- Email: [email protected]
Bifidobacteria identification based on 16S rRNA and pyruvate kinase partial gene sequence analysis
Anaerobe 8 (2002) 341–344
Taxonomy/systematics
Bifidobacteria identification based on 16S rRNA and pyruvate kinase partial gene sequence analysis$ Laurence Vaugiena, Fabien Prevotsb, Christine Roquesa,* a
Laboratoire de Bact!eriologie, Virologie et Microbiologie Industrielle, Facult!e de Pharmacie, 35 Chemin des Mara#ıchers, 31062 Toulouse cedex 4, France b Degussa, Ferments d’aromatisation, 35, Chemin des Mara#ıchers, 31062 Toulouse cedex 4, France Received 4 October 2002; received in revised form 31 January 2003; accepted 5 February 2003
Abstract The lack of a simple and rapid identification system for Bifidobacterium species makes them difficult to use in industrial applications. To obtain valuable discriminating factor, we studied different strains, and human isolates by two molecular taxonomy methods. First method was based on chrono-differentiation. A metabolic gene (pyruvate kinase) was chosen to be used as a systematic discriminating factor. A comparison of about 40 pyruvate kinase protein sequences allowed us to synthesize two oligonucleotides that were able to amplify a fragment of this corresponding gene in our strains. Based on these partial pyruvate kinase gene sequences, several clusters could be identified. The second method used in this study was based on 16S rRNA sequences analysis. We compared sequences present in GenBank database, and this allowed to separate bifidobacteria species into different clusters. They were different from those obtained with partial pyruvate kinase gene sequences analysis. So, by combining both methods, we were able to identify our isolates, when only 10% of them could be strictly identified using the 16S rRNA method. Moreover, pyruvate kinase analysis allowed to differentiate very ambivalent groups such as B. animalis/B. lactis or B. infantis/ B. longum, but created different clusters for B. infantis species group, questioning on the homogeneity of this species. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Bifidobacterium; Pyruvate kinase; Phylogeny; 16S rRNA; Probiotic
1. Introduction Bifidobacterium represents the third most common genus of the human intestine microflora after the genera Bacteroides and Eubacterium (up to 3% of the total fecal microflora of adults) and is considered involved in host’s health [1]. The ever increasing use of specialised strains of bifidobacteria in the treatment and prevention of gastrointestinal disorders requires careful attention to strain identification, even at the species level. To date, differentiation of bifidobacteria species has mainly been performed on the basis of phenotypic characteristics [2,3], expressing poor discriminative power and reproducibility. Molecular approaches have been more recently used to analyse and quantify the composition $ Paper from Anaerobe Olympiad 2002. The 6th Biennial Congress of the Anaerobe Society of the Americas, Park City, Utah, 29 June–2 July 2002. *Corresponding author. Tel.: +33-5-62-25-68-60; fax: +33-5-61-2595-72. E-mail address: [email protected] (C. Roques).
of bacterial groups within human gut microflora [4,5]. Despite of this, in 1999, Tannock [6] underlined that molecular methods for the comprehensive identification of bifidobacteria species are not yet available. Bifidobacteria can be differentiated from morphological similar bacteria by the use of genus-specific PCR primers or oligonucleotide probes, but only DNA– DNA hybridization provides a reliable means of species identification. Since, the use of genus-specific PCR primers targeting the 16S rRNA gene, coupled with denaturing gradient gel electrophoresis (DGGE), has provided a screening method for the detection and identification of most of bacterial species [7,12]. In a recent study, most of human species could be identified by a coupled method. Moreover, not all of human species was included in this study, and so this method was not easily useful for human isolates. In order to develop a convenient method for rapid identification of human Bifidobacterium strains at the species level, we have obtained DNA sequences of a short region of the pyruvate kinase gene, specific for 11
1075-9964/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S1075-9964(03)00025-8
342
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
Bifidobacterium species which are often detected in human. To control the specificity of our sequences, Bifidobacterium species of non-human origin (2), and human isolates (15) were included in the study.
