Molecular diversity of Bacteroides spp. in human fecal microbiota as determined by group-specific 16S rRNA gene clone library analysis

ARTICLE IN PRESS

Systematic and Applied Microbiology 32 (2009) 193–200 www.elsevier.de/syapm

Molecular diversity of Bacteroides spp. in human fecal microbiota as determined by group-specific 16S rRNA gene clone library analysis$ Min Lia, Haokui Zhoua, Weiying Huaa, Baohong Wangb, Shengyue Wangc, Guoping Zhaoc, Lanjuan Lib, Liping Zhaoa, Xiaoyan Panga, a

Laboratory of Molecular Microbial Ecology and Ecogenomics, College of Life Science and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, China b State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, The First Affiliated Hospital, Zhejiang University, Hangzhou 310003, China c Shanghai-MOST Key Laboratory for Disease and Health Genomics, Chinese National Human Genome Center, Shanghai 201203, China

Abstract Bacteroides spp. represent a prominent bacterial group in human intestinal microbiota with roles in symbiosis and pathogenicity; however, the detailed composition of this group in human feces has yet to be comprehensively characterized. In this study, the molecular diversity of Bacteroides spp. in human fecal microbiota was analyzed from a seven-member, four-generation Chinese family using Bacteroides spp. group-specific 16S rRNA gene clone library analysis. A total of 549 partial 16S rRNA sequences amplified by Bacteroides spp.-specific primers were classified into 52 operational taxonomic units (OTUs) with a 99% sequence identity cut-off. Twenty-three OTUs, representing 83% of all clones, were related to 11 validly described Bacteroides species, dominated by Bacteroides coprocola, B. uniformis, and B. vulgatus. Most of the OTUs did not correspond to known species and represented hitherto uncharacterized bacteria. Relative to 16S rRNA gene universal libraries, the diversity of Bacteroides spp. detected by the group-specific libraries was much higher than previously described. Remarkable inter-individual differences were also observed in the composition of Bacteroides spp. in this family cohort. The comprehensive observation of molecular diversity of Bacteroides spp. provides new insights into potential contributions of various species in this group to human health and disease. r 2009 Elsevier GmbH. All rights reserved. Keywords: Bacteroides spp.; 16S rRNA; Group-specific clone library; Microbial diversity; Human feces

Introduction Bacteroides spp. is one of the most prominent groups in the human intestinal microbiota, which accounts for $ Note: Nucleotide sequence data reported are available in the GenBank databases under the accession numbers FJ209735–FJ210283. Corresponding author. Room 3-519, Biology Building, 800 Dongchuan Road, Shanghai 200240, China. Tel./fax: +86 21 34204878. E-mail address: [email protected] (X. Pang).

0723-2020/$ - see front matter r 2009 Elsevier GmbH. All rights reserved. doi:10.1016/j.syapm.2009.02.001

14–40% of cultivable bacteria in human feces [10]. Most Bacteroides spp. strains belong to ten cultivated species including Bacteroides vulgatus, B. thetaiotaomicron, B. distasonis, B. caccae, B. eggerthii, B. fragilis, B. merdae, B. ovatus, B. stercoris, and B. uniformis [28]. These bacteria have significant effects on human health, most notably in carbohydrate fermentation and catabolism of polysaccharides [32]. Some strains are important opportunistic pathogens, such as B. fragilis [18,30]. However, because 60–80% of bacteria in the human gut

