|Home | About | Journals | Submit | Contact Us | Français|
Helicobacter pylori is a gram-negative microaerophilic bacterium that colonizes the gastric mucosa, leading to disease conditions ranging from gastritis to cancer. Toll-like receptors (TLRs) play a central role in innate immunity by their recognition of conserved molecular patterns on bacteria, fungi, and viruses. Upon recognition of microbial components, these TLRs associate with several adaptor molecules, including myeloid differentiation factor 88 (MyD88). To investigate the contribution of the innate immune system to H. pylori infection, bone marrow-derived macrophages from mice deficient in TLR2, TLR4, TLR9, and MyD88 were infected with H. pylori SS1 and SD4 for 24 or 48 h. We demonstrate that MyD88 was essential for H. pylori induction of all cytokines investigated except alpha interferon (IFN-α). The secretion of IFN-α was substantially increased from cells deficient in MyD88. H. pylori induced interleukin-12 (IL-12) and IL-10 through TLR4/MyD88 signaling. In addition, H. pylori induced less IL-6 and IL-1β in TLR2-deleted macrophages, suggesting that the MyD88 pathway activated by TLR2 stimulation is responsible for H. pylori induction of the host proinflammatory response (IL-6 and IL-1β). These observations are important in light of a recent report on IL-6 and IL-1β playing a role in the development of H. pylori-related gastric cancer. In conclusion, our study demonstrates that H. pylori activates TLR2 and TLR4, leading to the secretion of distinct cytokines by macrophages.
Helicobacter pylori colonizes the gastric mucosa, causing chronic inflammation characterized by mononuclear cell infiltration, including macrophages. This influx of immune cells is associated with cytokine production in the gastric mucosa (10, 41). Despite a strong host immune response, H. pylori is able to efficiently establish a persistent infection.
Toll-like receptors (TLRs) play a central role in the innate immune system, the first line of defense against invading pathogens. Studies show that macrophages express TLRs, including TLR2, TLR4, and TLR9, and are a main source of proinflammatory cytokine secretion (23, 45). These TLRs recognize conserved pathogen-associated molecular patterns expressed on bacteria, fungi, or viruses (1, 28). For example, TLR2 acts as signal-transducing receptor for bacterial lipoproteins (3, 16), peptidoglycans (39, 46), and lipoarabinomannan (27). TLR4 recognizes primarily lipopolysaccharide (LPS) (6, 17, 36), while TLR9 is the receptor for CpG-containing DNA (15). Upon recognition of microbial components, these TLRs associate with several adaptor molecules, including myeloid differentiation factor 88 (MyD88) (29, 30). MyD88 is an adaptor molecule that is common to all TLR signaling with the exception of TLR3 (45). The interaction of a TLR with its ligand initiates signaling cascades (reviewed in reference 2) that lead to the induction of several cytokines (reviewed in reference 14), including interleukin-1β (IL-1β), IL-6, IL-12, and IL-10.
Although it is now accepted that TLRs are involved in the recognition of pathogen-associated molecular patterns during H. pylori infection, there is still contradictory information as to which TLRs are important in mediating the host response. A few reports agree that TLR2 (8, 26, 40) and TLR4 (5, 21, 22, 25, 42) play a role in H. pylori infection. In contrast, some studies show that both TLR2 (12) and TLR4 (4, 12, 26, 40) are not involved in H. pylori-related stimulation of cytokines, clearly indicating that the role of TLRs in H. pylori infection is still an area of controversy. In addition, the TLRs involved in H. pylori-related production of specific cytokines have not been fully defined. To better clarify the role of TLRs in H. pylori infection, bone marrow-derived macrophages (BMDM) from TLR-deficient mice were used to determine the contributions of TLR2, TLR4, and TLR9 to cytokine secretion during H. pylori infection in vitro. In addition, the role of the accessory molecule MyD88 in mediating the production of these cytokines during H. pylori infection was investigated using MyD88-deficient BMDM.
We demonstrate that TLR2 and TLR4 are crucial as signaling receptors for H. pylori activation of the host immune response leading to the secretion of cytokines and that TLR9 is not required. In addition, we also show that MyD88 is essential for H. pylori induction of these cytokines.
Six- to 10-week-old mice deficient in TLR2 (TLR2−/−), TLR4 (TLR4−/−), TLR9 (TLR9−/−), and MyD88 (MyD88−/−) of the C57BL/6 background as well as wild-type (WT) mice were used in this study. WT and TLR2−/− mice were purchased from The Jackson Laboratory (Bar Harbor, ME). TLR4−/−, TLR9−/−, and MyD88−/− mice were obtained from our breeding colony, originally provided by Akira (Osaka University, Japan). All animal procedures were approved by the animal welfare committee at the University of California, San Diego.