2. Materials and methods 2.1. DNA extraction The specific genomic DNA was prepared from 50 ml of bacterial culture on M17 medium supplemented with 0.4% glucose, 1% lactose and 1% ascorbic acid, in the stationary phase. Cells were harvested by centrifugation (4000 rpm, 5 min), suspended in 5 ml TES (Tris 100 mM pH 8, EDTA 20 mM, sucrose 20%) containing lyzosyme (10 mg ml 1) and incubated at 37 C for 2 h. Next, 50 ml of achromopeptidase (solution at 60 mg ml 1 —Sigma Aldrich, Saint Quentin Fallavier, France) were added. The suspension was then incubated for 2 h at 37 C. After the incubation, 115 ml of a 20% SDS solution were added and further incubated at 65 C during 30 min, and then put on ice for 5 min. After phenol/chloroform (50/ 50 v/v) extraction, DNA was precipitated in absolute alcohol. The precipitate was dried in a centrifugal evaporator and resuspended in water at a concentration of 100 ng ml 1. 2.2. Pyruvate gene partial amplification Forty sequences of pyruvate kinase protein, obtained from the EMBL sequence database, were compared using the Omiga program (Omiga 1.1, Oxford Molecular). GenBank accession numbers were as follows: CAB74228, BAA98307, AAF39440, AAF40552, BAA95686, T35759, B75251, B72406, G69685, D64925, S73859, I40840, JC4219, H64223, JC4220, C64130, S29783, B40620, S29790, S27330, AAF30593, BAA89788, BAA75222, AAC71435, AAC66733, AAC45776, AAC28104, CAB08894, AAC23216, AAC12962, AAB66498, BAA06727, BAA0625, BAA02406, CAA50527, S77240, BAA10621, CAB70653, AAF25804, AAC67927. These data included Gram-positive, Gram-negative, and mycoplasma organisms. Two relatively conserved sequences were translated into DNA. The two corresponding oligonucleotides (TAATCGATGATGGCAAGGTCCG and TAATACGCGGGCTACCATGATGCC—LifeTechnologies Laboratories, Cergy Pontoise, France) were used for amplification. Amplification was performed as follows: 94 C—3 min, 45 C—1 min, 72 C—1 min for 1 cycle and 94 C—1 min, 45 C—1 min, 72 C—1 min for 29 cycles. PCR products obtained were run on 2% agarose gel (w/v) in TAE 1X for 1 h 30 min at 70 mV and made visible by ethidium bromide staining and UV transillumination. The PCR products length was
about 300 bp. They were purified using the Promega extraction Kit, quantified on agarose gel, and then sent for sequencing (Genome Express S.A., Grenoble, France). All sequences obtained were analysed using the Omiga program (Omiga 1.1, Oxford Molecular). Partial pyruvate kinase gene sequences correspond to the following GenBank accession numbers: AF384974 (B. breve CIP 64.69 T), AF384975 (B. catenulatum CIP 104175), AF384976 (B. infantis CIP 64.67 T), AF384977 (B. lactis CIP 105265 T), AF384978 (B. longum CIP 64.62 T), AF384979 (B. adolescentis CIP 64.59 T), AF384980 (B. angulatum CIP 104167), AF384981 (B. animalis DSM 20104), AF384982 (B. bifidum CIP 56.7 T), AF384983 (B. longum CIP 64.63), AF384984 (B. pseudolongum CIP 105352), AF384985 (B. thermophilum CIP 105420 T), AF384986 (B. infantis DSM 20090), AF384987 (B. infantis DSM 20223). 2.3. Obtention of 16S rDNA sequence In order to obtain 16S rRNA partial gene amplification (900 bp fragment length) and sequencing, universal oligonucleotides were used (ATAATGCGGCCGCACGGGCGGTGTGTRC and TAATAGCGGCCGCAGCMGCCGCGGTAATWC). These primers correspond to universal sites at about 1390 and 520 bp, respectively, of 16S rRNA sequences. The annealing mixture was prepared by combining 0.5 ml Taq DNA Polymerase (2.5 units) (Boehringer Mannheim), 5 ml PCR buffer containing dNTP (Boehringer Mannheim buffer supplemented with each dNTP at 10mM), 5 ml of each oligonucleotide solution (60 mg ml 1), 1 ml of template DNA, and 33.5 ml H2O. Amplification was performed as described before. PCR products were desalted and digested by 1 U of NotI for 2 h at 37 C. Then digestions were resolved by 1% gel agarose and visualised by ethidium bromide staining. The fragments were purified using a Promega extraction kit (Promega Corp., Madison, USA), and cloned in pGEM 5Z vector (Promega Corp.). Escherichia coli was electroporated at 25 mF and 2.5 kV (Gene pulser II, Bio-Rad, Ivry/ Seine, France). Plasmids were purified using QIAprep spin plasmid kit (Qiagen S.A., Hilden, Germany), and sent for an insert sequencing (Genome Express S.A., Grenoble, France). Amplification was performed on human isolates (except strains 17, 41, and 47).
