The TNFSF12/TWEAK Modulates Colonic Inflammatory Fibroblast Differentiation and Promotes Fibroblast-Monocyte Interactions

Abstract

Fibroblasts acquire a proinflammatory phenotype in inflammatory bowel disease, but the factors driving this process and how fibroblasts contribute to mucosal immune responses are incompletely understood. TNF superfamily member 12 (TNFSF12, or TNF-like weak inducer of apoptosis [TWEAK]) has gained interest as a mediator of chronic inflammation. In this study, we explore its role as a driver of inflammatory responses in fibroblasts and its contribution to fibroblast-monocyte interaction using human primary colonic fibroblasts, THP-1 and primary monocytes. Recombinant human TWEAK induced the expression of cytokines, chemokines, and immune receptors in primary colonic fibroblasts. The TWEAK upregulated transcriptome shared 29% homology with a previously published transcriptional profile of inflammatory fibroblasts from ulcerative colitis. TWEAK elevated surface expression of activated fibroblast markers and adhesion molecules (podoplanin [PDPN], ICAM-1, and VCAM-1) and secretion of IL-6, CCL2, and CXCL10. In coculture, fibroblasts induced monocyte adhesion and secretion of CXCL1 and IL-8, and they promoted a CD14high/ICAM-1high phenotype in THP-1 cells, which was enhanced when fibroblasts were prestimulated with TWEAK. Primary monocytes in coculture with TWEAK-treated fibroblasts had altered surface expression of CD16 and triggering receptor expressed on myeloid cells-1 (TREM-1) as well as increased CXCL1 and CXCL10 secretion. Conversely, inhibition of the noncanonical NF-κB pathway on colonic fibroblasts with a NF-κB-inducing kinase small molecule inhibitor impaired their ability to induce a CD14high phenotype on monocytes. Our results indicate that TWEAK promotes an inflammatory fibroblast-monocyte crosstalk that may be amenable for therapeutic intervention.

Graphical abstract

FIGURE 1.

TWEAK induces an inflammatory profile in colonic fibroblasts. (A) Volcano plots depicting gene distribution of upregulated and downregulated transcripts in three independent cultures of primary colonic fibroblasts treated with TWEAK for 24 h, compared with vehicle. The total number of downregulated and upregulated genes and some representative genes names are depicted. (B) Summary of the top 20 molecular functions associated to the TWEAK upregulated transcriptome (GO_Terms analysis). (C–E) Heat maps depicting a selection of genes representative of top pathways dysregulated by TWEAK, including cell adhesion molecules, cytokines, chemokines, immune receptors, and TNF signaling (C), Ag processing and presentation (D) and extracellular matrix remodeling (E). In all cases data are >1.5-fold, adjusted p < 0.05.

FIGURE 2.

The TWEAK-induced signature is transcriptionally aligned with the inflammatory stroma in UC. (A) Venn diagrams comparing the TWEAK-upregulated transcriptome (n = 3, >1.5-fold, adjusted p < 0.05) to each of the stromal populations (S1–S4) previously described by Kinchen et al. (3). (B) Summary of the percentage of alignment between the TWEAK-downregulated and -upregulated transcriptome and that of each stromal population. (C) Violin plot depicting the presence of TNFRSF12A/Fn14-positive cells across the four stromal populations defined by Kinchen et al. (3). (D) Dot plot presenting all genes common to TWEAK-upregulated and S4 transcriptomes, comparing the percentage of positive cells (thickness of dots) and expression level (darkness) comparing these genes between each stromal population.

FIGURE 3.

TWEAK induces the expression of surface adhesion molecules and inflammatory cytokines. (A and B) Representative flow cytometry pseudo-color plots with gates showing the frequency of PDGFRα-, PDPN-, ICAM-1–, and VCAM-1–high cells (A) and corresponding quantification normalized to 50,000 live events (B) in vehicle- and TWEAK-treated colonic fibroblasts. (C and D) Representative histograms showing the intensity of PDGFRα, PDPN, ICAM-1, and VCAM-1 (C) and the corresponding mean fluorescence intensity (D) relative to vehicle. Individual data points represent independent experiments. (E) Quantification of secreted CCL2, IL-6, and CXCL10 in vehicle and TWEAK-stimulated human colonic fibroblasts. In all experiments the individual data points represent independent experiments (n ≥ 4). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

FIGURE 4.

