Fecal Microbial Diversity in a Strict Vegetarian ... - Wiley Online Library

Microbiol. Immunol., 46(12), 819―831, 2002

Fecal Microbial Diversity in a Strict Vegetarian as Determined by Molecular Analysis and Cultivation Hidenori Hayashi*, Mitsuo Sakamoto, and Yoshimi Benno Japan Collection of Microorganisms, RIKEN, Wako, Saitama 351―0198, Japan Received June 26, 2002; in revised form, August 26, 2002. Accepted September 12, 2002

Abstract: Fecal microbial diversity in a strictly vegetarian woman was determined by the 16S rDNA library method, terminal restriction fragment length polymorphism (T-RFLP) analysis and a culture-based method. The 16S rDNA library was generated from extracted fecal DNA, using bacteria-specific primers. Randomly selected clones were partially sequenced. T-RFLP analysis was performed using amplified 16S rDNA. The lengths of T-RF were analyzed after digestion by HhaI and MspI. The cultivated bacterial isolates were used for partial sequencing of 16S rDNA. Among 183 clones obtained, approximately 29% of the clones belonged to 13 known species. About 71% of the remaining clones were novel “phylotypes” (at least 98% similarity of clone sequence). A total of 55 species or phylotypes were identified among the 16S rDNA library, while the cultivated isolates included 22 species or phylotypes. In addition, many new phylotypes were detected from the 16S rDNA library. The 16S rDNA library and isolates commonly included the Bacteroides group, Bifidobacterium group, and Clostridium rRNA clusters IV, XIVa, XVI and XVIII. T-RFLP analysis revealed the major composition of the vegetarian gut microbiota were Clostridium rRNA subcluster XIVa and Clostridium rRNA cluster XVIII. The dominant feature of this strictly vegetarian gut microbiota was the detection of many Clostridium rRNA subcluster XIVa and C. ramosum (Clostridium rRNA cluster XVIII). Key words: Vegetarian gut microbiota, 16S rDNA library, T-RFLP, Phylogenetic analysis

There are marked individual differences in the composition of human gut microbiota (10, 38). Dietary habits are considered to be one of the factors for the difference in microbiota (7, 16). A Western diet (e.g., high animal-protein, high fat and low fiber content) is considered to be high risk for colon cancer (22), while a vegetarian diet (e.g., little or no animal-protein, low fat and high fiber content) is considered low risk (16). There is a heated debate on the relationships between the diversity of human gut microbiota, dietary ingredients and the incidence of colon cancer. Several studies have analyzed the composition of fecal bacteria in the low-risk group (e.g., vegetarian, Japanese diet, and South African diet) (7, 9, 16, 21, 22). The results of these studies showed that Bacteroides spp. or Bifidobacterium spp. might be related to the incidence of colon cancer (16, 22). However, even after analysis of bacterial cultivation, it remains uncertain whether a particular type of bacterium is related to colon cancer.

Culture-based approaches have been used for analysis of human gut microbiota (9, 21). However, the number of bacteria that can be isolated and identified from the human large intestine is limited due to the anaerobic and complex environment. In fact, the bacteria that can be cultivated constitute only 20 to 40% of the total (10, 18, 32). These shortfalls have left many unresolved questions regarding the composition of human gut microbiota. The culture-independent approach based on molecular-biological techniques has allowed the phylogenetic analysis of bacterial 16S rRNA genes in environmental samples (1). Phylogenetic analysis based on PCR cloning strategy has been used to characterize human gut microbiota (12, 32, 35). Recently, we have reported 16S rDNA clone libraries and 16S rDNA sequences of cultivated isolates to define phylotypes and then compare the phylotype, distribution, and composition of isolates Abbreviations: CFB, Cytophaga-Flexibacter-Bacteroides; DDBJ, DNA Data Bank of Japan; EMBL, European Molecular Biology Laboratory; PCR, polymerase chain reaction; RDP, Ribosomal Database Project; T-RF, terminal restriction fragment; T-RFLP, terminal restriction fragment length polymorphism.

*Address correspondence to Dr. Hidenori Hayashi, Japan Collection of Microorganisms, RIKEN, 2―1 Hirosawa, Wako, Saitama 351―0198, Japan. Fax: ＋81―48―462―4619. E-mail: [email protected]

