SEARCH

SEARCH BY CITATION

Keywords:

  • irritable bowel syndrome;
  • microbiota;
  • mucosa associated

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

Background  There is increasing evidence to support a role for the gastrointestinal microbiota in the etiology of irritable bowel syndrome (IBS). Given the evidence of an inflammatory component to IBS, the mucosa-associated microbiota potentially play a key role in its pathogenesis. The objectives were to compare the mucosa-associated microbiota between patients with diarrhea predominant IBS (IBS-D), constipation predominant IBS (IBS-C) and controls using fluorescent in situ hybridization and to correlate specific bacteria groups with individual IBS symptoms.

Methods  Forty-seven patients with IBS (27 IBS-D and 20 IBS-C) and 26 healthy controls were recruited to the study. Snap-frozen rectal biopsies were taken at colonoscopy and bacterial quantification performed by hybridizing frozen sections with bacterial-group specific oligonucleotide probes.

Key Results  Patients with IBS had significantly greater numbers of total mucosa-associated bacteria per mm of rectal epithelium than controls [median 218 (IQR – 209) vs 128 (121) P = 0.007], and this was chiefly comprised of bacteroides IBS [69 (67) vs 14 (41) P = 0.001] and Eubacterium rectaleClostridium coccoides [52 (58) vs 25 (35) P = 0.03]. Analysis of IBS sub-groups demonstrated that bifidobacteria were lower in the IBS-D group than in the IBS-C group and controls [24 (32) vs 54 (88) vs 32 (35) P = 0.011]. Finally, amongst patients with IBS, the maximum number of stools per day negatively correlated with the number of mucosa-associated bifidobacteria (P < 0.001) and lactobacilli (P = 0.002).

Conclusions & Inferences  The mucosa-associated microbiota in patients with IBS is significantly different from healthy controls with increases in bacteroides and clostridia and a reduction in bifidobacteria in patients with IBS-D.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

Irritable bowel syndrome (IBS) is characterized by a triad of abdominal pain, bloating and change in bowel habit with an absence of any overt mucosal abnormality.1 Irritable bowel syndrome is the most frequent reason for referral to gastroenterology out-patient clinics2 and affects 10–20% of the population in the developed world.3 Irritable bowel syndrome is divided into three sub-groups: diarrhea predominant (IBS-D), constipation predominant (IBS-C) and a mixed type (IBS-M) where the stool frequency and consistency varies.4 Despite its high prevalence, the exact etiology remains only partially understood. Traditional models have focused on the role of stress, visceral hypersensitivity and dysfunction of the enteric nervous system and therapeutic paradigms have reflected this with the use of antispasmodics, antidepressants, cognitive behavioral therapy and newer serotenergic agents.5

A number of factors support the role for disturbances in the host gastrointestinal (GI) microbiota in IBS such as altered fecal microbiota in patients with IBS6 and the relationship with acute gastroenteritis.7 There is limited evidence suggesting a link between abnormal fermentation and IBS, either in the form of small intestinal bacterial overgrowth (SIBO)8 or colonic fermentation.9 Finally, there is a growing weight of evidence for the therapeutic manipulation of the host GI microbiota in IBS such as the use of probiotics,10 prebiotics11 and antibiotics.12

Direct evidence of an altered microbiota in IBS has thus far been limited to examination of the fecal microbiota. A microbiological analysis of fecal samples using 16s ribosomal DNA sequencing found IBS subgroups had distinct bacterial populations.13 The same group also analyzed fecal samples from all three IBS sub-groups and controls using quantitative real-time polymerase chain reaction and found the concentrations of lactobacilli were significantly lower in IBS-D compared with IBS-C although not when compared with controls.14 There have been several other studies of the fecal microbiota in IBS which have found either a reduction in lactobacilli15 or bifidobacteria,15–17 both of which are known to have anti-inflammatory effects and are used in probiotic preparations.18 Other studies have demonstrated increases in fecal enterobacteriaceae in IBS which include Escherichia coli and Salmonella spp.,17,19 and are potentially pro-inflammatory. Data from 16S ribosomal DNA sequencing studies have shown significant variation in the stool and mucosa-associated microbial populations in healthy people.20 Thus, to understand the complex interaction between the host immune system and the adjacent microbiota, analysis of fecal microbiota maybe insufficient. Furthermore, of the hundreds species of bacteria known to make up the human GI microbiota the majority are unculturable, making traditional microbiological techniques unsuitable for studying this environment.20

In addition to a disturbance in the GI microbiota, there has been increasing evidence of upregulation in the GI mucosal immune system in IBS. In particular increased numbers of mast cells and colon have been demonstrated and increases in intra-epithelial lymphocytes, enterochromaffin cells and human beta defensin-2.21 The cause of inflammation is not yet understood but a role for the gastrointestinal microbiota seems likely. Immune–microbiota interactions in the GI tract have been shown to be important in both health and disease.22 In particular, the microbiota has been shown to play a role in the pathogenesis of inflammatory bowel disease (IBD).23 Transgenic murine models of Crohn’s disease (CD) do not develop inflammation in germ-free conditions.24 Examination of the mucosa-associated microbiota in IBD has consistently revealed differences compared to controls.25 In particular, increases in bacteroides and a reduction in bifidobacteria in both ulcerative colitis and CD have been noted. It is therefore plausible that low grade mucosal inflammatory changes in IBS and thus symptoms, could be driven by alterations in the mucosa-associated microbiota.

