Chlamydia trachomatis Induces Expression of IFN--Inducible Protein 10 and IFN- Independent of TLR2 and TLR4, but Largely Dependent on MyD881
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
IFN--inducible protein 10 (IP-10) is a chemokine important in the attraction of T cells, which are essential for resolution of chlamydial genital tract infection. During infections with Gram-negative bacteria, the IP-10 response mediated through type I IFNs usually occurs as a result of TLR4 stimulation by bacterial LPS. However, we found that levels of IP-10 in genital tract secretions of Chlamydia trachomatis-infected female wild-type mice were similar to those of infected TLR2- and TLR4-deficient mice but significantly greater than those of infected MyD88-deficient mice. We investigated the mechanism of IP-10 and IFN- induction during chlamydial infection using mouse macrophages and fibroblasts infected ex vivo. The induction of IP-10 and IFN- was unchanged in Chlamydia-infected TLR2- and TLR4-deficient cells compared with wild-type cells. However, infection of MyD88-deficient cells resulted in significantly decreased responses. These results suggest a role for MyD88-dependent pathways in induction of IP-10 and IFN- during chlamydial infection. Furthermore, treatment of infected macrophages with an endosomal maturation inhibitor significantly reduced chlamydial-induced IFN-. Because endosomal maturation is required for MyD88-dependent intracellular pathogen recognition receptors to function, our data suggest a role for the intracellular pathogen recognition receptor(s) in induction of IFN- and IP-10 during chlamydial infection. Furthermore, the intracellular pathways that lead to chlamydial-induced IFN- function through TANK-binding kinase mediated phosphorylation and nuclear translocation of IFN regulatory factor-3.
Introduction
Chlamydia trachomatis is a common sexually transmitted bacterial pathogen (1) that can cause oviduct inflammation and subsequent tubal infertility (2). The first line of defense against chlamydiae is the innate immune response. The cytokines/chemokines produced during the innate response recruit inflammatory cells and T cells that are needed for adaptive immunity. The T cell-dependent adaptive immune response and subsequent chlamydial clearance is predominated by IL-12- and IFN--dependent mechanisms (3, 4, 5, 6, 7).
One of the major players of the innate immune response to pathogens is the TLR, which is expressed on immune cells (reviewed in Ref. 8) as well as multiple other cell types (9, 10). The recognition of pathogen-associated molecular patterns (PAMP)3 by TLRs leads to signaling events and coordinated activation of transcription factors that induce the expression of antimicrobial molecules, chemokines, cytokines, and costimulatory molecules (11). Therefore, elucidating the TLR involved in initiation of the immune response to chlamydial infection is clearly a primary step to understanding its molecular pathogenesis.
We reported previously that Tlr2 knockout mice (TLR2–/–) infected genitally with the mouse pneumonitis (MoPn) agent of C. trachomatis (also known as C. muridarum) had significantly lower inflammatory cytokine responses, including TNF- and MIP-2, and less oviduct pathology compared with infected wild-type (WT) mice (12). Despite lower cytokine responses, the TLR2–/– mice cleared infection at a rate similar to WT mice. Tlr4 knockout mice (TLR4–/–) infected with MoPn responded like WT mice with respect to both cytokine response and clearance of infection. Because it is well documented that T cells play a major role in clearing Chlamydia infection (7, 13, 14), the results from the TLR2–/– and TLR4–/– mice suggest the influx of T cells to the site of infection was unaffected.
IFN--inducible protein 10 (IP-10) is an important chemokine involved in T cell recruitment (15). IP-10 is induced by both type I (IFN- and IFN-) and type II IFN (16, 17). Besides its role in inducing IP-10, type I IFNs also induce expression of MCP-5, RANTES, and GARG-16 and has been implicated in Th1 maturation (18). IFN- has also been shown to augment dendritic cell (DC) expression of IL-12 and CD40, both essential in generating robust T cell responses (19). It has been reported that IP-10 is produced in the upper and lower genital tract of mice infected with chlamydiae (20). Furthermore, murine DC pulsed in vitro with C. trachomatis (21) and a murine oviduct epithelial cell line have been reported to up-regulate expression of IP-10 during MoPn infection (22). It has also been shown that IFN- is induced in McCoy cells during C. trachomatis infection (23). However, the molecular pathways involved in IP-10 and IFN- induction during chlamydial infection have not yet been described.
TLRs play a major role in induction of type I IFN, in response to bacterial and viral infection (reviewed in Ref. 24). Cross-linking TLR3 and TLR4 by viral dsRNA and LPS, respectively, has been shown to induce IFN- and IFN- (25, 26). Type I IFNs are also induced when TLR7 and TLR9 bind to their ligands, ssRNA and bacterial CpG DNA, respectively (27, 28). However, the signaling pathways by which the different TLRs mediate induction of type I IFNs differ distinctly. TLR3- and TLR4-mediated type I IFN induction is independent of the central downstream adaptor molecule MyD88 (reviewed in Ref. 29), whereas TLR7 and TLR9 mediate induction of type I IFNs in a MyD88-dependent manner (27, 28). However, both pathways lead to phosphorylation of IFN regulatory factor-3 (IRF3), a transcription factor essential for IFN- production by the TANK-binding kinase (TBK) (30).
In this study, the role of IFN- in the induction of IP-10 during MoPn (C. trachomatis) infection was established in vitro, and the contribution of TLR in IP-10 and IFN- induction was explored using murine macrophages and fibroblasts. Chlamydial-induced IP-10 and IFN- were found to be completely independent of TLR2 and TLR4 but significantly dependent on MyD88. The TLR4 independence of chlamydial-induced IFN- expression was surprising, because TLR4 is the major receptor that binds bacterial LPS to induce IFN- (31). Preliminary evidence using an endosomal maturation inhibitor suggests a role for intracellular TLR or other pathogen recognition receptors (PRRs) that function in a MyD88-dependent manner to induce IFN- during chlamydial infection. Furthermore, the role of TBK-mediated phosphorylation of IRF3 and its nuclear translocation for IFN- induction was established.
Materials and Methods
Reagents and immunochemicals
Escherichia coli LPS (InvivoGen) and Staphylococcus aureus peptidoglycan (PGN; Lee Labs) were used as positive controls for TLR4 and TLR2 stimulation, respectively. Although PGN is no longer considered a ligand for TLR2, contaminants in the S. aureus PGN preparation are known to be strong agonists for TLR2 (32). Bafilomycin was obtained from Sigma-Aldrich. Anti-mouse IFN- (PBL Biomedical Laboratories), anti-mouse actin (Sigma Immunochemicals), and anti-mouse IRF3 (Zymed Laboratories) Abs were purchased as indicated. All ELISA kits were from R&D Laboratories.
Animals and cell lines
Tlr2 gene knockout mice (TLR2–/–) were kind gifts from Dr. S. Akira (Department of Host Defense, Research Institute for Microbial Diseases, Osaka, Japan) (33). Because they have not been sufficiently backcrossed to the C57BL/6 background, the heterozygous F1 generations from a cross of C57BL/6 and TLR2–/– mice were used as controls for TLR2–/– mice. However, because there were no differences in infection rates or cytokine levels between C57BL/6 mice and F1 progeny of C57BL/6 x TLR2–/– mice (12), C57BL/6 mice were used as controls for macrophage experiments. MyD88 gene knockout mice or MyD88–/– mice (originally from Dr. Akira) (34) were purchased from the Centre de La Recherche Scientifique and have been previously backcrossed with C57BL/6 mice over 10 generations (35, 36); therefore, C57BL/6 mice were used as controls for MyD88–/– mice. TLR4lps-del (hereafter referred to as TLR4–/–) mice have a deletion in the TLR4 gene (C57BL/10SCN) and were obtained from The Jackson Laboratory. C57BL/10 mice (The Jackson Laboratory) or C57BL/6 mice were used as controls for TLR4–/– in vivo experiments, because they displayed similar infection kinetics. IFN- receptor–/– mice (A129) and their corresponding WT controls (129Sv/Ev) were obtained from B & K Universal. Mice between 8 and 12 wk of age were used and were age matched for all experiments. Mice were given food and water ad libitum in an environmentally controlled room with 12-h light and 12-h dark cycles. All animal experiments were preapproved by the Institutional Animal Care and Use Committee of the University of Arkansas Medical Sciences. WT and TBK–/– mouse embryonic fibroblasts were kind gifts from Dr. W.-C. Yeh (University Health Network, Toronto, Ontario, Canada) (37) and were maintained in DMEM with 10% FBS, 1 mM glutamine, and 50 μg/ml gentamicin.
Murine genital infection with MoPn and analysis of IP-10 in genital secretions
Mice received 2.5 mg of Depoprovera (medroxyprogesterone acetate) in 0.1 ml of saline (Upjohn) s.c., 7 days before vaginal infection. Mice were anesthetized using sodium pentobarbital and infected by placing MoPn (1 x 107 inclusion-forming units (IFU) in 30 μl of buffer (250 mM sucrose, 10 mM sodium phosphate, and 5 mM L-glutamic acid (pH 7.2)) into the vaginal vault. Mice were monitored by swabbing the vaginal vault and cervix with a calcium alginate swab (Spectrum Medical Industries) at various times after infection and by enumerating IFU using McCoy cell monolayers (38). Mice were infected in groups of five, and each experiment was repeated at least once. Genital tract secretions were collected as described previously (12), and IP-10 levels in individual genital tract sponge eluates were determined by ELISA.
Isolation of murine macrophages and lung fibroblasts
Peritoneal macrophages were obtained from mice that were given injections of 3% thioglycolate medium 3 days before harvesting. To harvest macrophages, mice were killed, and their peritoneal cavities were washed three times with culture medium (RPMI 1640 containing 10% FBS, 2 mM glutamine, 10 mM HEPES (pH 7.4), 100 μM nonessential amino acids, 1 mM sodium pyruvate, 50 μM 2-ME, and 50 μg/ml gentamicin). The macrophages in culture medium were plated either in 24-well plates (7 x 105 cells/well) containing coverslips to stain for chlamydial inclusions or in 6-well plates (2.5 x 106 cells/well) for RNA preparations. Mouse peritoneal macrophages collected after thioglycolate treatment were kept in culture conditions for 24–48 h before infection. Bone marrow-derived macrophages (BMDMs) were prepared by flushing the femur and tibia of mice using a 27.5-gauge needle and culturing the cells as described by Kambayashi et al. (39) in the presence of IL-4 (10 ng/ml) and GM-CSF (100 ng/ml) for 5 days at 37°C in a 7.5% CO2 incubator. After removing the nonadherent DCs, the adherent macrophages were used for infection. Lung fibroblasts from WT or TLR-deficient mice were isolated as described previously (12, 40) and cultured for 5–6 days before infecting with C. trachomatis.
