Abstract
TLRs are key sensors for conserved bacterial molecules and play a critical role in host defense against invading pathogens. Although the roles of TLRs in defense against pathogen infection and in maintaining gut immune homeostasis have been studied, the precise functions of different TLRs in response to pathogen infection in the gut remain elusive. The present study investigated the role of TLR signaling in defense against the Gram-negative bacterial pathogen Salmonella typhimurium. The results indicated that TLR9-deficient mice were more susceptible to S. typhimurium infection compared with wild-type and TLR2- or TLR4-deficient mice, as indicated by more severe intestinal damage and the highest bacterial load. TLR9 deficiency in intestinal epithelial cells (IECs) augmented the activation of NF-κB and NLRP3 inflammasomes significantly, resulting in increased secretion of IL-1β. IL-1β increased the expression of NKG2D on intestinal intraepithelial lymphocytes and NKG2D ligands on IECs, resulting in higher susceptibility of IECs to cytotoxicity of intestinal intraepithelial lymphocytes and damage to the epithelial barrier. We proposed that TLR9 regulates the NF-κB–NLRP3–IL-1β pathway negatively in Salmonella-induced NKG2D-mediated intestinal inflammation and plays a critical role in defense against S. typhimurium infection and in the protection of intestinal integrity.
Introduction
Innate immunity acts as the first line of defense against invading pathogens by limiting microbial replication and spread, before priming an adaptive response (1). The innate immune system recognizes conserved microbial features common to broad classes of microbes via a limited number of germline-encoded receptors, such as TLRs (2). Thirteen members of the TLR family have been characterized: TLR1–TLR9 are expressed in humans and mice, TLR10 is expressed only in humans, and TLR11–TLR13 are expressed only in mice (3). TLRs target a range of microbial ligands, including lipoproteins (the ligand of TLR2), dsRNA (the ligand of TLR3), LPS (the ligand of TLR4), flagellin (the ligand of TLR5), ssRNA (the ligand of TLR7 and TLR8), and bacterial unmethylated CpG DNA (the ligand of TLR9) (4, 5). TLRs function as sentinels of innate immunity. When recognizing their corresponding ligands, TLRs trigger the transcriptional activation of proinflammatory cytokines, chemokines, and costimulatory molecules, subsequently initiating defense mechanisms that are vital for host survival (6). The interplay between host innate immunity and pathogen virulence largely determines the outcome of most infections. For example, TLR4 mediates cellular activation in response to LPS derived from Escherichia coli and Salmonella minnesota (7). TLR5 contributes to host resistance to i.p. infection with S. typhimurium or i.n. infection with Pseudomonas aeruginosa, and this role could be masked by the presence of a functional TLR4 (8). TLR9 is required for the Th1-type inflammatory response that follows oral infection with Toxoplasma gondii (9) and is critical for the early control of lesion development and parasite burden during Leishmania braziliensis infection (10). Despite many recent advances, the role of TLR signaling and the proinflammatory transcription factor NF-κB in mucosal immune responses to pathogens remains unclear. Fully determining the role and mechanisms of TLRs in the defense against invading pathogens is imperative.
S. typhimurium, a facultative intracellular Gram-negative bacteria and an important food-borne pathogen, can cause a broad range of clinical manifestations in human and animal hosts. For example, it causes a self-limited gastroenteritis, which is characterized by fever, diarrhea, acute intestinal inflammation, and the presence of neutrophils in stool samples (11). In addition, S. typhimurium is a model organism to study bacterial genetics and microbial pathogenesis (12). Four TLR ligands have been found in S. typhimurium: bacterial lipoproteins are recognized by TLR2, LPS is recognized by TLR4, flagellin is recognized by TLR5, and CpG DNA is recognized by TLR9. Recent innovative research has shed some light on the molecular mechanisms of the potential interactions between S. typhimurium and TLR5. The results suggested that, on the one hand, TLR5 recognition promotes bacterial colonization in the intestinal and systemic infection; however, on the other hand, TLR5 recognition attenuates intestinal inflammation (13). The role of TLR2, TLR4, and TLR9 signaling also seems crucial for the recognition of S. typhimurium infection and the defense via host innate immunity; however, their precise role during S. typhimurium infection remains unclear. Recently, our group found that TLR2 and TLR9 play opposing roles in host innate immunity against S. typhimurium infection: TLR2 regulates the anti–S. typhimurium immune response negatively, whereas TLR9 is vital for host defense against S. typhimurium invasion (14). In the current study, we used TLR2-, TLR4-, and TLR9-deficient mice to further investigate the roles of TLRs in murine S. typhimurium oral infection. Our findings indicated a critical role for TLR9 in the protection of intestinal epithelium integrity and in defense against S. typhimurium infection.
Materials and Methods
Cell lines and cell culture
The murine intestinal epithelial cell (IEC) line MODE-K (generously provided by Dr. Caigan Du, University of British Columbia, Vancouver, BC Canada), the murine melanoma cell line B16, and the murine macrophage cell line RAW264.7 were cultured in DMEM (Life Technologies/BRL, Grand Island, NY) supplemented with 10% FBS at 37°C in a humidified 5% CO2 atmosphere.
Mice
Male C57BL/6 mice (6–8 wk) were purchased from Beijing Huafukang Bioscience (Beijing, China). Rag1−/− mice were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). TLR2−/− and TLR4−/− mice with the C57BL/6 background were kindly provided by S.B. Su (Sun Yat-Sen University, Guangzhou, China), and TLR9−/− mice with the C57BL/6 background were generously provided by Dr. S. Akira (Osaka University, Osaka, Japan).
NLRP3−/− mice were described previously (15). TLR9−/− and NLRP3−/− mice were crossed to generate TLR9 and NLRP3 double-deficient mice (TLR9−/− × NLRP3−/−). All mice were maintained under specific pathogen–free conditions. To prevent the influence of microbiota, TLR−/− mice and wild-type (WT) mice were fed in the same cages. All animal experiments and protocols were conducted in accordance with guidelines for experimental animals from Shandong University and were approved by the Committee on the Ethics of Animal Experiments of Shandong University.
Abs
The following mAbs were used in this study. FITC Armenian Hamster anti-γδTCR (clone GL3), PercpCy5.5-Rat anti-CD8α (clone 53-6.7), FITC-Rat anti-CD8β (clone H35-17.2), PE Armenian Hamster anti-CD69 (clone H1.2F3), allophycocyanin-Rat anti-NKG2D (clone CX5), allophycocyanin-Rat anti-CD107a (clone 1D4B), and LEAF Purified anti-NKG2D (clone C7) were purchased from BioLegend (San Diego, CA). PE anti-H60 (clone 205326), PerCP anti–MULT-1 (clone 237104), FITC anti–Rae-1 (clone 186107), and PE anti–IL-1RI (clone 129304) were purchased from R&D Systems (Minneapolis, MN). PE-goat anti-rabbit IgG (Santa Cruz Biotechnology, Dallas, TX) was used as a secondary reagent to identify NF-κB p65 and p–NF-κB p65 by flow cytometry analysis. Each Ab was titrated to determine the optimal staining concentration for maximal signal strength.
Bacterial strains and growth conditions
A virulent atrichia S. typhimurium strain (ATCC 14028) was cultured in Luria–Bertani (LB; 10 g of tryptone, 5 g of yeast extract, and 10 g of NaCl per liter) broth or on LB agar plates. The bacteria were cultured at 37°C overnight on LB agar plates and then a single colony was selected to culture in LB broth for 12 h. The bacterial cultures at saturation density were diluted (1/100) and cultured to the midlogarithmic growth phase (OD600, 0.4–0.6). The OD value of bacteria was tested using an UV spectrophotometer (Purkinje General, Beijing, China), and the number of bacteria was calculated from a standard curve relating CFU to OD600. Bacterial cultures were centrifuged and washed in LB broth twice before use.
Infection of mice
Mice were inoculated with suspensions (0.2 ml) of S. typhimurium (5 × 104) in sterile LB by oral gavage or with an equal volume of a sterile LB broth vehicle as mock-infected control. The small intestine, liver, mesenteric lymph node (MLN), and spleen were isolated from infected and control animals at different days postinfection.
