Aims To characterise the predominant species of bacterial populations associated with duodenal biopsies of paediatric patients with active and treated coeliac disease.
Methods 20 biopsy specimens from patients with active coeliac disease, 12 from patients with treated coeliac disease, and eight from age-matched controls were evaluated for comparative purposes. Bacteroides, Bifidobacterium and lactic acid bacteria (LAB) populations were analysed by PCR-denaturing gradient gel electrophoresis using group-specific primers.
Results Bacteroides diversity was higher in biopsy specimens from controls than in those from patients with active and treated coeliac disease. Bacteroides distasonis, Bacteroides fragilis/Bacteroides thetaiotaomicron, Bacteroides uniformis and Bacteroides ovatus were more abundant in controls than in patients with coeliac disease (p<0.05). Bacteroides vulgatus was more frequently detected in controls than in patients with treated coeliac disease (p<0.01). Bacteroides dorei was more common in patients with active coeliac disease than in those with treated coeliac disease and control children (p<0.01). Bifidobacterium diversity was higher in patients with coeliac disease than in controls. Bifidobacterium adolescentis and Bifidobacterium animalis subsp lactis were more prevalent in patients with active coeliac disease than in patients with treated coeliac disease and control children. A higher LAB diversity was found in patients with treated coeliac disease and controls than in patients with active coeliac disease. Weissella spp and Lactobacillus fermentum were more frequently detected in patients with treated coeliac disease than in controls and patients with active coeliac disease.
Conclusions Bacteroides, Bifidobacterium and LAB populations in the duodenum of Spanish children with typical coeliac disease (active and treated) and controls differ in diversity and species composition; this could contribute to features of the disease.
- coeliac disease
- intestinal microbiota
Statistics from Altmetric.com
The human gastrointestinal tract hosts a complex microbial community (microbiota) composed of hundreds of different bacterial species.1 2 The microbiota contributes to regulating energy metabolism, and epithelial cell and gut-associated lymphoid tissue functions.3 4 The intestinal immune system has to maintain a balance between the need to respond to pathogens and tolerance to the presence of a large community of commensal bacteria, whose disruption might contribute to the pathogenesis of inflammatory conditions.5 6 In fact, alterations to the composition of the intestinal microbiota have been implicated in diseases such as allergies,7 chronic inflammatory bowel conditions8 and cancer.9
Coeliac disease is a chronic inflammatory disorder of the small intestine that presents in genetically predisposed individuals following gluten consumption.10 Removal of gluten from the diet is currently the only treatment available. In recent years, intestinal dysbiosis has been reported in patients with coeliac disease. The microbiota of these patients is characterised by a breakdown in the balance between potentially protective bacteria (eg, Lactobacillus spp and Bifidobacterium spp) and potentially harmful bacteria (eg, Bacteroides spp and Enterobacteriaceae).11 12 Some differences in bacterial species composition have also been detected in stool samples of patients with active coeliac disease compared with those of healthy controls13; however, patients on a gluten-free diet were not included and duodenal samples were not analysed, limiting the conclusiveness of that preliminary study. A further report provided information on only the faecal and duodenal Bifidobacterium species composition in patients with coeliac disease, analysed using real-time PCR.14
Denaturing gradient gel electrophoresis (DGGE) has been successfully applied as a rapid culture-independent method for the analysis of intestinal microbiota composition.15 16 DGGE is based on sequence-specific separation of equal-sized PCR products of the 16S rRNA gene on a polyacrylamide gel. DGGE facilitates the identification of a wider number of bacterial species within a genus than real-time PCR, in which the selection of specific primers restricts the number of species that can be detected. DGGE also allows the detection of non-cultivable bacterial species, which could represent more than 50% of intestinal bacteria.17 18
The objective of this study was to determine the species composition of the genera Bacteroides and Bifidobacterium, and lactic acid bacteria (LAB), in duodenal biopsy specimens of patients with active and treated coeliac disease, and in those of control children, by PCR-DGGE, in order to obtain more detailed information on the possible contribution of specific species to the disease.