Chlamydial stocks and in vitro infection of macrophages
The MoPn agent of C. trachomatis (Nigg) was grown in mycoplasma-free McCoy cells. Elementary bodies (EBs) were harvested from infected cells, resuspended in buffer (250 mM sucrose, 10 mM sodium phosphate, and 5 mM L-glutamic acid (pH 7.2)), and quantified as IFU as described previously (41). One multiplicity of infection (MOI) is equivalent to one infectious EB per cell. To infect murine macrophages, culture medium (1–2 ml) containing the required amounts of chlamydial EBs corresponding to 1 or 5 MOI were added. UV-inactivated EBs prepared by exposing purified EBs to UV light in a laminar flow hood for 3 h (12), and confirmed free of infectious organisms, were applied to wells containing macrophages in an identical fashion. Cells were spun at 1800 x g for 60 min at 37°C. Media were replaced after centrifugation to remove nonadherent chlamydial EBs, and macrophages were incubated at 37°C in a 5% CO2 incubator for various time periods. Macrophage supernatants collected at 24 h were frozen at –80°C and analyzed later for cytokines/chemokines by ELISA. To determine the level of infection, cells on coverslips were fixed with methanol for at least 10 min at 24 h postinfection (p.i.) and stained for chlamydial inclusions using the Pathfinder FITC-conjugated murine anti-chlamydial mAb according to the manufacturer’s instructions (Bio-Rad). Inclusions were viewed using an Olympus fluorescent microscope (Olympus). Infection with a MOI of 1 and 5 MoPn in macrophages results in infection of 20–30% and >80% of the cells, respectively, whereas infection of murine fibroblasts with a MOI of 0.25 results in infection of 20–30% of the cells.
RNA extraction and real-time PCR
RNA was prepared using the RNeasy kit (Qiagen). Total RNA (500 ng) was DNase treated in buffer containing 1x PCR buffer (Applied Biosystems), 3 mM MgCl2, 0.5 mM dNTP, 1 μM oligo(dT), 1 μM random hexamer, 10 U of Rnasin, and 1 U of RNase-free DNaseI (Promega) in a total volume of 20 μl, at 37°C for 15 min. DNase was inactivated at 65°C for 10 min, and reverse transcription (RT) was conducted by adding 100 U of Superscipt RT II (Invitrogen Life Technologies) to the above reaction. RT reactions without Superscript II (–RT) were used as negative controls. Reactions were conducted as per the manufacturer’s instructions (Invitrogen Life Technologies). The reaction mixture was diluted to 200 μl after RT, and 2.5 μl of cDNA mixture was used for quantitative PCR. PCR primers for -actin, TNF-, IFN-, IFN-, IFN-, and IP-10 were generated using Beacon Designer software (Bio-Rad) (Table I). Quantitative PCR was conducted using the IQ-SYBR mix (Bio-Rad) in a Bio-Rad iCycler. Reaction volumes were limited to 12.5 μl containing 2.5 μl of cDNA mix, with 50 nM of sense and antisense primers. After an initial denaturation step at 95°C for 3 min, 40 cycles of two-step PCR with 10 s of denaturation at 95°C and 1 min of annealing/extension at 60°C was used for all reactions. A melt curve was performed for all reactions to check for product integrity and primer-dimer formation. Standard curves were generated for each gene of interest using dilutions of purified PCR products at known concentrations. All PCR primers had efficiencies of >95%. PCRs using primers for -actin mRNA were conducted to provide a normalization reference. The concentrations for all genes were normalized to -actin levels in each sample, and the data are presented as normalized cDNA in picograms or femtograms.
For lung fibroblasts, RNA preparation and RT from uninfected and infected fibroblasts were conducted as described above. Quantitative PCR was performed with 1/50 of the cDNA preparation in the Mx3000P (Stratagene) in 25-μl final volumes with the Brilliant QPCR Master Mix (Stratagene). cDNA was amplified using 300 nM of each specific sense and antisense primers. We also used 100 nM of the fluorogenic oligonucleotides specific for the gene segments in which a reporter fluorescent dye on 5' (FAM) and a quencher dye on 3' (TAMRA) were attached. For -actin, the primers used were as follows: sense primer, AGAGGGAAATCGTGCGTGAC; antisense primer, CAATAGTGATGACCTGGCCGT; probe, CACTGCCGCATCCTCTTCCTCCC. For IP-10, the primers used were as follows: sense primer, GCCGTCATTTTCTGCCTCAT; antisense primer, GCTTCCCTATGGCCCTCATT; probe, TCTCGCAAGGACGGTCCGCTG. Quantitative PCR was conducted at 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, 60°C for 1 min, and 72°C for 30 s. The expression levels of IP-10 cDNA were compared with -actin and normalized to uninfected WT fibroblast responses by the comparative cycle threshold method, as described by the manufacturer (Stratagene).
Subcellular fractionation and Western blots
Macrophages harvested at 3 or 8 h p.i. were processed to isolate nuclear fractions as described previously, with some modifications (42). Briefly, 300 μl of cold buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.5 mM PMSF) were added to the macrophages, and the plates were left on ice for 15 min, followed by the addition of 20 μl of 10% Nonidet P-40. The cell lysates were transferred to a microfuge tube, vortexed for 20–30 s, and spun at 3000 rpm for 5 min at 4°C to spin down the nuclei. The cytoplasmic fraction was saved, and the nuclear pellet was washed once with buffer A containing 1% Nonidet P-40. Nuclei were lysed in 100 μl of buffer B (20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF) and rocked at 4°C for 1 h. The lysate was cleared by centrifugation at 14,000 rpm for 10 min. Protein concentration of the nuclear extracts was determined using the Bradford dye method (Bio-Rad), and 10 μg of protein were separated on a 9% SDS-PAGE. Gels were processed for Western blotting as described previously (43). The immunoblots were stained with Abs to IRF3 or actin, followed by HRP-coupled anti-rabbit or anti-mouse secondary Abs (US Biologicals). Blots were developed using the ECL chemiluminescent system (Amersham Biosciences).
Statistics
Statistical comparisons between the murine strains for level of infection and cytokine production over the course of infection were made by a two-factor (days and murine strain) ANOVA with post hoc Tukey’s test as a multiple comparison procedure. The Wilcoxon rank sum test was used to compare the duration of infection in the respective strains over time. Quantitative PCR data are provided as mean values of triplicate samples with SD, from a representative experiment. One-way ANOVA was used to analyze differences in cDNA/protein levels among various in vitro groups. SigmaStat software was used for statistical analyses (SPSS Science).
Results
IP-10 is produced independent of TLR2 and TLR4 but partly dependent on MyD88 during in vivo infection of mice with MoPn
IP-10 is a predominant chemokine induced in the upper reproductive tract in female mice during infection with MoPn (20). We have shown previously that TLR2 plays a predominant role in inflammatory cytokine (TNF-, IL-6) production in cells infected with C. trachomatis (12). However, TLR4 is the major mediator of bacterial LPS-mediated IP-10 responses, and this induction normally occurs independently of MyD88 (44, 45). To begin identifying the specific TLR responsible for IP-10 expression during chlamydial infection in vivo, IP-10 levels in genital tract secretions from MoPn-infected WT, TLR2–/–, TLR4–/–, and MyD88–/– mice, and corresponding control female mice, were analyzed by ELISA. The overall IP-10 response in infected TLR2–/– mice was not significantly different from WT mice when compared by two-way ANOVA (Fig. 1A). Likewise, the levels of IP-10 detected in secretions from TLR4–/– mice paralleled those of the control group (Fig. 1B). In contrast, levels of IP-10 detected in secretions of the MyD88–/– group were significantly lower than in the WT control group (p < 0.001, two-way ANOVA), although their IP-10 response was not abolished (Fig. 1C). Significant decreases on days 2–20 were revealed by Tukey’s test (data not shown). These findings suggest induction of IP-10 after chlamydial infection in vivo may be independent of TLR2 and TLR4 but partly dependent on MyD88. Furthermore, the levels of IP-10 in the infected MyD88–/– mice correlated with their ability to clear MoPn infection at a reduced rate compared with WT controls (Fig. 1D). These results are in contrast to the results from TLR2–/– and TLR4–/– mice, which we have recently shown to clear infection at the same rate as their respective control groups (12).
To also confirm the role of IFN- in induction of the IP-10 response during chlamydial infection, macrophages from WT mice were infected with 1 MOI of MoPn in the presence of increasing concentrations of neutralizing anti-IFN- Ab. With increasing concentration of Ab, dose-dependent decreases in IP-10 mRNA (Fig. 3C) and protein (D) were observed. These results indicate IFN- mediates IP-10 induction during chlamydial infection of macrophages.
IP-10 induction during MoPn infection of macrophages is independent of TLR2 and TLR4 but partly dependent on MyD88
TLR4, but not TLR2, has been shown to play a major role in bacterial LPS-induced IP-10, and this induction is independent of the downstream adaptor molecule MyD88 (25). To investigate the role of TLR2 and TLR4 in chlamydial-induced induction of IP-10, we compared IP-10 mRNA levels in WT, TLR2–/–, and TLR4–/– macrophages after MoPn infection (Fig. 4A). In WT and TLR2–/– macrophages, IP-10 induction was detected as early as 3 h p.i. and progressively increased over time through 24 h (Fig. 4A). MoPn-infected TLR4–/– macrophages showed induction of IP-10 at levels similar to those seen with WT macrophages (Fig. 4A). At 8 h p.i, increased levels of the IP-10 message were observed in TLR2–/– macrophages compared with WT macrophages at 5 MOI (p = 0.033 for WT vs TLR2–/– by one-way ANOVA). Results with infection at a MOI of 1 were similar (data not shown). However, by 24 h, no significant differences were observed between WT, TLR2–/–, and TLR4–/– macrophages in IP-10 mRNA levels. Under similar conditions, E. coli LPS, a TLR4 ligand, was unable to turn on IP-10 mRNA in TLR4–/– macrophages (Fig. 4A), as expected. Protein levels for IP-10 determined in supernatants paralleled the mRNA response at 24 h (data not shown). These findings indicate chlamydial-induced IP-10 is not dependent on TLR2 and TLR4.