Bone marrow was isolated from WT, TLR2−/−, TLR4−/−, TLR9−/−, and MyD88−/− mice. Mice were sacrificed and their femurs excised and crushed under aseptic conditions. The bone marrow cells were washed with Dulbecco's modified Eagle's medium (DMEM; Cellgro, VA) and cultured in DMEM supplemented with glutamine, pyruvate, 10% heat-inactivated fetal calf serum, and 100 U/ml penicillin-100 μg/ml streptomycin (Pen/Strep; Cellgro). The supernatant of L929 fibroblasts at a final concentration of 30% (vol/vol) was a source of colony-stimulating factors. Cultures were incubated at 37°C for 7 days. BMDM were harvested by being washed three times with DMEM without antibiotics, resuspended in fresh DMEM containing 10% heat-inactivated fetal calf serum and macrophage colony-stimulating factor (PeproTech Inc., NJ), and seeded in 12-well culture plates at densities of 5 × 105 to 7 × 105 cells/ml. The process of isolating, culturing, and harvesting BMDM for infection was repeated three or four separate times for each MyD88- or TLR-deficient mouse with their WT counterpart.
H. pylori strains SS1 (24) and SD4, a clinical isolate (9), were used in the present study. In addition, an isogenic SD4 cagE mutant (13) was used. Bacteria were maintained on Columbia agar supplemented with 5% laked horse blood under microaerophilic conditions at 37°C, as previously described (32). For infections, bacteria were grown in brain heart infusion broth (Becton Dickinson, MD) containing 5% fetal calf serum and incubated at 37°C on a reciprocal shaker overnight. Before infections, spiral bacteria were counted using a Petroff-Hausser counting chamber.
H. pylori strains SS1 and SD4 were added to BMDM at a multiplicity of infection (MOI) of 100 and incubated for 24 or 48 h. BMDM and bacteria were incubated in cell culture medium under the cell culture conditions described above. Experiments were performed in triplicate. In subsequent experiments, cells were infected for 24 h, which is a commonly used time point for H. pylori infections of human and murine macrophages (12, 33, 34) and is considered optimal for the production of most cytokines investigated in the present study. In a separate experiment, H. pylori strain SD4 was added to BMDM at an MOI of 10, 50, or 100 for 24 h. Supernatants were collected and stored at −80°C for cytokine analysis at a later time.
BMDM in 12-well culture plates were incubated for 24 h with H. pylori. In addition, BMDM were stimulated with 2 μg/ml of LPS (Escherichia coli serotype 026:B6; Sigma-Aldrich, St. Louis, MO) or Pam3Cys-SK4 (EMC microcollections GmbH). All TLR ligands were diluted prior to the experiment with DMEM. Pam3Cys-SK4 served as a positive control for TLR2 and LPS for TLR4. The levels of IL-12p40, IL-6, IL-10 (R & D Systems, MN), alpha interferon (IFN-α), and IL-1β (Biosource, CA) produced by the TLR- and MyD88-deleted and WT control BMDM in culture supernatants were determined by commercially available enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturers' instructions. The minimum detectable cytokine concentrations were 4 pg/ml, 4 pg/ml, 1.8 pg/ml, 10 pg/ml, and 3 pg/ml for IL-12p40, IL-10, IL-6, IFN-α, and IL-1β, respectively.
Results are expressed as means ± standard errors of the means (SEM). Data were analyzed using a Student t test (two-tailed). A P value of ≤0.01 was considered statistically significant.
To determine the contribution of MyD88 to H. pylori induction of cytokines, BMDM deficient in MyD88 were infected with two strains of H. pylori, SS1 and SD4. BMDM were infected for 24 h, and IL-6, IL-12, IL-10, IL-1β, and IFN-α levels in culture supernatants were measured by ELISA. Cytokine levels in culture supernatants of BMDM from WT mice were also measured. The results shown in Fig. Fig.1A1A demonstrate that BMDM secreted IL-6, IL-12, IL-10, and IL-1β in response to H. pylori infection. The secretion of these cytokines was significantly higher (P < 0.01) in BMDM from WT mice than in those from mice deficient in MyD88 (Fig. (Fig.1A).1A). Both H. pylori SS1 and SD4 induced more IL-6, IL-12, IL-10, and IL-1β production in BMDM from WT mice than in those from MyD88−/− mice. IL-6 levels were undetectable in uninfected BMDM from both WT and MyD88−/− mice. In addition, less than 12 pg/ml of IL-12 or IL-10 was detected in uninfected BMDM while less than 50 pg/ml of IL-1β was detected in uninfected BMDM. Our data indicate that normal secretion of these cytokines in response to H. pylori infection requires signaling through MyD88.