819

820

H. HAYASHI ET AL

and 16S rDNA clone libraries from three healthy Japanese individuals (10). Our results showed that the bacteria in the human large bowel could be classified into three major groups consisting of Bacteroides group, Clostridium rRNA cluster IV (4)/the Clostridium leptum group, and Clostridium rRNA subcluster XIVa (4)/ the Clostridium coccoides group. The gut microbiota represented by the 16S rDNA library was slightly more complex than that represented by the culture method and the composition of the bacteria differs widely among individuals. Furthermore, it was also clear that a large number of phylotypes present in the human large bowel remains unknown. However, human gut microbiota are still indistinct, since only a few samples were analyzed. Recently, terminal restriction fragment length polymorphism (T-RFLP) has often been used for analysis of bacterial communities. This tool has the advantage of allowing rapid analysis of complex microbial communities as terminal restriction fragments (T-RFs) patterns or profiles (15). We have recently used this method to characterize the human gut microbiota in healthy Japanese (M. Sakamoto, H. Hayashi and Y. Benno, unpublished data). In the present study, we describe gut microbiota in a strictly vegetarian individual based on the analysis of 16S rDNA library and T-RFLP of fecal specimen and 16S rDNA sequences of cultivated isolates. Materials and Methods Fecal culture. Fecal samples were collected from a consenting vegetarian (35-year-old woman) who had been taking vegetable soup (200 ml/day) only for 13 years. She had not been treated with any antibiotics for one year prior to the study. A 0.5 g fecal sample was suspended in dilution buffer (10). Then, 50 µl samples of appropriate dilutions were plated anaerobically on medium 10 (3) using the “plate-in-bottle” method (20). DNA extraction. The fecal sample (0.5 g) was suspended in 5 ml of buffer A (10 mM Tris-HCl and 50 mM EDTA, pH 7.5), and washed four times using buffer A. The sample was then resuspended in 5 ml of buffer A containing lysozyme (final concentration, 5 mg/ml), Nacetylmuramidase (final concentration, 0.5 mg/ml), and achromopeptidase (final concentration, 0.5 mg/ml) (Wako Pure Chemical Industries, Ltd., Osaka, Japan). After incubation at 37 C for 2 hr, proteinase K and sodium dodecyl sulfate were added to a final concentration of 2 mg/ml and 1% (wt/vol), respectively. The mixture was incubated at 60 C for 3 hr. The following operations were carried out as described previously (28). Isolation of DNAs from bacterial colonies was performed by the following method. A single bacterial colony was touched with a sterile loop. The loop was

rotated briskly in 100 µl of sterile phosphate-buffer saline (PBS), 100 µl of 10% Triton X100 was added, and the mixture was heated at 100 C for 5 min and then cooled. PCR amplification and cloning. Two universal primers 27F (5' AGAGTTTGATCCTGGCTCAG 3') and 1492R (5' GGTTACCTTGTTACGACTT 3') (17) were used in PCR to amplify the 16S rDNA coding region. Amplification reactions were performed in a total volume of 100 µl containing 250 ng of DNA extracted from fecal sample or 2 µl sample of DNA extracted from colonies, 2.5 U of TaKaRa Ex Taq, 10 µl of Ex Taq buffer, 8 µl of dNTP mixture (2.5 mM each) and 50 pmol of each primer. The reaction mixtures were amplified in a Biometra PCR TGRADIENT using the following program: 95 C for 3 min, followed by 15 cycles consisting of 95 C for 30 sec, 50 C for 30 sec, 72 C for 1.5 min, and a final extension period of 72 C for 10 min. The amplification from colonies was performed with 30 cycles. The amplified 16S rRNA genes were purified using an UltraClean PCR Clean-up Kit (Mo Bio Laboratories, Inc., Calif., U.S.A.) or MultiScreen-PCR (Millipore, Bedford, Mass., U.S.A.). Purified amplicon from the fecal sample was ligated into the plasmid vector pCR® 2.1, then transformed into One Shot® INVαF' competent cells using the Original TA Cloning Kit (Invitrogen, San Diego, Calif., U.S.A.). Plasmid DNA of selected transformants was purified using MultiScreen 96-well filter plates (Millipore). DNA sequencing and phylogenetic analysis. Plasmid DNAs from 16S rDNA libraries and the amplicon of DNA from colonies were used as templates for sequencing. An equal portion (about 500 bp) of 16S rDNA (Escherichia coli position 27 to 520) was used for sequence analysis. The dideoxy chain termination reaction was conducted with a double-stranded DNA template and 27F or 520R (5' ACCGCGGCTGCTGGC 3') (17) primer using the Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, Calif., U.S.A.), and products were analyzed on a model ABI PRISM 3700 DNA analyzer system (Applied Biosystems). Nucleotide sequences were analyzed with FASTA search (27). All sequences were examined for possible chimeric artifacts by the CHECK CHIMERA program of the Ribosomal Database Project (RDP) (19). The previously determined 16S rRNA sequences used for comparisons in this study were retrieved from DDBJ, EMBL, and GenBank nucleotide sequence databases. Sequence data were aligned with the CLUSTAL W (34) package and corrected by manual inspection. Nucleotide substitution rates (Knuc values) were calculated (13) and the phylogenetic trees were constructed using the neighbor-joining method (29). Bootstrap resampling analysis

MOLECULAR ANALYSIS OF A VEGETARIAN GUT MICROBIOTA

821

Fig. 1. Phylogenetic tree showing the relationship of 16S rDNA sequences in vegetarian fecal samples within the Clostridium rRNA cluster IV (Clostridium leptum group). The tree was constructed using neighbor-joining analysis based on 16S rDNA sequences. Bootstrap values (n＝100 replicates) of ≧50 are considered as percentage. The scale bar represents 0.02 substitutions per nucleotide position. Clones from the 16S rDNA library appear as bold-faced underlined letters (JW). Isolates are shown in italics and bold-faced letters in boxes (CJ). The intestinal clone and intestinal strain represent clones and isolates that were detected in our previous study (9).