Because MyD88 is the downstream adaptor molecule for most TLRs (49), the role of MyD88 in IP-10 induction during MoPn infection was determined using MyD88–/– macrophages. MyD88–/– and WT macrophages were infected with MoPn, and their IP-10 responses were compared. A reduction in IP-10 transcripts (33% with 1 MOI (data not shown) and 75% with 5 MOI) was observed at the 24 h time point in MyD88–/– macrophages compared with WT (Fig. 4B). The decrease in IP-10 levels in MyD88–/– macrophages was statistically significant at both 1 and 5 MOI and at all three time points tested (p < 0.001, p = 0.002, and p = 0.002, at 3, 8, and 24 h, respectively, with 5 MOI by one-way ANOVA). The levels of IP-10 protein detected in supernatants revealed an even greater effect of MyD88 deficiency on IP-10 production (p < 0.001; data not shown). However, expression of IP-10 mRNA and protein was not abolished in MyD88–/– macrophages after infection, suggesting that some of its induction (25% of WT) may be mediated through a MyD88-independent pathway. To verify that the MyD88-independent pathway is intact in MyD88–/– macrophages, E. coli LPS was used as a control. The IP-10 mRNA levels detected from MyD88–/– macrophages treated with E. coli LPS for 3 h were similar to WT macrophages, confirming that the TLR4-mediated, MyD88-independent pathway was unaffected in these cells (Fig. 4B). Thus, chlamydial-induced IP-10 production is largely but not completely dependent on MyD88.
To verify that the TLR and MyD88 dependence of IP-10 production was not limited to macrophages, the expression of the IP-10 gene was also measured in primary lung fibroblasts isolated from TLR2–/–, TLR4–/–, and MyD88–/– mice. Infection of the fibroblasts with MoPn led to a large increase in transcription of the IP-10 gene (Fig. 5A). Lower levels of transcription were found consistently in TLR2 fibroblasts compared with WT fibroblasts (Fig. 5A). This finding is in agreement with the decrease in production of IP-10 mRNA and protein in TLR2 macrophages. A smaller difference was observed between WT and TLR4 fibroblasts infected with Chlamydia (Fig. 5B). Dramatically lower levels of IP-10 gene transcription were observed in infected MyD88–/– fibroblasts (Fig. 5C).
It is important to note that the differences in cytokine and chemokine levels discussed above were not due to differences in the level of in vitro chlamydial infection of macrophages. No significant differences were observed in chlamydial inclusions between WT, TLR2–/–, TLR4–/–, and MyD88–/– macrophages (Fig. 7A). To determine whether the chlamydial inclusion developed normally in the knockout macrophages, the EBs recovered from the macrophages were used to infect a McCoy cell monolayer, and IFU were enumerated. No significant differences (by one-way ANOVA) were observed in the recovered IFU from WT, TLR2–/–, or MyD88–/– macrophages, whereas a 2- to 4-fold reduction was observed in EBs from TLR4–/– macrophages in three independent experiments (Fig. 7B) (p = 0.007, one-way ANOVA). The difference in TLR4–/– vs WT macrophages could be due to the inhibitory role of TLR4 on phagosome-lysosome fusion (50). However, this difference does not influence IP-10 levels in TLR4–/– macrophages (Fig. 6A).
IFN- induction during chlamydial infection of macrophages is independent of TLR2 and TLR4 but mostly dependent on MyD88
It is well documented that IFN- transcription is induced by TLR3 stimulation by viral dsRNA or TLR4 cross-linking by LPS (29), whereas TLR2 has not been implicated in this pathway (25). We have shown that induction of chlamydial-induced IP-10 mRNA was dependent on IFN- (Fig. 3). Analysis of IFN- levels in infected TLR2–/– and TLR4–/– macrophages showed that the differences in IFN- mRNA levels in infected WT, TLR4–/–, and TLR2–/– macrophages were not significant (Fig. 8A). Thus, chlamydial-induced IFN- occurred independent of TLR2 or TLR4. However, analysis of IFN- mRNA in MoPn-infected MyD88–/– macrophages showed a significant reduction at all time points (3 h, p = 0.002; 8 and 24 h, p < 0.001; one-way ANOVA) compared with WT macrophages. A 70–80% inhibition of IFN- mRNA transcription was observed in MyD88–/– macrophages compared with WT macrophages at an infection of 5 MOI (Fig. 8B), whereas a 30–40% inhibition was observed with a MOI of 1 (data not shown). The E. coli LPS-mediated IFN- response was significantly elevated in MyD88–/– macrophages compared with WT macrophages, most likely because of an increased use of the Toll/IL-1R domain-containing adaptor inducing INF--mediated pathway in the absence of MyD88, a phenomenon observed previously for LPS-stimulated cells (31). Assays for detecting IFN- protein levels were not successful possibly due to low sensitivity of the Abs used. These data indicate that TLR2 and TLR4 do not contribute to chlamydial-induced IFN- in macrophages, but a MyD88-dependent stimulatory pathway is involved. However, it must be noted that although IFN- mRNA levels are reduced in MyD88–/– macrophages compared with WT macrophages, significant induction of IFN- mRNA still occurs in the absence of MyD88, in response to infection (Fig. 8B). This suggests chlamydiae stimulate other, as yet uncharacterized, MyD88-independent pathways. Nonetheless, chlamydial-induced induction of IFN- is largely dependent on MyD88.
MyD88-dependent IFN pathways and the role for endosomal maturation in IFN- synthesis during chlamydial infection
Data from the above studies indicate a prominent role for MyD88 in MoPn-induced IFN- in macrophages, suggesting involvement of intracellular PRRs. Intracellular TLRs such as TLR7, TLR8, and TLR9 have been shown to induce IFN- in a MyD88-dependent manner (27, 28). TLR7, TLR8, and TLR9 form a subgroup within the TLR family due to their exclusive intracellular localization and recognition of PAMP in endosomal/lysosomal compartments (51, 52). Activation of these receptors depends on acidification and maturation of endosomes. Furthermore, during their activation, MyD88 is targeted to vesicular structures with lysosomal characteristics (52).
Based on these reports, we postulated that disruption of endosomal maturation would influence intracellular TLR-dependent IFN- induction during chlamydial infection. To address the role of intracellular TLRs in chlamydial-induced IFN-, MoPn-infected macrophages were treated with varying concentrations of bafilomycin A1 during infection. Bafilomycin A1, an acidification inhibitor, is traditionally used to block endosomal maturation (53). A dose-dependent reduction in the induction of IFN- mRNA and IP-10 protein (Fig. 9, A and B) was observed in bafilomycin-treated cells. As seen in previous experiments, UV-inactivated MoPn failed to induce any response regardless of bafilomycin treatment.
In contrast to IFN- and IP-10 levels, bafilomycin treatment of MoPn-infected cells did not have a negative effect on induction of TNF-. In fact, TNF- mRNA levels were increased at higher doses of bafilomycin (Fig. 9C). Thus, bafilomycin treatment did not cause a global inhibition of chlamydial-induced stimulatory pathways but was specific for the IFN- pathway. Bafilomycin treatment also did not affect chlamydial inclusion formation or recovery of infectious EBs in a significant manner (data not shown). Thus, the differences observed in induction of IFN- and IP-10 is not due to differences in the rate of infection. These results imply a role for intracellular PRRs that recognize chlamydial PAMP in endosomes to induce IFN- synthesis.
Role of IRF3 and TBK in chlamydial-induced IFN- induction
IRF3 is a necessary transcription factor for inducible IFN- (reviewed in Ref. 54). An important step preceding the induction of IFN- transcription is the phosphorylation of the transcription factor IRF3 by the kinases TBK and IKK (55, 56). TLR3 and TLR4, after binding their respective ligands, activate IRF3 by these kinases. As a result, IRF3 gets phosphorylated and transported to the nucleus to induce IFN- gene expression (57). By a complex mechanism that involves MyD88, the receptors TLR7, TLR8, and TLR9 can also activate these kinases to phosphorylate IRF3 (58). To begin elucidating the molecular mechanism of chlamydial-induced IFN- gene expression, nuclear translocation of IRF3 was investigated using MoPn-infected macrophages.
Nuclear extracts were prepared from control and infected macrophages at 3 and 8 h p.i., and IRF3 translocation was analyzed by Western blot (Fig. 10A). IRF3 nuclear translocation was observed as early as 3 h and increased at 8 h p.i. (Fig. 10A). The blots were stripped and reprobed for actin protein that served as a loading control and showed equivalent protein amounts in all lanes (Fig. 10A). These results demonstrate that IRF3 translocates to the nucleus during chlamydial infection and suggests its role in induction of IFN-.
As described previously, MyD88-dependent and independent signaling pathways that lead to IFN- induction converge in phosphorylation of IRF3. Two kinases that can function interchangeably have been implicated in phosphorylation of IRF3: TBK (55) and IKK (56). TBK is expressed in all cells, whereas IKK is present only in immune cells (56). To confirm the role of IRF3 phosphorylation in chlamydial-induced IFN-, TBK–/– MEFs were used. TBK–/– MEFs were infected with MoPn, and their IFN- expression was compared with WT MEFs. IFN- and IP-10 expression was completely abolished in MoPn-infected TBK–/– MEFs compared with WT MEFs, suggesting a complete dependence on the TBK-mediated pathway (Fig. 10B). It must be noted that in fibroblasts, only TBK plays a major role in IRF3 phosphorylation, whereas in macrophages, both TBK and IKK play a role interchangeably (56), and the above experiments do not take into account the role of IKK. However, these results demonstrate collectively that IRF3 phosphorylation and nuclear translocation play an important role in regulating induction of IFN- and, in turn, IP-10 during chlamydial infection. Together, our results demonstrate that chlamydial-induced IFN- results from activation of a MyD88-dependent pathway that functions through TBK-mediated IRF3 phosphorylation.
Discussion
Primary chlamydial genital infection is mostly resolved by T cells that are recruited to the site of infection in response to T cell chemokines, such as IP-10. Induction of the chemokines in turn results from interaction of host TLRs with chlamydial PAMP. In this study, we show that C. trachomatis (MoPn) infection, in vivo and in vitro, induces IP-10 and IFN- independently of TLR2 and TLR4. Furthermore, we demonstrate a prominent role for the downstream adaptor molecule MyD88 in induction of IP-10 and IFN- during infection. This is the first report of an intracellular bacterium inducing IFN- in a TLR4-independent and MyD88-dependent manner. In addition, we have found that the intracellular mechanisms inducing IFN- production during chlamydial infection involve TBK-mediated IRF3 nuclear translocation, mediated possibly by a PRR that requires endosomal maturation.
During our studies of in vivo genital infection of TLR2–/– and TLR4–/– mice with MoPn, we observed that IP-10 levels in the genital tract secretions are independent of TLR4. This was surprising, because IP-10 responses during infections with Gram-negative bacteria are normally mediated through TLR4 stimulation by bacterial LPS. Recent studies have demonstrated the importance of IP-10 in generation of effective innate and adaptive immune responses to genital infection of mice with MoPn (59, 60, 61). Our study demonstrates that, in agreement with the lower levels of IP-10 in genital secretions of MyD88–/– mice during MoPn infection, the MyD88–/– mice clear infection at a slower rate than control mice. This is in contrast to our previous results obtained with TLR2–/– and TLR4–/– mice, which clear infection at the same rate as the control groups (12). However, the exact cell type(s) producing IP-10 or responding to IP-10 in vivo is not known. Moreover, the levels of IP-10 in vivo could result from both type I and type II IFN levels, and the specific contribution of type I IFNs has yet to be established using type I IFNs knockout mice.
