Changes in bacterial populations termed “dysbiosis” are thought central to ulcerative colitis (UC) pathogenesis. In particular, the possibility that novel Helicobacter organisms play a role in human UC has been debated but not comprehensively investigated. The aim of this study was to develop a molecular approach to investigate the presence of Helicobacter organisms in adults with and without UC.
A dual molecular approach to detect Helicobacter was developed. Oligonucleotide probes against the genus Helicobacter were designed and optimised alongside a validation of published H. pylori probes. A comprehensive evaluation of Helicobacter genus and H. pylori PCR primers was also undertaken. The combined approach was then assessed in a range of gastrointestinal samples prior to assessment of a UC cohort. Archival colonic samples were available from 106 individuals for FISH analysis (57 with UC and 49 non-IBD controls). A further 118 individuals were collected prospectively for dual FISH and PCR analysis (86 UC and 32 non-IBD controls). An additional 27 non-IBD controls were available for PCR analysis. All Helicobacter PCR-positive samples were sequenced. The association between Helicobacter and each study group was statistically analysed using the Pearson Chi Squared 2 tailed test. Helicobacter genus PCR positivity was significantly higher in UC than controls (32 of 77 versus 11 of 59, p = 0.004). Sequence analysis indicated enterohepatic Helicobacter species prevalence was significantly higher in the UC group compared to the control group (30 of 77 versus 2 of 59, p<0.0001). PCR and FISH results were concordant in 74 (67.9%) of subjects. The majority of discordant results were attributable to a higher positivity rate with FISH than PCR.
Citation: Thomson JM, Hansen R, Berry SH, Hope ME, Murray GI, Mukhopadhya I, et al. (2011) Enterohepatic Helicobacter in Ulcerative Colitis: Potential Pathogenic Entities? PLoS ONE 6(2): e17184. doi:10.1371/journal.pone.0017184
Editor: Tanya Parish, Queen Mary University of London, United Kingdom
Received: September 2, 2010; Accepted: January 24, 2011; Published: February 23, 2011
Copyright: © 2011 Thomson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was funded by the Broad Medical Research Program (Grant number IBD-0178). JMT was supported by grants from the GI Research Unit, Aberdeen Royal Infirmary; RH was supported by a Clinical Academic Training Fellowship, from the Chief Scientist Office (CAF/08/01). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ulcerative colitis (UC) is a chronic condition of the human colon which affects the superficial mucosal layer from the rectum and extending proximally for variable distances . This variable phenotype remains a puzzle, as does our difficulty in achieving long-term cure with current treatments. Recent developments in genetics have greatly improved our understanding of the inflammatory bowel diseases (Crohn's disease and UC), resulting in a renewed interest in the interplay between host immunology and bacteria at the mucosal surface; however genetic elements appear to be more important in Crohn's disease (CD) than UC. The possibility of infection as a trigger event for, or indeed as the cause of, inflammatory bowel disease (IBD) has long been debated with various organisms being suggested as pathogens. None of these organisms have been conclusively proven as causative agents. Studies examining the diversity of bacteria in IBD have shown increased cell counts of bacteria and reduced bacterial diversity. Changes in bacterial populations to the detriment of the host have been termed “dysbiosis” and this change is thought central to IBD pathogenesis. IBD onset following infectious episodes is well described and one possibility is that gastrointestinal infection may facilitate dysbiosis and ultimately IBD. Whether acute self-limiting infection is sufficient as a single entity, or whether chronic infection with as yet unknown agents is required to drive the chronicity of disease is unknown. UC is a stronger candidate than CD for a purely infectious aetiology because of the weaker genetic association, continuity of disease distribution and the relative limitation of disease to superficial tissue. It is likely however that a combination of host (genetic) susceptibility, a trigger event (which may be infectious) and the progression to dysbiosis are all likely required for the development of IBD.
The discovery that Helicobacter pylori was the causative agent underpinning gastric and duodenal ulceration and ultimately gastric cancer revolutionised our understanding of these conditions and resulted in a Nobel prize for Robin Warren and Barry Marshall. The tantalising possibility that a similar agent is responsible for IBD warrants consideration and exploration .
The family Helicobacteraceae contains the genera Helicobacter and Wolinella. The Helicobacter genus can be split into two groups, gastric Helicobacter, describing those that preferentially colonise the stomach, and enterohepatic Helicobacter, which preferentially colonise the intestinal or hepatobilliary system (Table 1). Enterohepatic Helicobacter organisms have been cultured from both Cotton-top tamarin monkeys (Saguinus oedipus) and rhesus monkeys (Macaca mulatta) with colitic disease (Helicobacter sp. Flexispira taxon 10, Helicobacter macacae and Helicobacter sp. Rhesus monkey 2), whilst Helicobacter hepaticus and Helicobacter bilis have been shown to be capable of causing IBD-like disease in immunodeficient rodent models –. Thus animal models demonstrating that infection with Helicobacter spp. on a background of host immunodeficiency can lead to colitis, and that “auto-immune” type reactions to commensal bacteria can be initiated by such organisms, would suggest the possibility of parallel mechanisms in humans resulting in IBD.
