P052 Comprehensive proteomic profiling of stool from UC and CD patients reveals expected and novel mucosal pathobiology, enabling identification of candidate non-invasive biomarkers to monitor mucosal disease activity

Abstract

Background Inflammatory bowel disease (IBD) is a multifactorial disorder characterized by chronic gastrointestinal (GI) inflammation. In clinical practice, physicians urgently need biomarkers to monitor changes in mucosal disease to manage treatment. Non-invasive tools such as fecal biomarkers could allow for frequent monitoring of mucosal disease activity and may reflect inflammation of the entire GI tract. Towards identification of novel non-invasive fecal biomarkers, we performed a comprehensive, unbiased, analysis of the human fecal proteome from ulcerative colitis (UC), Crohn’s disease (CD) patients, and healthy controls (HC), focusing on human secreted proteins using proteomic approaches. Methods Fecal extraction processes optimized for recovery of secreted proteins were applied to stool samples of mild to severe UC and CD patients and HCs, 20 samples per group. An unbiased survey of the human fecal proteome of the IBD and HC stool extracts was performed using both data-dependent acquisition and data-independent acquisition methods on QExactive HF and Fusion Lumos mass spectrometers. Data were analyzed in Spectronaut for peptide identification, followed by analysis in MS Stats for protein quantification. Fecal calprotectin (fCAL) levels were measured in these stool samples using the Buhlmann EK-CAL immunoassay kit to correlate fCAL levels to proteomic calprotectin (S100A8/S100A9 heterodimer). Results We identified and quantified 594 differentially expressed proteins in stool using proteomics, including known proteins perturbed in IBD pathobiology. Elevated levels of neutrophilic proteins, e.g. fCal, neutrophil elastase, lactoferrin and leucine-rich Α-2 glycoprotein 1 were detected as well as proteins linked to rectal bleeding and ulceration e.g. hemoglobin subunit alpha 1 and beta. Pancreatic enzymes, suggested to be associated with active disease, were also detected. Overall, UC and CD contained many proteins not detected in HC. Differential abundances were observed between UC and CD patients, possibly reflecting differences in mucosal pathophysiology. The abundance of proteomic calprotectin correlated highly with the fCAL levels as determined by immunoassay (Spearman rank order > 0.9 for S100A8 and S100A9 to fCAL), validating the fecal proteomic approach to uncover known and novel proteins of IBD biology. Conclusion We developed and qualified methods to perform fecal proteomics on IBD stool samples, enabling identification of known and novel proteins of mucosal UC and CD disease biology. In depth analysis of these proteomic datasets may reveal novel fecal biomarker candidates, targeted to reflect mucosal disease activity and potentially serve as non-invasive surrogates for endoscopy and help guide treatment in IBD patients.