Materials and methods
Biopsy specimens from three groups of children were included in this study: 20 patients with active coeliac disease, 12 patients with treated coeliac disease who had been following a gluten-free diet for at least 2 years, and 8 control children without a known gluten intolerance. Clinical characteristics of the children are shown in table 1. Untreated patients with coeliac disease were on a normal gluten-containing diet, showed clinical symptoms of the disease, positive coeliac serology markers (anti-gliadin and anti-transglutaminase antibodies) and signs of severe enteropathy, classified as type 3 according to the Marsh classification of coeliac disease based on examination of duodenal biopsies. Patients with treated coeliac disease had been on a gluten-free diet for at least 2 years, showed negative coeliac serology markers, and possessed normal mucosa or infiltrative lesions classified as type 0–1 according to the Marsh classification.19 Control children were presumptive patients with coeliac disease showing unspecific symptoms (eg, abdominal pain, difficulty to thrive, weight loss, etc), but the diagnosis was negative.
The children included in the study were not treated with antibiotics for at least 1 month before the sampling time. Ethics committee approval was secured for the study from Hospital Universitario de Valencia, Spain; Hospital La Fe Valencia, Spain, and Consejo Superior de Investigaciones Científicas, Madrid, Spain. Written informed consent was obtained from subjects (or their guardians).
Sampling preparation and DNA extraction
Duodenal biopsy specimens were obtained by upper intestinal endoscopy or capsule, frozen immediately at ‒80°C and kept until further processing. Each specimen (10–15 mg) was diluted 1:10 (w/v) in phosphate-buffered saline (pH 7.2), homogenised by thorough agitation and used for DNA extraction.
DNA from pure cultures of reference bacterial strains (table 2) and from biopsy samples were extracted by using the QIAamp DNA stool Mini kit (Qiagen, Hilden, Germany), following the manufacturer's instructions. Genomic DNA concentration was measured using the NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, Massachusetts, USA).
PCR amplification and DGGE analyses
The bacterial reference strains used as reference ladder for the identification of Bacteroides, Bifidobacterium and LAB species by PCR-DGGE are shown in table 2. The 16S rRNA gene of Bacteroides, Bifidobacterium and the LAB group present in biopsy specimens was partially amplified with the primer pair listed in table 3. For amplification of the 16S rRNA gene of Bacteroides, Bifidobacterium and the LAB group present in biopsy specimens, three separate PCR reactions were performed. Each PCR mixture (30 μl) contained 3 μl 10× buffer stock (containing 1.5 mM MgCl2), 1.5 μl bovine serum albumin (10 mg/ml), 0.5 mM of each deoxynucleoside triphosphate, 1 μM of each primer, 2.5 U Taq polymerase (Ecotaq; Ecogen, Barcelona, Spain) and 30 ng genomic DNA. One PCR core program was used for all amplifications: an initial denaturation step at 95°C for 3 min; 35 cycles of denaturation at 94°C for 30 s, annealing at a primer-specific temperature (table 3) for 1 min, and extension at 72°C for 1 min; and a final extension step at 72°C for 7 min. PCR amplification products of 293, 520 and 340 bp, representing Bacteroides, Bifidobacterium and LAB, respectively, were checked by electrophoresis in ethidium-bromide-stained 1.5% agarose gels and stored at −20°C. Visualisation of the LAB population required the inclusion of a nested PCR, using 27-δ and 1238-r primers23 as an initial amplification round.
