Toll-like receptors (TLRs) are among the first-line sentinels for immune detection and responsiveness to pathogens. The TLR2 subfamily of TLRs (TLR1, TLR2, TLR6) form heterodimers with each other and are thus able to recognize a broad range of components from several microbes such as yeast, Gram-positive bacteria and protozoa. Until now, TLR2 activation by bacterial ligands has long been associated with pro-inflammatory cytokines but not type I interferon responses.
Using a variety of transgenic mice, here we provide in vivo and in vitro data showing that TLR2 activation does in fact induce interferon-beta and that this occurs via MyD88-IRF1 and -IRF7 pathways. Interestingly, by microscopy we demonstrate that although a cell surface receptor, TLR2 dependent induction of type I interferons occurs in endolysosomal compartments where it is translocated to upon ligand engagement. Furthermore, we could show that blocking receptor internalization or endolysosomal acidification inhibits the ability of TLR2 to trigger the induction type I interferon but not pro-inflammatory responses.
The results indicate that TLR2 activation induces pro-inflammatory and type I interferon responses from distinct subcellular sites: the plasma membrane and endolysosomal compartments respectively. Apart from identifying and characterizing a novel pathway for induction of type I interferons, the present study offers new insights into how TLR signaling discriminates and regulates the nature of responses to be elicited against extracellular and endocytosed microbes. These findings may also have clinical implication. Excessive production of pro-inflammatory cytokines and type I IFNs following activation of TLRs is a central pathologic event in several hyper-inflammatory conditions. The discovery that the induction of pro-inflammatory and type I IFN responses can be uncoupled through pharmacological manipulation of endolysosomal acidification suggests new avenues for potential therapeutic intervention against inflammations and sepsis.
Citation: Dietrich N, Lienenklaus S, Weiss S, Gekara NO (2010) Murine Toll-Like Receptor 2 Activation Induces Type I Interferon Responses from Endolysosomal Compartments. PLoS ONE 5(4): e10250. doi:10.1371/journal.pone.0010250
Editor: Derya Unutmaz,
Received: February 11, 2010; Accepted: March 26, 2010; Published: April 20, 2010
Copyright: © 2010 Dietrich et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was supported in part by the German Research Council (DFG) via SFB 566, the Kultusministerium of Niedersachsen via the Lichtenberg PhD program, the centre for infection biology (ZIB) via contract number VH-KO-106, the Helmholtz International Research School for Infection Biology (HIRSIB) and the German Federal Ministry of Education and Research (BMBF). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Pattern recognition receptors (PRRs) serve as first-line sentinels for immune detection of pathogens. To date, the best characterized PRRs are the Toll-like receptors (TLRs) expressed either on the cell surface (TLR1, 2, 4, 5, 6 and 11) or on endosomal membranes (TLR3, 7, 8 and 9) . Stimulation of TLRs triggers activation of MyD88-dependent signaling pathways as well as TRIF-dependent pathways that are unique to TLR3 and TLR4 , , . These pathways activate various transcription factors like NF-κB, AP-1 and/or interferon regulatory factors (IRFs). This results in the transcription of genes encoding pro-inflammatory cytokines, chemokines and/or type I interferons (IFNs) that orchestrate anti-bacterial and anti-viral responses , , .
Although originally discovered based on their ability to interfere with viral replication, type I IFNs are also produced in response to bacterial and even parasitic infections , . Type I IFNs encompass a family of more than 13 IFN-α isotypes, a single IFN-β isotype, and others, such as IFN-ω, -ε and -κ , , , . They all signal through a common two-chain receptor (IFNAR) and activate a large set of IFN-inducible genes that exert widely overlapping pleiotropic immunomodulatory effects central to the activation of innate and adaptive immunity.
Of the PRRs, broadly, two categories of receptors are known to induce type I IFNs: the transmembrane TLRs and the cytosolic receptors , . The latter include the RIG-I helicase and MDA5 which recognize cytosolic RNA and DNA , , , 
The TLR2 subfamily of receptors (TLR2/TLR1 and TLR2/TLR6 heterodimers) recognizes a broad range of cell surface components from yeast, Gram-positive bacteria, viruses and even protozoa , . By virtue of their cell surface localization, in principle, these receptors are among the first to interact with and hence initiate immune reactions against such microbes. So far,these receptors have been assumed to elicit pro-inflammatory but not type I interferon responses in response to activation by bacterial ligands, , , , , , . We now show that ligands for TLR2/1 and TLR2/6 heterodimers do induce type I IFN responses. Although a cell surface receptor, we observed that upon activation, TLR2 is internalized and transported into endolysosomal compartments from where it induces interferon-β via MyD88-IRF1 and IRF7 dependent pathways. Consequently, inhibition of receptor internalization or endosomal acidification could block the induction of IFN-β and IFN-inducible genes but not pro-inflammatory cytokines like TNF-α. This suggests that TLR2 activation induces pro-inflammatory and type I interferon responses from distinct sub-cellular sites: the plasma membrane and endolysosomal compartments, respectively. The spatio-temporal diversification of TLR2 signaling offers new insights into how TLR signaling discriminates and regulates the nature of responses to be elicited against extracellular and endocytosed microbes.
TLR2 induces type I IFN responses
TLR2 activation by bacterial ligands is generally assumed to induce the transcription of canonical NF-κB dependent pro-inflammatory cytokines but not type I interferons (IFNs) or interferon-inducible genes , , , , . However, when we stimulated bone marrow derived macrophages (BMDMs) we noted that besides pro-inflammatory cytokines like TNF-α and IL-12 (Fig. 1A and D), ligands for TLR2/TLR1 (PAM3CSK4) and TLR2/TLR6 (MALP-2, PAM2CSK4), could also trigger the induction of type I IFN-inducible genes; IP-10 (CXCL-10), MX-2, IL-6 and iNOS (for NO production) (Fig. 1B, E and Fig. S1).