ARTICLE IN PRESS 194

M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

are uncultivable due to their strict requirement for anaerobic conditions and unknown nutritional needs [4,12], many studies of Bacteroides spp. have been limited to the known cultivable species, such as B. thetaiotaomicron [32]. Therefore, to understand the roles of these bacteria, it is essential to characterize the population composition of Bacteroides spp. in the human gut at a comprehensive level. Molecular ecological techniques based on 16S rRNA gene sequences provide detailed insights into complex microbial communities. Specific primers [18] or probes [3,25] have been designed to detect species of Bacteroides and to enumerate the Bacteroides population in human feces by dot blot hybridization and fluorescent in situ hybridization (FISH). Recently, a new molecular technique, termed hierarchical oligonucleotide primer extension (HOPE), has been developed to evaluate the relative abundances of Bacteroides spp. in human fecal samples [13]. However, these quantitative techniques provide little information about the diverse composition of Bacteroides spp. Group-specific fingerprinting techniques have been widely used to assess and monitor the composition of this group in gut microbiota. Bacteroides spp.-specific temperature gradient gel electrophoresis (TGGE) analysis has been developed to profile the composition of Bacteroides spp. in the human gut [23], and Bacteroides-Prevotella group-specific terminal restriction length polymorphism (T-RFLP) has been used to compare the Bacteroides spp. composition among human and other species [5]. Though these methods are suitable for quick profiling and comparison of Bacteroides diversity between samples, clone library analysis is needed for a detailed and comprehensive inventory of this bacterial group in human guts [4,17,31]. Recently, we used the universal 16S rRNA gene library to describe the overall structure of gut microbiomes of a seven-member, four-generation Chinese family, and found that 16% of the clones were Bacteroides spp.-related sequences [17]. However, some minor species might have been missed by this universal library analysis due to the preference for amplification of prevalent species during PCR. For example, Hayashi et al. [11] showed a much greater diversity of the Clostridium coccoides group using group-specific 16S rRNA gene libraries than was detected using universal libraries. Therefore, in order to obtain a more detailed phylogenetic inventory of Bacteroides spp. in the human gut, we used Bacteroides spp.-specific clone libraries with group-specific primers to determine the composition of Bacteroides spp. in human feces from the seven Chinese family members. This set of data describing the surprisingly diverse Bacteroides spp. group lays the foundation for further understanding the roles these species play in disease and health in the Chinese population.

Materials and methods Subject selection and sample preparation A four-generation Chinese family, including greatgrandmother (GG, aged 95 years), grandmother (GM, aged 58 years), grandfather (GF, aged 55 years), father (FA, aged 30 years), mother (MO, aged 30 years), uncle (UC, aged 18 years), and baby (BB, aged 1.5 years) was chosen as the study cohort [17]. GG, GM, and GF lived in the same house; FA, MO, and BB lived in another house in the same city. UC had been living with GG, GM, and GF but had spent 2 years in the UK prior to being sampled in China. None of the subjects had a past medical history of gastrointestinal diseases, and none had received probiotics or antibiotics within at least 3 months before sampling. The use of the subjects was approved by the First Affiliated Hospital of Zhejiang University Institutional Review Broad. All volunteers in this study provided informed consent. Fresh fecal samples were collected into sterile tubes (50 ml) and placed into an anaerobic jar (GENbox anaer, Biome´rieux), then transferred to the laboratory immediately in an ice-box and stored at 70 1C within 15 min of sample collection. All fresh samples were pretreated according to Li et al. [17]. A fresh sample (1 g wet weight) was suspended in 30 ml sterile ice-cold sodium phosphate buffer (0.1 M SPB: 1 l contained 1 M Na2HPO4 57.7 ml, 1 M NaH2PO4 42.3 ml, pH 7.4) followed by vortexing for 30 min in a 50 ml tube. The suspension was centrifuged three times at 200g for 5 min to remove coarse particles. The cells in the supernatant were collected and washed three times by centrifuging at 9000g for 5 min followed by resuspension in 30 ml fresh SPB. The washed cell pellets were resuspended in 10 ml sterile SPB, allocated into 1 ml aliquots and stored at 70 1C until needed for DNA extraction.

DNA extraction and Bacteroides spp.-specific PCR amplification Total genomic DNA from fecal samples was extracted from lysates prepared by the bead beating method (Biospec Products, Bartlesville, OK) and purified with phenol–chloroform extraction, as described earlier [17]. The amount of extracted DNA was determined using a DyNA Quant 200 fluorometer (Amersham Pharmacia Biotech, USA), and its integrity and size were checked by 0.8% (w/v) agarose gel electrophoresis in 0.5 mg/ml ethidium bromide. All DNA samples were stored at 20 1C until needed for further analysis. Primers Bfr-F (50 -CTGAACCAGCCAAGTAGCG-30 ) (Escherichia coli position 386–405) and Bfr-R (50 -CCGCAAACTTTCACAACTGACTTA-30 ) (E. coli position 593–617) [18,23] were used to amplify Bacteroides

ARTICLE IN PRESS M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

spp.-specific fragments of 16S rRNA genes from the extracted DNA. The 25 ml PCR reaction mixture contained 0.5 U of TaKaRa rTaq polymerase (Takara, Dalian, China), 1  PCR buffer (Mg2+ free), 2 mM MgCl2, 10 pmol of each primer, 200 mM deoxynucleoside triphosphate (dNTP), and 10 ng of template DNA. PCR amplification was performed in a thermocycler PCR system (PCR Sprint, Thermo electron, Corp., UK) using the following procedure: 95 1C for 2 min, 30 cycles at 95 1C for 30 s, 53 1C for 30 s, 72 1C for 1 min and finally, 72 1C for 5 min. The 230-bp PCR products were purified with DNA Gel Extraction Kit (V-gene, Hangzhou, China), according to the manufacturer’s instruction, and then analyzed by 1% (w/v) agarose gel electrophoresis.

