Gut microbiome has a profound effect on human health and disease. The burgeoning body of research investigating the gut microbiome has illustrated its important roles on not only the local inflammatory diseases, such as inflammatory bowel disease, but also on systemic immunological disorders, such as autoimmune diseases. Many of these associations were determined at the level of metagenome analysis. This provides categorical information on both taxonomic composition and predicted functional capacities of the microbial community. This thesis tackles the problem of how gut microbiota affect human diseases using both these categories of assessment.

Although the accumulating evidence suggested the important roles of gut microbiota in the human immune system and Crohn's disease (CD), the molecular mechanisms of how microbiome directly modulates immune responses and leads to CD progression remain unclear. To advance the current understanding of microbial roles in CD, we developed a systems biology approach to assess the human immune function in response to disease-associated microbial products. Progress in computational omics analysis of the microbiome in CD enables the functional inference and prediction of microbial metabolites. Both genetic and clinical studies suggested the involvement of human CD4+ T cells in CD progression. We therefore tabulated a set of microbial metabolites predicted to be differentially abundant in CD, and screened these microbial metabolites for their bioactivity in human CD4+ T cell functions. Our screen revealed 15 bioactive microbial metabolites, 3 previously reported and 12 unprecedented, with selective action on CD4+ T cell cytokine production. Mechanistic assessment of one novel microbe-derived metabolite, ascorbate, revealed apoptosis of activated human CD4+ T cells associated with selective inhibition of glycolytic energy metabolism. These findings suggest a substantial rate (11%) of bioactive metabolites among the predicted CD-associated metabolite reservoirs and provide evidence for novel modes of microbial activity targeting T cell metabolism.

Emerging studies have characterized the association of taxonomic composition with systemic autoimmune diseases with various clinical phenotypes and modes of pathologic immune responses. In this thesis, we assessed systemic sclerosis (SSc), a previously unstudied autoimmune disease with respect to the microbiome, notable for both systemic manifestations and gastrointestinal involvement. The metagenomic study of mucosal-luminal interface (MLI) samples revealed a unique microbial composition associated with disease state. Patients with SSc had decreased abundances of Faecalibacterium and Clostridium, and increased abundances of Bifidobacterium, Lactobacillus, Fusobacterium and γ-Proteobacteria compared with healthy controls. In addition, the increase of Fusobacterium species and the decrease of Bacteroides fragilis were associated with moderate/severe gastrointestinal symptoms. We conducted an independent study of faecal samples from two independent cohorts of patients, from United States and Norway, that revealed a similar group of microbial taxa associations, and a greater dysbiosis in patients from the United States. These studies enable the specific targets for intervention to avert or treat the gastrointestinal involvement in systemic sclerosis.

The study of gut microbiota is beginning to evolve from association toward causality. The functions of gut microbiota are not only restricted to local inflammatory responses but expanded to systemic immune regulations. We hypothesized that gut microbiome may regulate host immune cellular responses by producing metabolites. We investigated the bioactive microbial factors regulating Crohn's disease, and explored the potential microbial taxonomic biomarkers in a systemic autoimmune disorder, SSc. Those improved understandings of the microbiome activity and metabolism may inform therapeutic strategies and ultimately improve patient health.