Differences in faecal bacterial communities in coeliac and healthy children as detected by PCR and denaturing gradient gel electrophoresis


  • Editor: Willem van Eden

Correspondence: Yolanda Sanz, Instituto de Agroquímica y Tecnologia de Alimentos (CSIC), Apartado 73, 46100 Burjassot, Valencia, Spain. Tel.: +34 963900022; fax: +34 963636301;
e-mail: yolsanz@iata.csic.es


Coeliac disease (CD) is a chronic inflammatory disorder of the small intestinal mucosa. Scientific evidence supports a role of the gut microbiota in chronic inflammatory disorders; yet information is not specifically available for CD. In this study, a comparative denaturing gradient gel electrophoresis analysis of faecal samples from coeliac children and age-matched controls was carried out. The diversity of the faecal microbiota was significantly higher in coeliac children than in healthy controls. The presence of the species Lactobacillus curvatus, Leuconostoc mesenteroides and Leuconostoc carnosum was characteristic of coeliac patients, while that of the Lactobacillus casei group was characteristic of healthy controls. The Bifidobacterium population showed a significantly higher species diversity in healthy children than in coeliacs. In healthy children, this population was characterized by the presence of Bifidobacterium adolescentis. Overall, the results highlighted the need for further characterization of the microbiota in coeliac patients, and suggested a potential role of probiotics and/or prebiotics in restoring their gut microbial balance.


Coeliac disease (CD) is a chronic inflammatory disorder of the small intestinal mucosa that involves genetic and environmental factors. The classic form of the disease often manifests in early childhood (9–24 months) with gastrointestinal symptoms and malabsorption (Van Heel & West, 2006). CD is the result of an aberrant Th1−immune response to gluten peptides within the intestinal mucosa, where IFN-γ is the predominantly secreted cytokine, as well as an innate immune response mediated by IL-15 (Koning et al., 2005). Over the last decades, significant progress has been made in the understanding of the aetiology and pathogenesis of CD. In spite of that, little is known about the roles of environmental factors other than gluten in CD presentation (Barnich & Darfeuille-Michaud, 2007). It has been suggested that a transient infection could increase the permeability of the mucosal epithelial layer to gluten antigens by activating macrophages and dendritic cells with the production of pro-inflammatory cytokines (Kagnoff, 2005). Moreover, the inflammatory milieu originated by gluten antigens could lead to disturbances in the gut microbial composition that could in turn contribute to perpetuate inflammation (Collado et al., 2007). The development and maintenance of immune homeostasis depends on signals from the gut microbiota. Members of the genera Lactobacillus and Bifidobacterium are regarded as plausible significant players of gut health, and therefore are intensively investigated for probiotic uses (Isolauri et al., 2001; Thompson-Chagoyan et al., 2005). It has been demonstrated that components of the intestinal microbiota of animal models and inflammatory bowel disease (IBD) patients are involved in the abnormal T cell immune responses leading to loss of tolerance and mucosal inflammation (Barnich & Darfeuille-Michaud, 2007). Furthermore, the administration of probiotics has been found to exert beneficial effects in some disease models and IBD patients by decreasing the production of proinflammatory cytokines (e.g. IFN-γ, and TNF-α) and interfering with harmful bacterial adhesion (Dotan & Rachmilewitz, 2005). At present, little is known about the potential role of the microbiota in CD. Alterations in the composition of faecal short-chain fatty acids as well as the presence of rod-shaped bacteria associated with the mucosa have been reported in coeliac patients (Forsberg et al., 2004; Tjellstrom et al., 2005). More recently, a quantitative FISH analysis of the faecal microbiota of coeliac children and healthy controls revealed the existence of increased levels of Gram-negative bacteria and Staphylococcus, but significant differences in Lactobacillus and Bifidobacterium numbers were not detected (Collado et al., 2007). Here, denaturing gradient gel electrophoresis (DGGE) analyses using universal and primers specific for Bifidobacterium and Lactobacillus groups were carried out in order to detect differences in species composition that could be characteristic of an early stage of the disease, and support the use of probiotics and/or prebiotics for restoring the intestinal balance in these patients.

