Helicobacter pylori (H. pylori) induces chronic gastritis in humans, and infection can persist for decades. One H. pylori strain-specific constituent that augments disease risk is the cag pathogenicity island. The cag island encodes a type IV secretion system (T4SS) that translocates DNA into host cells. Toll-like receptor 9 (TLR9) is an innate immune receptor that detects hypo-methylated CpG DNA motifs. In this study, we sought to define the role of the H. pylori cag T4SS on TLR9-mediated responses in vivo. H. pylori strain PMSS1 or its cagE− mutant, which fails to assemble a T4SS, were used to infect wild-type or Tlr9−/− C57BL/6 mice. PMSS1-infected Tlr9−/− mice developed significantly higher levels of inflammation, despite similar levels of colonization density, compared with PMSS1-infected wild-type mice. These changes were cag dependent, as both mouse genotypes infected with the cagE− mutant only developed minimal inflammation. Tlr9−/− genotypes did not alter the microbial phenotypes of in vivo-adapted H. pylori strains; therefore, we examined host immunological responses. There were no differences in levels of TH1 or TH2 cytokines in infected mice when stratified by host genotype. However, gastric mucosal levels of IL-17 were significantly increased in infected Tlr9−/− mice compared with infected wild-type mice, and H. pylori infection of IL-17A−/− mice concordantly led to significantly decreased levels of gastritis. Thus loss of Tlr9 selectively augments the intensity of IL-17-driven immune responses to H. pylori in a cag T4SS-dependent manner. These results suggest that H. pylori utilizes the cag T4SS to manipulate the intensity of the host immune response.
- Helicobacter pylori
- Toll-like receptor 9
NEW & NOTEWORTHY
Helicobacter pylori (H. pylori) activates Toll-like receptor 9 (TLR9) via translocation of DNA through its main virulence constituent, the cag pathogenicity island. In this study, we utilized Tlr9- and IL-17-deficient mice to demonstrate that H. pylori-induced activation of TLR9 in vivo induces an anti-inflammatory phenotype suppressing IL-17-mediated responses. These results suggest that H. pylori utilizes the cag island to manipulate the intensity of the host immune response.
chronic gastritis induced by Helicobacter pylori (H. pylori) typically persists for the lifetime of the host (1, 10). One strain-specific H. pylori locus that augments risk for disease is the cag pathogenicity island. The cag island encodes for a bacterial type IV secretion system (T4SS) that translocates CagA, as well as peptidoglycan, into host cells (19, 27). Intracellular CagA activates multiple signaling cascades in host cells, including proinflammatory pathways. In contrast, chronic activation of Nod1 by H. pylori peptidoglycan can downregulate proinflammatory signaling (2, 25, 27). Recently, we reported that the cag T4SS can also translocate H. pylori DNA into host cells to activate Toll-like receptor 9 (TLR9) (26).
TLR9 is an intracellular receptor that recognizes hypomethylated CpG motifs (13), which are abundant in DNA of bacterial, viral, or synthetic origin (22). TLR9 is a multidimensional immune receptor based on its ability to mediate both pro- and anti-inflammatory responses (16). In the human intestinal tract, TLR9 activation by commensal organisms regulates, and even dampens, inflammatory responses as a means of maintaining homeostasis (16). In mouse models of colitis, TLR9 activation by H. pylori DNA directs immune responses toward an anti-inflammatory phenotype (12, 17, 21, 23), and one study of short-term infection reported that H. pylori gastritis was increased in the absence of TLR9 (20). However, the mechanism of H. pylori-induced TLR9 activation has remained elusive because TLR9 is an endosomal receptor and the majority of colonizing H. pylori remain extracellular.
Because our recent studies demonstrated that H. pylori utilizes the cag T4SS to translocate DNA and activate TLR9 (26), we sought to define the role of this virulence locus in manipulating TLR9 and downstream immune effectors in vivo.
