Comparative Transcriptomics of IBD Patients Indicates Induction of Type 2 Immunity Irrespective of the Disease Ideotype

Abstract

Inflammatory cytokines initiate and sustain the perpetuation of processes leading to chronic inflammatory conditions such as inflammatory bowel diseases (IBD). The nature of the trigger causing an inflammatory reaction decides whether type 1, type 17, or type 2 immune responses, typically characterized by the respective T- helper cell subsets, come into effect. In the intestine, Type 2 responses have been linked with mucosal healing and resolution upon an immune challenge involving parasitic infections. However, type 2 cytokines are frequently elevated in certain types of IBD in particular ulcerative colitis (UC) leading to the assumption that Th2 cells might critically support the pathogenesis of UC raising the question of whether such elevated type 2 responses in IBD are beneficial or detrimental. In line with this, previous studies showed that suppression of IL-13 and other type 2 related molecules in murine models could improve the outcomes of intestinal inflammation. However, therapeutic attempts of neutralizing IL-13 in ulcerative colitis patients have yielded no benefits. Thus, a better understanding of the role of type 2 cytokines in regulating intestinal inflammation is required. Here, we took a comparative transcriptomic approach to address how Th2 responses evolve in different mouse models of colitis and human IBD datasets. Our data show that type 2 immune-related transcripts are induced in the inflamed gut of IBD patients in both Crohn's disease and UC and across widely used mouse models of IBD. Collectively our data implicate that the presence of a type 2 signature rather defines a distinct state of intestinal inflammation than a disease-specific pathomechanism.

Keywords: IBD; chronic inflammation; intestine; type 2 immunity; ulcerative colitis.

Figure 1

Transcriptomic analyses of type 2 identifiers in multiple mouse models of colitis. (A) percent change in body weight from baseline and DSS treatment paradigm. Time points representing inflammation are represented in red (B) Heatmap depicting the normalized expression of the indicated genes from four different time-points of dextran sulfate sodium (DSS) induced colitis D3 = day 3, D8 = day 8, D12 = day 12, and D19 = day 19. The text boxes below represent the treatment type and duration. Hierarchical clustering shows clear segregation of two distinct groups of genes. (C) Representative photomicrographs of H&E-stained tissue sections from the indicated colitis mouse models. DSS, acute dextran sulfate sodium colitis; cDSS, chronic dextran sulfate sodium colitis; OxC, Oxazolone induced colitis; TC, T-cell adoptive transfer colitis. (D) Heatmap showing normalized expression levels of the indicated genes the type of regimen used to induce colitis is depicted below. Numbers in each quadrant represent the p-value and the legend denotes the fold changes in expression ratios.

Figure 2

Analysis of multiple IBD cohorts using the type 2 identifier gene set. (A) Heatmap of the normalized ileal gene expression profiles from the indicated cohorts of CD patients. Where available, gender, inflammation type and ulceration status are depicted below. (B) Heatmap of the normalized gene expression profiles of the indicated genes in the colonic and rectal tissues from IBD patients. Where available, gender, disease diagnosis (UC, Ulcerative Colitis; iCD, ileal Crohn's disease; and cCD, colonic Crohn's disease) and treatments (5-ASA, 5-aminosalicylic acid; CSIV, Cyclosporin intravenous; and CSOral, Cyclosporin oral) are indicated below. Numbers in each quadrant represent the p-value and the legend denotes the fold changes in expression ratios.

Figure 3

Network analysis of the six type 2 associated genes similarly regulated in human IBD and mouse models of colitis. Radial interaction map of type 2 associated marker genes common to both human IBD and mouse models of colitis (the six central circles, shaded gray). Each non-shaded gray circle represents a predicted pathway interactor (aqua lines) or physical interactor (mauve lines) of the six core factors. Size of each circle represents the rank generated by an automatically selected weighting method in Cytoscape.