Cloning and sequencing The purified PCR products were ligated into pGEM-T Easy Vector (Promega, Madison, WI) and then transformed into E. coli DH5a. Transformants were randomly selected by plating into Luria–Bertani (LB)/Ampicillin (100 mg/ml) agar plates with IPTG (isopropyl-b-D-thiogalactopyranoside) and X-Gal (5-bromo-4-chloro-3-indolylb-D-galactopyranoside), and incubated at 37 1C overnight. White clones from each adult subject (n ¼ 96) and baby (n ¼ 48) were randomly picked for sequencing. The insert DNA of each clone was prepared by alkaline lysis, and sequenced using the BigDye Terminator (Applied Biosystems) with T7 sequencing primers on ABI 3700xl sequencers (Applied Biosystems).

Phylogenetic analysis of Bacteroides spp.-related sequences in group-specific and universal clone libraries

195

vector sequences, then checked for chimeras using Chimera Check v2.7 on the Ribosomal Database Project II (RDP-II) website [21]. All sequences were aligned to the Greengenes database [2] using the ARB package [20]. Aligned sequences were grouped into operational taxonomic units (OTUs) at a threshold of 99% minimum similarity from an Olsen-corrected distance matrix by DOTUR with the furthest-neighbor algorithm and 0.001 precision [26]. Nearest neighbors with full-length 16S rRNA sequences of the representative clones of each OTU in the group-specific libraries were searched by the SILVA Aligner [24]. Bacteroides spp.-related sequences in previous universal libraries, which were constructed using universal primers 27f and 1492r [17], were selected based on a 100% match with two Bacteroides spp.-specific primers utilized in the present work. All sequences were aligned to the RDP-II database using the ARB package, and grouped into OTUs at a 99% sequence identity cut-off using the method described above. In order to identify the phylogenetic affiliation of Bacteroides spp. group-specific sequences, a reference phylogenetic tree including validly described Bacteroidetes isolated species [33], representative clones of each OTU in the universal libraries and nearest neighbors of groupspecific OTUs, was constructed by using a neighbor-joining algorithm from an Olsen-corrected distance matrix in the ARB. The representative clones of each OTU in the group-specific libraries were then inserted into the reference tree using the parsimony insertion algorithm in the ARB. The relative abundances of each OTU in individual and combined universal/group-specific libraries were indicated in the tree.

Diversity and richness estimation of group-specific libraries

All sequences obtained from group-specific libraries were manually checked and trimmed to exclude the

To determine whether the size of a clone library was large enough to represent the diversity of an original

Table 1. Diversity indices and coverage estimation of each individual and the combined Bacteroides spp. group-specific clone libraries.

No. of clones No. of total OTUs No. of host-specific OTUs No. of singletons Chao1 estimator of OTUs ACE estimator of OTUs Shannon’s index (H) Simpson’s index (1D) Evenness Good’s coverage (%) No. of Bacteroides related clones in the universal library No. of Bacteroides OTUs in the universal library

GG

GM

GF

UC

FA

MO

BB

Combined

77 14 5 4 16.3 17 2.17 0.84 0.63 94.8 244 13

78 13 4 5 16 18.2 1.68 0.68 0.41 93.6 124 14

88 13 4 6 18.4 23 1.5 0.61 0.35 93.2 173 6

85 14 6 8 28.3 26.5 1.62 0.7 0.36 90.6 109 9

93 18 5 8 25.6 29.2 2.04 0.77 0.43 91.4 68 11

83 18 5 9 25.4 29.8 1.85 0.69 0.35 89.2 63 12

45 2 2 0 2 2 0.3 0.16 0.67 100 388 1

549 52 31 19 67 74.3 2.74 0.88 0.3 96.5 1169 27

Analysis based on Bacteroides spp.-related sequences in the universal libraries [17] is also included here.