The primary aim of this study was to use culture independent techniques to examine the mucosa-associated microbiota, in patients with IBS and healthy controls. The majority of previous studies grouped all patients with IBS together.15,17,19 In the few papers that compared IBS subgroups,13,14 individual symptoms were not evaluated, which may be important as sub-group analysis is limited by the broad definitions used in the Rome criteria. Therefore, in addition to comparing the mucosa-associated microbiota between diarrhea and constipation predominant IBS sub-groups, we have also correlated these with individual IBS symptoms.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

Patients were recruited from the outpatient and endoscopy clinics at Guy’s and St Thomas’ NHS Foundation Trust and the London Bridge Hospital. The three study groups were patients with IBS-D, IBS-C and healthy controls. All patients met the Rome III criteria for IBS, classification of patients into IBS sub-groups was based on the Bristol stool chart as per the Rome III criteria.4 Patients were excluded if they had a previous diagnosis of IBD, celiac disease, diverticular disease, microscopic or infectious colitis or had taken antibiotics or probiotics regularly in the last 3 months. Control patients were selected from individuals undergoing endoscopy for surveillance of polyps or cancer, had not been taking probiotics or antibiotics, had no history of recurrent abdominal pain, a normal bowel habit and no recurrent bloating or distension. In addition there colonoscopies were normal with no colorectal cancer or polyps found. Ethical approval for the study was obtained from the St Thomas’ Hospital research and ethics committee (06/Q0702/74).

Patients with IBS completed a validated symptom score26 and the hospital anxiety and depression scale.27 The symptom score comprised of four prompted visual analog scales each marked out of 100 and the typical number of days patients suffer with abdominal pain over a 10-day period (multiplied by 10) giving a total score out of 500. In addition, the maximum and minimum stool frequencies were assessed (in order to estimate severity of diarrhea and constipation, respectively), and totals calculated per month.

Prior to ileo-colonoscopy, all patients received bowel preparation consisting of two sachets of sodium picosulphate (Picolax®; Ferring Pharmaceuticals UK, London, UK) the day before and two senna capsules (Senokot granules; Reckitt & Benckiser, Slough, Berk, UK). In addition, patients followed a low residue diet the day prior to the colonoscopy and drank only clear fluids on the day of examination. In patients with a diagnosis of IBS, inflammation was excluded macroscopically by the endoscopist and microscopically by an independent histopathologist. At colonoscopy, rectal biopsies were taken using Radial Jaw 3 (Boston Scientific, St Alban’s, Herts, UK) biopsy forceps, from a well-prepared area of rectum with no obvious fecal debris. Samples were transferred from the forceps, without washing, into dry, sterile, cryotubes (Alpha Laboratories Ltd, Eastleigh, Hants, UK), snap frozen in liquid nitrogen and stored at −80 °C until analysis.

Mucosa-associated microbiota were analyzed using fluorescent in situ hybridization. Firstly, 6 μm sections were cut and mounted onto 4-well poly l-(+) lycine-coated slides (Hendley, Essex, UK). Sections were air dried, fixed in 4% paraformaldehyde solution, washed in phosphate-buffered saline (PBS – Oxoid, Basingstoke, Hants, UK) and permeabilized in 0.2% Triton X-100 (Calbiochem, Nottingham, UK). The sections were incubated between 37 and 50 °C in a 1.8 mol L−1 NaCl, 40 mmol L−1 Tris HCl, 2% SDS solution at pH 7.0 for 30 min. They were then hybridized overnight with oligonucleotide probes specific to bacterial 16S rRNA sequences (Microsynth, Balgach, Switzerland). The probes used are detailed in Table 1 and were those used in previously published experiments28 (Table 1). These probes have previously been shown to be altered in studies examining the fecal microbiota13–15,17 in IBS and in studies examining the mucosa-associated microbiota of patients with IBD.25,29

Table 1.   Oligonucleotide probe sequences
Target bacterial groupProbeSequence (5′-3′)Optimal hybridization temp.
Universal bacteriaEUB42GCT GCC TCC CGT AGG AGT47
BifidobacteriaBif 16443CAT CCG GCA TTA CCA CCC50
Clostridium coccoides–Eubacterium rectale clusterEREC 48228GCT TCT TAG TCA RGT ACC G50
Bacteroides–Prevotella clusterBac 30344CCA ATG TGG GGG ACC TT45
Lactobacillus/EnterococciLab 15845GGT ATT AGC AYT GTT TCC A45
Escherichia coliE. coli 153146CAC CGT AGT GCC TCG TCA TCA37

All probes were conjugated with Cy3 fluoroform and diluted to a concentration of 50 ng μL−1 in sterile 10 mmol L−1 Tris-HCl (Trizma; Sigma, Dorset, UK) buffered to pH 7.0 as per the manufacturers’ instructions. Following hybridization, slides were washed and mounted with fluorescent mountant containing antifade (Dako A/S, Ely, Cambridgeshire, UK).