822

H. HAYASHI ET AL

Fig. 2.


823

Fig. 2. Phylogenetic tree showing the relationship of 16S rDNA sequences in vegetarian fecal samples within the Clostridium rRNA subcluster XIVa (Clostridium coccoides group). The tree was constructed using neighbor-joining analysis based on 16S rDNA sequences. Bootstrap values (n＝100 replicates) of ≧50 are considered as percentage. The scale bar represents 0.02 substitutions per nucleotide position. Clones from the 16S rDNA library appear as bold-faced underlined letters (JW). Isolates are shown in italics and bold-faced letters in boxes (CJ). The intestinal clone and intestinal strain represent the clones and isolates that were detected in our previous study (9).

824

H. HAYASHI ET AL

(6) of 100 replicates was performed to estimate the confidence of tree topologies. The term “phylotype” has been used for clusters of clone sequences that differed from known species by about 2% and were at least 98% similar to members of their cluster (26). T-RFLP analysis. 27F and 1492R were used for TRFLP. 27F was labeled with 6-FAM (6-carboxyfluorescein, Applied Biosystems). PCR conditions were the same as those for amplification of 16S rDNA sequences from colonies. PCR products were purified by polyethylene glycol (PEG 6000) (11). Purified PCR products were digested with HhaI or MspI. Digested PCR products were analyzed by electrophoresis on ABI PRISM 310 Genetic Analyzer (Applied Biosystems) in Genescan mode. GS-500 ROX and GS-1000 ROX (Applied Biosystems) were used as the internal standards. Nucleotide sequence accession number. The nucleotide sequences reported in this paper will appear in the DDBJ, EMBL, and GenBank nucleotide sequence databases with accession numbers AB080845―AB080905. Results Identity of Cultured Bacteria The number of total bacteria in the fecal sample was 4.8×1011 cells/g (wet weight). A total of 48 isolates in the fecal sample on medium 10 were subjected to sequence analysis with subsequent on-line homology searches using the aforementioned databases. These isolates were divided into 22 species or phylotypes. Of these isolates, 38.8% (19 isolates) shared less than 98% identity with known cultured bacterial strains. The results of phylogenetic analysis are shown in Figs. 1 to 3 and Table 1. The majority of sequences (64.4%, 31 isolates) were within the Firmicutes. 22.9% (11 isolates) were phylogenetically within the Clostridium rRNA cluster IV (Fig. 1 and Table 1). The isolates related to this cluster were classified into 4 phylotypes. Figure 2 shows the phylogenetic relationships of colonies in the Clostridium rRNA subcluster XIVa. About 16% of the isolates belonged to this cluster. The isolates related to this subcluster were classified into phylotypes or species. Three isolates (CJ60, 72, and 84) were closely related to the Ruminococcus torques (sequence similarity, 100%). Figure 3 shows the phylogenetic relationships of the colonies among Cytophaga-Flexibacter-Bacteroides (CFB) (25). Ten isolates (20.4% of total isolates) were located within this group. Five isolates (CJ78, 80, 85, 86 and 89) had a sequence similarity to known species (Bacteroides fragilis and B. caccae) of greater than 98%. Five isolates (10.2% of total isolates) belonged to the