Analysis of IFN- induction during chlamydial infection in vitro provides a direct means for studying the TLR intracellular signaling pathways in response to infection at a cellular level. Although more work will be needed to correlate the findings of the in vitro experiments with the in vivo results, we have observed that in vitro responses such as TNF- and IL-6 production in response to chlamydial infection using TLR2–/– and TLR4–/– macrophages and fibroblasts appear to parallel the in vivo responses (12). Other in vitro studies demonstrating a role for TLRs in cytokine/chemokine expression in chlamydial infection using C. pneumoniae showed a similar correlation with in vivo results (62).
Nonetheless, significant variations in responses to different chlamydial strains have been reported. Two independent groups (62, 63) have shown a predominant role for TLR2 in C. pneumoniae-induced activation of DCs and macrophages using strains CM-1 and TW-183, respectively, whereas the findings from another group using the AR39 strain of C. pneumoniae have suggested a predominant role for TLR4 (64). Recently, Rothfuchs et al. (47) have shown, using BMDMs infected with a clinical isolate of C. pneumoniae (Kajaani 6), a reduction in IFN- and IFN- mRNA accumulation in TLR4–/– macrophages and no change in the levels of TNF- in MyD88–/– macrophages. Our results using the MoPn strain of C. trachomatis are in contrast to these findings. These differences could be due to innate differences between the structural components of C. pneumoniae and C. trachomatis, or to the use of BMDMs in previous studies vs the peritoneal macrophages used in this study. To address the difference in cell type used, we infected BMDMs from MyD88–/– and WT mice with C. trachomatis. The results for TNF- and IFN- induction in infected WT vs MyD88 BMDMs were similar to those obtained for peritoneal macrophages (data not shown), suggesting that the differences between the C. pneumoniae and C. trachomatis results may be attributed to inherent differences in the pathogens and not the cell types used for analysis.
The specific role of TLR4 in IFN- induction is well established (25). As a result, E. coli LPS-stimulated MyD88–/– cells remain fully able to promote IFN-inducible genes such as IP-10, GARG-16, or IRG-1. Likewise, Yersinia enterocolitica, Bacillus anthracis, or the sip B mutant of Salmonella typhimurium mediate type I IFN induction through stimulation of TLR4 (65, 66). But we have observed that C. trachomatis-induced IFN- and the IFN response gene IP-10 are stimulated independently of TLR4. Furthermore, 70–80% of the IFN- induced during chlamydial infection is MyD88 dependent (Fig. 8B). The intracellular TLRs (TLR7, TLR8, and TLR9) are the only TLRs known to induce type I IFNs in a MyD88-dependent manner (28, 51), suggesting a role for these TLRs in chlamydial-induced type I IFNs. The intracellular TLRs require endosomal maturation to function (51). We therefore explored the possibility that they may be involved in chlamydial-induced type I IFN, using inhibitors of endosomal maturation, and found that they may in fact play a role. However, we cannot rule out a role for other PRRs in macrophages and fibroblasts, which may function in a MyD88-dependent manner and require endosomal maturation, because TLR7 and TLR9 are expressed at high levels only in plasmocytoid DCs (67, 68). Moreover, although most of the IFN- induction during MoPn infection was MyD88 dependent, a complete inhibition of IFN- induction in MyD88–/– macrophages was not observed, suggesting some stimulation through chlamydial-mediated MyD88-independent pathways for IFN- induction. Besides TLR4, MyD88-independent induction of IFN- may be stimulated by a number of other candidate receptors. One of them is TLR3, a viral dsRNA-recognizing TLR known to activate the IFN- pathway, the role of which has not been studied during chlamydial infection. In addition to the surface and endosomal TLRs, the nucleotide-binding oligomerization domain family proteins (69) or the RNA helicase RIG-1 (70) that are present in the cytoplasm of macrophages could also function in a MyD88-independent manner to induce IFN-, although these PRRs are not expected to be sensitive to inhibitors of endosomal maturation. Therefore, we speculate that the fraction of MyD88-independent, IFN- induction observed during chlamydial infection could be mediated by any of these unexplored receptors.
IFN- has been traditionally considered an antiviral cytokine (24), and the intracellular mechanisms regulating its expression have been studied extensively in response to viral infection. Induction of the IFN- gene by bacteria has attracted attention only recently, because of the pleiotropic functions of type I IFNs in immunity (71). We present evidence that chlamydial-induced IFN- is mediated by TBK-dependent nuclear translocation of IRF3 (Fig. 9). Recently, Stockinger et al. (72) and O’Connell et al. (73) have shown the requirement for IRF3 and TBK in type I IFN induction during infection of macrophages by another intracellular pathogen, Listeria monocytogenes. In this respect, intracellular bacterial infection begins to resemble viral infection. However, in contrast to our observations with Chlamydia, up-regulation of type I IFN in macrophages by Listeria was not only TLR4 independent but also MyD88 independent (72). One possibility for these differences could be the intracellular location of Chlamydia in inclusions compared with Listeria, which resides in the cytosol. Therefore, depending on their intracellular location, the bacteria may use different pathways to activate IRF3. Besides IRF3, another key transcription factor that plays a role in type I IFN induction is IRF7 (74). Two groups (58, 75) have recently shown that IRF7 interacts with MyD88 to form a complex that also involves IRAK4 and TRAF6 (components of NF-B activation). Together they provide the foundation for intracellular TLR-mediated IFN induction. This would suggest that in MoPn-infected MyD88–/– macrophages, nuclear IRF7 levels could be reduced significantly. Future studies will therefore need to investigate the levels of nuclear IRF3 and IRF7 in WT and MyD88–/– macrophages to determine the relative role of IRF3 and IRF7 in chlamydial-induced IFN- induction.
The physiological function of type I IFNs as a link between innate and adaptive immunity is well established. However, the physiological role of type I IFNs during chlamydial infection in vivo has not been investigated. The signaling pathways that lead to IFN- induction could be of major importance in determining the course of infection or pathology in vivo. Unexpectedly, the pathways of IFN- induction during chlamydial infection turn out to be more complex than the expected LPS-TLR4-mediated response. In summary, our findings indicate that IFN- and subsequent IP-10 induction during MoPn infection occurs through two pathways. The major pathway requires MyD88-dependent activation, whereas a minor pathway involves MyD88- independent activation; both function through IRF3 phosphorylation. Determining the specific PRR(s) that contributes to the induction of IFN- during chlamydial infection is the next critical step in understanding the ensuing cytokine/chemokine response.
Acknowledgments
We thank Dr. Wen-Chen Yeh (University Health Network, Toronto, Ontario, Canada) for providing TBK–/– MEFs. Technical help by Jim Sikes and Anne Bowlin on ELISA, animal husbandry, and Chlamydia stocks is gratefully acknowledged. We also thank Dr. Shanmugam Nagarajan for critical reading of this manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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 This work was supported by the Horace C. Cabe and the Bates-Wheeler Foundations, the Arkansas Children’s Hospital Research Institute, a Pilot grant award from the University of Arkansas for Medical Sciences, the Medical Research Endowment, University of Arkansas for Medical Sciences award to U.M.N., and Public Health Service Grant AI054624 from the National Institutes of Health to T.D.
2 Address correspondence and reprint requests to Dr. Uma M. Nagarajan, Department of Microbiology and Immunology, 4301 West Markham Street, BM506, University of Arkansas for Medical Science, Little Rock, AR 72205. E-mail address: nagarajanuma@uams.edu
3 Abbreviations used in this paper: PAMP, pathogen-associated molecular pattern; IP-10, IFN--inducible protein 10; EB, elementary body; MoPn, mouse pneumonitis agent of Chlamydia trachomatis; MOI, multiplicity of infection; PGN, peptidoglycan; p.i., postinfection; PRR, pattern recognition receptor; IRF3, IFN response factor-3; TBK, TANK-binding kinase; WT, wild type; DC, dendritic cell; IFU, inclusion-forming unit; BMDM, bone marrow-derived macrophage; RT, reverse transcription; MEF, murine embryonic fibroblast.
Received for publication January 12, 2005. Accepted for publication April 19, 2005.
References
Gerbase, A. C., J. T. Rowley, T. E. Mertens. 1998. Global epidemiology of sexually transmitted diseases. Lancet 351:(Suppl. 3): 2-4.
Swenson, C. E., J. Schachter. 1984. Infertility as a consequence of chlamydial infection of the upper genital tract in female mice. Sex. Transm. Dis. 11: 64-67.
Cain, T. K., R. G. Rank. 1995. Local Th1-like responses are induced by intravaginal infection of mice with the mouse pneumonitis biovar of Chlamydia trachomatis. Infect. Immun. 63: 1784-1789.
Ramsey, K. H., R. G. Rank. 1991. Resolution of chlamydial genital infection with antigen-specific T-lymphocyte lines. Infect. Immun. 59: 925-931.
Rank, R. G., K. H. Ramsey, E. A. Pack, D. M. Williams. 1992. Effect of interferon on resolution of murine chlamydial genital infection. Infect. Immun. 60: 4427-4429.
Perry, L. L., K. Feilzer, H. D. Caldwell. 1997. Immunity to Chlamydia trachomatis is mediated by T helper 1 cells through IFN--dependent and -independent pathways. J. Immunol. 158: 3344-3352.
Cotter, T. W., K. H. Ramsey, G. S. Miranpuri, C. E. Poulsen, G. I. Byrne. 1997. Dissemination of Chlamydia trachomatis chronic genital tract infection in interferon gene knockout mice. Infect. Immun. 65: 2145-2152.
Takeda, K., T. Kaisho, S. Akira. 2003. Toll-like receptors. Annu. Rev. Immunol. 21: 335-376.
Spitzer, J. H., A. Visintin, A. Mazzoni, M. N. Kennedy, D. M. Segal. 2002. Toll-like receptor 1 inhibits Toll-like receptor 4 signaling in endothelial cells. Eur. J. Immunol. 32: 1182-1187.
Shiraki, R., N. Inoue, S. Kawasaki, A. Takei, M. Kadotani, Y. Ohnishi, J. Ejiri, S. Kobayashi, K. Hirata, S. Kawashima, M. Yokoyama. 2004. Expression of Toll-like receptors on human platelets. Thromb. Res. 113: 379-385.
Barton, G. M., R. Medzhitov. 2003. Toll-like receptor signaling pathways. Science 300: 1524-1525.
