacterial species that are the predominant bacteria in the human intestinal tract. The present species-specific primers were applied to the identification of 43 isolated strains, consisting of six strains of B. adolescentis, eight of the B. catenulatum group, seven of B. bifidum, nine of B. breve, and 13 of the B. longum group.

Key words
  • 16S rRNA
  • Polymerase chain reaction primer
  • Identification
  • Bifidobacterium
  • Human intestinal flora

1 Introduction

The genus Bifidobacterium includes Gram-positive pleomorphic strict anaerobes, and certain species are predominant inhabitants of human intestines. The beneficial effects of bifidobacteria on human health have been demonstrated with regard to immunopotentiation, nutrition, the prevention of intestinal infections, and the reduction of intestinal putrefaction [1, 2].

Analyses of bifidobacteria in intestinal microflora require labor-intensive and time-consuming techniques such as the single-colony isolation of candidate isolates from a specific selective medium, followed by testing for multiple physiological and biochemical traits. Moreover, these tests do not always provide clear-cut results, and are sometimes unreliable. Therefore, there is a need for practical techniques that enable rapid and accurate analysis of intestinal bifidobacteria.

For some years now, the comparison of 16S rRNA sequences has attracted attention as a reliable method for the classification and identification of several bacterial species [3]. 16S rRNA-targeted hybridization probes or PCR primers enable rapid and specific detection of a wide range of bacterial species, and have become key procedures in the detection of microorganisms [48]. Wang et al. [4] developed species-specific primers based on 16S rRNA gene sequences for 12 anaerobic bacteria that are predominant in the human intestinal tract, and showed that the PCR procedure could be a powerful tool in the quantitative detection of these species in human feces. Kok et al. [5] prepared 16S rRNA-targeted primers for a probiotic Bifidobacterium strain and for the genus Bifidobacterium, and then attempted to monitor human fecal flora using these primers. Langendijk et al. [6] and Kaufmann et al. [7] developed genus-specific probes for Bifidobacterium, and applied them to quantitative fluorescence in situ hybridization or colony hybridization. Yamamoto et al. [8] developed species-specific probes for human intestinal bifidobacteria such as B. adolescentis, B. bifidum, B. breve, B. longum, and B. infantis, and these have become useful tools in the identification of isolated strains of bifidobacteria. However, species-specific primers for most human intestinal bifidobacteria have not yet been developed, apart from those for B. adolescentis and B. longum reported by Wang et al. [4].

The present report describes the development of 16S rRNA-targeted species- and group-specific primers for human intestinal bifidobacteria, including their application to the identification of Bifidobacterium strains isolated from human fecal samples.

2 Materials and methods

2.1 Bacterial strains and culture conditions

Strains listed in Table 1 were obtained from the American Type Culture Collection (ATCC), Japan Collection of Microorganisms (JCM), National Collection of Food Bacteria (NCFB), National Collection of Type Culture (NCTC), National Institute of Biosciences and Human-Technology (FERM), or the Yakult Central Institute for Microbiological Research Tokyo (YIT). The strains were cultured in GAM broth (Nissui Seiyaku, Tokyo, Japan) supplemented with 1% glucose under O2-free CO2 at 37°C overnight, except for E. coli, which was cultured aerobically in Tripticase Soy Broth (Difco, Detroit, MI, USA) at 37°C overnight.

View this table:
Table 1

List of bacterial strains and the results of PCR assays using species- and group-specific primer, BiADO, BiANG, BiBIF, BiBRE, BiCATg, and BiLONg

