| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Institute of Medical Microbiology, Canton Hospital Luzern, Switzerland
2 Medical Department I, University Hospital Dresden, Germany
3 Institute of Medical Microbiology and Hygiene, Technical University Dresden, Germany
4 Laboratory of Food Microbiology, Institute of Food Science and Nutrition, Swiss Federal Institute of Technology Zurich, ETH Center LFV B21, Schmelzbergstrasse 7, CH-8092 Zurich, Switzerland
Correspondence
Markus Schuppler
Markus.Schuppler{at}ilw.agrl.ethz.ch
Received 10 June 2005
Accepted 28 December 2005
Abbreviations: DIG, digoxigenin; IBD, inflammatory bowel disease; UC, ulcerative colitis.
The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA gene sequences determined in this study are AJ812137–AJ812212.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The above-mentioned studies clearly demonstrate that intestinal bacteria play a crucial role in the pathogenesis of IBD. Therefore, investigation of the mucosa-associated bacterial flora constitutes an important pre-requisite in order to be able to address the question of whether the enteric flora per se or a singular bacterial species contributes to IBD pathogenesis (Fiocchi, 1998).
The large intestine of humans harbours up to 1014 bacteria comprising an estimated 400–500 species (Berg, 1996), of which most indigenous microbial species remain unknown (Eckburg et al., 2005; Suau et al., 1999). The identification of bacteria on the basis of their 16S rRNA gene sequences using molecular techniques overcomes the limitations of culture-based techniques and represents a state-of-the-art tool for the culture-independent analysis of complex microbial communities (Amann et al., 1995).
The aim of this study was to analyse and compare the bacterial composition of the mucosa-associated microbial flora in mucosal tissue from patients with active UC and patients without IBD to determine the potential association of particular bacterial populations with intestinal inflammation in UC.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The colonoscopy preparation using Klean-Prep (Norgine) followed an evidence-based procedure (Ell et al., 2003) and was the same for all of the patients included in this study: patients started with a first lavage (2 l) the night before the endoscopy and received a second lavage (2–3 l) supervised by a hospital nurse on the endoscopy ward the following morning.
Biopsies were collected with a flexible videocoloscope (CFQ160AI; Olympus) and spiked biopsy forceps (FB24U; Olympus). In patients P77 and P178, biopsies were collected from macroscopically inflamed mucosal sites, as well as from adjacent mucosa that appeared normal macroscopically. For each patient, three to four biopsy samples from each location were pooled to minimize sampling bias. The sampled biopsies were immediately frozen in liquid nitrogen and stored at –80 °C until extraction of nucleic acids.
Amplification of 16S rRNA genes. Chromosomal DNA was extracted from bacterial cultures and gastrointestinal biopsy specimens using the DNeasy Tissue kit (Qiagen) following the manufacturer's recommendations. The 16S broad-range primers 63F (5'-CAGGCCTAACACATGCAAGTC-3') and 1387R (5'-GGGCGGWGTGTACAAGGC-3') were used to amplify almost-complete 16S rRNA genes from the mixed template preparations (Marchesi et al., 1998). Amplification was performed on a GeneAmp PCR System 2400 Thermal Cycler (PE Applied Biosystems). In brief, serial dilutions of the extracted DNA were used in 50 µl PCRs containing 1x GeneAmp PCR buffer comprising 1·5 mM MgCl2 (PE Applied Biosystems), 200 µM each dNTP, 1 µM each primer and 1·25 U AmpliTaq DNA Polymerase LD. PCR cycle conditions comprised an initial denaturation step of 2 min at 95 °C, followed by 30 cycles of denaturation for 60 s at 95 °C, annealing for 60 s at 55 °C and extension for 2 min at 72 °C, with a single final extension step of 10 min at 72 °C. Negative PCR controls containing no DNA were run in parallel. Amplified PCR products were analysed by 1·5 % agarose gel electrophoresis and purified using the QIAquick PCR Purification kit (Qiagen) following the manufacturer's specifications.
