TLRs trigger immunity by detecting microbe-associated molecular patterns (MAMPs). Flagellin is a unique MAMP because it harbors 1) an antigenic hypervariable region and 2) a conserved domain involved in TLR5-dependent systemic and mucosal proinflammatory and adjuvant activities. In this study, the contribution of the flagellin domains in TLR5 activation was investigated. We showed that TLR5 signaling can be neutralized in vivo by flagellin-specific Abs, which target the conserved domain. However, deletions of flagellin’s hypervariable region abrogated the protein’s intrinsic ability to trigger the production of neutralizing Abs. The fact that MAMP-specific Abs block TLR-mediated responses shows that this type of neutralization is a novel mechanism for down-regulating innate immunity. The stimulation of mucosal innate immunity and adjuvancy to foreign Ag was not altered by the hypervariable domain deletions. In contrast, this domain is essential to trigger systemic innate immunity, suggesting that there are distinct mechanisms for TLR5 activation in systemic and mucosal compartments. In summary, specific MAMP determinants control the production of neutralizing Abs and the compartmentalization of innate responses.

Toll-like receptors are instrumental in the coordinated induction of innate and adaptive immunity in mammals (1, 2). Because TLRs are expressed by a broad variety of cell types, they are able to trigger immunity throughout the body. Following infection by pathogenic microorganisms, TLRs recognize conserved motifs referred to as microbe-associated molecular patterns (MAMPs)3 (1). TLR engagement induces a gene expression program dedicated to both innate clearance of and acquired immunity to pathogenic microorganisms (2). For instance, TLRs induce the production of chemokines that, in turn, specifically attract the polymorphonuclear neutrophils (PMNs) directly involved in innate microbial clearance. Furthermore, TLRs promote the secretion of pleiotropic immune mediators (such as TNF-α) and the functional maturation of dendritic cells (DCs) that specialize in Ag presentation to lymphocytes. Consequently, TLR agonists not only stimulate “broadly specific” proinflammatory immune responses but also enhance the adaptive immune response to defined Ags and are, thus, considered to be adjuvants (3). Despite these potentially beneficial effects, the systemic toxicity of MAMPs has prompted efforts to develop derivatives that bias MAMP activity toward adjuvancy (4). Indeed, engineering molecules with unique properties is a major challenge in manipulating immune responses.

Bacterial flagellins (the major flagella components in many bacterial pathogens) are specific, unique agonists for TLR5 activation (5, 6). The FliC flagellin from Salmonella enterica serovar Typhimurium is the paradigm for studies on flagellum structure-function, immunity, and TLR5 signaling (6, 7, 8, 9). It is a 494-aa protein with two distinct domains. The amino- and carboxyl-terminal conserved regions (comprising ∼170 and 90 aa, respectively) form a domain that is essential for TLR5 activation. Furthermore, the motif 89–96 is absolutely required (8, 9, 10). The middle (outer) domain of flagellin FliC comprises amino acids from positions 170–400 and is not mandatory for TLR5 signaling (9, 10). It is designated as a hypervariable region, since the primary sequences greatly vary in composition and size from one bacterial species to another. In contrast, it is known that the hypervariable region is essential for flagellin antigenicity (11, 12). In fact, flagellins’ hypervariable regions carry H-Ag specificity used for serotyping enteropthopathogenic bacteria, especially Salmonella strains. Deletion of the central domain decreases the flagellins’ antigenicity. For example, the S. typhimurium flagellin FliCΔ204–292, which is truncated of 99 aa positioned in the middle part of FliC, is poorly recognized by Abs directed against the whole flagellin (11). The question of whether or not the antigenic and TLR5-activating domains can be functionally dissociated had not been addressed until now.

TLR5 agonists are potent activators of systemic and mucosal innate responses. We and others have shown that i.v. injection of flagellins promotes a systemic response, characterized by the production of proinflammatory mediators (such as TNF-α or IL-6) and DC activation (5, 6, 13, 14, 15, 16, 17). Furthermore, flagellins trigger mucosa-specific innate and adaptive defense mechanisms (18, 19, 20). For instance, epithelial cell lines and lung mucosa up-regulate the production of chemokines like CXCL8 (IL-8) and CCL20 which, in turn, recruit mucosal PMNs and DCs, respectively (18, 19). Various authors have reported that flagellins are potent systemic and mucosal adjuvants that elicit 1) serum and/or secretory Ab responses and 2) Th1 and Th2 cell responses to both the flagellins themselves and coadministered Ags (14, 17, 21, 22).

The goal of the present study was to determine whether or not MAMP domains could alter the outcome of innate and adaptive immune responses. Since MAMPs have adjuvant activity, they also promote intrinsic anti-MAMP responses which, in turn, can neutralize proinflammatory and adjuvant properties. We addressed this issue by using flagellin as a model. We found that 1) anti-flagellin Abs neutralize TLR5 signaling, 2) deletion of the hypervariable part of flagellin abrogates the latter’s ability to induce neutralizing Abs but does not significantly alter proinflammatory and adjuvant activities, and 3) systemic detection of flagellin does not involve the same molecular determinants as mucosal detection.

The recombinant flagellins originated from the Salmonella enterica serovar Typhimurium ATCC14028 flagellin FliC (accession number AAL20871). The flagellins FliC and FliCΔ205–293 were either isolated from the S. typhimurium strains SIN22 (fljB) and SJW46, as described previously (11, 15, 19), or purchased from Alexis Biochemicals. The constructs encoding FliCΔ174–400 and FliCΔ191–352 were generated by PCR on a pBR322-derived plasmid harboring the wild-type fliC gene under the control of its own promoter and using the following primer pairs: AGCACCattcagcgtatccagacc/GCTGGTgctacaaccaccgaaaac and TCGAGatatcctgtaacagttgcagcc/ACTCGAGgacggtacatccaaaactgcac (bases encoding a linker are in italics). Site-directed mutagenesis was also performed on the plasmid harboring FliCΔ174–400 to replace the residues 89–96 (QRVRELAV) involved in TLR5 detection by the corresponding sequences from a nonsignaling flagellin (DTVKVKAT); the resulting protein was thus FliCΔ174–400/89–96* (8). In FliCΔ174–400, FliCΔ191–352, and FliCΔ174–400/89–96*, the asparagine located six residues from the end was changed into a serine. The truncated flagellins were purified from the supernatant of recombinant S. typhimurium SIN41 (fliC fljB) as follows. Salmonella were grown in Luria-Bertani broth for 18 h at 37°C with agitation. The supernatant was filtered and saturated with 60% ammonium sulfate (Sigma-Aldrich). The precipitated materials were recovered by centrifugation, solubilization in 20 mM Tris-HCl (pH 7.5), and then dialysis. The proteins were further purified by successive rounds of hydroxyapatite and anion exchange chromatography (Bio-Rad). Lastly, the proteins were depleted of LPS using a polymyxin B column (Pierce). Using the Limulus assay (Associates of Cape Cod), the residual LPS concentration was determined to be <30 pg of LPS/μg of recombinant flagellin. When specified, flagellins were treated for 1 h at 37° with 0.017% trypsin-EDTA (Invitrogen) to totally hydrolyze the proteins, followed by heating at 70°C for 1 h to inactivate the trypsin. Proteins were analyzed using standard SDS-PAGE and immunoblotting with FliC-specific polyclonal Abs.