SpeciesStraina16S rRNA sequence accession no.Species-specific primers
B. adolescentisATCC 15703T[M58729]+
B. adolescentisNCFB 2229+
B. adolescentisNCFB 2230+
B. adolescentisNCFB 2231+
B. angulatumATCC 27535T[D86182]d+
B. angulatumJCM 1252+
B. bifidumATCC 29521T[S83624]+
B. bifidumATCC 15696+
B. bifidumATCC 11863+
B. bifidumStrain Yakultc+
B. breveATCC 15700T[AB006658]d+
B. breveATCC 15698+
B. breveStrain Yakultc+
B. catenulatumATCC 27539T[M58732]+
B. catenulatumJCM 7130+
B. pseudocatenulatumJCM 1200T[D86187]d+
B. longumATCC 15707T[M58739]+
B. longumATCC 15708+
B. longumFERM P-6548+
B. infantisATCC 15697T[D86184]d+
B. infantisATCC 15702+
B. infantisATCC 25962+
B. suisATCC 27533T[M58743]+
B. animalisATCC 25527T[D86185]d
B. asteroidesATCC 25910T[M58730]
B. boumJCM 1211T[D86190]d
B. choerinunATCC 27686T[D86186]d
B. coryneformeATCC 25911T[M58733]
B. cunniculiATCC 27916T[M58734]
B. denticolensDSM 10105T[D89331]d
B. dentiumATCC 27534T[D86183]d
B. gallicumJCM 8224T[D86189]d
B. gallinarumJCM 6291T[D86191]d
B. indicumATCC 25912T[D86188]d
B. inopinatumDSM 10107T[D86332]d
B. lactisDSM 10140T[X89513]
B. magnumJCM 1218T[D86193]d
B. merycicumJCM 8219T[D86192]d
B. minimumATCC 27538T[M58741]
B. globosum bATCC 25864T[D86194]d
B. pseudolongum bJCM 1205T[D86195]d
B. pullorumJCM 1214T[D86196]d
B. ruminantiumJCM 8222T[D86197]d+
B. saeculareDSM 6531T[D89328]d
B. subtileDSM 20096T[D89379]d
B. thermophilumATCC 25866T[U10151]

A total of 43 strains of Bifidobacterium listed in Table 3 were isolated in the following manner. Fecal samples of infants and adults were collected anaerobically, out of which approximately 0.5 g (wet weight) was transferred to a tared tube containing 4.5 ml of a prereduced anaerobically sterilized dilution fluid, flushed with O2-free CO2, and serially diluted with vigorous shaking. The 107, 108, and 109 dilutions were cultured in roll tubes of selective media for bifidobacteria [9] for four to six days at 37°C, and suspected colonies of Bifidobacterium were picked in succession into GAM broth.

View this table:
Table 3

Identification of isolated strains of Bifidobacterium through the use of species- and group-specific primers

StrainsSpecies-specific primerPhenotypic traitsDNA-DNA
MC-36, 37, 38, 39, 40, 41+B. adolescentis or similar speciesaB. adolescentis
MC-1, 2, 3, 4, 31, 32, 33+B. bifidumB. bifidum
MC-5, 6, 14, 15, 16, 18, 20, 21, 22+B. breveB. breve
MC-42, 43, 44, 45, 46, 47,48, 49+B. adolescentis or similar speciesaB. pseudocatenulatum
MC-10, 11, 12, 23, 24, 25, 26, 27, 28, 9, 30+B. longumB. longum
MC-8, 9+B. infantisB. infantis
  • a B. adolescentis or phenotypically similar Bifidobacterium species: B. adolescentis, B. angulatum, B. catenulatum, or B. pseudocatenulatum.

  • bDNA-DNA homologies of isolates to the respective reference strain were over 70%.

2.2 Preparation of template DNA

The genomic DNA of the strains used was extracted by heating at 94°C for 15 min in 50 μl of TE buffer, or by the method described by Zhu et al. [10]. Briefly, bacterial cells were suspended in extraction mixture (450 μl) consisting of 250 μl of extraction buffer (100 mM Tris-HCl, 40 mM EDTA, pH 9.0), 50 μl of 10% SDS, and 150 μl of benzyl chloride. Following incubation at 50°C for 30 min with shaking, the DNA was obtained by isopropanol precipitation.

2.3 Development of 16S rRNA-targeted species-specific primers

At present, 31 different species of bifidobacteria have been identified. 16S rRNA sequences of all species are available in DDBJ/GenBank/EMBL databases (Table 1). Based on a multiple aliment of 16S rRNA sequences of 31 Bifidobacterium created using the program Clustal W [11], potential primer target sites for species- or group-specific detection were identified in the variable regions V2 and V3 (Fig. 1). We then designed 16S rRNA targeted primers for B. adolescentis, B. angulatum, B. bifidum, B. breve, the B. catenulatum group (B. catenulatum and B. pseudocatenulatum), and the B. longum group (B. longum and B. infantis), which were selected because they are species commonly found in human intestinal microflora (Table 2). These primers were commercially synthesized by Rikaken (Tokyo, Japan).

Figure 1

Partial 16S rRNA sequences of human intestinal bifidobacteria. The target regions for species-specific primers are underlined. Numbering corresponds to the structure model of E. coli 16S rRNA [12]. A: V2 region for forward primers; B: V3 region for reverse primers.