All slides were viewed using a fluorescent Leica DMIRE2® confocal microscope. Five consecutive high power fields were selected along the epithelial edge, corresponding to a total of 1 mm of epithelium. A band of mucous lay adjacent, but not adherent to the colonic epithelial layer containing bacteria hybridized with probe. All fields were selected at random under light microscopy to minimize selection bias. The bacteria within the mucous layer were counted manually along the epithelial edge in each picture. All fields were counted in duplicate and an average was calculated. As these were frozen sections and not paraffin embedded, cutting artifact occasionally dislodged the mucin layer from the epithelium, in which case this was discarded and the next available section of epithelium was examined.

Statistical analysis

Statistical analysis was performed using spss version 15.0 (Chicago, IL, USA). Analysis demonstrated that the bacterial counts were not normally distributed. Therefore comparisons between bacterial numbers per mm of epithelium in all IBS patients and controls were performed using a Mann–Whitney U test and sub-group analysis was done using a Kruskall–Wallis test with inter-group analysis performed with a Mann–Whitney U test after Bonferoni post hoc correction. Correlations between bacterial numbers and symptom scores were performed using Pearson’s correlation coefficient. A P-value of <0.05 was accepted as statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

Clinical data

Rectal biopsies were obtained from 47 patients with IBS (IBS-D 27, IBS-C 20) and 26 controls. Patient demographics are summarized in Table 2. Patients with IBS were significantly younger and more were female than controls; however, we found no correlation between gender or age and any of the mucosa-associated bacteria probed. Patients with IBS-C had significantly higher mean bowel dissatisfaction (79.8 ± 13.6%vs 61.8 ± 16.5%P = 0.004) and anxiety scores (9.1 ± 3.9 vs 5.1 ± 3.3 P = 0.009) than patients with IBS-D. In contrast, the IBS-D group had a higher maximum stool frequency than IBS-C (121 ± 88 vs 47 ± 54 stools per month P = 0.01).

Table 2.   Patient demographics and mean symptom scores
 IBS-D (SD – unless otherwise stated)IBS-C (SD – unless otherwise stated)ControlP-value
  1. *One-way anova. Chi-squared test. All other statistical analysis was performed using a unpaired t-test.

Number272026 
Age, years (95% CI)36.2 (32.1–40.3)32.4 (28.1–36.7)46.1 (41.4–51.1)0.001*
Males (%)40 (57)66 (79)25 (43)0.010
Pain score28.4 (13.2–43.6)35.4 (23.7–47.0)ns
Number of days with pain in 10 day period5.3 (3.0–7.4)3.8 (1.9–5.6)ns
Bloating score43.6 (29.2–57.9)52.7 (37.5–67.9)ns
Bowel habit dissatisfaction score61.8 (51.2–72.2)79.8 (72.5–87.0)0.004
Impact on daily life score53.0 (42.1–63.9)60.8 (49.3–72.3)ns
Symptom severity score239 (183–295)266 (220–312)ns
Max stool freq (in a month)121 (65–177)47 (18–76)0.01
Min stool freq (in a month)34.2 (18–51)18.4 (2–35)0.16
Anxiety Score5.1 (3.0–7.2)9.1 (7.0–11.1)0.009
Depression Score2.8 (1.0–4.6)4.5 (3.7–6.4)ns

Bacterial counts

The spatial organization of the mucosa-associated bacteria appeared similar for all study groups. All bacteria were found in a band within the mucus layer adjacent to the epithelium of the rectal mucosa (Fig. 1). There were no differences in the thickness of the mucus layer between IBS and controls [in a sample of 10 IBS samples and 10 controls the mean mucous layer was 120 μm (95% CI 101–140 μm) and 125 μm (95% CI 102–146 μm) respectively, Student’s t-test P = 0.82]. There were scant bacteria directly adjacent to the epithelium or in the colonic crypts. Similarly a small number of bacteria were seen either in the epithelium or in the lamina propria and in such cases were associated with a broken area of the epithelium.

image

Figure 1.  Fluorescent in situ hybridization with the eubacteria probe showing the spatial organization of the mucosa-associated microbiota.

Download figure to PowerPoint

There were significantly more total bacteria [numbers per mm of epithelium (interquartile range, IQR)] in patients with IBS (218, IQR 209) than in controls (128, IQR 121) (P = 0.007). There were greater numbers of both bacteroides [69 (IQR 67) vs 14 (IQR 41 P = 0.001)] and Clostridia coccoides–Eubacterium rectale [52 (IQR 58) vs 25 (IQR 35 P = 0.003)], in the IBS group compared to controls, respectively. There were no differences in the numbers of bifidobacteria, lactobacillus-enterococci or E. coli between patients with IBS and healthy controls.