Materials and methods

Reference bacterial strains and growth conditions

The reference strains used as ladders for identification of Bifidobacterium spp. by DGGE were: Bifidobacterium adolescentis LMG 11037T, Bifidobacterium animalis ssp. animalis LMG 10508T, B. animalis ssp. lactis LMG 10140TBifidobacterium angulatum LMG 11039T, Bifidobacterium bifidum LMG 11041T, Bifidobacterium dentium LMG 11045T, Bifidobacterium longum biotype infantis LMG 11046T, B. longum biotype longum LMG 10497T, Bifidobacterium pseudocatenulatum LMG 10505T and Bifidobacterium ruminantium LMG 21811T. The reference strains used as DGGE ladders for identification of lactic acid bacteria were: Lactobacillus acidophilus DSM 20079T, Lactobacillus delbrueckii ssp. bulgaricus ATCC 11842T, Lactobacillus ruminis LMG 10756T, Lactobacillus curvatus LMG 9198T, Lactobacillus reuteri LMG 9213T, Lactobacillus brevis LMG 7944T, Lactobacillus gasseri LMG 9203T, Lactobacillus casei ATCC 393T, Lactobacillus sakei LMG 9468T, Lactobacillus plantarum ATCC 14917T and Pediococcus acidilactici LMG 11384T. All strains were grown in Man–Rogosa–Sharpe (MRS) medium, supplemented with 0.5 g L−1 cysteine to grow bifidobacteria, and incubated in anaerobiosis (Anaerocult, Merck, Darmstadt, Germany) at 37 °C.

Subjects and faecal sampling

Altogether, 20 children were included in the study: 10 coeliac patients (age 15–45 months; mean 28 months) and 10 age-matched healthy controls (age 11–40 months; mean 24 months). The group of patients included children who showed clinical symptoms of coeliac disease, positive serology markers (antigliadin, antiendomysial and antitransglutaminase antibodies) and signs of severe enteropathy by duodenal biopsy examination. The group of controls included healthy children with no known food intolerance. Samples from coeliac children were collected at the presentation of the disease, when they were still following a normal-gluten-containing diet. The children included in the study were not treated with antibiotics for at least 1 month before the sampling time. The study was approved by the Committee on Ethical Practice of General Universitary Hospital and CSIC, and children were enrolled in the study after written informed consent was obtained from their parents.

Faecal samples were collected from both groups of children and immediately maintained at 4 °C, under anaerobiosis (AnaeroGen, Oxoid, Hampshire, UK), and processed in <12 h as described previously (Collado et al., 2007). Briefly, faecal samples (2 g wet weight) were 10-fold diluted in phosphate-buffered saline [PBS, 130 mM sodium chloride, 10 mM sodium phosphate, (pH 7.2)], and homogenized in a Lab Blender 400 stomacher (Seward Medical, London, UK). An aliquot was submitted to low-spin centrifugation (300 g for 2 min) and the supernatant was maintained at −80 °C until analysed.

Nucleic acid extraction from bacterial cultures and faecal samples

The extractions of genomic DNA from pure cultures of Lactobacillus and Bifidobacterium used as reference strains as well as total DNA from faecal samples were carried out as described previously (Marzotto et al., 2006).

PCR amplification and DGGE analysis

PCR fragments of 200 bp representing total faecal bacterial were amplified with the universal primers HDA1-GC and HDA2 (Walter et al., 2000). PCR fragments of 340 bp representing Lactobacillus-related species were amplified with the primers Lac1 and Lac2-GC (Walter et al., 2001). PCR fragments of 520 bp representing species of Bifidobacterium were amplified with the primers Bif164 and Bif662-GC (Satokari et al., 2001). DGGE analysis of PCR amplicons was carried out on the Dcode Universal Mutation Detection System (Bio-Rad, Richmond, CA), essentially as described previously (Marzotto et al., 2006). The linear denaturing gradients of urea and formamide used for separation of amplicons from total microbiota, Bifidobacterium and Lactobacillus-related species were 30–50%, 45–55% and 30–50%, respectively. A 100% denaturant corresponds to 7 M urea and 40% (v/v) formamide. Selected unknown DGGE bands were excised from the denaturing gels and reamplified with the corresponding primers but without the GC-clamp. The PCR products were purified from agarose gel using the Qiaex II gel extraction kit (Qiagen, Hilden, Germany) and sequenced at the Bio Molecular Research Center (BMR), University of Padova (Italy). Search analyses to determine the closest relatives of the partial 16S rRNA gene sequences retrieved were conducted in GenBank using the blast algorithm.

