TLR3 and cytoplasmic RIG-I-like receptor (RLR) recognize virus-derived dsRNA and induce type I IFN production in a distinct manner. Human TLR3 localizes to the endosomal compartments in myeloid dendritic cells (mDCs), while it localizes to both the cell surface and interior in fibroblasts and epithelial cells. TLR3 signaling arises in the intracellular compartment in both cell types and requires endosomal maturation. The mechanisms by which extracellular dsRNA is delivered to the TLR3-containing organelle remain largely unknown. Among various synthetic dsRNAs, poly(I:C) is preferentially internalized and activates TLR3 in mDCs. In vitro transcribed dsRNAs hardly induce IFN-β production in mDCs. In this study, we demonstrate that the clathrin-dependent endocytic pathway mediates cell entry of poly(I:C) to induce IFN-β gene transcription. Furthermore, poly(I:C)-induced IFN-β production is inhibited by pretreatment of cells with B- and C-type oligodeoxynucleotides (ODNs) but not with TLR7/8 ligands. The binding and internalization of B-type ODNs by mDCs was reduced in the presence of poly(I:C), suggesting that poly(I:C) shares the uptake receptor with B- and C-type ODNs. Hence, foreign dsRNA is recognized by differently categorized receptors, cytoplasmic RIG-I-like receptor, membrane-bound TLR3 and cell-surface RNA capture. The endocytic pathway is critical for dsRNA-induced TLR3-mediated cell activation.
Type I IFNs (IFN-α/β) play essential roles in both innate and adaptive antiviral immune responses (1, 2). Many types of cells such as fibroblasts, epithelial cells, and dendritic cells (DCs)5 produce IFN-β upon viral infection or stimulation with poly(I:C), a synthetic analog of viral dsRNA (3). Membrane-bound TLR3 and cytoplasmic DEAD/H box RNA helicases, such as retinoic acid-inducible gene-I and melanoma differentiation-associated gene 5, participate in the recognition of virus-derived dsRNA and induction of IFN-α/β gene transcription (4, 5, 6, 7, 8, 9).
Human TLR3 localizes to the endosomal compartments in myeloid DCs (mDCs), while it localizes to both the cell-surface and interior of fibroblasts and epithelial cells (5, 10, 11). Anti-human TLR3 mAb inhibits poly(I:C)-induced IFN-β production in fibroblasts, indicating that TLR3 present on the cell surface participates in dsRNA recognition (5). However, in both cell types, TLR3 signaling arises in an intracellular compartment and requires endosomal maturation (10, 11). After dsRNA recognition, TLR3 homodimerizes, and this is followed by recruitment of an adaptor molecule, i.e., Toll-IL-1 receptor domain-containing adaptor molecule-1 (TICAM-1, also called Toll-IL-1 receptor domain-containing adaptor inducing IFN-β). This activates the NF-κB and interferon regulatory factor-3 transcription factors leading to IFN-β production (12, 13, 14, 15). However, the mechanism by which extracellular dsRNA is delivered to the TLR3-positive organelle is unknown.
A recent study has shown that CD14 directly binds to poly(I:C) and mediates poly(I:C) cellular uptake (16). Bone marrow-derived macrophages from CD14−/− mice exhibited impaired responses to poly(I:C) (16). CD14 is a well-known cell-surface pattern-recognition receptor that is involved in both LPS-mediated TLR4 signaling and in TLR2 signaling (17, 18). However, mDCs do not express CD14 (19), suggesting that other cell-surface molecules mediate the entry of dsRNA into mDCs. In this study, we used pharmacological inhibitors to analyze the mechanisms by which extracellular dsRNAs activate endosomal TLR3. We found that the clathrin-dependent endocytic pathway participates in poly(I:C)-induced IFN-β production in mDCs. Furthermore, an inhibition study with various nucleic acids revealed that poly(I:C) shares its uptake receptor with B- and C-type oligodeoxynucleotides (ODNs).