DGGE analysis of PCR amplicons was carried out on the Dcode Universal Mutation Detection System (Bio-Rad, Richmond, California, USA), essentially as described previously.24 The linear denaturing gradients of urea and formamide used for PCR product separation were 25–45%, 40–50% and 25%–45% for Bacteroides, Bifidobacterium and LAB species, respectively. A 100% denaturant corresponds to 7 M urea and 40% (v/v) formamide. Selected unknown DGGE bands (S1–S11) were excised from the denaturing gels and checked using a new DGGE. Then, the PCR products were purified using the GFXtm PCR DNA and Gel Band DNA Purification Kit (GE Healthcare Chalfont St Giles, UK). The identity of the unknown DGGE bands was carried out by DNA sequencing using an ABI PRISM-3130XL Genetic Analyser (Applied Biosystems, Foster City, California, USA). Search analyses to determine the closest relatives of the partial 16S rRNA gene sequences retrieved were conducted in GenBank using the Basic Local Alignment Search Tool (BLAST) algorithm. Sequences with more than 97% similarity were considered to be of the same species.
Statistical and clustering analyses
Similarities between the banding patterns generated by PCR-DGGE analyses were analysed using the Dice coefficient and the unweighted-pair group method with the arithmetic average clustering algorithm and were shown graphically as a dendrogram. The Shannon–Wiener index of diversity (H′)25 was used to determine the diversity of taxa present in biopsy specimens of patients with active and treated coeliac disease, and in those of control children. The diversity index was calculated using the following equation:where s is the number of species, and pi is the proportion of the species i in the sample. Diversity data were non-uniformly distributed, so nonparametric analysis using a Mann–Whitney U test was performed. Analyses were carried out with the PAST software (PAlaeontological Statistics, University of Oslo, Oslo, Norway). Differences in species composition between groups with active and treated coeliac disease and control group DGGE profiles were analysed using χ2 tests. Analyses were carried out with the STATGRAPHICS software (Manugistics, Rockville, Maryland, USA). In every case, statistical significance was established at p values <0.05.
DGGE analysis of Bacteroides species
The DGGE profiles of PCR amplicons obtained with Bacteroides-specific primers from patients with active and treated coeliac disease and control children are shown in figure 1. It was not possible to differentiate between the species pair Bacteroides fragilis and Bacteroides thetaiotaomicron because the two DNA bands had the same migration distance. Clustering analysis (Dice/unweighted-pair group method with the arithmetic average clustering algorithm) of the DGGE Bacteroides profiles showed two differentiated clusters: cluster I grouped DGGE Bacteroides profiles from the control group samples at 72% similarity; and cluster II was divided into two major subgroups, one containing DGGE Bacteroides profiles from patients with treated coeliac disease clustered at 64% similarity, and another containing profiles from patients with active and treated coeliac disease clustered at 58% similarity (figure 2).
The DGGE Bacteroides profiles revealed that control children showed the greatest biological diversity, with five to seven bands (H′=1.73), followed by patients with active coeliac disease with two to four bands (H′=1.19, p=0.005) and patients with treated coeliac disease with one to three amplicons (H′=0.82, p=0.002).
The PCR amplicons that were identified by sequencing are shown in table 4. The prevalence of different Bacteroides species detected by PCR-DGGE in biopsy specimens from the children included in this study is shown in table 5. Amplicons most closely related to Bacteroides distasonis, Bacteroides fragilis/Bacteroides thetaiotaomicron and Bacteroides uniformis were significantly more abundant in DGGE profiles from samples of the control group than from those of patients with coeliac disease, regardless of the status of the disease (p<0.05). The prevalence of Bacteroides ovatus was also higher in biopsy samples from the control group than in those from treated and active coeliac disease groups (p<0.05). Significant differences were also observed for Bacteroides ovatus between biopsy specimens of patients with treated and active coeliac disease (p<0.05). Bacteroides vulgatus was more frequently detected in control children than in samples from patients with treated coeliac disease (p<0.01). Amplicons most closely related to Bacteroides dorei were more frequently detected in samples from patients with active coeliac disease than in those from patients with treated coeliac disease and control children (p<0.01), and Bacteroides plebeius was more frequently detected in biopsy samples from patients with active disease than in those from patients with treated coeliac disease (p=0.03). In contrast, Bacteroides coprocola was more frequently detected in samples from patients with treated coeliac than in samples from patients with active coeliac disease and control samples (p<0.01 and p=0.02, respectively).