Data analysis

DGGE patterns were analysed with the software package 1d-manager and lanemanager v2.0 (TDI, Barcelona, Spain). The similarities between DGGE profiles obtained with universal primers were determined using the Dice coefficient and the unweighted-pair group method with the arithmetic average (UPGMA) clustering algorithm. The number of bands of each individual in every DGGE profile was considered as an indicator of diversity of the faecal microbiota. Differences in diversity between both children groups (coeliacs and controls) were analysed by applying the Mann–Whitney U-test. Differences in species composition between both population groups were analysed using the χ2 tests. In every case, analyses were carried out with the statgraphics software (Manugistics, Rockville, MD), and statistical significance was established at P values <0.05.

Results and discussion

DGGE analysis with universal primers

DGGE profiles of PCR amplicons obtained with universal primers were complex and unique for each individual. No amplicon could be unequivocally associated with the presence or the absence of coeliac disease although two major clusters were differentiated: one grouping most coeliac patients and the other most healthy controls (Figs 1 and 2). The individual DGGE profiles showed similarities ranging between 43% and 94%. The diversity of the faecal microbiota, according to the number of bands in DGGE profiles, was significantly higher in coeliac children than in healthy controls (mean 8.40 vs. 7.2, P<0.05; Figs 1 and 2). In contrast, the biodiversity of active faecal bacteria of patients with ulcerative colitis was found to be lower than that of healthy subjects when analysed by temporal temperature gradient gel electrophoresis (Sokol et al., 2006). A recent metagenomic approach also showed a reduction in the diversity of the phylum Firmicutes in the faecal microbiota of Crohn's disease patients (Manichanh et al., 2006). Whether these changes are pathogenic, secondary or methodological still remains to be investigated.

Figure 1.

 DGGE profiles of total bacteria of faecal DNA from healthy (a) or coeliac (b) children amplified with the universal primers HDA1-GC and HDA2.

Figure 2.

 Dendogram derived from DGGE analysis of total faecal bacteria of healthy (H) and coeliac (C) children using universal primers based on Dice's similarity index and the UPGMA clustering algorithm.

DGGE analysis with Lactobacillus group-specific primers

The DGGE profiles of PCR amplicons obtained with Lactobacillus specific primers are shown in Fig. 3. A cluster analysis of the DGGE profiles obtained with Lactobacillus-group-specific primers was not performed due to their simplicity. The PCR amplicons that were identified by sequencing are shown in Table 1. The profiles of healthy children tended to show a lower diversity than those of coeliac patients but the differences were not significant (P>0.05; Fig. 3. and Table 1). The DGGE profiles of healthy subjects showed between one and four different bands and, in most of them, an amplicon corresponding to the L. casei reference strain was the unique or dominant one. In contrast, the DGGE profiles of coeliac children revealed the presence of one to six different Lactobacillus group-specific bands and species belonging to genera other than Lactobacillus were dominant. The prevalence of the L. casei group amplicon was significantly higher (P<0.05) in healthy children than in coeliacs (Table 2). According to the specificity of the Lac1 and Lac2-CG primers, the L. casei group might correspond to the species L. casei, Lactobacillus paracasei, Lactobacillus rhamnosus or Lactobacillus zeae (Walter et al., 2000; Ahrne et al., 2005). In contrast, the prevalence of Lactobacillus curvatus, Leuconostoc mesenteroides and Leuconostoc carnosum was significantly higher (P<0.05) in coeliac patients (Table 2). Most lactic acid bacterial species detected in coeliac patients were likely transient (allochthonous) (Reuter, 2001; Walter et al., 2001), while in healthy samples L. gasseri and species of the L. casei group could be regarded as both endogenous and food-related bacteria (Walter et al., 2000; Reuter, 2001; Ahrne et al., 2005). Although most of the allochthonous bacterial species found in both groups have been identified previously in faeces of healthy subjects (Walter et al., 2000, 2001; Reuter, 2001), it might be relevant to consider their impact on the consumer as they transit the gut in the digesta (Walter et al., 2001). The proportions of different Lactobacillus species might be functionally important because they largely influence the biochemistry, immunology and population dynamics of the host intestinal tract (Walter et al., 2000; Reuter, 2001). DGGE analysis of faeces from Crohn's disease patients and controls indicated that the diversity of lactic acid bacteria varied significantly between the groups (Scanlan et al., 2006). Alterations in the composition of Lactobacillus and other lactic acid bacterial species have also been described in ulcerative colitis patients (Bullock et al., 2004). Although direct evidence of the role of specific Lactobacillus species in human disease has not yet been provided, their composition was found to be different in colitic animal mice (IL-10 – knock out) and controls, and so was the anti-TNF-α activity of the isolated species (Pena et al., 2004).