Darville, T., J. M. O’Neill, C. W. Andrews, Jr, U. M. Nagarajan, L. Stahl, D. M. Ojcius. 2003. Toll-like receptor-2, but not Toll-like receptor-4, is essential for development of oviduct pathology in chlamydial genital tract infection. J. Immunol. 171: 6187-6197.(Uma M. Nagarajan2,*,, Dav)
IFN--inducible protein 10 (IP-10) is a chemokine important in the attraction of T cells, which are essential for resolution of chlamydial genital tract infection. During infections with Gram-negative bacteria, the IP-10 response mediated through type I IFNs usually occurs as a result of TLR4 stimulation by bacterial LPS. However, we found that levels of IP-10 in genital tract secretions of Chlamydia trachomatis-infected female wild-type mice were similar to those of infected TLR2- and TLR4-deficient mice but significantly greater than those of infected MyD88-deficient mice. We investigated the mechanism of IP-10 and IFN- induction during chlamydial infection using mouse macrophages and fibroblasts infected ex vivo. The induction of IP-10 and IFN- was unchanged in Chlamydia-infected TLR2- and TLR4-deficient cells compared with wild-type cells. However, infection of MyD88-deficient cells resulted in significantly decreased responses. These results suggest a role for MyD88-dependent pathways in induction of IP-10 and IFN- during chlamydial infection. Furthermore, treatment of infected macrophages with an endosomal maturation inhibitor significantly reduced chlamydial-induced IFN-. Because endosomal maturation is required for MyD88-dependent intracellular pathogen recognition receptors to function, our data suggest a role for the intracellular pathogen recognition receptor(s) in induction of IFN- and IP-10 during chlamydial infection. Furthermore, the intracellular pathways that lead to chlamydial-induced IFN- function through TANK-binding kinase mediated phosphorylation and nuclear translocation of IFN regulatory factor-3.
Introduction
Chlamydia trachomatis is a common sexually transmitted bacterial pathogen (1) that can cause oviduct inflammation and subsequent tubal infertility (2). The first line of defense against chlamydiae is the innate immune response. The cytokines/chemokines produced during the innate response recruit inflammatory cells and T cells that are needed for adaptive immunity. The T cell-dependent adaptive immune response and subsequent chlamydial clearance is predominated by IL-12- and IFN--dependent mechanisms (3, 4, 5, 6, 7).
One of the major players of the innate immune response to pathogens is the TLR, which is expressed on immune cells (reviewed in Ref. 8) as well as multiple other cell types (9, 10). The recognition of pathogen-associated molecular patterns (PAMP)3 by TLRs leads to signaling events and coordinated activation of transcription factors that induce the expression of antimicrobial molecules, chemokines, cytokines, and costimulatory molecules (11). Therefore, elucidating the TLR involved in initiation of the immune response to chlamydial infection is clearly a primary step to understanding its molecular pathogenesis.
We reported previously that Tlr2 knockout mice (TLR2–/–) infected genitally with the mouse pneumonitis (MoPn) agent of C. trachomatis (also known as C. muridarum) had significantly lower inflammatory cytokine responses, including TNF- and MIP-2, and less oviduct pathology compared with infected wild-type (WT) mice (12). Despite lower cytokine responses, the TLR2–/– mice cleared infection at a rate similar to WT mice. Tlr4 knockout mice (TLR4–/–) infected with MoPn responded like WT mice with respect to both cytokine response and clearance of infection. Because it is well documented that T cells play a major role in clearing Chlamydia infection (7, 13, 14), the results from the TLR2–/– and TLR4–/– mice suggest the influx of T cells to the site of infection was unaffected.
IFN--inducible protein 10 (IP-10) is an important chemokine involved in T cell recruitment (15). IP-10 is induced by both type I (IFN- and IFN-) and type II IFN (16, 17). Besides its role in inducing IP-10, type I IFNs also induce expression of MCP-5, RANTES, and GARG-16 and has been implicated in Th1 maturation (18). IFN- has also been shown to augment dendritic cell (DC) expression of IL-12 and CD40, both essential in generating robust T cell responses (19). It has been reported that IP-10 is produced in the upper and lower genital tract of mice infected with chlamydiae (20). Furthermore, murine DC pulsed in vitro with C. trachomatis (21) and a murine oviduct epithelial cell line have been reported to up-regulate expression of IP-10 during MoPn infection (22). It has also been shown that IFN- is induced in McCoy cells during C. trachomatis infection (23). However, the molecular pathways involved in IP-10 and IFN- induction during chlamydial infection have not yet been described.
TLRs play a major role in induction of type I IFN, in response to bacterial and viral infection (reviewed in Ref. 24). Cross-linking TLR3 and TLR4 by viral dsRNA and LPS, respectively, has been shown to induce IFN- and IFN- (25, 26). Type I IFNs are also induced when TLR7 and TLR9 bind to their ligands, ssRNA and bacterial CpG DNA, respectively (27, 28). However, the signaling pathways by which the different TLRs mediate induction of type I IFNs differ distinctly. TLR3- and TLR4-mediated type I IFN induction is independent of the central downstream adaptor molecule MyD88 (reviewed in Ref. 29), whereas TLR7 and TLR9 mediate induction of type I IFNs in a MyD88-dependent manner (27, 28). However, both pathways lead to phosphorylation of IFN regulatory factor-3 (IRF3), a transcription factor essential for IFN- production by the TANK-binding kinase (TBK) (30).
In this study, the role of IFN- in the induction of IP-10 during MoPn (C. trachomatis) infection was established in vitro, and the contribution of TLR in IP-10 and IFN- induction was explored using murine macrophages and fibroblasts. Chlamydial-induced IP-10 and IFN- were found to be completely independent of TLR2 and TLR4 but significantly dependent on MyD88. The TLR4 independence of chlamydial-induced IFN- expression was surprising, because TLR4 is the major receptor that binds bacterial LPS to induce IFN- (31). Preliminary evidence using an endosomal maturation inhibitor suggests a role for intracellular TLR or other pathogen recognition receptors (PRRs) that function in a MyD88-dependent manner to induce IFN- during chlamydial infection. Furthermore, the role of TBK-mediated phosphorylation of IRF3 and its nuclear translocation for IFN- induction was established.
Materials and Methods
Reagents and immunochemicals
Escherichia coli LPS (InvivoGen) and Staphylococcus aureus peptidoglycan (PGN; Lee Labs) were used as positive controls for TLR4 and TLR2 stimulation, respectively. Although PGN is no longer considered a ligand for TLR2, contaminants in the S. aureus PGN preparation are known to be strong agonists for TLR2 (32). Bafilomycin was obtained from Sigma-Aldrich. Anti-mouse IFN- (PBL Biomedical Laboratories), anti-mouse actin (Sigma Immunochemicals), and anti-mouse IRF3 (Zymed Laboratories) Abs were purchased as indicated. All ELISA kits were from R&D Laboratories.
Animals and cell lines
Tlr2 gene knockout mice (TLR2–/–) were kind gifts from Dr. S. Akira (Department of Host Defense, Research Institute for Microbial Diseases, Osaka, Japan) (33). Because they have not been sufficiently backcrossed to the C57BL/6 background, the heterozygous F1 generations from a cross of C57BL/6 and TLR2–/– mice were used as controls for TLR2–/– mice. However, because there were no differences in infection rates or cytokine levels between C57BL/6 mice and F1 progeny of C57BL/6 x TLR2–/– mice (12), C57BL/6 mice were used as controls for macrophage experiments. MyD88 gene knockout mice or MyD88–/– mice (originally from Dr. Akira) (34) were purchased from the Centre de La Recherche Scientifique and have been previously backcrossed with C57BL/6 mice over 10 generations (35, 36); therefore, C57BL/6 mice were used as controls for MyD88–/– mice. TLR4lps-del (hereafter referred to as TLR4–/–) mice have a deletion in the TLR4 gene (C57BL/10SCN) and were obtained from The Jackson Laboratory. C57BL/10 mice (The Jackson Laboratory) or C57BL/6 mice were used as controls for TLR4–/– in vivo experiments, because they displayed similar infection kinetics. IFN- receptor–/– mice (A129) and their corresponding WT controls (129Sv/Ev) were obtained from B & K Universal. Mice between 8 and 12 wk of age were used and were age matched for all experiments. Mice were given food and water ad libitum in an environmentally controlled room with 12-h light and 12-h dark cycles. All animal experiments were preapproved by the Institutional Animal Care and Use Committee of the University of Arkansas Medical Sciences. WT and TBK–/– mouse embryonic fibroblasts were kind gifts from Dr. W.-C. Yeh (University Health Network, Toronto, Ontario, Canada) (37) and were maintained in DMEM with 10% FBS, 1 mM glutamine, and 50 μg/ml gentamicin.
Murine genital infection with MoPn and analysis of IP-10 in genital secretions
Mice received 2.5 mg of Depoprovera (medroxyprogesterone acetate) in 0.1 ml of saline (Upjohn) s.c., 7 days before vaginal infection. Mice were anesthetized using sodium pentobarbital and infected by placing MoPn (1 x 107 inclusion-forming units (IFU) in 30 μl of buffer (250 mM sucrose, 10 mM sodium phosphate, and 5 mM L-glutamic acid (pH 7.2)) into the vaginal vault. Mice were monitored by swabbing the vaginal vault and cervix with a calcium alginate swab (Spectrum Medical Industries) at various times after infection and by enumerating IFU using McCoy cell monolayers (38). Mice were infected in groups of five, and each experiment was repeated at least once. Genital tract secretions were collected as described previously (12), and IP-10 levels in individual genital tract sponge eluates were determined by ELISA.
Isolation of murine macrophages and lung fibroblasts
Peritoneal macrophages were obtained from mice that were given injections of 3% thioglycolate medium 3 days before harvesting. To harvest macrophages, mice were killed, and their peritoneal cavities were washed three times with culture medium (RPMI 1640 containing 10% FBS, 2 mM glutamine, 10 mM HEPES (pH 7.4), 100 μM nonessential amino acids, 1 mM sodium pyruvate, 50 μM 2-ME, and 50 μg/ml gentamicin). The macrophages in culture medium were plated either in 24-well plates (7 x 105 cells/well) containing coverslips to stain for chlamydial inclusions or in 6-well plates (2.5 x 106 cells/well) for RNA preparations. Mouse peritoneal macrophages collected after thioglycolate treatment were kept in culture conditions for 24–48 h before infection. Bone marrow-derived macrophages (BMDMs) were prepared by flushing the femur and tibia of mice using a 27.5-gauge needle and culturing the cells as described by Kambayashi et al. (39) in the presence of IL-4 (10 ng/ml) and GM-CSF (100 ng/ml) for 5 days at 37°C in a 7.5% CO2 incubator. After removing the nonadherent DCs, the adherent macrophages were used for infection. Lung fibroblasts from WT or TLR-deficient mice were isolated as described previously (12, 40) and cultured for 5–6 days before infecting with C. trachomatis.
