These authors contributed equally to this work and should be considered as co-first author.
Critical role of Toll-like receptor 2 in Bacteroides fragilis-mediated immune responses in murine peritoneal mesothelial cells
Article first published online: 23 OCT 2012
© 2012 The Societies and Wiley Publishing Asia Pty Ltd
Microbiology and Immunology
Volume 56, Issue 11, pages 782–788, November 2012
How to Cite
Kim, T.-H., Lee, K.-B., Kang, M.-J. and Park, J.-H. (2012), Critical role of Toll-like receptor 2 in Bacteroides fragilis-mediated immune responses in murine peritoneal mesothelial cells. Microbiology and Immunology, 56: 782–788. doi: 10.1111/j.1348-0421.2012.00505.x
- Issue published online: 23 OCT 2012
- Article first published online: 23 OCT 2012
- Accepted manuscript online: 31 AUG 2012 09:16AM EST
- Received 9 April 2012; revised 6 August 2012; accepted 23 August 2012.
- Bacteroides fragilis;
- immune responses;
- peritoneal mesothelial cell;
- Toll-like receptor 2
In this study, the role of Toll-like receptor 2 (TLR2) in immune responses of murine peritoneal mesothelial cells against Bacteroides fragilis was investigated. Enzyme linked immunosorbent assay was used to measure cytokines and chemokines. Activation of nuclear factor κB (NF-κB-α) and mitogen-activated protein kinases (MAP kinases) was investigated by western blot analysis. B. fragilis induced production of interleukin-6, chemokine (C-X-C motif) ligand 1 (CXCL1) and chemokine (C-C motif) ligand 2 (CCL2) in wild type peritoneal mesothelial cells; this was impaired in TLR2-deficient cells. In addition, in response to B. fragilis, phosphorylation of inhibitory NF-κB-α and c-Jun N-terminal kinase mitogen-activated protein kinase (MAPK) was induced in wild type mesothelial cells, but not in TLR2-deficient cells,. Inhibitor assay revealed that NF-κB and MAPKs are essential for B. fragilis-induced production of CXCL1 and CCL2 in mesothelial cells. These findings suggest that TLR2 mediates immune responses in peritoneal mesothelial cells in response to B. fragilis.
- B. fragilis
bone marrow-derived dendritic cell
chemokine (C-C motif) ligand
cluster of differentiation
chemokine (C-X-C motif) ligand
- E. coli
extracellular signal-regulated kinases
c-Jun N-terminal kinase
- MAP kinase
mitogen-activated protein kinase
multiplicity of infection
myeloid differentiating factor 88
nuclear factor κB
nucleotide-binding oligomerization domain
- P. gingivalis
sodium dodecyl sulfate
tumor necrosis factor
Mesothelial cells are flattened, squamous-like cells derived from mesoderm during development. They compose the mesothelium, which covers all the surfaces of serosal cavities and the organs in those cavities (1). The mesothelium provides a protective barrier against physical damage and a frictionless interface between adjacent organs. Mesothelial cells also have a role in immune response such as cytokine production and T cell activation against invading pathogens (2–3).
Toll-like receptors are one of the type 1 transmembrane proteins with intracellular Toll-interleukin 1 receptor domains, transmembrane domains, and ectodomains containing leucine-riche repeats; they are responsible for the recognition of ligands (4). TLRs play important roles in host innate immune responses by sensing pathogen associated molecular patterns of microorganisms, including bacteria, fungi and viruses (2). TLR activation leads to recruitment of adaptor molecules such as MyD88, TIR domain-containing adapter protein, TIR-domain-containing adapter-inducing interferon-β and TIR-domain-containing adapter-inducing interferon-β related adaptor molecule. This recruitment triggers the cascade of signaling pathway and activation of NF-κB, interferon-regulatory factors, and MAP kinases such as p38, JNKs and ERK1/2, which then induces transcription of inflammatory cytokines, type I interferons and chemokines (5).
Mesothelial cells have various functional TLRs. The genes of TLR1–6 are expressed constitutively in both human and murine peritoneal mesothelial cells (2, 6). Synthetic lipid A directly induces gene expression of CCL2 (monocyte chemoattractant protein-1) and CXCL2 (MIP-2) and activation of MAP kinases (p38, ERK, and JNK) in murine peritoneal mesothelial cells via a TLR4-dependent pathway (6). In addition, various TLR agonists including lipoteichoic acid (TLR2), LPS (TLR4), polyinosinic:polycytidylic acid (TLR3), and flagellin (TLR5) can induce production of CXCL1 and CCL2 in murine peritoneal mesothelial cells (7). These findings indicate that TLRs may act as immune sentinels in peritoneal mesothelial cells.