ARTICLE IN PRESS 196

M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

community, two non-parametric estimators, SChao1 and SACE, were calculated for each library using an online program (http://www.aslo.org/lomethods/free/ 2004/0114a.html) and the output data were treated using the method described by Kemp and Aller [15,16]. The coverage of each clone library and combined sequences set were calculated as [1(n/N)]  100 by Good’s method [8], where n is the number of singletons and N is the total number of sequences. The diversity of each sequence set was estimated using the Shannon Wiener index (H) and Simpson’s index (1D) with PAST software [9].

representing 29.5% of total clones (162 clones, 6 OTUs), B. uniformis, 20.4% of total clones (112 clones, 4 OTUs), and B. vulgatus, 8.6% of total clones (47 clones, 2 OTUs). The less prominent OTUs were related to B. caccae, B. dorei, B. eggerthii, B. massiliensis, B. cellulosilyticus, B. finegoldii, B. fragilis, and B. coprophilus. The remaining 29 OTUs (91 clones, 55.8% of total OTUs) had no nearest isolated relatives, and could be considered as uncharacterized.

Statistical comparison of group-specific libraries

On average, 15 OTUs were detected for each adult. FA and MO had the greatest number of observed OTUs among all the adults (18 OTUs, represented by 93 and 83 clones, respectively), while GM and GF had the lowest number of OTUs in the adults (13 OTUs, represented by 78 and 88 clones, respectively). GG and UC had 14 OTUs, represented by 77 and 85 clones, respectively. Only two OTUs (SOTU46 and SOTU47, including a total of 45 clones) were detected in the baby. Diversity estimation by the Shannon Wiener index and Simpson’s index also showed that the species diversity and evenness of Bacteroides spp. in the baby’s gut were much lower than that in the adults, and those of GG and FA were slightly higher than other adults (Table 1). The distribution and proportion of Bacteroides spp. OTUs in the group-specific libraries varied remarkably among individuals (denoted in Fig. 1 and Table S1 in Supplementary Material; the accession number of the clones of each OTU is shown in Table S2 in Supplementary Material). No OTU was present in all subjects. Only one OTU (SOTU5, including 30 clones) was common among all the adults, and it was closely related to B. dorei in the phylogenetic tree. The proportion of this OTU in each adult group-specific library varied from 1.2% to 9.1%. Three OTUs, SOTU 7, 9, and 10, existed in 5 out of 6 adults. SOTU9 and SOTU10, which were related to B. coprocola, were both detected in five adults, all except GF, and the proportions of these two OTUs were higher in young adults than in the elderly. SOTU7, which was related to B. uniformis, also existed in five adults, all except MO, and it had a higher frequency in the elderly. However, a greater number of OTUs (a total of 31) were individual-specific OTUs. The distribution of

To determine whether the differences between clone libraries were R significant or were caused by artifacts of sampling, -LIBSHUFF analysis [27,29] based on an Olsen-corrected dissimilarity matrix was performed for pairwise comparisons in each library, and the P value was estimated by 10,000 random permutations of sequences between libraries.

Results Overall diversity of Bacteroides spp. group-specific clone libraries A total of 549 clones were obtained from the seven Bacteroides spp. group-specific clone libraries. Based on a 99% sequence identity cut-off, the sequences were grouped into 52 operational taxonomic units with 19 singletons (Table 1). Good’s coverage of the whole library was 496%. In addition, Chao1 and ACE estimators of total species richness were 67 and 74, respectively, similar to the number of observed OTUs, indicating that the size of the library was large enough to represent the majority of Bacteroides species in the human gut used in the present study. The phylogenetic relationships between each OTU and Bacteroides spp. isolates are shown in the phylogenetic tree (Fig. 1). Based on the valid reference tree, from among a total of 52 OTUs, 23 (458 clones, 83.4% of total clones) were closely related to 11 validly described Bacteroides spp. isolates [33]. Most of them (58.5% of all Bacteroides spp. clones) corresponded to three species: Bacteroides coprocola,

OTU distribution of individual Bacteroides spp. group-specific clone libraries

Fig. 1. Phylogenetic tree showing the relationship of 16S rRNA operational taxonomic units (OTUs) obtained from Bacteroides spp. group-specific clone libraries and the previous universal clone libraries [17]. A reference phylogenetic tree including validly described Bacteroidetes isolates (bold) [33], representative clones of each OTU in the universal libraries, and nearest neighbors with full length 16S rRNA sequences of group-specific OTUs was constructed using a neighbor joining algorithm from an Olsencorrected distance matrix in the ARB. The representative clones of each OTU in the group-specific libraries were then inserted into the reference tree using the parsimony insertion algorithm in the ARB. SOTU represents those clones detected in the group-specific library, and UOTU those detected in the universal library. The proportion of each OTU in the universal/group-specific libraries is noted after its name. The individuals in which each OTU has been observed and the relative abundances of each OTU in each individual library are shown in parentheses. The scale bar represents 0.10 substitutions per nucleotide position.