Female NMRI mice (6–8 wk old) were purchased from Charles River Laboratories and maintained in a specific pathogen-free facility in an accredited establishment (no. A59107; Institut Pasteur de Lille, Lille, France). All experiments complied with current national and institutional regulations and ethical guidelines. For hyperimmunization, animals were injected s.c. with the flagellin FliC (1 μg/injection) emulsified in 200 μl of CFA/PBS on day 1 and IFA/PBS on days 21, 35, and 49. On day 63, mice were given 200 μl of flagellin/PBS i.v. and were sacrificed 2 h later by i.p. injection of 5 mg of sodium pentobarbital (CEVA Santé Animale) for serum and tissue sampling and analysis.

To characterize the mucosal innate response and adjuvant properties, 20 μl of PBS with or without proteins were administered intranasally (i.n.) to mice anesthetized i.p. with 1.5 mg of ketamine (Merial) and 0.3 mg of xylazine (Bayer) per 25-g animal. To study proinflammatory responses, mice were sampled either at 2 h (for RNA and gene expression assays) or 6 h (to test cytokine production). For immunization assays, mice were administered i.n. with PBS with or without LPS-depleted OVA (20 μg, grade VII; Sigma-Aldrich) with or without flagellins (1 μg) on days 1 and 21. Bronchoalveolar lavages (BALs) and serum were sampled on day 35.

To assess neutralization, immune and mock sera were heated for 30 min at 56°C to inactivate complement. Serial serum dilutions (in 200 μl of PBS) were passively transferred to animals by the i.v. route 1 h before systemic activation with flagellins. In some experiments, sera were mixed with flagellins diluted in PBS and administered i.n. to test mucosal neutralization.

BALs were collected after the intratracheal injection of 1 ml of PBS with Complete Protease Inhibitor Cocktail (Roche) and clarified by centrifugation. Blood samples were collected and clotted at room temperature, with the serum then being separated by centrifugation. Lung protein extracts were prepared by homogenizing tissue with 2 ml of T-PER Tissue Protein Extraction Reagent (Pierce) supplemented with protease inhibitors. All samples were stored at −80°C before analysis.

Levels of OVA- or flagellin- specific Abs in serum and BAL samples were assessed using ELISAs. Briefly, OVA (20 μg/well in phosphate buffer 0.2 M (pH 6.5)) and flagellin FliC (100 ng/well in PBS) were coated on MaxiSorp microplates (Nalge Nunc) overnight at 4°C. All microplates were washed with PBS/0.05% Tween 20 and then blocked with PBS/1% dry milk for 1 h at room temperature. Serial dilutions of samples were incubated for 1 h at room temperature before development. Biotinylated anti-mouse IgG or IgA Abs (Southern Biotechnology Associates), HRP-conjugated streptavidin (GE Healthcare), and 3,3′,5,5′- tetramethylbenzidine (BD Biosciences) were used as development reagents. The reaction was stopped by addition of H2SO4 and the OD at 450 nm was determined. The IgG titer was defined as the reciprocal of the highest sample dilution yielding an absorbance value of 0.15 OD for OVA and 0.5 OD for FliC and was systematically compared with a reference serum. Titers are given as geometrical means of titers from individual mice. Total IgA and OVA-specific IgA levels in BALs were measured and normalized using a calibration curve with commercial mouse IgA (Sigma-Aldrich). The specific IgA ratio (expressed in nanograms of OVA-specific IgA per microgram of total IgA) was determined for each mouse.

Mouse CXCL2 and CCL20 and human IL-8 (CXCL8) levels were measured in serum, BALs, total lung and/or cell culture supernatant using commercial ELISA kits (R&D Systems).

Total RNA from mouse lungs was extracted with a Nucleospin RNA II Kit (Macherey Nagel) and reverse-transcribed with the High-Capacity cDNA Archive Kit (Applied Biosystems). The resulting cDNA was amplified using SYBR Green-based real-time PCR (Applied Biosystems). The specific primers were CGTCATCCATGGCGAACTG/GCTTCTTTGCAGCTCCTTCGT (ACTB, coding for β-actin), TTTTGGGATGGAATTGGACAC/TGCAGGTGAAGCCTTCAACC (CCL20), and CCCTCAACGGAAGAACCAAA/CACATCAGGTACGATCCAGGC (CXCL2). Relative mRNA levels (2 − ΔΔCt) were determined by comparing 1) the PCR cycle thresholds (Ct) for the gene of interest and ACTB (ΔCt) and 2) ΔCt values for treated and control groups (ΔΔCt) as described previously (19).

The Caco-2 human colon adenocarcinoma cell line was stably transfected with the plasmid harboring a luciferase gene under the control of the human CCL20 promoter (23), giving rise to the Caco-Rumbo line. These intestinal epithelial cells were grown in DMEM supplemented with 10% FCS, 10 mM HEPES, 1× nonessential amino acids, 100 U/ml penicillin, and 100 U/ml streptomycin and (for transgene selection) 0.7 mg/ml G418 (Invitrogen). The human bronchial epithelial cell line BEAS-2B was cultured in Kaigh’s F12 nutrient medium supplemented as for Caco-Rumbo medium plus 1 mM sodium pyruvate and insulin-transferrin-selenium mix (Invitrogen).

Cells were stimulated with recombinant flagellins for 6 h for luciferase assays or for 16 h before harvesting the supernatant for ELISA. Luciferase activity in cell extracts was measured using the Bright Glo Luciferase Assay (Promega). Relative luminescence (RLU) was normalized as a percentage of the maximum activity with wild-type flagellin for the activation test with the recombinant flagellins. For the in vitro neutralization test, the RLU was normalized as a percentage of the maximum activity for each protein: [(RLUtreated/RLUuntreated)/(RLUmax/RLUuntreated)] × 100.

Statistical differences were analyzed using the Mann-Whitney U test and were considered to be significant for p < 0.05. Unless otherwise specified, results are expressed as arithmetic means ± SD.

Bacterial flagellins are known to elicit strong Ab responses, which are mainly directed against the hypervariable region (11, 12, 24). We hypothesized that anti-flagellin Abs would neutralize the flagellins’ TLR5-stimulating activity. Hence, mice were immunized s.c. with the flagellin FliC or a mock preparation (PBS alone or the irrelevant Ag OVA formulated in CFA), followed by boosts with IFA. ELISA analysis revealed that the anti-FliC sera exhibited specific IgG titers (>106), whereas mock sera titers were below the assay’s detection threshold (102).

We and others have previously used human intestinal epithelial cell lines as unique reporters of flagellin/TLR5 stimulatory activity, based on expression of the chemokine CCL20 (also known as “liver-activated and -regulated chemokine,” LARC) (19, 23, 25). Using Caco-Rumbo cells harboring the luciferase gene under the control of the CCL20 promoter (23), we demonstrated that an anti-FliC serum is able to fully neutralize FliC’s TLR5 agonist activity (Fig. 1,A). The neutralizing effect of FliC-specific Abs on TLR5 signaling was then directly assessed in immunized animals. To this end, systemic proinflammatory responses in mice (production of CCL20 and CXCL2 chemokines) were studied after i.v. injection of FliC (Fig. 1, B and C). In mock-immunized animals, a FliC challenge triggered a significant increase in serum levels of CCL20 and CXCL2, compared with a PBS challenge. In contrast, chemokine production in FliC-immunized animals was not enhanced by any of the challenges. Using passive serum transfer in naive animals, a close correlation was found between the amount of Ab injected and the systemic innate response, as shown in Fig. 1 D. In conclusion, preexisting immunity to flagellin can neutralize the latter’s TLR5-stimulating activity, both in vitro and in vivo.

FIGURE 1.

