Species Specific Identification of Nine Human Bifidobacterium spp. in Feces

System. Appl. Microbiol. 25, 536–543 (2002) © Urban & Fischer Verlag http://www.urbanfischer.de/journals/sam

Species Specific Identification of Nine Human Bifidobacterium spp. in Feces Jacques-Edouard Germond, Olivia Mamin, and Beat Mollet Nestlé Research Center, Nestec Ltd., Vers-chez-les-Blanc, Lausanne, Switzerland Received: August 22, 2002

Summary Based on the 16S rDNA sequences, species specific primers were designed for the rapid identification by DNA amplification of nine human Bifidobacterium spp., namely B. adolescentis, B. angulatum, B. bifidum, B. breve, B. catenulatum, B. dentium, B. infantis, B. longum, B. pseudocatenulatum. B. lactis currently included in dairy products was added to the series. The primers were designed to target different positions of the 16S rDNA, allowing the simultaneous identification of these ten species of Bifidobacterium using two mixtures of primers. The identification procedure described in this paper was validated by establishing a correlation with an AluI restriction pattern of the different full length amplified 16S rDNA. This multiple primer DNA amplification technique was applied for the identification of pure colonies of Bifidobacterium spp. or directly from total bacteria recovered from human fecal samples. The technique was shown to be useful to detect dominant species and, when primers were used in separate reactions, underrepresented species could be identified as well. Key words: Human Bifidobacterium spp. – Species identification – Multiplex PCR amplification – feces

Introduction Bifidobacterium spp. are gram-positive, non-motile bacteria belonging to the complex community of microorganisms living in the intestinal tract of mammals where they have been attributed different health-promoting properties such as prevention of diarrhea and intestinal infections [18]. Furthermore, they have been shown to play an important role in modulation of the host immune system, e.g. an alleviation of atopic eczema and allergy [1, 9]. Bifidobacteria belong to the predominant species of the human colonic microbiota. They colonize the neonatal intestine as from the first week after birth at least in breastfed babies [8]. In elderly people, the endogenous bifidobacteria population was shown to decrease [10]. A natural way to manage the bifidobacteria population in the gut is by nutrition. Therefore, food products have been developed containing bifidobacteria strains found naturally in the intestinal tract of animal and humans, as e.g. B. animalis, B. longum, B. bifidum and B. infantis species [2, 3]. The growing consumer awareness for healthier foods has prompted the dairy industry to screen bacterial culture collections and human intestinal isolates for new Bifidobacterium strains with potentially even better health beneficial effects [21].

Therefore, rapid and accurate identification of Bifidobacterium spp. is important. This would greatly facilitate the monitoring of such species in complex ecosystems and would allow the identification of probiotic strains included in food products after their consumption and passage through the gastro-intestinal tract [16]. Direct identification from food products such as yogurt, cheese or infant formula may also become an important quality issue. Several methods were developed for molecular typing of Bifidobacterium spp. Nowadays, they mostly rely on amplification and comparison of the 16S rDNA sequences, which remain one of the best available methods for the identification of bacterial species. DNA amplification techniques using two genus specific 16S rDNA derived primers allowed detection of Bifidobacterium spp. in infant feces [13]. Using species specific primers hybridizing in the same region of the 16S rDNA, several isolates of human Bifidobacterium spp. could be discriminated [15]. The use of primers targeting different regions of the 16S rDNA led to simultaneous detection of several isolates of Bifidobacterium spp. (7). But among the five primers reported for identification purpose, the sequence for the B. infantis primer does not match exactly to the 0723-2020/02/25/04-536 $ 15.00/0

Species Specific Identification of Bifidobacteria

16S rDNA sequence of the type strain of this species (accession number D86184). Also, the primer based on the B. longum sequence would recognize B. catenulatum and B. dentium. Furthermore, all the techniques reported so far are limited to the identification of single colonies of Bifidobacterium spp., except for the last one [7] which was applicable to a complex mixture of bacteria, but not when applied to fecal samples. In this study, a novel series of species specific primers were designed that extend the number of human Bifidobacterium spp. to nine, which can be easily and reliably identified from pure cultures. Identification was performed using two mixtures of primers. In addition, the described method allows also the identification of Bifidobacterium spp. directly from biological samples such as feces.