Figure 1. TLR2 directly induces pro-inflammatory and type I IFN dependent responses via MyD88.
WT, TLR2−/−, MyD88−/−, IFN-β−/− and IFNAR1−/− macrophages were stimulated with 1.5 µg/ml of MALP-2 or PAM2CSK4 or PAM3CSK4 then analyzed 6 h later by real-time RT-PCR for mRNAs of TNF-α (A), IP-10 (B), Mx-2 (E) and IFN-β (C). In parallel, culture supernatants of cells stimulated for 24 h or 10 h were analyzed by ELISA for concentrations of IL-12p70 (D) or IFN-β (F) respectively. Data are representative of at least three independent experiments (mean ± s.e.m).doi:10.1371/journal.pone.0010250.g001
An initial interpretation of type I IFN responses elicited by TLR2 ligands was contamination. To rule out such a possibility, macrophages from TLR2−/− and MyD88−/− mice were analyzed. Indeed, TLR2 ligands elicited no response in TLR2−/− and MyD88−/− macrophages (Fig. 1A–F and Fig. S1). We also analyzed macrophages from mice deficient in IFN-β or the type I interferon receptor (IFNAR). IFN-β−/− and IFNAR−/− macrophages were impaired in the induction of type I IFN-dependent genes IP-10, MX-2, iNOS and IL-6 but not in the pro-inflammatory cytokines TNF-α or IL-12 (Fig. 1A, B, D, E and Fig. S1). Interestingly, the impairment in IFN-β−/− macrophages was equivalent to that in IFNAR−/− macrophages (Fig. 1B, E, and Fig. S1). Since the IFNAR is a common receptor for all type I IFNs, this observation suggested that the type I IFN-dependent responses induced by such agonists was largely due to the activity of IFN-β and not the other isotypes like IFN-αs.
This prompted us to test for IFN-β induction. Quantitative real-time RT-PCR and ELISA assays demonstrated that all the TLR2 ligands tested could trigger the transcription and subsequent secretion of IFN-β in a TLR2-MyD88 dependent manner (Fig. 1C and F). Intriguingly, IFNAR−/− macrophage cultures were repeatedly found to contain more IFN-β (Fig. 1F). Such difference was however not reflected at the mRNA levels (Fig. 1C), suggesting that higher levels of IFN-β in IFNAR−/− cultures was due to lack of consumption via IFNAR. Taken as a whole, these results show that activation of the TLR2-MyD88 pathway leads to the production of IFN-β which in turn induces type I IFN-dependent responses.
These findings are contrary to several previous studies , , , . However, recently, while this study was under consideration elsewhere, Barbalat and co-workers could show TLR2 - driven type I IFN responses but only toward viral and not bacterial ligands . Whereas the explanation for the previous failures to detect TLR2-driven type I IFN responses is not very certain, several factors including sub-optimal activation of TLR2, low sensitivity of assays and/or measuring type I IFN response off peak period could have been responsible. For example low concentrations of TLR2 ligands and sub-optimal northern blotting and RT-PCR conditions detecting only very strong responses could account for some of the previous observations , , , . For instance, it is well known, and also shown herein (Fig. 2A and Fig S2A), that TLR ligands induce IFN-β with a rapid kinetic with mRNA peaks between 1 and 6 hours post stimulation followed by a rapid decline to near basal level by 12 hour post stimulation. In the study by Barbalat and colleagues, the conclusions that bacterial TLR2 ligands do not induce type I IFNs were largely based on experiments measuring IFN-β transcript 12–24 hours–a time window not optimal for detection of IFN-β transcript. The fact that IFN-β induced by vaccinia virus could be detected at this time point is probably a reflection of the potential difference in the kinetics with which the purified TLR ligand and whole microorganisms (in this case vaccinia virus) interacts with the host cells. On the other hand, while Toshchakov and co-workers  also concluded that TLR2 ligands do not induce type I IFNs in bone marrow macrophages, a critical analysis of their data suggests otherwise. Albeit lower than that by LPS, the authors could detect TLR2 driven transcriptional upregulation of IFN-β and the IFN dependent genes iNOS (Figure 5a and 1 respectively of that paper). The inability to also detect clear response for the other type I IFN genes like IP-10 could be due to low sensitivity of assays probably optimal for the detection of stronger responses triggered via TLR4 and not TLR2. A similar argument could also account the PCR and northern blot data in the other previous studies , . In short, several reasons could account for the poor detection of TLR2-driven type I IFNs responses in previous studies and the explanations may vary from one experiment/study to another. Equally noteworthy, as we have shown herein (Fig. 1F) the fact that secreted IFN-β is rapidly depleted from the system via IFNAR could probably have undermined its detection in other previous studies. This is especially so if assayed long after the peak of production.
Figure 2. TLR2 activation induces IFN-β transcription independent of an auto/paracrine loop.