MATERIALS AND METHODS
All animal studies were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the NIH. Vanderbilt University Institutional Animal Care and Use Committee approved all protocols, and all efforts were made to minimize animal suffering. Male and female C57BL/6 mice were purchased from Harlan Laboratories and housed in the Vanderbilt University Animal Care Facilities. Tlr9−/− C57BL/6 mice were generously provided by the laboratory of Dr. Gregory Barton. Mice were orogastrically challenged with Brucella broth alone, with wild-type cag+ H. pylori strain PMSS1 or with a PMSS1 cagE− isogenic mutant (3). Mice were euthanized at 6 wk after the challenge, and gastric tissue was harvested for quantitative culture, histology, immunohistochemistry, and Luminex assays. Serum samples collected from a cardiac puncture were used for ELISA. Results were obtained from two independent 6-wk infection experiments. For the first experiment, n = 5 mice per group, and for the second experiment, n = 10–12 mice per group. IL-17A−/− C57BL/6 mice were orogastrically challenged with Brucella broth alone or with H. pylori strain PMSS1 for 12 wk.
Quantitative H. pylori culture.
Gastric tissue was homogenized in sterile PBS. Samples were plated on selective Trypticase soy agar plates with 5% sheep blood (Hemostat Laboratories) for isolation of H. pylori as described (25). Plates were incubated for 3–5 days at 37°C with 5% CO2. Colony counts were expressed as log colony-forming units per milligram of tissue.
AGS human gastric epithelial cells (ATCC CRL-1739) were grown in RPMI 1640 (Life Technologies) with 10% fetal bovine serum (Atlanta Biologicals). HEK-Blue-hTLR9 (TLR9+) and HEK-Blue Null1 (parental) cells (Invivogen) were used as previously described (26). H. pylori coculture studies were conducted at a multiplicity of infection (MOI) of 100 for 4–24 h.
H. pylori were cocultured with AGS cells at an MOI of 30 for 4 h as previously described (22). Protein lysates were harvested in RIPA buffer, separated by SDS-PAGE, and transferred to PVDF membranes (Thermo). Levels of total CagA (1:5,000 anti-CagA antibody; Austral Biologicals) and phosphorylated CagA (1:5,000 anti-pY99 antibody; Santa Cruz Biotechnology) were determined via Western blotting. Protein intensities were quantified using the ChemiGenius Gel Bio Imaging system (Syngene). Experiments were repeated at least three times.
TLR9 activation assays were conducted as previously described (26). Briefly, HEK-Blue-hTLR9 cells and HEK-Blue-Null1 cells (Invivogen) were seeded in 96-well plates (Co-Star) and challenged with H. pylori (MOI 100). After 24 h, levels of secreted embryonic alkaline phosphatase (SEAP) were quantified by spectrophotometer (650 nm, Biotek) using Quanti-Blue (Invivogen). All experiments were performed in duplicate and repeated at least three times.
Gastric tissues were homogenized in immunoprecipitation (IP) lysis buffer (Pierce) and subsequently passed through a 21-gauge needle. Samples were then centrifuged at 4°C for 10 min, and the aqueous fraction was removed for analysis. Samples were assayed using a magnetic bead-based protein detection assay for murine cytokines in duplicate according to manufacturer's instructions (Millipore) and quantified by a FlexMap 3D plate reader (Luminex). Samples were then normalized to milligrams of protein.
ELISA assays were performed as previously described (25). Briefly, H. pylori strain PMSS1 was grown overnight, washed twice in 1× PBS, pelleted, and lysed in IP lysis buffer (Pierce). The lysate was diluted 1:50 in coating buffer (85 mM NaHCO3, 15 mM Na2CO3, pH 9.5) and incubated overnight at 4°C in 96-well ELISA plates (DYNEX). Following lysate binding, murine serum samples (1:20 dilution) were added for 2 h. Biotin-conjugated anti-mouse IgG1 (BD Pharmigen) or IgG2a (isoform b, BD Pharmigen) antibodies were diluted 1:5,000 and incubated for 1 h at room temperature. Streptavidin-horseradish peroxidase (HRP) (1:10,000 Life Technologies) was added for 1 h. For detection of total IgG, donkey anti-mouse-conjugated HRP (1:5,000 Santa Cruz Biotechnology) was applied for 1 h at room temperature. Reactions were stopped with 2 N H2SO4, and the OD450 was quantified by spectrophotometer (Biotek). All samples were performed in duplicate.
Quantitative reverse transcription real-time PCR.