3. Results and discussion The amplifications were performed on collection strains including the type strain of the 11 species tested (Table 1), and 15 faecal human isolates. According to this method, the collection strains could be separated into nine clusters (Table 2). Except strains 08, 14, 17, 41, and 47 all the studied isolates belonged to a specific
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
cluster. These five strains shared the same partial PK gene sequence, and could be grouped in a 10th unidentified cluster. Moreover, further experiments need Table 1 Collection strains tested Strain
Referencea
B. B. B. B. B. B. B. B. B. B. B. B. B. B. B.
CIP 64.59T CIP 104167T CIP 105419T DSM 20104 CIP 56.7T CIP64.69T CIP 104175T CIP 64.67T DSM 20090 DSM 20223 CIP 105265T CIP 64.62T CIP 64.63 CIP 105352T CIP 105420T
adolescentis angulatum animalis animalis bifidum breve catenulatum infantis infantis infantis lactis longum longum pseudolongum thermophilum a
CIP Pasteur Institute Collection DSM:Deutsche Sammlung von Mikrooganismen.
343
to be performed on the two Bifidobacterium species isolated from humans not yet included in our study (B. dentium and B. pseudocatenulatum). This would lead us to confirm their differentiation among the PK clusters. Despite of this, none of the 15 tested human isolates, belonged to B. adolescentis, B. angulatum, B. breve, or B. catenulatum species. The PK gene sequence analysis led to a sufficient discrimination at species level of human strains for four different clusters (cluster 4—B. longum, cluster 5—B. angulatum; cluster 6—B. breve; cluster 7—B. catenulatum). The specificity is indicated by the differentiation, without cross-reaction, of B. animalis (cluster 3) and B. thermophilum (cluster 9). Although the several B. infantis collection strains studied were classified in three different clusters (clusters 1, 2 and 8), they did not share the same sequence with the B. longum strains studied (cluster 4). So, partial PK gene sequence analysis allowed us to differentiate B. infantis from B. longum and also B. animalis from B. lactis. After 16S rDNA amplification, sequences were analysed using the FastA program. These data did not lead to a bifidobacteria identification at the species level
Table 2 Consensus grouping of bifidobacteria based on 16S rRNA and partial pyruvate kinase DNA sequences PK Analysis
16S rRNA analysis Cluster A B. adolescentis B. angulatum B. catenulatum B. dentium B. pseudocatenulatum
Cluster 1 B. bifidum CIP56.7T B. lactis CIP105265T B. infantis CIP 64.67T B. pseudolongum CIP105352 Cluster 2 B. adolescentis CIP64.59T B. infantis DSM 20090 Cluster 3 B. animalis CIP105419T B. animalis DSM20104 Cluster 4 B. longum CIP64.62T B. longum 64.63 Cluster 5 B. angulatum Cluster 6 B. breve CIP64.69T Cluster 7 B. catenulatum CIP104175T Cluster 8 B. infantis DSM 20223 Cluster 9 B. thermophilum CIP105420T a
Industrial strains.
B. adolescentis
Cluster B B. breve B. infantis B. longum B. pseudolongum
Cluster C B. animalis B. lactis
B. infantis B. pseudolongum 05, 56, 82
B. lactis 39, 63, 65a, 73a
Cluster D B. thermophilum
Cluster E B. bifidum
B. bifidum 02, 20
B. infantis
B. animalis 11 B. longum 29, 69a, 80 B. angulatum B. breve B. catenulatum B. infantis B. thermophilum
344
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
but only to a differentiation in clusters as already described by several teams [8,9]. They are in agreement with our PK clusters differentiation. Moreover, the combination of the two methods allow us to confirm the identification of the tested human isolates at the species level. In contrast to many genera, the bifidobacteria form a monophyletic cluster on the basis of rRNA sequences. This led to the development of genus-specific 16S rRNA probes [10] but did not allow all Bifidobacterium species to be differentiated. There are many recently developed approaches for the molecular fingerprinting of bifidobacteria, i.e. amplified ribosomal DNA restriction analysis [5], analyses of conserved gene sequences [11, 12] or triplicate arbitrarily primed-PCR procedure [13]. Such molecular methods have frequently a discriminatory power with strains differentiation for one species. But, on the other hand, differentiation of even closely related strains, like B. longum/B. infantis, and B. animalis/B. lactis remains confused [8,9]. Using oligonucleotides probes for RNA–DNA hybridisation, Yamamoto et al. [14] were able to detect and identify B. adolescentis, B. bifidum, B. breve, B. infantis and B. longum. Despite of this, probes for B. bifidum and B. infantis cross-reacted with a few strains of non-human origin. In the same way, in a recent report, Requena et al. [12] identified bifidobacteria by PCR amplification of a 301bp transaldolase gene sequence and DGGE. Under such conditions, the authors observed two subtypes of B. longum, five subtypes of B. adolescentis and two subtypes of B. pseudocatenulatum. On the other hand, B. angulatum and B. catenulatum type cultures could not be differentiated from each other demonstrating the difficulty to obtain accurate identification tool. To complete our study, experiments were also performed on three industrial strains (Table 2). The first one is confirmed as B. longum. The two others preliminary given as B. animalis were identified as B. lactis. Herreman et al. [15] have also identified and characterised three industrial strains by combining two molecular taxonomy methods (DNA–DNA hybridisation and DNA restriction XbaI fragment polymorphism). One strain is confirmed as B. longum. The two others presumed to belong to B. infantis and B. bifidum species, were identified as B. animalis. These observations underline the difficulty to develop acurate method for bifidobacteria identification. In this study, we designed partial pyruvate kinase gene analysis as a reliable, rapid and accurate manner for the identification of human intestinal bifidobacteria. Although complementary data are necessary (B. dentium
and B. pseudocatenulatum) to definitely purpose this method combined with 16S rRNA sequence analysis for bifidobacteria differentiation.