Chlamydial stocks and in vitro infection of macrophages
The MoPn agent of C. trachomatis (Nigg) was grown in mycoplasma-free McCoy cells. Elementary bodies (EBs) were harvested from infected cells, resuspended in buffer (250 mM sucrose, 10 mM sodium phosphate, and 5 mM L-glutamic acid (pH 7.2)), and quantified as IFU as described previously (41). One multiplicity of infection (MOI) is equivalent to one infectious EB per cell. To infect murine macrophages, culture medium (1–2 ml) containing the required amounts of chlamydial EBs corresponding to 1 or 5 MOI were added. UV-inactivated EBs prepared by exposing purified EBs to UV light in a laminar flow hood for 3 h (12), and confirmed free of infectious organisms, were applied to wells containing macrophages in an identical fashion. Cells were spun at 1800 x g for 60 min at 37°C. Media were replaced after centrifugation to remove nonadherent chlamydial EBs, and macrophages were incubated at 37°C in a 5% CO2 incubator for various time periods. Macrophage supernatants collected at 24 h were frozen at –80°C and analyzed later for cytokines/chemokines by ELISA. To determine the level of infection, cells on coverslips were fixed with methanol for at least 10 min at 24 h postinfection (p.i.) and stained for chlamydial inclusions using the Pathfinder FITC-conjugated murine anti-chlamydial mAb according to the manufacturer’s instructions (Bio-Rad). Inclusions were viewed using an Olympus fluorescent microscope (Olympus). Infection with a MOI of 1 and 5 MoPn in macrophages results in infection of 20–30% and >80% of the cells, respectively, whereas infection of murine fibroblasts with a MOI of 0.25 results in infection of 20–30% of the cells.
RNA extraction and real-time PCR
RNA was prepared using the RNeasy kit (Qiagen). Total RNA (500 ng) was DNase treated in buffer containing 1x PCR buffer (Applied Biosystems), 3 mM MgCl2, 0.5 mM dNTP, 1 μM oligo(dT), 1 μM random hexamer, 10 U of Rnasin, and 1 U of RNase-free DNaseI (Promega) in a total volume of 20 μl, at 37°C for 15 min. DNase was inactivated at 65°C for 10 min, and reverse transcription (RT) was conducted by adding 100 U of Superscipt RT II (Invitrogen Life Technologies) to the above reaction. RT reactions without Superscript II (–RT) were used as negative controls. Reactions were conducted as per the manufacturer’s instructions (Invitrogen Life Technologies). The reaction mixture was diluted to 200 μl after RT, and 2.5 μl of cDNA mixture was used for quantitative PCR. PCR primers for -actin, TNF-, IFN-, IFN-, IFN-, and IP-10 were generated using Beacon Designer software (Bio-Rad) (Table I). Quantitative PCR was conducted using the IQ-SYBR mix (Bio-Rad) in a Bio-Rad iCycler. Reaction volumes were limited to 12.5 μl containing 2.5 μl of cDNA mix, with 50 nM of sense and antisense primers. After an initial denaturation step at 95°C for 3 min, 40 cycles of two-step PCR with 10 s of denaturation at 95°C and 1 min of annealing/extension at 60°C was used for all reactions. A melt curve was performed for all reactions to check for product integrity and primer-dimer formation. Standard curves were generated for each gene of interest using dilutions of purified PCR products at known concentrations. All PCR primers had efficiencies of >95%. PCRs using primers for -actin mRNA were conducted to provide a normalization reference. The concentrations for all genes were normalized to -actin levels in each sample, and the data are presented as normalized cDNA in picograms or femtograms.
For lung fibroblasts, RNA preparation and RT from uninfected and infected fibroblasts were conducted as described above. Quantitative PCR was performed with 1/50 of the cDNA preparation in the Mx3000P (Stratagene) in 25-μl final volumes with the Brilliant QPCR Master Mix (Stratagene). cDNA was amplified using 300 nM of each specific sense and antisense primers. We also used 100 nM of the fluorogenic oligonucleotides specific for the gene segments in which a reporter fluorescent dye on 5' (FAM) and a quencher dye on 3' (TAMRA) were attached. For -actin, the primers used were as follows: sense primer, AGAGGGAAATCGTGCGTGAC; antisense primer, CAATAGTGATGACCTGGCCGT; probe, CACTGCCGCATCCTCTTCCTCCC. For IP-10, the primers used were as follows: sense primer, GCCGTCATTTTCTGCCTCAT; antisense primer, GCTTCCCTATGGCCCTCATT; probe, TCTCGCAAGGACGGTCCGCTG. Quantitative PCR was conducted at 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, 60°C for 1 min, and 72°C for 30 s. The expression levels of IP-10 cDNA were compared with -actin and normalized to uninfected WT fibroblast responses by the comparative cycle threshold method, as described by the manufacturer (Stratagene).
Subcellular fractionation and Western blots
Macrophages harvested at 3 or 8 h p.i. were processed to isolate nuclear fractions as described previously, with some modifications (42). Briefly, 300 μl of cold buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.5 mM PMSF) were added to the macrophages, and the plates were left on ice for 15 min, followed by the addition of 20 μl of 10% Nonidet P-40. The cell lysates were transferred to a microfuge tube, vortexed for 20–30 s, and spun at 3000 rpm for 5 min at 4°C to spin down the nuclei. The cytoplasmic fraction was saved, and the nuclear pellet was washed once with buffer A containing 1% Nonidet P-40. Nuclei were lysed in 100 μl of buffer B (20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF) and rocked at 4°C for 1 h. The lysate was cleared by centrifugation at 14,000 rpm for 10 min. Protein concentration of the nuclear extracts was determined using the Bradford dye method (Bio-Rad), and 10 μg of protein were separated on a 9% SDS-PAGE. Gels were processed for Western blotting as described previously (43). The immunoblots were stained with Abs to IRF3 or actin, followed by HRP-coupled anti-rabbit or anti-mouse secondary Abs (US Biologicals). Blots were developed using the ECL chemiluminescent system (Amersham Biosciences).
Statistics
Statistical comparisons between the murine strains for level of infection and cytokine production over the course of infection were made by a two-factor (days and murine strain) ANOVA with post hoc Tukey’s test as a multiple comparison procedure. The Wilcoxon rank sum test was used to compare the duration of infection in the respective strains over time. Quantitative PCR data are provided as mean values of triplicate samples with SD, from a representative experiment. One-way ANOVA was used to analyze differences in cDNA/protein levels among various in vitro groups. SigmaStat software was used for statistical analyses (SPSS Science).
Results
IP-10 is produced independent of TLR2 and TLR4 but partly dependent on MyD88 during in vivo infection of mice with MoPn
IP-10 is a predominant chemokine induced in the upper reproductive tract in female mice during infection with MoPn (20). We have shown previously that TLR2 plays a predominant role in inflammatory cytokine (TNF-, IL-6) production in cells infected with C. trachomatis (12). However, TLR4 is the major mediator of bacterial LPS-mediated IP-10 responses, and this induction normally occurs independently of MyD88 (44, 45). To begin identifying the specific TLR responsible for IP-10 expression during chlamydial infection in vivo, IP-10 levels in genital tract secretions from MoPn-infected WT, TLR2–/–, TLR4–/–, and MyD88–/– mice, and corresponding control female mice, were analyzed by ELISA. The overall IP-10 response in infected TLR2–/– mice was not significantly different from WT mice when compared by two-way ANOVA (Fig. 1A). Likewise, the levels of IP-10 detected in secretions from TLR4–/– mice paralleled those of the control group (Fig. 1B). In contrast, levels of IP-10 detected in secretions of the MyD88–/– group were significantly lower than in the WT control group (p < 0.001, two-way ANOVA), although their IP-10 response was not abolished (Fig. 1C). Significant decreases on days 2–20 were revealed by Tukey’s test (data not shown). These findings suggest induction of IP-10 after chlamydial infection in vivo may be independent of TLR2 and TLR4 but partly dependent on MyD88. Furthermore, the levels of IP-10 in the infected MyD88–/– mice correlated with their ability to clear MoPn infection at a reduced rate compared with WT controls (Fig. 1D). These results are in contrast to the results from TLR2–/– and TLR4–/– mice, which we have recently shown to clear infection at the same rate as their respective control groups (12).
To also confirm the role of IFN- in induction of the IP-10 response during chlamydial infection, macrophages from WT mice were infected with 1 MOI of MoPn in the presence of increasing concentrations of neutralizing anti-IFN- Ab. With increasing concentration of Ab, dose-dependent decreases in IP-10 mRNA (Fig. 3C) and protein (D) were observed. These results indicate IFN- mediates IP-10 induction during chlamydial infection of macrophages.
IP-10 induction during MoPn infection of macrophages is independent of TLR2 and TLR4 but partly dependent on MyD88
TLR4, but not TLR2, has been shown to play a major role in bacterial LPS-induced IP-10, and this induction is independent of the downstream adaptor molecule MyD88 (25). To investigate the role of TLR2 and TLR4 in chlamydial-induced induction of IP-10, we compared IP-10 mRNA levels in WT, TLR2–/–, and TLR4–/– macrophages after MoPn infection (Fig. 4A). In WT and TLR2–/– macrophages, IP-10 induction was detected as early as 3 h p.i. and progressively increased over time through 24 h (Fig. 4A). MoPn-infected TLR4–/– macrophages showed induction of IP-10 at levels similar to those seen with WT macrophages (Fig. 4A). At 8 h p.i, increased levels of the IP-10 message were observed in TLR2–/– macrophages compared with WT macrophages at 5 MOI (p = 0.033 for WT vs TLR2–/– by one-way ANOVA). Results with infection at a MOI of 1 were similar (data not shown). However, by 24 h, no significant differences were observed between WT, TLR2–/–, and TLR4–/– macrophages in IP-10 mRNA levels. Under similar conditions, E. coli LPS, a TLR4 ligand, was unable to turn on IP-10 mRNA in TLR4–/– macrophages (Fig. 4A), as expected. Protein levels for IP-10 determined in supernatants paralleled the mRNA response at 24 h (data not shown). These findings indicate chlamydial-induced IP-10 is not dependent on TLR2 and TLR4.
Because MyD88 is the downstream adaptor molecule for most TLRs (49), the role of MyD88 in IP-10 induction during MoPn infection was determined using MyD88–/– macrophages. MyD88–/– and WT macrophages were infected with MoPn, and their IP-10 responses were compared. A reduction in IP-10 transcripts (33% with 1 MOI (data not shown) and 75% with 5 MOI) was observed at the 24 h time point in MyD88–/– macrophages compared with WT (Fig. 4B). The decrease in IP-10 levels in MyD88–/– macrophages was statistically significant at both 1 and 5 MOI and at all three time points tested (p < 0.001, p = 0.002, and p = 0.002, at 3, 8, and 24 h, respectively, with 5 MOI by one-way ANOVA). The levels of IP-10 protein detected in supernatants revealed an even greater effect of MyD88 deficiency on IP-10 production (p < 0.001; data not shown). However, expression of IP-10 mRNA and protein was not abolished in MyD88–/– macrophages after infection, suggesting that some of its induction (25% of WT) may be mediated through a MyD88-independent pathway. To verify that the MyD88-independent pathway is intact in MyD88–/– macrophages, E. coli LPS was used as a control. The IP-10 mRNA levels detected from MyD88–/– macrophages treated with E. coli LPS for 3 h were similar to WT macrophages, confirming that the TLR4-mediated, MyD88-independent pathway was unaffected in these cells (Fig. 4B). Thus, chlamydial-induced IP-10 production is largely but not completely dependent on MyD88.