Materials and Methods
Cell culture and reagents
A human embryonic kidney cell line HEK293 was obtained from Sumitomo Pharmaceuticals and maintained in DMEM supplemented with 10% heat-inactivated FCS (JRH Biosciences) and antibiotics. The HEK293 cells have no TLR3. We prepared TLR3-expressing HEK293 cells by transient transfection of the expression plasmid for human TLR3, which predominantly express TLR3 intracellularly but possess some TLR3 molecules on the cell-surface. Chloroquine, chlorpromazine, cytochalasin D, methyl-β-cyclodextrin, 4′,6-diamidino-2-phenylindole (DAPI), propidium iodide, control ODN2006, LPS from Escherichia coli (serotype 0111:B4), and polymyxin B were purchased from Sigma-Aldrich. Alexa Fluor 488-acetylated low-density lipoprotein (AcLDL) and Alexa Fluor 488-cholera toxin B subunit (CTXB) were from Molecular Probes. Poly(I:C) was from Amersham Bioscience. Imiquimod, Gardiquimod, CL075, ODN2006, ODN2216, ODNM362, FITC-ODN2006, and FITC-ODN2216 were purchased from Invivogen. Poly I was provided from Dr. Nishikawa (Institute for Biological Resources and Function, Tsukuba, Japan). Anti-human TLR3 mAb (clone TLR3.7) was generated in our laboratory (5). Anti-dsRNA mAb (K1) (20) was purchased from BioLink. Mouse IgG1 and mouse IgG2a were from Sigma-Aldrich, anti-CD83 mAb was from Ancell, Alexa Fluor568- conjugated goat anti-mouse IgG was from Molecular Probes, and FITC-labeled goat anti-mouse IgG was from American Qualex. Cytochalasin D was dissolved in DMSO at the concentration of 1 mg/ml.
dsRNAs of various lengths (N100, N200, N500, and N1000) were synthesized using a MEGAscript RNA Kit (Ambion) as described previously (21). cDNA of the N-protein of the measles virus (MV) strain Edmonston was used as a template for the transcription reaction. The synthetic dsRNAs were treated with polymyxin B (final 10 μg/ml) for 1 h at 37°C before stimulation of the cells. Treatment of dsRNA with calf intestine alkaline phosphatase (CIAP) was performed as described (22). In brief, 10 μg dsRNA was treated with 30 U CIAP (Roche) for 3 h at 37°C in the presence of 10 U RNase inhibitor (Promega). The enzyme was removed by using the RNeasy Mini kit (Qiagen).
CD14+ monocytes were isolated from human PBMCs using the MACS system (Miltenyi Biotec). The monocytes were cultured for 6 days in RPMI 1640 supplemented with 10% heat-inactivated FCS and antibiotics in the presence of 500 U/ml GM-CSF and 100 U/ml IL-4 (PeproTech) to obtain monocyte-derived immature DCs (iDCs) (19). iDCs in 96-well round-bottom plates (1 × 106/ml) were stimulated with synthetic duplex RNA or poly(I:C) for 24 h in the presence of 500 U/ml GM-CSF. The amount of IFN-β present in the culture supernatants was measured by ELISA kit (TFB). In some experiments, dsRNA (1 μg) was mixed with 6 μl of N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate (DOTAP) (Roche) in OPTI-MEM (total 30 μl) and incubated for 10 min at room temperature. The mixture was added to iDCs (6 × 105/ml). In the case of inhibition assays, cells were preincubated with indicated concentrations of pharmacological inhibitors or nucleic acids for 1 h at 37°C and then stimulated with 10 μg/ml poly(I:C) for 24 h. At the time of supernatant collection, cells were washed twice with Dulbecco’s PBS (DPBS) and then stained with propidium iodide (50 μg/ml in DPBS) for 10 min at room temperature. The viability of the cells was estimated by flow cytometry.
Assay for clathrin-dependent and independent endocytosis
iDCs and HEK293 cells were pretreated with chlorpromazine (25 μg/ml for iDCs, 50 μg/ml for HEK293 cells), methyl-β-cyclodextrin (1 mM), or medium alone for 1 h at 37°C and subsequently incubated with Alexa Fluor 488-AcLDL (0.2 μM for iDC, 0.02 μM for HEK293) or Alexa Fluor 488-CTXB (5 μg/ml for iDC, 50 μg/ml for HEK293) for 30 min at 4°C. The cells were then warmed for 5 min at 37°C to allow endocytosis to occur (23). For quenching the fluorescence of uningested Alexa Fluor 488-AcLDL or Alexa Fluor 488-CTXB, the cell suspensions were mixed with trypan blue solution (2 mg/ml in DPBS) and analyzed by flow cytometry (19).