References [1] Shir A, Gramet G, Suau A, Rochet V, Pochart P, Dore J. Quantification of bacterial groups within the human fecal flora by oligonucleotide probe hybridization. Appl Environ Microbiol 2000;66:2263–6. [2] Mitsuoka T. Taxonomy and ecology of bifidobacteria. Bifidobacteria Microflora 1984;3:11–28. [3] Roy D, Ward P. Evaluation of rapid methods for differentiation of Bifidobacterium species. J Appl Bacteriol 1990;69:739–49. [4] Fanedl L, Nekrep FV, Augustin G. Random amplified polymorphic DNA analysis and demonstration of genetic variability among bifidobacteria isolated from rats fed with raw kidney beans. Can J Microbiol 1998;44:1094–101. [5] Roy D, Sirois S. Molecular differentiation of Bifidobacterium species with amplified ribosomal DNA restriction analysis and alignment of short regions of the ldh gene. FEMS Microbiol Lett 2000;191:17–24. [6] Tannock GW. Identification of lactobacilli and bifidobacteria. Curr Issues Mol Biol 1999;1:53–64. [7] Satokari RM, Vaughan EE, Akkermans ADL, Saarela M, De Vos WM. Bifidobacterial diversity in human feces detected by genus-specific PCR and denaturating gradient gel electrophoresis. Appl Environ Microbiol 2001;67:504–13. [8] Bahaka D, Neut C, Khattabi A, Monget D, Gavini F. Phenotype and genomic analyses of human strains belonging or related to Bifidobacterium longum, Bifidobacterium infantis and Bifidobacterium breve. Int J Syst Bacteriol 1993;43:565–73. [9] Meile L, Ludwig W, Rueger U, Gut C, Kaufmann P, Dasen G, Wenger S, Teuber M. Bifidobacterium lactis sp. Nov: a moderately oxygen tolerant species isolated from fermented milk. Syst Appl Microbiol 1997;20:57–64. [10] Langendijk PS, Schut F, Jansen GJ, Raangs GC, Kamphuis GR, Wilkinson MHF, Welling GW. Quantitative fluorescence in situ hybridisation of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples. Appl Environ Microbiol 1995;61:3069–75. [11] Kullen MJ, Brady LJ, O’Sullivan DJ. Evaluation of using a short region of recA gene for rapid and sensitive speciation of dominant bifidobacteria in the human large intestine. FEMS Microbiol Lett 1997;154:377–83. [12] Requena T, Burton J, Matsuki T, Munro K, Simon MA, Tanakaq R, Watanabe K, Tannock GW. Identification, detection, and enumeration of human Bifidobacterium species by PCR targeting the transaldolase gene. Appl Environ Microbiol 2002; 68:2420–7. [13] Cusik SM, O’Sullivan DJ. Use of a single, triplicate arbitrarily primed-PCR procedure for molecualr fingerprinting of lactic acid bacteria. Appl Environ Microbiol 2000;66:2227–31. [14] Yamamoto T, Morotomi M, Tanaka R. Species-specific oligonucleotides probes for five Bifidobacterium species detected in human intestinal microflora. Appl Environ Microbiol 1992;58:4076–9. [15] Herreman C, Asselin D, Novel G. Identification et caract!erisation de souches industrielles du genre Bifidobacterium par deux m!ethodes de taxonomie mol!eculaire. Lait 1994;74:479–84.