Quantification of bacterial burden
The liver, spleen, small intestine tissue, and MLN were obtained at different time points postinfection, homogenized using a tissue homogenizer (TIANGEN, Beijing, China), and lysed with 0.1% Triton X-100 in 1× PBS to release the bacteria. A dilution series in 1× PBS was plated onto LB agar plates, and colonies of bacteria from triplicates were counted after incubation overnight at 37°C.
Preparation of intestinal intraepithelial lymphocytes and IECs
Intestinal intraepithelial lymphocytes (iIELs) and IECs were isolated from the small intestine, as previously described (16, 17). To prepare iIELs, the small intestine was divided into 2–3-cm pieces, and the specimens were placed in prewarmed iIEL digestive juice and incubated with stirring at 37°C for 40 min, followed by vigorous shaking for 30 s. The supernatants were passed over two nylon wool columns to remove undigested tissue debris. The filtrate was collected and centrifuged, and the lymphocytes were enriched by a discontinuous (40 and 70%) Percoll density gradient (GE Healthcare, Uppsala, Sweden). The cells (iIELs) at the interface between the 40 and 70% fractions were collected. The viability, purity, and phenotype of iIELs were identified using FACS analysis.
To prepare IECs, Peyer’s patches were excised, and the small intestine was opened longitudinally and washed in 1× PBS containing 100 U/ml penicillin-streptomycin. The tissue was cut into 3–4-cm fragments and incubated at 24°C on a shaker platform in 1× PBS containing 0.02 mg/ml Dispase I and 60 U/ml Collagenase XIa (both from Sigma-Aldrich, St. Louis, MO) for 15 min. Serum (2%) was added to terminate the digestion process. Cells and small sheets of intestinal epithelium were separated from the intestinal fragments by harvesting supernatants after 2 min of deposition in medium. Cells were then centrifuged three times in DMEM plus 2% sorbitol, and the supernatants were discarded. The remaining pellet consisted mainly of cells in intact crypts and small sheets of intestinal epithelium (organoids). Cell viability was assessed using 7-aminoactinomycin D staining (>80%). IECs were cultured in epithelial cell medium containing equal volumes of phenol red–free DMEM and Ham’s F-12 medium with the following additives: 2% FBS, 5 mg/ml insulin (Sigma-Aldrich), 5 × 108 M dexamethasone (Sigma-Aldrich), 60 nM selenium (Sigma-Aldrich), 2 mM glutamine, 5 μg/ml transferrin (Sigma-Aldrich), 10 ng/ml epidermal growth factor (Sigma-Aldrich), 5 × 108 M triiodothyronine (Sigma-Aldrich), 20 mM HEPES, 100 mg/ml streptomycin, 100 U/ml penicillin, and 0.2% d-glucose. The primary IECs adhered well to plates coated with rat tail collagen and expressed cytokeratin, the marker of IECs.
Adoptive transfer of iIELs
iIELs were prepared and transferred adoptively as previously described (18). Briefly, iIELs (2.5 × 106) isolated from WT and TLR9−/− mice were transferred adoptively into naive WT recipients. Two days later, recipient mice were challenged with S. typhimurium.
Bone marrow chimeras
Lethally irradiated WT mice, and TLR9−/− mice (recipient mice) were reconstituted with 1 × 107 bone marrow (BM) cells from WT or TLR9−/− mice. For NLRP3−/− mice, lethally irradiated WT and NLRP3−/− mice (recipient mice) were reconstituted with 1 × 107 BM cells from WT or NLRP3−/− mice. After 6 wk, the recipient mice were infected with S. typhimurium.
Quantitative real-time PCR
Total RNA was extracted from purified IECs or MODE-K cells using a TRIzol RNA isolation kit (Invitrogen, Life Technologies, Carlsbad, CA). Total RNA was reverse transcribed to generate cDNA, according to the protocol provided by the manufacturer (Invitrogen). cDNA was quantified using real-time PCR analysis. The Actb gene (encoding β-actin) was used as an internal control. PCR reactions were carried out using SYBR Green Mix (Roche, Basel, Switzerland). Primer sequences are as follows: RAE-1, 5′-ATCAACTTCCCCGCTTCCA-3′ (forward) and 5′-AGATATGAAGATGAGTCCCACAGAGATA-3′ (reverse); H60, 5′-GAGCCACCAGCAAGAGCAA-3′ (forward) and 5′-CCAGTATGGTCCCCAGATAGCT-3′ (reverse); MULT-1, 5′-TTCACATAGTGCAGGAGACTAACACA-3′ (forward) and 5′-ACTGGCCACACACCTCAGC-3′ (reverse); IL-1β, 5′-TCTTTGAAGTTGACGGACCC-3′ (forward) and 5′-TGAGTGATACTGCCTGCCTG-3′ (reverse); IL-18, 5′-ACAACTTTGGCCGACTTCAC-3′ (forward) and 5′-GGGTTCACTGGCACTTTGAT-3′ (reverse); NLRP3, 5′-TCCACAATTCTGACCCACAA-3′ (forward) and 5′-ACCTCACAGAGGGTCACCAC-3′ (reverse); ASC, 5′-CTTGTCAGGGGATGAACTCAAAA-3′ (forward) and 5′-GCCATACGACTCCAGATAGTAGC-3′ (reverse); caspase-1, 5′-TCAGCTCCATCAGCTGAAAC-3′ (forward) and 5′-TGGAAATGTGCCATCTTCTTT-3′ (reverse); and β-actin, 5′-AGAGGGAAATCGTGCGTGAC-3′ (forward) and 5′-CAATAGTGATGACCTGGCCGT-3′ (reverse).
Lysis assay
The cytolytic activity of iIELs was assessed using a lactate dehydrogenase (LDH) release assay. IECs were used as target cells. IECs were isolated from noninfected or S. typhimurium–infected WT, TLR9−/−, or NLRP3−/− mice at day 5 postinfection, seeded in 96-well plates (2 × 104 per well) coated with rat tail collagen, and incubated for 24 h in epithelial cell medium. iIELs were isolated from noninfected or Salmonella-infected WT or TLR9−/− mice at day 5 postinfection and were added to target cells in 96-well plates at an E:T ratio of 10:1. The plate was incubated for 6 h, and the assay was performed according to the protocol provided by the manufacturer (LDH Cytotoxicity Assay Kit; Beyotime, Nantong, China). Cytotoxicity was calculated as follows (A, absorbance): percentage lysis = [1 − (A of cocultured cells − A of effector cells)/A of target cells] × 100%.
Western blotting analysis
IECs were isolated from S. typhimurium–infected WT or TLR9−/− mice at day 5 postinfection. Western blotting was carried out as described previously (19). The following Abs were used: rabbit anti–NF-κB p65 and rabbit anti–p-NF-κB p65 (Cell Signaling Technology, Danvers, MA).
H&E staining and evaluation
For histochemical analysis, tissues from the lower small intestine were excised at day 3, 5, or 7 postinfection, fixed in 4% neutral-buffered formalin, dehydrated, and embedded in paraffin. Four-micrometer sections were fixed to slides, deparaffinized, and stained with H&E. Morphological changes in the stained sections were examined using light microscopy. Small intestine epithelial damage was scored blindly, as described previously (20). Briefly, small intestine epithelial damage was scored as follows: 0 = normal; 1 = hyperproliferation, irregular crypts, and goblet cell loss; 2 = mild to moderate crypt loss (10–50%); 3 = severe crypt loss (50–90%); 4 = complete crypt loss, surface epithelium intact; 5 = small to medium–sized ulcer (<10 crypt widths); and 6 = large ulcer (>10 crypt widths). Infiltration with inflammatory cells was scored separately for mucosa (0 = normal, 1 = mild, 2 = modest, 3 = severe), submucosa (0 = normal, 1 = mild, 2 = modest, 3 = severe), and muscle and serosa (0 = normal, 2 = moderate to severe). Scores for epithelial damage and inflammatory cell infiltration were added.
ELISA
Mouse tissue culture supernatants of small intestines were stored at −80°C before cytokine measurement. The levels of IL-1β and IL-18 were measured using commercially available ELISA kits (CUSABIO, Wuhan, China), according to the manufacturer’s instructions.
