TGF-β Signaling in Dendritic Cells Governs Colonic Homeostasis by Controlling Epithelial Differentiation and the Luminal Microbiota

Sozaburo Ihara, Yoshihiro Hirata, Takako Serizawa, Nobumi Suzuki, Kosuke Sakitani, Hiroto Kinoshita, Yoku Hayakawa, Hayato Nakagawa, Hideaki Ijichi, Keisuke Tateishi and Kazuhiko Koike
J Immunol June 1, 2016, 196 (11) 4603-4613; DOI: https://doi.org/10.4049/jimmunol.1502548

Abstract

Dendritic cells (DCs) mediate host immune responses to gut microbes and play critical roles in inflammatory bowel disease. In this study, we examined the role of TGF-β signaling in DCs in colonic homeostasis. CD11c-cre Tgfbr2fl/fl mice developed spontaneous colitis, and CD11c-cre Tgfbr2fl/+ mice exhibited susceptibility to dextran sulfate sodium–induced colitis. Colitis in these mice was characterized by goblet cell depletion and dysbiosis caused by Enterobacteriaceae enrichment. Wild-type mice gavaged with Enterobacteriaceae from CD11c-cre Tgfbr2fl/fl mice feces showed severe colitis after dextran sulfate sodium treatment, whereas those treated with Notch inhibitor exhibited attenuated colonic injury with increased goblet cell numbers, thickened mucus layer, and fewer fecal Enterobacteriaceae. Wild-type mice transplanted with CD11c-cre Tgfbr2fl/fl bone marrow developed colitis showing increased Jagged1 and Jagged2 in DCs, increased Hes1 levels in epithelium, and goblet cell depletion. These findings suggest that TGF-β signaling in DCs regulates intestinal homeostasis by modulating epithelial cell differentiation and fecal microbiota.

Introduction

There are two main inflammatory bowel diseases (IBDs): ulcerative colitis (UC) and Crohn disease. IBD is a chronic intestinal inflammation of unknown etiology (1) and caused by the complex interaction between genetic and environmental factors, but its pathogenesis has not been elucidated fully (2). Genome-wide association studies identified >100 loci that are associated with the susceptibility to IBD (3), including Smad3, which plays a role in TGF-β signaling (4). TGF-β is an immunoregulatory cytokine that suppresses the inflammatory response in the normal gut and contributes to immune tolerance, whereas impaired TGF-β signaling is associated with the development of IBD (5, 6).

Microbial changes are also important for the pathogenesis of IBD (2, 7). Many studies have revealed inappropriate reactions to commensal bacteria as well as an altered bacterial composition during the pathogenesis of IBD (7, 8). For example, decreased microbial diversity such as the accumulation of Enterobacteriaceae (including Escherichia coli) or decreased levels of Bacteroides and Firmicutes are observed in active IBD (7, 9). Specific enteric bacteria were associated with immune regulation directly in an experimental model, such as segmented filamentous bacteria modulates the production of IL-17 by Th17 cells and Clostridium species affect the induction of regulatory T cells (Tregs) (10, 11). These changes in microbiota are thought to be important for the induction or progression of intestinal inflammation (12, 13). However, it is unknown how the host regulates microbiotic homeostasis.

Dendritic cells (DCs) are APCs that play crucial roles in establishing immune reactions to pathogens and tolerance to commensal bacteria (14). Intestinal DCs have diverse functions against luminal bacteria by patrolling the luminal contents directly via dendrites and responding via the intestinal epithelium or other immune cells indirectly (14, 15). Bacteria-stimulated intestinal DCs migrate to lymph nodes and present Ags to immune cells such as T cells to induce the immune response against enteric bacteria. The dysfunction of DCs causes an imbalance in gut immunity and induces intestinal inflammation (16). In addition, previous studies revealed that disrupted TGF-β signaling in DCs resulted in multiorgan autoimmunity (17), atherosclerosis with T cell activation (18), and the spontaneous maturation of cutaneous DCs with migratory potential (19).

In gut immunity, mice with DC-specific deletion of TGF-β signaling (CD11c-cre Tgfbr2fl/fl mice) were reported to show spontaneous colitis characterized by loss of goblet cells with lymphocytic infiltration. This colitis was described as one of the systemic autoimmune phenotypes due to altered Treg differentiation, activated T and B cells, and the increased secretion of inflammatory cytokines such as TNF-α and IFN-γ (17). However, the communication between TGF-β signal depleted DCs and enteric microbial community is largely unknown.

In the current study, we investigated the involvement of TGF-β signaling in DCs in the pathogenesis of IBD, with a particular focus on colonic homeostasis. Data revealed that TGF-β signaling in DCs inhibits colonic inflammation by regulating commensal Enterobacteriaceae and Notch signaling between DCs and the intestinal epithelium.

Materials and Methods

Mice

Tgfbr2 flox/flox C57BL/6 mice (20) were crossed with CD11c-cre/+ mice (B6N.Cg-Tg (Itgax-cre) 1-1Reiz/J) (21), and Cre-negative Tgfbr2 flox/+, Cre-negative Tgfbr2 flox/flox, CD11c-cre/+ Tgfbr2 flox/+ (CD11c-cre Tgfbr2fl/+), and CD11c-cre/+ Tgfbr2 flox/flox (CD11c-cre Tgfbr2fl/fl) mice were generated. Cre-negative littermates were used as controls. All mice were maintained in a conventional animal facility at the University of Tokyo, Tokyo, Japan. All experimental protocols were approved by the Ethics Committee for Animal Experimentation and conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals of the Department of Medicine, University of Tokyo. Mice were born at the facility and housed in autoclaved open cages with autoclaved materials. The mice were fed with a γ-irradiated basal diet (CE-2; CLEA, Tokyo, Japan) and bottled water. Sex-matched littermates had been cohoused in the same cages before they were sacrificed.

Dextran sulfate sodium colitis model

Sex-matched 8- to 14-wk-old Cre-negative control and CD11c-cre Tgfbr2fl/+ mice received 2.0% dextran sulfate sodium (DSS; m.w., 36–40 kDa; ICN Biomedicals, Irvine, CA) in drinking water for 5 d, followed by 4 d with regular water. The weight of each mouse was recorded daily. The mice were sacrificed 10 d after the initiation of DSS treatment, and tissues were prepared for histological analysis as described previously (22). Feces was collected from each mouse for culture and 16S rRNA gene analysis. Effects of antibiotics were examined by adding 1.0 g/l neomycin sulfate (Sigma-Aldrich, St. Louis, MO) and 1.0 g/l metronidazole (Sigma-Aldrich). Effects of inhibiting Notch signaling were examined by injecting the γ-secretase inhibitor dibenzazepine (DBZ) (Tocris BioScience) i.p. at a dose of 5 mg/kg body weight or 10% DMSO where indicated.

Bone marrow chimeric mice

Ten-week-old male C57BL/6 wild-type (WT) recipients (CLEA) were irradiated with a single dose of 9 Gy x-rays and injected 1 × 107 bone marrow cells from CD11c-cre Tgfbr2fl/fl mice or Cre-negative littermates i.v. After transplantation, 1.0 g/l neomycin sulfate and 10 mg/l polymyxin B sulfate (Sigma-Aldrich) were administered in drinking water for 2 wk. Mice were sacrificed at the indicated times after transplantation, and their tissues and feces were prepared. Reconstitution efficiency was evaluated by TGFBR2 expression in magnetically sorted CD11c+ cells from spleen and lamina propria (LP) of both recipient groups using flow cytometry.