Clostridium rRNA cluster XVIII (4) (Table 1). All isolates in this cluster were closely related to type strain of Clostridium ramosum (sequence similarity, 99.4%). There were 5 isolates (10.2% of total isolates) in the genus Bifidobacterium that were distributed within two species. Five isolates were closely related to type strain of Bifidobacterium (designated as Bif.) pseudocatenulatum (sequence similarity, 99.2%). One isolate was related to type strain of Bif. infantis (sequence similarity, 98.6%). Cloned Fecal 16S rDNA Sequences The 16S rDNA library was constructed from the total community DNA collected from vegetarian fecal sample using universally conserved 16S rRNA-targeted PCR primers. A total of 183 sequences from this library were subjected to sequence analysis. On the basis of sequence similarities, the clones were classified into several clusters corresponding to the major phyla of the domain Bacteria (37): the Frimicutes, CFB, Actinobacteria, and Proteobacteria subclass (Table 1). We detected 55 species or phylotypes from 16S rDNA library. Twenty-nine percent (53 clones) of the total clones closely related (＞98%) to the type strain of 13 species. Seventy-one percent (130 clones) of the remaining clones belonged to 42 phylotypes (Table 2). Furthermore, 32 (73 clones) of these phylotypes were newly discovered phylotypes (Table 2). Phylogenetic Position of Clonal Sequences The results of phylogenetic analysis are shown in Figs. 1 to 3 and Table 1. The clones were affiliated with the following phyla: Firmicutes (90.2%), CFB (6.0%), Actinobacteria (0.5%) and Proteobacteria (3.3%) (Table 1). Figure 1 shows the phylogenetic relationship of the clones in Clostridium rRNA cluster IV. The 24 clones related to this cluster were classified into 1 species and 14 phylotypes. Twelve out of 14 phylotypes were newly detected in this study. Two clones (JW1B3) were related to type strain of R. bromii (sequence similarity, 98.3%). Figure 2 shows the phylogenetic relationship of the clones in Clostridium rRNA subcluster XIVa. The 109 clones related to this subcluster were classified into 3 species and 18 phylotypes. Thirteen out of 18 phylotypes were newly detected in this study. Sixteen clones (JW1B7) were closely related to type strain of R. torques and three isolates (CJ60, 72, and 84) (sequence similarity, 100%; 8.7% of total clone population). Clone JW1A5 (25 clones) was the most numerous in this cluster and was not related to known species. Figure 3 shows the phylogenetic relationship of the clones in CFB. The 11 clones related to CFB were


825

Fig. 3. Phylogenetic tree showing the relationship of 16S rDNA sequences in vegetarian fecal samples within the Cytophaga-FlexibacterBacteroides (CFB) phylum. The tree was constructed using neighbor-joining analysis based on 16S rDNA sequences. Bootstrap values (n＝100 replicates) of ≧50 are considered as percentage. The scale bar represents 0.05 substitutions per nucleotide position. Clones from the 16S rDNA library appear as bold-faced underlined letters (JW). Isolates are shown by italics and bold-faced letters in boxes (CJ). The intestinal clone and intestinal strain are clones and isolates that were detected in our previous study (9).

826

H. HAYASHI ET AL Table 1. Distribution of 16S rDNA clone and cultivated bacteria detected in fecal sample Group Firmicutes Clostridium rRNA cluster IV (Clostridium leptum group) Clostridium rRNA subcluster XIVa (Clostridium coccoides group) Clostridium rRNA cluster XV Clostridium rRNA cluster XVI Clostridium rRNA cluster XVIII Others Actinobacteria Bifidobacterium Others Cytophaga-Flexibacter-Bacteroides Proteobacteria Epsilon subdivision Total

Clones (%)

Cultivated bacteria (%)

24 (13.1)

11 (22.9)

109 (59.6) 0 (0) 3 (1.7) 22 (12.0) 7 (3.8)

8 (16.7) 2 (4.2) 5 (10.4) 5 (10.2) 0 (0)

1 (0.5) 0 11 (6.0)

5 (10.2) 2 (4.2) 10 (20.4)

6 (3.3) 183

0 (0) 48

Table 2. Diversity of species and phylotypes No. of known No. of novel phylotypes species Known phylotypea) Unknown phylotypeb) (number of clones) (number of clones) 183 13 (53) 10 (57) 32 (73) a) Phylotype identified in previous studies. b) Phylotype discovered in the present study.

No. of clones analyzed

Fig. 4. T-RFLP analysis of 16S rDNA by primers 27F and 1492R in the intestinal tract of our vegetarian subject. PCR products were digested with HhaI (top) and MspI (bottom). The major T-RF bands are indicated by arrows.

classified into 3 species and 5 phylotypes. Two out of 5 phylotypes were newly detected in this study. Three clones (JW1E2) were related to three isolates (CJ69, 74, and 90) (sequence similarity, 99.8%). Clone JW1C2 was closely related to type strain of B. fragilis and isolates (CJ78, 80, 85, and 86) (sequence similarity, 98.8% and 100%, respectively). Twenty-two clones belonged to the Clostridium rRNA

cluster XVIII (Table 1). Twenty out of 22 clones closely related to type strain of C. ramosum and 5 isolates (sequence similarity, 99.2% and 99.8%). Only one clone of Bifidobacterium was detected, which is one of the predominant species in human intestinal tract from this library (Table 1). This clone was related to type strain of Bif. infantis and one isolate (CJ75) (sequence similarity, 99.1% and 100%, respectively).


827

Fig. 5. Comparison of the phylogenetic diversity of a vegetarian and healthy men. a The microbiota of three healthy men were analyzed by 16S rDNA libraries as reported previously (9).