Immunofluorescence staining
MODE-K cells were fixed with 4% paraformaldehyde for 10 min at room temperature postinfection with S. typhimurium for 24 h. MODE-K cells were blocked with blocking buffer (5% normal mouse serum and 0.3% Triton X-100 in 1× PBS) for 60 min at room temperature. Cells were then incubated with recombinant mouse NKG2D-Fc Chimera (4 μg/ml; R&D Systems) overnight at 4°C and secondary Ab mouse anti-human IgG DyLight 549 (1:100; Jackson ImmunoResearch, West Grove, PA) for 1 h at room temperature. Nuclei were stained with Hoechst 33342 (1:1000; Beyotime, Shanghai, China) for 2 min at room temperature. Negative controls were made by omission of the primary Ab.
To detect TLR9, MODE-K, RAW264.7, and primary IECs were fixed with 4% paraformaldehyde for 15 min at room temperature after overnight culture. To detect cytokeratin, IECs were fixed with 4% paraformaldehyde for 15 min at room temperature after culture for 6 and 24 h. Cells were washed with 1× PBS, blocked with blocking buffer (10% normal goat serum and 1% Triton X-100 in 1× PBS) for 60 min at room temperature, and incubated with Rabbit polyclonal Abs for TLR9 (ab37154; Abcam, Cambridge, MA) for 6 h at 4°C, followed by secondary Ab Alexa Fluor 549–conjugated goat anti-rabbit IgG (1:100; Proteintech, Rosemont, IL) for 1 h at room temperature, or were incubated with FITC–anti-pan cytokeratin (clone PCK-26; Sigma-Aldrich) for 30 min at 37°C. Nuclei were stained with DAPI for 3 min at room temperature.
To detect TCRγδ+ iIELs in the small intestine, tissue sections were subjected to Ag retrieval after deparaffinization and dehydration and were blocked with 5% BSA for 30 min at 37°C. The Ab solution (FITC–anti-TCRγδ mAb, clone GL-3; Abcam) was added to cover the tissue sections and incubated overnight at 4°C. Nuclei were stained with DAPI for 4 min at room temperature. Tissue sections were mounted with glycerol mounting medium.
Statistical analysis
All analyses were carried out using GraphPad Prism 5 (GraphPad, La Jolla, CA). Data were analyzed for statistical significance using one-way ANOVA. Student t tests were performed to compare values obtained from two groups. A p value <0.05 was considered statistically significant.
Results
TLR9 protects intestinal integrity during S. typhimurium infection
To explore the roles of TLR2, TLR4, and TLR9 in the defense against S. typhimurium infection, age-matched WT, TLR2−/−, TLR4−/−, and TLR9−/− mice were infected orally with S. typhimurium at 5 × 104 CFU. With increasing time, the body weights of TLR9−/− mice were reduced markedly compared with WT mice; this was accompanied by diminished activity, eating less, and lower energy levels. However, there were no significant changes in TLR2−/− and TLR4−/− mice compared with WT mice. We further observed changes in intestinal inflammation and bacterial load in WT, TLR2−/−, TLR4−/−, and TLR9−/− mice. The histopathology of the small intestines and S. typhimurium burden were assessed at days 3, 5, and 7 postinfection. Histopathological observation after H&E staining revealed greater inflammatory cell infiltration in the mucosa and lamina propria of the lower small intestine in TLR9−/− mice compared with WT mice at days 3 and 5. Moreover, the integrity of the epithelial barrier was damaged severely in TLR9−/− mice at day 7 (Fig. 1A). In comparison, the damage in TLR4−/− mice was similar to that in WT mice. Moreover, the inflammatory signs and the damage to intestinal villi in TLR2−/− mice were the lowest at the different days postinfection (Fig. 1A). The histological score of the small intestine sections also revealed more significant epithelial damage and inflammation in TLR9−/− mice compared with WT mice at days 3, 5, and 7 postinfection (Fig. 1B). The least amount of small intestinal damage resulting from S. typhimurium infection was observed in TLR2−/− mice. More obvious intestinal damage in TLR4−/− mice was found only at day 3 compared with WT mice (Fig. 1B). Higher S. typhimurium loads in the small intestine and the MLNs were observed in TLR9−/− mice, but not in TLR2−/− or TLR4−/− mice, compared with WT mice at days 5 and 7 postinfection (Fig. 1C, 1D). Collectively, these results demonstrated that TLR9-knockout (KO) mice were significantly more susceptible to oral S. typhimurium infection compared with WT mice, as indicated by severe intestinal damage and higher bacterial load in the small intestine and MLN. Therefore, TLR9 might play a critical role in the host defense against S. typhimurium infection.
TLR9−/− mice are more sensitive to S. typhimurium infection compared with WT controls. (A) WT, TLR2−/−, TLR4−/−, and TLR9−/− mice were infected orally with 5 × 104 CFU S. typhimurium. Histological sections of their small intestines were prepared and stained with H&E at days 3, 5, and 7 postinfection (original magnification ×200). (B) Histological scoring for intestinal inflammation and injury from H&E-stained small intestine sections obtained at days 3, 5, and 7 postinfection with S. typhimurium from WT, TLR2−/−, TLR4−/−, and TLR9−/− mice (n = 4). Bacterial burden measured at days 3, 5, and 7 postinfection in the small intestine (C) and MLNs (D) of WT, TLR2−/−, TLR4−/−, and TLR9−/− mice (n = 5). Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, Student t test.
TLR9−/− mice are more sensitive to S. typhimurium infection compared with WT controls. (A) WT, TLR2−/−, TLR4−/−, and TLR9−/− mice were infected orally with 5 × 104 CFU S. typhimurium. Histological sections of their small intestines were prepared and stained with H&E at days 3, 5, and 7 postinfection (original magnification ×200). (B) Histological scoring for intestinal inflammation and injury from H&E-stained small intestine sections obtained at days 3, 5, and 7 postinfection with S. typhimurium from WT, TLR2−/−, TLR4−/−, and TLR9−/− mice (n = 4). Bacterial burden measured at days 3, 5, and 7 postinfection in the small intestine (C) and MLNs (D) of WT, TLR2−/−, TLR4−/−, and TLR9−/− mice (n = 5). Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, Student t test.
TLR9 deficiency promotes the activation and NKG2D-mediated cytotoxicity of iIELs against S. typhimurium–infected IECs
iIELs in the small intestine are the first elements of the T cell compartment available to respond to oral infection with pathogens. To explore the roles of iIELs in the prevention and clearance of S. typhimurium infection in WT and TLR9−/− mice, iIELs were isolated from the small intestine, and their viability, purity, and phenotype were assessed. As shown in Supplemental Fig. 1A and 1B, the cell viability was >90%, as assessed by 7-aminoactinomycin D staining, and CD3+ T cells accounted for >90%. The three subpopulations of iIELs were identified as TCRγδ+ iIELs, TCRαβ+CD8αα+ iIELs, and TCRαβ+CD8αβ+ iIELs, accounting for ∼35, 53, and 17%, respectively (Supplemental Fig. 1C), which was consistent with a previous report (21). These results indicated that we had isolated genuine iIELs. Flow cytometry analysis was performed to detect the percentages of different iIELs subpopulations at days 3, 5, and 7 postinfection. We found that the percentages of TCRγδ+ iIELs increased in WT and TLR9−/− mice at each day postinfection, whereas, at day 7, only the percentage of TCRγδ+ iIELs in TLR9−/− mice was increased significantly relative to WT mice (Supplemental Fig. 1D, 1E). The percentages of CD8αα+ and CD8αβ+ iIELs did not change during S. typhimurium infection in WT or TLR9−/− mice. The absolute numbers of TCRγδ+ iIELs in TLR9−/− mice also increased significantly relative to WT mice at day 7 postinfection (Fig. 2A). The expression of CD69, the activation marker for iIELs, on TCRγδ+ iIELs increased significantly from day 5 to day 7 postinfection in TLR9−/− mice compared with WT mice, whereas, at day 7, CD8αα+ iIELs in TLR9−/− mice showed significantly increased CD69 expression (Fig. 2B). Through the use of immunofluorescence staining, we further confirmed the increase in TCRγδ+ iIELs in the epithelium of the intestine postinfection (Supplemental Fig. 1F). These results demonstrated that S. typhimurium infection promoted the recruitment or proliferation of iIELs and stimulated the activation of iIEL subpopulations. Interestingly, these effects were more prominent when TLR9 was depleted.
