IRF-8/Interferon (IFN) Consensus Sequence-binding Protein Is Involved in Toll-like Receptor (TLR) Signaling and Contributes to the Cross-talk between TLR and IFN-γ Signaling Pathways*

Toll-like receptor (TLR) and interferon-γ (IFN-γ) signaling pathways are important for both innate and adaptive immune responses. However, the cross-talk between these two signaling pathways is incompletely understood. Here we show that IFN-γ and LPS synergistically induce the expression of proinflammatory factors, including interleukin-1 (IL-1), IL-6, IL-12, NO, and tumor necrosis factor-α (TNF-α). Comparable synergism was observed between IFN-γ and peptidoglycan (PGN; a TLR2 ligand) and poly(I:C) (a TLR3 ligand) in the induction of IL-12 promoter activity. IFN-γ enhanced lipopolysaccharide (LPS)-induced ERK and JNK phosphorylation but had no effect on LPS-induced NF-κB activation. Interestingly, we found that IRF-8–/– macrophages were impaired in the activation of LPS-induced ERK and JNK and the production of proinflammatory cytokines induced by LPS or IFN-γ plus LPS. Retroviral transduction of IRF-8 into IRF-8–/– macrophages rescued ERK and JNK activation. Furthermore, co-immunoprecipitation experiments show that IRF-8 physically interacts with TRAF6 at a binding site between amino acid residues 356 and 305 of IRF-8. Transfection of IRF-8 enhanced TRAF6 ubiquitination, which is consistent with a physical interaction of IRF-8 with TRAF6. Taken together, the results suggest that the interaction of IRF-8 with TRAF6 modulates TLR signaling and may contribute to the cross-talk between IFN-γ and TLR signal pathways.

interacts with the TLR signaling pathways resulting in the cross-talk between TLR and IFN-␥ signaling pathways.
In this study, we show that IFN-␥ and LPS synergistically activate ERK and JNK kinases resulting in the synthesis of macrophage proinflammatory factors. The activation of ERK and JNK kinases is significantly impaired in macrophages from IRF-8-deficient mice. Furthermore, IRF-8 binds to TRAF6 and regulates TRAF6 ubiquitination. These results show that IRF-8 is an important molecule mediating cross-talk between the TLR and IFN-␥ signaling pathways and therefore identify IRF-8 as a potential target of therapeutic intervention to control harmful immune responses.

MATERIALS AND METHODS
Reagents and Antibodies-Lipopolysaccharide (LPS) was from Sigma and IFN-␥ was obtained from R&D System (Minneapolis, MN). Antibodies against IRF-8, TRAF6, MyD88, ubiquitin, and ␤-actin were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Hemagglutinin (HA) monoclonal antibody 12CA5 was from Roche Applied Science; the FLAG monoclonal antibody and antibody against FLAG was from Sigma; antibody against AU1 was from Imgenex. Antibodies against ERK, JNK, and phosphorylated ERK and JNK were obtained from Cell Signaling. Protein G-Sepharose beads were from Amersham Biosciences. Recombinant mouse macrophage-colony stimulating factor and granulocyte macrophage-colony stimulating factor were from Peprotech.
Plasmid Constructs-IRF-8 full-length cDNA and IRF-8 C-terminal deletion mutants were inserted into the mammalian expression vector pcDNA3.1. The HA-ubiquitin plasmid was kindly provided by Dr. Ze'ev Ronai (Mount Sinai School of Medicine). The TRAF6 expression plasmid was from Dr. Adrian Ting (Mount Sinai School of Medicine).
Cell Lines and Peritoneal Macrophage-The RAW264.7 murine macrophage cell line and 293T cell line were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and penicillin-streptomycin. CL-2 cells (IRF-8Ϫ/Ϫ macrophages) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, antibiotics, recombinant mouse macrophagecolony stimulating factor (6 ng/ml), and recombinant mouse granulocyte macrophage-colony stimulating factor (6 ng/ml). Peritoneal macrophages were isolated from C3H/OUJ mice 3 days after intraperitoneal injection of 2 ml of thioglycollate medium. Cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and antibiotics.