T-RFLP Analysis of Bacterial Population Present in Vegetarian Intestinal Tract Vegetarian gut microbiota were analyzed by T-RFLP. The T-RFLP profiles generated by HhaI digestions had three major T-RFs (Fig. 4). Three of these corresponded to Clostridium rRNA subcluster XIVa based on the result of a computer simulation, which was performed using 16S rRNA sequences registered into the RDP and determined by this study. Another major T-RF corresponded to Clostridium rRNA cluster XVIII (C. ramosum). The minor T-RFs of about 100 bp and 395 bp were considered to be the Bacteroides group and the Clostridium rRNA cluster IV, respectively. Two major TRFs were detected in the T-RFLP profiles generated by MspI digestion. The T-RF of around 220 bp was derived form the Clostridium rRNA subcluster XIVa. Another major T-RF (∼298 bp) was derived from the Clostridium rRNA cluster XVIII (C. ramosum). In addition, the TRFs of about 95 bp and 299 bp were considered to be the Bacteroides group and the Clostridium rRNA cluster IV, respectively. Discussion Previous studies have considered that gut microbiota of a strict vegetarian are a low risk for colon cancer (22). However, it is impossible to identify the species that are related to colon cancer, although many researchers analyzed low-risk groups using the culturebased method (7, 16, 22). One of the reasons for this

limitation is the difficulty in culturing many bacteria in the human intestinal tract. Based on this background, molecular analysis and cultivation were used in the present study to characterize the gut microbiota in a vegetarian who consumed only vegetable soup over a long period of time. We investigated the composition of the vegetarian gut microbiota by determining phylotypes, distribution, and similarities for 16S rDNA library and isolates from fecal specimens (Table 1). In addition, the human gut bacterial community was analyzed by T-RFLP (Fig. 4). We detected many phylotypes from the 16S rDNA library. Recently, we reported that 744 clones present in fecal samples of three healthy men could be classified into 130 species or phylotypes (10). Of these, 99 (74% of total clone population) were phylotypes. Suau et al. (32) and Hold et al. (12) also reported that approximately 75% of the total clone population detected from 16S rDNA libraries were phylotypes. These percentage values are in agreement with our data. Furthermore, several new phylotypes (new 16S rDNA sequences) were identified from the 16S rDNA library (Table 3). Most of them belonged to the Clostridium cluster. These may be specific phylotypes in vegetarians, and have had some influence on the ecology of the alimentary tract. Figure 5 shows the proportions of various bacteria in the intestinal tract of the vegetarian individual. The dominant feature of this library was the detection of many Clostridium rRNA subcluster IV, Clostridium rRNA subcluster XIVa, and C. ramosum (Clostridium

828

H. HAYASHI ET AL

Table 3. Homology values and frequencies of new phylotypes discovered in isolated colonies and retrieved clones from vegetarian intestinal tract Phylum Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Firmicutes Bacteroidetes Bacteroidetes Bacteroidetes Bacteroidetes Proteobacteria

Group Clostridium cluster IVa) Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium cluster IV Clostridium subcluster XIVab) Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium subcluster XIVa Clostridium cluster XVIII Unclass. Clost clusterc) Unclass. Clost cluster Unclass. Clost cluster Unclass. Clost cluster Unclass. Clost cluster Unclass. Clost cluster Unclass. Clost cluster Bacteroidaceae Bacteroidaceae Bacteroidaceae Bacteroidaceae Epsilon subdivision

Clones/ strains JW1D4 JW1C11 JW1D6 JW1G9 JW2A8 JW2D11 JW2G1 CJ31 CJ36 JW1B12 JW1C7 JW1H4 JW1D8 JW1C1 JW1A12 JW1D1 JW2H4 JW2C7 JW1A2 JW2G3 CJ67 JW1B8 JW1B11 JW1H7 JW2E12 JW1D7 JW1G2 JW2B4 JW1H11 JW1B2 JW2F12 JW2H12 JW1C9 CJ44 CJ47 JW2G9 JW1G4

Homology No. of (%) clones/strains Anaerofilum pentosovorans 86 2 Clostridium orbiscindens 89 1 Clostridium orbiscindens 95 1 Clostridium orbiscindens 92 1 Clostridium orbiscindens 93 2 Clostridium orbiscindens 88 2 Clostridium orbiscindens 97 1 Clostridium orbiscindens 94 4 Clostridium viride 90 1 Eubacterium siraeum 86 7 Fusobacterium prausnitzii 86 1 Desulfotomaculum guttoideum 97 4 Clostridium clostridiiformes 96 2 Clostridium indolis 94 2 Clostridium oroticum 94 6 Clostridium oroticum 94 1 Clostridium oroticum 94 1 Clostridium polysaccharolyticum 92 1 Clostridium symbiosum 96 1 Eubacterium contortum 93 1 Eubacterium contortum 97 1 Eubacterium oxidoreducens 93 8 Ruminococcus obeum 95 2 Ruminococcus obeum 93 5 Clostridium spiroforme 94 2 Anaerovorax odorimutans 89 1 Anaerovorax odorimutans 90 1 Clostridium cellobioparum 82 1 Clostridium sphenoides 83 1 Clostridium thermocellum 84 1 Fusobacterium prausnitzii 82 1 Ruminococcus obeum 94 1 Bacteroides putredinis 91 1 Bacteroides putredinis 92 1 Bacteroides putredinis 90 1 Bacteroides thetaiotaomicron 90 1 Helicobacter pylori strain J99 77 6 Nearest species

a)