Microbial transplantation experiments

Donor feces from Cre-negative control or CD11c-cre Tgfbr2fl/fl mice was collected and stored in 20% glycerol at −80°C. Each stock was titered using blood agar cultures at 5 × 108 CFU/recipient before transplantation. Male WT recipient mice (8–12 wk old) received drinking water containing antibiotics (1.0 g/l neomycin sulfate and 1.0 g/l metronidazole) for 3 wk. At 3 d after the cessation of antibiotics, recipients were gavaged orogastrically with frozen fecal stocks from controls or CD11c-cre Tgfbr2fl/fl mice and received DSS treatment 3 d after transplantation. The recipients were sacrificed 13 d after transplantation.

In another experiment, antibiotic-pretreated male WT recipients were gavaged orogastrically with 5 × 108 CFU/recipient of bacterial cultures from the feces of CD11c-cre Tgfbr2fl/fl mice grown on MacConkey or blood agar.

Histological analysis and immunohistochemistry

For histological analysis, formalin-fixed and paraffin-embedded colon tissues were stained with H&E, and histological score was evaluated as described previously (22).

For immunohistochemistry, 5-μm sections of frozen colon tissues prepared in Tissue-Tek solutions were fixed with acetone, blocked with 2% BSA in PBS and avidin biotin (Avidin Biotin blocking kit; DAKO, Glostrup, Denmark) if needed. They were incubated with the following Abs: biotinylated hamster anti-mouse CD11c (BD Biosciences, San Jose, CA), rat anti-mouse Hes1 (MBL, Nagoya, Japan), and rabbit anti-Jagged1 and Jagged2 (Novus Biologicals, Littleton, CO). For staining for Hes1, formalin-fixed and paraffin-embedded colon tissues were deparaffinized, rehydrated, pretreated with Ag-retrieval solution (S1700; DakoCytomation), and then incubated with the primary Ab.

To analyze goblet cells and mucus layer, colon tissues were fixed in Carnoy’s fixative and embedded in paraffin, as described previously (23). Sections were stained with rhodamine-labeled UEA-1 (Vector Laboratories, Burlingame, CA). Goblet cell numbers and the mucus thickness in UEA-1–stained colon sections were determined in three randomly selected locations from three mice from each group.

Bacterial culture, titration, and isolation

Fecal bacteria were cultured and titered on blood agar in an anaerobic chamber containing 5% CO2 at 37°C for 3 d or MacConkey agar under aerobic conditions at 37°C for 24 h. The E. coli used in microbial transplantation experiments were isolated from feces of CD11c-cre Tgfbr2fl/fl mice on MacConkey agar and identified using 16S rRNA gene sequence analysis.

Fecal DNA extraction and 16S rRNA gene pyrosequencing analysis

Fecal DNA was extracted using QIAamp DNA Stool Mini Kit (Qiagen, Valencia, CA) for quantitative PCR (qPCR) analysis. Amplicons of the V4 region of bacterial 16S rRNA genes were generated using PCR primer set (515F-806R) and sequenced on an Ion PGM system (Thermo Fisher). Taxonomic assignments of these sequence data were performed by BLAST search using the National Center for Biotechnology Information 16S rRNA database. BLAST results and principal component analysis (PCA) were analyzed by Metagenome@KIN software (World Fusion, Tokyo, Japan).

Isolation and treatment of DCs and bone marrow–derived DCs

Leukocytes were isolated from the spleen, mesenteric lymph node (MLN), and colon LP, as described previously (16). To isolate CD11c+ DCs from leukocytes, CD11c+ cells were enriched by positive selection using MACS Microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). In our experiments, CD11c+ cells were positively selected to >96% purity using CD11c MACS Microbeads. Bone marrow–derived DCs (BMDCs) were prepared from Cre-negative controls, CD11c-cre Tgfbr2fl/+ mice, and CD11c-cre Tgfbr2fl/fl mice as described previously (22). After incubation with GM-CSF (10 ng/ml) in RPMI 1640 medium containing 10% FBS for 5–8 d, BMDCs were harvested and analyzed using qPCR and Western blotting. To analyze the effect of inhibiting TGF-β signaling, BMDCs derived from Cre-negative Tgfbr2fl/fl mice were infected with adenovirus expressing Cre or LacZ. Short hairpin RNA (shRNA)–mediated knockdown of Tgfbr2 in BMDCs derived from WT B6 mice was performed by lentiviral approach using Tgfbr2 shRNA (Thermo Scientific).

Flow cytometry

Isolated cells were prepared, and the surface markers were detected using anti-CD11c (BD Biosciences), anti-CD45, I-Ab, CD11b, CD11c, CD103, CD4, F4/80, Jagged1, Jagged2 (BioLegend, San Diego, CA), and TGFBR2 (R&D Systems, Minneapolis, MN) Abs with appropriate control Abs. Live cells were identified using 7-aminoactinomycin D (BioLegend). Data were acquired on Guava easyCyte cytometer (Merck Millipore, Darmstadt, Germany) or FACSAria III (BD Biosciences) and analyzed using GuavaSuite software (Merck Millipore) or FlowJo software (Tree Star, Ashland, OR).

RT-PCR, qPCR, and PCR arrays

Total RNA was extracted from colon tissues, DCs, and BMDCs using an RNeasy Mini Kit (Qiagen). The first-strand cDNA was synthesized using an ImProm-II Reverse Transcription System (Promega, Madison, WI), and PCR amplification was performed using a StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA) and normalized to GAPDH. For PCR arrays (mouse signal transduction pathway finder PCR array; Qiagen), RNA samples from five mice were pooled and analyzed.

Statistical analysis

Data were analyzed statistically using Student t tests, Mann–Whitney U tests, or one-way ANOVAs with Bonferroni post hoc tests; p < 0.05 was considered to indicate statistical significance.

Results

TGF-β signaling in DCs protects against colitis

To investigate the involvement of TGF-β signaling in DCs in colonic inflammation, we used CD11c-cre Tgfbr2fl/fl mice. First, we confirmed whether CD11c-driven cre recombinase specifically abrogates TGF-β signaling in DCs. The expression of TGFBR2 and the phosphorylation of Smad2 in CD11c+ cells were attenuated in CD11c-cre Tgfbr2fl/+ mice and CD11c-cre Tgfbr2fl/fl mice compared with controls (Supplemental Fig. 1A, 1B, 1E), whereas TGFBR2 expression on CD4+ T cells and F4/80+ macrophages and phosphorylation of Smad2 in CD4+ T cells were unaffected (Supplemental Fig. 1C, 1D). We also examined TGFBR2 expression on each intestinal DC subset (CD103+CD11b+, CD103+CD11b−, and CD103−CD11b+ cells) of control and CD11c-cre Tgfbr2fl/fl mice. TGFBR2 was similarly expressed by each DC subset of control mice (Supplemental Fig. 1F).

CD11c-cre Tgfbr2fl/+ and control mice appeared normal phenotypically. In contrast, similar to a previous report (17), CD11c-cre Tgfbr2fl/fl mice showed signs of sickness and wasting and died aged 5–14 wk (Fig. 1A). CD11c-cre Tgfbr2fl/fl mice exhibited a shorter colon than did controls (Fig. 1B, 1C). Histological injury of the colon, cellular infiltration, and goblet cell depletion were more severe in CD11c-cre Tgfbr2fl/fl mice than in CD11c-cre Tgfbr2fl/+ and control mice (Fig. 1D, 1E). These results suggest that TGF-β signaling in DCs is important for preventing colonic inflammation.

FIGURE 1.
FIGURE 1.