View this table:
Table 2

Bifidobacterium species- and group-specific primers based on 16S rRNA sequences

Name of primers SequenceaLengthTarget sitebProduct sizeAimed human intestinal bifidobacteria
BiADO-1CTCCAGTTGGATGCATGTC19182–200279B. adolescentis
BiANG-1CAGTCCATCGCATGGTGGT19185–203275B. angulatum
BiCATg-1CGGATGCTCCGACTCCT17176–192289B. catenulatum
BiCATg-2CGAAGGCTTGCTCCCGAT18476–442B. pseudocatenulatum
BiLONg-2TCSCGCTTGCTCCCCGAT18478–441B. infantis
  • aS=G:C=1:1.

  • bNumbering corresponds to the structure model of E. coli 16S rRNA [12].

2.4 PCR amplification

The PCR reaction mixture (25 μl) was composed of 50 mM Tris-HCl, pH 8.8; 2.5 mM MgCl2; 15 mM (NH4)2SO4; 0.45% Triton X-100; 200 μM of each dNTP; 0.25 μM of each species-specific primer; 10 ng of DNA of 1 μl of the heated bacterial cells described above; and 0.9 U of Taq DNA polymerase (Biotech International, Australia). The PCR was carried out in a Touchdown Thermal Cycler (Hybaid, Middlesex, UK). The amplification program consisted of one cycle of 94°C for 5 min, then 35 cycles of 94°C for 20 s, 55°C for 20 s, and 72°C for 30 s, and finally one cycle of 72°C for 5 min. Amplification products were subjected to gel electrophoresis in 1% agarose, followed by ethidium bromide staining.

2.5 Physiological and biochemical traits

To confirm whether the isolated strains belonged to the genus Bifidobacterium, cellular morphology, Gram-staining, and the fermentation products of glucose were analyzed. Carbohydrate fermentation patterns were tested using the API 50 CHL system (La Balme les Grottes, France), in accordance with the instructions provided by the manufacturers.

2.6 DNA-DNA homology testing

DNA-DNA homology tests were carried out in accordance with the method described by Ezaki et al. [13]. The following strains were used as references: B. adolescentis ATCC 15703T, B. angulatum ATCC 27535T, B. bifidum ATCC 29521T, B. breve ATCC 15700T, B. catenulatum ATCC 27539T, B. pseudocatenulatum JCM 1200T, B. gallicum JCM 8224T, B. longum ATCC 15707T, and B. infantis ATCC 15697T.

3 Results

3.1 The specificity of primers

The specificity of the primers was confirmed by PCR using both chromosomal DNA extracted from 46 strains belonging to 31 Bifidobacterium species, and that from 14 non-Bifidobacterium species that are commonly found in human intestinal microflora (Table 1). The BiBIF, BiBRE, and BiCATg primers were able to detect the target species specifically, providing PCR products with the expected size (Fig. 2). However, the BiADO primers cross-reacted with B. ruminantium, and BiLONg showed a positive PCR result for DNA of B. suis.

Figure 2

PCR product for the eight species of Bifidobacterium with their specific primers. Lanes: M: DNA size markers (sizes are shown to the left); 1: B. adolescentis ATCC 15703T; 2: B. angulatum ATCC 27535T; 3: B. bifidum ATCC 29521T; 4: B. breve ATCC 15700T; 5: B. catenulatum ATCC 27539T; 6: B. pseudocatenulatum JCM 1200T; 7: B. longum ATCC 157071T; 8: B. infantis ATCC 15697T; 9: negative control (PCR result of BiADO with E. coli ATCC 11775T).

3.2 Identification of isolated strains of Bifidobacterium through the use of species-specific primers

The species-specific PCR technique was used to identify Bifidobacterium strains. DNA extraction procedures, both by heating and by the method described by Zhu et al. [10], worked well in identifying the isolated strains (Table 3). Forty-three isolates were clearly identified as six strains of B. adolescentis (MC-36, 37, 38, 39, 40, 41), eight of the B. catenulatum group (MC-42, 43, 44, 45, 46, 47, 48, 49), seven of B. bifidum (MC-1, 2, 3, 4, 31, 32, 33), nine of B. breve (MC-5, 6, 14, 15, 16, 18, 20, 21, 22), and 13 of the B. longum group (MC-10, 11, 12, 23, 24, 25, 26, 27, 28, 29, 30, 8, 9).

3.3 Phenotypic characteristics of isolated strains

The carbohydrate fermentation patterns of the 43 strains were analyzed, and eight strains in the B. catenulatum group were found not to differ significantly from those of B. adolescentis (Table 3). The fermentation patterns of the 14 strains identified as B. adolescentis or members of the B. catenulatum group by species-specific primers, which are shown in Table 4, indicate that it is difficult to distinguish between B. adolescentis and the B. catenulatum group.