A comparison across all three study groups (IBS-D, IBS-C, healthy controls) is summarized in Table 3. There were significant differences in the number of mucosa-associated bifidobacteria (P = 0.01), bacteroides (P = 0.001) and clostridia (P = 0.003) across the IBS-D, IBS-C and control groups (Fig. 2). Comparison between sub-groups demonstrated greater numbers of total bacteria and bacteroides in IBS-D compared with controls and fewer bifidobacteria compared with IBS-C patients (Table 4). There were also greater numbers of bacteroides, C. coccoides–E. rectale and bifidobacteria in IBS-C compared with controls (Table 4).

Table 3.   Median (IQR) number of mucosa-associated bacteria per millimeter of epithelium
 IBS n = 47Control n = 26P-value Mann–Whitney U-test
Universal bacteria218 (209)128 (121)0.007
Bacteroides–Prevotella cluster69 (67)14 (41)0.001
Clostridium coccoides–Eubacterium rectale cluster52 (58)25 (35)0.003
Bifidobacteria38 (60)32 (35)0.31
Lactobacillus/Enterococci6 (22)10 (24)0.97
Escherichia coli8 (20)9 (19)0.41
image

Figure 2.  (A–C) Box and Whisker plot showing the median number of rectal mucosa-associated bifidobacteria (A), bacteroides (B) and Clostridia coccoidesE. rectale (C) in IBS-D, IBS-C and control patients. Boxes denote the 25th–75th centile and the Whiskers the extreme outlying values. An * denotes a significant difference.

Download figure to PowerPoint

Table 4.   Median (IQR) number of mucosa-associated bacteria per millimeter of epithelium by subgroup
 IBS-D n = 27IBS-C n = 20Control n = 26P-value Kruskall–Wallis
  1. P-values = 0.05 were considered significant. *P < 0.025 between IBS-D and controls. P < 0.025 IBS-C and controls. P < 0.025 between IBS-D and IBS-C all using Mann–Whitney U test after Bonferoni correction.

Universal bacteria*226 (184)203 (240)128 (121)0.025
Bacteroides–Prevotella cluster*,67 (62)73 (81)14 (41)0.001
Clostridium coccoides–Eubacterium rectale cluster†40 (57)68 (69)25 (36)0.003
Bifidobacteria†,‡24 (32)54 (88)32 (35)0.011
Lactobacillus/Enterococci6 (40)6 (23)9 (18)0.97
Escherichia coli12 (25)8 (17)10 (24)0.67

Correlations of bacterial numbers and symptom scores in the IBS sub-groups are shown in Table 5. There were significant, negative correlations between the stool frequency (illustrated by maximum and minimum number of stools per month) and numbers of total bacteria (P = 0.035), C. coccoides–E. rectale (P = 0.042) bifidobacteria (P < 0.001) and lactobacilli enterococci (P = 0.002). There was a positive correlation between anxiety and depression scores and the number of E. coli (P = 0.044 and P = 0.021 respectively). There was also a negative correlation between numbers of E. coli and pain scores (P = 0.014) and a negative correlation between total numbers of bacteria, C. coccoides–E. rectale and bifidobacteria and the average number of days with pain or discomfort (P = 0.009, P = 0.15, P = 0.31 respectively).

Table 5.   Correlations between individual symptom score outcomes and numbers of bacteria using Pearson’s correlation
 PainDays of pain out of 10BloatingBowel habit dissatisfactionImpact on daily lifeSSSMaximumMinimumAnxietyDepression
  1. *P < 0.05, **P < 0.01.

Total number of bacteria
 Correlation coefficient−0.442*−0.485**0.122−0.1310.190−0.246−0.400*−0.3060.2580.161
 Sig. (2-tailed)0.0180.0090.5360.5070.3330.2080.0350.1130.1860.412
Number of Bacteroides spp.
 Correlation coefficient−0.149−0.258−0.079−0.161−0.028−0.217−0.234−0.1290.105−0.145
 Sig. (2-tailed)0.4490.1840.6900.4130.8880.2680.2310.5130.5940.461
Number of Clostridia coccoides–Eubacteria rectale
 Correlation coefficient−0.189−0.453*0.2670.2170.220−0.068−0.386*−0.3150.1670.254
 Sig. (2-tailed)0.3350.0150.1700.2660.2610.7320.0420.1020.3950.193
Number of bifidobacteria
 Correlation coefficient−0.171−0.409*0.3050.1560.2290.014−0.698**−0.539**0.3290.329
 Sig. (2-tailed)0.3830.0310.1140.4280.2410.9440.0000.0030.0870.087
Number of E. coli
 Correlation coefficient−0.460*−0.2410.2380.0900.201−0.051−0.333−0.2250.384*0.435*
 Sig. (2-tailed)0.0140.2160.2220.6470.3060.7970.0840.2510.0440.021
Number of lactobacilli
 Correlation coefficient−0.124−0.1960.2390.0670.2480.096−0.550**−0.3450.2610.203
 Sig. (2-tailed)0.5310.3170.2210.7350.2020.6270.0020.0720.1790.300

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

This study is the first detailed analysis of the mucosa-associated microbiota in IBS. It not only examines the results by phenotypic sub-group but also correlates findings with individual symptoms and severity. The study has three key findings. Firstly, patients with IBS have a global expansion of the mucosa-associated microbiota and this expansion is largely comprised of bacteroides and clostridia. Secondly, analysis of phenotypic sub-group of IBS reveals that each has a distinct bacterial population. Finally, correlating bacterial counts with symptoms reveals that there was an inverse correlation between the number of many of the mucosa-associated bacteria including bifidobacteria and stools frequency.