TGF-β signaling in DCs protects against colitis. (A) Survival curves (n = 12/group). (B) Colon lengths (8–13-wk-old mice), n = 5. (C) Representative images of the ceca and colons (13-wk-old mice). (D) Representative histology of colon sections. Scale bars, 100 μm. (E) Histological scores of colon sections (n = 10). (A)–(E) all show data from littermate control, CD11c-cre Tgfbr2fl/+, and CD11c-cre Tgfbr2fl/fl mice. *p < 0.05 versus littermate control mice, **p < 0.005 versus littermate control mice by one-way ANOVA; data are presented as mean ± SE. (F) Body weight curves (n = 9–12 for each data point). (G) Colon lengths (n = 3–9 for each data point). (H) Representative histology of colon sections. Scale bars, 200 μm. (I) Histological scores of colon sections (n = 3–9 for each data point). Data in (F)–(I) are from littermate control and CD11c-cre Tgfbr2fl/+ mice with DSS treatment (8–12 wk old) at the indicated times. See also Supplemental Figs. 1 and 4. *p < 0.05 versus littermate control mice by Student t test; data are presented as mean ± SE.

To further investigate the role of TGF-β signaling in DCs in colonic homeostasis, control and CD11c-cre Tgfbr2fl/+ mice were treated with 2.0% DSS. Although unchallenged CD11c-cre Tgfbr2fl/+ mice were normal (Fig. 1C, 1D), CD11c-cre Tgfbr2fl/+ mice treated with DSS exhibited severe loss of body weight (Fig. 1F), a shorter colon (Fig. 1G), histological injury such as severe ulceration (Fig. 1H), and a higher histological score (Fig. 1I) compared with controls. These results suggest that TGF-β signaling in DCs is also protective against DSS-induced colitis.

Deficient TGF-β signaling in DCs is associated with altered colonic microbiota and the subsequent induction of colitis

To understand how TGF-β signaling in DCs affects colonic homeostasis, we compared the number of bacteria in the feces of control and CD11c-cre Tgfbr2fl/fl mice using culture analysis. Enterobacteriaceae in feces, as counted from MacConkey agar cultures, were enriched ∼100-fold in CD11c-cre Tgfbr2fl/fl mice compared with controls, although there was no significant difference in the total number of bacteria between controls and CD11c-cre Tgfbr2fl/fl mice on blood agar cultures (Fig. 2A). Quantitative PCR (qPCR) analysis of fecal DNA using specific primer for bacterial family also showed that Enterobacteriaceae were enriched in CD11c-cre Tgfbr2fl/fl mice, compared with controls, although there was no significant difference in the number of total and other families of bacteria between control and CD11c-cre Tgfbr2fl/fl mice (Supplemental Fig. 2A). Furthermore, to investigate their fecal microbiota, we performed pyrosequencing analysis of V4 region amplicons of 16S rRNA genes. PCA of taxonomically classified 16S rRNA transcripts showed that marked separation of fecal microbial composition between controls and CD11c-cre Tgfbr2fl/fl mice (Fig. 2B). In bacterial family-level analysis of 16S rRNA pyrosequencing, Enterobacteriaceae were significantly more abundant in CD11c-cre Tgfbr2fl/fl mice compared with controls (Fig. 2C). Among 208 families evaluated, the change in Enterobacteriaceae was most significant (Fig. 2C). These results suggest that Enterobacteriaceae enrichment is the major factor in microbial changes in CD11c-cre Tgfbr2fl/fl mice. To identify the specific Enterobacteriaceae bacterial species in CD11c-cre Tgfbr2fl/fl feces, 52 colonies were isolated, and culture-dependent analysis and 16S rRNA sequencing were performed. Both analyses revealed that 50 colonies (96%) were E. coli (data not shown).

FIGURE 2.
FIGURE 2.

Microbial analysis of the feces from CD11c-cre Tgfbr2fl/fl mice and CD11c-cre Tgfbr2fl/+ mice. (A) The numbers of bacteria grown on blood agar and MacConkey agar from the feces of littermate control and CD11c-cre Tgfbr2fl/fl mice (8–13 wk old); n = 5/group. Data are presented as mean ± SE. *p < 0.05 versus littermate control mice by Mann–Whitney U test. (B) PCA of 16S rRNA sequencing libraries from the feces of littermate control and CD11c-cre Tgfbr2fl/fl mice (8 wk old); n = 4/group. (C) Comparison of family-level composition in littermate control and CD11c-cre Tgfbr2fl/fl fecal microbial communities analyzed by 16S rRNA sequencing (n = 4/group). Average relative abundance of bacterial families (left panel) and Log10 fold changes of each family in the feces of CD11c-cre Tgfbr2fl/fl mice divided by that of littermate controls (right panel) are shown. The families <1% relative abundance were included in Other fraction. Error bars indicate SE. (D) Body weight (BW) curves (n = 6/group). (E) Representative histology of colon sections (12-wk-old mice). Scale bars, 100 μm. (F) Colon lengths (n = 6/group). (G) Bacterial number grown on MacConkey agar from the feces (n = 4/group). Data in (D)–(G) are from littermate control and CD11c-cre Tgfbr2fl/+ mice with DSS plus antibiotics treatment (8–12 wk old) at the indicated times. Data are presented as mean ± SE. See also Supplemental Fig. 2. *p < 0.05 versus littermate control mice by Mann–Whitney U test.

To investigate whether intestinal bacteria are associated with the severity of DSS-induced colitis in CD11c-cre Tgfbr2fl/+ mice, the mice were treated both with broad-spectrum antibiotics and DSS. CD11c-cre Tgfbr2fl/+ and control mice that received antibiotics together with DSS exhibited no significant loss of body weight (Fig. 2D) or severe inflammation (Fig. 2E, 2F), suggesting that antibiotic treatment prevented colitis to a similar extent in both CD11c-cre Tgfbr2fl/+ and control mice. These results suggest that TGF-β signaling in DCs is associated with bacteria-mediated colitis. In DSS-treated mice (Fig. 1), Enterobacteriaceae species were enriched 10-fold in CD11c-cre Tgfbr2fl/+ mice compared with control (Fig. 2G). In contrast, the abundance of Enterobacteriaceae did not significantly differ between control and CD11c-cre Tgfbr2fl/+ mice not treated with DSS or treated with DSS and antibiotics. These results suggest that TGF-β signaling in DCs might be associated with intestinal homeostasis by regulating the density of commensal Enterobacteriaceae.

To investigate the contribution of microbial alteration (dysbiosis) on the pathogenesis of colitis in CD11c-cre Tgfbr2fl/fl mice, fecal microbial transplantation experiments were performed (Fig. 3A). Antibiotic-pretreated WT mice gavaged with feces from CD11c-cre Tgfbr2fl/fl mice exhibited more severe loss of body weight (Fig. 3B) and histological injury than did mice gavaged with feces from control mice upon DSS exposure (Fig. 3C, 3D). Next, the contribution of the specific bacteria within the feces of CD11c-cre Tgfbr2fl/fl mice on the regulation of colitis was examined. Antibiotic-pretreated WT mice were gavaged with 5 × 108 CFU total bacteria, Enterobacteriaceae, and a single colony of E. coli cultured from the feces of CD11c-cre Tgfbr2fl/fl mice. Mice gavaged with Enterobacteriaceae or E. coli were more susceptible to DSS-induced colitis than were mice gavaged with total bacteria (Fig. 3E–G). These results suggest that enteric bacteria in CD11c-cre Tgfbr2fl/fl mice, particularly Enterobacteriaceae and E. coli, are associated with the development of colitis.

FIGURE 3.
FIGURE 3.