In A, BMDMs were stimulated with either MALP-2 (1.5 µg/ml) or Poly I:C (5 µg/ml) for the indicated time periods then analyzed by quantitative real time RT-PCR for IFN-β mRNA. B–D; TLR2 driven IFN-β induction is independent of an auto/paracrine mechanisms. BMDMs pre-treated for 1 h with or without 0.5 µg/ml CHX were stimulated for 6 h with either MALP-2 (1.5 µg/ml) or Poly I:C (5 µg/ml) then analyzed by real-time RT-PCR for IFN-β (B, please note logarithmic scale) and TNF-α (C). Culture supernatants of cells stimulated in parallel for 24 hours in the presence of cycloheximide were analyzed for nitric oxide (D). Data are representative of three independent experiments (mean ± s.e.m).doi:10.1371/journal.pone.0010250.g002
TLR2 induces IFN-β directly
We also tested whether TLR2 induces IFN-β directly as opposed to a delayed response via auto/paracrine mechanisms. To that end, first, we did a time course analysis by real-time RT-PCR, to determine the kinetic of IFN-β gene induction. TLR2 agonists induced IFN-β transcription with a rapid kinetic (peak at 2–3 h). This fast response was comparable to that by Poly:IC (TLR3 ligand) which is known to activate IFN-β directly (Fig. 2A). As depicted below, rapid IFN-β induction upon TLR2 activation was also evident in vivo (Fig. 3A, C and Fig. S2A). Nonetheless, to compellingly test whether IFN-β induction is independent of de novo protein synthesis, cells were stimulated with TLR2 and TLR3 agonists in the presence or absence of the protein synthesis inhibitor cycloheximide (CHX). In spite of complete inhibition of protein synthesis, as depicted by the absence of NO and IL-6 in supernatants (Fig. 2D and data not shown), TLR2 and TLR3 ligands still induced the transcription of IFN-β and TNF-α in the presence of CHX (Fig. 2B–C). Thus, much like TLR3, activation of TLR2 does induce both pro-inflammatory and IFN-β responses directly.
Figure 3. TLR2-driven IFN-β and type I IFN dependent responses in vivo.
A–B; Five mice per group WT, TLR2−/− and IFNAR−/− were injected with control medium or PAM3CSK4 (100 µg/mouse) then sacrificed at different time points between 0–18 hours and their sera (A) or peritoneal exudates (B) analyzed by ELISA for IFN-β and IP-10 respectively Data representative of two independent experiment show ± s.e.m. C; Three mice per group of TLR2+/+IFN-β+/Δ-luc (TLR2+/+) and TLR2−/−IFN-β+/Δ-luc (TLR2−/−) mice were injected with MALP-2 (100 µg/mouse) and expression of luciferase was monitored for 24 h by luminescence imaging. C shows two representative mice out of a group of four mice before (0 hours) and after (3 hours) MALP-2 administration. Fold increase and kinetics of luciferase induction of corresponding mice are depicted in Fig. S2A. D–I; TLR2 mediated IFN-β induction in macrophages. Macrophages matured from bone marrows of TLR2+/+IFN-β+/Δ-luc or TLR2−/−IFN-β+/Δ-luc mice were stimulated with indicated concentrations of MALP-2 (D, G) or PAM2CSK4 (E, H) or PAM3CSK4 (F, I) for 4 hours and analyzed for luciferase expression. D–F; show representative luminescence images of macrophages while G–I indicate the fold increase in luciferase in corresponding cells. Luciferase expression is represented by a colour shift from blue to red. Similar results were obtained from three independent experiments (mean ± s.e.m). Note that the background response in reporter mice, mainly in the liver region, might be due to the liver's role in trapping and detoxifying noxious substances from the system. Such substances and the metabolic products thereof could account for the higher MALP-2 associated background in TLR2−/− mice.doi:10.1371/journal.pone.0010250.g003
TLR2-driven IFN-β induction in vivo
To investigate TLR2-driven type I IFN responses in vivo, WT, TLR2−/− and IFNAR−/− mice were inoculated with the TLR2 ligand PAM3CSK4 and then analyzed for secretion of IFN-β and type I IFN-dependent mediators. ELISA assay readily revealed an accumulation of IFN-β in the serum of IFNAR−/− but not TLR2−/− or WT mice (Fig. 3A and data not shown). This is consistent with the in vitro data showing that upon TLR triggering, IFN-β is secreted which in turn binds to and is thus removed from the system via IFNAR. Indeed, further analysis revealed elevated levels of IP-10 and NO in WT but not in TLR2−/− or IFNAR−/− mice (Fig. 3B and data not shown).
To monitor the in vivo induction of IFN-β in more details, we constructed a transgenic IFN-β-luciferase reporter mouse (Δβ-luc). In these mice, the IFN-β gene was replaced by the firefly luciferase. For the present analysis, we used Δβ-luc heterozygous mice containing a functional IFN-β allele and a luciferase expression cassette (TLR2+/+IFN-β+/Δ-luc). For control, the Δβ-luc mice were crossed with TLR2−/− mice to generate TLR2−/−IFN-β+/Δ-luc mice. When challenged with MALP-2 or PAM3CSK4, TLR2+/+IFN-β+/Δ-luc mice responded with luminescence while mice lacking TLR2 (Fig. 3C and Fig. S2A). Consistent with the rapid kinetics of IFN-β induction observed in vitro (Fig. 2A) the peak of the in vivo IFN-β response by such TLR2 ligands was also 3-4 hours post stimulation (Fig. 3C and Fig. S2A).
Over the years it had been suggested that only the TLR9 subfamily (TLR7, TLR8 and TLR9) could activate type I IFNs via MyD88. In fact, even the MyD88 pathways for IFN induction by such TLRs was believed to be functional only in plasmacytoid dendritic cells (pDCs) , . However recently Mancuso and co-workers were able to demonstrate a role for TLR7-MyD88 dependent pathway of type I IFN induction in conventional DCs in response to group B streptococcus . To investigate whether the TLR2-MyD88 pathway of IFN-β induction is also functional in other cell types, peritonial macrophages, bone marrow derived macrophages (BMDM) and dendritic cells (BMDCs) were also tested in an in vitro stimulation assay with graded concentrations of TLR2 and TLR4 agonists. TLR2 agonists were found to induce luciferase in both, macrophages and DCs from TLR2+/+ but not TLR2−/− mice (Fig. 3D–I and Fig. S2B and data not shown). In contrast, LPS induced responses in both, TLR2+/+ and TLR2−/−, cells (Fig. S2C&D). Furthermore, LPS stimulation depicted stronger responses as compared to TLR2 which is consistent with the activation of the TRIF pathway associated with very strong type I IFN induction.