Bacteroides fragilis, a Gram-negative, anaerobic bacterium, is part of normal intestinal flora (8). This bacterium is one of the major causes of peritonitis induced by intestinal microflora gaining access to the peritoneal cavity (9–10). When the intestinal contents gain access to the peritoneal cavity because of infarction, obstruction, or direct trauma, aerobic bacteria such as E. coli are the predominant cause of infection. Once the amount of oxygen decreases significantly, anaerobic B. fragilis begins to predominate in the peritoneal infection (9). Although there is still controversy concerning this, B. fragilis seems to activate the TLR2 pathway. A symbiosis factor PSA of B. fragilis directs regulatory T cells to promote immunologic tolerance through the TLR2 pathway (11). In addition, a transfection assay showed that purified B. fragilis LPS extracted by various methods promote luciferase activity driven by NF-κB-regulated gene promoter through TLR2, but not through TLR4 (12). Peritoneal mesothelial cells secret cytokines and chemokines and induce gene expression of adhesion molecules in response to various stimuli, including bacteria (13–15). However, the role of TLR2 in the B. fragilis-induced immune response in peritoneal mesothelial cells remains unclear. In this study, we examined whether TLR2 is required for production of cytokines and chemokines and activation of NF-κB and MAP kinases in murine peritoneal mesothelial cells in response to B. fragilis.
MATERIALS AND METHODS
Preparation of murine peritoneal mesothelial cells
Wild type and TLR2-deficient mice on a C57BL/6 background (Jackson Laboratory, Bar Harbor, ME, USA) were used for isolation of peritoneal mesothelial cells. Mesothelial cells were prepared from the peritoneum and external surfaces of liver, spleen, and kidneys of adult mice as previously described (16). Briefly, samples of peritoneum and intact organs were obtained from killed mice and digested with 0.25%-trypsin-EDTA (Invitrogen, Grand Island, NY, USA) solution for 50 min at 37°C. Intact tissues and tissue debris were discarded and the cell suspension centrifuged at 90 g for 5 min. The pellet was resuspended in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 15% heat-inactivated FBS (Invitrogen) and antibiotics and cultured overnight. The next day, floating cells were removed by washing twice with PBS and adherent cells cultured for five additional days. Mesothelial cells were used between passages 2 and 4.
Reagents and bacterial culture
Ultrapure LPS from E. coli O111:B4 and Pam3CSK4 were purchased from InvivoGen (San Diego, CA, USA). For inhibitor assays, PD98059 (an ERK inhibitor) and SB203580 (a p38 inhibitor) were obtained from Selleck Chemicals (Houston, TX, USA) and Bay 11–7082 (an NF-κB inhibitor) and SP60012 (a JNK inhibitor) from Calbiochem (La Jolla, CA, USA). B. fragilis (ATCC 25285) was grown in brain heart infusion broth supplemented with hemin and vitamin K1 under anaerobic conditions at 37°C. Bacterial cultures were pelleted by centrifugation at 890 g for 20 mins, washed, and resuspended in PBS. The bacterial concentrations of the suspensions were adjusted to 1 × 109 CFU/mL. The bacteria were diluted to the desired concentration and used in subsequent experiments.
Measurement of cytokines and chemokines
The concentrations of IL-6, CXCL1, and CCL2 in culture supernatants was determined using a commercial ELISA kit (R & D Systems, Minneapolis, MN, USA).
The cells were lysed in buffer containing 1% Nonidet-P40 supplemented with complete protease inhibitor ‘cocktail’ (Roche, Mannheim, Germany) and 2 mM dithiothreitol. Lysates were separated by 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes by electro-blotting. The membranes were immunoblotted with primary antibodies such as regular- and phospho-IκB-α, p38, ERK, and JNK (Cell Signaling Technology, Beverly, MA, USA). After immunoblotting with secondary antibodies, proteins were detected with enhanced chemiluminescence reagent (Intron Biotechnology, Seong-Nam, Korea).
Mesothelial cells were pretreated with various doses of each inhibitor 2 hrs before B. fragilis infection. After washing twice with PBS, the cells were infected with B. fragilis at MOI 1/50 and cotreated with and without inhibitors. At 24 hrs after infection, the culture supernatants were collected and the concentrations of CXCL1 and CCL2 measured by ELISA.
The differences among the mean values of the different groups were tested and the values expressed as the mean ± SD. All of the statistical calculations were performed by one-way analysis of variance followed by the Bonferroni post-hoc test for multigroup comparisons using GraphPad Prism version 5.01 (GraphPad Software, San Diego, California, USA). Values of P < 0.05 were considered significant.