Cloning of PCR products and colony-blot analysis. Purified PCR products were cloned using the TOPO TA Cloning kit (Invitrogen) as indicated by the manufacturer. From each cloning reaction, 1000 recombinant clones were plated on to Luria–Bertani agar supplemented with 0·1 mg ampicillin ml–1.
To quantify the number of recombinant clones representing the 16S rRNA gene from Bacteroides spp. or Prevotella spp., colony hybridization using the 5' digoxigenin (DIG)-labelled oligonucleotide probe BAC303 (5'-CCAATGTGGGGGACCTT-3') (Manz et al., 1996) was performed. Bacterial colonies were transferred to positively charged nylon membrane discs (Millipore) and lysed according to standard protocols. In brief, two layers of Whatman 3MM paper were soaked for each of the following: denaturation solution (0·5 M NaOH, 1·5 M NaCl), neutralization solution (1·5 M NaCl, 1·0 M Tris/HCl, pH 7·5) and 2x SSC buffer (300 mM NaCl, 30 mM sodium citrate, pH 7·0). Membrane discs were placed colony-side up on filter paper soaked with denaturation solution for 15 min, followed by blotting of the membrane discs briefly on Whatman 3MM paper. The discs were then placed for 15 min on to prepared filter paper soaked in neutralization solution. After blotting of the membrane discs on Whatman 3MM paper, they were placed on to prepared filter paper soaked with 2x SSC buffer for 10 min. Fixation of the DNA to the nylon membranes was performed by baking the dried membrane discs for at least 30 min at 80 °C.
After immobilization of the DNA, proteinase K treatment was performed to remove cellular and agar debris. For this purpose, the membrane discs were placed on to a clean piece of aluminium foil, 0·5 ml proteinase K solution (2 mg ml–1) was added and evenly distributed, and the discs were incubated at 37 °C for 1 h. The membrane discs were blotted between Whatman 3MM paper and pressure was applied by passing a bottle over the area before the upper filter paper, with the debris sticking to it, was gently pulled off.
Hybridization was performed in a hybridization oven (Biometra) using roller bottles. Membrane discs were placed in a roller bottle and pre-hybridized with 60 ml hybridization solution without oligonucleotide probe [5x SSC, 0·1 % (w/v) N-laurylsarcosine, 0·02 % (w/v) SDS, 1·0 % (w/v) blocking reagent (Boehringer Mannheim)] for 1 h at 43 °C. The DIG-labelled oligonucleotide probe BAC303 was denatured by boiling for 5 min at 95–100 °C and subsequently diluted in pre-hybridization solution to a concentration of 10 pmol ml–1. Hybridization was performed using 6 ml of this hybridization solution at 43 °C. After 2 h, the membranes were washed twice in 50 ml washing buffer [5x SSC, 0·1 % (w/v) SDS] for 15 min at the hybridization temperature. Detection of DIG-labelled probe was conducted using anti-DIG antibodies coupled to alkaline phosphatase (DIG Luminescent Detection kit; Boehringer Mannheim) as indicated by the manufacturer. Chemiluminescence was determined by exposure of membranes to X-ray films (X-Omat AR; Kodak).
16S rRNA gene analysis. The 16S rRNA gene insert of recombinant clones was amplified and sequenced as described elsewhere (Schuppler et al., 2004). To determine the approximate phylogenetic relationships, the partial 16S rRNA gene sequences were subjected to a BLAST search against the EMBL and GenBank nucleotide sequence databases as described previously (Schuppler et al., 1995). By using >95 and >97 % sequence similarity, respectively, to delimit a genus and a species, the sequences were assigned to the respective bacterial phylotypes.
|
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows the relative abundance of the different phylogenetic groups represented by the clone libraries. The majority of the recovered sequences belonged to members of the Bacteroidetes and ‘Clostridia’. All 16S rRNA gene libraries contained members of the Bacteroidetes with Bacteroides vulgatus as the major representative. Sequences from other species such as Bacteroides caccae, Bacteroides distasonis, Bacteroides fragilis, Bacteroides merdae, Bacteroides ovatus, Bacteroides putredinis, Bacteroides thetaiotaomicron, Bacteroides uniformis and the recently described Bacteroides massiliensis (Fenner et al., 2005) were also frequently identified. There was one exception: the libraries from patient P178 were dominated by a 16S rRNA gene sequence from a hitherto-unknown member of the Bacteroidetes. Several clones revealed high sequence similarity to previously described sequences from hitherto-uncultured bacteria such as the uncultured bacterium adhufec367 reported by Suau et al. (1999).