Materials and Methods Bacterial strains and growth conditions B. lactis and 33 strains belonging to nine human Bifidobacterium species used in this study are presented in Table 1. Bacteria originated from the ATCC and DSMZ collections, or the Nestlé Culture Collection (NCC). The strains were grown in Brain Heart Infusion (BHI) (Oxoid, Basel, Switzerland) containing 0.05% cysteine under anaerobic conditions in jars containing the anaerobic system AnaeroGen™ (Oxoid) at 37 °C. Fecal samples Fecal samples were collected from seven human volunteers (subjects I to VII) differing in age (21 to 50 years) and sex (four women and three men). Samples were collected and directly processed or frozen at –80 °C until analysis. They were resuspended in oxygen free 0.9% NaCl containing 0.05% cysteine and Bifidobacterium spp. were selectively grown under anaerobic conditions on agar plate containing 40% canned tomato juice (Campbell’s or Libby’s), 4.5% Eugon agar (BBL, Becton Dickinson, USA), 1% maltose, 0.5% agar and 5 µg/ml hemin (20). DNA preparation Chromosomal DNA was purified from liquid cultures by a modification of the spooling method [5]. Cells were grown for 24 to 48 h, centrifuged and resuspended in 1/4 to 1/3 of the original culture volume in SuTE buffer (10 mM Tris pH 8, 5 mM EDTA, 25% sucrose) containing 1 mg/ml lysozyme (Serva, Heidelberg, Germany) and 0.025 mg/ml mutanolysin (Sigma, Buchs, Switzerland). After 30 min of incubation at 37 °C with shaking (200 rpm), the cells were centrifuged, resuspended in the same volume of TE (10 mM Tris pH8, 10 mM EDTA) and carefully lysed by addition of SDS at a final concentration of 1%, directly followed by gentle mixing in order to obtain a clear and homogeneous viscous mass. After complete lysis, RNAse was added to a final concentration of 50 µg/ml. After incubation for 30 min at 37 °C, proteinase K was added to a final concentration of 200 µg/ml and mixture was further incubated for 60 min at 55 °C. After addition of sodium acetate (final concentration of 0.3 M) and one volume of isopropanol, precipitated DNA was spooled on a toothpick, washed in 70% ethanol, drained and resuspended in TE (10 mM Tris pH 8, 1 mM EDTA). Nucleic acids were isolated from feces by using the QIAmp®DNA Stool Mini Kit (Qiagen GmbH, Germany) according to the supplier’s instructions or by disruption of the bacterial cells using glass beads. Fecal samples were diluted ten times in 0.9% NaCl, centrifuged and washed a second

537

time in the same buffer. DNA was extracted from the bacterial pellets as described elsewhere [22]. Cells were disrupted using a Mini-Beadbeater™ (Biospec, USA) in the presence of 50% of the volume of glass beads (106 µm, Sigma, St. Louis, Mo, USA). DNA amplification and analysis All primers used in this study were custom synthesized (Microsynth GmbH, Balgach, Switzerland) and are listed in Table 2. Bifidobacterium genus specific DNA amplification was obtained using the 16S rDNA targeted primers Bif164-f and Bif662-r [13], which produced a 520 bp amplification product. Amplification of the entire 16S rDNA (for AluI restriction) was performed using primers P0-f and P6-r located at the ends of the 16S rDNA gene [6]. Bifidobacterium species specific DNA amplifications were obtained using the 16S rDNA targeted forward primers designed in this study and lm3 as reverse primer [12]. DNA was amplified by PCR in a final volume of 25 µl containing: 50–200 ng DNA, 200 µM of each dNTP (Amersham Pharmacia biotech, Dübendorf, Switzerland), 0.4 to 2.0 µM primers (Table 2), 1.5 mM MgCl2 and 0.25 U Taq polymerase Platinium (Life Technologies, Basel, Switzerland) in standard reaction buffer. Prior to amplification reactions, the polymerase was heat activated (5 min at 95 °C). Then, the amplification proceeded for 30 cycles of 30 sec at 94 °C, 30 sec at 63 °C or 65 °C (for Table 1. List of Bifidobacterium spp. used in this study. NCC nbr

origin

Name

341 294 318 365 376 230 243 251 286 314 435 293 335 344 481 381 268 390 420 453 446 327 370 419 466 2715 2750 2562 291 312 325 280