To verify that the TLR and MyD88 dependence of IP-10 production was not limited to macrophages, the expression of the IP-10 gene was also measured in primary lung fibroblasts isolated from TLR2–/–, TLR4–/–, and MyD88–/– mice. Infection of the fibroblasts with MoPn led to a large increase in transcription of the IP-10 gene (Fig. 5A). Lower levels of transcription were found consistently in TLR2 fibroblasts compared with WT fibroblasts (Fig. 5A). This finding is in agreement with the decrease in production of IP-10 mRNA and protein in TLR2 macrophages. A smaller difference was observed between WT and TLR4 fibroblasts infected with Chlamydia (Fig. 5B). Dramatically lower levels of IP-10 gene transcription were observed in infected MyD88–/– fibroblasts (Fig. 5C).
It is important to note that the differences in cytokine and chemokine levels discussed above were not due to differences in the level of in vitro chlamydial infection of macrophages. No significant differences were observed in chlamydial inclusions between WT, TLR2–/–, TLR4–/–, and MyD88–/– macrophages (Fig. 7A). To determine whether the chlamydial inclusion developed normally in the knockout macrophages, the EBs recovered from the macrophages were used to infect a McCoy cell monolayer, and IFU were enumerated. No significant differences (by one-way ANOVA) were observed in the recovered IFU from WT, TLR2–/–, or MyD88–/– macrophages, whereas a 2- to 4-fold reduction was observed in EBs from TLR4–/– macrophages in three independent experiments (Fig. 7B) (p = 0.007, one-way ANOVA). The difference in TLR4–/– vs WT macrophages could be due to the inhibitory role of TLR4 on phagosome-lysosome fusion (50). However, this difference does not influence IP-10 levels in TLR4–/– macrophages (Fig. 6A).
IFN- induction during chlamydial infection of macrophages is independent of TLR2 and TLR4 but mostly dependent on MyD88
It is well documented that IFN- transcription is induced by TLR3 stimulation by viral dsRNA or TLR4 cross-linking by LPS (29), whereas TLR2 has not been implicated in this pathway (25). We have shown that induction of chlamydial-induced IP-10 mRNA was dependent on IFN- (Fig. 3). Analysis of IFN- levels in infected TLR2–/– and TLR4–/– macrophages showed that the differences in IFN- mRNA levels in infected WT, TLR4–/–, and TLR2–/– macrophages were not significant (Fig. 8A). Thus, chlamydial-induced IFN- occurred independent of TLR2 or TLR4. However, analysis of IFN- mRNA in MoPn-infected MyD88–/– macrophages showed a significant reduction at all time points (3 h, p = 0.002; 8 and 24 h, p < 0.001; one-way ANOVA) compared with WT macrophages. A 70–80% inhibition of IFN- mRNA transcription was observed in MyD88–/– macrophages compared with WT macrophages at an infection of 5 MOI (Fig. 8B), whereas a 30–40% inhibition was observed with a MOI of 1 (data not shown). The E. coli LPS-mediated IFN- response was significantly elevated in MyD88–/– macrophages compared with WT macrophages, most likely because of an increased use of the Toll/IL-1R domain-containing adaptor inducing INF--mediated pathway in the absence of MyD88, a phenomenon observed previously for LPS-stimulated cells (31). Assays for detecting IFN- protein levels were not successful possibly due to low sensitivity of the Abs used. These data indicate that TLR2 and TLR4 do not contribute to chlamydial-induced IFN- in macrophages, but a MyD88-dependent stimulatory pathway is involved. However, it must be noted that although IFN- mRNA levels are reduced in MyD88–/– macrophages compared with WT macrophages, significant induction of IFN- mRNA still occurs in the absence of MyD88, in response to infection (Fig. 8B). This suggests chlamydiae stimulate other, as yet uncharacterized, MyD88-independent pathways. Nonetheless, chlamydial-induced induction of IFN- is largely dependent on MyD88.
MyD88-dependent IFN pathways and the role for endosomal maturation in IFN- synthesis during chlamydial infection
Data from the above studies indicate a prominent role for MyD88 in MoPn-induced IFN- in macrophages, suggesting involvement of intracellular PRRs. Intracellular TLRs such as TLR7, TLR8, and TLR9 have been shown to induce IFN- in a MyD88-dependent manner (27, 28). TLR7, TLR8, and TLR9 form a subgroup within the TLR family due to their exclusive intracellular localization and recognition of PAMP in endosomal/lysosomal compartments (51, 52). Activation of these receptors depends on acidification and maturation of endosomes. Furthermore, during their activation, MyD88 is targeted to vesicular structures with lysosomal characteristics (52).
Based on these reports, we postulated that disruption of endosomal maturation would influence intracellular TLR-dependent IFN- induction during chlamydial infection. To address the role of intracellular TLRs in chlamydial-induced IFN-, MoPn-infected macrophages were treated with varying concentrations of bafilomycin A1 during infection. Bafilomycin A1, an acidification inhibitor, is traditionally used to block endosomal maturation (53). A dose-dependent reduction in the induction of IFN- mRNA and IP-10 protein (Fig. 9, A and B) was observed in bafilomycin-treated cells. As seen in previous experiments, UV-inactivated MoPn failed to induce any response regardless of bafilomycin treatment.
In contrast to IFN- and IP-10 levels, bafilomycin treatment of MoPn-infected cells did not have a negative effect on induction of TNF-. In fact, TNF- mRNA levels were increased at higher doses of bafilomycin (Fig. 9C). Thus, bafilomycin treatment did not cause a global inhibition of chlamydial-induced stimulatory pathways but was specific for the IFN- pathway. Bafilomycin treatment also did not affect chlamydial inclusion formation or recovery of infectious EBs in a significant manner (data not shown). Thus, the differences observed in induction of IFN- and IP-10 is not due to differences in the rate of infection. These results imply a role for intracellular PRRs that recognize chlamydial PAMP in endosomes to induce IFN- synthesis.
Role of IRF3 and TBK in chlamydial-induced IFN- induction
IRF3 is a necessary transcription factor for inducible IFN- (reviewed in Ref. 54). An important step preceding the induction of IFN- transcription is the phosphorylation of the transcription factor IRF3 by the kinases TBK and IKK (55, 56). TLR3 and TLR4, after binding their respective ligands, activate IRF3 by these kinases. As a result, IRF3 gets phosphorylated and transported to the nucleus to induce IFN- gene expression (57). By a complex mechanism that involves MyD88, the receptors TLR7, TLR8, and TLR9 can also activate these kinases to phosphorylate IRF3 (58). To begin elucidating the molecular mechanism of chlamydial-induced IFN- gene expression, nuclear translocation of IRF3 was investigated using MoPn-infected macrophages.
Nuclear extracts were prepared from control and infected macrophages at 3 and 8 h p.i., and IRF3 translocation was analyzed by Western blot (Fig. 10A). IRF3 nuclear translocation was observed as early as 3 h and increased at 8 h p.i. (Fig. 10A). The blots were stripped and reprobed for actin protein that served as a loading control and showed equivalent protein amounts in all lanes (Fig. 10A). These results demonstrate that IRF3 translocates to the nucleus during chlamydial infection and suggests its role in induction of IFN-.
As described previously, MyD88-dependent and independent signaling pathways that lead to IFN- induction converge in phosphorylation of IRF3. Two kinases that can function interchangeably have been implicated in phosphorylation of IRF3: TBK (55) and IKK (56). TBK is expressed in all cells, whereas IKK is present only in immune cells (56). To confirm the role of IRF3 phosphorylation in chlamydial-induced IFN-, TBK–/– MEFs were used. TBK–/– MEFs were infected with MoPn, and their IFN- expression was compared with WT MEFs. IFN- and IP-10 expression was completely abolished in MoPn-infected TBK–/– MEFs compared with WT MEFs, suggesting a complete dependence on the TBK-mediated pathway (Fig. 10B). It must be noted that in fibroblasts, only TBK plays a major role in IRF3 phosphorylation, whereas in macrophages, both TBK and IKK play a role interchangeably (56), and the above experiments do not take into account the role of IKK. However, these results demonstrate collectively that IRF3 phosphorylation and nuclear translocation play an important role in regulating induction of IFN- and, in turn, IP-10 during chlamydial infection. Together, our results demonstrate that chlamydial-induced IFN- results from activation of a MyD88-dependent pathway that functions through TBK-mediated IRF3 phosphorylation.
Discussion
Primary chlamydial genital infection is mostly resolved by T cells that are recruited to the site of infection in response to T cell chemokines, such as IP-10. Induction of the chemokines in turn results from interaction of host TLRs with chlamydial PAMP. In this study, we show that C. trachomatis (MoPn) infection, in vivo and in vitro, induces IP-10 and IFN- independently of TLR2 and TLR4. Furthermore, we demonstrate a prominent role for the downstream adaptor molecule MyD88 in induction of IP-10 and IFN- during infection. This is the first report of an intracellular bacterium inducing IFN- in a TLR4-independent and MyD88-dependent manner. In addition, we have found that the intracellular mechanisms inducing IFN- production during chlamydial infection involve TBK-mediated IRF3 nuclear translocation, mediated possibly by a PRR that requires endosomal maturation.
During our studies of in vivo genital infection of TLR2–/– and TLR4–/– mice with MoPn, we observed that IP-10 levels in the genital tract secretions are independent of TLR4. This was surprising, because IP-10 responses during infections with Gram-negative bacteria are normally mediated through TLR4 stimulation by bacterial LPS. Recent studies have demonstrated the importance of IP-10 in generation of effective innate and adaptive immune responses to genital infection of mice with MoPn (59, 60, 61). Our study demonstrates that, in agreement with the lower levels of IP-10 in genital secretions of MyD88–/– mice during MoPn infection, the MyD88–/– mice clear infection at a slower rate than control mice. This is in contrast to our previous results obtained with TLR2–/– and TLR4–/– mice, which clear infection at the same rate as the control groups (12). However, the exact cell type(s) producing IP-10 or responding to IP-10 in vivo is not known. Moreover, the levels of IP-10 in vivo could result from both type I and type II IFN levels, and the specific contribution of type I IFNs has yet to be established using type I IFNs knockout mice.