TWEAK-induced fibroblasts recruit and polarize THP-1 cells. (A) Schematic representation of the direct coculture model. (B) Representative flow cytometry plots showing the percentage of CD11b^high THP-1 cells among the suspended fraction, alone, or cocultured with fibroblasts. (C) Quantification of CD11b^high population frequency in (B) normalized to 50,000 cells. (D and E) Representative flow cytometry plots and quantification of the percentage of CD14^high THP-1 cells in suspension in monoculture and in cocultures with either vehicle or TWEAK-stimulated fibroblasts. (F and G) Representative flow cytometry plots and quantification of the percentage of ICAM-1^high THP-1 cells in the suspension fraction when cocultured with vehicle or TWEAK-stimulated colonic fibroblasts and heatmap overlay representing the fluorescence intensity of CD14. (H) Representative flow cytometry plots showing the gating strategy used to segregate CCR2^high/CD90^low THP-1 cells from CCR2^low/CD90^high colonic fibroblasts in the adherent fraction in coculture. (I) Quantification of the frequency of CCR2^high/CD90^low THP-1 cells adhered when cocultured with vehicle or TWEAK-stimulated fibroblasts. (J and K) Representative flow cytometry plots and quantification of the percentage of CD14^high cells within the CCR2^high (THP-1 cells) population identified in the adherent coculture fraction. (L) ELISA of CXCL8/IL-8 and CXCL1 in supernatants from fibroblasts and THP-1 cells in monoculture or in coculture with either vehicle or TWEAK. Individual data points represent independent experiments with a minimum of three independent fibroblast and monocyte cultures used. Symbols represent THP-1 cell coculture with two different strains of primary colonic fibroblasts: CCD-18Co (○) or PB-H-6231 (▽). For cell frequency, data are normalized to 50,000 live events per condition. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

FIGURE 5.

TWEAK-induced fibroblasts polarize primary monocytes. (A–C) Representative flow cytometry plots showing the frequency of CD90^high and CD90^low (A), CCR2^high and CCR2^low (within CD90^low, B) and CD16^high, CD16^int and CD16^low (within CD90^low/CCR2^high) primary monocytes cultured alone (with or without TWEAK) or in direct coculture with vehicle- or TWEAK-treated fibroblasts. The color map overlay in (B) represents the CD16 fluorescence intensity. (D) Quantification of relative frequency of CD16^int and CD16^high subpopulations (CD16^int/high) normalized to 50,000 CD90⁻/CCR2^high events. (E) Representative flow cytometry plots showing the frequency of TREM-1^high monocytes (within CD90^low/CCR2^high) cocultured with vehicle- or TWEAK-treated fibroblasts. (F) Quantification of TREM-1^high cells in (E) normalized to 50,000 CD90⁻/CCR2^high events. (G) ELISA of CXCL1 and CXCL10 in supernatants from vehicle- or TWEAK-treated fibroblast in monoculture or coculture with monocytes. Individual data points represent independent experiments (n ≥ 3). *p < 0.05, **p < 0.01. ns, not significant.

FIGURE 6.

Noncanonical NF-κB regulates the interaction between TWEAK-primed fibroblasts and monocytes. (A) String analysis depicting K-means clustering of common genes between TWEAK-upregulated and S4 transcriptomes, illustrating a cluster associated with inflammatory signaling via NF-κB. (B) Representative immunoblot and densitometry analysis for p100/p52 in whole-cell lysates from human colonic fibroblasts treated with vehicle or TWEAK for 1, 4, 24, and 48 h. (C) Quantification of surface expression of VCAM-1 and ICAM-1 by flow cytometry in fibroblasts treated with vehicle or TWEAK in the presence of the NIK inhibitor (NIK SMI1). (D) Quantification of CCL2 secretion (ELISA) by fibroblasts treated with vehicle or TWEAK in the presence of the NIK inhibitor. (E) Representative flow cytometry plots showing the frequency of CD14^high THP-1 cells in cocultures with fibroblasts treated with vehicle or TWEAK in the presence of the NIK inhibitor. (F) Quantification of the population frequency in (E) normalized to 50,000 live events. Individual data points represent independent experiments (n ≥ 3). Symbols represent two different strains of human primary colonic fibroblasts: CCD-18Co (○) or PB-H-6231 (▽). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.