ATCC15697T NCC NCC ATCC17930 ATCC25962 ATCC15703T ATCC15706 ATCC15704 NCC NCC ATCC15707T NCC NCC NCC NCC ATCC29521T NCC ATCC15696 DSM20215 NCC ATCC15700T NCC NCC NCC NCC ATCC27535T ATCC27539T ATCC27919T NCC NCC NCC ATCC27534T

B. infantis B. infantis B. infantis B. infantis B. infantis B. adolescentis B. adolescentis B. adolescentis B. adolescentis B. adolescentis B. longum B. longum B. longum B. longum B. longum B. bifidum B. bifidum B. bifidum B. bifidum B. bifidum B. breve B. breve B. breve B. breve B. breve B. angulatum B. catenulatum B. pseudo catenulatum B. pseudo catenulatum B. pseudo catenulatum B. pseudo catenulatum B. dentium

362

ATCC 27536

B. animalis (B. lactis: DSM10140T)

T

: Type strain, NCC: Nestlé Culture Collection

538

J.-E. Germond et al.

primer Mix1 or 2, respectively) and 2 min at 72 °C in a Perkin Elmer thermocycler 9700 (Perkin Elmer, Foster City, USA). Amplification products obtained in the same way, using primers located at the ends of the 16S rDNA (P0 and P6, Table 2), were digested with AluI (Roche, Rotkreutz, Switzerland) according to the manufacturer’s conditions. Amplification and digestion products were analyzed by agarose gel (1%) electrophoresis in TBE (100 mM Tris-HCl, 100 mM boric acid and 2 mM EDTA). After staining with ethidium bromide, the banding patterns were recorded with a Kodak digital camera (Kodak, New York, USA). Sequence alignments of the 16S rDNA were analyzed with ClustalW (ver. 1.8) and the phylogenic tree was generated using the boostrap method. Nucleotide sequence accession numbers The sequence of the 16S rDNA of B. catenulatun was deposited in the GenBank database and has been assigned the accession number AF432082.

Results and Discussion Design of species-specific primers for human Bifidobacterium spp. The complete sequences of the Bifidobacterium spp. 16S rDNA available from GenBank were aligned. The deduced evolutionary tree showed that the Bifidobacterium spp. of human origin are clustered in two major groups (Fig. 1). One group includes B. breve, B. infantis and B. longum; the second one comprises B. angulatum, B. pseudocatenulatum, B. catenulatum, B. adolescentis, and B. dentium. B. bifidum does not really cluster with the above-mentioned species, which is in accordance with its physiological characteristics [17]. The 16S rDNA genes were used as target for species identification in combination with DNA amplification.

The forward primer was designed for each of the nine human Bifidobacterium spp. and B. lactis (DSM10140T) which is included in several dairy products (Table 2). The hybridization sites of these primers were carefully distributed along the 16S rDNA sequence in order to obtain species and size specific amplicons for the different human Bifidobacterium spp. The location of the different primers relative to the 16S rDNA sequence is reported in Fig. 2A. The overall high sequence similarity between the different 16S rDNA and the constraint imposed by the position of the primers along the sequence left only very few nucleotide differences available for primer design. In some cases, there was only one nucleotide difference between one species and the others. But, such a single base pair difference, when located as last nucleotide at the 3′-end of the primer, was sufficient to lead to a specific amplification product. In addition, special care was taken to avoid that the designed primers would hybridize to the 16S rDNA of other Bifidobacterium spp. retrieved from the GenBank database. The sequence of the primers designed for the nine human Bifidobacterium spp. showed identity with only few of the 23 available 16S rDNA sequences of Bifidobacterium spp. isolated from animals or sewage. Indeed, the primers designed for B. infantis and B. bifidum could hybridize with the B. subtile (a species found in sewage) 16S rDNA. Similarly, the B. longum primer would target B. minimum (sewage), while the B. dentium primer would react with two dental caries isolates, B. denticoleus and B. inopinatum, or with B. cuniculi found in rabbits. Validation of the primers The different primers were distributed in two groups in order to identify unknown species in two steps. The first mixture (Mix1) contained the species specific primers

