In this study we describe the establishment of the enteric microbial community in the ileal pouch of patients with a history of UC who have undergone total colectomy with IPAA. Our study is unique in examining the relationship of the microbiota and pouchitis and offers several advantages: (1) a clearly defined starting point when all patients are free of disease and off medications, (2) the opportunity to observe the development of a nascent pouch microbiota relative to the maturation of the ileal pouch, (3) sampling of mucosa-associated pouch microbiota without colonic lavage  and (4) the identification of pouch microbial communities that may be associated with health or increased risk of disease.
We were able to determine that in all four patients the pouch microbial community evolved over time, but it did so in a manner that was unique to each individual. As expected, the shift in the microbial community in each individual was most evident immediately after closure of the diverting ileostomy and reinstitution of the fecal stream through the pouch. While the microbial communities continued to evolve over time, the changes were less pronounced. The microbial community appeared to move toward a more “healthy” colonic community at each subsequent visit in patients 206, 207 and 210. However, in patient 200, while the microbial community initially moved toward a “healthy” colonic community, it drifted away from the healthy colonic community structure 1 and 2 months after ileostomy closure. Interestingly, this patient whose microbial community was less similar over time to the communities of the healthy colons developed chronic pouchitis 7 months after visit 4 and was diagnosed with C. difficile of the ileal pouch 10 months following visit 4. Patients 206 and 207 both had episodes of acute pouchitis at 16 and 17 months, respectively, following closure of the ileostomy. Both patients responded well to antibiotic treatment. Patient 210’s microbial community structure moved closest to the community structure observed in healthy colon samples, and this patient has had the best clinical outcome of the four subjects. Two years into the study, he continues to do well with no episodes of pouchitis.
Although our sample size was small, our findings suggest that the failure to develop a mature microbial community similar to that found in the healthy colon in the months following ostomy closure increases susceptibility to pouchitis. Alternatively, if an individual’s pouch microbiota develop into a diverse, mature community similar to that found in the healthy colon, we propose the risk of pouchitis is decreased. Falk and colleagues also performed a longitudinal study in two patients with a history of UC undergoing colectomy with IPAA . They followed these two patients for 1 year after stoma closure. Although the microbial profiles were quite different between the two patients, both individuals’ mucosal-associated microbiota evolved over time and became more “colon-like.” Neither patient developed pouchitis.
Others have compared the pouch microbiota in patients with and without pouchitis, although the results from these various studies have varied, and no clear association between particular bacterial profiles has been identified [23–26]. Komanduri et al. pooled mucosal samples from patients with UC healthy pouch, UC pouchitis and non-IBD controls. They described shifts in the microflora (dysbiosis) in patients with pouchitis, with an increase in the proportion of Fusobacter (phylum Proteobacteria) and a decrease in Streptococci (phylum Firmicutes). In another study examining the mucosal-associated microbiota from ileal pouches, diversity was greater in the UC patients without pouchitis compared to the UC patients with pouchitis. There were no specific phylotypes associated with pouchitis .
Dysbiosis of the intestinal microbiota is commonly found in patients with IBD , but in almost all cases, the question remains as to whether these changes are causes or consequences of the activated immune and inflammatory condition. Typically, large changes in 16S rRNA gene-based profiles are observed at the phylum level, characterized by blooms of Proteobacteria and changes in the relative abundance of Firmicutes and Bacteroidetes . Significant differences in the functional metagenomes of enteric microbes from healthy and IBD patients have also been shown . However, the observed structural and functional changes are not unique to human IBD and can be found in other non-IBD inflammatory states [30, 31] and in experimental colitis [32–34], suggesting that these changes are largely a consequence of the altered immune and inflammatory state. These large, descriptive data sets have shed little light on fundamental mechanisms of IBD etiopathogenesis and are often confounded by difficult to control variables inherent to clinical studies, including differences in individual microbiomes, medication (including antibiotics), diet, genetics, and environmental and lifestyle factors.
