Gut microbial proinflammatory proteolytic activity is detected before ulcerative colitis onset Heather J Galipeau1, Alberto Caminero1, Williams Turpin2,3, Miriam Bermudez-Brito1, Alba Santiago1, Josie Libertucci1, Marco Constante1, Juan Antonio Raygoza Garay2,3, Gaston Rueda1, Sarah Armstrong1, Alex Clarizio1, Michelle I. Smith2, Michael G. Surette1, Premysl Bercik1, The CCC GEM Project Research Consortium, Kenneth Croitoru2,3, Elena F. Verdu1 1. Farncombe Family Digestive Health Research Institute, McMaster University, Hamilton, Canada. 2. Zane Cohen Centre for Digestive Diseases, Lunenfeld Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada. 3. Division of Gastroenterology, Department of Medicine, University of Toronto, Toronto, Ontario, Canada Introduction: Altered gut microbiota composition and function has been associated with inflammatory bowel diseases (IBD) including ulcerative colitis (UC), but causality and mechanisms remain unknown. Most studies have examined patients with active or treated disease, making it difficult to discern if the described changes are attributable to the effect of inflammation. Little is known about microbial compositional or functional changes that occur before disease onset. Aims: We studied a longitudinal cohort of subjects at risk for IBD to define the fecal microbial composition and function in subjects prior to UC onset (pre-UC) and at diagnosis (post-UC), and in matched at-risk subjects that remained healthy. Methods: Fecal samples were collected from healthy individuals at-risk for IBD (pre-UC; n=13) and subjects were followed longitudinally until UC diagnosis (post-UC; n=9), at which point another fecal sample was collected. Fecal samples from a cohort of matched at-risk individuals that did not develop UC were used as healthy controls (HC; n=48). We applied 16S rRNA gene sequencing, next generation shotgun sequencing, in vitro proteolytic assays and gnotobiotic colonizations to define the microbial fecal composition and proteolytic function. Results: Microbiota composition of post-UC subjects was different from HC and pre-UC. Post-UC subjects clustered separately from pre-UC and HC, based on bray-curtis and unweighted UniFrac, had reduced alpha-diversity compared to pre-UC and HC and had reduced abundance of Adlercreutzia compared to pre-UC and HC. In vitro functional analysis revealed increased fecal proteolytic and elastase activity in pre-UC and post-UC samples, compared to HC. Metagenomics identified pathways and gene families related to protein metabolism and proteases/peptides that were significantly different between HC and pre-UC samples, suggesting a bacterial component to the pre-UC proteolytic signature. Elastase activity inversely correlated with the relative abundance of Adlercreutzia, and other potentially beneficial taxa, and directly correlated with Bacteroides vulgatus, a known proteolytic taxon. High elastase activity was confirmed in Bacteroides isolates from fecal samples. Bacterial contribution and functional significance of the proteolytic signature was investigated in germ-free adults and litters born from dams colonized with HC, pre-UC or post-UC microbiota. Mice colonized as adults, or neonatally with pre-UC microbiota developed higher fecal proteolytic activity and an inflammatory immune tone compared with HC colonized mice. Conclusion: We have identified increased fecal proteolytic activity that precedes clinical diagnosis of UC and associates with gut microbiota changes. This may constitute a non-invasive biomarker of inflammation to monitor at-risk populations that can be targeted therapeutically with anti-proteases.