Table 2. Bifidobacterium genus and species specific primers on 16S rDNA used in this study. Species Bifidobacterium genusd Bifidobacterium genusd B. breve B. angulatum B. longum B. adolescentis B. bifidum B. infantis B. catenulatun and B. pseudocatenulatum B. catenulatun B. dentium B. pseudocatenulatum B. lactis Universal primere Universal primerf Universal primerf a

code

Primera

Sequence

posb

Conc. (µM)c

Bbr Bag Blo Bad Bbi Bin

Bif164 f Bif662 r B787 f G009 f G027 f G028 f G003 f B791 f

5′GGGTGGTAATGCCGGATG 5′CCACCGTTACACCGGGAA 5′GATGCGACAGTGCGAGC 5′CGTGTTGCCAGCACATG 5′GACATGTTCCCGACGGT 5′GGGACCATTCCACGGTC 5′AAGGGCTCGTAGGCGGC 5′TATCGGGGAGCAAGCGT

168 649 1245 1110 983 823 561 443

0.4-m1 0.8-m1 0.8-m1 0.4-m1 0.8-m1 1.2-m1

B798 f G055 f B790 f G020 f G022 f lm3 r P0 f P6 r

5′CGGATGCTCCGACTCCT 5′AAGTCGAACGGGATCAG 5′CATCGCTTAACGGTGGG 5′GACAGCCGTAGAGATAT 5′TGGCCGGTACAACGCGG 5′CGGGGTGCTGCCCACTTTCATG 5′GAGAGTTTGATCCTGGCTCAG 5′GTACGGCTACCTTGTTACGA

179 65 601 994 1230 1387 16 1473

Bc/p Bca Bde Bps Bla

: Code of the primer, f and r: forward and reverse primers : Position of the 3′end of the different primers on the 16S rDNA sequence (B. breve, GenBank AB006658) c : Concentration of the primers in the primer Mix1 and Mix2 (m1, m2) d : [13], e: [12], f: [6] b

0.8-m1 0.4-m2 0.4-m2 0.4-m2 0.8-m2 2.0-m1+m2

Species Specific Identification of Bifidobacteria

for the seven major human intestinal Bifidobacterium spp. (B. adolescentis, B. angulatum, B. bifidum, B. breve, B. infantis, B. longum and B. catenulatum/B. pseudocatenulatum). The second mixture (Mix2) comprised species specific primers for B. catenulatum, B. pseudocatenulatum, B. dentium and B. lactis (Fig. 2A, Table 2). In both mixtures, the reverse primer was the universal primer lm3 [12]. Chromosomal DNA was prepared from the nine type strains of human Bifidobacterium spp. and B. lactis (Table 1) and subjected to DNA amplification using primer Mix 1 and 2. To minimize the presence of aspecific bands, the multiplex PCR reaction conditions were optimized as to temperature, time of annealing and relative concentration of primers in each mixtures (Table 2). Agarose gel electrophoresis showed that a specific DNA fragment is obtained for the seven human intestinal Bifidobacterium spp. using primer Mix1 (Fig. 2B). Individual fragments ranged between 0.15 and 1.30 Kb and corresponded to the expected size as deduced from the 16S rDNA sequence. As expected B. catenulatum and B. pseudocatenulatum generated a fragment of same size. In addition, a non specific band was obtained for B. den-