Although MyD88 was required for H. pylori induction of IL-6, IL-12, IL-10, and IL-1β, we found that it was not required for H. pylori induction of IFN-α (Fig. (Fig.1B).1B). Furthermore, a deficiency of MyD88 resulted in substantially increased IFN-α levels. The IFN-α levels detected from uninfected BMDM were less than 10 pg/ml. As mentioned in Materials and Methods, all assays were performed in triplicate wells and each experiment was repeated three separate times using independent BMDM preparations. In all three independent experiments, IFN-α was detected only in the BMDM from MyD88−/− mice. Both H. pylori strains SS1 and SD4 stimulated BMDM from MyD88−/− mice to produce IFN-α but not BMDM from WT mice. Therefore, H. pylori induced more IFN-α in BMDM from MyD88−/− mice than in those from WT mice.
MyD88 is a common adaptor molecule to all TLR signaling pathways except TLR3. We demonstrated in the present study that MyD88 is important in H. pylori induction of all the cytokines that we studied except IFN-α (Fig. (Fig.1B).1B). We next determined which TLRs were required for the response to H. pylori infection. Using available TLR knockout mice (TLR2−/−, TLR4−/−, and TLR9−/−), we investigated the role of these TLRs in H. pylori induction of these cytokines. We show that TLR2 was important in H. pylori induction of IL-6 (Fig. (Fig.2A)2A) and IL-1β (Fig. (Fig.2B).2B). In addition, we demonstrate that H. pylori induction of IL-6 and IL-1β was dose dependent (Fig. (Fig.3).3). The control TLR2 agonist Pam3Cys-SK4 induced both IL-6 and IL-1β in BMDM from WT mice but not in those from TLR2−/− mice (Fig. (Fig.3).3). Both H. pylori strains SS1 and SD4 induced significantly (P < 0.01) greater IL-6 (Fig. (Fig.2A)2A) and IL-1β (Fig. (Fig.2B)2B) levels in BMDM from WT mice than in those from TLR2−/− mice. BMDM secretion of IL-6 and IL-1β in response to infection with H. pylori SD4 was slightly less than that in response to infection with H. pylori SS1. IL-6 levels were undetectable from uninfected BMDM from both WT and TLR2−/− mice, while less than 50 pg/ml IL-1β was detected in uninfected BMDM. TLR2−/− macrophages and WT macrophages did not show a consistent difference in the secretion of IL-10 and IL-12 following stimulation by H. pylori (data not shown).
We next used BMDM from WT and TLR4−/− mice to investigate the role of TLR4 in H. pylori induction of cytokines. The activation of TLR4 is known to lead to the recruitment of MyD88 (reviewed in reference 44). As shown in Fig. Fig.4,4, both strains of H. pylori induced more IL-12 (Fig. (Fig.4A)4A) and IL-10 (Fig. (Fig.4B)4B) in BMDM from WT mice than in those from TLR4−/− mice, indicating that TLR4 is involved in H. pylori-related induction of these cytokines. Less than 40 pg/ml of IL-12 or IL-10 was detected in uninfected BMDM from both WT and TLR4−/− mice. H. pylori induction of IL-12 and IL-10 was dose dependent (Fig. (Fig.5).5). The control TLR4 agonist LPS induced both IL-12 and IL-10 in BMDM from WT mice, and the levels of these cytokines were much lower in TLR4−/− BMDM. IL-10 production increased with an increase in inoculum size (Fig. (Fig.5B),5B), while for IL-12, the reverse was observed (Fig. (Fig.5A).5A). The decline in IL-12 levels with an increase in the number of bacteria was also observed in BMDM from TLR2−/− mice (data not shown). As shown in Fig. Fig.5A,5A, the optimal MOI for the maximum production of IL-12 was 10. Nevertheless, the data still show that TLR4 is required for this maximum induction.
As shown in Fig. Fig.22 and and4,4, H. pylori stimulated BMDM to secrete cytokines via TLR2 and TLR4. H. pylori is known to release DNA in the culture and could provide a ligand for TLR9. However, TLR9 deficiency had no effect on cytokine secretion induced by H. pylori. Infection with either strain of H. pylori resulted in the production of similar levels of cytokines by BMDM from WT and TLR9−/− mice (data not shown). Thus, TLR2 and TLR4, but not TLR9, were important for the cytokine secretion response to H. pylori that we observed.