Various groups have examined human IBD for the presence of Helicobacter spp., from the negative studies of Bell and Grehan , , through to studies by Bohr, Zhang and Laharie which have successfully demonstrated PCR evidence of non-pylori Helicobacter (npH) in both IBD and controls –. The methodologies used, the variable rates of positivity reported between groups, and the small study numbers included in some, mean discussions at Helicobacter species level have been limited. Unfortunately, no-one has successfully cultured non-pylori Helicobacter organisms from IBD tissue (although 7 enterohepatic Helicobacter spp. have been cultured from the gastrointestinal tract of humans with diarrhoea or systemic disease ). The difficulties in isolating and culturing non-pylori Helicobacter from human colonic tissue highlight the importance of molecular approaches as viable alternatives to facilitate the study of the role of Helicobacter spp. in extra-gastric diseases. However, it is vital that these molecular methods are suitably sensitive, specific and applicable to a diverse range of samples. The purpose of the present study was to design a combined molecular approach to identify Helicobacteraceae organisms within a variety of gastrointestinal sample types. Our specific aim was to examine colonic tissue from IBD patients to assess the prevalence of Helicobacteraceae organisms against tissue from controls largely undergoing colorectal cancer screening. We elected to analyse UC cases rather than CD cases for the reasons outlined above regarding UC as a stronger candidate for an infectious aetiology.
Ethical approval for the archival specimen analysis and the biopsy study was granted by North of Scotland Research Ethics Service and written informed consent was obtained from all subjects in the biopsy study.
Development of a Combined Molecular Approach for the Detection of Helicobacteraceae Organisms
Development of PCR Methodology.
A bacterial reference panel was used to screen a series of primer combinations. The bacterial strains used in this study included: Helicobacter bilis (ATCC 51630), H. canadensis (ATCC 700968), H. canis (ATCC 51402), H. cholecystus (ATCC 700242), H. cinaedi (CCUG 18818), H. felis (ATCC 49179), H. hepaticus (ATCC 51449), H. pullorum (NCTC 12824), H. pylori (ATCC 700392), Pseudomonas fluorescens (clinical isolate), Listeria monocytogenes (clinical isolate), Aeromonas caviae (clinical isolate), Aeromonas sobria (clinical isolate), Campylobacter jejuni (clinical isolate), Proteus mirabilis (NCTC 3177), Enterobacter aerogenes (NCIMB 10102), Yersinia enterocolitica (NCIMB 2124), Bifidobacterium longum (NCIMB 8809), Bifidobacterium infantis (DSM 20088), Eubacterium rectale (NCIMB 14373), Roseburia intestinalis (DSM 14610), Bacteroides vulgatus (DSM 1447), Bacteroides thetaiotaomicron (NCTC 10582), Eubacterium hallii (DSM 17630), Enterococcus faecalis (NCIMB 13280), Pseudomonas aeruginosa (NCIMB 8626), Enterobacter cloacae (NCIMB 8556), Proteus vulgaris (NCTC 4175), Salmonella enteritidis (NCTC 12694), Salmonella poona (NCTC 4840), Salmonella typhimurium (NCIMB 13284), Escherichia coli (NCIMB 12210), Shigella sonnei (ATCC 25931), Staphylococcus epidermidis (NCIMB 8853), Bacillus subtilis (NCIB 8054), Bacillus cereus (ATCC 10876), Klebsiella pneumoniae (NCIMB 13281), Staphylococcus aureus (NCIMB 12702), Streptococcus gordonii (ATCC 35105), Faecalibacterium prausnitzii strain A2-165 (DSM 17677), Megasphaera elsdenii (ATCC 25940), Bifidobacterium adolescentis isolate L2-32, Lactococcus lactis strain MG1363, Enterococcus faecalis strain JH2-2, Ruminococcus albus strain SY3, Ruminococcus flavefaciens strain 17, Eubacterium cylindroides strain T2-87, Coprococcus spp L2-50, Methanobrevibacter smithii ATCC 35061, Acinetobacter baumanii, Lactobacillus acidophilus (ATCC 43561) and Streptococcus bovis strain Z6. Aerobic and microaerobic strains were grown at 37°C on Columbia agar with 10% horse blood. Anaerobic strains were grown at 37°C on M2GSC , MRS or M17 media (Becton Dickinson, Oxford, UK). All strains were used in fluorescent in-situ hybrisation (FISH) and PCR optimisation studies.
Initial assessment used a universal 16S bacterial PCR described previously . To allow identification of the family Helicobacteraceae (genera Helicobacter and Wolinella), 8 Helicobacteraceae PCR primer pairs were assessed. The nested PCR combination of C05 and C97  followed by a reverse complement of primer C98  and1067r  was selected as it yielded a final product of suitable length (∼400 bp) for sequence analysis. For H. pylori specific PCR, numerous H. pylori specific primer sets targeting the 16S rRNA gene were assessed with the most successful pairing being identified as 27f (5′-AGAGTTTGATCMTGGCTCAG-3′)  and HPY (5′-CTGGAGAGACTAAGCCCTCC-3′) . Both PCRs utilised the following conditions: denaturation at 94°C for 5 min, followed by 35 cycles of 94°C for 1 min, 66°C for 1 min, 72°C for 2 min. Final extension 72°C for 10 min. To determine the sensitivity of the genus PCR, decreasing amounts of H. pylori and H. hepaticus-derived DNA were spiked into faecal samples which were previously analysed by FISH and PCR and found to be negative for Helicobacteraceae. Dilutions ranged from 500 pg to 0.05 pg of Helicobacter DNA with the detection level of 0.5 pg Helicobacter DNA (representing approximately 30 bacteria) consistently being achieved.
Development of FISH Methodology.