Complementary DNA expression vector
Complementary DNA for human TLR3 was cloned in our laboratory by RT-PCR and was ligated into the cloning site of the expression vector pEFBOS, a gift from Dr. S. Nagata (Osaka University, Osaka, Japan).
Reporter gene assay
HEK293 cells (2 × 105 cells per well) seeded in 24-well plates were transiently transfected with pEFBOS/TLR3 (0.1 μg) or empty vector together with a luciferase-linked p-125 luc reporter plasmid (0.1 μg) using Lipofectamine 2000 reagent (Invitrogen). The p-125 luc reporter containing the human IFN-β promoter region (−125 to +19) was provided by Dr. T. Taniguchi (University of Tokyo, Tokyo, Japan). The total amount of transfected DNA (0.8 μg) was kept constant by adding empty vector. The plasmid phRL-TK (2.5 ng) was used as an internal control. Twenty-four hours after transfection, cells were washed and stimulated with medium alone or polymyxin B-treated dsRNA for 6 h. In some experiments, cells were incubated with dsRNA complexed with DOTAP for 6 h. In the inhibition assays, cells were preincubated with inhibitors or nucleic acids for 1 h at 37°C and then stimulated with 10 μg/ml poly(I:C) for 6 h. The cells were lysed in lysis buffer (Promega), and dual luciferase activities were measured according to the manufacturer’s instructions. The firefly luciferase activity was normalized by Renella luciferase activitiy and is expressed as the fold induction relative to the activity of unstimulated vector-transfected cells.
Monocyte-derived iDCs were pretreated with medium alone or 5 μg/ml chloroquine for 2 h at 37°C and then stimulated with 100 ng/ml LPS or 10 μg/ml polymyxin B treated-poly(I:C) for 24 h. After washing, cells were incubated with mouse IgG1, or anti-CD83 mAb (1 μg) in the presence of human IgG (10 μg) for 30 min at 4°C in FACS buffer (DPBS containing 0.5% BSA and 0.1% sodium azide). After the cells were washed twice with the above buffer, FITC-labeled secondary Ab (American Qualex) was added and the cells were further incubated for 30 min at 4°C. In the case of dsRNA binding assay, cells were incubated with the indicated concentrations of poly(I:C) or N500 in culture medium for 30 min at 4°C. After washing, cells were labeled with anti-dsRNA mAb (K1) or control mouse IgG2a (1 μg) for 30 min at 4°C and then incubated with FITC-labeled secondary Ab. The cells were analyzed on a FACSCalibur (BD Biosciences). For examination of binding and internalization of ODNs, iDCs were incubated with FITC-labeled ODN2006 or ODN2216 in the presence or absence of poly(I:C) for 2 h at 37°C. After washing, cells were analyzed on FACSCalibur.
Monocyte-derived iDCs (1 × 106/ml) were incubated with 2 μM FITC-ODN2006 for 30 min at 37°C. Cells were washed three times and treated with permeabilizing solution (BD Biosciences) for 10 min at room temperature. After washing, cells were stained with mouse IgG1 or anti-TLR3 mAb (TLR3.7) (20 μg/ml) in FACS buffer for 1 h at room temperature. Alexa Fluor 568-conjugated secondary Ab (1/400 diluted with PBS containing 10% BlocAce and 10% goat serum) was used to visualize staining of the primary Abs. Nuclei were stained with DAPI (2 μg/ml in PBS) for 10 min before mounting onto glass slides using PBS containing 2.3% DABCO and 50% glycerol. Cells were visualized at a magnification of ×63 with an LSM510 META microscope (Zeiss).