Clostridium leptum group, b) Clostridium coccoides group, c) Unclass. Clost cluster, Unclassified Clostridium cluster. The underlined characters are isolated colonies.

rRNA cluster XVIII). These were also detected by TRFLP analysis (Fig. 4). However the frequency of Clostridium rRNA cluster IV detected by T-RFLP was lower than that detected by the 16S clone rDNA library. Bonnet et al. (2) reported that a lack of diversity when fecal DNA samples were amplified with a large number of PCR cycles. The amplification in T-RFLP was performed for 30 cycles. Therefore, a difference in the percentage of species between T-RFLP and library may appear. C. ramosum was isolated at high frequency in the culture-based method. Finegold et al. (7, 9) reported that C. ramosum was recovered from 52% of fecal samples of

vegetarians and was one of the major bacteria that belong to the Clostridium cluster, in the human intestinal tract. This species may play an important role within the vegetarian intestinal tract. Clostridium rRNA subcluster XIVa is one of the most important clusters in the human intestinal tract (Figs. 2 and 5, and Table 1). Eight human gut microbiota have been analyzed by the 16S rDNA library method (10, 12, 32, 35). The rate of this cluster ranged from 10 to 59%. In addition, the type of species that constitute the cluster differed among individuals (Fig. 2). The difference in the composition of species among individuals is similarly seen in other clusters


(Figs. 1 to 3). The diet consumed daily is considered as an important cause for the different composition of microbiota among different individuals. We detected two high clone number phylotypes (JW1A5 and JW1A1), which belonged to Clostridium rRNA subcluster XIVa. These phylotypes may play an important role in the vegetarian intestinal tract. It is not clear that the species or phylotypes are related to a vegetarian diet. However, vegetarian gut microbiota clearly differ from other human gut microbiota, which were previously analyzed using 16S rDNA libraries (10, 12, 32, 35). Moore and Moore (22) reported that a higher total number of Bifidobacterium group was associated with a higher risk of colon cancer. On the other hand, these genera were detected with high counts in rural Japanese people (low-risk diet group) (16). Noack-Loebel et al. (23) reported that many Bifidobacterium spp. were detected in a lacto-ovo-vegetarian diet consumed by children. We also detected many Bifidobacterium spp. in the culture-based method. On the other hand, few Bifidobacterium spp. were detected from the 16S rDNA library (only one clone) and T-RFLP. When Bifidobacterium spp. are detected using molecular methods, some biases of PCR may influence the results (32, 36). Our results indicate that the relation between Bifidobacterium spp. count and vegetarian diet is not clear. The Bacteroides group is also one of the important constituents of human gut microbiota (21). A high intake of meat and fat is considered a high risk for colon cancer, and fat is known to stimulate bile flow, which in turn specifically stimulates Bacteroides spp (5, 8). Many strains of Bacteroides spp. convert bile to metabolites and fecapentaenes, which are considered as cocarcinogens or mutagens (14). Koornhof et al. (16) reported that the proportions of B. vulgatus and B. distasonis correlated with the risk for colon cancer. On the other hand, an inverse relationship was found with B. fragilis. Moore and Moore (22) reported that B. vulgatus and B. stercoris were significantly associated with high risk for colon cancer as opposed to low risk, and analysis of total density of Bacteroides spp. showed only a slight increase in risk with increased proportions of these species. We also detected certain strains of B. fragilis using culturebased method and 16S rDNA library. Vegetarian gut microbiota in this study resembled gut microbiota of individuals at low risk of colon cancer (16, 22). Fusobacterium prausnitzii is one of the most frequent and numerous species detected in the human large bowel using 16S rDNA library and culture-based methods (approximately 4―10% of the total clone population and approximately 7% of total isolates) (21, 32, 33) as shown in Fig. 1. However, we did not detect F. prausnitzii in the fecal specimen of our strict vegetarian.