View this table:
Table 4

Phenotypical characteristics of isolated strains of Bifidobacterium

StrainsCarbohydrate fermentation patternPrimers
  • Symbols: +, positive; −, negative; w, weakly positive; d, doubt.

On the other hand, the phenotypic characteristics of 29 strains of B. bifidum, B. breve, and the B. longum group (B. longum and B. infantis) showed the same results as those obtained by the species-specific PCR technique (Table 3).

3.4 DNA-DNA hybridization test

The DNA-DNA hybridization test was performed on the 43 isolates. The results were the same as those obtained by the species-specific PCR technique (Table 3).

4 Discussion

In the present study, we developed useful 16S rRNA-targeted species- and group-specific primers for B. adolescentis, B. angulatum, B. bifidum, B. breve, the B. catenulatum group (B. catenulatum and B. pseudocatenulatum), and the B. longum group (B. longum and B. infantis). These primers cover most of the bifidobacterial species commonly found in the human intestinal tract, except for the recently discovered but uncommon species B. gallicum[14]. They are therefore an excellent means for identifying Bifidobacterium isolated from the human intestinal tract.

Yamamoto et al. [8] reported species-specific probes for some human intestinal bifidobacteria, but not for B. catenulatum, B. pseudocatenulatum, or B. angulatum. Furthermore, the identification of isolates through the use of probes is more labor-intensive than when species-specific primers are used. Using species-specific primers developed by Wang et al. [4], it was difficult to identify the isolated strains of the genus Bifidobacterium. This is due to the fact that they do not cover all of the human intestinal bifidobacteria, and that the specificity of their primers was not tested in the other Bifidobacterium species.

Although the BiCATg primers (Table 2) could not be used to distinguish between B. catenulatum and B. pseudocatenulatum we consider that those two species should be treated as members of the B. catenulatum group, because they share the same murein type, ferment a similar range of carbohydrates, have a closely related DNA-DNA homology of 70–76%[15], and the similarity of the 16S rRNA sequences was over 99%[16]. Furthermore, the BiLONg primers did not distinguish between B. longum, B. infantis, and B. suis, which should be treated as members of the B. longum group due to the following characteristics: unique murein type, a DNA-DNA homology of 62–72%[15], and a similarity of the 16S rRNA sequences of over 99%[16]. The BiADO primers developed for B. adolescentis cross-reacted with B. ruminantium, which is isolated from the rumen of cattle, because B. ruminantium has the same sequence as B. adolescentis at the primer target position (data not shown). It was reported that the DNA-DNA homology between B. adolescentis and B. ruminantium is relatively high at 40–72%[17] and the similarity of the 16S rRNA sequence was 99%[16], which indicates that they are closely related species and that B. ruminantium can be included in the B. adolescentis group. As B. ruminantium has not been detected in the human intestinal tract, the BiADO primers are useful for identifying isolates from the human intestinal tract.

Testing the phenotypic characteristics of isolated strains of Bifidobacterium is labor-intensive and time-consuming, and these tests did not always produce clear-cut results, particularly those on strains of B. adolescentis and the B. catenulatum group (Tables 3 and 4). It was also suggested that strains of B. catenulatum and B. pseudocatenulatum exhibited fermentation patterns similar to those of B. adolescentis[18, 19]. A sorbitol-positive fermentation pattern could not be a key characteristic for distinguishing B. pseudocatenulatum from B. adolescentis and B. catenulatum, as it was reported that B. adolescentis biovar a, c, and B. catenulatum also ferment sorbitol [18, 19]. Mannose-positive and starch-positive fermentation patterns could be key characteristics for distinguishing between B. pseudocatenulatum and B. catenulatum, but not for distinguishing between B. pseudocatenulatum and B. adolescentis[20, 21]. Therefore, strains of B. pseudocatenulatum were, on a phenotypic basis, hardly distinguishable from B. adolescentis.

The species- and group-specific PCR technique, which was applied to the identification of isolated strains of Bifidobacterium, yielded the same results as the DNA-DNA hybridization test. The PCR method always gave clear-cut results, even for the species that could not be clearly identified based on normal phenotypic traits (Tables 3 and 4). In addition, it is possible to identify the isolate of bifidobacteria within three hours by the PCR technique. Therefore, the species-specific PCR technique that we developed is a useful identification method, since it is rapid, convenient, accurate, and cost-effective.


We would like to thank Dr. T. Mitsuoka, Professor Emeritus University of Tokyo, for his valuable advice. This work was supported by the Yakult Bio-Science Foundation (Tokyo, Japan).


View Abstract