539

tium and B. lactis, whose size was similar to that of B. bifidum. The use of primer Mix2 allowed the discrimination of B. catenulatum from B. pseudocatenulatum and also of B. dentium and B. lactis from B. bifidum. In addition, primers from Mix2 generated no amplification products for the six other gut Bifidobacterium spp. All nine human Bifidobacterium spp. could thus be distinguished by two simple DNA amplification reactions using the newly designed mixtures of primers. In order to validate our identification procedure, a series of amplifications using primer Mix1 were performed with DNA extracted from several human bifidobacteria strains from the Nestlé Culture Collection (NCC, Table 1). In each case, a single amplification product of the expected size was obtained for all strains belonging to a given species, including the type strain (Fig. 3). It is worth mentioning that the entire NCC collection of Bifidobacterium spp. has recently been reclassified using recognized identification techniques such as ribotyping [14] and Amplified Ribosomal DNA Restriction Analysis (ARDRA) [23] (unpublished data). Using these two genetic typing techniques, B. lactis was shown to exhibit identical ribotyping

Fig. 1. Unrooted tree including bootstrap values of the type strains of the Bifidobacterium spp. available from GenBank according to the sequence of their 16S rDNA (accession numbers of the sequences used are given for each species).

540

J.-E. Germond et al.

Fig. 2. Scheme of the Bifidobacterium spp. 16S rDNA sequence locating the genus specific primers (Bif164, Bif662), the species specific primers of Mix1 and Mix2 (underlined) and the universal primers (lm3, P0, P6). The size of the amplification products are indicated in base pairs (bold) (A). Agarose gel electrophoresis (negative picture) of the 16S rDNA amplified products using primer Mix1 or Mix2 and DNA extracted from the type strains of the different Bifidobacterium spp. (Bbr: B. breve, Bag: B. angulatum, Blo: B. longum, Bad: B. adolescentis, Bbi: B. bifidum, Bin: B. infantis, Bca: B. catenulatum, Bps: B. pseudocatenulatum, Bde: B. dentium, Bla: B. lactis). M: 100 bp DNA ladder (B).

and ARDRA patterns with those of B. animalis ATCC27536 isolated by Scardovi from chicken feces [19]. The similarity of these two species was already suggested by Cai et al [4]. Validation of species identification by ARDRA In order to validate our identification procedure, a correlation was established with the restriction pattern of the full length 16S rDNA. Chromosomal DNA from the different human Bifidobacterium spp. was subjected to DNA amplification using the 16S rDNA terminal universal primers (P0 and P6, Fig. 2A, Table 2, [6]). The amplification products were then digested with the restriction enzyme AluI (ARDRA) and analyzed by agarose gel electrophoresis (Fig. 4A). The patterns of DNA fragments obtained, ranging from 50 to 800 bp, were all shown to be different for the eight human intestinal Bifidobacterium spp, while B. catenulatum and B. pseudocatenulatum had the same pattern. The eight different patterns also perfectly matched the AluI restriction map of the 16S rDNA sequences from the different Bifidobacterium spp. type strains (Fig. 4B) used to design the species specific primers. ARDRA can be used as confirmation or alterna-

Fig. 3. Agarose gel electrophoresis (negative picture) of the 16S rDNA amplified products using primer Mix1 and DNA extracted from five strains of B. breve (Bbr), one strain of B. angulatum (Bag), five strains of B. longum (Blo), B. adolescentis (Bad), B. bifidum (Bbi), B. infantis (Bin) and four strains of B. pseudocatenulatum (Bps) (Table 1). M: 100 bp DNA ladder.

tive for the identification of the human Bifidobacterium spp., but is much more tedious and not applicable to large series of isolates generated in human or animal studies or to identify a single species in a mixture of Bifidobacterium spp.

Species Specific Identification of Bifidobacteria

541

Fig. 5. Agarose gel electrophoresis (negative picture) of the 16S rDNA amplified products using primer Mix1 and DNA extracted from Bifidobacterium spp. isolated as single colonies from human feces. One to three colonies were selected from five human subjects (I to V). M: 100 bp DNA ladder.