In a recently study, Barbalat and co-workers also proposed TLR2 - dependent type I IFN production only in response to viral but not toward bacterial ligands. The authors also proposed that such responses were only feasible in inflammatory monocytes . The functional role of inflammatory monocytes in the study in question was based on depletion of cells expressing CD11b - a molecule generally expressed on several cell types including monocytes, macrophages, granulocytes, and dendritic cells. This raises the possibility that such responses were in fact the sum of several cell types and not just inflammatory monocytes. Whichever the case and in view of our findings it can safely be concluded that TLR2 pathway of IFN-β induction is functional in several host cell types and that this pathway is not a preserve of viral activation but rather a consequence of TLR2 activation by various ligands including those from bacteria, mycoplasma and possibly fungi and protozoa.
TLR2 induces type I IFN-β via IRF1 and IRF7
The transcription of IFN-α/β is primarily controlled by the IRF family of transcription factors. To elucidate the IRFs downstream of the TLR2-MyD88 pathway of IFN-β induction, we investigated the role of IRF1, IRF3 and IRF7. These transcription factors have been implicated as positive regulators of the transcription of type I IFN genes . Macrophages from IRF1−/− and IRF7−/− mice were highly diminished in IFN-β production following stimulation with TLR2 ligands, while IRF3−/− macrophages had a normal response (Fig. 4A–C). Based on the present data, TLR2 activation induces type I IFN responses via MyD88-IRF1 and -IRF7 pathways. These findings suggest that the TLR2-driven pathway of type I IFN induction is in a way similar to that by the TLR9 subfamily of TLRs also shown to trigger MyD88-IRF1 and -IRF7 pathways , , , , .
Figure 4. TLR2 activation induces type I IFN-β via IRF1 and IRF7.
A–C, WT, TLR2−/−, IRF1−/−, IRF3−/−, and IRF7−/− BMDMs were stimulated with ligands for TLR2/6 (1.5 µg/ml MALP-2 or PAM2CSK4) or TLR2/1 (1.5 µg/ml PAM3CSK4). After 10 hours, supernatants were analyzed for IFN-β by ELISA. Similar data represent an average of two independent experiments (mean ± s.e.m, *: p<0.05).doi:10.1371/journal.pone.0010250.g004
TLR2 induces IFN-β from endolysosomes
One of the emerging themes in the induction of type I IFNs is the intracellular localization of receptors. Potent inducers of type I IFNs like TLR3, TLR7/8, TLR9, RIG-I, MDA5, DAI, are all intracellular receptors that sense nucleic acids located in the endosome or cytosol , . Using confocal microscopy, we first studied the sub-cellular localization of TLR2-GFP in transiently transfected RAW 264.7 macrophages. Consistent with previous observations , in unstimulated cells, TLR2-GFP was found to be localized either at the plasma membrane or in the perinuclear region (most likely in ER and trans-golgi networks) and hardly co-localized with the endolysosomal markers lysotracker or dextran (Fig. 5A and Fig. S3A). However by 3 h after MALP-2 stimulation, TLR2 was found enriched in clusters which co-localized with lysotracker or Dextran (Fig. 5B and Fig. S3B). Such co-localization was however severely impaired when the endocytic maturation pathway was blocked using Bafilomycin A (BafA), an inhibitor of the endosomal proton pump (Fig. 5C and Fig. S4C). These observations suggest that TLR2 activates endocytic internalization and endosome maturation, as proposed previously , , .
Figure 5. Ligand induced internalization and trafficking of TLR2 into endolysosomal compartments.
RAW 264.7 macrophages transfected with hTLR2-GFP were labelled with lysotracker then left untreated (A) or stimulated with 1 µg/ml of MALP-2 and imaged by confocal microscopy after 3 h (B). Panel C shows cells first labelled with lysotracker then incubated for 1 hour with 50 µM Bafilomycin A before stimulation with MALP-2. The co-localization fluorographs displayed in the right panels show the intensities and scatter pattern of all pixels within the merged images. Pixels with mostly one fluorescent component are placed along the axes 1 and 2 while the pixels with equal fluorescence intensity from both components (due to co-localization) are placed along the diagonal. Axis 1 and 2 show the green (TLR2-GFP) and red (Lysotracker) fluorescence intensities on an arbitrary scale of 0 to 250.doi:10.1371/journal.pone.0010250.g005
To investigate a potential link between the translocation of TLR2 into endolysosomal compartments and the induction of type I IFN responses, macrophages were stimulated with TLR2 ligands in the presence of Cytochalasin D (CytD). CytD blocks endocytic internalization by inhibiting actin polymerization. CytD almost completely inhibited the induction of IFN-β and type I IFN-inducible genes (IP-10, and iNOS) by TLR2 ligands (Fig. 6A–C). In addition to blocking receptor internalization and vesicular trafficking, CytD may also interfere with other signalling processes. Therefore, we resorted to a more specific inhibitor of the endocytic maturation pathway: Bafilomycin A (BafA). BafA inhibits the endosomal proton pump hence endosome acidification and subsequent fusion with lysosomes. BafA severely inhibited the TLR2 driven IFN-β and type I IFN dependent responses but not the pro-inflammatory responses (Fig. 6D–F). Similar results were obtained with Dynasore (Dyn), an inhibitor of the endocytic effector protein dynamin (Fig. S4). Collectively, these data suggest that TLR2 induces pro-inflammatory cytokines like TNF-α from the plasma membrane while IFN-β is triggered from endolysosomal compartments after receptor internalization.
Figure 6. Blockers of receptor internalization and endosome maturation pathways abrogate TLR2 driven type I IFN responses.