The expansion of bacteroides and clostridia in IBS demonstrated by this study may be due to a number of different causes. Changes in mucous production, colonic motility or the mucosal microenvironment could all contribute to alterations in the mucosa-associated microbiota. In vitro studies designed to mimic the colonic mucous layer have demonstrated that bacteroides and clostridia avidly colonize mucin and can rapidly lyse the glycoproteins within.30 An increase in bacteroides and clostridia in patients with IBS may therefore be secondary to an increased production of rectal mucous. Many patients with IBS describe mucous associated with defecation irrespective of whether they have diarrhea or constipation.1 Secondly because the controls used were undergoing colonoscopy because they were at risk of colorectal cancer it is possible that they do not represent the microbiota of the general healthy population. One study examined numbers of mucosa-associated bifidobacteria in surgical specimens of patients with colorectal cancer, diverticular disease and IBD.31 They demonstrated a relative reduction in bifidobacteria in the cancer patients compared to those with diverticular disease. No studies have compared the microbiota in patients with a family history of cancer to those with established cancer.

It is possible that the global increase in bacterial numbers is secondary to changes in either the mucosal innate defensive systems or gastrointestinal motility. However, the IBS-D and IBS-C subgroups represent opposite ends of this spectrum and yet both had an increase in bacteria compared with controls. Fecal stasis in patients with IBS-C might increase the numbers of both luminal and mucosa-associated bacteria, whereas in contrast, in IBS-D, faster GI transit leading to alterations in mucosal environment such as changes in pH, oxygen concentrations and available fermentable substrates, could also lead to alterations in the microbiota.32 It might be that the rectal mucosa-associated microbiota in states of diarrhea becomes more like the right colon albeit studies of the mucosa-associated microbiota from the right and left colon in healthy controls have shown little differences within individuals.20,33 Changes in environment may be secondary to numerous causes such as differences in dietary intake in IBS, GI transit time or low grade inflammation. The increase in mucosal microbiota shown here may imply that symptoms of IBS are amenable to modification with antibiotics. The success of the recent trial of rifaximin in the treatment of IBS may in part be explained by a global increase in colonic microbiota.12 Rifaximin has a broad spectrum of antimicrobial activity and is likely to have a major impact on both the luminal and mucosa-associated microbiota. In addition it is not clear as to the mechanism of antibiotics in IBS, whether there is a reduction in mucosal inflammation, an alteration in colonic fermentation profile or a direct antimicrobial effect against specific pathogens.

Several of the earlier studies comparing the fecal microbiota of patients with IBS and controls did not distinguish between IBS sub-groups.15,17,34 However, it is likely that these differing phenotypes would lead to differences in the GI environment and that the mucosa-associated microbiota would reflect this. One of the few studies which did study IBS sub-groups, used 16S rDNA sequencing, in pooled fecal samples and discovered that each subgroup and controls had a distinct bacterial population.13 This study also examined the third IBS sub-group the mixed or alternating group (IBS-M or IBS-A depending on Rome criteria) in which bowel habit fluctuates from constipation to diarrhea which was not included in our study because insufficient numbers were recruited. Kassinen et al., did find that IBS-A patients had a distinct fecal microbiota from controls and the other IBS sub-groups and it would have been interesting to compare the mucosa-associated bacteria in this group.

Our findings in the rectal mucosa-associated microbiota support this, demonstrating significant differences in three of the key bacterial groups (bacteroides, clostridia and bifidobacteria), and in the total numbers of bacteria across all three study groups. Bacteroides and clostridia were found in greater numbers in the IBS-D and IBS-C groups compared to controls which is mirrored in the data comparing all IBS patients and controls. Increased levels of bacteroides have been associated with a ‘colonic dysbiosis’. In studies in which phyla level data is available, the relative quantities of Bacteroidetes and Firmicutes have been used as a marker of dysbiosis.13 Phyla level data are not available in our study in particular for Firmicutes, however as there appears to be an expansion of both clostridia and bacteroides it would appear unlikely that there is a significant alteration in their ratio. However bifidobacteria were found in the greatest number in IBS-C patients and lowest in IBS-D. The proportion of bifidobacteria of the total number of bacteria in the IBS-C group and controls was roughly similar (27% and 25% respectively) whereas in the IBS-D group it is less than half of this (11%).

Mucosal FISH has been used in a small group of IBS patients in one additional study in which they made up a sub-group of controls in a comparison of the mucosa-associated microbiota in IBD.29 However this study was therefore not statistically powered to examine the IBS group independently and did not differentiate between IBS sub-groups, there are a number of parallels between it and our study. There was a trend towards greater total numbers of mucosa-associated bacteria in the IBS group compared to controls, and there were greater numbers of clostridia in the IBS group than controls. In contrast, the paper also described a significant increase in the total mucosa-associated bacterial population in patients with IBD than in controls (and IBS patients), which showed greater adherence to the mucosal surface. In the IBD samples, >60% of this bacterial ‘biofilm’ hybridized with Bacteroides fragilis compared to <15% in controls. The oligonucleotide probe used to detect Bacteroides spp. in our study will hybridize with B. fragilis. Although our study does not include an IBD group, it is interesting that there appear to be parallels between patients with IBS and IBD. Given the increasing evidence suggesting an upregulation of the GI mucosal immune system in IBS an overgrowth or dysbiosis of bacteroides in the gut might drive the inflammation through microbial interactions with toll-like receptors and other pattern recognition receptors on the luminal surface of epithelial cells.