Neutralization of TLR5 signaling by flagellin-specific Abs. NMRI mice were immunized s.c. at week 1 with 1 μg of flagellin FliC and CFA, followed by boosts at weeks 3, 5, and 7 with FliC and IFA. In mock conditions, animals were similarly treated with OVA and adjuvants or adjuvants alone. Experiments were conducted at week 9. A, In vitro TLR5-neutralizing activity of flagellin-specific immune serum. Caco-Rumbo epithelial cells harboring the reporter construct CCL20-luc were activated for 6 h with the flagellin FliC incubated with 50% v/v FliC hyperimmune (○) or mock (•) sera. Luciferase activity was determined and normalized to the activity obtained with 100 ng/ml FliC. Results are representative of one of three independent experiments. B and C, In vivo TLR5-neutralizing activity of flagellin-specific immune serum. Immunized animals (n = 3) were injected i.v. with PBS (▪) or 0.1 μg (▦) or 1 μg of flagellin FliC (□). Sera were collected 2 h later and the concentrations of CCL20 (B) and CXCL2 (C) were determined by ELISA. D, The neutralizing activity of immune serum. Animals (n = 3 per dose) were passively transferred i.v. with various amounts of flagellin-specific or mock serum and treated 1 h later i.v. with recombinant flagellins, as indicated. Chemokine production in serum 2 h postchallenge was measured by ELISA. Statistical significance (p > 0.05) was determined using the Mann-Whitney U test.

FIGURE 1.

Neutralization of TLR5 signaling by flagellin-specific Abs. NMRI mice were immunized s.c. at week 1 with 1 μg of flagellin FliC and CFA, followed by boosts at weeks 3, 5, and 7 with FliC and IFA. In mock conditions, animals were similarly treated with OVA and adjuvants or adjuvants alone. Experiments were conducted at week 9. A, In vitro TLR5-neutralizing activity of flagellin-specific immune serum. Caco-Rumbo epithelial cells harboring the reporter construct CCL20-luc were activated for 6 h with the flagellin FliC incubated with 50% v/v FliC hyperimmune (○) or mock (•) sera. Luciferase activity was determined and normalized to the activity obtained with 100 ng/ml FliC. Results are representative of one of three independent experiments. B and C, In vivo TLR5-neutralizing activity of flagellin-specific immune serum. Immunized animals (n = 3) were injected i.v. with PBS (▪) or 0.1 μg (▦) or 1 μg of flagellin FliC (□). Sera were collected 2 h later and the concentrations of CCL20 (B) and CXCL2 (C) were determined by ELISA. D, The neutralizing activity of immune serum. Animals (n = 3 per dose) were passively transferred i.v. with various amounts of flagellin-specific or mock serum and treated 1 h later i.v. with recombinant flagellins, as indicated. Chemokine production in serum 2 h postchallenge was measured by ELISA. Statistical significance (p > 0.05) was determined using the Mann-Whitney U test.

Close modal

Since flagellin’s antigenic domain (i.e., the hypervariable central region of the molecule) is not mandatory for TLR5 signaling, we sought to engineer flagellin variants which could not be neutralized by Abs (11, 24). These recombinant molecules were designed on the basis of flagellin’s three-dimensional structure and reported immunological properties (7, 8, 9). Two novel flagellin molecules (FliCΔ191–352 and FliCΔ174–400, composed of 336 and 271 aa, respectively) were constructed by internal deletion (Fig. 2,A). As a control, we used the previously characterized variant FliCΔ204–292, which has a partial deletion in the antigenic domain (11) (Fig. 2,A). As a negative control for in vitro and in vivo experiments, mutations that impair TLR5 signaling were introduced into FliCΔ174–400, yielding the recombinant protein FliCΔ174–400/89–96*. The predicted structures of the respective flagellins indicated that the motif 89–96 and the overall structure of the conserved regions were unchanged (Fig. 2 A). With the exception of FliCΔ204–292, the variants were unable to complement the motility of flagellin-deficient bacteria and were secreted into the culture supernatant.

FIGURE 2.

Characteristics and cross-reactivity of hypervariable region-deleted flagellins. A, A schematic three-dimensional view of the recombinant flagellins. The structure of wild-type flagellin FliC is presented in the left-hand panel using Pymol (http://www.pymol.org). In the monomer, terminal regions (1–170 and 400–494) are tightly folded in α helixes and form a structural domain involved in flagellum function. The motif 89–96 (black) is essential for TLR5 signaling. The FliC “hypervariable” domain is mainly constituted of β structures and β turns. Using Swiss-Model (http://www.expasy.org/spdbv/), an overall structure was predicted for FliCΔ204–292 and FliCΔ174–400, showing partial and total deletion of the hypervariable region, respectively. For FliCΔ191–352, the positions of amino acids delineating the deletion are shown in the left-hand panel. FliCΔ174–400 and FliCΔ191–352 contain GAAG and LELE linkers at the deletion junction, respectively. B and C, Cross-reactivity of FliC-specific sera. Hyperimmune sera were obtained after s.c. administration of flagellin formulated with CFA for priming, followed by IFA boosts. Serum was titrated in ELISAs for FliC, FliCΔ204–292, FliCΔ191–352, and FliCΔ174–400. The results are representative of two experiments. B, Cross-reactivity of anti-FliC serum. C, Cross-reactivity of anti- FliCΔ174–400 serum. Statistical significance (p > 0.05 in a Mann-Whitney U test) is indicated by an asterisk.

FIGURE 2.

Characteristics and cross-reactivity of hypervariable region-deleted flagellins. A, A schematic three-dimensional view of the recombinant flagellins. The structure of wild-type flagellin FliC is presented in the left-hand panel using Pymol (http://www.pymol.org). In the monomer, terminal regions (1–170 and 400–494) are tightly folded in α helixes and form a structural domain involved in flagellum function. The motif 89–96 (black) is essential for TLR5 signaling. The FliC “hypervariable” domain is mainly constituted of β structures and β turns. Using Swiss-Model (http://www.expasy.org/spdbv/), an overall structure was predicted for FliCΔ204–292 and FliCΔ174–400, showing partial and total deletion of the hypervariable region, respectively. For FliCΔ191–352, the positions of amino acids delineating the deletion are shown in the left-hand panel. FliCΔ174–400 and FliCΔ191–352 contain GAAG and LELE linkers at the deletion junction, respectively. B and C, Cross-reactivity of FliC-specific sera. Hyperimmune sera were obtained after s.c. administration of flagellin formulated with CFA for priming, followed by IFA boosts. Serum was titrated in ELISAs for FliC, FliCΔ204–292, FliCΔ191–352, and FliCΔ174–400. The results are representative of two experiments. B, Cross-reactivity of anti-FliC serum. C, Cross-reactivity of anti- FliCΔ174–400 serum. Statistical significance (p > 0.05 in a Mann-Whitney U test) is indicated by an asterisk.

Close modal

Next, we assessed the intrinsic antigenicity of the recombinant flagellins. To this end, saturating concentrations of flagellins were coated onto microplates and probed by ELISA, using a hyperimmune serum specific for FliC or FliCΔ174–400. As illustrated in Fig. 2 B, we observed 3- to 10-fold lower Ab titers when anti-FliC serum was titrated against FliC variants than against wild-type FliC. In contrast, the reactivity of hyperimmune serum specific for FliCΔ174–400 was similar, whatever the target flagellin. These results suggest that the central hypervariable region is the major target for anti-flagellin Abs.