Oral infection with S. typhimurium causes different changes in iIEL populations in WT and TLR9−/− mice. At various time points postinfection, the small intestines were removed, and iIELs were isolated and stained with fluorescently labeled mAbs for analysis by FACS. (A) Statistical analysis of the absolute number of TCRγδ+ iIELs, CD8αα+ iIELs, and CD8αβ+ iIELs at days 3, 5, and 7 after S. typhimurium infection. (B) Statistical analysis of CD69 expression (MFI) by TCRγδ+ iIELs, CD8αα+ iIELs, and CD8αβ+ iIELs. n = 4 or 5 for each group in (A) and (B). Data are representative of at least three independent experiments (mean ± SD). *p < 0.05, Student t test. MFI, mean fluorescent intensity.
Oral infection with S. typhimurium causes different changes in iIEL populations in WT and TLR9−/− mice. At various time points postinfection, the small intestines were removed, and iIELs were isolated and stained with fluorescently labeled mAbs for analysis by FACS. (A) Statistical analysis of the absolute number of TCRγδ+ iIELs, CD8αα+ iIELs, and CD8αβ+ iIELs at days 3, 5, and 7 after S. typhimurium infection. (B) Statistical analysis of CD69 expression (MFI) by TCRγδ+ iIELs, CD8αα+ iIELs, and CD8αβ+ iIELs. n = 4 or 5 for each group in (A) and (B). Data are representative of at least three independent experiments (mean ± SD). *p < 0.05, Student t test. MFI, mean fluorescent intensity.
iIELs are located between the IECs (22) and play an important role in maintaining the barrier function and the integrity of the intestinal epithelium (23). iIELs exhibit cytotoxic activity to eliminate infected IECs, but they may also lead to intestinal damage when the cytotoxicity is too strong. To investigate whether the signs of severe epithelial damage and inflammation in TLR9−/− mice were derived from too strong cytotoxicity from iIELs against infected IECs, we assessed the levels of NKG2D and CD107a in iIEL subpopulations from WT and TLR9−/− mice postinfection with S. typhimurium for 5 d. The results showed that the levels of NKG2D and CD107a were much higher in each iIEL subpopulation from TLR9−/− mice compared with those from WT mice (Fig. 3A, 3B). Primary IECs from the small intestine were then isolated, and the viability, adhesion features, phenotype, and TLR9 levels were determined (Supplemental Fig. 2). The levels of NKG2D ligands on epithelial cell line MODE-K cells and primary IECs, which both express functional TLR9 on their membranes and intracellularly (Supplemental Fig. 2D, 2E), were examined after S. typhimurium infection using immunofluorescence and FACS, separately. We found that the expression of NKG2D ligands was upregulated after S. typhimurium infection (Fig. 3C, 3D). The protein levels of NKG2D ligands (MULT-1, RAE-1, and H60) on primary IECs were increased in WT and TLR9−/− mice, but the extent of upregulation of RAE-1 and H60 in TLR9−/− mice was more obvious than that in WT mice (Fig. 3D). Finally, the cytotoxic potential of iIELs against IECs was measured. As shown in Fig. 3E, there was almost no cytotoxicity of iIELs from noninfected or infected mice against normal uninfected IECs; however, the infected iIELs exhibited higher cytolytic activity against infected IECs than did the uninfected iIELs. The iIELs from TLR9−/− mice displayed much stronger cytotoxicity against infected IECs than did the iIELs from WT mice.
TLR9 deficiency enhances the cytolytic activity of iIELs against S. typhimurium–infected IECs in an NKG2D-dependent manner. Total iIELs and IECs were isolated from S. typhimurium–infected WT or TLR9−/− mice at day 5 postinfection. (A) The expression levels of NKG2D and CD107a in different subsets of iIELs were examined by FACS. (B) Statistical analysis of the percentages of NKG2D+ and CD107a+ iIEL subsets. (C) The levels of NKG2D ligands on MODE-K cells postinfection with S. typhimurium (MOI = 10:1) for 24 h were determined using immunofluorescence. (D) The surface expression (percentages of positive cells) of MULT-1, RAE-1, and H60 on primary IECs from WT or TLR9−/− mice was analyzed by FACS. (E) The cytotoxicity of iIELs against IECs was measured using an LDH assay (E:T = 10:1). (F) The cytotoxicity of iIELs against S. typhimurium–infected IECs after blocking with anti-NKG2D Ab was detected at an E:T ratio of 25:1 using an LDH assay. iIELs were cocultured with saturating concentrations of anti-NKG2D Ab (20 μg/ml) or an isotype-matched control Ab and then washed for use as effector cells. (G) WT or TLR9−/− mice (n = 3) were pretreated with anti-NKG2D mAb or isotype-control Ab (400 g i.v. per mouse) at 24 h before S. typhimurium infection. The intestinal inflammation and injury of the small intestine at day 5 postinfection were then determined by H&E staining (original magnification ×100). (H) Histological scores of the small intestine sections. Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test. MOI, multiplicity of infection.
TLR9 deficiency enhances the cytolytic activity of iIELs against S. typhimurium–infected IECs in an NKG2D-dependent manner. Total iIELs and IECs were isolated from S. typhimurium–infected WT or TLR9−/− mice at day 5 postinfection. (A) The expression levels of NKG2D and CD107a in different subsets of iIELs were examined by FACS. (B) Statistical analysis of the percentages of NKG2D+ and CD107a+ iIEL subsets. (C) The levels of NKG2D ligands on MODE-K cells postinfection with S. typhimurium (MOI = 10:1) for 24 h were determined using immunofluorescence. (D) The surface expression (percentages of positive cells) of MULT-1, RAE-1, and H60 on primary IECs from WT or TLR9−/− mice was analyzed by FACS. (E) The cytotoxicity of iIELs against IECs was measured using an LDH assay (E:T = 10:1). (F) The cytotoxicity of iIELs against S. typhimurium–infected IECs after blocking with anti-NKG2D Ab was detected at an E:T ratio of 25:1 using an LDH assay. iIELs were cocultured with saturating concentrations of anti-NKG2D Ab (20 μg/ml) or an isotype-matched control Ab and then washed for use as effector cells. (G) WT or TLR9−/− mice (n = 3) were pretreated with anti-NKG2D mAb or isotype-control Ab (400 g i.v. per mouse) at 24 h before S. typhimurium infection. The intestinal inflammation and injury of the small intestine at day 5 postinfection were then determined by H&E staining (original magnification ×100). (H) Histological scores of the small intestine sections. Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test. MOI, multiplicity of infection.
To further investigate whether the increased levels of NKG2D and its ligands were responsible for the stronger cytotoxicity of iIELs against S. typhimurium–infected IECs, saturating concentrations of an anti-NKG2D Ab were added to block the interaction of NKG2D and its ligands. As shown in Fig. 3F, blocking NKG2D attenuated the cytotoxicity of iIELs against S. typhimurium–infected IECs significantly, and the cytolytic activity of iIELs from TLR9−/− mice was reduced to nearly the same level as that of iIELs from WT mice. Furthermore, 400 μg of anti-NKG2D–blocking Ab was injected i.v. into WT or TLR9−/− mice 24 h before S. typhimurium infection to block the effect of NKG2D in vivo. Five days postinfection, intestinal damage was measured. The results showed that blocking NKG2D attenuated the intestinal damage significantly in WT and TLR9−/− mice, with similar degrees of damage in both types of mice (Fig. 3G, 3H). These results showed that iIELs indeed exert their cytolytic activity against S. typhimurium–infected IECs in an NKG2D-dependent manner. Collectively, these results indicated that TLR9 deficiency aggravates the NKG2D-dependent killing of iIELs against S. typhimurium–infected IECs, which accounted for the severe epithelial damage.
TLR9 deficiency augments the NF-κB–NLRP3–IL-1β pathway in IECs
To further determine the exact mechanism of TLR9 in the protection of intestinal integrity, we attempted to analyze whether TLR9 deficiency in IECs or iIELs contributes to the severe intestinal damage during infection. iIELs isolated from WT or TLR9−/− mice were adoptively transferred into Rag1−/− mice, separately, via tail i.v. administration. Two days later, the recipient mice were infected with S. typhimurium, and the bacterial burden and inflammatory injury status were detected 2 d postinfection (Fig. 4A). H&E staining and histological scores showed no differences in inflammation or intestinal injury in the two groups (Fig. 4B, 4C). There also were no significant differences in bacterial burden in the liver, spleen, or MLN between the two groups (Fig. 4D). These results suggested that TLR9 deficiency in iIELs did not influence the susceptibility and intestinal injury of mice in response to S. typhimurium infection.