Taxonomy/systematics
Bifidobacteria identification based on 16S rRNA and pyruvate kinase partial gene sequence analysis$ Laurence Vaugiena, Fabien Prevotsb, Christine Roquesa,* a
Laboratoire de Bact!eriologie, Virologie et Microbiologie Industrielle, Facult!e de Pharmacie, 35 Chemin des Mara#ıchers, 31062 Toulouse cedex 4, France b Degussa, Ferments d’aromatisation, 35, Chemin des Mara#ıchers, 31062 Toulouse cedex 4, France Received 4 October 2002; received in revised form 31 January 2003; accepted 5 February 2003
Abstract The lack of a simple and rapid identification system for Bifidobacterium species makes them difficult to use in industrial applications. To obtain valuable discriminating factor, we studied different strains, and human isolates by two molecular taxonomy methods. First method was based on chrono-differentiation. A metabolic gene (pyruvate kinase) was chosen to be used as a systematic discriminating factor. A comparison of about 40 pyruvate kinase protein sequences allowed us to synthesize two oligonucleotides that were able to amplify a fragment of this corresponding gene in our strains. Based on these partial pyruvate kinase gene sequences, several clusters could be identified. The second method used in this study was based on 16S rRNA sequences analysis. We compared sequences present in GenBank database, and this allowed to separate bifidobacteria species into different clusters. They were different from those obtained with partial pyruvate kinase gene sequences analysis. So, by combining both methods, we were able to identify our isolates, when only 10% of them could be strictly identified using the 16S rRNA method. Moreover, pyruvate kinase analysis allowed to differentiate very ambivalent groups such as B. animalis/B. lactis or B. infantis/ B. longum, but created different clusters for B. infantis species group, questioning on the homogeneity of this species. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Bifidobacterium; Pyruvate kinase; Phylogeny; 16S rRNA; Probiotic
1. Introduction Bifidobacterium represents the third most common genus of the human intestine microflora after the genera Bacteroides and Eubacterium (up to 3% of the total fecal microflora of adults) and is considered involved in host’s health [1]. The ever increasing use of specialised strains of bifidobacteria in the treatment and prevention of gastrointestinal disorders requires careful attention to strain identification, even at the species level. To date, differentiation of bifidobacteria species has mainly been performed on the basis of phenotypic characteristics [2,3], expressing poor discriminative power and reproducibility. Molecular approaches have been more recently used to analyse and quantify the composition $ Paper from Anaerobe Olympiad 2002. The 6th Biennial Congress of the Anaerobe Society of the Americas, Park City, Utah, 29 June–2 July 2002. *Corresponding author. Tel.: +33-5-62-25-68-60; fax: +33-5-61-2595-72. E-mail address: [email protected] (C. Roques).
of bacterial groups within human gut microflora [4,5]. Despite of this, in 1999, Tannock [6] underlined that molecular methods for the comprehensive identification of bifidobacteria species are not yet available. Bifidobacteria can be differentiated from morphological similar bacteria by the use of genus-specific PCR primers or oligonucleotide probes, but only DNA– DNA hybridization provides a reliable means of species identification. Since, the use of genus-specific PCR primers targeting the 16S rRNA gene, coupled with denaturing gradient gel electrophoresis (DGGE), has provided a screening method for the detection and identification of most of bacterial species [7,12]. In a recent study, most of human species could be identified by a coupled method. Moreover, not all of human species was included in this study, and so this method was not easily useful for human isolates. In order to develop a convenient method for rapid identification of human Bifidobacterium strains at the species level, we have obtained DNA sequences of a short region of the pyruvate kinase gene, specific for 11
1075-9964/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S1075-9964(03)00025-8
342
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
Bifidobacterium species which are often detected in human. To control the specificity of our sequences, Bifidobacterium species of non-human origin (2), and human isolates (15) were included in the study.