Several studies have suggested that the overall diversity of the gut microbiota is decreased in the setting of IBD. However, our results indicate that diversity in and of itself (as measured by a metric such as the Shannon diversity) is not necessarily a predictor of disease. While it is true that patients 210 and 200 both developed pouch mucosal communities that were similar in overall diversity to that seen in the normal colon, patient 200 developed severe, refractory pouchitis, while patient 210 remains healthy after 2 years of follow-up. Rather than using overall diversity as a measure of “health” of the microbiome, it is likely that the specific composition and thus function of the community are the key predictors. In this regard, the community in patient 210 is most similar in composition and structure to that seen in the healthy colon, while the community in patient 200 is quite distinct, despite having relatively high overall diversity. This is also reflected in the relative potential of the communities to produce the beneficial SCFA butyrate where the community in patient 210 had a much higher potential of butyrogenesis compared to that in patient 200.
The importance of specific microbial functions is further indicated in our analysis of the butyrogenic potential of an individual patient’s microbiome. Abundant butyrate-producing communities were established in all patients after ileostomy takedown, but only patient 210 exhibited both taxa Roseburia sp. and Faecalibacterium sp. (Figure 6), which are considered to be the main butyrate producers in healthy colons [19, 20]. In our companion work (Microbiome, submitted in parallel to this manuscript), we specifically investigated the diversity (via a pyrosequencing strategy) and abundance (via quantitative PCR) of butyrate-producing genes in corresponding luminal samples taken at the same time as mucosa samples. All patients established an abundant butyrate-producing community (approximately 5–26% of the total community) after ileostomy takedown, but with distinct profiles between patients, where patient 210 was the only individual exhibiting a but/k profile similar to those of the control samples. Whereas the overall patterns between 16S rRNA gene analysis (of luminal samples) and the functional gene-targeted approach were consistent, only the latter could reveal butyrate-producing gene families and their inferred taxa in detail to more directly evaluate the butyrogenic potential. Unfortunately, it was not possible to use the functional gene-targeted approach in this mucosal study because of the low amount of bacterial DNA in most biopsy samples. However, 16S rRNA gene patterns specific for butyrate producers from Vital et al.  were identical to results presented here (Figure 6), suggesting consistency in functional gene profiles between mucosa and luminal derived samples as well. Indeed, we were able to retrieve qPCR data from mucosal samples of patient 210 (only visit 3 and 4) and found but genes linked to Roseburia sp./Eubacterium sp. and F. prausnitzii (data not shown) at abundances similar to luminal aspirates and consistent with the 16S rRNA gene data presented in Figure 6. The observed progression of patient 210 toward a “healthy type” bacterial community based on global 16S rRNA analysis was consistent with the development of a specific butyrate-producing community similar to that of healthy colons. The question of how individual bacterial profiles relate to the actual production of butyrate and whether abnormal communities perform equally compared to regular ones is under study.
This study describes the establishment of the pouch microbiota in a longitudinal manner. This is the first step toward understanding how the intestinal microbiota influence or trigger the development of the inflammatory response seen in IBD. Despite our small sample size, we can confirm that there is an evolution of the microbial community in all individuals. These changes were apparent when the communities were followed by multiple methodologies. This includes the use of culture-independent surveys based on retrieval of 16S rRNA-encoding gene sequences, specific culture-based techniques and targeted analysis of functional genes.
Structurally and functionally, we find that the mucosal microbial community of the pouch moves toward a community more similar to a colonic microbial community following ileostomy takedown. Nonetheless, the membership within the community is distinct from the colonic microbial community, likely secondary to the influence of the small intestinal epithelium on the development of the pouch microbiota and the overall anatomic differences inherent to IPAA. The establishment of a pouch microbiota similar to a healthy colonic microbial community is potentially protective against pouchitis. In the future, longitudinal studies, such as this one, may provide the opportunity to identify patterns of the microbial community, both in terms of structure and function, that predict the onset of inflammation or disease in at-risk individuals.