ARTICLE IN PRESS M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

197

ARTICLE IN PRESS 198

M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

unique OTUs in each individual is shown in Fig. 1 and Table S1. GG, FA and MO had 5 unique OTUs in their group-specific libraries, represented by 39, 6, and 7 clones, respectively. GM and GF had 4 unique OTUs, represented by 6 and 4 clones, respectively. UC had 6 unique OTUs, represented by 11 clones. The baby only had two OTUs, both of which were unique, and the majority of the clones (91% of the total) were related to B. caccae. Most of the individual-specific OTUs had no nearest isolated relatives. R -LIBSHUFF analysis was performed to determine statistically the inter-individual differences in the composition of Bacteroides spp. The results showed that most of the group-specific libraries were significantly different (po0.05), with the exception of the Bacteroides spp. compositions of FA and MO, which were not significant (p40.05). These two subjects shared similar proportions of B. coprocola-like OTUs, which were the most predominant Bacteroides spp. in their guts (51.6% in FA and 59% in MO).

Comparison of diversity of Bacteroides spp. in the group-specific library and universal library A total of 1169 Bacteroides spp.-related sequences in our previous universal libraries were grouped into 27 OTUs with a 99% sequence identity cut-off. The number of total OTUs was less than that for the group-specific libraries, even though the number of sequenced clones was almost twice that of the specific libraries. The numbers of observed OTUs in most of the individuals were also less in the universal libraries than in the corresponding group-specific libraries, except for GM who had 14 OTUs in the universal library but 13 in the group-specific library. However, the number of sequenced clones in GM’s universal library was much higher (124 clones and 78 clones), so that the proportion of OTUs to the total number of clones was still higher in the group-specific library (Table 1). The OTU distribution of each individual universal library and the accession numbers of clones for each OTU is shown in Tables S1 and S2, respectively. The phylogenetic distributions of all OTUs detected by the two clone library approaches are shown in the phylogenetic tree (Fig. 1). OTUs nearest to eight species, including B. coprocola, B. uniformis, B. vulgatus, B. dorei, B. eggerthii, B. massiliensis, B. fragilis, and B. finegoldii were detected in the adults by both groupspecific and universal libraries. In the baby’s gut, only one prominent OTU, which was related to B. caccae, was detected in both sets of libraries. Some minor OTUs were only detected in one of the library sets. For example, B. coprophilus and B. cellulosilyticus-like OTUs only existed in the groupspecific libraries, while B. thetaiotaomicron, B. ovatus, B. intestinalis, B. stercoris, and B. xylanisolvens-like

OTUs were only detected in the universal libraries. In addition, a greater number of uncharacterized OTUs were detected in the group-specific libraries (Fig. 1).

Discussion In previous studies, the diversity of Bacteroides spp. in the human gut has been reported using universal 16S rRNA gene clone libraries [10,12,31]. A total of 23 Bacteroides OTUs (98% identity cut-off) were detected from a universal library set containing a total of 744 clones constructed from three adults [10]. In our previous universal library, 27 Bacteroides OTUs (99% identity cut-off) were grouped from a total of 7255 clones in seven individuals [17]. However, the significant increase of library coverage does not lead to detection of much higher diversity of Bacteroides spp. in the human gut. This is possibly due to the tendency to preferentially amplify highly represented species during PCR, which leads to the difficult detection of minor members in the gut with universal primers. Even if the coverage is large enough to reflect the overall diversity of the microbial community, the diversity of a specific group would still be underestimated by the universal library. It was suggested that a group-specific clone library approach could exhibit higher diversity of a specific group in the human gut than a universal library analysis [11]. In the present study, with less than half of the clones, the number of Bacteroides OTUs detected by the groupspecific libraries was still almost twice that of the universal libraries. Most of the group-specific OTUs were uncharacterized, showing that a higher, and previously undetected diversity of Bacteroides spp., was represented by the group-specific library. Therefore, group-specific clone library analysis can provide more comprehensive descriptions of the Bacteroides spp. composition of the human gut. OTUs closely related to B. dorei, B. coprocola, and B. uniformis in the phylogenetic tree were shared in most of the adult group-specific libraries, indicating that these bacteria might be core Bacteroides species of gut microbiota in this study cohort. The B. dorei OTU was common among all the adults, but with low abundance in the group-specific libraries. B. dorei was only recently isolated and identified from human feces [1], and its physiology and function has still not been studied clearly. However, the high frequency of this bacteria detected in the gut of adults suggests the potential importance of this species, and it should be studied in larger scale samples. B. coprocola-related OTUs were detected with the highest abundance in adult group-specific libraries. In our previous work, we found that this species was statistically related to the variation of two metabolites in human urine, dihydrothymine and 3,5-hydroxybenzoate, which are produced by gut bacteria [17]. The high abundance of this species and