TLR9 deficiency in IECs, but not BM-derived iIELs, exacerbated the intestinal injury. (A and E) Study flowcharts. (B and F) H&E histology of the representative mouse small intestines (original magnification ×100). (C and G) Histological scores of the small intestine sections (n = 4 or 5). (D) Bacterial burden of the liver, spleen, and MLN from the following iIEL-transplanted mice on day 5 of infection: WT→Rag1−/− and TLR9−/−→Rag1−/− (n = 4). (H) Bacterial burden of the liver, spleen, and MLN from the following BM-transplanted mice on day 5 of infection: WT→WT (lethally irradiated), TLR9−/−→WT (lethally irradiated), and WT→TLR9−/− (lethally irradiated) (n = 5). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
TLR9 deficiency in IECs, but not BM-derived iIELs, exacerbated the intestinal injury. (A and E) Study flowcharts. (B and F) H&E histology of the representative mouse small intestines (original magnification ×100). (C and G) Histological scores of the small intestine sections (n = 4 or 5). (D) Bacterial burden of the liver, spleen, and MLN from the following iIEL-transplanted mice on day 5 of infection: WT→Rag1−/− and TLR9−/−→Rag1−/− (n = 4). (H) Bacterial burden of the liver, spleen, and MLN from the following BM-transplanted mice on day 5 of infection: WT→WT (lethally irradiated), TLR9−/−→WT (lethally irradiated), and WT→TLR9−/− (lethally irradiated) (n = 5). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
BM chimeras were then used to confirm the role of TLR9 in IECs. Lethally irradiated WT and TLR9−/− recipient mice were reconstituted with BM cells from WT or TLR9−/− mice. Six weeks later, the recipient mice were infected with S. typhimurium, and the bacterial burden and inflammatory injury status were assessed at 5 d postinfection (Fig. 4E). We observed more inflammatory cell infiltration in the mucosa and lamina propria, as well as more severe intestinal injury, in TLR9−/−-recipient mice than in WT-recipient mice (Fig. 4F, 4G). The bacterial burden in the liver, spleen, and MLN was significantly higher in TLR9−/−-recipient mice than in WT-recipient mice (Fig. 4H). However, there was no significant difference in inflammatory cell infiltration, intestinal injury, or bacterial burden between the TLR9−/−→WT group and the WT→WT group (Fig. 4F–H). These results further confirmed that TLR9 deficiency in BM-derived iIELs did not influence the susceptibility and intestinal injury of mice in response to S. typhimurium infection, whereas TLR9 deficiency in IECs exacerbated the intestinal injury, possibly leading to greater bacterial spread into the liver, spleen, and MLN. Collectively, these data showed that TLR9 in IECs plays a crucial role in protecting mice from excessive injury by iIELs during S. typhimurium infection.
IECs not only constitute a physical barrier to the external environment, they also discriminate commensal and pathogenic bacteria and play an essential role in regulating immune responses and maintaining normal homeostasis in the small intestine. IECs express a broad spectrum of TLRs, including TLR2, TLR3, TLR4, TLR7, and TLR9. When pathogenic bacteria are recognized by TLRs, activation of the NF-κB pathway promotes a cascade of signaling events that results in the expression of proinflammatory cytokines and chemokines (24, 25). However, TLR ligand–induced cross-tolerance to TLR activation has been described recently in IECs stimulated with bacterial DNA motifs binding to TLR9 (20, 26). To observe the inflammatory response in the IECs of WT and TLR9−/− mice after S. typhimurium infection, the gene and protein levels of inflammatory cytokines IL-1β and IL-18 were assessed using real-time PCR and ELISA. The mRNA levels of IL-1β and IL-18 were upregulated in WT and TLR9−/− mice after S. typhimurium infection, and more obvious upregulation was observed in TLR9−/− mice (Fig. 5A). However, only the protein level of IL-1β in tissue culture supernatants exhibited a similar phenomenon, whereas the level of IL-18 exhibited no significant changes (Fig. 5B). The production of IL-1β is dependent on NF-κB signaling; therefore, the activation of NF-κB was further assessed by Western blotting and flow cytometry. A higher level of NF-κB activation was observed in the IECs from TLR9−/− mice compared with those from WT mice (Fig. 5C, 5D). To determine whether the activation of NF-κB signaling promoted the expression of NKG2D ligands in IECs, NF-κB signaling was blocked using pyrrolidine dithiocarbamate, an inhibitor of NF-κB, before MODE-K cells were infected with S. typhimurium. As shown in Fig. 5E, the blockage of NF-κB signaling completely eliminated the increase in NKG2D ligands in the IECs. These data revealed that TLR9 deficiency leads to stronger inflammatory responses, which enhance NKG2D ligand expression in/on S. typhimurium–infected IECs, thus resulting in increased susceptibility of IECs to the cytotoxicity of iIELs, ultimately resulting in severe intestinal damage.
TLR9 deficiency augmented the expression of IL-1β and the activation of NF-κB in IECs in response to S. typhimurium infection. IECs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. (A) The mRNA levels of IL-1β and IL-18 in mock-infected or S. typhimurium–infected IECs were determined by quantitative real-time PCR. Results are shown as the fold increase in the expression of IL-1β and IL-18 compared with that of Actb. Gene expression values were then calculated based on the ΔΔCT method, using the mean of results for the mock-infected IECs as a calibrator. Relative quantities (RQs) were determined using the equation RQ = 2−ΔΔCT (n = 5 or 6). (B) The levels of IL-1β and IL-18 in the supernatants of small intestine culture were measured using ELISA (n = 5 or 6). (C) The total proteins were extracted from the infected or mock-infected IECs and subjected to Western blotting analysis to determine the amount of NF-κB p65, p–NF-κB p65, and β-actin. β-actin was used as a loading control. (D) The levels of NF-κB p65 and p–NF-κB p65 in infected or mock-infected IECs were analyzed by FACS (n = 5 or 6). (E) MODE-K cells were infected with S. typhimurium (multiplicity of infection = 10:1), with or without pyrrolidine dithiocarbamate (PDTC), for 6 h. Total RNA was extracted, and the mRNA levels of Mult-1, Rae-1, and H60 were determined by quantitative PCR. Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
TLR9 deficiency augmented the expression of IL-1β and the activation of NF-κB in IECs in response to S. typhimurium infection. IECs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. (A) The mRNA levels of IL-1β and IL-18 in mock-infected or S. typhimurium–infected IECs were determined by quantitative real-time PCR. Results are shown as the fold increase in the expression of IL-1β and IL-18 compared with that of Actb. Gene expression values were then calculated based on the ΔΔCT method, using the mean of results for the mock-infected IECs as a calibrator. Relative quantities (RQs) were determined using the equation RQ = 2−ΔΔCT (n = 5 or 6). (B) The levels of IL-1β and IL-18 in the supernatants of small intestine culture were measured using ELISA (n = 5 or 6). (C) The total proteins were extracted from the infected or mock-infected IECs and subjected to Western blotting analysis to determine the amount of NF-κB p65, p–NF-κB p65, and β-actin. β-actin was used as a loading control. (D) The levels of NF-κB p65 and p–NF-κB p65 in infected or mock-infected IECs were analyzed by FACS (n = 5 or 6). (E) MODE-K cells were infected with S. typhimurium (multiplicity of infection = 10:1), with or without pyrrolidine dithiocarbamate (PDTC), for 6 h. Total RNA was extracted, and the mRNA levels of Mult-1, Rae-1, and H60 were determined by quantitative PCR. Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
The maturation of IL-1β requires the activation of NLRP3 inflammasomes and further cleavage by caspase-1. Therefore, the activation of NLRP3 inflammasomes in IECs was further detected by real-time PCR and Western blotting. The mRNA levels of Nlrp3 were upregulated in WT and TLR9−/− mice after S. typhimurium infection, and no changes were found in the expression of Asc or the gene encoding caspase-1 (Supplemental Fig. 3A). However, the protein levels of NLRP3, procaspase-1, caspase-1, pro–IL-1β, and IL-1β in the IECs of WT and TLR9−/− mice were augmented postinfection and more obvious upregulation was observed in TLR9−/− mice (Fig. 6A, Supplemental Fig. 3B). MODE-K cells were then used to further confirm the cleavage of caspase-1. MODE-K cells were infected with S. typhimurium, with or without the addition of the caspase-1 inhibitor Z-YVAD-FMK or IL-1β. The results show that Z-YVAD-FMK treatment attenuated S. typhimurium infection–induced NKG2D ligand expression significantly, whereas IL-1β stimulation obviously upregulated the expression of NKG2D ligands (Fig. 6B). These data demonstrated that the secretion of IL-1β was caspase-1 dependent during S. typhimurium infection and that IL-1β could stimulate the expression of NKG2D ligands. Furthermore, we determined the cytotoxicity of iIELs against IECs from NLRP3−/− mice to demonstrate a functional link with NLRP3 activation. As shown in Fig. 6C, NLRP3 deficiency significantly attenuated the cytotoxicity of iIELs against infected IECs. These results further confirmed our findings that the activation of NLRP3 inflammasomes results in the secretion of IL-1β, which promotes the recruitment of iIELs into the infected intestine and augments the cytotoxicity of iIELs against infected IECs.