BMDM, untreated or pre-treated with either 25 µM Cytochalasin D (A–C) or 50 µM Bafilomycin A (D–F) for 1 hour, were stimulated with 1.5 µg/ml of the indicated TLR2 ligands. After 6 h cell pellets were analyzed by quantitative real-time RT-PCR for IFN-β (A and D). Supernatants of cells stimulated in parallel for 24 hours were also analyzed for Nitric Oxide (Nitrite) by Griess reaction (B and E) or for TNF-α. Data represent similar results from three independent experiments (mean ± s.e.m, *: p<0.05). G; Model for TLR2 signaling. Upon ligand engagement at the plasma membrane TLR2 activates a MyD88-NF-κB pathway leading to the induction of pro-inflammatory cytokines like TNF-α and IL-12. The initial trigger at the plasma membrane also induces endocytic internalization and transportation of the TLR2 complex into endolysosomal compartments from where it activates MyD88-IRF1/IRF7 pathways of IFN-β induction. Once produced, IFN-β in turn activates STAT1 which alone or in synergy with other transcription factors like NF-κB activates the transcription of IFN-inducible genes; Mx-2, IP-10, iNOS, IL-6.doi:10.1371/journal.pone.0010250.g006
So far, the signalling model for the TLR2 subfamily of TLRs has been based on the dogma that these receptors induce pro-inflammatory cytokines but not type I interferons upon activation by bacterial ligands. Using different in vitro and in vivo approaches, we now show that in addition to pro-inflammatory cytokines, TLR2 heterodimers also elicit interferon-β and hence type I interferon dependent responses. Interestingly, the induction of pro-inflammatory responses and type I interferons by TLR2 ligands occur from distinct sub-cellular sites: the plasma membrane and endolysosomal compartments, respectively.
As schematically illustrated in Fig. 6G, our results suggest a new model for TLR2 signalling. We propose that analogous to that for TLR4, TLR2 activation at the plasma membrane first induces the MyD88-NF-κB-dependent pathway. This activates not only the transcription of genes encoding pro-inflammatory cytokines but also induces endocytic internalization and trafficking of the TLR2 complex into endolysosomal compartments, where it additionally activates MyD88-dependent IRF1 and IRF7 pathways for IFN-β induction. Consequently, as we have shown, pharmacological blockers of receptor internalization and endosome maturation are able to inhibit the induction of type I interferons but not pro-inflammatory cytokines upon TLR2.
Type I IFNs support the transcription of several mediators that promote protection against microbes. Some of the type I IFN-inducible mediators such as IP-10 and NO, herein shown to be induced in response to TLR2 activation play an important role not only in the activation and recruitment of inflammatory cells to sites of infection but also have direct anti-microbial effects. For example, TLR2-driven NO production, is known to be critical for protective immunity against Toxoplasma gondii . Similarly, NO production, also a critical protective component against mycoplasmas , is known to be largely due to TLR2 activation by MALP-2 or its homologs , .
In a recent study, Barabalat and co-workers reported a role for TLR2 in the recognition of vaccinia virus . However since such studies employed UV irradiated virus particles, incapable of activating cytosolic signalling pathways, whether type I IFN responses induced by vaccinia virus are mainly due to TLR2 activation as proposed , as opposed to the TLR independent cytosolic pathways as reported by Zhu and others  remains to be resolved. Nonetheless, together with the present findings it is clear that although so far ignored, TLR2–driven pathway of type I IFN induction is a crucial component of host defences against a variety of microbes.
The mechanistic explanation for the induction of type I IFNs from endolysosomal compartments by TLR2 is not clear at the present. However, we are tempted to speculate that such intracellular compartments could serve as docking sites on which key signalling molecules involved in type I IFN induction are marshalled to a sufficient concentration to mediate downstream signalling processes. We also infer that the acidic environment in endolysosomal compartments probably facilitates conformational changes in TLRs that initiate interactions with downstream adaptors, kinases and IRFs. In line with this hypothesis, DNA structures that target different endocytic compartments are known to induce distinct biological responses , , , and have been shown to trigger different conformational changes in TLR9 . A more recent study showing that proteolytic processing of TLR7 and TLR9 by endolysosomal proteases is a prerequisite for their signalling  is also consistent with the above idea. Additional studies are required to provide more mechanistic insights into how intracellular localization and an acidic environment facilitate the induction of type I IFNs by TLRs and other PRRs.
Finally, our results have clinical implications. Although the TLR triggering does mediate protective host responses against microbes, excessive production of pro-inflammatory cytokines and type I IFNs due to such activation is known to inflict severe self-injury. Thus the discovery that pro-inflammatory and type I IFN responses are induced from distinct sub-cellular sites and that these two responses can be uncoupled by properly manipulating the endosome maturation pathway suggest new avenues for potential therapeutic intervention against inflammations.
Materials and Methods
All mice were maintained under pathogen free conditions and experiments were approved and carried out according to the guidelines set out by LAVES (Niedersaechsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit).
C57BL/6 mice were purchased from Harlan- Germany. IRF1−/−, IFNAR−/−  and IFN-β−/−  mice were bred in our animal facility. TLR2−/−  and MyD88−/−  mice, originally generated in Dr. S Akira's laboratory, were obtained from MPI-Berlin. IRF3−/−, and IRF7−/−  mice were obtained from Dr. T. Decker (University of Vienna, Austria) with permission from Dr. T. Taniguchi (University Tokyo, Japan). The transgenic IFN-β-luciferase reporter mice (Δβ-luc)  were generated by replacing IFN-β with the firefly luciferase using the standard homologous recombination protocol. The Δβ-luc mice were then backcrossed to TLR2−/− mice  to generate TLR2-/-IFN-β+/Δ-luc and TLR2+/+IFN-β+/Δ-luc mice.