Enterobacteriaceae enrichment in CD11c-cre Tgfbr2fl/fl feces is associated with the development of colitis. (A) Experimental timeline of bacterial transplantation and DSS treatment. (B) Body weight (BW) curves. Data are presented as mean ± SE. *p < 0.05 versus littermate control feces group by Student t test. (C) Representative histology of colon sections (12-wk-old mice). Scale bars, 200 μm. (D) Histological scores of colon sections. Data are presented as mean ± SE. *p < 0.05 versus littermate control feces group by Student t test. The data in (B)–(D) were obtained from mice that received 5 × 108 CFU feces from littermate control (n = 7–9) and CD11c-cre Tgfbr2fl/fl mice (n = 5–9). (E–G) Mice gavaged with 5 × 108 CFU total bacteria (n = 4–9), Enterobacteriaceae (n = 5–9), or E. coli (n = 4–9) cultured from the feces of CD11c-cre Tgfbr2fl/fl mice. (E) Body weight (BW) curves. Data are presented as mean ± SE. *p < 0.05 versus the total bacteria group by Student t test. (F) Representative histology of colon sections. Scale bars, 200 μm. (G) Histological scores of the colon sections. Data are presented as mean ± SE. *p < 0.05 versus the total bacteria group by one-way ANOVA.

Goblet cell and mucin depletion is associated with dysbiosis

Goblet cell depletion is a histological hallmark of human UC (24, 25). Because goblet cell depletion was observed in the histology of CD11c-cre Tgfbr2fl/fl mice (Fig. 1D), we next investigated the contribution of goblet cell depletion to the pathogenesis of colitis. Staining for UEA-1 in the colon tissues revealed that the number of goblet cells was depleted severely, and the mucus layer was thinner in CD11c-cre Tgfbr2fl/fl mice compared with controls (Fig. 4A–C). To assess the effect of goblet cell depletion, goblet cell–specific gene expression in the colon was analyzed using qPCR. Muc2, TFF3, and Relmβ were downregulated significantly in CD11c-cre Tgfbr2fl/fl mice compared with controls, as expected (Fig. 4D). Furthermore, immunostaining for E. coli showed that significantly more E. coli were directly accumulated on colonic epithelium in CD11c-cre Tgfbr2fl/fl mice, whereas fewer E. coli were localized in the thick mucus layer distant from the epithelium in control mice (Fig. 4E). These results suggest that TGF-β signaling in DCs is important for the homeostasis of goblet cells, the mucus layer, and enteric microbiota.

FIGURE 4.
FIGURE 4.

Goblet cell and mucus depletion is associated with dysbiosis and colitis. (A) Representative images of UEA-1 staining (red) in Carnoy-fixed colon sections from littermate control and CD11c-cre Tgfbr2fl/fl mice (13-wk-old mice). Hoechst (blue) was used as a counter stain. Scale bars, 100 μm. (B and C) Quantification of goblet cells and mucus thickness in UEA-1–stained colon sections from littermate control and CD11c-cre Tgfbr2fl/fl mice (8–13-wk-old mice). Data are presented as mean ± SE. *p < 0.05 versus littermate control mice by Student t test. (D) Muc2, TFF3, and Relmβ expression in colon tissues from littermate control and CD11c-cre Tgfbr2fl/fl mice was analyzed using qPCR (n = 5/group) (8–13-wk-old mice). *p < 0.05 versus littermate control mice by Student t test. (E) Representative images of E. coli (green) and UEA-1 (red) staining in frozen sections of colons from littermate control and CD11c-cre Tgfbr2fl/fl mice (13-wk-old mice). Scale bars, 50 μm. (F–J) Ten-week-old WT mice were treated with DSS and DBZ or DMSO. (F) Body weight (BW) curves (n = 7 to 8 for each data point). Data are presented as mean ± SE. *p < 0.05 versus DMSO + DSS group by Student t test. (G) Representative histology of colon sections. Scale bars, 200 μm. (H) Representative images of UEA-1 (red) staining in Carnoy-fixed colon sections. Scale bars, 50 μm. (I) Quantification of the mucus thickness in UEA-1–stained colon sections. Data are presented as mean ± SE. *p < 0.05 versus DMSO group by Student t test. (J) The number of fecal bacteria grown on MacConkey agar from mice treated with DBZ or DMSO with or without DSS for 10 d (n = 7/group). See also Supplemental Fig. 3. *p < 0.05 versus DMSO group by Mann–Whitney U test.

To investigate the contribution of goblet cells and the colon mucus layer to the regulation of enteric bacteria and intestinal inflammation, we used the γ-secretase inhibitor DBZ to induce goblet cell differentiation (26). WT mice treated with DBZ plus DSS exhibited significantly attenuated loss of body weight (Fig. 4F) and histological injury (Fig. 4G) compared with those treated with DMSO and DSS. Without DSS treatment, the thickness of the mucus layer was increased in WT mice treated with DBZ compared with those that received DMSO (Fig. 4H, 4I). After DSS treatment, the thickness of the mucus layer was greater in WT mice treated with DBZ than in those treated with DMSO (Fig. 4H, 4I). After DSS treatment, the number of Enterobacteriaceae was significantly lower in the feces of WT mice that received DBZ compared with that received DMSO (Fig. 4J). These results suggest that the mucus layer composed of goblet cells is important for controlling the number of enteric bacteria, which might protect against colitis.

To determine whether inhibiting TGF-β signaling in DCs causes goblet cell depletion and alters the microbial population, especially in the healthy adult mice with normal epithelium and microbiota, we generated bone marrow chimeric mice by reconstituting irradiated WT mice with bone marrow cells from Cre-negative control or CD11c-cre Tgfbr2fl/fl mice. WT mice that received bone marrow cells from control and CD11c-cre Tgfbr2fl/fl mice had a similar number of goblet cells in their colon and fecal Enterobacteriaceae at the time of transplantation. Recipients that received CD11c-cre Tgfbr2fl/fl bone marrow died 1–4 wk after transplantation, whereas 9 of 10 recipients that received Cre-negative bone marrow survived >4 wk (Fig. 5A). We confirmed the efficient reconstitution following 2 wk after transplantation by revealing attenuated TGFBR2 expression on CD11c+ DCs from the spleen and LP of recipients received CD11c-cre Tgfbr2fl/fl bone marrow compared with that received Cre-negative bone marrow (data not shown). Recipients of CD11c-cre Tgfbr2fl/fl bone marrow showed severe histological injury, goblet cell depletion, a thinner mucus layer (Fig. 5B–D), and a higher number of Enterobacteriaceae in their feces after transplantation (Fig. 5E). We found goblet cell–specific gene expressions (Muc2, TFF3, and Relmβ) in the colon of recipients transplanted CD11c-cre Tgfbr2fl/fl bone marrow were downregulated at 1 wk after transplantation compared with recipients transplanted Cre-negative bone marrow, although gene expressions associated with cellular infiltration (CD4, Gr-1, and CD11c) were not upregulated at this time point and upregulated later (Fig. 5F). These results suggest that TGF-β signal depleted DCs induce goblet cell depletion, dysbiosis, and intestinal inflammation in healthy intestine.

FIGURE 5.
FIGURE 5.

Epithelial and microbial changes in bone marrow chimeric mice. Ten-week-old WT mice received littermate control and CD11c-cre Tgfbr2fl/fl bone marrow at the indicated weeks. (A) Survival curves (n = 10/group). (B) Representative histology of colon sections. Scale bars, 100 μm. (C) Representative images of UEA-1 staining (red) in Carnoy-fixed colon sections; Hoechst (blue) was used as a counterstain. Scale bars, 100 μm. (D) Quantification of mucus thickness. *p < 0.05 versus littermate control bone marrow transferred mice by Student t test. (E) The number of fecal bacteria grown on MacConkey agar (n = 5–10 for each data point); data are presented as mean ± SE. *p < 0.05, **p < 0.005 by Mann–Whitney U test. (F) Gene expression associated with goblet cells and cellular infiltration in the colon tissues analyzed by qPCR (n = 4–6/group). Data are presented as mean ± SE. See also Supplemental Fig. 3. *p < 0.05 versus littermate control bone marrow–transferred mice by Student t test.