TLR-mediated signal transduction plays a central role in the innate immune system, initiating the inflammatory reaction to bacteria. However, the specific TLR pathways involved in H. pylori activation of the innate immune response remain controversial. The MyD88 adaptor protein is a key mediator in signal transduction for many TLRs, including TLR2, TLR4, and TLR9. We explored the role of MyD88 and these TLRs during H. pylori infection by using BMDM isolated from mice deficient in these factors. H. pylori activated IL-12, IL-10, IL-6, and IL-1β production in macrophages by a MyD88-dependent mechanism, indicating that MyD88 is essential for the production of these cytokines in response to H. pylori components present in the whole organism. In contrast to our findings, Gobert et al. (12) reported that the production of IL-6 in murine macrophages stimulated by H. pylori heat shock protein 60 or intact bacteria was TLR2, TLR4, and MyD88 independent. In our study, we could barely detect IL-6 in BMDM from MyD88−/− mice. In addition, we observed that H. pylori induced less IL-6 in TLR2-deleted macrophages than in WT macrophages, suggesting that TLR2 plays a key role in eliciting the signaling cascade that leads to IL-6 production. One difference between our experiment and that of Gobert et al. (12) was the origin of the macrophages. We used macrophages derived from the bone marrow, while they used macrophages isolated from the peritoneum. Our results are in agreement with several published reports that suggest that MyD88 is essential in the production of IL-6 mediated by microbial components through TLRs that use MyD88 as an adaptor molecule (reviewed in reference 45).
Contrary to the importance of MyD88 in H. pylori-related induction of IL-6, IL-12, IL-10, and IL-1β, a functional MyD88 resulted in less IFN-α production. Our data provide evidence that MyD88 deletion leads to the production of IFN-α by BMDM when infected with H. pylori. IFN-α is a type I IFN, first described as an antiviral protein essential in the clearance of viral pathogens (reviewed in reference 7). We showed that H. pylori induced IFN-α only in the absence of MyD88. MyD88-independent induction of type I IFNs occurs through the adaptor TRIF-dependent pathway (48, 49). Since we did not detect IFN-α in macrophages from WT mice when infected with H. pylori, our data indicate a novel down-regulatory effect on type I IFN mediated by a MyD88-dependent mechanism that could affect both innate and acquired immune responses.
H. pylori induction of IL-1β and IL-6 secretion in macrophages was decreased in the absence of TLR2 at all bacterial doses used. TLR2 has been suggested to play a role in controlling infection with H. pylori (35). Our results are in agreement with those of Mandell et al. (26), who showed that H. pylori activates innate immunity via TLR2. In addition, our present observations on the role of TLR2 in H. pylori induction of IL-6 and IL-1β are important in light of a recent report on these two cytokines' role in the development of gastric cancer (20).
Data on the role of TLR4 in H. pylori infection are contradictory. While some studies report no involvement of TLR4 in the immune response to H. pylori (12, 40), others show that TLR4 plays an essential role in mucosal immunity to H. pylori (18, 21, 37, 38). BMDM from TLR4-deficient mice showed less induction of IL-10 and IL-12 than WT BMDM at every dose of bacteria used. Since IL-12 is pivotal in directing the development of Th1 responses (11, 47), this suggests that TLR4 may be important for mounting a Th1 response to H. pylori. In addition, we demonstrated that TLR4 is important in H. pylori induction of IL-10. It is well documented that IL-10 suppresses the host immune response (19). Therefore, our studies showing the partial dependence of IL-10 induction by H. pylori on TLR4 suggest that H. pylori suppresses the host immune response through TLR4. In mice, the production of IL-10 was associated with a lack of gastric inflammation (43). Data from our recent in vitro study using dendritic cells suggest that H. pylori modulates the host immune response through the induction of IL-10, which then down-regulates IL-12 (31). Expression of TLR4 has been demonstrated in the human stomach (38). Indeed, TLR4 has been detected in gastric pit cells, where it has been suggested to play a role in the regulation of the gastric pit cell response (21).
In the present study, we have shown that H. pylori activation of TLR2 and TLR4 resulted in the induction of distinctly different cytokines. TLR2 was important for the secretion of the proinflammatory cytokines IL-1β and IL-6, while TLR4 was required for full production of the immune regulatory cytokines IL-10 and IL-12. These results suggest a functional specificity for the activation of the immune system through TLR ligation during H. pylori infection, with TLR2 being involved in proinflammatory stimuli and TLR4 mediating a regulatory effect through a bacterial dose-dependent reciprocal induction of IL-10 and decrease of IL-12.
We thank Maripat Corr for providing key mouse strains.
This study is supported by a Public Health Service grant (K01 CA96709) from the National Cancer Institute to Marygorret Obonyo.
Editor: F. C. Fang
Published ahead of print on 12 March 2007.