MALP-2 was a kind gift from Dr. Werner Tegge, HZI. PAM2CSK4, PAM3CSK4, and LPS Salmonella minnesota were from Invivogen. Poly I:C was from Amersham Biosciences. Cytochalasin D was purchased from FLUKA, Cycloheximide, Bafilomycin A and Dynasore were purchased from Sigma.
Bone marrow derived macrophages were cultured for 7 days in the presence of M-CSF and bone marrow derived dendritic cells were cultured for 10 days in the presence of GM-CSF and IL-4 according to standard protocols.
Cell transfection and confocal imaging
hTLR2-GFP was bought from Invivogen. RAW 264.7 cells were transiently transfected using MicroPorator (PM-100) according to the manufacturer's instructions. hTLR2-GFP transfected cells were loaded with 1 µM of Lysotracker Red DND-99 or Dextran-Alexa Fluor 647 (both from Invitrogen) for 30 minutes, washed, then incubated with Bafilomycin A (in some cases) before stimulation with TLR2 ligands. Confocal imaging was done as described previously .
Supernatants were analyzed by ELISA using rat anti-mouse IL-12p70 and rat anti-mouse IL-6 antibodies together with the respective recombinant mouse IL-6 and IL-12 standards from BD Pharmingen. The IFN-β ELISA kit was purchased from PBL Biomedical Laboratories and the IP-10 ELISA was from Bender MedSystems. Both were performed according to the manufacturer's instructions. Nitric oxide was detected using the Griess reagent kit (Promega).
In vivo luciferase assays
Mice were injected intraperitoneally with TLR2 ligands. At different time points, mice were anaesthetized with isoflurane using the XGI-8 gas anaesthesia system (Caliper). Prior to image acquisition 3 mg of luciferin (Caliper) dissolved in 100 ml PBS was injected intraperitoneally. Images were obtained at consecutive time points thereafter using the IVIS-200 system (Caliper) according to instructions of the manufacturer. The software Living image 3.0 (Caliper) was used for image analysis and quantification of emission intensities.
In-vitro luciferase assays
Cells were stimulated with TLR2 agonists for 4 hours. The cells were lysed in Reporter Lysis Buffer (Promega). 5 µl of cell lysate was added to 25 µl of Luciferase Assay Reagent II (Promega) and Images were obtained using the IVIS-200 system (Caliper) Luminescence was quantified as described above.
Data in the text and figures are expressed as the mean +/− SD. Statistical comparisons were carried out using the Student's t-test. P values less than 0.05 were considered significant.
RNA extraction was performed with the Quiagen Rneasy Mini Kit according to the manufacturer's instructions. RT-PCR was performed with the GoTaq Flexi DNA Polymerase from Promega and Real time PCR was carried out using Power SYBR Green PCR Master Mix from Applied Biosystems with the following primers (operon):
TNF-α sense (5′-TCTCATCAGTTCTATGGCCC-3′),
TNF-α antisense (5′-GGGAGTAGACAAGCTACAAC-3′),
IP-10 sense (5′-GATGACGGGCCAGTGAGAATGAG-3′),
IP-10 antisense (5′-CTGGGTAAAGGGGAGTGATGGAGA-3′),
Mx-2 sense (5′-GAGAATGTCGCCTATTCACCAGGCTC-3′),
Mx-2 antisense (5′-GCCATGGATGAGGTCTGTAAATCTTGGG-3′),
IFN-β sense (5′-CATCAACTATAAGCAGCTCCA-3′),
IFN-β antisense (5′-TTCAAGTGGAGAGCAGTTGAG-3′),
RPS9 sense (5′-CTGGACGAGGGCAAGATGAAGC-3′)
and RPS9 antisense (5′-TGACGTTGGCGGATGAGCACA-3′).
TLR2 activation induces type I IFN dependent responses via MyD88. WT, TLR2−/−, MyD88−/−, IFN-β−/− and IFNAR1−/− macrophages were stimulated (or unstimulated) with 1.5 µg/ml of MALP-2, PAM2CSK4 or PAM3CSK4. After 24 hours respective culture supernatants were analyzed for IL-6 by ELISA (A) or Nitric oxide/Nitrite by Griess reaction (B). Data are representative of more than four independent experiments (mean ± s.e.m.).
(1.41 MB TIF)
In vivo and in vitro analysis of IFN-β induction using the Luciferase reporter system. A shows the kinetics of IFN-β induction in TLR2+/+IFN-β+/Δβ-luc (TLR2+/+) and TLR2−/−IFN-β+/Δβ-luc (TLR2−/−) mice injected with MALP-2 (*: p<0.05). The corresponding luminescence images are depicted in Fig. 3C–D. Dendritic cells (BMDCs) or macrophages (BMDMs) derived from TLR2+/+IFN-β+/Δβ-luc (TLR2+/+) and TLR2−/−IFN-β+/Δβ-luc (TLR2−/−) mice were stimulated with indicated concentrations of MALP-2 or LPS for 4 hours and analyzed for luciferase expression. B shows quantification of fold increase in luciferase expression (i.e., IFN-β induction) in BMDCs (mean ± s.e.m.). C show representative luminescence images of macrophages while D indicates the fold increase in luciferase in corresponding cells (mean ± s.e.m.). Note that the luminescence colour scale bar for LPS and MALP-2 are different.
(3.10 MB TIF)
Ligand induced internalization and trafficking of TLR2 into endolysosomal compartments. RAW 264.7 macrophages transfected with hTLR2-GFP were loaded with fluorescent Dextran-Alexa Fluor 647, then left untreated (A) or stimulated with 1.5 µg/ml of MALP-2 and imaged by confocal microscopy after 3 hours (B). Panel C shows cells first labelled with Detran-Alexa and then pre-incubated for 1 h with 50 µM Bafilomycin A before stimulation with MALP-2 for 3 hours.