|
For the Proteobacteria, most sequences were derived from the Gammaproteobacteria and were closely related to previously cultured bacteria including Acinetobacter johnsonii, Haemophilus parainfluenzae, Stenotrophomonas maltophilia and Escherichia coli, which was the dominant species in most libraries of the Gammaproteobacteria. However, unlike the Bacteroidetes and ‘Clostridia’, not all libraries harboured sequences from members of the Enterobacteriaceae. Members of the Alpha- and Betaproteobacteria were found only sporadically in the libraries and represented known species.
Beside sequences with high similarities to database sequences, the libraries also harboured numerous sequences revealing only weak homologies (<95 %). A considerable fraction of sequences derived from hitherto-uncultivated bacteria that are known only by their 16S rRNA gene sequences recovered using molecular approaches from gut samples of humans (Eckburg et al., 2005; Hold et al., 2002; Suau et al., 1999; Wang et al., 2003) or other mammals (Leser et al., 2002; Pryde et al., 1999; Salzman et al., 2002). Sequences from recently identified bacteria, such as a butyrate-producing bacterium cultured from a human gut sample (Barcenilla et al., 2000), were also found.
Comparison of the 16S rRNA gene clone libraries from the UC patients and the healthy individuals did not determine sequences that occurred exclusively in all libraries from the patients and not in the control group. However, an obvious difference was the abundance of members of the Bacteroidetes in the libraries from the UC patients. As demonstrated in Table 1
, all clone libraries from UC patients were characterized by a predominance of sequences from strict anaerobes belonging to the Bacteroides–Prevotella cluster of the Cytophaga–Flavobacter–Bacteroides group. Consequently, sequences from members of the Bacteroidetes represented 50–82 % of the total number of analysed clones in the libraries from UC patients, whereas the 16S rRNA gene libraries from the non-IBD patients P128 and P2110 contained 16 and 25 %, respectively.
Colony hybridization of 16S rRNA gene clone libraries
Sequence analysis of 60–80 recombinant clones from each of the established 16S rRNA gene clone libraries revealed a predominance of members of the Bacteroidetes in the mucosal biopsy samples from UC patients. In order to verify this observation, 1000 recombinant clones from the clone libraries of UC patient P178 and non-IBD patient P2110 were analysed by colony-blot hybridization. Oligonucleotide probe BAC303 was used for specific detection of clones containing 16S rRNA gene sequences from Bacteroides spp. or Prevotella spp. The specificity of BAC303 was demonstrated by the fact that all recombinant clones previously identified as Bacteroides or Prevotella sequences gave a positive hybridization signal with this probe. In contrast, none of the previously analysed clones harbouring sequences from other bacteria resulted in a positive hybridization signal. Due to the specificity of probe BAC303, which was restricted to sequences of the genera Bacteroides and Prevotella, colony hybridization resulted in slightly lower counts compared with the values for members of the Bacteroidetes determined by sequence analysis of the respective clone libraries. However, as is apparent from Table 2, the relative proportions proved to be consistent with the results from the comparative 16S rRNA gene analysis, confirming the abundance of Bacteroides spp. and Prevotella spp. (45 and 56 % for the inflamed and non-inflamed UC areas, respectively) in the clone libraries from UC patient P178 compared with the non-IBD patient P2110 (14 %).
|
). Almost exclusively, they harboured sequences from members of the Bacteroidetes and ‘Clostridia’, with only a few sequences representing other taxa. In contrast, the samples from patient P77 revealed differences in the bacterial composition of libraries from inflamed and non-inflamed mucosae. Both libraries harboured additionally sequences from E. coli, but revealed noticeable differences with respect to the qualitative and quantitative distribution of sequences from members of the Bacteroidetes, ‘Clostridia’ and the Enterobacteriaceae (Table 3). According to the results from the UC patients showing complete mucosal inflammation of the colon, the libraries from inflamed mucosa as well as from non-inflamed mucosa revealed an abundance of members of the Bacteroidetes.