Unresponsiveness of mDCs to synthetic virus-derived dsRNA
mDCs express TLR3 intracellularly and produce IFN-β in response to poly(I:C), which is a synthetic TLR3 ligand (10). To analyze the mechanism by which mDCs recognize extracellular dsRNA, we examined whether synthetic virus-derived dsRNA activates DCs to produce IFN-β. dsRNAs of various lengths were in vitro transcribed using MV cDNA encoding N-protein as the template. First, we examined the abilities of synthetic dsRNAs to activate the IFN-β promoter in HEK293 cells transiently expressing human TLR3, which predominantly express TLR3 intracellularly but possess some TLR3 molecules on the cell surface. The TLR3-acitivating abilities of these dsRNAs were remarkably lower than that of poly(I:C) when extracellularly added to the cells (Fig. 1,A, left panel), while they significantly activated TLR3 when introduced into the cells using a cationic liposome DOTAP (Fig. 1,A, center and right panel). Next, the DC-activating abilities of these dsRNAs were analyzed by measuring IFN-β production in monocyte-derived iDCs. The synthetic dsRNAs failed to induce IFN-β production in iDCs (Fig. 1,B, left panel), as previously observed (21). In contrast, when dsRNA was targeted to endosomal TLR3 using DOTAP, all MV-N-derived dsRNAs induced IFN-β production at a level that was similar to or higher than that from poly(I:C) stimulation (Fig. 1 B, right panel). Since these in vitro-transcribed dsRNAs contain 5′-ppp, we treated dsRNA with CIAP. Once again, extracellular CIAP-treated dsRNAs did not induce IFN-β production in iDCs, suggesting that the DC-activating ability of dsRNA is independent of its 5′ structure (Fig. 1 C).
The clathrin-mediated endocytic pathway participates in dsRNA-induced IFN-β production
Poly(I:C)-induced IFN-β production and costimulatory molecule (CD83) up-regulation were inhibited by pretreatment of the cells with chloroquine, an inhibitor of endosomal maturation (Fig. 2,A). The endocytic pathway that participates in TLR3-mediated signaling was analyzed using the pharmacological inhibitors cytochalasin D (a phagocytosis inhibitor), methyl-β-cyclodextrin (a caveolae-mediated endocytosis inhibitor) (23, 24, 25), and chlorpromazine (a clathrin-mediated endocytosis inhibitor) (23, 26). We first evaluated various concentrations of these inhibitors to maximize their specificity and eliminate toxic side effects. Under the experimental setting, cell damage was negligible within the concentrations of the inhibitors used in this study. Under optimized conditions, chlorpromazine inhibited the internalization of AlexaFluor 488-AcLDL, a marker for the clathrin-dependent pathway (27), but had no effect on the uptake of AlexaFluor 488-CTXB, a marker for caveolar-mediated internalization (28). In contrast, methyl-β-cyclodextrin significantly inhibited the internalization of AlexaFluor 488-CTXB, but had few effects on AlexaFluor 488-AcLDL uptake (Fig. 2,B). IFN-β production was inhibited by pretreatment of iDCs with cytochalasin D and chlorpromazine but not with methyl-β-cyclodextrin, suggesting that the clathrin-dependent endocytic pathway mediates cell entry of poly(I:C) to induce IFN-β gene transcription (Fig. 2,C). In addition, poly I, an inhibitor of scavenger receptors (29), did not block poly(I:C)-induced IFN-β production in iDCs (Fig. 2,D). Although poly I has been shown to activate TLR3 in murine B cells and bone marrow-derived DCs (30), human monocyte-derived iDCs could not produce IFN-β in response to extracellular poly I. Similar results were obtained with TLR3-expressing HEK293 cells (Fig. 2, E and F). Poly(I:C)-induced TLR3-mediated IFN-β promoter activation was inhibited by pretreatment of cells with cytochalasin D and chlorpromazine but not with methyl-β-cyclodextrin or poly I.
Poly(I:C) binding to mDCs
The potent IFN-β-inducing ability of poly(I:C) may be due to the uptake of the latter by iDCs. To test poly(I:C) binding to unknown cell-surface receptors, iDCs and HEK293 cells were incubated with various poly(I:C) concentrations at 4°C for 30 min. As shown in Fig. 3, poly(I:C) bound to iDCs and HEK293 cells in a dose-dependent manner. The extent of poly(I:C) binding depended on the cell type and on the individual cells in the case of iDCs. In contrast, in vitro-transcribed dsRNAs hardly bound to iDCs and HEK293 cells.