DGGE analysis of bifidobacterial species
DGGE profiles for Bifidobacterium species contained zero to four bands (data not shown) and revealed that the Bifidobacterium community associated with biopsy samples in control children was significantly less diverse (0 or 1 amplicons, H′=0.00) than that from patients with treated (0–3 amplicons, H′=0.3, p=0.04) and active coeliac disease (0–4 amplicons, H′=0.48, p=0.03). Due to the simplicity of Bifidobacterium-specific DGGE profiles, no cluster analysis was performed.
The bifidobacterial PCR products identified by sequencing are shown in table 4. The prevalence of the different Bifidobacterium species detected by PCR-DGGE in biopsy samples of patients with active and treated coeliac disease and the control group is shown in table 5. Amplicons most closely related to Bifidobacterium adolescentis and Bifidobacterium animalis subsp lactis were more commonly detected in biopsy samples from patients with active coeliac disease than in samples from patients with treated coeliac disease and controls (p=0.03 and p<0.001, respectively). Bifidobacterium catenulatum was only detected in biopsy samples of the control group.
DGGE analysis of LAB
The DGGE profiles of PCR amplicons obtained with LAB primers from patients with active and treated coeliac disease and control children are shown in figure 3. Clustering analysis did not allow differentiation of samples according to disease status (data not shown).
DGGE profiles showed that the LAB diversity was higher in biopsy specimens from patients with treated coeliac disease (H′=3.58) and control children (H′=3.25) than in those from patients with active coeliac disease (H′=2.05, p=0.003 and p=0.01, respectively).
The PCR amplicons that were identified by sequencing are shown in table 4. The prevalence of the different LAB species detected by PCR-DGGE in biopsy samples of patients with active and treated coeliac disease and the control group is shown in table 5. The prevalence of amplicons related to the genus Weissella and to Lactobacillus fermentum was significantly higher in patients with treated coeliac disease than in controls and patients with active coeliac disease (p<0.05). The same trend was found for Lactobacillus reuteri and Pediococcus pentosaceus, but differences were only significant between patients with treated and active coeliac disease (p=0.01). The control group showed higher prevalences of Lactobacillus casei, Lactobacillus plantarum, Lactobacillus sakei, Lactobacillus ruminis, Leuconostoc citreum and Pediococcus acidilactici amplicons than children with active and treated coeliac disease, but these differences were not statistically significant.
The present study confirms the hypothesis that typical cases of coeliac disease are associated with intestinal dysbiosis, which involves specific bacterial species. In this study, a decrease in the Bacteroides diversity index was shown for duodenal samples from patients with coeliac disease compared with those from control individuals. This may mean that the increased Bacteroides counts detected in biopsy specimens from children with coeliac disease11 12 could involve only a limited number of Bacteroides species that could magnify their individual effects in these patients. Furthermore, it was shown that the duodenal Bacteroides species profiles of patients with treated coeliac disease were more similar to those of patients with active coeliac disease than to those of control children. This finding indicates that a gluten-free diet does not lead to a complete restoration of the balance of Bacteroides species in the duodenal mucosa of patients with coeliac disease. In patients with active coeliac disease, histological lesions can lead to nutrient malabsorption, which could result in modifications in the composition of the intestinal microbiota. However, alterations in the duodenal microbiota were detected in untreated and treated patients without mucosal lesions; therefore, nutrient malabsorption cannot be considered the only reason for such microbial changes. With a few exceptions,26 27 in patients with inflammatory bowel disease (IBD) the concentrations of Bacteroides associated with the mucosa were also shown to be higher and increased with the severity of the disease.28–31
Due to the different potential pathogenicity of Bacteroides species, related to a variety of virulence factors,32 33 the identification of the specific Bacteroides species associated with a given disease is epidemiologically important. In this context, some studies have suggested a relationship between enterotoxigenic B fragilis and clinically-active IBD in humans34–36 and in animal models.36 37 Patients with Crohn disease have also been shown to have a lower relative abundance of B uniformis and higher prevalence of B ovatus and B vulgatus.38 It has been proposed that B vulgatus and B ovatus could be implicated in the disruption of the integrity of the intestinal epithelial barrier, thereby contributing to the initiation of the inflammatory response of IBD.39–41 In contrast, the present study reports an increased prevalence of B dorei and a decreased prevalence of B ovatus and B uniformis in patients with coeliac disease.