Fig. 4. Agarose gel electrophoresis (negative picture) of AluI cleaved 16S rDNA amplified products using primers P0 and P6 and DNA extracted from the type strains of the different Bifidobacterium spp. (Table 1 and Fig.1 for abbreviations), M: 100 bp DNA ladder (A). Schematic representation of the AluI restriction maps of the corresponding Bifidobacterium spp. 16S rDNA, indicating the position of the restriction sites (AluI sites) and the size of the different fragments in bp (bold)(B).

Fig. 6. Agarose gel electrophoresis (negative picture) of the 16S rDNA amplified products using DNA extracted from a fecal samples of two human subjects (VI and VII). DNA was amplified using either primer Mix1 or the individual primers specific for the different Bifidobacterium spp. (B787 for B. breve, G009 for B. angulatum, G027 for B. longum, G028 for B. adolescentis, G003 for B. bifidum, B791 for B. infantis, B798 for B. catenulatum / B. pseudocatenulatum, Table 2). M: 100 bp DNA ladder.

542

J.-E. Germond et al.

Identification of Bifidobacterium spp. in human feces In order to evaluate the efficacy of our technique, fecal samples of human volunteers were analyzed, starting first from single colonies isolated by plating on selective medium. Chromosomal DNA was extracted from each isolate and subjected to DNA amplification using Bifidobacterium genus specific primers (Bif164 and Bif662, Fig. 2A, Table 2). The generation of an amplification product of 520 bp indicates that the isolates belong to the Bifidobacterium genus [13]. In the next step, DNA amplification using chromosomal DNA and primer Mix1 and 2 determined the species of each isolate. The size of the amplification products obtained allowed an unambiguous identification of the isolates. Figure 5 illustrates examples of Bifidobacterium spp. isolated from five volunteers (I to V). One to three colonies were taken from each volunteer (1–3). B. adolescentis was identified in two volunteers (I and IV-2), B. bifidum in one (II), B. breve two times in the same volunteer (III-1 and III-3) and B. longum in three (III-2, IV-1 and V). These results show that when starting from a single colony, a unique DNA amplification was sufficient to determine the species. In order to verify if the designed primers would amplify other 16S rDNA, major representative of human intestinal bacteria were individually subjected to DNA amplification using primer Mix1 and 2. No products was generated from any of the bacteria tested (Lactobacillus acidophilus, L. johnsonii, L. casei, Clostridium perfringens, Bacteroides distasonis, Collinsella aerofaciens and some propionibacteria) using the two primer mixtures (data not shown). It indicates that at least these representative enteric bacteria would not interfere with our Bifidobacterium spp. identification, when applied to fecal samples. Fecal samples were then tested for the presence of bifidobacteria. Total fecal DNA was prepared from two selected volunteers (VI and VII) and subjected to DNA amplification using primer Mix1 (Fig. 6, Mix1). For subject VI, two amplification products were obtained which corresponded to B. longum and B. adolescentis. In order to determine if other Bifidobacterium spp. were present in the fecal sample, DNA amplification was performed using the different species specific primers in separate reactions. Under these conditions, B. bifidum was additionally identified (Fig. 6-VI, primer G003). This species was not detected with primer Mix1, due to its lower concentration in the feces (see below). For the other selected subject (VII), two major amplification products were obtained for B. longum and B. adolescentis and a weaker one for B. angulatum. Using the species specific primers in separate reactions revealed in addition specific products for B. bifidum and B. cat/pseudocatenulatum (Fig. 6VII, primers G003 and B798). A none specific product was obtained with primer B791. B. cat/pseudocatenulatum was further subjected to DNA amplification using the respective specific primers from primer Mix2 and identified as B. catenulatum (not shown). In order to confirm these results, 30 single colonies of Bifidobacterium