Lastly, we sought to establish whether or not the recombinant molecules retained any TLR5-stimulating activity. A dose-response analysis was performed using Caco-Rumbo reporter cells and the lung epithelial cell line BEAS-2B. Activation was assessed by measuring luciferase activity in Caco-Rumbo cells and IL-8 secretion by BEAS-2B cells. As shown in Fig. 3, A and B, FliCΔ204–292, FliCΔ191–352 and FliCΔ174–400 were all potent cell activators. The flagellins’ respective EC50 values varied slightly with the cell type but fell within the previously described nanogram per milliliter range (9). The recombinant flagellins’ activity was found to be fully dependent on TLR5, since FliCΔ174–400/89–96* was unable to activate epithelial cells. The requirement for TLR5 signaling was further confirmed by using bone marrow macrophages derived from TLR5-deficient mice. The cells did not synthesize any detectable IL-12 p40 (<18 pg/ml) upon stimulation with 0.5 μg/ml recombinant flagellins in contrast to cells derived from wild-type C57BL/6 animals (ranging from 500.7 to 709.5 pg/ml IL-12p40).

FIGURE 3.

Epithelial and mucosal proinflammatory activity of hypervariable region-deleted flagellins. A and B, Activation of epithelial cells by recombinant flagellins. Human epithelial cells were activated with flagellins FliC, FliCΔ204–292, FliCΔ191–352, FliCΔ174–400, or FliCΔ174–400/89–96* at the indicated concentrations. Caco-Rumbo cells harboring the reporter fusion CCL20-luc were activated for 6 h and luciferase activity was normalized to the maximal activity measured with saturating FliC levels (A). BEAS-2B bronchial epithelial cells were stimulated for 16 h before measuring IL-8 levels in the supernatant (B). Results are representative of one of two independent experiments. C and D, Stimulation of the mucosal innate response by deleted flagellins. Recombinant flagellins or trypsin-treated preparations (1-μg equivalent) were administrated i.n. to anesthetized mice (n = 3–5). CCL20-specific mRNA levels in the whole lungs were determined 2 h later using real-time quantitative RT-PCR (C). Six hours after instillation, BALs (▪) and lungs (□) were sampled to measure the CCL20 concentration (D). Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

FIGURE 3.

Epithelial and mucosal proinflammatory activity of hypervariable region-deleted flagellins. A and B, Activation of epithelial cells by recombinant flagellins. Human epithelial cells were activated with flagellins FliC, FliCΔ204–292, FliCΔ191–352, FliCΔ174–400, or FliCΔ174–400/89–96* at the indicated concentrations. Caco-Rumbo cells harboring the reporter fusion CCL20-luc were activated for 6 h and luciferase activity was normalized to the maximal activity measured with saturating FliC levels (A). BEAS-2B bronchial epithelial cells were stimulated for 16 h before measuring IL-8 levels in the supernatant (B). Results are representative of one of two independent experiments. C and D, Stimulation of the mucosal innate response by deleted flagellins. Recombinant flagellins or trypsin-treated preparations (1-μg equivalent) were administrated i.n. to anesthetized mice (n = 3–5). CCL20-specific mRNA levels in the whole lungs were determined 2 h later using real-time quantitative RT-PCR (C). Six hours after instillation, BALs (▪) and lungs (□) were sampled to measure the CCL20 concentration (D). Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

Close modal

TLR5 stimulation by recombinant flagellins was then studied in vivo by the mucosal route. To this end, CCL20 and CXCL2 expression in the lungs of mice treated i.n. with flagellins was quantified using quantitative RT-PCR (Fig. 3,C). Within 2 h, CCL20 mRNA pulmonary levels were ∼30-fold higher in animals treated with wild-type or recombinant flagellins than in mock-treated animals. Furthermore, CCL20 chemokine production was detected at 6 h after instillation, both in lung homogenates and BALs (Fig. 3 D). In control experiments, FliCΔ174–400/89–96* and trypsin-digested flagellins did not induce this type of effect. Similar findings were observed for CXCL2 (data not shown). These results confirmed that the in vivo proinflammatory response was exclusively due to the recombinant flagellins. Overall, flagellins with deletions in the hypervariable region displayed mucosal proinflammatory properties equivalent to those of the wild-type FliC counterpart.

Intranasal administration of flagellins is known to promote mucosal adaptive immunity (14, 22, 26). To characterize the adjuvant properties of our recombinant molecules, Ab responses in serum and secretions were studied after i.n. immunizations. OVA was used as a model Ag, formulated with or without the various flagellins or with cholera toxin (CT) as a gold standard mucosal adjuvant. The coadministration of FliC with OVA significantly increased the OVA-specific IgG response (both in serum and the BAL, ∼300- and 100-fold, respectively), compared with animals immunized with OVA alone (Fig. 4, A and B). Moreover, the OVA-specific IgA response was enhanced in BAL, thereby suggesting that FliC promotes the archetypal secretory Ab response of a mucosal adjuvant (Fig. 4,C). Interestingly, FliC’s effect was similar to that of CT. Like FliC, the recombinant flagellins FliCΔ204–292, FliCΔ191–352, and FliCΔ174–400 were thus able to potentate systemic and mucosal responses. In contrast, FliCΔ174–400/89–96* and trypsin-treated flagellins lacked potency (Fig. 4 and Table I). Hence, the deletion of flagellin’s hypervariable region did not significantly influence the TLR5-mediated mucosal adjuvant properties. Our data also showed that the recombinant molecules’ respective effects on innate and adaptive immunity are correlated.

FIGURE 4.

Adjuvant effect of flagellins with hypervariable region deletion. Mice (n = 8) were immunized i.n. with OVA with or without flagellins or CT on days 1 and 21. On day 35, OVA-specific IgG titers were measured in the serum (A) and BALs (B). The concentration of OVA-specific IgA in BALs was determined (C). Results are representative of one of two independent experiments. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

FIGURE 4.

Adjuvant effect of flagellins with hypervariable region deletion. Mice (n = 8) were immunized i.n. with OVA with or without flagellins or CT on days 1 and 21. On day 35, OVA-specific IgG titers were measured in the serum (A) and BALs (B). The concentration of OVA-specific IgA in BALs was determined (C). Results are representative of one of two independent experiments. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

Close modal
Table I.

Protease-sensitive immune responses induced by recombinant flagellins

Intranasal ImmunizationaAnti-OVA IgGbAnti-FliC IgGb
SerumBALSerumBAL
MeanSDMeanSDMeanSDMeanSD
PBS ND 0.0 ND 0.0 ND 0.0 ND 0.0 
OVA 2.4 0.8 1.1 0.7 ND 0.0 ND 0.0 
FliC + OVA 5.7 0.1 3.9 0.5 5.9 0.6 3.0 0.7 
FliCΔ204–292 + OVA 5.5 0.9 3.4 0.8 3.3 0.8 1.0 0.5 
FliCΔ191–352 + OVA 4.5 1.3 2.9 0.9 2.2 0.3 ND 0.0 
FliCΔ174–400 + OVA 4.9 0.9 2.7 0.9 2.0 0.1 0.1 0.2 
TRP + OVA 3.4 1.2 1.3 0.8 ND 0.0 ND 0.0 
FliC/TRP + OVA 2.8 0.5 1.1 0.6 ND 0.0 0.3 0.3 
FliCΔ204–292/TRP + OVA 3.0 1.4 0.8 1.0 ND 0.0 ND 0.0 
FliCΔ191–352/TRP + OVA 2.6 0.7 0.8 0.9 ND 0.0 ND 0.0 
FliCΔ174–400/TRP + OVA 2.6 0.4 1.1 0.9 ND 0.0 ND 0.0 
Intranasal ImmunizationaAnti-OVA IgGbAnti-FliC IgGb
SerumBALSerumBAL
MeanSDMeanSDMeanSDMeanSD
PBS ND 0.0 ND 0.0 ND 0.0 ND 0.0 
OVA 2.4 0.8 1.1 0.7 ND 0.0 ND 0.0 
FliC + OVA 5.7 0.1 3.9 0.5 5.9 0.6 3.0 0.7 
FliCΔ204–292 + OVA 5.5 0.9 3.4 0.8 3.3 0.8 1.0 0.5 
FliCΔ191–352 + OVA 4.5 1.3 2.9 0.9 2.2 0.3 ND 0.0 
FliCΔ174–400 + OVA 4.9 0.9 2.7 0.9 2.0 0.1 0.1 0.2 
TRP + OVA 3.4 1.2 1.3 0.8 ND 0.0 ND 0.0 
FliC/TRP + OVA 2.8 0.5 1.1 0.6 ND 0.0 0.3 0.3 
FliCΔ204–292/TRP + OVA 3.0 1.4 0.8 1.0 ND 0.0 ND 0.0 
FliCΔ191–352/TRP + OVA 2.6 0.7 0.8 0.9 ND 0.0 ND 0.0 
FliCΔ174–400/TRP + OVA 2.6 0.4 1.1 0.9 ND 0.0 ND 0.0 
a