Figure 3.

 DGGE profiles of 16S rRNA gene fragments of Lactobacillus spp. and related lactic acid bacteria of faecal DNA from healthy (a) or coeliac (b) children amplified with the primers Lac1 and Lac2-GC. Lanes M1 and M2, lactic acid bacteria identification ladder (L1, Lactobacillus plantarum; L2, Lactobacillus brevis; L3, Lactobacillus acidophilus; L4, Lactobacillus ruminis; L5, Pediococcus acidilactici; L6, Lactobacillus reuteri; L7, Lactobacillus sakei; L8, Lactobacillus curvatus; L9, Lactobacillus gasseri; L10, Lactobacillus delbrueckii ssp. bulgaricus; L11, Lactobacillus casei). Numbered arrows refer to sequenced fragments, whose amplicon IDs and closest relatives are shown in Table 1.

Table 1. Lactobacillus and related lactic acid bacterial species identified by sequencing the DGGE bands amplified from faecal DNA of healthy and coeliac children using the Lactobacillus group-specific primers
Amplicon ID*Closest relative (Accession number)Identity (%)
  • *

    Identification labels corresponding to DGGE bands shown in Fig. 3a and b.

  • Accession numbers of the closets relatives as determined by searching analyses conducted in GenBank using the blast algorithm (Koning et al., 2005).

Healthy children
 H1 Lactobacillus algidus (AB033209)97
 H2 Leuconostoc mesenteroides (AY675249)99
 H3 Leuconostoc inhae (AY675244)98
 H4 Lactobacillus sakei (AY204898)100
 H5 Lactobacillus curvatus (AY204894)100
 H6 Lactobacillus delbrueckii ssp. bulgaricus (AY735407)98
 H7 Pediococcus sp. (AM040655)100
 H8 Leuconostoc mesenteroides (DQ105649)99
Coeliac children
 C1 Weissella viridescens (M23040)100
 C2 Leuconostoc mesenteroides (AY675249)99
 C3 Leuconostoc carnosum (AB022925)99
 C4 Lactobacillus delbrueckii ssp. lactis (AY773950)96
 C5 Leuconostoc mesenteroides (AY675249)100
 C6 Lactobacillus fermentum (AJ575812)97
 C7 Lactobacillus delbrueckii ssp. lactis (AY773950)99
 C8 Lactobacillus paracasei (DQ199664)100
Table 2.   Lactic acid bacteria and Bifidobacterium species detected by DGGE analysis of faecal DNA from healthy and coeliac children using the Lactobacillus group-specific primers and the Bifidobacterium species-specific primers, respectively
Bacterial groupCoeliac
  • *

    Significant difference established at P<0.05 by using the χ2 test.

