Search

Previous studies have demonstrated the importance of the microbiome in the pathogenesis of inflammatory bowel disease (IBD), with a dysbiosis being a common feature. This change in the microcbiome has metabolic consequences, and metabonomic profiling can be used to detect these metabolic signals along with changes in the patient's own metabolic profile. Metabonomics have been applied clinically in oncology to augment diagnosis, treatment, and prognostication, and the aim is for it to have a similar role in IBD. Most studies in IBD have been performed on stool, and have observed a number of discriminatory metabolites, but more recently urinary metabonomics has been studied, as this is a much more convenient biofluid to use both in terms of sample collection and preparation. Here too, discriminatory metaboliteshave been observed, several of which are produced by intestinal microbial or host-microbial metabolism. These changes in the IBD metabolome give an insight into the complex pathophysiology of this disease, and have significant potential impact in both diagnosis and stratification of patients into different disease phenotypes and in the assessment of treatment outcome. However, up until now IBD metabonomic research has been restricted to homogeneous clinical cohorts where subjects with comorbidities have been excluded, and for the technique to be clinically useful it has to provide consistent results in real-life cohorts, including factors that may influence the metabolome such as obesity and the effects of bowel cleansing pre-colonoscopy, both of which apply to a real-life IBD clinical population. This project assessed these factors on the IBD metabolome, examining principally urine due to its ease of use. Multivariate analysis showed that changes seen in the IBD metabolome are present with the inclusion of subjects with comorbidities in IBD and Crohn's disease cohorts relative to controls. Targeted analysis showed differences in hippurate and 4-cresol sulfate between IBD cohorts relative to controls, likely associated with reduced abundance of Clostridia species seen on microbiomic analysis. No changes could be seen in ulcerative colitis (UC) cohorts, but UC participant numbers were relatively low following removal of patients taking 5- aminosalicylates that influenced the NMR spectrum. Obesity may have an effect on the urinary IBD metabolome, as although there was no clustering on unsupervised multivariate analysis, a reduction in hippurate excretion was not seen in the obese IBD relative to control cohorts, and hippurate has been the most consistently reported discriminatory metabolite associated with IBD. Longitudinal analysis of the urinary and faecal metabolome showed resistance at day 3 and week 6 following bowel cleansing; a temporary reduction of alpha diversity was observed on microbiomic analysis. Overall, this project has shown that whilst many elements of the IBD related urinary metabolome are preserved when patients with comorbidities are included into analysis, obesity may be a significant confounder. The studies of the urinary and faecal metabolome after bowel cleansing show that the metabolome appears resistant to significant changes at day 3 following bowel cleansing, which may be helpful in analysing future metabolic and microbiomic studies after colonoscopy, and help to associate changes identified from sampling during those procedures.

The pathogenesis of IBD, involving dynamic interactions between the microbiome, innate and adaptive immune systems, genetics and environmental factors, is a major focus of academic interest, in order to reveal more about the heterogeneous clinical course of the disease and in pursuit of improved therapeutic targets. Metabonomics has been previously used with a variety of biofluids to successfully distinguish IBD from controls, but the complex metabolic data also have potential to unlock insights into pathogenesis and better understand how to better stratify patients for personalised clinical care. In the largest urinary metabonomics IBD study to date, changes in the white European cohort confirmed previous published findings, highlighting discriminatory metabolites of gut microbial and inflammatory pathway sources. Significant metabolic differences were seen when comparing IBD patients and controls from South Asia to white North Europeans, demonstrating the influence of ethnicity on the metabolic profile and showing metabolite changes related to host-nutrition-microbiome interactions. Results from longitudinal measurements of the IBD metabolome in the same individuals over several years indicate relative stability despite the relapsing-remitting course of the disease and different treatments. This early finding suggests clinical outcomes may only have subtly discernible changes on metabolic profiles, potentially limiting its application as a disease-monitoring tool. 16S rRNA profiling, employed to characterise the microbiome, showed reduced microbial diversity in IBD and 4 key bacterial genera - Veillonella, Acidaminococcus, Lactobacillus and Streptococcus - associated with disease. Significant urinary and faecal metabolites in the same patients were correlated with these bacteria to demonstrate the feasibility of multi-omic integration in IBD. Furthermore, the breath VOC profiles of IBD patients obtained by SIFT-MS were distinct from those of heathy controls, with the significant compounds originating from microbial sources, and inflammatory pathways, demonstrating the potential of this technology and another facet to metabolic profiling in IBD.