2. Materials and methods 2.1. DNA extraction The specific genomic DNA was prepared from 50 ml of bacterial culture on M17 medium supplemented with 0.4% glucose, 1% lactose and 1% ascorbic acid, in the stationary phase. Cells were harvested by centrifugation (4000 rpm, 5 min), suspended in 5 ml TES (Tris 100 mM pH 8, EDTA 20 mM, sucrose 20%) containing lyzosyme (10 mg ml 1) and incubated at 37 C for 2 h. Next, 50 ml of achromopeptidase (solution at 60 mg ml 1 —Sigma Aldrich, Saint Quentin Fallavier, France) were added. The suspension was then incubated for 2 h at 37 C. After the incubation, 115 ml of a 20% SDS solution were added and further incubated at 65 C during 30 min, and then put on ice for 5 min. After phenol/chloroform (50/ 50 v/v) extraction, DNA was precipitated in absolute alcohol. The precipitate was dried in a centrifugal evaporator and resuspended in water at a concentration of 100 ng ml 1. 2.2. Pyruvate gene partial amplification Forty sequences of pyruvate kinase protein, obtained from the EMBL sequence database, were compared using the Omiga program (Omiga 1.1, Oxford Molecular). GenBank accession numbers were as follows: CAB74228, BAA98307, AAF39440, AAF40552, BAA95686, T35759, B75251, B72406, G69685, D64925, S73859, I40840, JC4219, H64223, JC4220, C64130, S29783, B40620, S29790, S27330, AAF30593, BAA89788, BAA75222, AAC71435, AAC66733, AAC45776, AAC28104, CAB08894, AAC23216, AAC12962, AAB66498, BAA06727, BAA0625, BAA02406, CAA50527, S77240, BAA10621, CAB70653, AAF25804, AAC67927. These data included Gram-positive, Gram-negative, and mycoplasma organisms. Two relatively conserved sequences were translated into DNA. The two corresponding oligonucleotides (TAATCGATGATGGCAAGGTCCG and TAATACGCGGGCTACCATGATGCC—LifeTechnologies Laboratories, Cergy Pontoise, France) were used for amplification. Amplification was performed as follows: 94 C—3 min, 45 C—1 min, 72 C—1 min for 1 cycle and 94 C—1 min, 45 C—1 min, 72 C—1 min for 29 cycles. PCR products obtained were run on 2% agarose gel (w/v) in TAE 1X for 1 h 30 min at 70 mV and made visible by ethidium bromide staining and UV transillumination. The PCR products length was
about 300 bp. They were purified using the Promega extraction Kit, quantified on agarose gel, and then sent for sequencing (Genome Express S.A., Grenoble, France). All sequences obtained were analysed using the Omiga program (Omiga 1.1, Oxford Molecular). Partial pyruvate kinase gene sequences correspond to the following GenBank accession numbers: AF384974 (B. breve CIP 64.69 T), AF384975 (B. catenulatum CIP 104175), AF384976 (B. infantis CIP 64.67 T), AF384977 (B. lactis CIP 105265 T), AF384978 (B. longum CIP 64.62 T), AF384979 (B. adolescentis CIP 64.59 T), AF384980 (B. angulatum CIP 104167), AF384981 (B. animalis DSM 20104), AF384982 (B. bifidum CIP 56.7 T), AF384983 (B. longum CIP 64.63), AF384984 (B. pseudolongum CIP 105352), AF384985 (B. thermophilum CIP 105420 T), AF384986 (B. infantis DSM 20090), AF384987 (B. infantis DSM 20223). 2.3. Obtention of 16S rDNA sequence In order to obtain 16S rRNA partial gene amplification (900 bp fragment length) and sequencing, universal oligonucleotides were used (ATAATGCGGCCGCACGGGCGGTGTGTRC and TAATAGCGGCCGCAGCMGCCGCGGTAATWC). These primers correspond to universal sites at about 1390 and 520 bp, respectively, of 16S rRNA sequences. The annealing mixture was prepared by combining 0.5 ml Taq DNA Polymerase (2.5 units) (Boehringer Mannheim), 5 ml PCR buffer containing dNTP (Boehringer Mannheim buffer supplemented with each dNTP at 10mM), 5 ml of each oligonucleotide solution (60 mg ml 1), 1 ml of template DNA, and 33.5 ml H2O. Amplification was performed as described before. PCR products were desalted and digested by 1 U of NotI for 2 h at 37 C. Then digestions were resolved by 1% gel agarose and visualised by ethidium bromide staining. The fragments were purified using a Promega extraction kit (Promega Corp., Madison, USA), and cloned in pGEM 5Z vector (Promega Corp.). Escherichia coli was electroporated at 25 mF and 2.5 kV (Gene pulser II, Bio-Rad, Ivry/ Seine, France). Plasmids were purified using QIAprep spin plasmid kit (Qiagen S.A., Hilden, Germany), and sent for an insert sequencing (Genome Express S.A., Grenoble, France). Amplification was performed on human isolates (except strains 17, 41, and 47).