CpG ODNs but not synthetic TLR7/8 ligands inhibit poly(I:C)-induced IFN-β production in mDCs
Previous studies demonstrated that CpG ODNs are endocytosed into mice bone marrow-derived DCs (31, 32). To examine the effects of nucleic acids on poly(I:C)-induced IFN-β production, inhibition analysis was performed using synthetic ligands of TLR7, 8, and 9. iDCs were preincubated with increasing concentrations of nucleic acids for 1 h and then stimulated with poly(I:C). The TLR7/8 ligands, i.e., Imiquimod, Gardiquimod, and CL075, did not affect poly(I:C)-induced IFN-β production (Fig. 4,A, data not shown). In contrast, the TLR9 ligands, i.e., CpG ODNs, inhibited poly(I:C)-induced IFN-β production in iDCs (Fig. 4,B). Both B-type ODN2206 and C-type ODN M362 that induce IL-6 production and B cell proliferation completely inhibited poly(I:C)-induced IFN-β production in the 1 to 10 μg/ml range, while A-type ODN2216 that induces robust IFN-α production in plasmacytoid DCs (pDCs) retarded poly(I:C) function in a dose-dependent manner (Fig. 4,B). In addition, control ODN2006, a non-TLR9 ligand, also completely inhibited poly(I:C)-induced IFN-β production in a manner similar to that of ODN2006 (Fig. 4,B), indicating that the inhibitory effect of the ODNs appears to depend on their specific sequences, not on their immunostimulatory features. In the 1 to 100 ng/ml range, ODN2006, ODN M362, and control ODN2006 inhibited poly(I:C)-induced IFN-β production in a dose-dependent manner (Fig. 4 C). Since 10 μg/ml poly(I:C) (approximate length, 1500 bps) and 67 ng/ml ODN2006 each correspond to 8 pmol/ml, it is possible that poly(I:C) competes with ODNs for binding to a cell surface molecule that mediates endocytosis.
In contrast, the inhibitory effects of ODNs on poly(I:C)-induced TLR3 signaling in TLR3-expressing HEK293 cells were somewhat different from those in iDCs. In the 2 to 50 μg/ml range, B-type ODN2006 and C-type ODN M362 inhibited poly(I:C)-induced IFN-β promoter activation in a dose-dependent manner, while A-type ODN2216 did not affect the poly(I:C) function (Fig. 5). In TLR3-expressing HEK293 cells, it is possible that some TLR3 molecules that were present on the cell surface contribute to poly(I:C) internalization in conjunction with the uptake receptor.
Poly(I:C) recognition by endosomal TLR3 in the presence of ODNs
The inhibitory effect of ODN2006 on poly(I:C)-induced IFN-β production was abrogated when the cells were washed after preincubation with ODNs and then stimulated with poly(I:C) (data not shown). It is necessary for ODNs and poly(I:C) to be present together to inhibit poly(I:C) function. To rule out the possibility that ODNs affect TLR3-mediated poly(I:C) recognition by forming a complex with poly(I:C), poly(I:C) was introduced into the endosomal compartment together with ODNs by using DOTAP, and IFN-β production or promoter activation was measured. As shown in Fig. 6, poly(I:C) induced IFN-β promoter activation through TLR3 regardless of whether the ODNs were present or absent. These results imply that ODNs inhibit poly(I:C) function during the binding and uptake processes.
ODN2006 shares the uptake receptor with poly(I:C)
To test the possibility that B- and C-type ODN compete for surface binding with poly(I:C), iDCs and HEK293 cells were incubated with poly(I:C) in the presence or absence of ODN2006 for 30 min at 4°C. The binding of poly(I:C) on the cell surface was assessed by flow cytometry with anti-dsRNA mAb and FITC-labeled secondary Ab. The poly(I:C) binding level was reduced in the presence of ODN2006 in both iDCs and HEK293 cells (Fig. 7). Similar results were obtained with ODNM362 (data not shown). We next analyzed the binding and internalization of FITC-labeled ODNs in iDCs by using flow cytometry. The binding and internalization of FITC-ODN2006 was observed in iDCs, and this was markedly reduced in the presence of poly(I:C) (Fig. 8,A, upper panel). In contrast, FITC-labeled ODN2216 hardly bound to iDCs (Fig. 8,A, lower panel), suggesting that an unknown membrane receptor mediates cell entry of ODN2006 and poly(I:C) but not that of ODN2216. Similar results were obtained with HEK293 cells (data not shown). To examine whether ODN2006 is delivered to the TLR3-positive organelle, iDCs were incubated with FITC-ODN2006 for 30 min at 37°C and then stained with anti-TLR3 mAb. Cells were analyzed by confocal microscopy. As shown in Fig. 8 B, ODN2006 partly colocalized with TLR3.