spp. were isolated by plating a fecal sample of volunteer VII on selective medium. Analysis of these colonies with primer Mix1 showed that they corresponded mostly to B. adolescentis (20 colonies) and in a lower proportion to B. longum (5 colonies), B. angulatum (4 colonies) and B. bifidum (1 colony). No B. catenulatum was found. Notably, the relative number of colonies found for each species was in accordance with the amount of amplicons obtained for the same species using total fecal DNA and primer Mix1. In other human fecal samples, only one specific amplification product was obtained using the primer mixtures and all the different primers in separate reactions, indicating the presence of a single Bifidobacterium sp. In addition, it can be mentioned that the global intestinal microbiota did not interfere with the identification of Bifidobacterium spp, since no unexpected amplicons were observed. It can be concluded that primer Mix1 allows the detection of the most abundant Bifidobacterium spp. present in a fecal sample, whereas detection of all Bifidobacterium spp. requires the use of individual primers. Limits of the identification method The next step was to determine the concentration range of a given Bifidobacterium sp. that can be detected relative to a dominant species. DNA was extracted from a fecal sample, which contained B. longum and B. adolescentis at the same concentration (108 bacteria per gram of feces). This DNA sample was serially diluted with DNA extracted from a fecal sample containing only B. longum, at the same concentration. These DNA mixtures were then amplified using primer Mix1. An amplification product specific for B. adolescentis was detected even when it was 5 ×102 times less abundant than B. longum (data not shown). The limit of our technique in the detection of one Bifidobacterium sp. in feces was also determined. DNA from a fecal sample containing 5 ×108 B. longum per gram of feces was serially diluted and then amplified using the B. longum specific primers. An amplification product was still obtained at a dilution equivalent of 104 B. longum per gram of feces. Even though about 102 bacteria per gram of feces can be detected by plating on selective media (11), this is only working when a single Bifidobacterium sp. is present. On the contrary, our technique is independent of the presence of other Bifidobacterium spp. In a background of 1011–1012 bacteria per gram of intestinal content, concentrations below 104 bifidobacteria per gram should not have a real physiological relevance. In conclusion, this paper reports the design of species specific primers targeting different positions of the 16S rDNA, that allows simultaneous identification by DNA amplification of the known human Bifidobacterium spp. Identification can be performed either from single colonies or from feces. In fecal samples, underrepresented Bifidobacterium spp., which are difficult to isolate from the dominant ones by direct plating, can be clearly detected and identify.

Species Specific Identification of Bifidobacteria Acknowledgements We gratefully acknowledge Dr. Rainer Simmering and David Vilanova for helpful discussions, Dr. Annick Mercenier and Dr. Bruno Pot for critical reading of the manuscript and Dr. Elizabeth Prior for reviewing the manuscript.

12.

References

13.

1. Apostolou, E., Pelto, L., Kirjavainen, P. V., Isolauri, E., Salminen, S. J., & Gibson, G. R.: Differences in the gut bacterial flora of healthy and milk- hypersensitive adults, as measured by fluorescence in situ hybridization. FEMS Immunol.Med.Microbiol. 30, 217–221 (2001). 2. Ballongue, J.: Bifidobacteria and probiotic action. In Lactic acid bacteria, microbiology and functional aspects, eds. Salminen, S. & von Wright, A., pp. 519–587. Marcel Dekkers, Inc, New York, N.Y (1998). 3. Ballongue, J., Grill, J., & Baratte-Euloge, P.: Action sur la flore intestinale de laits fermentés au Bifidobacterium. Lait 73, 249–256 (1993). 4. Cai, Y., Matsumoto, M., & Benno, Y.: Bifidobacterium lactis Meile et al. 1997 is a subjective synonym of Bifidobacterium animalis (Mitsuoka 1969) Scardovi and Trovatelli 1974. Microbiol. Immunol. 44, 815–820 (2000). 5. Delley, M., Mollet, B., & Hottinger, H.: DNA probe for Lactobacillus delbrueckii. Appl. Environ. Microbiol. 56, 1967–1970 (1990). 6. Di Cello, F. & Fani, R.: A molecular strategy for the study of natural bacterial communities by PCR-based techniques. Minerva Biotecnologica 8, 126–134 (1996). 7. Dong, X., Cheng, G., & Jian, W.: Simultaneous identification of five Bifidobacterium species isolated from human beings using multiple PCR primers. Systematic and Applied Microbiology 23, 386–390 (2000). 8. Harmsen, H. J., Wildeboer-Veloo, A. C., Raangs, G. C., Wagendorp, A. A., Klijn, N., Bindels, J. G., & Welling, G. W.: Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr. Gastroenterol. Nutr. 30, 61–67 (2000). 9. He, F., Ouwehand, A. C., Isolauri, E., Hashimoto, H., Benno, Y., & Salminen, S.: Comparison of mucosal adhesion and species identification of bifidobacteria isolated from healthy and allergic infants. FEMS Immunol. Med. Microbiol. 30, 43–47 (2001a). 10. He, F., Ouwehand, A. C., Isolauri, E., Hosoda, M., Benno, Y., & Salminen, S.: Differences in composition and mucosal adhesion of bifidobacteria isolated from healthy adults and healthy seniors. Curr. Microbiol. 43, 351–354 (2001b). 11. Hengstler, A., Kronenberg, M. U., & Fahr, A. M.: A reliable method for the removal of PCR inhibitors during the prepa-