(1.38 MB TIF)
Dynasore, an inhibitor of dynamin 1, abrogates TLR2-driven IFN-β induction. BMDM from TLR2+/+IFN-β+/Δβ-luc mice pre-incubated for 1 h with indicated concentrations of Dynasore were stimulated with or without 1.5 µg/ml of MALP-2 for 4 h and analyzed for luciferase expression. A shows representative luminescence images of stimulated macrophages while B indicates the corresponding fold increase in luciferase activity. Luciferase expression is represented by a colour shift from blue to red. Data represent similar results from two independent experiments (mean ± s.e.m.). *: p<0.05 indicate statistical significance in the deference between Dynasore treated and untreated cells that were stimulated with MALP-2.
(3.00 MB TIF)
We wish to thank Susanne zur Lage, Martina Krey and Regina Lesch for expert technical assistance and P. Mueller and M. Probst-Kepper for critical reading of the manuscript. We are also grateful to U. Klemm, MPI-Berlin for providing us with MyD88−/− mice and A. Kroeger HZI- Braunschweig for supplying us with IRF1−/− mice. We are grateful to T. Taniguchi (Tokyo Univ., Tokyo, Japan) for the permission to use IRF3−/− IRF7−/− mice.
Conceived and designed the experiments: NOG. Performed the experiments: ND NOG. Analyzed the data: ND SL NOG. Contributed reagents/materials/analysis tools: SW. Wrote the paper: ND NOG.
- 1. Kawai T, Akira S (2009) The roles of TLRs, RLRs and NLRs in pathogen recognition. Int Immunol 21: 317–337.
- 2. Fitzgerald KA, Rowe DC, Barnes BJ, Caffrey DR, Visintin A, et al. (2003) LPS-TLR4 signaling to IRF-3/7 and NF-kappaB involves the toll adapters TRAM and TRIF. J Exp Med 198: 1043–1055.
- 3. Kenny EF, O'Neill LA (2008) Signalling adaptors used by Toll-like receptors: an update. Cytokine 43: 342–349.
- 4. O'Neill LA, Bowie AG (2007) The family of five: TIR-domain-containing adaptors in Toll-like receptor signalling. Nat Rev Immunol 7: 353–364.
- 5. Uematsu S, Akira S (2007) Toll-like receptors and Type I interferons. J Biol Chem 282: 15319–15323.
- 6. Noppert SJ, Fitzgerald KA, Hertzog PJ (2007) The role of type I interferons in TLR responses. Immunol Cell Biol 85: 446–457.
- 7. Bogdan C, Mattner J, Schleicher U (2004) The role of type I interferons in non-viral infections. Immunol Rev 202: 33–48.
- 8. Takeuchi O, Akira S (2009) Innate immunity to virus infection. Immunol Rev 227: 75–86.
- 9. Vilcek J (2006) Fifty years of interferon research: aiming at a moving target. Immunity 25: 343–348.
- 10. Ablasser A, Bauernfeind F, Hartmann G, Latz E, Fitzgerald KA, et al. (2009) RIG-I-dependent sensing of poly(dA:dT) through the induction of an RNA polymerase III-transcribed RNA intermediate. Nat Immunol 10: 1065–1072.
- 11. Chiu YH, Macmillan JB, Chen ZJ (2009) RNA polymerase III detects cytosolic DNA and induces type I interferons through the RIG-I pathway. Cell 138: 576–591.
- 12. Bafica A, Santiago HC, Goldszmid R, Ropert C, Gazzinelli RT, et al. (2006) Cutting edge: TLR9 and TLR2 signaling together account for MyD88-dependent control of parasitemia in Trypanosoma cruzi infection. J Immunol 177: 3515–3519.
- 13. Doyle S, Vaidya S, O'Connell R, Dadgostar H, Dempsey P, et al. (2002) IRF3 mediates a TLR3/TLR4-specific antiviral gene program. Immunity 17: 251–263.
- 14. Hoshino K, Kaisho T, Iwabe T, Takeuchi O, Akira S (2002) Differential involvement of IFN-beta in Toll-like receptor-stimulated dendritic cell activation. Int Immunol 14: 1225–1231.
- 15. Kagan JC, Su T, Horng T, Chow A, Akira S, et al. (2008) TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon-beta. Nat Immunol 9: 361–368.
- 16. Kawai T, Takeuchi O, Fujita T, Inoue J, Muhlradt PF, et al. (2001) Lipopolysaccharide stimulates the MyD88-independent pathway and results in activation of IFN-regulatory factor 3 and the expression of a subset of lipopolysaccharide-inducible genes. J Immunol 167: 5887–5894.
- 17. Toshchakov V, Jones BW, Perera PY, Thomas K, Cody MJ, et al. (2002) TLR4, but not TLR2, mediates IFN-beta-induced STAT1alpha/beta-dependent gene expression in macrophages. Nat Immunol 3: 392–398.
- 18. Barbalat R, Lau L, Locksley RM, Barton GM (2009) Toll-like receptor 2 on inflammatory monocytes induces type I interferon in response to viral but not bacterial ligands. Nat Immunol 10: 1200–1207.
- 19. Lienenklaus S, Cornitescu M, Zietara N, Lyszkiewicz M, Gekara N, et al. (2009) Novel reporter mouse reveals constitutive and inflammatory expression of IFN-beta in vivo. J Immunol 183: 3229–3236.
- 20. Coccia EM, Severa M, Giacomini E, Monneron D, Remoli ME, et al. (2004) Viral infection and Toll-like receptor agonists induce a differential expression of type I and lambda interferons in human plasmacytoid and monocyte-derived dendritic cells. Eur J Immunol 34: 796–805.
- 21. Kerkmann M, Rothenfusser S, Hornung V, Towarowski A, Wagner M, et al. (2003) Activation with CpG-A and CpG-B oligonucleotides reveals two distinct regulatory pathways of type I IFN synthesis in human plasmacytoid dendritic cells. J Immunol 170: 4465–4474.