TGF-β signaling in DCs regulates colonic differentiation and mucus secretion via Notch signaling

To identify the gene expression profile of DCs in CD11c-cre Tgfbr2fl/fl mice, we used PCR arrays related to TGF-β and other immunological signaling pathways in LP DCs. The Notch ligand Jagged1 was upregulated 3-fold in LP DCs from CD11c-cre Tgfbr2fl/fl mice compared with those from control mice (Supplemental Fig. 3A). To assess Notch signaling in the intestine, we next analyzed the expression of Notch target genes and Notch ligands in control and CD11c-cre Tgfbr2fl/fl mice. The results of qPCR revealed that the expression of Hes1 in colon tissues and the Notch ligands Jagged1 and Jagged2 in DCs from the colonic LP were increased in CD11c-cre Tgfbr2fl/fl mice compared with controls (Fig. 6A, 6B). Immunostaining and flow cytometric analysis indicated most Jagged1- and Jagged2-positive cells were LP DCs in CD11c-cre Tgfbr2fl/fl mice, in contrast with undetectable Jagged1 and Jagged2 expression on LP leukocytes of control mice (Fig. 6C–F). Immunostaining showed that the number of Hes1-positive cells was increased significantly in the base of the goblet cell–depleted colonic crypts of CD11c-cre Tgfbr2fl/fl mice (Fig. 6G). In the bone marrow chimera experiment (Fig. 5), Jagged1 and Jagged2 expression was increased in the colon tissues of CD11c-cre Tgfbr2fl/fl bone marrow transferred mice 1 wk after bone marrow reconstitution, and Hes1 expression was also elevated at 1 wk and later (Supplemental Fig. 3B). These results suggest that the increased expression of Notch ligands in DCs could activate Notch signaling in the colonic epithelium by the direct contact, which eventually inhibited goblet cell differentiation. Furthermore, to evaluate the effect of Notch signal activation on intestinal inflammation of DC TGF-β signal–deficient mice, we performed DBZ treatment on CD11c-cre Tgfbr2fl/fl mice and DBZ plus DSS treatment on CD11c-cre Tgfbr2fl/+ mice. We found DBZ treatment markedly reduced colonic inflammation with increase of goblet cell number and mucus layer of CD11c-cre Tgfbr2fl/fl mice (Supplemental Fig. 3C). We found DBZ treatment also attenuated DSS-induced colitis in CD11c-cre Tgfbr2fl/+ mice (Supplemental Fig. 3D, 3E). These results suggest that treatment targeting to Notch signaling partially rescue the phenotype of CD11c-cre Tgfbr2fl/fl mice and would have some impact on the potential therapeutics in IBD.

FIGURE 6.
FIGURE 6.

Notch signaling between DCs and the colonic epithelium in CD11c-cre Tgfbr2fl/fl mice. (A) Hes1 transcripts in colon tissues from littermate control and CD11c-cre Tgfbr2fl/fl mice (8–13 wk old) were analyzed using qPCR (n = 5 for each). Data are presented as mean ± SE. *p < 0.05 versus littermate control mice by Student t test. (B) Jagged1 and Jagged2 transcripts in LP CD11c+ cells from littermate control and CD11c-cre Tgfbr2fl/fl mice were analyzed using qPCR (n = 4 for each). *p < 0.05 versus littermate control mice by Student t test. Flow cytometric analysis for Jagged1 and Jagged2 expression on LP leukocytes from littermate control and CD11c-cre Tgfbr2fl/fl mice (8 wk old) (C) or on CD11c-positive/negative cells from LP leukocytes of CD11c-cre Tgfbr2fl/fl mice (8 wk old) (D). (E–G) Representative images of CD11c staining (red) and Jagged1 or Jagged2 staining (green) in colon sections and Hes1 (green) and UEA-1 staining (red) in colon sections from littermate control and CD11c-cre Tgfbr2fl/fl mice (13 wk old). Hoechst (blue) was used as a counterstain. Scale bars, 50 μm. (H) Jagged1 and Jagged2 expression in BMDCs from WT mice stimulated with the indicated concentrations of TGF-β was analyzed using qPCR. *p < 0.05 versus unstimulated control cells by one-way ANOVA. Data are presented as mean ± SE. (I) Jagged1 and Jagged2 expression in BMDCs from Tgfbr2fl/fl mice infected with Cre-expressing or LacZ-expressing adenovirus was analyzed using qPCR. *p < 0.05 versus adeno-LacZ–infected BMDCs by Student t test. (J) Jagged1 and Jagged2 expression in BMDCs from WT mice infected with lentivirus expressing Tgfbr2-specific shRNA or control shRNA was analyzed using qPCR. Experiments were performed in triplicate. *p < 0.05 versus control shRNA-transfected BMDCs by Student t test. (K) Graphical summary of the role of TGF-β signaling in DCs on colonic homeostasis clarified in this study. See also Supplemental Fig. 3.

Finally, we examined the contribution of TGF-β signaling to the expression of Notch ligands in DCs in vitro by treating BMDCs derived from WT mice with rTGF-β1. The expression of Jagged1 and Jagged2 was downregulated in BMDCs after exposure to TGF-β1 (Fig. 6H). Furthermore, the expression of Jagged1 and Jagged2 was upregulated by the in vitro deletion or knockdown of Tgfbr2 (Fig. 6I, 6J). These results suggest that TGF-β signaling plays a critical role in suppressing Jagged1 and Jagged2 expression in DCs.

Discussion

DCs play important roles in regulating the innate immune responses to gut microbes (27). Generally, intestinal CD11c+MHC class II+ DCs can be divided into three major subsets; CD103+CD11b−, CD103+CD11b+, and CD103−CD11b+ cells, and the functions of each subset are intensively analyzed in many studies (15, 28–30). In this study, we used CD11c-cre transgenic mice to examine the role of DCs and identified magnetically sorted CD11c+ cells as DCs, although CD11c is reported to express on other immune cells such as CD8+ T cells in the intestine (31). TGF-β is an important cytokine in the pathogenesis of inflammatory diseases, including IBD (32). In humans, activated TGF-β1 is observed in healthy and IBD colonic tissues, whereas the phosphorylation of Smad3 is attenuated in the LP mononuclear cells of patients with IBD (5). In addition, a recent clinical trial demonstrated that restoration of TGF-β signaling was effective for Crohn disease, suggesting a critical role of this signaling (33). In a mouse model, TGF-β signaling in T cells was investigated intensively, and its critical role in Treg differentiation and autoimmunity was clarified (34, 35). In the current study, we demonstrated that TGF-β signaling in DCs regulates the expression of Notch ligands, goblet cell differentiation, and maintenance of the mucus layer in the colon and prevents the overgrowth of enteric bacteria and the development of colitis.