Five broad-specificity probes were designed to target the small subunit rRNA of the family Helicobacteraceae (Table 2). The new probes were designed with the Primrose software package , checked against the Ribosomal Database Project (RDP) and EMBL databases, and were named according to the nomenclature suggested by the Oligonucleotide Probe Database (OPD) . One of the designed Helicobacteraceae probe sequences S-G-Hel-1047-a-A-21 had been previously described as a Helicobacter genus specific PCR primer  but to our knowledge it has not been used as a probe. Of note, in-silico analysis indicated that this probe detects several bacteria of the genera Sulfurimonas (17 of 21), Sulfurovum (9 of 42) and Wolinella (4 of 5). The specificity of the 5 newly designed Helicobacteraceae probes along with four previously published H. pylori specific probes (Table 2) was tested by whole-cell in situ hybridization against a panel of 60 reference strains derived from the human and animal gastrointestinal tract (see above) including a panel of 9 Helicobacter type strains . As a positive control for the presence of bacteria, the bacterium-specific probe S-D-Bact-0338-a-A-18 (termed Eub338) was used . Following assessment of the various Helicobacteraceae probes, it was identified that the Helicobacteraceae probe S-G-Hel-1047-a-A-21 and the H. pylori specific probe Hp16S2 hybridized only to the respective target organisms but not to any of the other organisms tested. Both the Eub338 and Hp16S2 hybridised at 50°C and could be co-hybridised using discriminating fluorescent labels (Rhodamine red and Oregon green 488 respectively). S-G-Hel-1047-a-A-21 hybridised at 52°C.
Validation of Molecular Methods (PCR/FISH) for the Detection of Helicobacteraceae Organisms.
In order to calculate the sensitivity and specificity of the dual molecular approach 100 gastric samples were selected on the basis of H. pylori status (50 positive and 50 negative). H. pylori status had been confirmed previously by CLO test and histology; however this was blinded to researchers until after molecular assessment. FISH was performed on archival paraffin tissue sections and PCR was performed on DNA extracted from fresh biopsies.
Biopsy blocks were cut to a thickness of 4 µm using a Leica RM2125RT rotary microtome with sections cut per block, and mounted on ChemMate capillary gap slides, 75 µm, (DakoCytomation, Cambridgeshire, UK). Following microtome sectioning and mounting of tissue, slides were dried vertically at room temperature and incubated overnight at 37°C to ensure that the tissue was adhered to the slide. Slides were then arranged by patient and block number and sections 1, 3 and 5 were used for assessing the presence of H. pylori coupled with the universal bacterial probe. Sections 2, 4 and 6 were used for assessing the presence of all Helicobacteraceae. Biopsy sections were deparaffinised using xylene and ethanol . For glass slides carrying deparaffinised tissue sections, 50 µl of hybridisation buffer was added and coverslips were used to minimise evaporation. Hybridisation was performed for 16 hours for all tissue sections and Vectashield Hardmount (Vector Laboratories, Peterborough, UK) was used.
DNA extraction of mucosal biopsies was performed using the commercially available Qiagen QIAamp Mini kit (Qiagen Crawley UK) with the following amendments. Biopsy samples were kept frozen until the addition of ATL buffer before allowing biopsies to equilibrate to room temperature, an additional 10 µl of Proteinase K was added for an initial lysis period of 18 hours to ensure complete lysis of the biopsy material prior to DNA extraction. PCR was performed as described above with biopsy DNA initially subjected to universal bacterial PCR  to confirm the suitability of the DNA for further analysis.
One hundred infectious diarrhoea samples were also collected for inclusion in the validation cohort. Samples were obtained from the Department of Medical Microbiology (Aberdeen Royal Infirmary, Aberdeen) and DNA was extracted using the Nucleon phytopure DNA extraction kit. PCR was performed as described above with faecal DNA initially subjected to universal bacterial PCR  to confirm the suitability of the DNA for further analysis.
Assessment of Helicobacteraceae Prevalence in Human Colonic Tissue Using the Combined Molecular Approach
Archival colonic tissue specimens.
Paraffin embedded colonic specimens from a total of 106 patients were obtained from the Department of Pathology (Aberdeen Royal Infirmary). Fifty-seven ulcerative colitis (UC) patients and forty-nine healthy controls (HC) were included. All UC patients were assessed during active disease and analysis was performed on all available colonic sites that were biopsied at the time of colonoscopy. The HC subjects comprised individuals who had undergone a colonoscopy in which colonic tissue was macroscopically normal, and subsequently confirmed as microscopically normal by histology. This cohort was examined exclusively by the FISH method outlined above. This sample cohort was not amenable to PCR methodologies.
Fresh colonic biopsy specimens.
A total of 145 individuals were recruited for the prospective fresh biopsy study, 86 formed the ulcerative colitis (UC) cohort and were analysed alongside a cohort of 59 healthy controls (HC). Of the UC cohort 9 individuals were excluded, 3 could not undergo colonoscopy for clinical reasons and 6 had an alternative final diagnosis. The HC cohort comprised two groups. The first group (N = 32) were recruited specifically for this study and had biopsies collected for both FISH and PCR studies (as outlined above). The second group (N = 27) had been recruited previously and all 27 had biopsies collected from normal colon whilst undergoing polypectomy.
Biopsies were collected during colonoscopy using standard endoscopic forceps (Boston Scientific Nanterre Cedex France). The colonic mucosa was rinsed with sterile water via the colonoscope to remove residual faecal material. Biopsies were immediately snap frozen in liquid nitrogen and then transferred to a −80°C freezer until used for DNA analysis. Additional biopsies were also sent for histopathology assessment and FISH analysis.