Interleukin-6 and chemokine (C-C motif) ligand 2 production in wild type and Toll-like receptor 2-deficient peritoneal mesothelial cells in response to Toll-like receptor ligands
To confirm the phenotype of TLR2-deficient mesothelial cells, we examined production of IL-6 and CCL2 in culture supernatants from WT and TLR2-deficient mesothelial cells in response to TLR ligands. Pam3CSK4 (a synthetic TLR1/2 ligand) increased production of IL-6 and CCL2 in WT mesothelial cells; this was abolished in TLR2-deficient cells (Fig. 1a, b). However, as compared with WT mesothelial cells, LPS (a TLR4 ligand) produced comparable amounts of IL-6 and CCL2 in TLR2-deficient cells (Fig. 1a, b), suggesting that TLR2 is a critical factor for bacterial lipoprotein-induced production of cytokines and chemokines in peritoneal mesothelial cells.
Production of interleukin-6, chemokine (C-X-C motif) ligand 1 and chemokine (C-C motif) ligand 2 in Bacteroides fragilis-infected peritoneal mesothelial cells
We next examined production of IL-6, CXCL1, and CCL2 induced by B. fragilis in peritoneal mesothelial cells. At 4 hrs after infection, B. fragilis had increased IL-6 production in both WT and TLR2-deficient mesothelial cells, but the amount was slightly lower in TLR2-deficient than WT cells (Fig. 2a). CXCL1 and CCL2 were also upregulated in WT cells in response to B. fragilis; this was impaired in TLR2-deficient cells (Fig. 2b, c). At 24 hrs after infection, the amounts of IL-6, CXCL1, and CCL2 upregulated by B. fragilis were significantly greater in WT mesothelial cells (only at MOI 1/10 and 1/50 for IL-6 and CCL2) than in TLR2-deficient cells (Fig. 2d–f). When we infected WT cells with MOI 1/100 of B. fragilis, production of IL-6 and CCL2 was slightly decreased compared with infection with MOI 1/50. However, their production by B. fragilis was increased in TLR2-deficient cells in a dose-dependent manner, which resulted in no significant difference between WT and TLR2-deficient cells in B. fragilis-induced IL-6 and CCL2 production at MOI 1/100 (Fig. 2d, f). These findings suggest that TLR2 signaling is involved in B. fragilis-induced production of cytokines and chemokines in peritoneal mesothelial cells.
Activation of nuclear factor-κB and mitogen-activated protein kinase by Bacteroides fragilis in wild type and Toll-like receptor 2-deficient peritoneal mesothelial cells
Nuclear factor F-κB and MAP kinases are critical factors for induction of immune responses in various cell types. Therefore, we used western blot analysis to determine whether B. fragilis induces activation of NF-κB and MAP kinases and whether TLR2 is required for these events. Phosphorylation of IκB-α and JNK MAP kinase was increased in B. fragilis-infected WT mesothelial cells; this was abolished in TLR2-deficient cells (Fig. 3). There was no significant difference in phosphorylation of p38 and ERK MAP kinases between WT and TLR2-deficient cells in response to B. fragilis (Fig. 3).
Role of nuclear factor-κB and mitogen-activated protein kinases on Bacteroides fragilis-induced production of chemokine (C-X-C motif) ligand 1 and chemokine (C-C motif) ligand 2 in peritoneal mesothelial cells
To determine whether NF-κB and MAPKs are critical for B. fragilis-induced production of chemokines, we performed an inhibitor assay. Production of CXCL1 and CCL2 by B. fragilis was completely abolished in peritoneal mesothelial cells treated with BAY11–7082 (an NF-κB inhibitor) at doses of 10 and 20 μM (Fig. 4a, b). Inhibitors for p38 (SB203580) and JNK (SP600125) also reduced B. fragilis-induced production of CXCL1 and CCL2 in mesothelial cells in a dose-dependent manner (Fig. 4c–f). ERK inhibitor PD98059 also decreased the production of chemokines, but the effect was weaker than that of other inhibitors (Fig. 4g, h).
There is controversy concerning TLR activity of B. fragilis LPS. Mancuso et al. reported that purified B. fragilis-derived LPS leads to IL-8 production in HEK 293 cells cotransfected with TLR4/CD14/MD2, but not in cells with TLR2/CD14 (8). In addition, B. fragilis LPS reportedly fails to produce TNF-α in macrophages from C3H/HeJ mice, which have a point mutation in TLR4 (Pro712His) and have been widely used as a LPS nonresponsive mouse strain. It also fails to induce lethality in these mice (8), suggesting that B. fragilis LPS induces production of cytokines and chemokine and lethal toxicity via TLR4 signaling. In contrast, B. fragilis LPS induces CXCL2 (MIP-2) production in 10ScNCr/23 cells, a cell line derived from C57BL/10ScNCr mouse strain which has a deletion of the TLR4 gene (17), whereas LPS from E. coli or Salmonella does not (18), indicating that B. fragilis LPS induces TLR4-independent signaling. This is supported by a study by Hasegawa et al. showing that B. fragilis whole cell extract and culture supernatant has no TLR4 activity (19). Although differences in purification or test methods and contamination can be considerable, the precise reasons for this discrepancy remain unknown.
