Biomarker Discovery and Extracellular Vesicle Proteomic Signatures in Pediatric Inflammatory Bowel Disease

Abstract

Background: Reliable biomarkers are needed to evade the risk of injury, invasiveness and discomfort of endoscopies, which are required for inflammatory bowel disease (IBD) diagnosis and extent of disease assessment in ulcerative colitis (UC) patients. The need for biomarkers is accentuated in children, wherein the most frequently used IBD biomarker yields low specificity. Proteomics of clinical samples or their enriched components is a means to evaluate and identify alterations in proteins reflective of disease, with the potential for use as biomarkers and for providing insight on disease pathogenesis. Methods: Proteins were isolated from the intestinal mucosal-luminal interface (MLI), collected from the ascending and descending colon of pediatric treatment-naive patients. The intestinal MLI proteomes of 42 IBD and 18 control patients were analyzed by high resolution mass spectrometry (HRMS). Multivariate analysis and receiver operating characteristics curves were performed to develop protein biomarker panels to discriminate IBD from control, and for UC extent of disease. ELISAs were used to assess a subset of biomarker candidates in stool samples from an independent pediatric cohort (n=24). Extracellular vesicles (EVs) were isolated by ultracentrifugation from the intestinal MLI of 11 IBD and seven control patients, and analyzed by electron microscopy, nanoparticle tracking analysis and HRMS. Results: A biomarker panel of four proteins classified patients as either controls or active IBD with 97.5% accuracy. A second biomarker panel correctly classified 100% of UC patients as presenting with pancolitis or non-pancolitis. The differential protein expression of two biomarker candidates (catalase and leukotriene A-4 hydrolase) identified from the intestinal MLI was comparable in stool samples. Comparison of EV proteomes isolated from IBD patients and controls identified differential expression of processes related to host defense and immunity. Conclusions: Proteomic analysis of clinical samples identified differentially expressed proteins that can classify IBD patients from non-IBD controls and distinguish UC patients with pancolitis from those without pancolitis; proteins identified in intestinal aspirates displayed consistent differential expression in stool. Furthermore enrichment of EVs from the intestinal MLI indicates that these may contribute to the dysregulated host response against the intestinal microbiota which is observed in IBD.