Analysis of IFN- induction during chlamydial infection in vitro provides a direct means for studying the TLR intracellular signaling pathways in response to infection at a cellular level. Although more work will be needed to correlate the findings of the in vitro experiments with the in vivo results, we have observed that in vitro responses such as TNF- and IL-6 production in response to chlamydial infection using TLR2–/– and TLR4–/– macrophages and fibroblasts appear to parallel the in vivo responses (12). Other in vitro studies demonstrating a role for TLRs in cytokine/chemokine expression in chlamydial infection using C. pneumoniae showed a similar correlation with in vivo results (62).
Nonetheless, significant variations in responses to different chlamydial strains have been reported. Two independent groups (62, 63) have shown a predominant role for TLR2 in C. pneumoniae-induced activation of DCs and macrophages using strains CM-1 and TW-183, respectively, whereas the findings from another group using the AR39 strain of C. pneumoniae have suggested a predominant role for TLR4 (64). Recently, Rothfuchs et al. (47) have shown, using BMDMs infected with a clinical isolate of C. pneumoniae (Kajaani 6), a reduction in IFN- and IFN- mRNA accumulation in TLR4–/– macrophages and no change in the levels of TNF- in MyD88–/– macrophages. Our results using the MoPn strain of C. trachomatis are in contrast to these findings. These differences could be due to innate differences between the structural components of C. pneumoniae and C. trachomatis, or to the use of BMDMs in previous studies vs the peritoneal macrophages used in this study. To address the difference in cell type used, we infected BMDMs from MyD88–/– and WT mice with C. trachomatis. The results for TNF- and IFN- induction in infected WT vs MyD88 BMDMs were similar to those obtained for peritoneal macrophages (data not shown), suggesting that the differences between the C. pneumoniae and C. trachomatis results may be attributed to inherent differences in the pathogens and not the cell types used for analysis.
The specific role of TLR4 in IFN- induction is well established (25). As a result, E. coli LPS-stimulated MyD88–/– cells remain fully able to promote IFN-inducible genes such as IP-10, GARG-16, or IRG-1. Likewise, Yersinia enterocolitica, Bacillus anthracis, or the sip B mutant of Salmonella typhimurium mediate type I IFN induction through stimulation of TLR4 (65, 66). But we have observed that C. trachomatis-induced IFN- and the IFN response gene IP-10 are stimulated independently of TLR4. Furthermore, 70–80% of the IFN- induced during chlamydial infection is MyD88 dependent (Fig. 8B). The intracellular TLRs (TLR7, TLR8, and TLR9) are the only TLRs known to induce type I IFNs in a MyD88-dependent manner (28, 51), suggesting a role for these TLRs in chlamydial-induced type I IFNs. The intracellular TLRs require endosomal maturation to function (51). We therefore explored the possibility that they may be involved in chlamydial-induced type I IFN, using inhibitors of endosomal maturation, and found that they may in fact play a role. However, we cannot rule out a role for other PRRs in macrophages and fibroblasts, which may function in a MyD88-dependent manner and require endosomal maturation, because TLR7 and TLR9 are expressed at high levels only in plasmocytoid DCs (67, 68). Moreover, although most of the IFN- induction during MoPn infection was MyD88 dependent, a complete inhibition of IFN- induction in MyD88–/– macrophages was not observed, suggesting some stimulation through chlamydial-mediated MyD88-independent pathways for IFN- induction. Besides TLR4, MyD88-independent induction of IFN- may be stimulated by a number of other candidate receptors. One of them is TLR3, a viral dsRNA-recognizing TLR known to activate the IFN- pathway, the role of which has not been studied during chlamydial infection. In addition to the surface and endosomal TLRs, the nucleotide-binding oligomerization domain family proteins (69) or the RNA helicase RIG-1 (70) that are present in the cytoplasm of macrophages could also function in a MyD88-independent manner to induce IFN-, although these PRRs are not expected to be sensitive to inhibitors of endosomal maturation. Therefore, we speculate that the fraction of MyD88-independent, IFN- induction observed during chlamydial infection could be mediated by any of these unexplored receptors.
IFN- has been traditionally considered an antiviral cytokine (24), and the intracellular mechanisms regulating its expression have been studied extensively in response to viral infection. Induction of the IFN- gene by bacteria has attracted attention only recently, because of the pleiotropic functions of type I IFNs in immunity (71). We present evidence that chlamydial-induced IFN- is mediated by TBK-dependent nuclear translocation of IRF3 (Fig. 9). Recently, Stockinger et al. (72) and O’Connell et al. (73) have shown the requirement for IRF3 and TBK in type I IFN induction during infection of macrophages by another intracellular pathogen, Listeria monocytogenes. In this respect, intracellular bacterial infection begins to resemble viral infection. However, in contrast to our observations with Chlamydia, up-regulation of type I IFN in macrophages by Listeria was not only TLR4 independent but also MyD88 independent (72). One possibility for these differences could be the intracellular location of Chlamydia in inclusions compared with Listeria, which resides in the cytosol. Therefore, depending on their intracellular location, the bacteria may use different pathways to activate IRF3. Besides IRF3, another key transcription factor that plays a role in type I IFN induction is IRF7 (74). Two groups (58, 75) have recently shown that IRF7 interacts with MyD88 to form a complex that also involves IRAK4 and TRAF6 (components of NF-B activation). Together they provide the foundation for intracellular TLR-mediated IFN induction. This would suggest that in MoPn-infected MyD88–/– macrophages, nuclear IRF7 levels could be reduced significantly. Future studies will therefore need to investigate the levels of nuclear IRF3 and IRF7 in WT and MyD88–/– macrophages to determine the relative role of IRF3 and IRF7 in chlamydial-induced IFN- induction.
The physiological function of type I IFNs as a link between innate and adaptive immunity is well established. However, the physiological role of type I IFNs during chlamydial infection in vivo has not been investigated. The signaling pathways that lead to IFN- induction could be of major importance in determining the course of infection or pathology in vivo. Unexpectedly, the pathways of IFN- induction during chlamydial infection turn out to be more complex than the expected LPS-TLR4-mediated response. In summary, our findings indicate that IFN- and subsequent IP-10 induction during MoPn infection occurs through two pathways. The major pathway requires MyD88-dependent activation, whereas a minor pathway involves MyD88- independent activation; both function through IRF3 phosphorylation. Determining the specific PRR(s) that contributes to the induction of IFN- during chlamydial infection is the next critical step in understanding the ensuing cytokine/chemokine response.
Acknowledgments
We thank Dr. Wen-Chen Yeh (University Health Network, Toronto, Ontario, Canada) for providing TBK–/– MEFs. Technical help by Jim Sikes and Anne Bowlin on ELISA, animal husbandry, and Chlamydia stocks is gratefully acknowledged. We also thank Dr. Shanmugam Nagarajan for critical reading of this manuscript.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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 This work was supported by the Horace C. Cabe and the Bates-Wheeler Foundations, the Arkansas Children’s Hospital Research Institute, a Pilot grant award from the University of Arkansas for Medical Sciences, the Medical Research Endowment, University of Arkansas for Medical Sciences award to U.M.N., and Public Health Service Grant AI054624 from the National Institutes of Health to T.D.
2 Address correspondence and reprint requests to Dr. Uma M. Nagarajan, Department of Microbiology and Immunology, 4301 West Markham Street, BM506, University of Arkansas for Medical Science, Little Rock, AR 72205. E-mail address: nagarajanuma@uams.edu
3 Abbreviations used in this paper: PAMP, pathogen-associated molecular pattern; IP-10, IFN--inducible protein 10; EB, elementary body; MoPn, mouse pneumonitis agent of Chlamydia trachomatis; MOI, multiplicity of infection; PGN, peptidoglycan; p.i., postinfection; PRR, pattern recognition receptor; IRF3, IFN response factor-3; TBK, TANK-binding kinase; WT, wild type; DC, dendritic cell; IFU, inclusion-forming unit; BMDM, bone marrow-derived macrophage; RT, reverse transcription; MEF, murine embryonic fibroblast.
Received for publication January 12, 2005. Accepted for publication April 19, 2005.
References
Gerbase, A. C., J. T. Rowley, T. E. Mertens. 1998. Global epidemiology of sexually transmitted diseases. Lancet 351:(Suppl. 3): 2-4.
Swenson, C. E., J. Schachter. 1984. Infertility as a consequence of chlamydial infection of the upper genital tract in female mice. Sex. Transm. Dis. 11: 64-67.
Cain, T. K., R. G. Rank. 1995. Local Th1-like responses are induced by intravaginal infection of mice with the mouse pneumonitis biovar of Chlamydia trachomatis. Infect. Immun. 63: 1784-1789.
Ramsey, K. H., R. G. Rank. 1991. Resolution of chlamydial genital infection with antigen-specific T-lymphocyte lines. Infect. Immun. 59: 925-931.
Rank, R. G., K. H. Ramsey, E. A. Pack, D. M. Williams. 1992. Effect of interferon on resolution of murine chlamydial genital infection. Infect. Immun. 60: 4427-4429.
Perry, L. L., K. Feilzer, H. D. Caldwell. 1997. Immunity to Chlamydia trachomatis is mediated by T helper 1 cells through IFN--dependent and -independent pathways. J. Immunol. 158: 3344-3352.
Cotter, T. W., K. H. Ramsey, G. S. Miranpuri, C. E. Poulsen, G. I. Byrne. 1997. Dissemination of Chlamydia trachomatis chronic genital tract infection in interferon gene knockout mice. Infect. Immun. 65: 2145-2152.
Takeda, K., T. Kaisho, S. Akira. 2003. Toll-like receptors. Annu. Rev. Immunol. 21: 335-376.
Spitzer, J. H., A. Visintin, A. Mazzoni, M. N. Kennedy, D. M. Segal. 2002. Toll-like receptor 1 inhibits Toll-like receptor 4 signaling in endothelial cells. Eur. J. Immunol. 32: 1182-1187.
Shiraki, R., N. Inoue, S. Kawasaki, A. Takei, M. Kadotani, Y. Ohnishi, J. Ejiri, S. Kobayashi, K. Hirata, S. Kawashima, M. Yokoyama. 2004. Expression of Toll-like receptors on human platelets. Thromb. Res. 113: 379-385.
Barton, G. M., R. Medzhitov. 2003. Toll-like receptor signaling pathways. Science 300: 1524-1525.
Darville, T., J. M. O’Neill, C. W. Andrews, Jr, U. M. Nagarajan, L. Stahl, D. M. Ojcius. 2003. Toll-like receptor-2, but not Toll-like receptor-4, is essential for development of oviduct pathology in chlamydial genital tract infection. J. Immunol. 171: 6187-6197.(Uma M. Nagarajan2,*,, Dav)