Lactic acid bacteria
 Lactobacillus algidus 0 (0%)2 (20%)0.14
 Lactobacillus delbrueckii ssp bulgaricus3 (30%)2 (20%)0.60
 Lactobacillus brevis 0 (0%)2 (20%)0.13
 Lactobacillus casei 0 (0%)5 (50%)0.01*
 Lactobacillus curvatus 4 (40%)0 (0%)0.02*
 Lactobacillus gasseri 0 (0%)1 (10%)0.30
 Lactobacillus fermentum 2 (20%)0 (0%)0.14
 Lactobacillus paracasei 2 (20%)0 (0%)0.14
 Lactobacillus plantarum 1 (10%)0 (0%)0.30
 Lactobacillus sakei 2 (20%)2 (20%)1.00
 Leuconostoc mesenteroides 7 (70%)2 (20%)0.02*
 Leuconostoc carnosum 4 (40%)0 (0%)0.02*
 Pediococcus acidolactici 1 (10%)0 (0%)0.30
 Pediococcus pentosaceus0 (0%)2 (20%)0.14
 Weissella viridescens 2 (20%)0 (0%)0.14
 Leuconostoc inhae 0 (0%)1 (10%)0.30
Bifidobacterium spp.
 Bifidobacterium adolescentis 0 (0%)4 (40%)0.03*
 Bifidobacterium bifidum 4 (40%)2 (20%)0.33
 Bifidobacterium infantis 3 (30%)3 (30%)1.00
 Bifidobacterium longum 8 (80%)10 (100%)0.14
 Bifidobacterium dentium 0 (0%)2 (20%)0.14
 Bifidobacterium pseudocatenulatum 6 (60%)7 (70%)0.64

DGGE analysis with Bifidobacterium species-specific primers

The DGGE profiles of PCR amplicons obtained with Bifidobacterium specific primers are shown in Fig. 4. The prevalence of different Bifidobacterium species is shown in Table 2. All Bifidobacterium species could be identified by comparing the migration distances of their respective PCR amplicons with those of reference strains used as ladders. In addition, the identity of six DNA bands was confirmed by sequencing (data not shown). The diversity of Bifidobacterium species was significantly higher (P<0.05) in healthy children than in coeliacs. Most DGGE profiles of the healthy subjects (7 out of 10) showed between three and four different Bifidobacterium species whereas most DGGE profiles of coeliac children (8 out of 10) only showed one or two species. The prevalence of B. longum, B. pseudocatenulatum and B. dentium tended to be higher in healthy children while that of B. bifidum did so in coeliac children, although these differences were not significant. B. dentium and B. adolescentis were not detected in any coeliac samples and for the last species, the difference was statistically significant (P<0.05, Table 2). Overall, the bifidobacterial population of healthy controls combined both infant- and adult-type features while that of coeliac patients mainly consisted of infant-type species (B. bifidum and Bifidobacterium infantis). Bifidobacterium species composition is thought to influence host-immune responses as a consequence of the differential immunomodulatory properties shown by diverse Bifidobacterium species (Young et al., 2004). Although not in every report (Penders et al., 2006), changes in the composition of Bifidobacterium species in the infant gut microbiota have often been related to the development of allergy (Kirjavainen et al., 2001, 2002). In addition, it has been speculated that typical adult-type Bifidobacterium species could favour Th2-biased immune responses characteristic of allergy inflammation (Young et al., 2004). Furthermore, the administration of probiotics and prebiotics to stimulate the proliferation of bifidobacteria has also been linked to beneficial effects in alleviating some IBD and allergic inflammation (Kirjavainen et al., 2002; Dotan & Rachmilewitz, 2005).

Figure 4.

 DGGE profiles of 16S rRNA gene fragments of Bifidobacterium spp. of faecal DNA from healthy (a) or coeliac (b) children amplified with the primers Bif164 and Bif662-GC. Lanes M1 and M2, Bifidobacterium identification ladder (B1, Bifidobacterium adolescentis; B2, Bifidobacterium infantis; B3, Bifidobacterium dentium; B4, Bifidobacterium animalis ssp. lactis; B5, Bifidobacterium bifidum; B6, Bifidobacterium longum; B7, Bifidobacterium pseudocatenulatum; B8, Bifidobacterium animalis ssp. animalis).

In summary, the diversity of total faecal microbiota as well as the species composition of lactic acid bacteria and Bifidobacterium have been shown to differ between coeliac children and age-matched controls. The results obtained highlight the need for further characterization of the gut microbiota in coeliac patients and suggest the potential role of probiotic and prebiotics in restoring their microbial gut balance.


This work was supported by grants AGL2005-05788-C02-01 from CICYT (Spain) and 200570F0091 from CSIC (Spain). The grant PTR95-0987.OP.01 of E. Sánchez (CICYT, Spain) and the mobility grant CTESPP/2005/055 of Y. Sanz from GVA (Valencia, Spain) are fully acknowledged.