Biopsies were only collected for PCR based studies and so FISH analysis was not studied in this cohort. Therefore mucosal biopsies were obtained from 136 individuals, 77 with a clinical and histological diagnosis of UC (55 established disease – three had antibiotic therapy in the 6 months prior to study recruitment, 22 de-novo) and 59 healthy controls. The entire UC group had biopsies collected for both fluorescent in-situ hybridisation (FISH) and PCR studies. Samples that generated positive PCR results with Helicobacteraceae or H. pylori PCR primers were sequenced to confirm identity. Based on sequence analysis results, samples that were suspected of containing multiple Helicobacter sequences were cloned  and 5 clones per sample were sent for sequence analysis (400 bp).
Ethical approval for the archival specimen analysis and the biopsy study was granted by North East of Scotland Research Ethics Service and written informed consent was obtained from all subjects in the biopsy study.
Validation of Molecular Methods (PCR/FISH) for the Detection of Helicobacteraceae Organisms
Fifty of the gastric biopsy samples demonstrated FISH positivity using both Helicobacteraceae and H. pylori specific probes along with the universal Eub338 probe. The same 50 samples tested positive by PCR for Helicobacteraceae, H. pylori specific and universal bacterial primer sets with a subset of results (N = 10) subjected to sequence analysis which confirmed the presence of H. pylori (sequence identities >99%). The molecular results were 100% concordant with the findings of clinical investigation indicating that the combined approach had a high diagnostic sensitivity and specificity.
The combined molecular approach was then applied to 100 infectious diarrhoea samples. All infectious diarrhoea samples showed positivity with the universal bacterial FISH probe (Eub338) and the universal bacterial PCR. One diarrhoea sample was positive for Helicobacter by FISH and both Helicobacteraceae and H. pylori specific PCR with sequencing confirming the presence of H. pylori. As non-pylori Helicobacter organisms have been isolated from diarrheal samples previously, we performed a series of spiking experiments to confirm that the diarrheal samples were not inhibiting PCR amplification. Ten samples which were negative for Helicobacteraceae by both FISH and PCR were spiked with H. pylori DNA (500 pg to 0.5 pg) prior to PCR amplification. All spiked samples yielded positive PCR and subsequent sequencing results (>99% sequence similarity, over 360 bp, to the spiked H. pylori 16S rDNA gene sequence) confirming that if present these organisms would have been detected.
Helicobacteraceae Prevalence in Human Colonic Tissue
The UC archival cohort (n = 57, 44% Male) had a median age of 40 (range 15–82) at the time of colonoscopy and were classified as extensive (40%), left sided (37%) and proctitis (23%) according to the Montreal criteria . A total of 284 biopsies were analysed with 46% of subjects having biopsies from inflamed and un-involved mucosa available (E1 n = 8, E2 n = 9, E3 n = 9) and the remainder having biopsies only from inflamed mucosa. The HC archival cohort (n = 49, 35% Male) had a median age of 42 (range 14–80) at the time of colonoscopy. A total of 127 biopsies were processed in triplicate to assess the presence of Helicobacteraceae from the available pathology blocks of the right and left colon including rectum. Subjects were considered to be positive if appropriate fluorescent organisms were observed in at least 1 slide. All fluorescent in-situ hybridisations were assessed alongside H. pylori positive gastric biopsy reference slides and a selection of Helicobacter reference strains. All of the 106 subjects had bacteria detected with the universal bacterial probe. Of the 55 subjects positive for Helicobacteraceae, 29 were from the UC cohort and 26 were from the HC cohort. In both groups there was no statistically significant correlation with gender or age of the subjects and in the UC cohort the extent of disease did not correlate with the rate of positivity either as detailed in Table 3. Interestingly there was a statistically significant difference in the presence of Helicobacter pylori with 13 of the Helicobacteraceae positive subjects from the HC cohort also having a positive result for Helicobacter pylori compared with only 3 of the UC cohort (p = 0.002). Assuming that these Helicobacter pylori positive results represent transported gastric Helicobacter pylori we consequently removed these from the Helicobacteraceae positive results, thus creating a new category of non-pylori Helicobacter positive organisms. By this approach there was also a statistically significant difference in the presence of non-pylori Helicobacter organisms between the UC and HC cohorts (p = 0.04). There was no correlation with age, gender or severity of disease as detailed in Table 3. Because of limitations of the archival FISH technique it was not possible to ascertain if the H. pylori organisms were entirely responsible for the genus positivity or if the H. pylori were cohabiting with other non-pylori Helicobacter. Indeed, it was also not possible to ascertain if more than one Helicobacteraceae species was present within a sample. Attempts were made to extract microbial DNA from the archival tissue but these were unsuccessful and therefore a prospectively collected cohort was established with samples taken for PCR based analyses and FISH analysis.