Given the prevalence of IBS and the current diagnostic criteria, it is likely that the clinical label ‘IBS’ encompasses a heterogeneous group of underlying etiologies. Causes of rapid transit include neuromotility disorders, stress, carbohydrate malabsorption, inflammation, SIBO and bile salt malabsorption. The aim of correlating symptoms with bacterial numbers was to investigate whether symptoms such as bloating or diarrhea might independently correlate with alterations in microbiota (Table 5). In particular we had hypothesized bloating might correlate to a particular bacterial genera, as studies have shown a link between increased intra-luminal bowel gas and bloating.35 The lack of correlation with bloating maybe because it is a small bowel rather than colonic phenomenon36 or that fermentation occurs primarily in the lumen rather than mucosal layer. Alternatively gas production may not play a role in bloating at all, with some studies suggesting bloating and distension is secondary to alterations in abdominal wall tone.37

However there were some significant correlations between numbers of mucosal bacteria and symptoms, the most striking of which is the relationship between stool output (as measured by the maximum number of stools per day) and several of the mucosa-associated species. The total number of bacteria, C. coccoides–E. rectale and in particular bifidobacteria (P < 0.001) and lactobacilli (P = 0.002) all negatively correlate with stool frequency. As the minimum number of stools per day is also a direct measurement of stool frequency bifidobacteria also negatively correlate with this variable (P = 0.003). There is relatively linear relationship between stool frequency and the number of bifidobacteria, suggesting that the overall reduction in bifidobacteria seen in the IBS-D group is simply secondary to diarrhea. However bifidobacteria are the only group of bacteria where this is the case, IBS-D patients have greater numbers of total bacteria, bacteroides and C. coccoides–E. rectal than controls. A further argument to suggest that alterations in the GI microbiota such as those found in this study may be primary rather than secondary is the success of therapeutic agents designed to alter the GI microbiota in IBS. There have been a number of randomized placebo controlled trials of antibiotics,12 probiotics10 and prebiotics11 which have shown symptomatic benefit in patients with IBS. The probiotic strain Bifidobacteria infantis 35624 was significantly superior to placebo in two randomized controlled trials in patients with IBS.38 Other probiotics that have shown benefit over placebo in RCT’s in IBS include a probiotic mix containing Lactobacillus GG, Lactobacillus rhamnosus LC705, Bifidobacterium breve Bb99, Propionibacterium freudenreichii spp. Shermanii JS39 and E. coli DSM17252.40 However the beneficial properties of probiotic bacteria are often strain specific and efficacy in the treatment of one disease cannot be extrapolated to another. Probiotic strains have been shown to often have highly specific properties such as amelioration of pain or reduction of inflammation. Thus interpretation of mucosa-associated microbiota with respect to probiotic strains and their benefits in IBS is complex. However the relative reduction in bifidobacteria in IBS-D in our study, when taken in context with earlier culture-based studies on fecal samples15,17 supports a role for the use of clinically proven probiotics as a therapeutic modality in IBS.

There are some limitations of this study. Firstly, there were differences in the demographics between patients with IBS and controls. These differences are explained by the epidemiology of IBS, which predominantly affects individuals under 50 and has a male to female ratio of 1 : 2. In contrast, finding patients who were undergoing full colonoscopy with normal stool output that matched the age of the IBS group was difficult. One pan-European study of the fecal microbiota of healthy controls found that there were significantly greater numbers of Bacteroides–Prevotella group in men than women.41 However, we found no correlation between either age or sex and any of the mucosa-associated microbiota in this study. In addition, the bias in the previous study is opposite to our findings of an increase in bacteroides in the female predominant IBS group. A disadvantage of studying the mucosa-associated microbiota using FISH is that it reveals bacteria at a genus level only. Studies have demonstrated that immune modulation can vary at the species and even strain level, rather than genus level, and can be highly specific to certain disease states. Whether it is because of altered protein expressions, enzymatic properties and even DNA sequences (such as CpG repeats) the benefits of probiotics appear to be extremely specific.

This study supports the hypothesis that differing IBS sub-groups have distinct microbial populations. It shows an increase in bacteroides and C. coccoidesE. rectale in IBS compared to controls and reduction in bifidobacteria. For the first time an attempt has been made to correlate these changes with some of the symptoms of IBS and in particular it has demonstrated a inverse relationship between numbers of lactobacillus enterococci, bifidobacteria and stool frequency. Although it is highly unlikely that all IBS patients have the same underlying pathophysiological mechanism this study supports a role for the GI microbiota in the etiology of IBS emphasizing its potential for therapeutic benefit.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

The funding for this study came from the Foundation for Allergy Information and Research.