TLR9 deficiency augmented the activation of NLRP3 inflammasomes in IECs in response to S. typhimurium infection. IECs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. (A) The total proteins were extracted from the IECs and subjected to Western blotting analysis to determine the levels of NLRP3, procaspase-1, caspase-1, pro–IL-1β, IL-1β, and β-actin. (B) The levels of NKG2D ligands on MODE-K cells postinfection with S. typhimurium (or in combination with Z-YVAD-FMK) or poststimulation with IL-1β for 24 h were determined by immunofluorescence (original magnification ×200). (C) Total iIELs and IECs were isolated from S. typhimurium–infected WT or NLRP3−/− mice at day 5 postinfection. The cytotoxicity of iIELs against IECs was measured using an LDH assay. (D) TLR9−/− and NLRP3 double-deficient mice (TLR9−/− × NLRP3−/−) were infected with S. typhimurium. At day 5, the changes in intestinal injury were assessed. H&E histology of the small intestines and histological scores are presented (original magnification ×100) (n = 4). (E) Study flowchart. (F) H&E histology of the representative mouse small intestines (original magnification ×100). (G) Histological scores of small intestine sections (n = 4). (H) The levels of IL-1β in the supernatants of small intestine culture were measured using ELISA (n = 4). Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, Student t test.
TLR9 deficiency augmented the activation of NLRP3 inflammasomes in IECs in response to S. typhimurium infection. IECs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. (A) The total proteins were extracted from the IECs and subjected to Western blotting analysis to determine the levels of NLRP3, procaspase-1, caspase-1, pro–IL-1β, IL-1β, and β-actin. (B) The levels of NKG2D ligands on MODE-K cells postinfection with S. typhimurium (or in combination with Z-YVAD-FMK) or poststimulation with IL-1β for 24 h were determined by immunofluorescence (original magnification ×200). (C) Total iIELs and IECs were isolated from S. typhimurium–infected WT or NLRP3−/− mice at day 5 postinfection. The cytotoxicity of iIELs against IECs was measured using an LDH assay. (D) TLR9−/− and NLRP3 double-deficient mice (TLR9−/− × NLRP3−/−) were infected with S. typhimurium. At day 5, the changes in intestinal injury were assessed. H&E histology of the small intestines and histological scores are presented (original magnification ×100) (n = 4). (E) Study flowchart. (F) H&E histology of the representative mouse small intestines (original magnification ×100). (G) Histological scores of small intestine sections (n = 4). (H) The levels of IL-1β in the supernatants of small intestine culture were measured using ELISA (n = 4). Data are representative of three independent experiments (mean ± SD). *p < 0.05, **p < 0.01, Student t test.
We also generated TLR9 and NLRP3 double-deficient mice (TLR9−/− × NLRP3−/−), by crossing TLR9−/− and NLRP3−/− mice, and observed the changes in intestinal injury after S. typhimurium infection. As shown in Fig. 6D, NLRP3 deficiency significantly attenuated TLR9 deficiency–induced intestinal damage, suggesting that NLRP3 deficiency abrogated IL-β production and, thus, reduced TLR9 deficiency–induced severe intestinal damage. To further confirm that the IL-1β production is actually caused by epithelial NLRP3 activation, we established BM chimeras from NLRP3−/− mice. Lethally irradiated WT and NLRP3−/− recipient mice were reconstituted with BM cells from WT or NLRP3−/− mice. Six weeks later, the recipient mice were infected with S. typhimurium, and the inflammatory injury, bacterial burden, and IL-1β production were assessed at 5 d postinfection (Fig. 6E). As shown in Fig. 6F–H, NLRP3 deficiency in BM-derived iIELs did not impair the intestinal injury and IL-1β production, whereas NLRP3−/− recipient mice displayed significantly lower IL-1β production, indicating that NLRP3 deficiency in IECs or other stroma cells leads to lower IL-1β production. However, unexpectedly, the lower production of IL-1β was not accompanied by attenuated intestinal damage, but it resulted in more severe injury and higher bacterial load (Fig. 6F, 6G, Supplemental Fig. 3C). We deduced that there might be other inflammatory factors (such as TNF-α) produced by the activation of the NF-κB pathway (independent of NLRP3 activation) during S. typhimurium infection that would aggravate the intestinal injury. Nonetheless, these results demonstrated that the production of IL-1β is indeed caused by epithelial NLRP3 activation, but it is not derived from BM-derived iIELs. Thus, TLR9 deficiency augments the expression of NKG2D ligands on IECs in an NF-κB–NLRP3–IL-1β–dependent manner.
IL-1β promotes the recruitment of iIELs into the infected intestine and augments the cytotoxicity of iIELs against infected IECs
As described above, the numbers of iIELs, particularly TCRγδ+ iIELs, in the epithelium of the small intestine increased significantly after S. typhimurium infection (Fig. 2A, Supplemental Fig. 1E, 1F). To clarify whether they were attracted into the infected intestine by chemotaxis, we assessed the expression of chemokines, which attract T cells into the intestine in MODE-K cells postinfection with S. typhimurium. The data showed that the expression of MIP-2 and CCL25, but not CCL20, was upregulated after S. typhimurium infection (Supplemental Fig. 4A). We also found increased expression of CCR9, the receptor for CCL25, on iIELs postinfection with S. typhimurium (Supplemental Fig. 4B). This suggested that infected IECs might recruit iIELs into the infected intestine. To further confirm this, we cocultured CFSE-labeled murine iIELs with S. typhimurium–infected MODE-K cells in a Transwell system. By counting the CFSE-labeled iIELs that entered the lower chamber, we observed that increasing numbers of iIELs were recruited by S. typhimurium–infected MODE-K cells (Supplemental Fig. 4C, 4D). These results suggested that infected IECs could indeed recruit iIELs to infected sites by expressing chemokines.
To further confirm whether IL-1β impairs the function of the iIELs that were recruited into infected tissue, we first assessed the expression of the IL-1β receptor (IL-1RI) on iIELs using FACS. The results showed that iIELs express IL-1RI and that S. typhimurium infection upregulated the expression of IL-1RI on iIELs in WT and TLR9−/− mice (Fig. 7A). Then, iIELs isolated from WT mice were stimulated with IL-1β in vitro, which significantly augmented the expression of CD69, NKG2D, and CD107a on TCRγδ+ iIELs (Fig. 7B). To further validate the role of IL-1β on iIELs, anti–IL-1β Ab and a recombinant IL-1R antagonist (rIL-1RA) were used to block the function of IL-1β in vivo when mice were infected with S. typhimurium. Five days postinfection, iIELs were isolated, and the levels of CD69, NKG2D, and CD107a on TCRγδ+ iIELs were determined. The results revealed that the levels of CD69, NKG2D, and CD107a were upregulated significantly postinfection with S. typhimurium. However, the increase was completely eliminated when IL-1β’s function was blocked with anti–IL-1β Ab and IL-1RA (Fig. 7C). As expected, IL-1β stimulation significantly enhanced the cytotoxicity of iIELs to MODE-K cells, whereas anti–IL-1β Ab blockade markedly attenuated the iIEL-mediated cytolysis (Fig. 7D). Importantly, there was less damage to the epithelial barrier in the anti–IL-1β Ab or IL-1RA blockade groups than in the mock-treated infection group (Fig. 7E, 7F). These results indicated that IL-1β could enhance the cytotoxicity of iIELs against infected IECs, leading to severe damage to the epithelial barrier (Fig. 8).