Cytokine Release and Nitric Oxide Production-Cytokine secretion was quantified by enzyme-linked immunosorbent assay (R&D Systems) and nitric oxide (NO) production was measured using Greiss reagent (Sigma).
Transfection-293T cells were transiently transfected using DNAcalcium phosphate precipitate as described previously. For each transfection, 10 -20 g of plasmid was used.
Retroviral Transduction-For retroviral transduction we used a derivative of the Moloney murine leukemia virus vector Pmmp412. 293T cells seeded at a density of 4 ϫ 10 6 cells per 10-cm dish were cultured overnight and then transfected by calcium phosphate precipitate with 2.5 g of plasmid pMD.G, encoding vesicular stomatitis virus G protein, 7.5 g of pM-D.OGP, encoding gag-pol, and 10 g of retroviral expressing construct, encoding IRF-8, TRAF6, or HA-tagged ubiquitin. After 48 h the viral supernatant was collected, centrifuged at 800 ϫ g, and used to transduce cells. RAW264.7 cells (5 ϫ 10 6 ) were resuspended in 10 ml of viral supernatant in the presence of 5 g/ml Polybrene (Sigma), and cells were cultured for 48 h before assay.
IL-12 p40 Promoter Analysis-RAW264.7 cells were transiently transfected using SuperFect (Qiagen). For each transfection, 2.5 g of plasmid was mixed with 100 l of Dulbecco's modified Eagle's medium (without serum and antibiotics) and 10 l of SuperFect reagent, incubated at room temperature for 10 min, mixed with 600 l of Dulbecco's modified Eagle's complete medium, and immediately added to the cells in 6-well plates. Luciferase activity was measured 16 -24 h later. When indicated, LPS (1 g/ml) was added to the culture for 6 -12 h before harvest. The cells were extracted with reporter lysis buffer (Promega), and 20 l of extract was assayed for luciferase as described. Cells were co-transfected with a constitutively active cytomegalovirus promoter-␤-galactosidase reporter plasmid to normalize experiments for transfection efficiency.
Western Blotting-Cell lysates and prestained molecular weight markers were subjected to SDS-PAGE and transferred to a nitrocellulose membrane. The membranes were blocked with 5% nonfat milk in TBST (Triton-containing Tris-buffered saline), incubated with various antibodies (1:2000) for 1-2 h, washed with TBST, and stained with anti-rabbit, or anti-goat, or anti-mouse IgG conjugated to peroxidase (1:5000). Immunoreactivity was visualized by enhanced chemiluminescence (ECL kit, Santa Cruz Biotechnology).
Electrophoretic Mobility Shift Assay-RAW264.7 cells nuclear extracts were prepared as described previously (18,27). Electrophoretic mobility shift assay probes were prepared by annealing complementary single-stranded oligonucleotides with 5Ј-GATC overhangs (Genosys Biotechnologies, Inc.) and were labeled by filling in with [␣-32 P]dGTP and [␣-32 P]dCTP using Klenow enzyme. Labeled probes were purified with Nuctrap purification columns (Roche Molecular Biochemicals). Electrophoretic mobility shift assays were performed as described previously, using 10 5 cpm of probe and 5 g of nuclear extracts per reaction. DNA-binding complexes were separated by electrophoresis on a 5% polyacrylamide-Tris/glycine-EDTA gel, which was dried and exposed to x-ray film.
Co-immunoprecipitation-Cells were lysed in 0.5 ml of ice-cold buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 10% glycerol, 1 mM EGTA, 0.2 mM EDTA, 1% Nonidet P-40, 1 mM dithiothreitol) with protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 5 g/ml pepstatin, 10 g/ml chymostatin, and leupeptin). Lysates were clarified by centrifugation (14,000 rpm, 15 min at 4°C). Aliquots of lysate (500 g of protein) were incubated with 2 g of normal rabbit IgG or normal goat IgG for 4 h with intermittent agitation at 4°C, and the lysates were precleared by the addition of 15 l of protein G-Sepharose (Amersham Biosciences) for 2 h at 4°C. After centrifugation, the resulting supernatant was incubated with 2 g of anti-TRAF6 antibody or anti-IRF-8 antibody overnight at 4°C with gentle rocking. Immune complexes were absorbed on protein G-Sepharose and washed five times with lysis buffer. Finally immunoblotting was performed as described above.