RNA was isolated from frozen mouse gastric tissue using the Qiagen RNeasy kit. cDNA was synthesized using the high-capacity cDNA reverse transcription kit (ABI). Quantitative PCR was performed using a TaqMan Universal PCR master mix and interferon regulatory factor 4 (IRF-4) TaqMan primer (Mm00516431_m1) in a 7300 Real-Time PCR system (ABI). All samples were normalized to expression of GAPDH.
Histology and immunohistochemistry.
Linear strips of murine gastric tissue were fixed in 10% neutral-buffered formalin (Azer Scientific), embedded in paraffin, and stained with hematoxylin and eosin. A pathologist, blinded to treatment groups, assessed indices of inflammation. Severity of acute and chronic inflammation was graded 0–3 in both the gastric antrum and corpus, as previously described (11), for a cumulative score ranging from 0–12. To quantify the abundance of Treg cells, tissue samples were stained with an antibody directed against FoxP3 (Novus), and the numbers of FoxP3+ cells were enumerated from five high-power fields each (×400 magnification) from both the antrum and corpus by a single pathologist (M. Piazuelo).
Statistical analyses were performed using Prism 6.0 (GraphPad Software). When comparisons between multiple groups were made, ANOVA with Bonferroni correction was performed. Student's t-test was performed when comparisons between only two groups were made. Spearman's correlation was performed to determine linear correlation. For Luminex results, values were square root transformed before statistical analysis. A P value of <0.05 was considered significant. In all figures, means ± SE are shown.
Six- to eight-wk-old wild-type or Tlr9−/− C57BL/6 mice were infected with either the cag+ H. pylori strain PMSS1, a PMSS1 cagE− isogenic mutant that lacks a functional T4SS, or vehicle (Brucella broth) for 6 wk. A significant (P < 0.01) increase in the severity of inflammation was observed in H. pylori-infected Tlr9−/− mice vs. infected wild-type mice (Fig. 1, A and B). As expected, levels of inflammation were cag dependent, as the cagE− mutant induced significantly less inflammation compared with the wild-type H. pylori strain, irrespective of host backgrounds (Fig. 1, A and B).
To define the mechanisms that underpin these differences, we first investigated whether microbial factors contributed to this phenotype. No significant differences in levels of H. pylori colonization were identified between wild-type and Tlr9−/− mice. There was an increase in colonization between wild-type H. pylori and the cagE− isogenic mutant within the same host genotype, which is concordant with previous reports (11) (Fig. 2A).
We next determined whether selective pressure exerted by the genetic loss of Tlr9 affected H. pylori phenotypes. H. pylori strains were recovered from PMSS1-infected wild-type or Tlr9−/− mice and were assessed for the ability to activate TLR9 and translocate CagA in vitro. To quantify TLR9 activation, we challenged HEK-293-hTLR9 reporter cells, which overexpress TLR9 and contain an NF-κB/AP-1-linked SEAP reporter, with in vivo-adapted H. pylori isolates. To quantify cag T4SS function, we challenged AGS gastric epithelial cells with the in vivo-adapted H. pylori strains and quantified levels of phosphorylated CagA by Western blot. Host Tlr9 status had no effect on TLR9 activation (Fig. 2B) or T4SS function (Fig. 2C); however, there was a significant correlation between the intensity of these phenotypic responses induced by individual strains (Fig. 2D). These data suggest that H. pylori colonization density per se did not contribute to the increase in inflammation in infected Tlr9−/− mice and that Tlr9 deficiency did not alter H. pylori T4SS phenotypes.