Mice (n = 8) were immunized i.n. with PBS, OVA, OVA plus flagellins, or OVA plus trypsin-treated flagellins (TRP) on days 1 and 21. On day 35, OVA- and FliC- specific IgG titers were measured in the serum and BALs. Statistical significance (p > 0.05) was determined in a Mann-Whitney test.

b

Values are expressed as log 10 (reciprocal titers) ± SD. In serum and BAL, limit of detection is 2 and 0.3 (1/100 serum dilution and 1/2 BAL dilution), respectively. ND, Not detected.

Deletion of the antigenic domain is expected to decrease the flagellin-specific immune response and thereby any neutralization of TLR5-mediated immunity, especially with repeated administration. We therefore decided to assess the efficacy of i.n. immunization with respect to the induction of FliC-specific Abs. As expected, FliC elicited a strong IgG response in serum and BALs (Fig. 5). In contrast, FliCΔ204–292 triggered 10-fold lower Ab levels in both fluids than did FliC, and a more pronounced effect was observed after immunization with FliCΔ191–352 and FliCΔ174–400 to reach nondetectable levels as for flagellin treated with protease (Table I). In conclusion, the flagellins’ antigenic and immunostimulatory domains are functionally uncoupled. Therefore, FliCΔ191–352 and FliCΔ174–400 are molecules of interest for preventing or attenuating the generation of flagellin-specific Abs with neutralizing activity.

FIGURE 5.

Intrinsic antigenic properties of flagellins lacking a hypervariable region. Mice (n = 8) were immunized i.n. with OVA with or without flagellins or CT or LPS on days 1 and 21. On day 35, FliC-specific IgG titers were measured in the serum (A) and BALs (B). Results are representative of one of two independent experiments. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

FIGURE 5.

Intrinsic antigenic properties of flagellins lacking a hypervariable region. Mice (n = 8) were immunized i.n. with OVA with or without flagellins or CT or LPS on days 1 and 21. On day 35, FliC-specific IgG titers were measured in the serum (A) and BALs (B). Results are representative of one of two independent experiments. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

Close modal

To determine whether or not FliCΔ174–400 can escape Ab-mediated neutralization, we performed experiments in vitro and in vivo. First, we found that an anti-FliC hyperimmune serum is able to neutralize to a similar extent the stimulation of epithelial cells by the TLR5 agonists FliC and FliCΔ174–400 (Fig. 6, A and B), indicating that deletion of the hypervariable domain does not impair neutralization. We further investigated the capacity of neutralization in vivo. The effective doses needed to initiate TLR5-mediated innate responses by the i.n. route was determined. FliC and FliCΔ174–400 displayed similar dose-response profiles and the 0.1-μg dose was selected for subsequent neutralization assays (Fig. 6, C and D). To this end, animals were hyperimmunized i.n. with FliC to elicit strong, FliC-specific mucosal IgG responses (mean titer ∼45,000) and then challenged i.n. with 0.1 μg of FliC or FliCΔ174–400 flagellins. Proinflammatory chemokine production in BALs was monitored. Challenge with FliC or FliCΔ174–400 led to CCL20 production (4.28 ± 1.98 vs 1.08 ± 0.54 ng/ml and 2.48 ± 1.22 vs 0.93 ± 0.48 ng/ml in mock- and FliC-immunized mice, respectively) as observed in naive animals (Fig. 6, C and D). Assuming that anti-FliC ELISA titers correlate to neutralizing titers (Fig. 1,D), we predicted that the neutralizing activity in BALs was too low to promote blockade of TLR5 signaling in the lungs, whatever the type of flagellin. To definitely define whether neutralization can operate in the lungs, flagellins were incubated with anti-FliC sera before administration (Fig. 6, E and F). Both TLR5 agonistic activities of FliC and FliCΔ174–400 were blocked in these conditions. In conclusion, deletion of the flagellin hypervariable region did not enable the resulting molecules to escape from Ab neutralization of TLR5 immune responses. These data also indicated that the FliC neutralization epitopes are located within the conserved sequences 1–173 and 401–494.

FIGURE 6.

Neutralization of TLR5 signaling induced by hypervariable region-deleted flagellin FliCΔ174–400. A and B, Epithelial neutralization of TLR5 signaling by flagellin-specific immune serum. Caco-Rumbo epithelial cells harboring the reporter construct CCL20-luc were activated for 6 h with 10 ng/ml FliC (A) or FliCΔ174–400 (B) incubated with various dilutions of FliC hyperimmune (○) or mock (•) sera. Luciferase activity was determined and normalized to the activity obtained with 10 ng/ml FliC or FliCΔ174–400 in the absence of serum. Results are representative of one of two independent experiments. C and D, Intranasal dose-response activity of flagellins. Various amounts of flagellin FliC (▪) or FliCΔ174–400 (□) were administrated i.n. The concentrations of CCL20 (C) and CXCL2 (D) were determined 6 h later in BALs using an ELISA. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test. E and F, In vivo neutralization of TLR5 activity stimulated by FliCΔ174–400. Animals (n = 3 per dose) were instilled i.n. with 50 ng of FliC (E) or FliCΔ174–400 (F) supplemented with various quantities of anti-FliC or mock sera as indicated. Chemokine production in BAL was assessed by ELISA 6 h postchallenge. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

FIGURE 6.