14. 15.

16. 17. 18.

19.

20. 21. 22.

23.

543

ration of bacterial DNA from stool specimens. Clin Lab 44, 447–450 (1998). Kaufmann, P., Pfefferkorn, A., Teuber, M., & Meile, L.: Identification and quantification of Bifidobacterium species isolated from food with genus-specific 16S rRNA-targeted probes by colony hybridization and PCR. Appl. Environ. Microbiol. 63, 1268–1273 (1997). Kok, R. G., de Waal, A., Schut, F., Welling, G. W., Weenk, G., & Hellingwerf, K. J.: Specific detection and analysis of a probiotic Bifidobacterium strain in infant feces. Appl. Environ. Microbiol. 62, 3668–3672 (1996). Mangin, I., Bourget, N., & Decaris, B.: Ribosomal DNA polymorphism in the genus Bifidobacterium. Res. Microbiol. 147, 183–192 (1996). Matsuki, T., Watanabe, K., Tanaka, R., & Oyaizu, H.: Rapid identification of human intestinal bifidobacteria by 16S rRNA-targeted species- and group-specific primers. FEMS Microbiol. Lett. 167, 113–121 (1998). O’Sullivan, D. & Kullen, M. J.: Tracking of probiotic bifidobacteria in the intestine. Int.Dairy J. 8, 513–525 (1998). Roissart, H. & Luquet, F. M.: Bactéries Lactiques I, Lorica (Uriage, France) (1994). Saavedra, J. M., Bauman, N. A., Oung, I., Perman, J. A., & Yolken, R. H.: Feeding of Bifidobacterium bifidum and Streptococcus thermophilus to infants in hospital for prevention of diarrhoea and shedding of rotavirus. Lancet 344, 1046–1049 (1994). Scardovi, V. & Trovatelli, L. D.: Bifidobacterium animalis (Mitsulka) comb. nov. and the “minimum” and “subtile” groups of new bifidobacteria found in sewage. Int. J Syst. Evol. Microbiol. 24, 21–28 (2002). Sutter, V., Citron, D., & Finegold, S.: Wadsworth Anaerobic Bacteriology Manual. Mosby Ed. St Louis (1980). Tamime, A. Y., Marshall, V. M., & Robinson, R. K.: Microbiological and technological aspects of milks fermented by bifidobacteria. J Dairy Res. 62, 151–187 (1995). Ventura, M., Elli, M., Reniero, R., & Zink, R.: Molecular microbial analysis of Bifidobacterium isolates from different environments by the species-specific amplified ribosomal DNA restriction analysis (ARDRA). FEMS Microbiol. Ecol. 36, 113–121 (2001). Vincent, D., Roy, D., Mondou, F., & Dery, C.: Characterization of bifidobacteria by random DNA amplification. Int. J Food Microbiol. 43, 185–193 (1998).

Corresponding author: J.-E. Germond, Nestlé Research Center, Vers-chez-les-Blanc, P.O. Box 44, CH-1000 Lausanne 26, Switzerland Tel.: ++41-21-785 88 35; Fax: ++41-21-785 89 25; e-mail: [email protected]