Role of innate immunity and the microbiome in colitis associated dysplasia

PROJECT ABSTRACT The two greatest fears expressed by patients with ulcerative colitis (UC) are developing colon cancer and losing their colon. Although colitis-associated cancer is less common, dysplasia is more common and often results in colectomy. Our proposal leverages the work of the previous funding period and advances in the field to take a translational approach to unraveling dysplasia in colitis. The focus of our studies has been the link between innate immune signaling and intestinal bacteria leading to colitis-associated neoplasia. Our results have led us to focus on the role of dual oxidase 2 (DUOX2) in dysplasia. DUOX2 is a NADPH oxidase that catalyzes the conversion of oxygen into hydrogen peroxide (H2O2) upon interaction with the maturation factor DUOXA2. It is consistently upregulated in biopsies from IBD patients and DUOX2/DUOXA2 expression is further increased in patients who have had dysplasia. Our group has shown that both inflammatory and microbial signals induce the expression and activity (H2O2 production) of DUOX2 in colonic epithelial cells. We have proven that chronic activation of DUOX2 leads to the formation of tumors, which is almost totally abrogated by inactivating DUOX2 in the epithelium. In the current proposal, we hypothesize that IBD-associated dysbiosis activates Duox2 and local production of H2O2 leading to epithelial barrier dysfunction, recruitment of tumor-promoting myeloid derived suppressor cells (MDSCs), and generation of a tumorigenic microbiome in a feed forward loop. This is pursued in the following specific aims: 1) Determine the dependence of inflammatory colonic dysplasia on epithelial Duox2 signaling. Here we will investigate epithelial barrier dysfunction and DNA damage pathways caused by epithelial Duox2 activation using colonoid models (human and murine). Using humanized germ-free mice, we will determine if the UC-dysplasia microbiome is sufficient to cause DUOX2-mediated permeability defects and DNA damage. 2) Dissect the role of tumor-promoting MDSCs on dysplasia development in the setting of DUOX2-mediated oxidative stress. Our preliminary data demonstrate that TLR4-driven tumors have enhanced recruitment of tumor supporting MDSCs. Here we will identify DUOX2-dependent epithelial factors (chemokines/cytokines) and the transcriptional (single cell-RNA seq) and functional (T cell suppressive) characteristics of MDSCs in our murine models of tumorigenesis and UC patients with dysplasia. 3) Identify targetable microbial pathways linked to Duox2 activation and dysplasia. Transfer of the microbiome from tumor-susceptible villin-TLR4 mice is sufficient to transfer CAC susceptibility. New metagenomic and metabolomic data show clear differences between UC dysplasia and UC without dysplasia. We will use our mouse models and UC dysplasia samples to perform deep sequencing and metabolomic characterization of the tumor-promoting microbiome. Metabolites will be tested for their ability to induce or inhibit H2O2/DUOX2 in colonoids. The work proposed herein will provide the mechanistic justification for subsequent human studies to target Duox2 and specific microbial pathways to halt progression from UC inflammation to dysplasia.