ARTICLE IN PRESS M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

its relatives in the human gut suggests its significant roles in human metabolism. B. uniformis OTUs were the second most abundant OTUs in the group-specific libraries. B. uniformis is a prominent species in the human gut that constitutes 13–36% of the Bacteroides population of normal human feces [12,23,31]. This species has also been found to correlate with host urinary metabolites, such as citrate and taurine [17]. Therefore, the high proportion and the metabolic relationship with the human host of these common species indicate that they would be prominent and beneficial bacteria for human health. Although there were some shared Bacteroides species in most of the family members, the seven Chinese individuals still had R significant host-specific Bacteroides spp. composition. -LIBSHUFF analysis also showed the high interindividual variability, which is in accordance with the result of the comparisons of the universal 16S rRNA gene clone library in our previous work [17]. A substantial proportion of total OTUs in each individual was hostspecific and mostly had no nearest isolated relatives. This indicates that there are still a large number of uncharacterized bacteria in the human gut. The host specificity of gut microbial composition has been well recognized by several diversity surveys of human gut microbiota [4,7,17]. The individual differences in the composition of Bacteroides spp. may be due to the different age, gender, diet, and other environmental factors. The relationship between these confounding factors and Bacteroides spp. should be studied with larger scale samples. The host-specific Bacteroides spp. composition in the baby’s gut is particularly interesting. Only two OTUs existed in the baby’s gut, and low diversity of gut microbiota in the baby was detected in our universal library [17]. The majority of clones were related to B. caccae, which was absent in adult guts, although B. caccae was originally isolated from adult feces [14]. However, Garcia-Rodriguez et al. [6] also isolated this species from the feces of two newborns, whereas Pang et al. [23] detected 13% of Bacteriodes spp. as B. caccae in a 10-year-old boy’s feces. In a recent large diversity survey of human mucosal and fecal microbiota from three adults, 2.8% of the total sequences belonged to B. caccae [4]. Although to our knowledge no study has reported that B. caccae is dominant in the baby’s gut, the predominance of the species observed in our work suggests that the potential significance of this microorganism during microbiota development should be investigated further. In this study, we used a 230-bp 16S rRNA gene fragment amplified by Bacteroides group-specific primers to reflect the structure of Bacteroides spp. However, the 230-bp partial 16S rRNA gene sequence seems to be short for accurate phylogenetic analysis. One way to circumvent this shortfall is to insert the short sequences into a validated and optimized phylo-

199

genetic reference tree with high-quality full length sequences to ensure reliable phylogenetic identification [19]. The validly described species in our reference tree also provide accurate taxonomic classification of the inserted short sequences [33]. This method has been widely applied in diversity analysis of gut microbiota based on short fragments of 16S rRNA genes, such as short reads generated by pyrosequencing [22]. In conclusion, an extensive diversity of Bacteroides spp. in human feces was derived from group-specific 16S rRNA gene clone library analysis in the present study. The group-specific library approach allows a more elaborate description of the composition of Bacteroides spp. in the Chinese gut than the universal approach. Most of the species detected had not been previously characterized. This study indicates the need for further work with larger populations to understand the symbiotic roles of these new Bacteroides species in human health and disease.

Acknowledgements This work was supported by the National Basic Research Program of China (2007CB513002), and the National Natural Science Foundation of China (30730005), as well as Shanghai International Cooperation Program Grants 075407001 and 075407064.

Appendix A. Supplemental data Supplementary data associated with this article can be found in the online version at doi:10.1016/j.syapm. 2009.02.001.