A previous report using same CD11c-cre Tgfbr2fl/fl mice showed that they developed spontaneous colitis characterized by loss of goblet cells with lymphocytic infiltration and systemic autoimmunity due to altered Treg differentiation, characterized by attenuated Foxp3 expression in Tregs and expansion of CD25− Treg subset (17). We also found systemic inflammation in multiple organs such as the lung, liver, and pancreas in CD11c-cre Tgfbr2fl/fl mice (Supplemental Fig. 4A). However, in our experiment, there were no significant difference of CD4+Foxp3+ Treg cell number in the spleen and MLN between control and CD11c-cre Tgfbr2fl/fl mice (Supplemental Fig. 4B). This result is quite different from the previous study showing the increased proportion of CD4+Foxp3+ Tregs in spleen and MLN of CD11c-cre Tgfbr2fl/fl mice (17). To evaluate the effect of Tregs on CD11c-cre Tgfbr2fl/fl mice, we performed in vivo depletion of CD4+ cells (containing CD4+CD25− Treg subset) to CD11c-cre Tgfbr2fl/fl mice. CD4-Ab–treated CD11c-cre Tgfbr2fl/fl mice did not show improvement in intestinal inflammation (Supplemental Fig. 4C), goblet cell depletion (Supplemental Fig. 4D), and Enterobacteriaceae overgrowth (Supplemental Fig. 4E) compared with isotype-Ab-treated CD11c-cre Tgfbr2fl/fl mice. Thus, the proportion and effect of Tregs may vary with experimental conditions, although inflammatory phenotype of CD11c-cre Tgfbr2fl/fl mice is consistent. Therefore, we focused on other factors involved in IBD pathogenesis.

It has been postulated that inappropriate reactions to enteric bacteria and changes in bacterial composition are involved in the pathogenesis of IBD (2). We found that only Enterobacteriaceae were significantly enriched among the bacterial families examined in the feces of CD11c-cre Tgfbr2fl/fl mice and DSS-treated CD11c-cre Tgfbr2fl/+ mice (Fig. 2). We also searched for specific bacteria that modulate host immune systems, such as segmented filamentous bacteria and Clostridia, but observed no differences in these bacteria in our experimental models (Supplemental Fig. 2B). The overgrowth of Enterobacteriaceae has been reported in patients with active IBD (7, 9) and in multiple experimental mouse models of IBD (8, 36). Several studies reported that inflammation-induced changes in the microbial composition such as reduced bacterial diversity or Enterobacteriaceae enrichment (7, 37). Enterobacteriaceae are a family of Gram-negative enteric bacteria that includes both pathogenic and commensal species (37, 38). Some species belonging to Enterobacteriaceae, such as Klebsiella pneumonia, Proteus mirabilis, and specific strains of E. coli, caused colitis in other IBD mouse models (8, 36). In the current study, 96% of Enterobacteriaceae colonies grown on MacConkey agar from the feces of colitogenic CD11c-cre Tgfbr2fl/fl mice were nonpathogenic E. coli (data not shown). Thus, the microbes responsible for dysbiosis and colitis might vary depending on host genetic deficiencies and environmental factors, although we did not fully unravel the mechanisms why E. coli increased selectively in CD11c-cre Tgfbr2fl/fl mice, and we cannot exclude the possibility that secondary effects of inflammation such as increased inducible NO synthase expression caused the overgrowth of E. coli (39, 40).

In the current study, we observed goblet cell depletion in colitic mice. Goblet cell depletion and a thinner mucus layer are characteristics of the colonic pathology of UC (41), although it is unknown clinically whether goblet cell depletion is a marker or pathogenic cause of UC. Intestinal goblet cells secrete the mucins that form the mucus layer (42, 43), which serves as a barrier that separates enteric bacteria from the epithelium (42). The thinner mucus layer observed in active UC makes enteric bacteria easy to locate and allows them to grow close to intestinal epithelium and invade the submucosal area (42), suggesting that it plays a causal role in the disease. Our results showed a close association between goblet cell depletion and the overgrowth of Enterobacteriaceae. Because forced goblet cell differentiation and mucus secretion by DBZ reduced the Enterobacteriaceae load (Fig. 4), goblet cell depletion might cause dysbiosis. Mucin- deficient mice (Muc2−/− mice) were also reported to show spontaneous colitis with depletion of goblet cell and mucus layer and dysbiosis characterized by decreased Bacteriodetes and increased Proteobacteria, which were supportive data of our findings (44), although further study is needed to clarify the mechanisms between altered mucus layer and dysbiosis.

Notch signaling plays a critical role in inhibiting the differentiation of the goblet cell lineage (26). Notch signaling is transferred by the direct contact between a cell expressing Notch ligands and a cell expressing Notch receptors (45). Some studies have reported Notch-associated crosstalk between immune cells and the epithelium (46, 47); however, Notch signaling between DCs and the epithelium has not been demonstrated. The expression of Notch ligands was increased in the intestines of human IBD, and activated Notch signaling was associated with downregulated goblet cell differentiation and mucin formation (48). In the current study, the expression of Jagged1 and Jagged2 was increased particularly in colonic LP DCs, and the expression of Hes1 was enhanced in the colonic epithelium of CD11c-cre Tgfbr2fl/fl mice. In contrast, control mice exhibited Smad phosphorylation and low Jagged expression in LP DCs (Fig. 6E, 6F, Supplemental Fig. 1E). Although we could not demonstrate direct interaction between DCs and epithelium, these findings suggested that steady-state TGF-β signaling in DCs suppresses the expression of Notch ligands, thereby inhibiting Notch activation in colonic epithelial cells and promoting goblet cell differentiation.

In summary, the current study revealed that the abrogation of TGF-β signaling in DCs leads to colonic inflammation. This is associated with the overexpression of Notch ligands in DCs, goblet cell depletion in the epithelium, decreased colonic mucus, and increased titer of Enterobacteriaceae (Fig. 6K). These results clarify the essential role of TGF-β signaling in DCs on the maintenance of colonic homeostasis among host immune cells, epithelial cells, and enteric microbiota.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Dr. A. Enomoto and Dr. K. Miyagawa (The University of Tokyo) for technical assistance with the radiation treatment. We also thank Y. Ishii (Institute of Medical Science–The University of Tokyo FACS Core Laboratory) for assistance with flow cytometry.

Footnotes

  • This work was supported by Japan Society for the Promotion of Science KAKENHI Grant 23590933, a research grant from Yakult Bio-Science Foundation, and the Japan Society for the Promotion of Science Core-to-Core Program Cooperative International Framework in TGF-β Family Signaling.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    BMDC
    bone marrow–derived dendritic cell
    DBZ
    dibenzazepine
    DC
    dendritic cell
    DSS
    dextran sulfate sodium
    IBD
    inflammatory bowel disease
    LP
    lamina propria
    MLN
    mesenteric lymph node
    PCA
    principal component analysis
    qPCR
    quantitative PCR
    shRNA
    short hairpin RNA
    Treg
    regulatory T cell
    UC
    ulcerative colitis
    WT
    wild-type.

  • Received December 7, 2015.
  • Accepted March 21, 2016.