The prospective UC cohort (n = 77, 46% Male) had a median age of 42 (range 16–84) at the time of index colonoscopy and were classified as extensive (25%), left sided (62%), proctitis (13.0%). A total of 137 biopsy sites were analysed. Twenty one (27%) subjects had a single site analysed, all of which represented inflamed mucosa and 56 (73%) had more than one biopsy site assessed, of which 42 both had inflamed and uninvolved mucosa. The prospective HC cohort (n = 59, 59% Male) had a median age of 63 (range 30–75) at the time of index colonoscopy. There was a statistically significant median age difference between the UC cohort and the control groups (Mann Whitney U test p<0.001). A single biopsy site (n = 59) was analysed from each control subject. All 136 subjects (77 UC and 59 controls) had colonic biopsies available for PCR analysis and all were positive for universal bacterial PCR indicating the presence of bacterial DNA within all samples. Of the 136 subjects, 43 (32%) were PCR positive for Helicobacteraceae and 3 (2%) were PCR positive for both Helicobacter pylori and Helicobacter genus (Table 4). Subsequent sequence analysis of the Helicobacteraceae PCR products confirmed the presence of only Helicobacter pylori in these 3 samples. Thus 40 subjects were PCR positive for Helicobacteraceae but not H. pylori. Sequence analysis revealed a further 4 subjects (3 controls and 1 UC) with only Helicobacter pylori identified.
In the remaining 36 subjects sequence analysis identified the presence of a further 8 Helicobacter species and a Wollinella succinogenes with 6 subjects having more than one species identified (Table 5). Of these subjects, 3 had multiple Helicobacter species present within the same biopsy sample and 3 had different Helicobacter species spread between the samples analysed (Table S1). Helicobacter pylori were not found to co-exist with any other Helicobacteraceae species. The species identified and the number of samples they were identified in is detailed in Table S1.
Helicobacteraceae PCR positivity was significantly higher in UC than controls 32 of 77 (42%) versus 11 of 59 (19%), p = 0.004). By analysing the sequences obtained and including only those Helicobacter species classified as “enterohepatic,” the prevalence was 29 of 77 (38%) in the UC group versus 2 of 59 (3%) in the controls (p<0.0001). There was also a negative association between the identification of gastric Helicobacter species in UC (2 of 77) versus controls (9 of 59, p = 0.007) (Table 4). There was no correlation between the age, gender or extent of disease.
The effect of bowel preparation on Helicobacteraceae PCR positivity was also considered. Of the 77 UC subjects, 32 had full bowel preparation prior to colonoscopy along with all 59 control subjects. Helicobacteraceae PCR positivity in subjects with bowel preparation was significantly higher in UC than controls [20 of 32 (63%) versus 11 of 59 (19%), (Pearson Chi squared test p<0.0001)]. There was also a significant difference in Helicobacteraceae PCR positivity within the UC cohort based on bowel preparation 20 of 32 (63% full bowel preparation), 1 of 5 (20% phosphate enema preparation) vs 11 of 40 (28% no bowel preparation), Pearson Chi squared test p = 0.007). There was also no statistically significant association with antibiotic usage.
For 109 (77 UC, 32 HC) of the 136 subjects, samples were available for both FISH and PCR based analyses. All samples analysed by FISH were Eub338 positive indicating the presence of bacteria. 62 of 77 (81%) UC were Helicobacteraceae positive whilst 1 was also H. pylori positive. In the 32 controls, 12 (38%) were Helicobacteraceae positive whilst 2 were also H. pylori positive (Table 6). The Helicobacteraceae positivity was significantly higher in the UC cohort (p<0.0001) but no negative association with Helicobacter pylori as seen in the archival FISH and prospective PCR studies. As in the other studies there was no correlation between the age, gender or extent of disease. Correlation between PCR and FISH results for these 109 subjects were examined which demonstrated concordance in 74 (68%) of subjects. The majority of discordant results were attributable to a higher positivity rate for Helicobacteraceae with FISH than PCR (Table 7).
During the development of our combined molecular approach, both the PCR and FISH techniques were highly sensitive and specific (100% each) when interrogating gastric biopsies with known H. pylori status. A further validation cohort utilising diarrhoeal samples demonstrated limited Helicobacteraceae positivity. However spiking experiments on this cohort indicate that organisms would have been identified had they been present. These findings suggest that Helicobacteraceae are not a prominent causative agent in infectious diarrhoea in our setting although they have been isolated from diarrhoeal samples by other investigators –. Part of the rationale behind our developing a combined approach to identify Helicobacteraceae, rather than one based solely on either PCR or FISH, was that the combination of techniques allows visualisation of organisms in-situ and species-level identification from sequencing. PCR-only studies can be criticised based on the possibility that contaminant environmental DNA could bias results. In the gastrointestinal tract for instance, DNA could be transited to the colon in the faecal stream from foodstuffs. FISH addresses these concerns by allowing direct visualisation and localisation of organisms to the colonic mucosa. FISH-only studies however are limited by the constraints of designing species-specific probes and therefore they lack species-level sensitivity at times. By utilising both approaches Helicobacteraceae species present could be visualised and also identified. Based on the strength of these validation studies, we considered that this combined methodology was suitable for investigating Helicobacteraceae prevalence in UC colonic biopsies. A combined FISH/PCR approach has also been utilised to examine Helicobacteraceae prevalence in a small cohort of children with IBD (n = 12) (Crohn's disease n = 11), irritable bowel syndrome (IBS; n = 5) and controls (n = 4) . This small study identified a strikingly high prevalence in both IBD (11/12) and IBS (5/5) versus controls (1/4). Through sequence analysis of DGGE bands several Helicobacteraceae were identified including Helicobacter ganmani, Wolinella succinogenes, H. hepaticus and H. pylori. Two further bands were identified as Helicobacter although equal sequence similarity was attributed to multiple species.