IFN-␥ and TLR Ligands Synergistically Induce the Production of Proinflammatory Factors in
Macrophages-To analyze the cross-talk between IFN-␥ and TLR signaling, we incubated peritoneal macrophages for 6 h (mRNA) or 24 h (protein release) with IFN-␥ (10 ng/ml), LPS (1 g/ml), or IFN-␥ (10 ng/ml) and LPS (1 g/ml). The results showed that IFN-␥ and LPS synergistically induced the mRNAs for IL-1␤, TNF-␣, iNOS, IL-12 p40, and IL-6 ( Fig. 1A). To ascertain if the synergism was at the transcriptional level, we transfected RAW264.7 cells with an IL-12 p40 promoter luciferase reporter construct for 12-18 h, and activated with IFN-␥, LPS, and IFN-␥ and LPS for 12 h. IFN-␥ and LPS synergistically activated the IL-12 p40 promoter (data not shown). We next wanted to determine if IFN-␥ synergizes with PGN (TLR2) and poly(I:C) (TLR3). We transfected RAW264.7 cells with an IL-12 p40 promoter luciferase reporter plasmid for 12-18 h and activated cells with PGN (TLR2) and poly(I:C) (TLR3) in the presence or absence of IFN-␥. The addition of IFN-␥ with the TLR ligands synergistically induced IL-12 p40 promoter activation (Fig. 1B). These results indicate that IFN-␥ synergistically interacts with various TLR ligands to induce proinflammatory factors, suggesting that there exists cross-talk between the IFN-␥ and TLR signaling pathways.
MAP Kinase, but Not NF-B Activation, Is Important for the Crosstalk between IFN-␥ and TLR Signaling-Stimulation of the TLR signal pathway leads to the activation of NF-B and MAP kinases, resulting in the induction of proinflammatory factors. To determine the effects of IFN-␥ and LPS on NF-B and MAP kinase activation in macrophages, RAW264.7 cells were incubated for various intervals in the presence of IFN-␥, LPS, or IFN-␥ plus LPS, and cell lysates were analyzed for IB, ERK, and JNK phosphorylation by immunoblotting. Although IFN-␥ alone did not induce ERK and JNK phosphorylation, IFN-␥ and LPS synergistically induced ERK and JNK phosphorylation ( Fig. 2A). LPS induced IB phosphorylation, but there was no synergism between IFN-␥ and LPS in the induction of IB phosphorylation ( Fig. 2A). To confirm this result, we performed electrophoretic mobility shift assay experiments to analyze nuclear NF-B DNA-binding activity in RAW264.7 cells stimulated with LPS or IFN-␥ plus LPS (Fig. 2B). There was no synergism, nor was there synergism between IFN-␥ and LPS in the induction of NF-B reporter activity (Fig. 2C). These results suggest that MAP kinase but not NF-B activation is important for the crosstalk between IFN-␥ and TLR signaling pathways.
MAP Kinase Activation Is Impaired in IRF-8Ϫ/Ϫ Macrophages-We previously reported that IRF-8 is important for IL-12 and iNOS gene activation (18). We also found that IFN-␥ and LPS acted synergistically to induce IRF-8 protein expression (18). To extend these observations, we analyze the synthesis of IRF-8 in RAW264.7 macrophages incubated with IFN-␥ alone and in combination with TLR ligands PGN and poly(I: C). IFN-␥ acted synergistically with PGN and poly(I:C) to induce IRF-8 FIGURE 1. IFN-␥ and TLR ligands induce proinflammatory factors. A, expression of cytokine and iNOS mRNAs in peritoneal macrophages after IFN-␥ and LPS stimulation. Thioglycollate-elicited peritoneal macrophages from C57Bl/6 mice were activated with LPS (1 g/ml) or IFN-␥ (10 ng/ml) plus LPS (1 g/ml) for 6 h, and cytokine and iNOS mRNAs were analyzed by RT-PCR. B, the synergistic effect of IFN-␥ with PGN and poly(I:C) in the activation of IL-12 p40 promoter. RAW264.6 cells were transfected with an IL-12 p40 promoter luciferase reporter for 18 -24 h. The transfected cells were activated with IFN-␥, PGN, and poly(I:C) as indicated for 12 h. Extracts were analyzed for luciferase activity. Results are expressed as the luciferase activity (normalized for ␤-galactosidase activity).