Author contribution

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

GP, LP, KW and JS designed the study; GP & ML wrote the ethics application; GP, BN & JS recruited patients; GP & NR performed the FISH; GP analyzed the data and wrote the paper; JS, KW & JB edited the paper.

Competing interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References

The author’s declare that there are no competing interests.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Author contribution
  9. Competing interests
  10. References
  • 1
    Thompson WG, Longstreth GF, Drossman DA et al. Functional bowel disorders and functional abdominal pain. Gut 1999; 45(Suppl. 2): II43II47.
  • 2
    Bommelaer G, Poynard T, Le PC et al. Prevalence of irritable bowel syndrome (IBS) and variability of diagnostic criteria. Gastroenterol Clin Biol 2004; 28 (6-7 Pt 1): 55461.
  • 3
    Hungin AP, Whorwell PJ, Tack J et al. The prevalence, patterns and impact of irritable bowel syndrome: an international survey of 40,000 subjects. Aliment Pharmacol Ther 2003; 17: 64350.
  • 4
    Longstreth GF, Thompson WG, Chey WD et al. Functional bowel disorders. Gastroenterology 2006; 130: 148091.
  • 5
    Ford AC, Talley NJ, Spiegel BM et al. Effect of fibre, antispasmodics, and peppermint oil in the treatment of irritable bowel syndrome: systematic review and meta-analysis. BMJ 2008; 337: a2313.
  • 6
    Parkes GC, Brostoff J, Whelan K et al. Gastrointestinal microbiota in irritable bowel syndrome: their role in its pathogenesis and treatment. Am J Gastroenterol 2008; 103: 155767.
    Direct Link:
  • 7
    Marshall JK, Thabane M, Garg AX et al. Eight year prognosis of postinfectious irritable bowel syndrome following waterborne bacterial dysentery. Gut 2010; 59: 60511.
  • 8
    Pimentel M, Chow EJ, Lin HC. Eradication of small intestinal bacterial overgrowth reduces symptoms of irritable bowel syndrome. Am J Gastroenterol 2000; 95: 35036.
    Direct Link:
  • 9
    King TS, Elia M, Hunter JO. Abnormal colonic fermentation in irritable bowel syndrome. Lancet 1998; 352: 11879.
  • 10
    Moayyedi P, Ford AC, Talley NJ et al. The efficacy of probiotics in the therapy of irritable bowel syndrome: a systematic review. Gut 2010; 59: 32532.
  • 11
    Silk DB, Davis A, Vulevic J et al. Clinical trial: the effects of a trans-galactooligosaccharide prebiotic on faecal microbiota and symptoms in irritable bowel syndrome. Aliment Pharmacol Ther 2009; 29: 50818.
  • 12
    Pimentel M, Lembo A, Chey WD et al. Rifaximin therapy for patients with irritable bowel syndrome without constipation. N Engl J Med 2011; 364: 2232.
  • 13
    Kassinen A, Krogius-Kurikka L, Makivuokko H et al. The fecal microbiota of irritable bowel syndrome patients differs significantly from that of healthy subjects. Gastroenterology 2007; 133: 2433.
  • 14
    Malinen E, Rinttila T, Kajander K et al. Analysis of the fecal microbiota of irritable bowel syndrome patients and healthy controls with real-time PCR. Am J Gastroenterol 2005; 100: 37382.
    Direct Link:
  • 15
    Balsari A, Ceccarelli A, Dubini F et al. The fecal microbial population in the irritable bowel syndrome. Microbiologica 1982; 5: 18594.
  • 16
    Kerckhoffs AP, Samsom M, van der Rest ME et al. Lower bifidobacteria counts in both duodenal mucosa-associated and fecal microbiota in irritable bowel syndrome patients. World J Gastroenterol 2009; 15: 288792.
  • 17
    Si JM, Yu YC, Fan YJ et al. Intestinal microecology and quality of life in irritable bowel syndrome patients. World J Gastroenterol 2004; 10: 18025.
  • 18
    O’Mahony L, McCarthy J, Kelly P et al. Lactobacillus and bifidobacterium in irritable bowel syndrome: symptom responses and relationship to cytokine profiles. Gastroenterology 2005; 128: 54151.
  • 19
    Wyatt GM, Bayliss CE, Lakey AF et al. The faecal flora of two patients with food-related irritable bowel syndrome during challenge with symptom-provoking foods. J Med Microbiol 1988; 26: 2959.
  • 20
    Eckburg PB, Bik EM, Bernstein CN et al. Diversity of the human intestinal microbial flora. Science 2005; 308: 16358.
  • 21
    Langhorst J, Wieder A, Rueffer A et al. Activated innate immune system in irritable bowel syndrome? Gut 2007; 56: 13256.
  • 22
    Finegold S, Sutter VL, Mathisen GE. Normal indigenous intestinal flora. In: Hentges DJ, ed. Human Intestinal Flora in Health and Disease. New York: Academic Press, 1983: 331. Ref Type: Serial (Book, Monograph).
  • 23
    Sartor RB, Muehlbauer M. Microbial host interactions in IBD: implications for pathogenesis and therapy. Curr Gastroenterol Rep 2007; 9: 497507.