The Bifidobacterium population associated with the duodenum of the children included in this study was relatively simple. This simplicity in mucosa-associated bifidobacterial populations has also been reported by Nielsen et al42 in biopsy samples from the ascending, transverse and descending parts of the colon of healthy individuals. In our study, the prevalence of B adolescentis and B animalis subsp lactis was higher in patients with active coeliac disease than in patients with treated coeliac disease and controls. In contrast, a previous study using real-time PCR showed that the prevalence of B animalis subsp lactis was higher in groups with treated coeliac disease groups than in those with active coeliac disease and control groups.14 Also, in that study, the prevalence of B catenulatum was found to be higher in biopsy samples from controls than in those from patients with active and non-active coeliac disease.14 B catenulatum was the only species found in control children in the present study. It is thought that the composition of Bifidobacterium species could influence host immune responses in inflammatory conditions.43 44 The prevalence of different Bifidobacterium species in the microbiota of infants has been suggested to be related to the incidence of allergic diseases.45 Bifidobacterium species have been demonstrated to have a species-specific and strain-specific influence on immunity, and could exert different effects on the T-helper 1 pro-inflammatory response characteristic of the coeliac disease.46 However, the association between the prevalence of different Bifidobacterim species and coeliac disease is still unclear.
Although the LAB present in the gut depend greatly on dietary intake, these species were analysed because they are also part of the indigenous microbiota, and are of interest for their possible roles as probiotics for the treatment of inflammatory conditions. DGGE analysis showed that LAB profiles were complex and specific to each host. The diversity index was higher for the patients with treated coeliac disease than for patients with active coeliac disease and controls. In healthy subjects, lack of stability in this bacterial group has been described, and has been attributed to the fact that a significant proportion of LAB populations are related to food-associated species.22 47 Lactobacillus species showing differences in prevalence between patients with treated coeliac disease and those with active coeliac disease have been reported to be regularly present in fermented foods and are considered to be allochthonous.48 Therefore, some of the differences detected in our study could be due to the disease, and also to dietary differences.
In summary, this study has demonstrated changes in diversity and species composition of the genera Bacteroides, Bifidobacterium and LAB in the duodenal microbiota of Spanish children with typical coeliac disease compared with those of control children. The changes detected in the microbiota of patients with active and treated coeliac disease do not seem to be completely dependent on the inflammatory status of the mucosa, particularly in the case of Bacteroides, whose potential role in the pathogenesis of the disease deserves further investigation.
Typical cases of coeliac disease are associated with intestinal dysbiosis, characterized by alterations in Bacteroides diversity and species composition.
Bacteroides diversity is reduced in duodenal biopsies from patients with active and treated coeliac disease in comparison with control children.
Adherence to a gluten-free diet does not lead to a complete restoration of the balance of Bacteroides species in the duodenal mucosa of patients with coeliac disease.
The findings suggest that these microbial alterations are not only a secondary consequence of the inflammation associated with the active phase of the disease.
Funding This work was supported by grants AGL2007-66126-C03-01/ALI and Consolider Fun-C-Food CSD2007-00063 from the Spanish Ministry of Science and Innovation. The scholarship to E Sánchez from Institute Danone is fully acknowledged.
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Hospital General Universitario, Hospital La Fe and CSIC.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.