3. Results and discussion The amplifications were performed on collection strains including the type strain of the 11 species tested (Table 1), and 15 faecal human isolates. According to this method, the collection strains could be separated into nine clusters (Table 2). Except strains 08, 14, 17, 41, and 47 all the studied isolates belonged to a specific
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
cluster. These five strains shared the same partial PK gene sequence, and could be grouped in a 10th unidentified cluster. Moreover, further experiments need Table 1 Collection strains tested Strain
Referencea
B. B. B. B. B. B. B. B. B. B. B. B. B. B. B.
CIP 64.59T CIP 104167T CIP 105419T DSM 20104 CIP 56.7T CIP64.69T CIP 104175T CIP 64.67T DSM 20090 DSM 20223 CIP 105265T CIP 64.62T CIP 64.63 CIP 105352T CIP 105420T
adolescentis angulatum animalis animalis bifidum breve catenulatum infantis infantis infantis lactis longum longum pseudolongum thermophilum a
CIP Pasteur Institute Collection DSM:Deutsche Sammlung von Mikrooganismen.
343
to be performed on the two Bifidobacterium species isolated from humans not yet included in our study (B. dentium and B. pseudocatenulatum). This would lead us to confirm their differentiation among the PK clusters. Despite of this, none of the 15 tested human isolates, belonged to B. adolescentis, B. angulatum, B. breve, or B. catenulatum species. The PK gene sequence analysis led to a sufficient discrimination at species level of human strains for four different clusters (cluster 4—B. longum, cluster 5—B. angulatum; cluster 6—B. breve; cluster 7—B. catenulatum). The specificity is indicated by the differentiation, without cross-reaction, of B. animalis (cluster 3) and B. thermophilum (cluster 9). Although the several B. infantis collection strains studied were classified in three different clusters (clusters 1, 2 and 8), they did not share the same sequence with the B. longum strains studied (cluster 4). So, partial PK gene sequence analysis allowed us to differentiate B. infantis from B. longum and also B. animalis from B. lactis. After 16S rDNA amplification, sequences were analysed using the FastA program. These data did not lead to a bifidobacteria identification at the species level
Table 2 Consensus grouping of bifidobacteria based on 16S rRNA and partial pyruvate kinase DNA sequences PK Analysis
16S rRNA analysis Cluster A B. adolescentis B. angulatum B. catenulatum B. dentium B. pseudocatenulatum
Cluster 1 B. bifidum CIP56.7T B. lactis CIP105265T B. infantis CIP 64.67T B. pseudolongum CIP105352 Cluster 2 B. adolescentis CIP64.59T B. infantis DSM 20090 Cluster 3 B. animalis CIP105419T B. animalis DSM20104 Cluster 4 B. longum CIP64.62T B. longum 64.63 Cluster 5 B. angulatum Cluster 6 B. breve CIP64.69T Cluster 7 B. catenulatum CIP104175T Cluster 8 B. infantis DSM 20223 Cluster 9 B. thermophilum CIP105420T a
Industrial strains.
B. adolescentis
Cluster B B. breve B. infantis B. longum B. pseudolongum
Cluster C B. animalis B. lactis
B. infantis B. pseudolongum 05, 56, 82
B. lactis 39, 63, 65a, 73a
Cluster D B. thermophilum
Cluster E B. bifidum
B. bifidum 02, 20
B. infantis
B. animalis 11 B. longum 29, 69a, 80 B. angulatum B. breve B. catenulatum B. infantis B. thermophilum
344
L. Vaugien et al. / Anaerobe 8 (2002) 341–344
but only to a differentiation in clusters as already described by several teams [8,9]. They are in agreement with our PK clusters differentiation. Moreover, the combination of the two methods allow us to confirm the identification of the tested human isolates at the species level. In contrast to many genera, the bifidobacteria form a monophyletic cluster on the basis of rRNA sequences. This led to the development of genus-specific 16S rRNA probes [10] but did not allow all Bifidobacterium species to be differentiated. There are many recently developed approaches for the molecular fingerprinting of bifidobacteria, i.e. amplified ribosomal DNA restriction analysis [5], analyses of conserved gene sequences [11, 12] or triplicate arbitrarily primed-PCR procedure [13]. Such molecular methods have frequently a discriminatory power with strains differentiation for one species. But, on the other hand, differentiation of even closely related strains, like B. longum/B. infantis, and B. animalis/B. lactis remains confused [8,9]. Using oligonucleotides probes for RNA–DNA hybridisation, Yamamoto et al. [14] were able to detect and identify B. adolescentis, B. bifidum, B. breve, B. infantis and B. longum. Despite of this, probes for B. bifidum and B. infantis cross-reacted with a few strains of non-human origin. In the same way, in a recent report, Requena et al. [12] identified bifidobacteria by PCR amplification of a 301bp transaldolase gene sequence and DGGE. Under such conditions, the authors observed two subtypes of B. longum, five subtypes of B. adolescentis and two subtypes of B. pseudocatenulatum. On the other hand, B. angulatum and B. catenulatum type cultures could not be differentiated from each other demonstrating the difficulty to obtain accurate identification tool. To complete our study, experiments were also performed on three industrial strains (Table 2). The first one is confirmed as B. longum. The two others preliminary given as B. animalis were identified as B. lactis. Herreman et al. [15] have also identified and characterised three industrial strains by combining two molecular taxonomy methods (DNA–DNA hybridisation and DNA restriction XbaI fragment polymorphism). One strain is confirmed as B. longum. The two others presumed to belong to B. infantis and B. bifidum species, were identified as B. animalis. These observations underline the difficulty to develop acurate method for bifidobacteria identification. In this study, we designed partial pyruvate kinase gene analysis as a reliable, rapid and accurate manner for the identification of human intestinal bifidobacteria. Although complementary data are necessary (B. dentium
and B. pseudocatenulatum) to definitely purpose this method combined with 16S rRNA sequence analysis for bifidobacteria differentiation.