In this study, we found that TLR2 is involved in B. fragilis-mediated immune responses in peritoneal mesothelial cells. TLR2 is required for B. fragilis-induced production of IL-6, CXCL1, and CCL2 and activation of NF-κB and JNK MAP kinase in peritoneal mesothelial cells. B. fragilis PSA and LPS are considered the major factors that express TLR2 activity. Wang et al. showed that B. fragilis PSA induces production of TNF-α and nitric oxide in RAW macrophages and that this is restored by anti-TLR2 antibody treatment (20). In addition, B. fragilis PSA leads to production of cytokines (TNF-α and IL-12p40) and enhances expression of surface molecules such as CD86 and major histocompatibility complex II in WT BMDCs; this is impaired in TLR2-deficient BMDCs (20). These findings indicate that bacterial carbohydrate can play an important role in innate and adaptive immune responses via a TLR2-dependent pathway.
It has long been known that enterobacterial LPS utilizes a signaling complex consisting of CD14, TLR4, and co-receptor MD-2 to activate downstream signals (21). However, several studies have demonstrated that LPS of non-enterobacteria, including P. gingivalis and Leptospira interrogans, can signal independently of TLR4, instead using TLR2-mediated signaling (22–23). B. fragilis lipid A is known to be structurally similar to P. gingivalis lipid A (24–25). Both have monophosphorylated disaccharide backbones and relatively long fatty acids. These structural characteristics could explain the weaker activity of their LPS relative to enterobacterial LPS. In fact, LPS from B. fragilis and P. gingivalis is less active in producing TNF-α in human monocytes than is E. coli LPS (26).
In one study, the authors also clarified that recognition of B. fragilis LPS absolutely depends on TLR2 (26).
In our study, IL-6 and CCL2 production by B. fragilis, although delayed, was still greater in TLR2-deficient-mesothelial cells. In addition, B. fragilis induced phosphorylation of p38 and ERK MAP kinases in TLR2-deficient-mesothelial cells. Moreover, a previous study has shown that intraperitoneal infection with B. fragilis (108 CFU) does not induce kill WT and MyD88-deficient mice, although TLR2 and MyD88 are required for bacteria-induced IL-10 production in peritoneal exudate cells (27). These findings suggest that other factors beside TLR2 may participate in B. fragilis-induced immune responses in peritoneal mesothelial cells. Nod1 is very likely a candidate for TLR2-independent immune responses by B. fragilis in mesothelial cells. Nod1 is a cytosolic receptor that recognizes D-glutamyl-meso-diaminophimelic acid derived from bacterial peptidoglycan (28). This structure exists in all Gram negative bacteria and some Gram positive bacteria, such as Listeria and Bacillus (28). A recent study revealed that Nod1 mRNA is expressed in murine peritoneal mesothelial cells and that Nod1 agonist KF1B induces substantial production of chemokines, such as CXCL1 and CCL2, and activation of NF-κB and MAP kinases in these cells (7). Because B. fragilis is a Gram negative bacterium, this bacterium presumably possess Nod1 activity. Therefore, whether Nod1 is involved in B. fragilis-induced immune responses in peritoneal mesothelial cells requires clarification.
It seems that TLR2-mediated immune responses against B. fragilis have both beneficial and harmful effects on the host. In a mouse model of intra-abdominal abscess formation induced by B. fragilis infection, TLR2 deficiency led to reduced abscess formation (20). In contrast, TLR2 is required for persistent intestinal colonization of B. fragilis. B. fragilis PSA signals through TLR2 on CD4+ T cells to induce immune tolerance by suppressing Th17 cell responses during intestinal colonization of B. fragilis (11).
In conclusion, we demonstrate here that TLR2 mediates immune responses in peritoneal mesothelial cells in response to B. fragilis. Because B. fragilis does induce IL-6 and CCL2 production and some MAP kinases activation in a TLR2-independent manner, further study is necessary to determine what other factors are involved and to clarify the precise mechanism for B. fragilis-induced immune responses in peritoneal mesothelial cells.
This work was supported by Konyang University Myung-Gok Research Fund (Grant No. 2010–04).
The authors have no financial conflict of interest.