IL-1β promoted the cytotoxicity of iIELs. (A) iIELs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. The expression of IL-1RI on infected or mock-infected iIELs was analyzed by FACS. (B) iIELs were isolated from mock-infected mice and stimulated with IL-1β in vitro. The levels of CD69, NKG2D, and CD107a were analyzed by FACS. (C) iIELs were isolated from mock-infected or Salmonella-infected mice (with or without anti–IL-1β Ab and rIL-1RA) at day 5 postinfection. The levels of CD69, NKG2D, and CD107a were analyzed by FACS. (D) iIELs were pretreated with IL-1β for 24 h, or anti–IL-1β Ab was added to the cytolytic coculture system. The cytotoxicity of iIELs against S. typhimurium–infected MODE-K cells was then measured using an LDH assay at different E:T ratios (n = 3). (E) Mice were infected with S. typhimurium (with or without anti–IL-1β Ab and rIL-1RA). Histological sections of the small intestines were prepared and stained with H&E at day 5 postinfection (original magnification ×100). (F) Histological scores of small intestine sections (n = 3 or 4). Data in (A), (B), and (D) are representative of three independent experiments [mean ± SD in (D)]. Data in (C) and (E) are representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
IL-1β promoted the cytotoxicity of iIELs. (A) iIELs were isolated from mock-infected or Salmonella-infected mice at day 5 postinfection. The expression of IL-1RI on infected or mock-infected iIELs was analyzed by FACS. (B) iIELs were isolated from mock-infected mice and stimulated with IL-1β in vitro. The levels of CD69, NKG2D, and CD107a were analyzed by FACS. (C) iIELs were isolated from mock-infected or Salmonella-infected mice (with or without anti–IL-1β Ab and rIL-1RA) at day 5 postinfection. The levels of CD69, NKG2D, and CD107a were analyzed by FACS. (D) iIELs were pretreated with IL-1β for 24 h, or anti–IL-1β Ab was added to the cytolytic coculture system. The cytotoxicity of iIELs against S. typhimurium–infected MODE-K cells was then measured using an LDH assay at different E:T ratios (n = 3). (E) Mice were infected with S. typhimurium (with or without anti–IL-1β Ab and rIL-1RA). Histological sections of the small intestines were prepared and stained with H&E at day 5 postinfection (original magnification ×100). (F) Histological scores of small intestine sections (n = 3 or 4). Data in (A), (B), and (D) are representative of three independent experiments [mean ± SD in (D)]. Data in (C) and (E) are representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, Student t test.
Schematic illustration of the model of the critical role of TLR9 in S. typhimurium (ST) infection based on our results. TLR9 plays an important role in protection against S. typhimurium infection. After recognizing the unmethylated CpG-containing DNA of S. typhimurium, the apical TLR9 on IECs activates its intracellular signaling, leading to accumulation of IκBα, which blocks NF-κB activation. This negative-regulation mechanism is required to sustain intestinal homeostasis and control of tissue repair. When TLR9 is missing, the negative regulation is interrupted, and the overactivation of the NF-κB signaling pathway in IECs leads to excessive pro–IL-1β expression and increased expression of NKG2D ligands. Meanwhile, TLR9 deficiency caused excessive activation of NLRP3 inflammasomes, leading to the maturation of pro–IL-1β through further cleavage by caspase-1. IL-1β induced the expression of NKG2D on iIELs and NKG2D ligands on IECs, resulting in higher susceptibility of IECs to the cytotoxicity of iIELs through NKG2D’s recognition of its ligands, thus leading to severe destruction of the epithelial barrier and systemic spread of S. typhimurium. In addition, infected IECs could recruit iIELs to infected areas via the interaction between CCL25 and CCR9.
Schematic illustration of the model of the critical role of TLR9 in S. typhimurium (ST) infection based on our results. TLR9 plays an important role in protection against S. typhimurium infection. After recognizing the unmethylated CpG-containing DNA of S. typhimurium, the apical TLR9 on IECs activates its intracellular signaling, leading to accumulation of IκBα, which blocks NF-κB activation. This negative-regulation mechanism is required to sustain intestinal homeostasis and control of tissue repair. When TLR9 is missing, the negative regulation is interrupted, and the overactivation of the NF-κB signaling pathway in IECs leads to excessive pro–IL-1β expression and increased expression of NKG2D ligands. Meanwhile, TLR9 deficiency caused excessive activation of NLRP3 inflammasomes, leading to the maturation of pro–IL-1β through further cleavage by caspase-1. IL-1β induced the expression of NKG2D on iIELs and NKG2D ligands on IECs, resulting in higher susceptibility of IECs to the cytotoxicity of iIELs through NKG2D’s recognition of its ligands, thus leading to severe destruction of the epithelial barrier and systemic spread of S. typhimurium. In addition, infected IECs could recruit iIELs to infected areas via the interaction between CCL25 and CCR9.
Discussion
TLRs act as key sensors for conserved bacterial molecules and play a critical role in preventing host infection by invading pathogens (27). In the intestine, TLRs are expressed by cells of the mucosal epithelium and associated immune cells, just like macrophages and dendritic cells. Increasing evidence shows that, in addition to TLR5, three other TLRs (TLR2, TLR4, and TLR9) participate in the innate mucosal responses to S. typhimurium. However, the precise role of these three TLRs in the recognition and defense of S. typhimurium infection remains undetermined. In the current study, we found that TLR9-deficient mice were more susceptible to S. typhimurium infection than were WT or TLR2- or TLR4-deficient mice, as indicated by the highest bacterial load and the most severe intestinal damage. We found that the deficiency in TLR9 in IECs resulted in overactivation of NF-κB and NLRP3 inflammasome signaling pathways, leading to excessive IL-1β secretion. IL-1β increased the expression of NKG2D ligands on IECs, which resulted in the higher susceptibility of IECs to the cytotoxicity of iIELs, which caused serious destruction of the epithelial barrier, and, finally, systemic dissemination of S. typhimurium. In addition, infected IECs could recruit iIELs to infected intestine via secretion of chemokines. IL-1β in the infected sites further stimulated the activation of iIELs and promoted the expression of NKG2D on iIELs. These findings demonstrated the negative-regulatory effect of TLR9 in IECs on the NF-κB–NLRP3–IL-1β pathway and the critical role of TLR9 in the protection of intestinal integrity and in the defense against S. typhimurium infection (Fig. 8).
