Direct signaling of TL1A-DR3 on fibroblasts induces intestinal fibrosis in vivo

Abstract

Tumor necrosis factor-like cytokine 1A (TL1A, TNFSF15) is implicated in inflammatory bowel disease, modulating the location and severity of inflammation and fibrosis. TL1A expression is increased in inflamed mucosa and associated with fibrostenosing Crohn's disease. Tl1a-overexpression in mice causes spontaneous ileitis, and exacerbates induced proximal colitis and fibrosis. Intestinal fibroblasts express Death-receptor 3 (DR3; the only know receptor for TL1A) and stimulation with TL1A induces activation in vitro. However, the contribution of direct TL1A-DR3 activation on fibroblasts to fibrosis in vivo remains unknown. TL1A overexpressing naïve T cells were transferred into Rag^-/- , Rag^-/- mice lacking DR3 in all cell types (Rag^-/-Dr3^-/-), or Rag^-/- mice lacking DR3 only on fibroblasts (Rag^-/-Dr3^∆Col1a2) to induce colitis and fibrosis, assessed by clinical disease activity index, intestinal inflammation, and collagen deposition. Rag^-/- mice developed overt colitis with intestinal fibrostenosis. In contrast, Rag^-/-Dr3^-/- demonstrated decreased inflammation and fibrosis. Despite similar clinical disease and inflammation as Rag^-/-, Rag^-/-Dr3^∆Col1a2 exhibited reduced intestinal fibrosis and attenuated fibroblast activation and migration. RNA-Sequencing of TL1A-stimulated fibroblasts identified Rho signal transduction as a major pathway activated by TL1A and inhibition of this pathway modulated TL1A-mediated fibroblast functions. Thus, direct TL1A signaling on fibroblasts promotes intestinal fibrosis in vivo. These results provide novel insight into profibrotic pathways mediated by TL1A paralleling its pro-inflammatory effects.

Figure 1

Fibroblast-selective DR3 deficiency results in severe disease activity: (A–D) Percent body weight loss, stool consistency, stool blood, and composite disease activity index are shown for the adoptive transfer model of colitis, at indicated time points for transfers of naïve Tl1a-Tg T cells into Rag^−/−, Rag^−/−Dr3^−/−, or Rag^−/−Dr3^∆Col1a2 mice. Data are represented as means ± SEM; * indicates p < 0.05, **p < 0.01 Rag^−/−Dr3^−/−vs Rag^−/−; # indicates p < 0.05 Rag^−/− Dr3^−/− vs Rag^−/−Dr3^∆Col1a2; n = 9–10 mice per group. (E–F) Representative gross colonic specimens, quantitated macroscopic pathology scores, and colon lengths; representative H&E cecal sections and quantitated histopathological scores are shown; * indicates p < 0.05, **p < 0.01. Pooled data of 3 independent experiments are represented.

Figure 2

Fibroblast-selective DR3-deficiency attenuates intestinal collagen deposition: Representative Sirius red staining of collagen deposition in cecal sections and quantitated percent mucosal area by ImageJ software shown for the adoptive transfer model of colitis of Rag^−/−, Rag^−/−Dr3^−/−, or Rag^−/−Dr3^∆Col1a recipient mice. * indicates p < 0.05, **p < 0.01; ns indicates not significant. Pooled data of 3 independent experiments are represented.

Figure 3

Fibroblast-selective DR3-deficiency reduces intestinal fibroblast activation: (A) Immunofluorescent staining of vimentin (green) and αSMA (red) from cecal sections of adoptive transfer Rag^−/−, Rag^−/−Dr3^−/−, or Rag^−/−Dr3^∆Col1a2 recipient mice. White arrows denote myofibroblasts that co-localize vimentin and αSMA at 400× magnification. (B) Total numbers and percentage of myofibroblasts over total vimentin-positive cells per mm² of tissue area were quantitated in an automated fashion by TissueGnostics software and displayed; * indicates p < 0.05, **p < 0.01; ns indicates not significant.

Figure 4

Fibroblast-selective DR3-deficiency attenuates intestinal fibroblast migration ex vivo: (A) Number of adherent cells/visual field indicated during adhesion assays at time points of 20 min and 80 min for fibroblasts isolated ex vivo from adoptive transfer Rag^−/−, Rag^−/−Dr3^−/−, or Rag^−/−Dr3^∆Col1a2 recipient mice. (B) Representative images of gap-closure assay after simulated wound on fibroblasts isolated ex vivo from adoptive transfer Rag^−/−, Rag^−/−Dr3^−/−, or Rag^−/−Dr3^∆Col1a2 recipient mice at initial wound (t₀) and 8 h after migration (t₈) with relative % area of gap closed quantitated; representative of 3 independent experiments; * indicates p < 0.05.

Figure 5

Transcriptomic analysis of Tl1a-stimulated intestinal fibroblasts identifies cytoskeletal contraction-related genes and the Rho signaling pathway: (A) Heatmap generated from transcriptome analysis of isolated cecal fibroblasts from three individual mice shown as untreated (a, b, c) and Tl1a-treated (a + Tl1a, etc.) pairs; BRB array tools, (brb.nci.nih.gov/BRB-ArrayTools). (B) Independent validation cohort was used to confirm several genes in the index group; matched pairs displayed as untreated (ut) or TL1A-treated fibroblasts with direct increase in expression for each sample shown, * indicates p < 0.05. (C) Genetic ontology pathway enrichment of differentially expressed genes induced by TL1A in (A).

Figure 6

Inhibition of Rho kinase pathway modulates morphology and function of intestinal fibroblasts: (A) Gap closure assay and morphology of migrating intestinal fibroblasts isolated from three individual Tl1a-Tg mice, in the presence (or absence) of 25uM ROCK inhibitor Y27632, shown as treated and un-treated pairs. Blue arrow denotes direction of migration with solid vs dotted line highlighting cell extensions at migrating front; upper panels 100× objective, lower panels 200× objective. (B) Collagen contraction assay of intestinal fibroblasts isolated and treated as in (A). (C) Collagen contraction assay of WT or DR3^−/− intestinal fibroblasts treated with 100 ng/mL mouse recombinant Tl1a, 25uM ROCK inhibitor Y27632, or both. Data normalized to genotype control with bars representative of means; * indicates p < 0.05, **p < 0.01.