Neutralization of TLR5 signaling induced by hypervariable region-deleted flagellin FliCΔ174–400. A and B, Epithelial neutralization of TLR5 signaling by flagellin-specific immune serum. Caco-Rumbo epithelial cells harboring the reporter construct CCL20-luc were activated for 6 h with 10 ng/ml FliC (A) or FliCΔ174–400 (B) incubated with various dilutions of FliC hyperimmune (○) or mock (•) sera. Luciferase activity was determined and normalized to the activity obtained with 10 ng/ml FliC or FliCΔ174–400 in the absence of serum. Results are representative of one of two independent experiments. C and D, Intranasal dose-response activity of flagellins. Various amounts of flagellin FliC (▪) or FliCΔ174–400 (□) were administrated i.n. The concentrations of CCL20 (C) and CXCL2 (D) were determined 6 h later in BALs using an ELISA. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test. E and F, In vivo neutralization of TLR5 activity stimulated by FliCΔ174–400. Animals (n = 3 per dose) were instilled i.n. with 50 ng of FliC (E) or FliCΔ174–400 (F) supplemented with various quantities of anti-FliC or mock sera as indicated. Chemokine production in BAL was assessed by ELISA 6 h postchallenge. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

Close modal

We also wanted to study the neutralization by flagellin-specific Abs of TLR5-dependent responses induced after i.v. injection of the recombinant flagellins. To analyze the systemic activation of innate immunity, the production in circulating proinflammatory chemokines CCL20 and CXCL2 was measured by ELISA in serum (Fig. 7). Unexpectedly, we observed that FliCΔ174–400 was ∼100-fold impaired in its ability to trigger systemic proinflammatory effects, compared with the wild-type FliC. Whereas 10 μg of FliCΔ174–400 stimulated a slight chemokine production, the variant mutated within the TLR5 motif FliCΔ174–400/89–96* was devoid of activity (0.85 ± 0.27 vs 0.02 ± 0.00 ng/ml for CCL20). This contrasted with FliCΔ204–292 and FliCΔ191–352, which were both potent activators like FliC (data not shown). Hence, certain molecular determinants on the hypervariable region (or dependent on the latter) are required for systemic TLR5 stimulation but not mucosal TLR5 stimulation. Taken as a whole, our results indicate that TLR5 activation within the mucosal and the systemic compartments is controlled by distinct mechanisms.

FIGURE 7.

Alteration of the systemic activation ability of hypervariable region-deleted flagellin FliCΔ174–400. Various amounts of flagellin FliC (▪) or FliCΔ174–400 (□) were administrated i.v. The concentrations of CCL20 (A) and CXCL2 (B) were determined 2 h later in the serum using an ELISA. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

FIGURE 7.

Alteration of the systemic activation ability of hypervariable region-deleted flagellin FliCΔ174–400. Various amounts of flagellin FliC (▪) or FliCΔ174–400 (□) were administrated i.v. The concentrations of CCL20 (A) and CXCL2 (B) were determined 2 h later in the serum using an ELISA. Statistical significance (p > 0.05) was determined in a Mann-Whitney U test.

Close modal

Over recent years, flagellins have been the focus of many studies on the role of TLR5 in systemic and mucosal immunity (6). In addition to TLR5-dependent stimulatory activity, flagellins display strong antigenic potency. Thus, flagellins are immunodominant Ags in the body’s responses to pathogenic bacteria and in chronic inflammatory disorders, since they elicit prominent T cell and Ab responses (6, 27, 28, 29). This dual nature as an innate immunity activator and an Ag means that flagellins are attractive immunological models. It is known that flagellins are constituted by a conserved domain (with a TLR5-activating motif) and a hypervariable region (assigned to antigenicity, especially Ab response) (9, 11, 12). In this study, we sought to establish whether TLR5 signaling can be regulated by flagellin-specific Abs and whether TLR5-stimulating and antigenic activities are linked and affect each other. Using truncated forms of the S. typhimurium flagellin FliC and anti-FliC hyperimmune serum, we showed for the first time that preexisting flagellin-specific Abs are capable of neutralizing TLR5 signaling effects in vivo. Additionally, we demonstrated that deletion of flagellin’s hypervariable region promotes escape from neutralization by decreasing the protein’s potency for generating antagonistic Abs. These data support that the flagellin TLR5-stimulating and antigenic domains can be dissociated but that their respective activities can affect the final outcome of immune responses. Lastly, we found that TLR5 signaling is compartmentalized, since the FliCΔ174–400 flagellin (i.e., lacking the hypervariable region) stimulated immunity in the mucosa but was devoid of any systemic activity.

Whereas innate TLR signaling clearly orchestrates adaptive immunity, the reverse process has been little explored. Although recent evidence supports an Ag-independent role for T lymphocytes in the regulation of innate immunity (30), the question of how B cells and, in particular, TLR agonist-specific Abs influence innate responses has not been resolved. Our study indicates that animals can develop flagellin-specific Abs that efficiently neutralize the onset of TLR5-mediated responses in vitro and in vivo (Figs. 1 and 7). Therefore, in vivo blockade of TLR signaling by MAMP-specific neutralization Abs is a novel mechanism for down-regulating innate immunity. Our results are consistent with the report by Saha et al. (31), which suggested that ex vivo Ab blockade of the TLR5 activation motif of Pseudomonas aeruginosa flagellin reduces its efficacy to induce lung innate responses. Interestingly, we found that neutralizing epitopes in S. typhimurium FliC are embedded within the conserved region (regions 1–173 and 401–494) that also carries the activation motif 89–96 (8, 9). We did not perform any cross-neutralization experiments using flagellins isolated from other bacteria: however, if neutralizing Abs indeed target the conserved TLR5 signaling motif with high affinity, one can expect blockade of the innate responses to any flagellins. We further demonstrated that the mechanisms of action of flagellin-specific Abs rely on immediate neutralization of TLR5 signaling, since NF-κB-dependent chemokine gene transcription was not turned on early after challenge (Fig. 3). It is known that LPS can be sequestered by secretory IgA within endosomes in intestinal epithelial cells, thereby blocking TLR4-mediated NF-κB activation (32). Hence, one can assume that high-affinity flagellin-specific neutralizing Abs bind to the flagellin signaling motif and thereby prevent any interaction between flagellin and its cognate detector TLR5. Further investigations with flagellin-specific mAbs are needed to dissect the mode of action and the Abs targets.

TLR signaling neutralization is a major strategy for managing uncontrolled inflammation in sepsis or chronic disease (33). Different targets can promote a TLR signaling blockade, including TLRs themselves and cognate MAMPs. Most efforts seek to block the TLRs and a recent study showed that an anti-TLR4 mAb efficiently inhibited LPS-induced immune responses in acute polymicrobial infections (34). As shown in the present work, MAMP targeting represents an effective and attractive neutralization strategy. Chronic TLR stimulation may contribute to the physiopathology of some diseases and, when conjugated with the intrinsic adjuvant and antigenic activities of MAMPs, it may elicit anti-MAMP-neutralizing Abs. In turn, the neutralization of TLR-specific responses could fully suppress innate responses. Flagellin’s hypervariable region is not essential for signaling (Figs. 3 and 4), a finding which is consistent with the known TLR5 stimulatory activity of Listeria monocytogenes flagellin, which almost completely lacks a variable region (5). Thus, flagellated bacteria could evade host defenses by facilitating the production of Abs that reduce the host’s ability to mount an innate immune response. The high antigenicity of the flagellin variable domain may be critical in the potentiation of this type of Ab production. Remarkably, the study by Honko et al. (22) showed that i.n. administered anti-flagellin Abs were unable to interfere with TLR5. Accordingly, we were unable to detect neutralization in similar conditions.

In contrast to profilin (that depends on TLR11 for effective antigenicity (35)), flagellin’s reduced immunogenicity following hypervariable region deletion does not rely on a TLR5 signaling failure. The deletion of major Th or B epitopes may explain the decreased ability of FliCΔ174–400 to elicit Abs. Previous studies identified a dominant Th epitope in flagellin’s conserved domain and highlighted a major role for CD4 T cells in Ab responses (29, 36). Deletion of flagellin’s hypervariable region is therefore not absolutely required for effective help for B cell responses. Likewise, i.n. coadministration of FliCΔ174–400 with OVA (which provides external help) did not enhance anti-FliC Ab titers, compared with instillation of FliCΔ174–400 alone. It seems that deletion of a dominant B cell epitope on the hypervariable region is essential for presentation of a subdominant, neutralizing epitope located within the conserved region.