Having shown that H. pylori increases inflammation in Tlr9−/− mice (Fig. 1) despite similar levels of colonization (Fig. 2A), we next assessed the role of TLR9-mediated immune effectors. The abundance of H. pylori-specific IgG2a and IgG1 serum antibodies was assessed as markers of TH1 or TH2 immune responses, respectively (Fig. 3A). There were no significant differences in the levels of IgG2a or IgG1 between H. pylori-infected mice when stratified on the basis of Tlr9 genotypes. Because serum IgG markers reflect systemic responses to inflammation, we also quantified immune effectors within uninfected or infected gastric tissue. Infection with H. pylori induced an increase in levels of a subset of specific TH1 and TH2 cytokines in both wild-type and Tlr9-deficient mice compared with uninfected controls. However, there were no significant differences in gastric mucosal levels of the archetypal TH1-secreted cytokines IFN-γ or IL-2 (Fig. 3, B and C), the TH1-associated proinflammatory chemokines keratinocyte chemoattractant or IP-10 (Fig. 3, D and E), or TH2 cytokine profiles (including IL-4 and IL-10; Fig. 3, F and G) when H. pylori-infected wild-type and Tlr9-deficient mice were compared. In addition to these prototype cytokines and chemokines, there were no differences in gastric mucosal levels of IL-1α, IL-1β, IL-6, IL-9, IL-12p40, IL-13, IL-15, granulocyte colony-stimulating factor, granulocyte macrophage colony-stimulating factor, monocyte chemoattractant protein-1, macrophage inflammatory protein-1α (MIP-1α), MIP-1β, or RANTES between H. pylori-infected wild-type vs. Tlr9-deficient mice (data not shown).
On the basis of these results, we next examined the role of TH17-mediated immune responses by quantifying levels of a prototypical marker of this response, IL-17, as a means to understand the observed differences in inflammation. In contrast to TH1 or TH2 effectors, levels of IL-17 were significantly increased in H. pylori-infected Tlr9−/− mice compared with H. pylori-infected wild-type mice (Fig. 4A). To investigate this in more depth, we examined an upstream regulator of TH17 cell differentiation, the transcription factor IRF-4, which has been shown to be constitutively expressed by certain cell types in the absence of TLR9 (5, 18). Using real-time PCR, we demonstrated a significant increase in levels of IRF-4 expression in H. pylori-infected Tlr9−/− mice compared with H. pylori-infected wild-type mice (Fig. 4B), which corresponds to the increase in IL-17 expression in TLR9-deficient mice (Fig. 4A). Previous reports have shown that H. pylori alters the TReg/TH17 balance toward a TReg-biased response, resulting in ineffective immune clearance of bacteria and thus perpetuating persistence (14). Because of the increase in levels of IL-17 in H. pylori-infected Tlr9-deficient mice, we next determined relative abundance of gastric mucosal FoxP3+ TReg cells in H. pylori-infected wild-type and Tlr9−/− mice as a means to understand differences in the intensity of the respective inflammatory responses. Concordant with previous reports (14), a significant increase in FoxP3+ cell abundance was present in mice infected with wild-type H. pylori, compared with uninfected or cagE−-infected mice (Fig. 4, C and D). However, no differences were present between infected wild-type and Tlr9−/− mice, indicating that the cag T4SS, but not TLR9, is required for modulating the TReg immune response to H. pylori. Finally, to more definitively implicate IL-17 in mediating inflammatory responses to H. pylori, we infected wild-type or IL-17A−/− mice with wild-type cag+ H. pylori strain PMSS1 for 12 wk. A significant reduction in the severity of acute inflammation was observed in infected IL-17A−/− mice compared with infected wild-type mice (Fig. 4E). Collectively, these data suggest that H. pylori-induced TLR9 activation suppresses IL-17-mediated immunity in a cag T4SS-dependent manner, but this mechanism is independent of regulatory T cells.
In this study, we capitalized on a mouse-adapted cag+ H. pylori strain that readily infects rodents and maintains cag T4SS function for weeks to investigate the effects of this virulence constituent within the context of TLR9 activation (3, 9, 15). Our results indicate that the mouse Tlr9−/− genotype did not alter H. pylori colonization density or the microbial phenotype of H. pylori output derivatives in terms of T4SS function. However, infected Tlr9-deficient mice developed more severe inflammation compared with infected wild-type mice. Although levels of TH1 and TH2 cytokines remained unchanged between mouse genotypes, levels of IL-17, a TH17 cytokine, were differentially abundant between H. pylori-infected wild-type vs. Tlr9−/− mice, raising the tantalizing possibility that TLR9 can suppress TH17 responses to H. pylori. We recognize that differences in inflammation may not be entirely attributable to IL-17, as TLR9 has been shown to mediate both pro- and anti-inflammatory signaling via multiple mechanisms (16, 20). In our experiments, the absence of TLR9 biased cellular responses toward proinflammatory phenotypes. Because TLR9 has been shown to upregulate Cox2 expression and prostaglandin E2 release, these events may also contribute to the increased inflammation seen in our model (6, 7).