The results of our study show that rates of Helicobacteraceae positivity are significantly higher in the colonic tissue of UC patients than in controls. When sequencing data is analysed and species identities attributed, in control patients, the species identified are almost exclusively gastric (namely H. pylori and H. mustelae (99–100% sequence similarity); comprising 9 of 11). This finding is in stark contrast to the Helicobacter sequences from the UC cohort where 32 patients had Helicobacteraceae species identified, although only 2 of these were attributed to H. pylori and curiously, H. mustelae was absent. Our study was not designed to obtain gastric Helicobacter species. As such it was not possible to confirm or refute the notion that Helicobacter species detected in the colon are truly colonising the mucus or merely transiting from the stomach. For the former, Helicobacter grown from both sites (stomach and colon) would be necessary in order to undertake a detailed strain comparison, which was not feasible within this study. As indicated in the results section there was also a statistically significant difference in age between the UC and control groups. The difference in age between the two groups is a result of the control group predominantly being recruited from a colorectal cancer screening programme whose lower age limit is 50 years.
Interestingly, bowel preparation appeared to increase the detection of Helicobacter species. It might be anticipated that the wash-out effect of bowel preparation might reduce the positivity. This surprising result may be the result of an unidentified confounding factor or could be a surrogate disease severity marker. This was due to it being a clinical decision whether subjects received bowel preparation or not. As no bowel preparation was given when there was the clinical impression of severe disease. This suggests that Helicobacteraceae positivity is associated with less severe disease. However no association between Helicobacteraceae positivity and the Montreal classification of disease extent and severity was observed.
Six of the UC patients appear to have multiple Helicobacter species present within their colonic tissue. Two of the six had different Helicobacter sequences identified from the same biopsy whereas the other four had single Helicobacter species identified from biopsies taken from different regions of the colon. Mixed species were not identified in the control cohort although it should be acknowledged that only single colonic sites were investigated by biopsy. Nevertheless our findings suggest that more than one enterohepatic Helicobacter species can be present in the same human host although this does not appear to be the case for gastric species.
It should be noted that allocation of species should not rely solely on 16S rRNA sequencing as comparison of these sequences can be misleading and does not always provide conclusive evidence for species level identification. Helicobacter species identity cannot be firmly established by 16S sequencing and the 400 bp product of the nested PCR further compounded this by only allowing sequencing over this short segment. However there is confidence that the sequence belonged to the genus Helicobacter based on the sequenced product. For example, although our nested PCR technique amplified a hyper-variable region of the 16S rRNA Helicobacter genome which equated to an estimated average evolutionary diversity of 14 base pairs within the 9 species identified , it is not possible to be certain of the Helicobacter species without additional genotypic or phenotypic characterisation . There is always the possibility that mixed Helicobacter organisms were present that had identical sequences over the 400 bp 16S rDNA region analysed. It is possible that alternative identification approaches including denaturing gradient gel electrophoresis (DGGE) could have been used to address this potential issue. Where sequence analysis indicated that mixed species were present, the additional cloning and sequencing approach was undertaken which demonstrated the presence of multiple Helicobacter sequences within a few samples.
The use of nested PCR is known to increase the sensitivity of the PCR test; however it is also known that without strict use of appropriate control strategies including sequencing of positive results, it can lead to false positives. In the current study, all positive nested PCR findings were sequenced in order to eliminate the query of false positives due to nested PCR. Extensive attempts were made to isolate Helicobacter species from the prospective cohort, however this was not successful. Helicobacter species are notoriously fastidious and although novel Helicobacter species have been isolated from human faeces, to date none have been isolated from colonic tissue. Examining both FISH and PCR analyses on a large number of samples (n = 109) revealed a correlation rate of ∼68%. In the majority of cases, discordant results showed that FISH was more likely to yield a positive result than PCR. This was not seen in our gastric validation cohort. The most likely explanation is that the number of H. pylori strains present in infected gastric tissue is higher than the corresponding number of Helicobacter species in the colon. These results suggest that the FISH technique is more sensitive but lacks the specificity to identify the Helicobacteraceae species. Regardless of these technical issues, both techniques demonstrate a statistically significant correlation between the presence of Helicobacteraceae species, particularly the enterohepatic Helicobacter species and the UC cohort.
The PCR methodology that we developed used a nested PCR approach for Helicobacteraceae but a single PCR was used for H. pylori. It is likely that a nested PCR for H. pylori would have been more sensitive in the colonic samples and would have increased the detection rate. This is a plausible explanation for at least 3 of the four samples which were shown by sequence analysis of the Helicobacteraceae PCR product to contain H. pylori despite H. pylori PCR results being negative. Nevertheless, in order to identify every Helicobacter that was detected, we chose to sequence every positive PCR reaction including those that were H. pylori positive in order to determine whether multiple species were present.
Attributing causation to putative pathogens has always been a difficult endeavour with the gold standard remaining fulfilment of Koch's postulates . In an era of molecular biology and an increasing awareness of the “unculturable” microbiota of the human colon however, these postulates are perhaps outdated. Swidsinski and colleagues recently proposed alternative postulates for a modern era :
- There must be a clear link between a pathogen and a disease,
- The pathogenic organism should be identified and characterised (by traditional culture and phenotyping or by “reliable” modern methods such as PCR, DNA sequencing and FISH),
- There should be positive evidence of the chain of infection (this can be from individual transfections or from epidemiological observation)
- Knowledge of a specific pathogen should assist the development of new diagnostic methods and treatment
We would add that host factors, in particular genetic or immunological susceptibility should be considered, particularly in the context of IBD. We believe that our data adds considerable weight to fulfilling Swidsinski's second postulate and that the first has already been firmly established in animal models. Further work is required to address the third postulate which would clearly be aided by successful culture of these organisms from the colonic tissue of UC patients.