We previously suggested that early epithelial CCL20 production correlates with mucosal adjuvant properties, probably through DC recruitment within the mucosa (19). Our study supports this paradigm, since the mucosal adjuvancy of flagellin molecules correlates with early CCL20 production in lung tissue and BALs (Figs. 3, 6, and 7). TLR5 signaling is absolutely required for flagellin-induced enhancement of immune responses to coinstilled Ags. The deleted flagellin FliCΔ174–400 also retains adjuvant properties when administered s.c, suggesting that mucosal and dermal responses behave similarly and thus differ from the systemic response (data not shown). Studies with the S. typhimurium flagellin FljB (which harbors a similar deletion in the hypervariable region) have been recently performed with a view to using this type of molecule as an Ag carrier (14, 21). Similar findings were obtained using the deleted flagellin FljB as a carrier for foreign Ags in s.c. immunization (21).

The use of wild-type flagellin as an adjuvant could lead to harmful effects because production of neutralizing Abs may attenuate both the booster effect of the adjuvant and the innate responses to pathogenic flagellated bacteria. We established that FliCΔ174–400 has more prominent beneficial properties, due to its poor capacity to generate neutralizing Abs. In addition, we found that FliCΔ174–400 is strongly attenuated for systemic signaling compared with wild-type flagellin, whereas mucosal activity was unaffected. This type of effect has been already observed for the TLR4 agonist LPS (4). Recent observations indicated that LPS’s molecular features are essential for its biological activity (4, 37, 38). The LPS’s O chain composition, the number and the length of the acyl chains, and the type of substitutions all affect the outcome of TLR4 signaling. Discrimination relies on a specific combination of coreceptors and adaptor molecules. Indeed, the monophosphoryl lipid A is a portion of LPS that preferentially signals through the adaptor Toll/IL-1R domain-containing adapter inducing IFN-β (TRIF) but not MyD88, thereby rendering the molecule less toxic but nevertheless adjuvant (4). Moreover, signaling in the mucosa only has indeed been observed for LPS. For instance, bladder epithelial cells are devoid of the LPS coreceptor CD14 but can detect uropathogenic Escherichia coli LPS by using alternative mechanisms involving fimbriae (38). Interestingly, asialo-GM1 has been proposed as a flagellin coreceptor in the lung (39). Lastly, it has been shown that intracellular detectors like IL-1-converting enzyme protease-activating factor (IPAF) and neuronal apoptosis inhibitory protein 5 (NAIP5) participate in flagellin detection (40). Whether flagellin-mediated activation operates according to the same mechanisms in both mucosal and systemic compartments remains to be determined.

Our findings open up new prospects for the development of antagonistic strategies for manipulating host innate responses and specific inflammatory disorders. Further studies will have to establish whether or not the neutralization of various MAMPs protects or exacerbates bacterial infections or inflammatory diseases.

We thank Prof. Michel Simonet and Dr. François Trottein for critical reading of this manuscript. We thank Dr. Bernhard Ryffel for the gift of TLR5-deficient bone marrow.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

C.N., D.C., and J.C.S. were funded by the Institut National de la Santé et de la Recherche Médicale (including Avenir Grant R02344ES), the Institut Pasteur de Lille, the Région Nord Pas de Calais and FEDER (ARCirémergence), the Franco-Argentinean ECOS-SETCIP Program (A04B03), and the European Community (STREP Grant VaccTIP LSHP-CT-2005-012161).

3

Abbreviations used in this paper: MAMP, microbe-associated molecular pattern; PMN, polymorphonuclear neutrophil; DC, dendritic cell; i.n., intranasal(ly); BAL, bronchoalveolar lavage; Ct, cycle threshold; RLU, relative luminescence; CT, cholera toxin.