|
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A comparison of results from the UC patients and non-IBD patients in this study did not lead to the identification of a specific sequence or group of sequences exclusively harboured in the samples from the UC patients. Therefore, it was not possible to relate certain bacterial species to the presence or absence of the colonic illness. This correlates well with reports from denaturing and temporal temperature gradient gel electrophoretic analyses of mucosal biopsies from healthy and diseased individuals (Seksik et al., 2003; Zoetendal et al., 2002). The generated profiles of the predominant bacterial communities appeared to be complex and unique for each individual, and no specific single band could be related to the presence or absence of colonic illness.
The most obvious difference in the mucosa-associated flora from the UC and non-IBD patients was the abundance of Bacteroides and Prevotella in the 16S rRNA gene libraries from the UC patients. It is well known that the quantitative distribution of sequences in 16S rRNA gene libraries does not necessarily reflect actual quantities of the respective bacterial populations in the original samples. However, all samples investigated in this study were processed equally in terms of sample collection, DNA isolation procedure, PCR amplification and the kits used for establishing the clone libraries. Therefore, the predominance of sequences from members of the Bacteroidetes in the libraries from the UC patients points to an abundance of members of this phylum in the mucosa of the UC patients compared with the non-IBD patients.
Although care was taken that all biopsy samples included in this study were obtained and investigated in a standardized manner, it must be taken into account that the mucosal flora is exposed to factors such as purgatives, laxatives, antimicrobial agents and diet, which may affect the residual flora prior to colonoscopy. We tried to avoid this by accurate choice of the individuals included in this study and by applying a standardized and evidence-based preparation procedure (Ell et al., 2003). None of the individuals included in the study received antimicrobial therapy within the 6 months prior to colonoscopy and none of the patients received a special diet that would provoke changes in the faecal flora. Two patients (P77 and P118) received therapy with mesalazine; no therapy was administered to any of the other patients included in this study. Despite reports that the preparation procedure mainly affects the luminal flora, whilst the mucosa-associated flora seems to be less influenced by such procedures (Marks et al., 1979), a potential influence on the residual flora could not be excluded totally. However, all patients included in this study received the same preparation prior to colonoscopy.
Increased counts for members of the Bacteroidetes in the colonic mucosa of IBD patients have also been reported from culture-based investigations. Ruseler-van Embden & Both-Patoir (1983) observed an increase in predominantly anaerobic flora, especially Bacteroides spp., in active Crohn's disease, whilst Hartley et al. (1992) isolated a higher level of Bacteroides spp. from patients with active UC, whether at first onset or upon release, than from patients with inactive disease. Swidsinski et al. (2002) reported a preferential increase in Bacteroides spp. and E. coli in IBD patients. Similar to the results from this study, the highest mean concentrations for Bacteroides spp. were reported for patients with UC or Crohn's disease. Interestingly, the abundance of members of the Bacteroidetes in the mucosa of UC patients was not restricted to inflamed mucosal sites. This is in concordance with the results from this study, which revealed high levels of members of the Bacteroidetes for both inflamed and non-inflamed mucosal areas. Furthermore, Swidsinski et al. (2002) found no significant differences in the bacterial spectrum between inflamed and non-inflamed regions of the mucosa, but demonstrated higher numbers of bacteria in the mucosa-associated flora of UC patients compared with healthy individuals in correlation to the severity of the disease. The observation of low numbers of members of the Bacteroidetes in the mucosal tissue from the healthy controls in this study is also consistent with other reports on the composition of the mucosal flora of healthy humans (Ott et al., 2004).