FIGURE 2. Effect of IFN-␥ on LPS-induced MAP kinase and NF-B activation in macrophages.
A, synergistic effect of IFN-␥ and LPS on ERK and JNK activation. RAW264.7 cells were activated with IFN-␥ (10 ng/ml), LPS (1 g/ml), or IFN-␥ plus LPS for various time intervals (10,20,30, and 60 min). Phosphorylation of ERK, JNK, and IB was analyzed by immunoblotting with phosphorylation-specific and control antibodies. B, IFN-␥ has no effect on LPS-induced NF-B DNA binding activity. RAW264.7 cells were activated with LPS (1 g/ml) or IFN-␥ (10 ng/ml) plus LPS for 4 h, and nuclear protein was extracted for electrophoretic mobility shift assay. 5 g of nuclear protein was added to the 32 P-labeled NF-B consensus sequence. Extract and probe were incubated with 0.5 g of Poly(dI:dC) at room temperature for 30 min. C, effect of IFN-␥ and LPS on NF-B activation. RAW264.7 cells were transfected with NK-B reporter luciferase plasmid for 18 -24 h, and the transfected cells were activated with IFN-␥ (10 ng/ml), LPS (1 mg/ml), or IFN-␥ plus LPS for 12 h. Extracts were analyzed by luciferase activity. Results are expressed as the luciferase activity (normalized for ␤-galactosidase activity). APRIL 14, 2006 • VOLUME 281 • NUMBER 15 JOURNAL OF BIOLOGICAL CHEMISTRY 10075 (Fig. 3A). These results parallel the synergistic induction of proinflammatory factors by IFN-␥ and LPS. Therefore we hypothesize that IRF-8 is pivotal for the cross-talk between IFN-␥ and TLR signaling pathways. To address this question directly, we assayed cytokines released by thioglycollate-elicited peritoneal macrophages from wild-type and IRF-8Ϫ/Ϫ mice activated with IFN-␥ (10 ng/ml), LPS (1 g/ml), or IFN-␥ (10 ng/ml) plus LPS (1 g/ml) for 24 h. IL-12, IL-1, TNF-␣, and IL-6 titers were lower in all cases from IRF-8Ϫ/Ϫ macrophages (Fig. 3, B-E).

IRF-8 Interacts with TLR Signaling
We next analyzed if MAP kinase activation was impaired in IRF-8Ϫ/Ϫ macrophages. Thioglycollate-elicited peritoneal macrophages from IRF-8Ϫ/Ϫ and wild-type mice activated with LPS (1 g/ml) were assayed for phosphorylation of ERK and JNK by Western blot. Activation of ERK and JNK by LPS was strongly impaired in macrophages derived from IRF-8Ϫ/Ϫ mice and the IRF-8-deficient macrophage cell line CL-2 (Fig. 4, A and B). To confirm further the importance of IRF-8, we transduced CL-2 cells with IRF-8 or GFP for 2 days, activated the  A, ERK and JNK activation was impaired in IRF-8Ϫ/Ϫ macrophages. Thioglycollate-elicited peritoneal macrophages from both wild-type and IRF-8Ϫ/Ϫ mice were activated with LPS (1 g/ml) for 10, 20, and 60 min, and phosphorylation of ERK and JNK was analyzed by immunoblotting with phosphorylation-specific antibodies and control antibodies. B, ERK and JNK activation was impaired in CL-2 cells (IRF-8Ϫ/Ϫ macrophages). RAW264.7 cells and CL-2 cells were activated with LPS (1 g/ml) for 10, 20, 30, 60 min, and phosphorylation of ERK and JNK was detected as in A. C, retroviral transduction of IRF-8 into CL-2 cells rescued ERK and JNK activation. CL-2 cells were transduced with IRF-8 or GFP for 48 h, the cells were activated with LPS (1 g/ml) for 10 and 30 min, and phosphorylation of ERK and JNK was detected as in A. cells with LPS (1 g/ml) at intervals, and analyzed ERK and JNK phosphorylation. Transduction of IRF-8 into IRF-8Ϫ/Ϫ macrophages rescued LPS-induced ERK and JNK activation compared with the GFP control (Fig. 4C). Similar results were found using the TLR ligands, PGN and poly(I:C) (Fig. 5, A and B). The inhibition of both MAP kinase activation and production of proinflammatory factors in IRF-8Ϫ/Ϫ macrophages highlight the central role for IRF-8 in TLR signaling.