1
Akira, S., S. Uematsu, O. Takeuchi.
2006
. Pathogen recognition and innate immunity.
Cell
124
:
783
-801.
2
Beutler, B., Z. Jiang, P. Georgel, K. Crozat, B. Croker, S. Rutschmann, X. Du, K. Hoebe.
2006
. Genetic analysis of host resistance: Toll-like receptor signaling and immunity at large.
Annu. Rev. Immunol.
24
:
353
-389.
3
McKee, A. S., M. W. Munks, P. Marrack.
2007
. How do adjuvants work? Important considerations for new generation adjuvants.
Immunity
27
:
687
-690.
4
Mata-Haro, V., C. Cekic, M. Martin, P. M. Chilton, C. R. Casella, T. C. Mitchell.
2007
. The vaccine adjuvant monophosphoryl lipid A as a TRIF-biased agonist of TLR4.
Science
316
:
1628
-1632.
5
Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R. Goodlett, J. K. Eng, S. Akira, D. M. Underhill, A. Aderem.
2001
. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5.
Nature
410
:
1099
-1103.
6
Rumbo, M., C. Nempont, J. P. Kraehenbuhl, J. C. Sirard.
2006
. Mucosal interplay among commensal and pathogenic bacteria: lessons from flagellin and Toll-like receptor 5.
FEBS Lett.
580
:
2976
-2984.
7
Yonekura, K., S. Maki-Yonekura, K. Namba.
2003
. Complete atomic model of the bacterial flagellar filament by electron cryomicroscopy.
Nature
424
:
643
-650.
8
Andersen-Nissen, E., K. D. Smith, K. L. Strobe, S. L. Barrett, B. T. Cookson, S. M. Logan, A. Aderem.
2005
. Evasion of Toll-like receptor 5 by flagellated bacteria.
Proc. Natl. Acad. Sci. USA
102
:
9247
-9252.
9
Smith, K. D., E. Andersen-Nissen, F. Hayashi, K. Strobe, M. A. Bergman, S. L. Barrett, B. T. Cookson, A. Aderem.
2003
. Toll-like receptor 5 recognizes a conserved site on flagellin protofilament formation and bacterial motility.
Nat. Immunol.
4
:
1247
-1253.
10
Eaves-Pyles, T. D., H. R. Wong, K. Odoms, R. B. Pyles.
2001
. Salmonella flagellin-dependent proinflammatory responses are localized to the conserved amino and carboxyl regions of the protein.
J. Immunol.
167
:
7009
-7016.
11
Yoshioka, K., S. Aizawa, S. Yamaguchi.
1995
. Flagellar filament structure and cell motility of Salmonella typhimurium mutants lacking part of the outer domain of flagellin.
J. Bacteriol.
177
:
1090
-1093.
12
Kuwajima, G..
1988
. Flagellin domain that affects H antigenicity of Escherichia coli K-12.
J. Bacteriol.
170
:
485
-488.
13
Liaudet, L., K. G. Murthy, J. G. Mabley, P. Pacher, F. G. Soriano, A. L. Salzman, C. Szabo.
2002
. Comparison of inflammation, organ damage, and oxidant stress Salmonella enterica serovar Muenchen flagellin and serovar lipopolysaccharide.
Infect. Immun.
70
:
192
-198.
14
Pino, O., M. Martin, S. M. Michalek.
2005
. Cellular mechanisms of the adjuvant activity of the flagellin component FljB of Salmonella enterica serovar Typhimurium to potentiate mucosal and systemic responses.
Infect. Immun.
73
:
6763
-6770.
15
Didierlaurent, A., I. Ferrero, L. A. Otten, B. Dubois, M. Reinhardt, H. Carlsen, R. Blomhoff, S. Akira, J. P. Kraehenbuhl, J. C. Sirard.
2004
. Flagellin promotes myeloid differentiation factor 88-dependent development of Th2-type response.
J. Immunol.
172
:
6922
-6930.
16
Means, T. K., F. Hayashi, K. D. Smith, A. Aderem, A. D. Luster.
2003
. The Toll-like receptor 5 stimulus bacterial flagellin induces maturation and chemokine production in human dendritic cells.
J. Immunol.
170
:
5165
-5175.
17
McSorley, S. J., B. D. Ehst, Y. Yu, A. T. Gewirtz.
2002
. Bacterial flagellin is an effective adjuvant for CD4+ T cells in vivo.
J. Immunol.
169
:
3914
-3919.
18
Gewirtz, A. T., P. O. Simon, C. K. Schmitt, L. J. Taylor, C. H. Hagedorn, A. D. O'Brien, A. S. Neish, J. L. Madara.
2001
. Salmonella typhimurium translocates flagellin across intestinal epithelia, inducing a proinflammatory response.
J. Clin. Invest.
107
:
99
-109.
19
Sierro, F., B. Dubois, A. Coste, D. Kaiserlian, J. P. Kraehenbuhl, J. C. Sirard.
2001
. Flagellin stimulation of intestinal epithelial cells triggers CCL20-mediated migration of dendritic cells.
Proc. Natl. Acad. Sci. USA
98
:
13722
-13727.
20
Hawn, T. R., A. Verbon, K. D. Lettinga, L. P. Zhao, S. S. Li, R. J. Laws, S. J. Skerrett, B. Beutler, L. Schroeder, A. Nachman, et al
2003
. A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires’ disease.
J. Exp. Med.
198
:
1563
-1572.
21
McDonald, W. F., J. W. Huleatt, H. G. Foellmer, D. Hewitt, J. Tang, P. Desai, A. Price, A. Jacobs, V. N. Takahashi, Y. Huang, et al
2007
. A West Nile virus recombinant protein vaccine that coactivates innate and adaptive immunity.
J. Infect. Dis.
195
:
1607
-1617.
22
Honko, A. N., N. Sriranganathan, C. J. Lees, S. B. Mizel.
2006
. Flagellin is an effective adjuvant for immunization against lethal respiratory challenge with Yersinia pestis.
Infect. Immun.
74
:
1113
-1120.
23
Rumbo, M., F. Sierro, N. Debard, J. P. Kraehenbuhl, D. Finke.
2004
. Lymphotoxin β receptor signaling induces the chemokine CCL20 in intestinal epithelium.
Gastroenterology
127
:
213
-223.
24
He, X. S., M. Rivkina, B. A. Stocker, W. S. Robinson.
1994
. Hypervariable region IV of Salmonella gene fliCd encodes a dominant surface epitope and a stabilizing factor for functional flagella.
J. Bacteriol.
176
:
2406
-2414.
25
Bambou, J. C., A. Giraud, S. Menard, B. Begue, S. Rakotobe, M. Heyman, F. Taddei, N. Cerf-Bensussan, V. Gaboriau-Routhiau.
2004
. In vitro and ex vivo activation of the TLR5 signaling pathway in intestinal epithelial cells by a commensal Escherichia coli strain.
J. Biol. Chem.
279
:
42984
-42992.
26
Lee, S. E., S. Y. Kim, B. C. Jeong, Y. R. Kim, S. J. Bae, O. S. Ahn, J. J. Lee, H. C. Song, J. M. Kim, H. E. Choy, et al
2006
. A bacterial flagellin, Vibrio vulnificus FlaB, has a strong mucosal adjuvant activity to induce protective immunity.
Infect. Immun.
74
:
694
-702.
27
Lodes, M. J., Y. Cong, C. O. Elson, R. Mohamath, C. J. Landers, S. R. Targan, M. Fort, R. M. Hershberg.
2004
. Bacterial flagellin is a dominant antigen in Crohn disease.
J. Clin. Invest.
113
:
1296
-1306.
28
Gewirtz, A. T., M. Vijay-Kumar, S. R. Brant, R. H. Duerr, D. L. Nicolae, J. H. Cho.
2006
. Dominant-negative TLR5 polymorphism reduces adaptive immune response to flagellin and negatively associates with Crohn’s disease.
Am. J. Physiol.
290
:
G1157
-G1163.
29
McSorley, S. J., S. Asch, M. Costalonga, R. L. Reinhardt, M. K. Jenkins.
2002
. Tracking Salmonella-specific CD4 T cells in vivo reveals a local mucosal response to a disseminated infection.
Immunity
16
:
365
-377.
30
Kim, K. D., J. Zhao, S. Auh, X. Yang, P. Du, H. Tang, Y. X. Fu.
2007
. Adaptive immune cells temper initial innate responses.
Nat. Med.
13
:
1248
-1252.
31
Saha, S., F. Takeshita, T. Matsuda, N. Jounai, K. Kobiyama, T. Matsumoto, S. Sasaki, A. Yoshida, K. Q. Xin, D. M. Klinman, et al
2007
. Blocking of the TLR5 activation domain hampers protective potential of flagellin DNA vaccine.
J. Immunol.
179
:
1147
-1154.
32
Fernandez, M. I., T. Pedron, R. Tournebize, J. C. Olivo-Marin, P. J. Sansonetti, A. Phalipon.
2003
. Anti-inflammatory role for intracellular dimeric immunoglobulin a by neutralization of lipopolysaccharide in epithelial cells.
Immunity
18
:
739
-749.
33
Kanzler, H., F. J. Barrat, E. M. Hessel, R. L. Coffman.
2007
. Therapeutic targeting of innate immunity with Toll-like receptor agonists and antagonists.
Nat. Med.
13
:
552
-559.
34
Daubeuf, B., J. Mathison, S. Spiller, S. Hugues, S. Herren, W. Ferlin, M. Kosco-Vilbois, H. Wagner, C. J. Kirschning, R. Ulevitch, G. Elson.
2007
. TLR4/MD-2 monoclonal antibody therapy affords protection in experimental models of septic shock.
J. Immunol.
179
:
6107
-6114.
35
Yarovinsky, F., H. Kanzler, S. Hieny, R. L. Coffman, A. Sher.
2006
. Toll-like receptor recognition regulates immunodominance in an antimicrobial CD4+ T cell response.
Immunity
25
:
655
-664.
36
Sanders, C. J., Y. Yu, D. A. Moore, 3rd, I. R. Williams, A. T. Gewirtz.
2006
. Humoral immune response to flagellin requires T cells and activation of innate immunity.
J. Immunol.
177
:
2810
-2818.
37
Jiang, Z., P. Georgel, X. Du, L. Shamel, S. Sovath, S. Mudd, M. Huber, C. Kalis, S. Keck, C. Galanos, et al
2005
. CD14 is required for MyD88-independent LPS signaling.
Nat. Immunol.
6
:
565
-570.
38
Fischer, H., M. Yamamoto, S. Akira, B. Beutler, C. Svanborg.
2006
. Mechanism of pathogen-specific TLR4 activation in the mucosa: fimbriae, recognition receptors, and adaptor protein selection.
Eur. J. Immunol.
36
:
267
-277.
39
McNamara, N., M. Gallup, A. Sucher, I. Maltseva, D. McKemy, C. Basbaum.
2006
. AsialoGM1 and TLR5 cooperate in flagellin-induced nucleotide signaling to activate Erk1/2.
Am. J. Respir. Cell Mol. Biol.
34
:
653
-660.
40
Miao, E. A., C. M. Alpuche-Aranda, M. Dors, A. E. Clark, M. W. Bader, S. I. Miller, A. Aderem.
2006
. Cytoplasmic flagellin activates caspase-1 and secretion of interleukin 1β via Ipaf.
Nat. Immunol.
7
:
569
-575.