Project Summary Crohn’s disease (CD) and ulcerative colitis (UC) affect over 2 million individuals in the United States and are associated with considerable morbidity. Existing treatments achieve remission in fewer than 50% of patients. Further, despite achieving endoscopic remission, up to 30% of patients with CD or UC will relapse over the subsequent three years. Pathophysiologic mechanisms leading to relapse have not been well established. The central premise of our proposal is that despite endoscopic, there exists a local pro-inflammatory microbial milieu and transcriptional profile that favors disease relapse. We hypothesize that current clinical tools do not have sufficient resolution to capture this state. Existing cohorts, by recruiting patients in a heterogeneous state of active inflammation, cannot be used to infer mechanisms of loss of remission and inception of inflammation. A targeted effort that comprehensively and longitudinally profiles a homogeneous cohort of patients in deep remission is essential to define the dynamic relationship between microbial alterations, metabolomic, transcriptional, and proteomic perturbations, and onset of inflammation. Identifying deficient components favoring relapse also allows the development of intervention to replace these deficiencies, thereby extending remission. They will also provide clues and serve as starting points for development of novel therapies. In the first aim, we will recruit 300 patients with IBD in clinical and endoscopic remission and prospectively, systematically follow them for 3 years. We will comprehensively characterize such patients through serial sampling of mucosal and fecal microbiome, serum and fecal metabolome, and proteome in addition to detailed environmental exposure assessment and measurement of drug pharmacokinetics. We will determine the dynamic predictive utility of each of these parameters in defining future relapse from a state of quiescence. In the second aim, we will define the role of pro-inflammatory changes at the cellular level by performing single cell transcriptomic analysis from colonic and ileal biopsies in patients with quiescent CD and UC recruited as above. This will provide important insights into loss of control of inflammation at the tissue level that determines future clinical activity. The final study aim will train and validate a machine-learning predictive model to define the contribution of each additional biologic layer to inception of inflammation and to identify more robust biomarkers of a state of sustained remission. Defining the molecular basis of future relapse in patients in deep remission will provide insights into the ‘pre-disease’ state, allowing for identification of immune pathways of relevance in preventing disease. Defining the fundamental mechanisms through which disease inception occurs from quiescence is critically important to inform key steps in the pathogenesis of these complex diseases, which in turn, will offer oppo...