TLR9 can also induce tolerogenic responses based on its cellular localization at the time of activation. Apical (luminal) TLR9 stimulation inhibits the function of the E3 ubiquitin ligase β-TrCP and downregulates inflammatory responses. In the canonical NF-κB signaling cascade, β-TrCP catalyzes the destruction of IκB and p105 (NF-κB1, p50 precursor) (16). Similarly, ERK activation also requires the ubiquitination of p105 (16). However, basolateral TLR9 activation leads to robust NF-κB activation. Therefore, H. pylori-mediated apical TLR9 activation may promote tolerogenic responses by blocking canonical NF-κB and MAPK pathways that could result from basolateral TLR9 stimulation or from other proinflammatory stimuli that utilize the NF-κB and/or MAPK pathways (16). In this study, we have observed an increase in inflammation and IL-17 expression in Tlr9−/− mice, which may result from such a lack of TLR9-mediated tolerogenic signaling.
Exciting new data have also supported a role for TLR9-mediated suppression of inflammation in the stomach (8). These investigations demonstrated that TLR9 and a key target, IFN-α, are required for the migration of myeloid-derived suppressor cells into gastric tissue, where they then suppress T-cell activation. Therefore, H. pylori activation of TLR9 provides a mechanism through which inflammation is suppressed that may facilitate the establishment of persistent infection, and selective targeting of TLR9 may represent a therapeutic strategy to optimize effective immune clearance to prevent prolonged periods of colonization.
Despite these findings that TLR9 can mediate an anti-inflammatory phenotype, there are limitations in translating these data to human infections. Inherent differences are present in mouse TLR9 compared with human TLR9. Both mouse and human TLR9 recognize CpG DNA motifs; however, these DNA sequences are distinct. Additionally, human TLR9 is expressed in epithelial cells, B cells, and plasmacytoid dendritic cells, whereas mouse TLR9 is expressed, not only in those same cell types, but also in myeloid dendritic cells, as well as macrophages (4, 24).
In conclusion, this study demonstrates that H. pylori modulates the host immune system toward an IL-17- and TLR9-mediated, anti-inflammatory response in a cag T4SS-dependent manner. This modulation could contribute to the ability of H. pylori to persist long term within the stomach, and, importantly, these studies lay the foundation for further exploration into the role of TLR9-H. pylori interactions in human hosts.
This work was supported by NIH R01-DK58587, R01-CA77955, P01-CA116087, and P30-DK058404 (R. Peek Jr.); R01-DK053620, R01-CA190612, and P01-CA028842 (K. Wilson); Department of Veterans Affairs Merit Review grants I01BX001453 (K. Wilson); IBX000915A (H. Algood); and IFNB-024-13F (E. Skaar).
No conflicts of interest, financial or otherwise, are declared by the authors.
M.G.V., E.P.S., K.T.W., H.M.S.A., and R.M.P. conception and design of research; M.G.V., J.R.-G., A.G.D., M.E.W., and R.V.P. performed experiments; M.G.V., M.B.P., J.R.-G., M.E.W., R.V.P., and H.M.S.A. analyzed data; M.G.V., M.B.P., A.G.D., G.S., M.E.W., U.S.K., K.T.W., H.M.S.A., and R.M.P. interpreted results of experiments; M.G.V. and R.M.P. prepared figures; M.G.V. and R.M.P. drafted manuscript; M.G.V., M.B.P., J.R.-G., A.G.D., G.S., M.E.W., U.S.K., E.P.S., K.T.W., H.M.S.A., and R.M.P. edited and revised manuscript; M.G.V., M.B.P., J.R.-G., A.G.D., G.S., M.E.W., U.S.K., R.V.P., E.P.S., K.T.W., H.M.S.A., and R.M.P. approved final version of manuscript.
We acknowledge Dr. Gregory Barton of the University of California at Berkley for providing the Tlr9−/− mice and the following Vanderbilt core laboratories and personnel: Tissue Acquisition and Pathology Core, Division of Animal Care, and the Digestive Disease Research Center.
- Copyright © 2016 the American Physiological Society