References

  1. ↵
    1. Abraham C.,
    2. J. H. Cho
    . 2009. Inflammatory bowel disease. N. Engl. J. Med. 361: 2066–2078.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Xavier R. J.,
    2. D. K. Podolsky
    . 2007. Unravelling the pathogenesis of inflammatory bowel disease. Nature 448: 427–434.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Jostins L.,
    2. S. Ripke,
    3. R. K. Weersma,
    4. R. H. Duerr,
    5. D. P. McGovern,
    6. K. Y. Hui,
    7. J. C. Lee,
    8. L. P. Schumm,
    9. Y. Sharma,
    10. C. A. Anderson,
    11. et al,
    12. International IBD Genetics Consortium (IIBDGC)
    . 2012. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature 491: 119–124.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Lees C. W.,
    2. J. C. Barrett,
    3. M. Parkes,
    4. J. Satsangi
    . 2011. New IBD genetics: common pathways with other diseases. Gut 60: 1739–1753.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Di Sabatino A.,
    2. K. M. Pickard,
    3. D. Rampton,
    4. L. Kruidenier,
    5. L. Rovedatti,
    6. N. A. Leakey,
    7. G. R. Corazza,
    8. G. Monteleone,
    9. T. T. MacDonald
    . 2008. Blockade of transforming growth factor beta upregulates T-box transcription factor T-bet, and increases T helper cell type 1 cytokine and matrix metalloproteinase-3 production in the human gut mucosa. Gut 57: 605–612.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Monteleone G.,
    2. A. Kumberova,
    3. N. M. Croft,
    4. C. McKenzie,
    5. H. W. Steer,
    6. T. T. MacDonald
    . 2001. Blocking Smad7 restores TGF-beta1 signaling in chronic inflammatory bowel disease. J. Clin. Invest. 108: 601–609.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Sartor R. B.
    2008. Microbial influences in inflammatory bowel diseases. Gastroenterology 134: 577–594.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Garrett W. S.,
    2. C. A. Gallini,
    3. T. Yatsunenko,
    4. M. Michaud,
    5. A. DuBois,
    6. M. L. Delaney,
    7. S. Punit,
    8. M. Karlsson,
    9. L. Bry,
    10. J. N. Glickman,
    11. et al
    . 2010. Enterobacteriaceae act in concert with the gut microbiota to induce spontaneous and maternally transmitted colitis. Cell Host Microbe 8: 292–300.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Qin J.,
    2. R. Li,
    3. J. Raes,
    4. M. Arumugam,
    5. K. S. Burgdorf,
    6. C. Manichanh,
    7. T. Nielsen,
    8. N. Pons,
    9. F. Levenez,
    10. T. Yamada,
    11. et al,
    12. MetaHIT Consortium
    . 2010. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 464: 59–65.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Goto Y.,
    2. C. Panea,
    3. G. Nakato,
    4. A. Cebula,
    5. C. Lee,
    6. M. G. Diez,
    7. T. M. Laufer,
    8. L. Ignatowicz,
    9. I. I. Ivanov
    . 2014. Segmented filamentous bacteria antigens presented by intestinal dendritic cells drive mucosal Th17 cell differentiation. Immunity 40: 594–607.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Atarashi K.,
    2. T. Tanoue,
    3. T. Shima,
    4. A. Imaoka,
    5. T. Kuwahara,
    6. Y. Momose,
    7. G. Cheng,
    8. S. Yamasaki,
    9. T. Saito,
    10. Y. Ohba,
    11. et al
    . 2011. Induction of colonic regulatory T cells by indigenous Clostridium species. Science 331: 337–341.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Heimesaat M. M.,
    2. S. Bereswill,
    3. A. Fischer,
    4. D. Fuchs,
    5. D. Struck,
    6. J. Niebergall,
    7. H. K. Jahn,
    8. I. R. Dunay,
    9. A. Moter,
    10. D. M. Gescher,
    11. et al
    . 2006. Gram-negative bacteria aggravate murine small intestinal Th1-type immunopathology following oral infection with Toxoplasma gondii. J. Immunol. 177: 8785–8795.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Darfeuille-Michaud A.,
    2. J. Boudeau,
    3. P. Bulois,
    4. C. Neut,
    5. A. L. Glasser,
    6. N. Barnich,
    7. M. A. Bringer,
    8. A. Swidsinski,
    9. L. Beaugerie,
    10. J. F. Colombel
    . 2004. High prevalence of adherent-invasive Escherichia coli associated with ileal mucosa in Crohn’s disease. Gastroenterology 127: 412–421.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Cader M. Z.,
    2. A. Kaser
    . 2013. Recent advances in inflammatory bowel disease: mucosal immune cells in intestinal inflammation. Gut 62: 1653–1664.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Farache J.,
    2. I. Koren,
    3. I. Milo,
    4. I. Gurevich,
    5. K. W. Kim,
    6. E. Zigmond,
    7. G. C. Furtado,
    8. S. A. Lira,
    9. G. Shakhar
    . 2013. Luminal bacteria recruit CD103+ dendritic cells into the intestinal epithelium to sample bacterial antigens for presentation. Immunity 38: 581–595.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Hirata Y.,
    2. L. Egea,
    3. S. M. Dann,
    4. L. Eckmann,
    5. M. F. Kagnoff
    . 2010. GM-CSF-facilitated dendritic cell recruitment and survival govern the intestinal mucosal response to a mouse enteric bacterial pathogen. Cell Host Microbe 7: 151–163.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Ramalingam R.,
    2. C. B. Larmonier,
    3. R. D. Thurston,
    4. M. T. Midura-Kiela,
    5. S. G. Zheng,
    6. F. K. Ghishan,
    7. P. R. Kiela
    . 2012. Dendritic cell-specific disruption of TGF-β receptor II leads to altered regulatory T cell phenotype and spontaneous multiorgan autoimmunity. J. Immunol. 189: 3878–3893.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Lievens D.,
    2. K. L. Habets,
    3. A. K. Robertson,
    4. Y. Laouar,
    5. H. Winkels,
    6. T. Rademakers,
    7. L. Beckers,
    8. E. Wijnands,
    9. L. Boon,
    10. M. Mosaheb,
    11. et al
    . 2013. Abrogated transforming growth factor beta receptor II (TGFβRII) signalling in dendritic cells promotes immune reactivity of T cells resulting in enhanced atherosclerosis. Eur. Heart J. 34: 3717–3727.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Kel J. M.,
    2. M. J. Girard-Madoux,
    3. B. Reizis,
    4. B. E. Clausen
    . 2010. TGF-beta is required to maintain the pool of immature Langerhans cells in the epidermis. J. Immunol. 185: 3248–3255.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Chytil A.,
    2. M. A. Magnuson,
    3. C. V. Wright,
    4. H. L. Moses
    . 2002. Conditional inactivation of the TGF-beta type II receptor using Cre:Lox. Genesis 32: 73–75.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Caton M. L.,
    2. M. R. Smith-Raska,
    3. B. Reizis
    . 2007. Notch-RBP-J signaling controls the homeostasis of CD8- dendritic cells in the spleen. J. Exp. Med. 204: 1653–1664.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Hayakawa, Y., Y. Hirata, H. Nakagawa, K. Sakamoto, Y. Hikiba, M. Otsuka, H. Ijichi, T. Ikenoue, K. Tateishi, M. Akanuma, et al. 2010. Apoptosis signal-regulating kinase 1 regulates colitis and colitis-associated tumorigenesis by the innate immune responses. Gastroenterology 138: 1055–1067.e1051–e1054. doi:10.1053/j.gastro.2009.11.015.
  23. ↵
    1. Johansson M. E.,
    2. M. Phillipson,
    3. J. Petersson,
    4. A. Velcich,
    5. L. Holm,
    6. G. C. Hansson
    . 2008. The inner of the two Muc2 mucin-dependent mucus layers in colon is devoid of bacteria. Proc. Natl. Acad. Sci. USA 105: 15064–15069.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. McCormick D. A.,