In addition to TLR5, TLR2-, TLR4-, and TLR9-mediated innate immune responses to S. typhimurium have received increasing attention. Several studies have shown, using TLR-KO mice, that TLR4, TLR2, TLR9, and MyD88 are involved in murine host defense against Salmonella in vivo (5, 28, 29). Weiss et al. (29) demonstrated that TLR2−/− mice infected orally with S. typhimurium had slightly lower or similar bacterial loads compared with WT mice, in the MLN and spleen, which is consistent with our results. Later, Arpaia et al. (5) found that, although TLR2 and TLR4 double-deficient mice were highly susceptible to S. typhimurium infection, the TLR2-KO × TLR4-KO × TLR9-KO mice were less susceptible to infection than TLR2-KO × TLR4-KO mice. They further confirmed that TLR signaling is required for S. typhimurium virulence and intracellular growth. They deduced that mice lacking sufficient TLR signaling are less susceptible to S. typhimurium infection because of reduced bacterial growth; however, they did not determine which TLR signaling is more important for this effect. Our results showed that TLR2−/− mice appear to be less susceptible to S. typhimurium infection, as indicated by the reduced intestinal injury and lower bacterial loads compared with WT mice after S. typhimurium infection (Fig. 1). Thus, whether TLR2 signaling is required for S. typhimurium intracellular growth requires further investigation. Indeed, our previous research found that TLR2 and TLR9 play opposing roles in macrophage-mediated innate immunity against S. typhimurium infection (14). We proposed that TLR2 signaling might promote an inflammatory reaction to S. typhimurium infection by suppressing the production of TGF-β or other anti-inflammatory factors. However, whether the role of TLR2 in IECs has a similar effect needs to be clarified further. With regard to the role of TLR4, Weiss et al. (29) showed that TLR4-deficient macrophages were more susceptible to S. typhimurium; the bacterial burden of TLR4−/− mice was 10- and 100-fold higher than that in WT mice. Further investigation demonstrated that TLR4 is critical for early cytokine production and killing of bacteria by murine macrophages, whereas TLR2 is required for the late stages of the macrophage response. However, our experiments showed that there were no significant differences in survival or bacterial load between TLR4−/− and WT groups (14), which is inconsistent with other reports. We speculated that these data might have resulted from different infection doses of S. typhimurium: Weiss et al. (29) showed that, at a low infectious dose, WT, TLR4−/−, and TLR2−/− mice exhibited equal susceptibility to Salmonella infection. Collectively, TLRs might play different roles in different models (inflammation, damage or infection; or different pathogens), as well as different pathological regions (upper or lower intestine). We showed that TLR9−/− mice exhibited significantly shorter survival (14) and higher bacterial loads compared with WT mice in response to S. typhimurium infection (Fig. 1C, 1D). Moreover, severe intestinal damage and increased dissemination to the MLN, spleen, and liver in TLR9−/− mice were noted (Figs. 1A, 1B, 4H). These results demonstrated the critical role of TLR9 in the defense against S. typhimurium infection, which is in line with other reports (5, 28, 30).
IECs represent the first line of defense against potentially harmful bacteria present in the lumen. They discriminate commensal and pathogenic bacteria and play a crucial role in regulating immune responses and maintaining normal homeostasis via innate immune receptors (31). iIELs are a population of T lymphocytes that reside within the epithelium of the intestine and form one of the main branches of the immune system. They are located between IECs and combine with IECs to form the front line of immune defense against invading pathogens. Through killing or eradication of infected or malignant epithelial cells, iIELs play important roles in immunosurveillance of the intestinal mucosa (16, 32). In the current study, we found that TLR9 deficiency aggravated the NKG2D-dependent cytolytic capacity of iIELs against infected IECs (Fig. 3E) and intestinal injury (Figs. 1B, 3G). In vivo, NKG2D blockade significantly attenuated intestinal damage in WT and TLR9−/− mice (Fig. 3G). These results indicated that the killing of infected IECs by iIELs indeed leads to severe intestinal damage that is not just derived from S. typhimurium infection–induced inflammation. Through adoptively transferring iIELs isolated from WT or TLR9−/− mice into Rag1−/− mice, we found that TLR9 deficiency in iIELs did not influence the susceptibility and intestinal injury of mice in response to S. typhimurium infection (Fig. 4A–D). The BM chimeras showed that TLR9 deficiency in IECs (or other stromal cells) exacerbated the intestinal injury, whereas TLR9 deficiency in BM-derived iIELs did not influence the susceptibility and intestinal injury of mice in response to S. typhimurium infection (Fig. 4E–H). These results confirmed that intrinsic TLR9 deficiency in IECs, but not in iIELs, leads to the greater susceptibility of IECs to the cytotoxicity of iIELs and, finally, severe intestinal damage.
Although studies of the roles of TLR pathways have mostly focused on their proinflammatory and immunostimulatory properties in APCs and immune cells, their role in IECs has attracted great interest, especially with regard to regulating tissue homeostasis and immune tolerance. TLR9 is usually expressed in intracellular endosomal organelles (33). However, TLR9 is expressed on the cell surface and in endosomes of IECs (20, 31, 34). Interestingly, the activation of TLR9 leads to different effects based on its expression on the apical or basolateral membranes of IECs. Stimulation of basolateral TLR9 induces NF-κB activation and a proinflammatory response, whereas signaling through surface or endosomal TLR9 at the apical pole of IECs leads to accumulation of IκBα in the cytoplasm, which blocks NF-κB activation (20, 31). Therefore, apical TLR9 stimulation seems to contribute to maintain intestinal homeostasis and prevent intestinal injury. These unique roles of apical TLR9 have been confirmed by its involvement in maintaining colonic homeostasis and protection against colonic inflammation by preventing epithelial barrier dysfunction; therefore, it plays an important role in regulating intestinal tolerance and inflammation (20, 31, 34–36). In this study, in the S. typhimurium infection model, we found that TLR9 deficiency promotes the infection-induced inflammatory reaction in IECs. We propose that TLR9 deficiency interrupts apical TLR9-sustained intestinal homeostasis by relieving the suppression of NF-κB signaling, leading to excessive inflammatory responses and intestinal injury. In addition, NF-κB has been reported to upregulate the expression of the NKG2D ligand MICA on activated T cells and HeLa tumor cells (37). In this study, we found that overactivation of NF-κB signaling augmented the expression of murine NKG2D ligands (Fig. 5E), which further led to higher susceptibility of IECs to the cytotoxicity of iIELs and severe damage to the epithelial barrier.
IL-1β, as a proinflammatory cytokine, promotes the activation and differentiation of lymphocytes, regulates the infiltration of inflammatory cells, induces chemotaxis and activation of other inflammatory factors, regulates the function of epithelial cells, and might bring about tissue damage by mediating inflammation (38). Our study found that IL-1β induced the expression of NKG2D and promoted the activation and cytotoxicity of iIELs (Fig. 7). IL-1β also upregulated the expression of NKG2D ligands on IECs (Fig. 6B). We confirmed that IL-1β promoted the NKG2D-mediated cytotoxicity of iIELs against IECs. We concluded that TLR9 deficiency induced excessive secretion of IL-1β during S. typhimurium infection, and the secreted IL-1β further promoted the cytotoxicity of iIELs against infected-IECs, leading to severe damage to the epithelial barrier.
Activation of the NF-κB signaling pathway only induced production of pro–IL-1β, and the maturation of IL-1β requires the activation of NLRP3 inflammasomes and further cleavage by caspase-1 (39–41). TLR-induced NF-κB activation promotes NLRP3 expression through binding to the NLRP3 promoter in murine macrophages (42). We found excessive activation of NF-κB, NLRP3 inflammasomes, and IL-1β secretion in IECs postinfection with S. typhimurium in TLR9-deficient mice (Fig. 6). We then generated TLR9−/− × NLRP3−/− mice and an NLRP3−/− BM chimera model, which confirmed that NLRP3 deficiency in IECs or other stroma cells, but not BM-derived iIELs, leads to lower IL-1β production. This suggests that TLR9 deficiency in IECs causes loss of control of the NF-κB signaling pathway, leading to excess activation of NLRP3 inflammasomes and IL-β production, the latter of which further promotes the activation and cytolysis of iIELs against infected IECs, ultimately resulting in severe intestinal injury. Unexpectedly, however, the lower production of IL-1β in NLRP3−/−-recipient BM chimeras was not accompanied by attenuated intestinal damage but resulted in more severe injury and a higher bacterial load (Fig. 6F, 6G). We deduced that there might be other inflammatory factors (such as TNF-α) produced by activation of the NF-κB pathway during S. typhimurium infection that aggravated the intestinal injury. In addition, the decreased IL-1β production might weaken the antibacterial activity, thus promoting the spread of bacteria. In addition, inflammasome activation–induced epithelial pyroptosis might also contribute to the intestinal damage.
Taken together, the current study demonstrated the critical role of TLR9 in protecting against S. typhimurium infection and in maintaining intestinal homeostasis. TLR9 expression on the surface or endosomally at the apical pole of IECs plays important roles in the maintenance of intestinal homeostasis. Disruption of the TLR9 signaling balance will result in exaggerated IL-1β secretion, severe intestinal injury, and finally, systemic spread of S. typhimurium. Our results provide important insights into the mechanisms of TLRs in sustaining intestinal homeostasis and defending against pathogens and provide novel insights for potential therapeutics that target infectious and inflammatory diseases of the intestine.
Footnotes
This work was supported by the National 973 Basic Research Program of China (Grant 2013CB944901) and the National Natural Science Foundation of China (Grants 91442114 and 81472646).
The online version of this article contains supplemental material.
References
Disclosures
The authors have no financial conflicts of interest.