829

Koornhof et al. (16) reported that the incidence of F. prausnitzii in a high-risk group for colon cancer was higher than that in a low-risk group. Finegold et al. (9) also failed to detect this species in a vegetarian. These results suggest that the low numbers of this species may have affected the composition of fecal microbiota in a vegetarian with a low-risk diet. A subpopulation of Clostridium spp. and Ruminococcus spp. belonging to Clostridium rRNA cluster IV and subcluster XIVa and Bacteroides spp. has fibrolytic activities (cellulase and xylanase) (24, 30, 31). Furthermore, B. eggerthii, B. fragilis, B. ovatus, and B. thetaiotaomicron have xylanolytic activity (31). We detected some strains of B. fragilis by both methods. In fact, some isolates had xylanase or cellulase activities (data not shown). These species may play an important role in fiber degradation in a strict vegetarian. We used 16S rDNA library and T-RFLP together in our analysis. T-RFLP provides a rapid and reproducible way to compare microbial communities and assess a diversity of complex communities (15). Although major T-RFs have been identified species from computer simulation, minor T-RFs of about 500―750 bp have not been identified by species (Fig. 4). These minor T-RFs may be phylotypes which were not detected from the library. T-RFLP also has difficulty in identification of TRF. In conclusion, we demonstrated bacterial diversity in the gut microbiota of a vegetarian, using the 16S rDNA library, T-RFLP, and cultivation. Many phylotypes that had not been discovered previously were detected in our subject. Our findings indicate that many bacteria that are yet unknown are present in the intestinal tract, and that certain phylotypes may be closely related to vegetarian intestinal tract ecology. Several of our results were similar to those found in a low-risk cancer group. Further studies are necessary to determine the exact species related to colon cancer. This study was supported by a grant from the Special Postdoctoral Research Program of RIKEN, Saitama, Japan. We thank Drs. Kudo and Ohkuma and Ms. Yuzawa of RIKEN, for their help in DNA sequencing. We also thank Mr. Mitsui of Japan Collection of Microorganisms in RIKEN, for his help in phylogenetic analysis.

References 1) Amann, R.I., Ludwing, W., and Schleifer, K.H. 1995. Phylogenetic identification and in situ detection of individual microbial cell without cultivation. Microbiol. Rev. 59: 143― 169. 2) Bonnet, R., Suau, A., Doré, J., Gibson, G.R., and Collins, M.D. 2002. Differences in rDNA libraries of faecal bacteria

830

3)

4)

5) 6) 7)

8) 9)

10)

11)

12)

13)

14)

15)

16)

17)

18)

19)

H. HAYASHI ET AL derived from 10- and 25-cycle PCRs. Int. J. Syst. Evol. Microbiol. 52: 757―763. Caldwell, D.R., and Bryant, M.P. 1966. Medium without rumen fluid for nonselective enumeration and isolation of rumen bacteria. Appl. Microbiol. 14: 794―801. Collins, M.D., Lawson, P.A., Willems, A., Cordoba, J.J., Fernandez-Garayzabal, J., Garcia, P., Cai, J., Hippe, H., and Farrow, J.A. 1994. The phylogeny of the genus Clostridium: proposal of five new genera and eleven new species combinations. Int. J. Syst. Bacteriol. 44: 812―826. Drasar, B.S., and Hill, M.J. 1974. Human intestinal flora, p. 193―222, Academic Press, Inc., New York. Felsenstein, J. 1985. Confidence limits of phylogenies: an approach using the bootstrap. Evolution 39: 783―791. Finegold, S.M., Sutter, V.L., Sugihara, P.T., Elder, H.A., Lenhman, S.M., and Phillips, R.L. 1977. Fecal microbial flora in seventh day Adventist populations and control subjects. Am. J. Clin. Nutr. 30: 1781―1792. Finegold, S.M. 1977. Anaerobic bacteria in human disease, p. 586―588, Academic Press, Inc., New York. Finegold, S.M., Sutter, V.L., and Mathisen, G.E. 1983. Normal indigenous flora, p. 3―31. In Hentges, D.J. (ed), Human intestinal microflora in health and disease, Academic Press, New York. Hayashi, H., Sakamoto, M., and Benno, Y. 2002. Phylogenetic analysis of the human gut microbiota using 16S rDNA clone libraries and strictly anaerobic culture-based methods. Microbiol. Immunol. 46: 535―548. Hiraishi, A., Kamagata, Y., and Nakamura, K. 1995. Polymerase chain reaction amplification and restriction fragment length polymorphism analysis of 16S rRNA gene from methanogens. J. Ferment. Bioeng. 79: 523―529. Hold, G.L., Pryde, S.E. Russell, V.J., Furrie, E., and Flint, H.J. 2002. Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis. FEMS Microbiol. Ecol. 39: 33―39. Kimura, M. 1980. A simple method for estimating the evolutionary rate of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16: 111―120. Kingston, D.G.I., Van Tassell, R.L., and Wilkins, T.D. 1990. The fecapentaenes, potent mutagens from human feces. Chem. Res. Toxicol. 3: 391―400. Kitts, C.L. 2001. Terminal restriction fragment patterns: a tool for comparing microbial communities and assessing community dynamics. Curr. Issues Intest. Microbiol. 2: 17― 25. Koornhof, H.J., Richardson, D.M., Wall, D.M., and Moore, W.E.C. 1979. Fecal bacteria in south African rural blacks and other population groups. Isr. J. Med. Sci. 15: 335―340. Lane, D.J. 1991. 16S/23S rRNA sequencing, p. 115―175. In Stackebrandt, E., and Goodfellow, M. (eds), Nucleic acid techniques in bacterial systematics, Wiley, New York. Langendijk, P.S., Schut, F., Jansen, G.J., Raangs, G.C., Kamphuis, G.R., Wilkinson, M.H., and Welling, G.W. 1995. Quantitative fluorescence in situ hybridization of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples. Appl. Environ. Microbiol. 61: 3069―3075. Maidak, B.L., Cole, J.R., Parker, C.T., Jr., Garrity, G.M.,