Clinical scores, serum and faecal markers, and endoscopy all have limitations in their use as instruments to monitor disease activity in Crohn's disease (CD). Recently, 18F-FDG-PET and novel MRI techniques have been proposed as sensitive and specific methods to quantify the inflammatory load. The aim of this work was to assess the reliability, responsiveness and, to an extent, the validity of outcome measures in these modalities, in monitoring inflammatory activity over a 12-week interval. In addition, two receptors, TSPO and IL-2R were assessed on tissue specimens ex-vivo for their potential to act as alternative targets for molecular imaging in CD. Three distinct groups of patients were recruited, 2 of which participated in the clinical imaging study, and one to donate tissue for the laboratory work. Dual timepoint FDG-PET and MRI scanning was performed within 1 week (Group 1) to assess the test-retest reliability of the imaging outcome measures, and before, and twelve weeks into anti-TNFα therapy (Group 2) to assess their responsiveness indices. The third group contributed tissue during scheduled intestinal resection for assessment of TSPO and IL-2R interactions with their corresponding radioligands. To support the latter study, stored tissue sections were also obtained for immunohistochemical assessment of target receptor expression. Results on 22 patients show that PET endpoints such as SUVMAX and SUVMEAN have high responsiveness and reliability indices and demonstrated significant differences in anti-TNF responders compared to non-responders. The finding of luminal FDG signal may affect the face validity of the scan. MRI modalities appeared less responsive at three months. Analysis on ex-vivo specimens showed increased abundance of TSPO in normal bowel, but a relative over-expression in inflamed specimens which was not statistically significant. IL-2R appeared more abundant in transmural sections containing severe CD, but autoradiographic corroboration was not achieved for technical reasons.

Memory T cell recruitment to the non-lymphoid tissue of the intestine (the gut lamina propria) requires the interactions between integrin a4B7 and its ligand Mucosal Addressin Cell Adhesion Molecule-1 (MAdCAM-1), which is exclusively expressed in the gut microvasculature. T cell homing to small intestine also requires the expression of chemokine receptor CCR9 on T cells, whose ligand - the gut chemokine CCL25 - is abundantly expressed in the small intestine. The a4B7 integrin and chemokine receptor CCR9 are known as the "homing" molecules whose expression is "imprinted" on T cells by intestinal dendritic cells during antigen presentation. In the first part of this thesis, data from in vitro studies showed that other than retinoic acid, soluble tissue-derived factors such as the gut chemokine CCL25 and T cell growth factor IL-2 also contributed to the induction of gut-homing receptor acquisition. Moreover, the expression of a4B7 integrin and CCR9 appeared to be modulated differently. The present study suggests that under the influence of tissue-derived factors, T cell acquisition of a4B7 integrin is the "default" response at steady state, and that only fully activated T cells are able to express high levels of CCR9. In the second part of this thesis, T cell gut homing was investigated in the context of disease, i.e. inflammatory bowel disease (IBD), including Crohn's disease (CD) and ulcerative colitis (UC). IBD is characterized by excessive lymphocyte infiltration to the intestine. By quantitatively analyzing peripheral blood and intestinal lymphocyte subsets, proportional changes in the circulating gut-homing T cells (a4B7+) was found to be associated with CD. In addition, my data suggest that downregulation of T cell receptor zeta chain (TcRC) correlates to disease severity in both CD and UC. Measuring TcRC expression may thus provide an additional tool for the monitoring and management of IBD.