References [1] M.A. Bakir, M. Sakamoto, M. Kitahara, M. Matsumoto, Y. Benno, Bacteroides dorei sp. nov., isolated from human faeces, Int. J. Syst. Evol. Microbiol. 56 (2006) 1639–1643. [2] T.Z. DeSantis, P. Hugenholtz, N. Larsen, M. Rojas, E.L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, G.L. Andersen, Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB, Appl. Environ. Microbiol. 72 (2006) 5069–5072. [3] J. Dore, A. Sghir, G. Hannequart-Gramet, G. Corthier, P. Pochart, Design and evaluation of a 16S rRNAtargeted oligonucleotide probe for specific detection and quantitation of human faecal Bacteroides populations, Syst. Appl. Microbiol. 21 (1998) 65–71. [4] P.B. Eckburg, E.M. Bik, C.N. Bernstein, E. Purdom, L. Dethlefsen, M. Sargent, S.R. Gill, K.E. Nelson, D.A. Relman, Diversity of the human intestinal microbial flora, Science 308 (2005) 1635–1638. [5] L.R. Fogarty, M.A. Voytek, Comparison of BacteroidesPrevotella 16S rRNA genetic markers for fecal samples from different animal species, Appl. Environ. Microbiol. 71 (2005) 5999–6007.

ARTICLE IN PRESS 200

M. Li et al. / Systematic and Applied Microbiology 32 (2009) 193–200

[6] J.A. Garcia-Rodriguez, J.E. Garcia-Sanchez, M.I. Heras de Pedro, Isolation of newly described species of the Bacteroides fragilis group from the faeces of newborns, Microb. Ecol. Health. Dis. 3 (1990) 1–6. [7] S.R. Gill, M. Pop, R.T. Deboy, P.B. Eckburg, P.J. Turnbaugh, B.S. Samuel, J.I. Gordon, D.A. Relman, C.M. Fraser-Liggett, K.E. Nelson, Metagenomic analysis of the human distal gut microbiome, Science 312 (2006) 1355–1359. [8] I.J. Good, The population frequencies of species and the estimation of population parameters, Biometrika 40 (1953) 237–264. [9] O. Hammer, D. Harper, P. Ryan, PAST: palaeontological statistics software package for education and data analysis, Palaeontol. Electron. 4 (2001) 1–9. [10] H. Hayashi, M. Sakamoto, Y. Benno, Phylogenetic analysis of the human gut microbiota using 16S rDNA clone libraries and strictly anaerobic culture-based methods, Microbiol. Immunol. 46 (2002) 535–548. [11] H. Hayashi, M. Sakamoto, M. Kitahara, Y. Benno, Diversity of the Clostridium coccoides group in human fecal microbiota as determined by 16S rRNA gene library, FEMS Microbiol. Lett. 257 (2006) 202–207. [12] H. Hayashi, M. Sakamoto, M. Kitahara, Y. Benno, Molecular analysis of fecal microbiota in elderly individuals using 16S rDNA library and T-RFLP, Microbiol. Immunol. 47 (2003) 557–570. [13] P.Y. Hong, J.H. Wu, W.T. Liu, Relative abundance of Bacteroides spp. in stools and wastewaters as determined by hierarchical oligonucleotide primer extension, Appl. Environ. Microbiol. 74 (2008) 2882–2893. [14] J.L. Johnson, W.E. Moore, L.V. Moore, Bacteroides caccae sp. nov., Bacteroides merdae sp. nov., and Bacteroides. stercoris sp. nov. isolated from human feces, Int. J. Syst. Bacteriol. 36 (1986) 499–501. [15] P.F. Kemp, J.Y. Aller, Bacterial diversity in aquatic and other environments: what 16S rDNA libraries can tell us, FEMS Microbiol. Ecol. 47 (2004) 161–177. [16] P.F. Kemp, J.Y. Aller, Estimating prokaryotic diversity: when are 16S rDNA libraries large enough?, Limnol. Oceanogr. Methods 2 (2004) 114–125. [17] M. Li, B. Wang, M. Zhang, M. Rantalainen, S. Wang, H. Zhou, Y. Zhang, J. Shen, X. Pang, M. Zhang, H. Wei, Y. Chen, H. Lu, J. Zuo, M. Su, Y. Qiu, W. Jia, C. Xiao, L.M. Smith, S. Yang, E. Holmes, H. Tang, G. Zhao, J.K. Nicholson, L. Li, L. Zhao, Symbiotic gut microbes modulate human metabolic phenotypes, Proc. Natl. Acad. Sci. USA 105 (2008) 2117–2122. [18] C.X. Liu, Y.L. Song, M. McTeague, A.W. Vu, H. Wexler, S.M. Finegold, Rapid identification of the species of the Bacteroides fragilis group by multiplex PCR assays using group- and species-specific primers, FEMS Microbiol. Lett. 222 (2003) 9–16. [19] W. Ludwig, O. Strunk, S. Klugbauer, N. Klugbauer, M. Weizenegger, J. Neumaier, M. Bachleitner, K.H. Schleifer, Bacterial phylogeny based on comparative sequence analysis, Electrophoresis 19 (1998) 554–568. [20] W. Ludwig, O. Strunk, R. Westram, L. Richter, H. Meier, Y. Kumar, A. Buchner, T. Lai, S. Steppi, G. Jobb, W. Forster, I. Brettske, S. Gerber, A.W. Ginhart, O. Gross, S. Grumann, S. Hermann, R. Jost, A. Konig, T. Liss, R.