20)

21)

22)

23)

24)

25)

26)

27)

28)

29)

30)

31)

32)

33)

34)

Larsen, N., Li, B., Lilburn, T.G., McCaughey, M.J., Olsen, G.J., Overbeek, R., Pramanik, S., Schmidt, T.M., Tiedje, J.M., and Woese, C.R. 1999. A new version of the RDP (Ribosomal Database Project). Nucleic Acids Res. 27: 171― 173. Mitsuoka, T., Morishita, Y., Terada, A., and Yamamoto, S. 1969. A simple method (“plate-in-bottle method”) for the cultivation of fastidious anaerobes. Jpn. J. Microbiol. 13: 383― 385. Moore, W.E., and Holdeman, L.V. 1974. Human fecal flora: the normal flora of 20 Japanese-Hawaiians. Appl. Microbiol. 27: 961―979. Moore, W.E., and Moore, L.H. 1995. Intestinal floras of populations that have a high risk of colon cancer. Appl. Environ. Microbiol. 61: 3202―3207. Noack-Loebel, C., Küster, E., Rusch, V., and Zimmermann, K. 1983. Influence of different dietary regimen upon the composition of the human fecal flora. Prog. Food Nutr. Sci. 7: 127―131. Ohmiya, K., Sakka, K., Karita, S., and Kimura, T. 1997. Structure of cellulases and their application. Biotechnol. Genet. Eng. Rev. 14: 365―414. Paster, B.J., Dewhirst, F.E., Olsen, I., and Fraser, G.J. 1994. Phylogeny of Bacteroides, Prevotella, and Porphyromonas spp. and related bacteria. J. Bacteriol. 176: 725―732. Paster, B.J., Boches, S.K., Galvin, J.L., Ericson, R.E., Lau, C.N., Levanos, V.A., Sahasrabudhe, A., and Dewhirst, F.E. 2001. Bacterial diversity in human subgingival plaque. J. Bacteriol. 183: 3770―3783. Pearson, W.R., and Lipman, D.J. 1988. Improved tools for biological sequence comparison. Proc. Natl. Acad. Sci. U.S.A. 85: 2444―2448. Sakamoto, M., Umeda, M., Ishikawa, I., and Benno, Y. 2000. Comparison of oral bacterial flora in saliva from a healthy subject and two periodontitis patients by sequence analysis of 16S rDNA libraries. Microbiol. Immunol. 44: 643―652. Saitou, N., and Nei, M. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4: 406―425. Salyers, A.A., Vercellotti, J.R., West, S.E.H., and Wilkins, T.D. 1977. Fermentation of mucin and plant polysaccharides by strains of Bacteroides from the human colon. Appl. Environ. Microbiol. 33: 319―322. Salyers, A.A., Balascio, J.R., and Palmer, J.K. 1981. Breakdown of xylan by enzyme from human colonic bacteria. J. Food Biochem. 6: 39―55. Suau, A., Bonnet, R., Sutren, M., Godon, J.-J., Gibson, G.R., Collins, M.D., and Dore, J. 1999. Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. Appl. Environ. Microbiol. 65: 4799―4807. Suau, A., Rochet, V., Sghir, A., Gramet, G., Brewaeys, S., Sutren, M., Rigottier-Gois, L., and Dore, J. 2001. Fusobacterium prausnitzii and related species represent a dominant group within the human fecal flora. Syst. Appl. Microbiol. 24: 139―145. Thompson, J.D., Higgins, D.G., and Gibson, T.J. 1994. CLUSTAL W: improving the sensitivity of progressive mul-

MOLECULAR ANALYSIS OF A VEGETARIAN GUT MICROBIOTA tiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673―4680. 35) Wilson, K.H., and Blitchington, R.B. 1996. Human colonic biota studied by ribosomal DNA sequence analysis. Appl. Environ. Microbiol. 62: 2273―2278. 36) Wintzingerode, F.V., Göel, U.B., and Stackebrandt, E. 1997. Determination of microbial diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbi-

831

ol. Rev. 21: 213―229. 37) Woese, C.R. 1987. Bacterial evolution. Microbiol. Rev. 51: 221―271. 38) Zoetendal, E.G., Akkermans, A.D., and De Vos, W.M. 1998. Temperature gradient gel electrophoresis analysis of 16S rRNA from human fecal samples reveals stable and host-specific communities of active bacteria. Appl. Environ. Microbiol. 64: 3854―3859.