207 may prevent experimental colitis or act as danger signals by mediating immune cell infiltration in the intestine.45,46 Although experimental evidence seems to be inconclusive, there is ample evidence indicating a role for aberrant MT homeostasis in IBD.47 This mechanism depends on the intracellular accumulation of zinc, which induces autophagy under chronic NOD2-stimulation. In IBD, the mucosal microbiota may contribute to the regulation of MT expression, intracellular zinc homeostasis and autophagy, thereby regulating intracellular bacterial clearance by intestinal macrophages. Findings from this study may support a putative role for Bacteroides in modulating MT activation, thereby contributing to intracellular redox homeostasis, zinc levels, macrophage autophagy, or even host defense against pathogens. Importantly, MTs and zinc regulation constitute potential therapeutic targets in IBD.44-47, 95-97 Individual gene–bacteria association analysis revealed distinctmucosal host–microbe interactions that largely overlap with those from the sparse-CCA analysis, but these provide more granular insight into the observed associations. Key examples of individual host gene–bacteria interactions are listed in Box 1. Amongst others, we demonstrate several host–microbe interactions that are putatively involved in immunological tolerance and prevention of autoimmunity (e.g. bifidobacteria and FOSL1/KLF2 expression), colorectal carcinogenesis (e.g. Anaerostipes and SMAD4, Akkermansia and YDJC) and inflammatory signaling (e.g. Oscillibacter and OSM expression). Notably, many of these associations are dependent on fibrostenotic disease, TNF-α-antagonist use and the degree of mucosal dysbiosis. In addition, deconvolution of the mucosal RNA-seq data reveal cell type–specific patterns of microbial interactions that warrant further study, for example through single-cell RNA-seq studies. Mucosal host–microbiota interactions have been investigated previously in both cohort (e.g. the HMP2 and Irish IBD) and experimental studies.12-16 Alongside several observations consistent with previous findings, we identify many novel host–microbe interactions. Differences in sample size, patient phenotypes and sample handling may be at least partially responsible for these observations. In our study, large groups of gene–bacteria associations are revealed that cover a wide range of molecular mechanisms potentially relevant in the context of IBD, including immune response pathways, cellular processes and a variety of metabolic pathways. Moreover, our study features the largest sample size so far,12-15 and this enabled us to perform an integrative analysis with respect to the large disease heterogeneity and identify novel host–microbiota crosstalk related to different clinical characteristics. However, several limitations also warrant recognition. As our study is of cross-sectional origin, we cannot assess the longitudinal dynamics of host– microbe interactions to discover signatures for therapy responsiveness or disease prognosis. Consequently, our associative results cannot establish potential causality between microbial abundances and host gene expression. Functional experiments are thus required to validate the biological relevance of the individual host–microbe interactions, as well as their behavior in microbial ecosystems. Finally, bowel preparation prior to the endoscopic procedure or crosscontamination between biopsy sites during endoscopy can affect the mucosal microbiota composition.21,50,98 Our results demonstrate a complex and heterogeneous interplay between mucosal microbiota and mucosal gene expression patterns that is concomitant with the strong impact of specific patient traits in a large cohort of patients with IBD. Our findings may guide development of Mucosal host-microbe interactions in IBD
RkJQdWJsaXNoZXIy MjY0ODMw