References [1] Shir A, Gramet G, Suau A, Rochet V, Pochart P, Dore J. Quantification of bacterial groups within the human fecal flora by oligonucleotide probe hybridization. Appl Environ Microbiol 2000;66:2263–6. [2] Mitsuoka T. Taxonomy and ecology of bifidobacteria. Bifidobacteria Microflora 1984;3:11–28. [3] Roy D, Ward P. Evaluation of rapid methods for differentiation of Bifidobacterium species. J Appl Bacteriol 1990;69:739–49. [4] Fanedl L, Nekrep FV, Augustin G. Random amplified polymorphic DNA analysis and demonstration of genetic variability among bifidobacteria isolated from rats fed with raw kidney beans. Can J Microbiol 1998;44:1094–101. [5] Roy D, Sirois S. Molecular differentiation of Bifidobacterium species with amplified ribosomal DNA restriction analysis and alignment of short regions of the ldh gene. FEMS Microbiol Lett 2000;191:17–24. [6] Tannock GW. Identification of lactobacilli and bifidobacteria. Curr Issues Mol Biol 1999;1:53–64. [7] Satokari RM, Vaughan EE, Akkermans ADL, Saarela M, De Vos WM. Bifidobacterial diversity in human feces detected by genus-specific PCR and denaturating gradient gel electrophoresis. Appl Environ Microbiol 2001;67:504–13. [8] Bahaka D, Neut C, Khattabi A, Monget D, Gavini F. Phenotype and genomic analyses of human strains belonging or related to Bifidobacterium longum, Bifidobacterium infantis and Bifidobacterium breve. Int J Syst Bacteriol 1993;43:565–73. [9] Meile L, Ludwig W, Rueger U, Gut C, Kaufmann P, Dasen G, Wenger S, Teuber M. Bifidobacterium lactis sp. Nov: a moderately oxygen tolerant species isolated from fermented milk. Syst Appl Microbiol 1997;20:57–64. [10] Langendijk PS, Schut F, Jansen GJ, Raangs GC, Kamphuis GR, Wilkinson MHF, Welling GW. Quantitative fluorescence in situ hybridisation of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples. Appl Environ Microbiol 1995;61:3069–75. [11] Kullen MJ, Brady LJ, O’Sullivan DJ. Evaluation of using a short region of recA gene for rapid and sensitive speciation of dominant bifidobacteria in the human large intestine. FEMS Microbiol Lett 1997;154:377–83. [12] Requena T, Burton J, Matsuki T, Munro K, Simon MA, Tanakaq R, Watanabe K, Tannock GW. Identification, detection, and enumeration of human Bifidobacterium species by PCR targeting the transaldolase gene. Appl Environ Microbiol 2002; 68:2420–7. [13] Cusik SM, O’Sullivan DJ. Use of a single, triplicate arbitrarily primed-PCR procedure for molecualr fingerprinting of lactic acid bacteria. Appl Environ Microbiol 2000;66:2227–31. [14] Yamamoto T, Morotomi M, Tanaka R. Species-specific oligonucleotides probes for five Bifidobacterium species detected in human intestinal microflora. Appl Environ Microbiol 1992;58:4076–9. [15] Herreman C, Asselin D, Novel G. Identification et caract!erisation de souches industrielles du genre Bifidobacterium par deux m!ethodes de taxonomie mol!eculaire. Lait 1994;74:479–84.