Finally, the presence of H. mustelae in the colonic tissue of controls but not UC patients warrants further consideration. Since H. mustelae is a gastric organism (previously only identified in ferrets ), it would be interesting to see if this species is co-colonising the human stomach and colon, the colon alone or simply being transited to the colon from the stomach. H. mustelae has a similar morphology to H. pylori and is also a urease positive organism so it could easily be mistaken with current clinical testing (CLO test) for H. pylori in human gastric disease. It may be that H. mustelae represents a less pathogenic organism in the human host which nonetheless confers the IBD protective benefits of H. pylori . In terms of the extra-gastric Helicobacters our findings clearly demonstrate compelling molecular evidence for their presence in the human colon. Their presence in the human host is not well established compared to animal models however; there is no reason to suggest that they do not reside in the human intestinal tract. Clearly these hypotheses require further exploration.
Bacterial Sequence Identification* of Helicobacteraceae positive samples.
We are grateful to Dr Sylvia Duncan (Rowett Institute for Nutrition and Health, Aberdeen) for supplying many of the bacterial isolates used in this work. We gratefully acknowledge technical assistance from Mrs Nicky Fyfe.
Conceived and designed the experiments: JMT RH GIM JGF EEO GLH. Performed the experiments: JMT RH SHB MEH IM MHM ZS. Analyzed the data: JMT RH GLH. Contributed reagents/materials/analysis tools: JMT RH ZS JGF EEO GLH. Wrote the paper: JMT RH GLH.
- 1. Lennard-Jones JE (1989) Classification of inflammatory bowel disease. Scand J Gastroenterol 24(S170): 2–6.
- 2. Hansen R, Thomson JM, Fox JG, El-Omar EM, Hold GL (2011) Could Helicobacter organisms cause inflammatory bowel disease? FEMS Imm Med Micro 61(1): 1–14.
- 3. Cahill RJ, Foltz CJ, Fox JG, Dangler CA, Powrie F, et al. (1997) Inflammatory bowel disease: An immunity-mediated condition triggered by bacterial infection with Helicobacter hepaticus. Infect Immun 65(8): 3126–3131.
- 4. Kullberg MC, Ward JM, Gorelick PL, Caspar P, Hieny S, et al. (1998) Helicobacter hepaticus triggers colitis in specific-pathogen-free interleukin-10 (IL-10)-deficient mice through an IL-12-and gamma interferon-dependent mechanism. Infect Immun 66(11): 5157–5166.
- 5. Jergens AE, Wilson-Welder JH, Dorn A, Henderson A, Liu Z, et al. (2007) Helicobacter bilis triggers persistent immune reactivity to antigens derived from the commensal bacteria in gnotobiotic C3H/HeN mice. Gut 56(7): 934–940.
- 6. Bell SJ, Chisholm SA, Owen RJ, Borriello SP, Kamm MA (2003) Evaluation of helicobacter species in inflammatory bowel disease. Aliment Pharmacol Ther 18(5): 481–486.
- 7. Grehan M, Danon S, Lee A, Daskalopoulos G, Mitchell H (2004) Absence of mucosa-associated colonic helicobacters in an australian urban population. J Clin Microbiol 42(2): 874–876.
- 8. Bohr URM, Glasbrenner B, Primus A, Zagoura A, Wex T, et al. (2004) Identification of enterohepatic helicobacter species in patients suffering from inflammatory bowel disease. J Clin Microbiol 42(6): 2766–2768.
- 9. Zhang L, Day A, McKenzie G, Mitchell H (2006) Nongastric helicobacter species detected in the intestinal tract of children. J Clin Microbiol 44(6): 2276–2279.
- 10. Laharie D, Asencio C, Asselineau J, Bulois P, Bourreille A, et al. (2009) Association between entero-hepatic helicobacter species and crohn's disease: A prospective cross-sectional study. Aliment Pharmacol Ther 30(3): 283–293.
- 11. Miyazaki K, Martin JC, Marinsek-Logar R, Flint HJ (1997) Degradation and utilization of xylans by the rumen anaerobe prevotella bryantii (formerly P. ruminicola subsp. brevis) B14. Anaerobe 3(6): 373–381.
- 12. Hold GL, Pryde SE, Russell VJ, Furrie E, Flint HJ (2002) Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis. FEMS Microbiol Ecol 39(1): 33–39.
- 13. Fox JG, Dewhirst FE, Shen Z, Feng Y, Taylor NS, et al. (1998) Hepatic helicobacter species identified in bile and gallbladder tissue from chileans with chronic cholecystitis. Gastroenterology 114(4): 755–763.
- 14. Grehan M, Tamotia G, Robertson B, Mitchell H (2002) Detection of Helicobacter colonization of the murine lower bowel by genus-specific PCR-denaturing gradient gel electrophoresis. Appl Environ Microbiol 68(10): 5164–5166.
- 15. Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173(2): 697–703.
- 16. Moreno Y, Ferrus MA, Alonso JL, Jimenez A, Hernandez J (2003) Use of fluorescent in situ hybridization to evidence the presence of helicobacter pylori in water. Water Res 37(9): 2251–2256.