Evidence for a pro-inflammatory role for members of the Bacteroidetes also comes from animal models for IBD where Bacteroides spp. appear to play a pivotal role in the induction of colitis (Onderdonk et al., 1983; Rath et al., 1996). The incidence of members of the Bacteroidetes would also explain the therapeutic response to metronidazole in IBD patients and animal models for IBD. Metronidazole is particularly active against strictly anaerobic bacteria and it is reported that treatment of active disease in IBD patients with metronidazole results in complete or partial clinical remission and also prevents recurrence after ileal resection (Greenbloom et al., 1998; Prantera et al., 1998; Rutgeerts et al., 1995). Furthermore, the protective effect of metronidazole in experimental UC has been demonstrated and suggests that anaerobic bacteria play a role in the initial events of experimental colitis in the guinea pig model (Onderdonk et al., 1978).
The mechanisms by which Bacteroides spp. may contribute to the chronic inflammatory process are chiefly unknown. Breeling et al. (1988) reported an association of outer-membrane antigens from B. vulgatus with carrageenan-induced colitis in guinea pigs. Another possible mechanism by which members of the Bacteroidetes probably contribute to the pathogenesis of colitis is by the production of mucin-degrading sulphatases. Elevated levels of bacterial mucin-desulphating sulphatases have been reported for patients suffering from active UC (Tsai et al., 1995). Furthermore, the existence of enzymes that will partially desulphate mucins has been demonstrated for B. thetaiotaomicron, B. fragilis (Tsai et al., 1992) and for Prevotella spp. (Wright et al., 2000), suggesting that members of the Bacteroidetes could contribute to chronic inflammation by an impairment of the barrier function of the epithelial cell layer.
E. coli is also frequently encountered as potentially being involved in the pathogenesis of UC, and several studies on animal models for IBD support an association of E. coli with the disease (Schuppler et al., 2004; Waidmann et al., 2003). However, the detection of sequences from E. coli in the samples from the non-IBD patients together with the observed absence of E. coli in several samples from UC patients does not substantiate a primary role for E. coli in human UC. Mycobacterium avium subsp. paratuberculosis, Helicobacter spp., Listeria monocytogenes and several other bacterial pathogens have also often been related to UC. In particular, Campylobacter, Salmonella, Shigella and Yersinia have been reported as responsible for an inflammatory relapse in UC (Campirei & Gionchetti, 2001), but the 16S rRNA gene libraries established in this study did not contain sequences from known bacterial pathogens.
In conclusion, the results from this study point to an increase in the abundance of members of the Bacteroidetes in the colonic mucosa of UC patients. This is an important finding, as it suggests that certain bacterial populations are altered in active UC and therefore may play a role in the pathogenesis of the disease. However, the observation of an increase in members of the Bacteroidetes in the mucosal flora of IBD patients in this and other studies does not necessarily establish an aetiological connection. Therefore, it cannot be deduced whether the increase in anaerobes in the mucosal flora is causative or a consequence of the disease and whether specific members of the Bacteroidetes play a primary role in initiation of the disease. Further studies of an extended number of patients are needed to determine to what extent the changes reported in this study are primary or secondary to alterations of the mucosal microenvironment. In order to investigate the aetiological mechanisms of alterations in the bacterial microflora, longitudinal studies may provide important information.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
ß-deficient mice fail to develop colitis in the absence of a microbial environment. Am J Pathol 150, 91–97.[Abstract]This article has been cited by other articles:
|
|
 |
|
  J. F. Petrosino, S. Highlander, R. A. Luna, R. A. Gibbs, and J. Versalovic Metagenomic Pyrosequencing and Microbial Identification Clin. Chem., May 1, 2009; 55(5): 856 - 866. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Nadal, E. Donant, C. Ribes-Koninckx, M. Calabuig, and Y. Sanz Imbalance in the composition of the duodenal microbiota of children with coeliac disease J. Med. Microbiol., December 1, 2007; 56(12): 1669 - 1674. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Zhang, B. Liu, Y. Zhang, H. Wei, Y. Lei, and L. Zhao Structural Shifts of Mucosa-Associated Lactobacilli and Clostridium leptum Subgroup in Patients with Ulcerative Colitis J. Clin. Microbiol., February 1, 2007; 45(2): 496 - 500. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