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

Lussmann, M. May, B. Nonhoff, B. Reichel, R. Strehlow, A. Stamatakis, N. Stuckmann, A. Vilbig, M. Lenke, T. Ludwig, A. Bode, K.H. Schleifer, ARB: a software environment for sequence data, Nucleic Acids Res. 32 (2004) 1363–1371. B.L. Maidak, J.R. Cole, T.G. Lilburn, C.T. Parker, P.R. Saxman, R.J. Farris, G.M. Garrity, G.J. Olsen, T.M. Schmidt, J.M. Tiedje, The RDP-II (Ribosomal Database Project), Nucleic Acids Res. 29 (2001) 73–74. P. McKenna, C. Hoffmann, N. Minkah, P.P. Aye, A. Lackner, Z. Liu, C.A. Lozupone, M. Hamady, R. Knight, F.D. Bushman, The macaque gut microbiome in health, lentiviral infection, and chronic enterocolitis, PLoS Pathog. 4 (2008) e20. X. Pang, D. Ding, G. Wei, M. Zhang, L. Wang, L. Zhao, Molecular profiling of Bacteroides spp. in human feces by PCR-temperature gradient gel electrophoresis, J. Microbiol. Methods 61 (2005) 413–417. E. Pruesse, C. Quast, K. Knittel, B.M. Fuchs, W. Ludwig, J. Peplies, F.O. Glockner, SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB, Nucleic Acids Res. 35 (2007) 7188–7196. L. Rigottier-Gois, V. Rochet, N. Garrec, A. Suau, J. Dore, Enumeration of Bacteroides species in human faeces by fluorescent in situ hybridisation combined with flow cytometry using 16S rRNA probes, Syst. Appl. Microbiol. 26 (2003) 110–118. P.D. Schloss, J. Handelsman, Introducing DOTUR, a computer program for defining operational taxonomic units and estimating species richness, Appl. Environ. Microbiol. 71 (2005) 1501–1506. P.D. Schloss, B.R. Larget, J. Handelsman, Integration of microbial ecology and statistics: a test to compare gene libraries, Appl. Environ. Microbiol. 70 (2004) 5485–5492. H.N. Shah, M.D. Collins, Proposal to restrict the genus Bacteroides (Castellani and Chalmers) to Bacteroides fragilis and closely related species, Int. J. Syst. Bacteriol. 39 (1989) 85–87. D.R. Singleton, M.A. Furlong, S.L. Rathbun, W.B. Whitman, Quantitative comparisons of 16S rRNA gene sequence libraries from environmental samples, Appl. Environ. Microbiol. 67 (2001) 4374–4376. C.J. Smith, D.R. Callihan, Analysis of rRNA restriction fragment length polymorphisms from Bacteroides spp. and Bacteroides fragilis isolates associated with diarrhea in humans and animals, J. Clin. Microbiol. 30 (1992) 806–812. A. Suau, R. Bonnet, M. Sutren, J.J. Godon, G.R. Gibson, M.D. Collins, J. Dore, Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut, Appl. Environ. Microbiol. 65 (1999) 4799–4807. J. Xu, M.K. Bjursell, J. Himrod, S. Deng, L.K. Carmichael, H.C. Chiang, L.V. Hooper, J.I. Gordon, A genomic view of the human–Bacteroides thetaiotaomicron symbiosis, Science 299 (2003) 2074–2076. P. Yarza, M. Richter, J. Peplies, J. Euzeby, R. Amann, K.H. Schleifer, W. Ludwig, F.O. Glockner, R. RosselloMora, The all-species living tree project: a 16S rRNAbased phylogenetic tree of all sequenced type strains, Syst. Appl. Microbiol. 31 (2008) 241–250.