The mechanism by which exogenously added dsRNA can activate endosomal TLR3 is unknown. By using pharmacological inhibitors, we demonstrated that the clathrin-dependent endocytic pathway participates in dsRNA-mediated TLR3 activation. Among the synthetic dsRNAs, poly(I:C) is preferentially internalized and activates TLR3 in monocyte-derived iDCs resulting in potent IFN-β induction. In vitro transcribed MV-originated dsRNA hardly activates TLR3 when iDCs are stimulated extracellularly. Putative RNA capture at the cell surface may involve the binding and transfer of poly(I:C) but not that of MV-originated synthetic dsRNA. In addition, B- and C-type ODNs compete with poly(I:C) for cellular uptake and inhibit poly(I:C)-induced IFN-β production in iDCs in a dose-dependent manner. This indicates that in mDCs, the uptake receptor is shared by poly(I:C) and B- and C-type ODNs.
Lee et al. (16) reported that CD14 directly bound poly(I:C) small fragments and mediated the cellular uptake of poly(I:C) small fragments in mice bone marrow-derived macrophages. In the human fibroblast cell line MRC5, which expresses TLR3 and CD14 on the cell surface, poly(I:C)-induced IFN-β production was inhibited by pretreatment of the cells with chloroquine (M. Matsumoto, S. Okahira, and T. Seya, unpublished data). Anti-human TLR3 mAb (TLR3.7) inhibits poly(I:C)-induced IFN-β production in MRC5 cells, indicating that cell-surface TLR3 might be involved in poly(I:C) recognition (5). Furthermore, surface expression of both TLR3 and CD14 was decreased 30 min after poly(I:C) stimulation (M. Matsumoto, unpublished data), suggesting that CD14 and TLR3 participate in poly(I:C) uptake in human fibroblasts. In contrast, monocyte-derived iDCs and HEK293 cells do not express CD14 (19). Thus, another cell surface molecule must be involved in the cellular uptake of poly(I:C) in both cell types.
Our results showed that poly(I:C)-induced IFN-β production in iDCs was inhibited only when ODNs were also present and that the synthetic ligands for TLR7/8 did not inhibit this production (Fig. 4). Since TLR9 was not expressed on human mDCs (data not shown and Ref. 33), it is unlikely that the inhibitory effect of ODNs is caused by TLR9-mediated signaling. Additionally, all types of ODNs did not affect poly(I:C)-mediated IFN-β production when poly(I:C) and ODNs were delivered to endosomal compartments with DOTAP (Fig. 6). Finally, surface binding of poly(I:C) was reduced in the presence of B- and C-type ODNs (Fig. 7 and data not shown) and also, binding and internalization of FITC-labeled ODN2006 (type B) was decreased in the presence of an equimolar concentration of poly(I:C) (Fig. 8 A). These results suggest that poly(I:C) shares the uptake receptor with B- and C-type ODNs in iDCs and that the inhibitory effects of B- and C-type ODNs rely on competitive binding between poly(I:C) and B- and C-type ODNs to the cell-surface receptor. In contrast, the binding and internalization of A-type ODN2216 was hardly observed in iDCs. A-type CpG ODNs (CpG-A) are potent IFN-α inducers in pDCs (34, 35). TLR9 activation by CpG-A occurs in the early endosome and leads to IFN-α production, whereas B-type CpG ODNs (CpG-B) localize to the lysosome and promote pDCs maturation (31, 32). In contrast, both CpG-A and CpG-B are delivered to the lysosome in mouse conventional DCs that express TLR9 intracellularly (31). These results imply that CpG-A and CpG-B are endocytosed through different cell surface receptors and pDCs possess specific machinery that retains the uptake receptor for CpG-A into the early endosome. In human monocyte-derived iDCs, the uptake receptor for A-type ODNs is hardly expressed, suggesting that the expression patterns of uptake receptors differ between human and mouse cells. The inhibitory effect of ODN2216 observed at excess concentrations in iDCs may reflect the weak affinity of A-type ODNs to the poly(I:C) uptake receptor.
