INFLAMMATORY CYTOKINES AND MICROBIAL LIGANDS AND METABOLITES INTERACT TO MODULATE DUOX2 EXPRESSION AND ACTIVITY

Abstract

Abstract Background The pathogenesis of inflammatory bowel diseases (IBD) is characterized by a dysregulated crosstalk between the host and the microbiome that leads to the development of inflammation and dysbiosis. Dysbiosis in IBD involves an expansion of Proteobacteria and a reduction of Firmicutes, particularly of butyrate-producing species such as Faecalibacterium prausnitzii. The epithelial NADPH oxidase dual oxidase 2 (DUOX2), which prevents bacterial colonization of the mucosa through the production of hydrogen peroxide (H2O2), is the only gene consistently altered in IBD patients before the onset of disease. However, the involvement of DUOX2 in IBD is not well understood. We aimed to define how inflammation and the microbiota regulate DUOX2 activity. Methods C57BI/6J males raised in specific-pathogen free (SPF) and germ-free (GF) conditions underwent a model of dextran sulfate sodium (DSS)-induced colitis for 6 days. We obtained colon specimens for histopathology and isolation of colon epithelial cells (CEC). We stimulated colonoids from wild-type (WT), toll-like receptor 4-KO (Tlr4-KO), and Duox2-KO mice with IFNγ, heat-killed adherent invasive Escherichia coli (AIEC) strain LF82, and heat-killed F. prausnitzii strain A2-165 and determined expression of Duox2, Duoxa2, and production of H2O2. We also treated colonoids with butyrate, a microbial metabolite with anti-inflammatory properties. Extracellular H2O2 production was analyzed by means of the Amplex Red assay, whereas gene expression was determined by qPCR. Results SPF mice undergoing DSS-induced colitis developed overt inflammation that was accompanied by upregulation of Duox2 and Duoxa2, as well as increased production of H2O2 in freshly isolated CECs. DSS-treated GF mice developed a mild inflammation that also caused increased H2O2 production and Duoxa2 upregulation. WT colonoid stimulation with IFNγ and the Crohn’s disease-associated pathobiont AIEC induced Duox2 and Duoxa2 expression, whereas heat-killed F. prausnitzii did not. Similarly, both IFNγ and AIEC promoted epithelial production of H2O2 in WT colonoids but not Duox2-KO colonoids, indicating that epithelial release of H2O2 in response to these stimuli is mediated by DUOX2. Response to AIEC additionally required functional TLR4. Although heat-killed F. prausnitzii did not alter H2O2 production, its metabolite butyrate caused a significant blockade in the release of H2O2 in response to both IFNγ and AIEC. Conclusions Our results show that both inflammation and pathobionts induce the expression and activity of DUOX2, which begets more inflammation. We posit that specific depletion of pathobionts or restitution with butyrate-producing bacteria such as F. prausnitzii may be beneficial in IBD.