  • 24
    Sellon RK, Tonkonogy S, Schultz M et al. Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10-deficient mice. Infect Immun 1998; 66: 522431.
  • 25
    Mylonaki M, Rayment NB, Rampton DS et al. Molecular characterization of rectal mucosa-associated bacterial flora in inflammatory bowel disease. Inflamm Bowel Dis 2005; 11: 4817.
  • 26
    Francis CY, Morris J, Whorwell PJ. The irritable bowel severity scoring system: a simple method of monitoring irritable bowel syndrome and its progress. Aliment Pharmacol Ther 1997; 11: 395402.
  • 27
    Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67: 36170.
  • 28
    Franks AH, Harmsen HJ, Raangs GC et al. Variations of bacterial populations in human feces measured by fluorescent in situ hybridization with group-specific 16S rRNA-targeted oligonucleotide probes. Appl Environ Microbiol 1998; 64: 333645.
  • 29
    Swidsinski A, Weber J, Loening-Baucke V et al. Spatial organization and composition of the mucosal flora in patients with inflammatory bowel disease. J Clin Microbiol 2005; 43: 33809.
  • 30
    Macfarlane S, Woodmansey EJ, Macfarlane GT. Colonization of mucin by human intestinal bacteria and establishment of biofilm communities in a two-stage continuous culture system. Appl Environ Microbiol 2005; 71: 748392.
  • 31
    Gueimonde M, Ouwehand A, Huhtinen H et al. Qualitative and quantitative analyses of the bifidobacterial microbiota in the colonic mucosa of patients with colorectal cancer, diverticulitis and inflammatory bowel disease. World J Gastroenterol 2007; 13: 39859.
  • 32
    McCoubrey H, Parkes GC, Sanderson JD et al. Nutritional intakes in irritable bowel syndrome. J Hum Nutr Diet 2008; 21: 3967.
    Direct Link:
  • 33
    Green GL, Brostoff J, Hudspith B et al. Molecular characterization of the bacteria adherent to human colorectal mucosa. J Appl Microbiol 2006; 100: 4609.
  • 34
    Matto J, Maunuksela L, Kajander K et al. Composition and temporal stability of gastrointestinal microbiota in irritable bowel syndrome – a longitudinal study in IBS and control subjects. FEMS Immunol Med Microbiol 2005; 43: 21322.
  • 35
    Koide A, Yamaguchi T, Odaka T et al. Quantitative analysis of bowel gas using plain abdominal radiograph in patients with irritable bowel syndrome. Am J Gastroenterol 2000; 95: 173541.
    Direct Link:
  • 36
    Salvioli B, Serra J, Azpiroz F et al. Origin of gas retention and symptoms in patients with bloating. Gastroenterology 2005; 128: 5749.
  • 37
    Accarino A, Perez F, Azpiroz F et al. Abdominal distention results from caudo-ventral redistribution of contents. Gastroenterology 2009; 136: 154451.
  • 38
    Whorwell PJ, Altringer L, Morel J et al. Efficacy of an encapsulated probiotic Bifidobacterium infantis 35624 in women with irritable bowel syndrome. Am J Gastroenterol 2006; 101: 158190.
    Direct Link:
  • 39
    Kajanda K, Myllyluoma E, Rajilic-Stojanovic M et al. Clinical trial: multispecies probiotic supplementation alleviates the symptoms of irritable bowel syndrome and stabilizes intestinal microbiota. Aliment Pharmacol Ther 2008; 27: 4857.
  • 40
    Enck P, Zimmermann K, Menke G et al. A mixture of Escherichia coli (DSM 17252) and Enterococcus faecalis (DSM 16440) for treatment of the irritable bowel syndrome – a randomized controlled trial with primary care physicians. Neurogastroenterol Motil 2008; 20: 11039.
  • 41
    Mueller S, Saunier K, Hanisch C et al. Differences in fecal microbiota in different European study populations in relation to age, gender, and country: a cross-sectional study. Appl Environ Microbiol 2006; 72: 102733.
  • 42
    Amann RI, Binder BJ, Olson RJ et al. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microbiol 1990; 56: 191925.
  • 43
    Langendijk PS, Schut F, Jansen GJ et al. Quantitative fluorescence in situ hybridization of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples. Appl Environ Microbiol 1995; 61: 306975.
  • 44
    Manz W, Amann R, Ludwig W et al. Application of a suite of 16S rRNA-specific oligonucleotide probes designed to investigate bacteria of the phylum cytophaga-flavobacter-bacteroides in the natural environment. Microbiology 1996; 5: 1097106.
  • 45
    Harmsen HJM, Elfferich P, Schut F et al. A 16S rRNA-targeted probe for detection of lactobacilli and enterococci in faecal samples by fluorescent in situ hybridization. Microb Ecol Health Dis 1999; 11: 312.
  • 46
    Poulsen LK, Licht TR, Rang C et al. Physiological state of Escherichia coli BJ4 growing in the large intestines of streptomycin-treated mice. J Bacteriol 1995; 177: 58405.