Host intestinal immune gene signatures and microbial dysregulations expose potential mechanisms in the pathogenesis of inflammatory bowel diseases (IBD). Profiling of mucosa-attached microbiota allows the understanding of locally present microbial communities and their immediate impact on the host. This study evaluated interactions between host mucosal gene expression and intestinal mucosa-attached microbiota in IBD.
Intestinal mucosal bulk RNA-sequencing data was combined with mucosal 16S rRNA gene sequencing data from 696 intestinal biopsies derived from 337 patients with IBD (181 with Crohn’s disease [CD] and 156 with ulcerative colitis [UC]) and 16 non-IBD controls. Hierarchical all-against-all associations testing (HAllA) was used to assess factors affecting host gene expressions and microbiota. Mucosal cell enrichments were predicted by deconvolution. Linear mixed interaction models were used to investigate host-microbiota interactions, adjusting for age, sex, BMI and batch effects. Variation explanation analysis was performed by Lasso regression.
In total, 15,934 intestinal genes and 113 microbial taxa were identified and included in subsequent analyses. Host intestinal gene expressions were characterized by tissue- and inflammation-specificity, whereas intraindividual variability of the mucosal microbiota dominated over disease location and inflammation effects. We observed forty associations between the mucosal expression of genes and the abundance of specific microbes independent of dysbiosis (FDR<0.05). Examples include a positive association between aryl hydrocarbon receptor (AHR) and Bifidobacterium, and a negative association between interleukin 18 receptor 1 (IL18R1) and Lachnoclostridium. Furthermore, 112 gene-microbiota interactions changed in patients with microbial dysbiosis compared to non-dysbiosis (FDR<0.05). These interactions were enriched in immune-related and extracellular matrix organization pathways. For example, the IL1R1 gene was positively associated with Collinsella abundance in non-dysbiotic patients, whereas an inverse association was observed in high dysbiosis. Finally, the presence of mucosal microbial taxa explained up to 10% of the variation in cell type enrichment, affecting epithelial cells, macrophages and regulatory T-cells.
Interactions between host intestinal gene expressions and mucosa-attached microbiota are disrupted in patients with IBD. Furthermore, mucosal microbiota are highly personalized and potentially contribute to intestinal cell type alterations. Our study unravels key immune-mediated molecular pathways and relevant bacteria in intestinal tissue, which may guide drug development and precision medicine in IBD.