- 17. Ashelford KE, Weightman AJ, Fry JC (2002) PRIMROSE: A computer program for generating and estimating the phylogenetic range of 16S rRNA oligonucleotide probes and primers in conjunction with the RDP-II database. Nucleic Acids Res 30(15): 3481–3489.
- 18. Loy A, Maixner F, Wagner M, Horn M (2007) probeBase–an online resource for rRNA-targeted oligonucleotide probes: New features 2007. Nucleic Acids Res. 35. (Database issue): pp. D800–D804.
- 19. Schwiertz A, Le Blay G, Blaut M (2000) Quantification of different eubacterium spp. in human fecal samples with species-specific 16S rRNA-targeted oligonucleotide probes. Appl Environ Microbiol 66(1): 375–382.
- 20. Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, et al. (1990) Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microbiol 56(6): 1919–1925.
- 21. Trebesius K, Panthel K, Strobel S, Vogt K, Faller G, et al. (2000) Rapid and specific detection of Helicobacter pylori macrolide resistance in gastric tissue by fluorescent in situ hybridisation. Br Med J 46(5): 608–614.
- 22. Silverberg MS, Satsangi J, Ahmad T, Arnott IDR, Bernstein CN, et al. (2005) Towards an integrated clinical, molecular and serological classification of inflammatory bowel disease: Report of a working party of the 2005 montreal world congress of gastroenterology. Can J Gastroenterol 19: Suppl A5A–36A.
- 23. Tee W, Anderson BN, Ross BC, Dwyer B (1987) Atypical campylobacters associated with gastroenteritis. J Clin Microbiol 25(7): 1248–1252.
- 24. Vandamme P, Falsen E, Pot B, Kersters K, De Ley J (1990) Identification of Campylobacter cinaedi isolated from blood and feces of children and adult females. J Clin Microbiol 28(5): 1016–1020.
- 25. Burnens AP, Stanley J, Schaad UB, Nicolet J (1993) Novel campylobacter-like organism resembling helicobacter fennelliae isolated from a boy with gastroenteritis and from dogs. J Clin Microbiol 31(7): 1916–1917.
- 26. Stanley J, Linton D, Burnens AP, Dewhirst FE, Owen RJ, et al. (1993) Helicobacter canis sp. nov., a new species from dogs: An integrated study of phenotype and genotype. Microbiology 139(10): 2495–2504.
- 27. Romero S, Archer JR, Hamacher ME, Bologna SM, Schell RF (1988) Case report of an unclassified microaerophilic bacterium associated with gastroenteritis. J Clin Microbiol 26(1): 142–143.
- 28. Stanley J, Linton D, Burnens AP, Dewhirst FE, On SLW, et al. (1994) Helicobacter pullorum sp. nov.-genotype and phenotype of a new species isolated from poultry and from human patients with gastroenteritis. Microbiology 140(12): 3441–3449.
- 29. Burnens AP, Stanley J, Morgenstern R, Nicolet J (1994) Gastroenteritis associated with Helicobacter pullorum. Lancet 344(8936): 1569–1570.
- 30. Steinbrueckner B, Haerter G, Pelz K, Weiner S, Rump JA, et al. (1997) Isolation of Helicobacter pullorum from patients with enteritis. Scand J Infect Dis 29(3): 315–318.
- 31. Melito PL, Munro C, Chipman PR, Woodward DL, Booth TF, et al. (2001) Helicobacter winghamensis sp. nov., a novel helicobacter sp. isolated from patients with gastroenteritis. J Clin Microbiol 39(7): 2412–2417.
- 32. Lastovica AJ (2000) Efficient isolation of campylobacteria from stools. J Clin Microbiol 38(7): 2798–2799.
- 33. Fox JG, Chien CC, Dewhirst FE, Paster BJ, Shen Z, et al. (2000) Helicobacter canadensis sp. nov. isolated from humans with diarrhea as an example of an emerging pathogen. J Clin Microbiol 38(7): 2546–2549.
- 34. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24(8): 1596–1599.
- 35. Whary MT, Fox JG (2004) Natural and experimental helicobacter infections. Comp Med 36: 139–177.
- 36. Koch R (1884) Die aetiologie der tuberkulose. Mittbeilungen Kaiserlichen Gesundbeitsamte 2: 1–88.
- 37. Swidsinski A, Doerffel Y, Loening-Baucke V, Theissig F, Ruckert JC, et al. (2010) Acute appendicitis is characterized by local invasion with Fusobacterium nucleatum/necrophorum. Gut:doi:10.1136/gut.2009.191320.
- 38. Fox JG, Correa P, Taylor NS, Lee A, Otto G, et al. (1990) Helicobacter mustelae-associated gastritis in ferrets. An animal model of Helicobacter pylori gastritis in humans. Gastroenterology 99(2): 352–361.
- 39. Luther J, Dave M, Higgins PDR, Kao JY (2009) Association between Helicobacter pylori infection and inflammatory bowel disease: A meta-analysis and systematic review of the literature. Inflamm Bowel Dis 16(6): 1077–1084.
- 40. Daims H, Brühl A, Amann R, Schleifer KH, Wagner M (1999) The domain-specific probe EUB 338 is insufficient for the detection of all bacteria: Development and evaluation of a more comprehensive probe set. Syst Appl Microbiol 22(3): 434–444.
- 41. Morotomi M, Hoshina S, Green P, Neu HC, LoGerfo P, et al. (1989) Oligonucleotide probe for detection and identification of Campylobacter pylori. J Clin Microbiol 27(12): 2652–2655.