- 22. Mancuso G, Gambuzza M, Midiri A, Biondo C, Papasergi S, et al. (2009) Bacterial recognition by TLR7 in the lysosomes of conventional dendritic cells. Nat Immunol 10: 587–594.
- 23. Honda K, Taniguchi T (2006) IRFs: master regulators of signalling by Toll-like receptors and cytosolic pattern-recognition receptors. Nat Rev Immunol 6: 644–658.
- 24. Honda K, Ohba Y, Yanai H, Negishi H, Mizutani T, et al. (2005) Spatiotemporal regulation of MyD88-IRF-7 signalling for robust type-I interferon induction. Nature 434: 1035–1040.
- 25. Honda K, Yanai H, Negishi H, Asagiri M, Sato M, et al. (2005) IRF-7 is the master regulator of type-I interferon-dependent immune responses. Nature 434: 772–777.
- 26. Honda K, Yanai H, Mizutani T, Negishi H, Shimada N, et al. (2004) Role of a transductional-transcriptional processor complex involving MyD88 and IRF-7 in Toll-like receptor signaling. Proc Natl Acad Sci U S A 101: 15416–15421.
- 27. Negishi H, Fujita Y, Yanai H, Sakaguchi S, Ouyang X, et al. (2006) Evidence for licensing of IFN-gamma-induced IFN regulatory factor 1 transcription factor by MyD88 in Toll-like receptor-dependent gene induction program. Proc Natl Acad Sci U S A 103: 15136–15141.
- 28. Triantafilou M, Manukyan M, Mackie A, Morath S, Hartung T, et al. (2004) Lipoteichoic acid and toll-like receptor 2 internalization and targeting to the Golgi are lipid raft-dependent. J Biol Chem 279: 40882–40889.
- 29. Blander JM, Medzhitov R (2004) Regulation of phagosome maturation by signals from toll-like receptors. Science 304: 1014–1018.
- 30. Sanjuan MA, Dillon CP, Tait SW, Moshiach S, Dorsey F, et al. (2007) Toll-like receptor signalling in macrophages links the autophagy pathway to phagocytosis. Nature 450: 1253–1257.
- 31. Underhill DM, Ozinsky A, Hajjar AM, Stevens A, Wilson CB, et al. (1999) The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401: 811–815.
- 32. Mun HS, Aosai F, Norose K, Chen M, Piao LX, et al. (2003) TLR2 as an essential molecule for protective immunity against Toxoplasma gondii infection. Int Immunol 15: 1081–1087.
- 33. Hickman-Davis J, Gibbs-Erwin J, Lindsey JR, Matalon S (1999) Surfactant protein A mediates mycoplasmacidal activity of alveolar macrophages by production of peroxynitrite. Proc Natl Acad Sci U S A 96: 4953–4958.
- 34. Cole BC, Mu HH, Pennock ND, Hasebe A, Chan FV, et al. (2005) Isolation and partial purification of macrophage- and dendritic cell-activating components from Mycoplasma arthritidis: association with organism virulence and involvement with Toll-like receptor 2. Infect Immun 73: 6039–6047.
- 35. Muhlradt PF, Kiess M, Meyer H, Sussmuth R, Jung G (1997) Isolation, structure elucidation, and synthesis of a macrophage stimulatory lipopeptide from Mycoplasma fermentans acting at picomolar concentration. J Exp Med 185: 1951–1958.
- 36. Zhu J, Martinez J, Huang X, Yang Y (2007) Innate immunity against vaccinia virus is mediated by TLR2 and requires TLR-independent production of IFN-beta. Blood 109: 619–625.
- 37. Gilliet M, Cao W, Liu YJ (2008) Plasmacytoid dendritic cells: sensing nucleic acids in viral infection and autoimmune diseases. Nat Rev Immunol 8: 594–606.
- 38. Verthelyi D, Ishii KJ, Gursel M, Takeshita F, Klinman DM (2001) Human peripheral blood cells differentially recognize and respond to two distinct CPG motifs. J Immunol 166: 2372–2377.
- 39. Latz E, Verma A, Visintin A, Gong M, Sirois CM, et al. (2007) Ligand-induced conformational changes allosterically activate Toll-like receptor 9. Nat Immunol 8: 772–779.
- 40. Ewald SE, Lee BL, Lau L, Wickliffe KE, Shi GP, et al. (2008) The ectodomain of Toll-like receptor 9 is cleaved to generate a functional receptor. Nature 456: 658–662.
- 41. Matsuyama T, Kimura T, Kitagawa M, Pfeffer K, Kawakami T, et al. (1993) Targeted disruption of IRF-1 or IRF-2 results in abnormal type I IFN gene induction and aberrant lymphocyte development. Cell 75: 83–97.
- 42. Muller U, Steinhoff U, Reis LF, Hemmi S, Pavlovic J, et al. (1994) Functional role of type I and type II interferons in antiviral defense. Science 264: 1918–1921.
- 43. Erlandsson L, Blumenthal R, Eloranta ML, Engel H, Alm G, et al. (1998) Interferon-beta is required for interferon-alpha production in mouse fibroblasts. Curr Biol 8: 223–226.
- 44. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, et al. (1999) Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cell wall components. Immunity 11: 443–451.
- 45. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, et al. (1998) Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9: 143–150.
- 46. Sato M, Suemori H, Hata N, Asagiri M, Ogasawara K, et al. (2000) Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for IFN-alpha/beta gene induction. Immunity 13: 539–548.
- 47. Gekara NO, Westphal K, Ma B, Rohde M, Groebe L, et al. (2007) The multiple mechanisms of Ca2+ signalling by listeriolysin O, the cholesterol-dependent cytolysin of Listeria monocytogenes. Cell Microbiol 9: 2008–2021.