It has been shown that poly(I:C) is delivered to the late endosome and then to the lysosome in CHO cells (16). In contrast to TLR9-MyD88 signaling, it appears that TLR3-TICAM-1 signaling does not require endosomal retention of poly(I:C). Once endosomal TLR3 is activated by poly(I:C), an adaptor molecule TICAM-1 transiently colocalizes with TLR3 and then dissociates from the receptor and forms the TICAM-1 signalosome (36). We have observed the colocalization of FITC-labeled ODN2006 and TLR3 in iDCs, indicating that the intracellular trafficking of poly(I:C) and ODN2006 is the same.
Natural and synthetic TLR7 ligands inhibit CpG-A- and CpG-C-ODN-induced IFN-α production in human pDCs (37). The TLR7 signal appears to regulate the outcome of TLR7 ligand/CpG-A-ODN costimulation. Although human mDCs express TLR8, the synthetic TLR8 ligand CL075 did not induce IFN-β production; further it did not affect poly(I:C)-induced IFN-β production in iDCs (Fig. 4 B). Similar results were obtained with TLR3-expressing HEK293 cells and from the reporter assay for IFN-β promoter activation. The mechanisms by which extracellular TLR7/8 ligands are delivered to intracellular TLR7/8 may differ from those of TLR3/9.
Our data clearly demonstrated that mDCs possess the uptake receptor for poly(I:C) but not that for in vitro-transcribed dsRNA. B- and C-type ODNs share the internalization receptor with poly(I:C). Both poly(I:C) and ODNs are synthetic nucleic acids that do not represent natural virus products, and there are no common structural motifs. The molecular structure of nucleic acids required for recognition by the cell surface receptor should be different from that required for TLR3 activation. It is important to investigate whether virus-derived RNA activates TLR3 extracellularly. TLR3-mediated mDCs activation results in the production of IL-12p70 and IFN-α/β and DC maturation that in turn activates NK cells and CTL (38, 39). At least in mDCs in which TLR3 is localized intracellularly, the ligand properties recognized by the catch-up receptor are critical for TLR3 activation. Identification of virus-derived RNA that is recognized by both the catch-up receptor and TLR3 would clarify the in vivo function of TLR3 during viral infection.
We thank Drs. T. Ebihara, H. Oshiumi, M. Shingai, M. Sasai, and A. Matsuo in our laboratory for valuable discussions and H. Itoh for providing the materials. We also thank Dr. T. Taniguchi (University of Tokyo, Japan) and Dr. S. Nishikawa (Institute for Biological Resource and Function, Tsukba, Japan) for providing the reagents.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by Grants-in-Aid from the Ministry of Education, Science, and Culture, the Ministry of Health, Labor, and Welfare of Japan, and by the Uehara Memorial Foundation, the Mitsubishi Foundation, the NorthTec Foundation, and the Akiyama Foundation. Financial supports by the Sapporo Biocluster “Bio-S” the Knowledge Cluster Initiative of the MEXT, and the Program of Founding Research Centers for Emerging and Reemerging Infectious Diseases, MEXT are gratefully acknowledged.
Abbreviations used in this paper: DC, dendritic cell; mDC, myeloid DC; TICAM-1, Toll-IL-1 receptor-containing adaptor molecule-1; ODN, oligodeoxynucleotide; AcLDL, acetylated low density lipoprotein; CTXB, cholera toxin subunit B; MV, measles virus; CIAP, calf intestine alkaline phosphatase; iDC, immature DC; pDC, plasmacytoid DC; CpG-B, B-type CpG ODN; CpG-A, a-type CpG ODN; DPBS, Dulbecco’s PBS; DOTAP, N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate; DAPI, 4′,6-diamidino-2-phenylindole.