    2. L. W. Horton,
    3. A. S. Mee
    . 1990. Mucin depletion in inflammatory bowel disease. J. Clin. Pathol. 43: 143–146.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Podolsky D. K.,
    2. K. J. Isselbacher
    . 1983. Composition of human colonic mucin. Selective alteration in inflammatory bowel disease. J. Clin. Invest. 72: 142–153.
    OpenUrlCrossRefPubMed
  26. ↵
    1. van Es J. H.,
    2. M. E. van Gijn,
    3. O. Riccio,
    4. M. van den Born,
    5. M. Vooijs,
    6. H. Begthel,
    7. M. Cozijnsen,
    8. S. Robine,
    9. D. J. Winton,
    10. F. Radtke,
    11. H. Clevers
    . 2005. Notch/gamma-secretase inhibition turns proliferative cells in intestinal crypts and adenomas into goblet cells. Nature 435: 959–963.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Hammer G. E.,
    2. A. Ma
    . 2013. Molecular control of steady-state dendritic cell maturation and immune homeostasis. Annu. Rev. Immunol. 31: 743–791.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Mildner A.,
    2. S. Jung
    . 2014. Development and function of dendritic cell subsets. Immunity 40: 642–656.
    OpenUrlCrossRefPubMed
    1. Cerovic V.,
    2. S. A. Houston,
    3. C. L. Scott,
    4. A. Aumeunier,
    5. U. Yrlid,
    6. A. M. Mowat,
    7. S. W. Milling
    . 2013. Intestinal CD103(-) dendritic cells migrate in lymph and prime effector T cells. Mucosal Immunol. 6: 104–113.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Scott C. L.,
    2. C. C. Bain,
    3. P. B. Wright,
    4. D. Sichien,
    5. K. Kotarsky,
    6. E. K. Persson,
    7. K. Luda,
    8. M. Guilliams,
    9. B. N. Lambrecht,
    10. W. W. Agace,
    11. et al
    . 2015. CCR2(+)CD103(-) intestinal dendritic cells develop from DC-committed precursors and induce interleukin-17 production by T cells. Mucosal Immunol. 8: 327–339.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Fujiwara D.,
    2. L. Chen,
    3. B. Wei,
    4. J. Braun
    . 2011. Small intestine CD11c+ CD8+ T cells suppress CD4+ T cell-induced immune colitis. Am. J. Physiol. Gastrointest. Liver Physiol. 300: G939–G947.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Konkel J. E.,
    2. W. Chen
    . 2011. Balancing acts: the role of TGF-β in the mucosal immune system. Trends Mol. Med. 17: 668–676.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Monteleone G.,
    2. M. F. Neurath,
    3. S. Ardizzone,
    4. A. Di Sabatino,
    5. M. C. Fantini,
    6. F. Castiglione,
    7. M. L. Scribano,
    8. A. Armuzzi,
    9. F. Caprioli,
    10. G. C. Sturniolo,
    11. et al
    . 2015. Mongersen, an oral SMAD7 antisense oligonucleotide, and Crohn’s disease. N. Engl. J. Med. 372: 1104–1113.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Li M. O.,
    2. S. Sanjabi,
    3. R. A. Flavell
    . 2006. Transforming growth factor-beta controls development, homeostasis, and tolerance of T cells by regulatory T cell-dependent and -independent mechanisms. Immunity 25: 455–471.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Marie J. C.,
    2. D. Liggitt,
    3. A. Y. Rudensky
    . 2006. Cellular mechanisms of fatal early-onset autoimmunity in mice with the T cell-specific targeting of transforming growth factor-beta receptor. Immunity 25: 441–454.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Kim S. C.,
    2. S. L. Tonkonogy,
    3. C. A. Albright,
    4. J. Tsang,
    5. E. J. Balish,
    6. J. Braun,
    7. M. M. Huycke,
    8. R. B. Sartor
    . 2005. Variable phenotypes of enterocolitis in interleukin 10-deficient mice monoassociated with two different commensal bacteria. Gastroenterology 128: 891–906.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Lupp C.,
    2. M. L. Robertson,
    3. M. E. Wickham,
    4. I. Sekirov,
    5. O. L. Champion,
    6. E. C. Gaynor,
    7. B. B. Finlay
    . 2007. Host-mediated inflammation disrupts the intestinal microbiota and promotes the overgrowth of Enterobacteriaceae. Cell Host Microbe 2: 119–129.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Pédron T.,
    2. P. Sansonetti
    . 2008. Commensals, bacterial pathogens and intestinal inflammation: an intriguing ménage à trois. Cell Host Microbe 3: 344–347.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Winter S. E.,
    2. M. G. Winter,
    3. M. N. Xavier,
    4. P. Thiennimitr,
    5. V. Poon,
    6. A. M. Keestra,
    7. R. C. Laughlin,
    8. G. Gomez,
    9. J. Wu,
    10. S. D. Lawhon,
    11. et al
    . 2013. Host-derived nitrate boosts growth of E. coli in the inflamed gut. Science 339: 708–711.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Novitskiy S. V.,
    2. M. W. Pickup,
    3. A. Chytil,
    4. D. Polosukhina,
    5. P. Owens,
    6. H. L. Moses
    . 2012. Deletion of TGF-β signaling in myeloid cells enhances their anti-tumorigenic properties. J. Leukoc. Biol. 92: 641–651.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Danese S.,
    2. C. Fiocchi
    . 2011. Ulcerative colitis. N. Engl. J. Med. 365: 1713–1725.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Johansson M. E.,
    2. J. K. Gustafsson,
    3. J. Holmén-Larsson,
    4. K. S. Jabbar,
    5. L. Xia,
    6. H. Xu,
    7. F. K. Ghishan,
    8. F. A. Carvalho,
    9. A. T. Gewirtz,
    10. H. Sjövall,
    11. G. C. Hansson
    . 2014. Bacteria penetrate the normally impenetrable inner colon mucus layer in both murine colitis models and patients with ulcerative colitis. Gut 63: 281–291.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Kim J. J.,
    2. W. I. Khan
    . 2013. Goblet cells and mucins: role in innate defense in enteric infections. Pathogens 2: 55–70.
    OpenUrlCrossRefPubMed
  44. ↵
    1. Alkadhi S.,
    2. D. Kunde,
    3. R. Cheluvappa,
    4. S. Randall-Demllo,
    5. R. Eri
    . 2014. The murine appendiceal microbiome is altered in spontaneous colitis and its pathological progression. Gut Pathog. 6: 25.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Miyamoto S.,
    2. D. W. Rosenberg
    . 2011. Role of Notch signaling in colon homeostasis and carcinogenesis. Cancer Sci. 102: 1938–1942.
    OpenUrlCrossRefPubMed
  46. ↵
    1. Sakata-Yanagimoto M.,
    2. T. Sakai,
    3. Y. Miyake,
    4. T. I. Saito,
    5. H. Maruyama,
    6. Y. Morishita,
    7. E. Nakagami-Yamaguchi,
    8. K. Kumano,
    9. H. Yagita,
    10. M. Fukayama,
    11. et al
    . 2011. Notch2 signaling is required for proper mast cell distribution and mucosal immunity in the intestine. Blood 117: 128–134.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Mathern D. R.,
    2. L. E. Laitman,
    3. Z. Hovhannisyan,
    4. D. Dunkin,
    5. S. Farsio,
    6. T. J. Malik,
    7. G. Roda,
    8. A. Chitre,
    9. A. C. Iuga,
    10. G. Yeretssian,
    11. et al
    . 2014. Mouse and human Notch-1 regulate mucosal immune responses. Mucosal Immunol. 7: 995–1005.
    OpenUrlPubMed
  48. ↵
    1. Gersemann M.,
    2. S. Becker,
    3. I. Kübler,
    4. M. Koslowski,
    5. G. Wang,
    6. K. R. Herrlinger,
    7. J. Griger,
    8. P. Fritz,
    9. K. Fellermann,
    10. M. Schwab,
    11. et al
    . 2009. Differences in goblet cell differentiation between Crohn’s disease and ulcerative colitis. Differentiation 77: 84–94.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section