To further analyze whether transactivation of IRF-8 is important for MAP kinase activation, we transduced the IRF-8 DNA binding domain mutant (K79E) (16,17) or GFP into CL-2 cells for 2 days, activated the cells with LPS (1 g/ml) at intervals, and analyzed ERK phosphorylation as above. The results showed that transduction of IRF-8 DNA binding domain mutant still rescued ERK activation (Fig. 5C), suggesting that the transactivation of IRF-8 is not essential for MAP kinase activation.
IRF-8 Interacts Physically with TRAF6-To investigate how IRF-8 regulates TLR signaling, we performed co-immunoprecipitation experiments between IRF-8 and MyD88 or TRAF6, which revealed an interaction between IRF-8 and TRAF (Fig. 6A) but not MyD88 (data not shown). However, co-immunoprecipitation experiments showed that MyD88 interacted with IRF-7 (Fig. 6B). In addition, gel filtration of lysates from 293T cells co-transfected with IRF-8 and TRAF6 showed that the two proteins co-eluted in a high molecular weight complex (data not shown). To confirm the interaction of IRF-8 with TRAF6 without protein overexpression, we incubated thioglycollate-elicited peritoneal macrophages with IFN-␥ (10 ng/ml) and LPS (1 g/ml) for 4 h and then collected cell lysates for co-immunoprecipitation experiments. Significantly we found that the physical interaction between IRF-8 and TRAF6 existed in peritoneal macrophages only after activation with IFN-␥ and LPS (Fig. 6C). On a functional level, when IRF-8 and TRAF6 were co-transfected into RAW264.7 cells with an IL-12 p40 promoter luciferase reporter plasmid, we observed synergism in IL-12 p40 promoter activation (Fig. 6D).
IRF-8 Enhances TRAF6 Ubiquitination-TRAF6, an E3 ubiquitin ligase, plays a pivotal role in the activation of TLR signaling (1, 3) by catalysis of Lys-63-coupled polyubiquitination. Signal transduction by the IL-1 receptor, which shares many signaling components with TLR4, including TRAF6, is inhibited by a ubiquitin K63R mutant (28). Because we find that IRF-8 and TRAF6 interact, we next analyzed the effect of IRF-8 on TRAF6 ubiquitination. Cell lysates from 293T cells co-transfected with TRAF6, HA-ubiquitin, and IRF-8 for 48 h were immunoprecipitated with anti-TRAF6 and immunoblotted with anti-HA or antiubiquitin. Co-transfection of IRF-8 significantly enhanced TRAF6 ubiquitination visualized by staining HA-ubiquitin (Fig. 7A). Similarly, transfection of IRF-8 also enhanced TRAF6 ubiquitination visualized by anti-ubiquitin staining (Fig. 7B). To analyze the region in IRF-8 that is required for enhancement of TRAF6 ubiquitination, we co-transfected various IRF-8 C-terminal truncation mutants into 293T cells with TRAF6 and HA-Ub plasmids for 48 h. In agreement with the experiments to map IRF-8 TRAF6 interaction (Fig. 6E) IRF-8, IRF-8-(1-390), and IRF-8-(1-356) clearly enhanced TRAF6 ubiquitination, but IRF-8-(1-305) and IRF-8-(1-253) were without effect. Thus the enhancement of TRAF6 ubiquitination by IRF-8 is dependent on physical interaction of IRF-8 with TRAF6 (Fig. 7C). To confirm these experiments, we cotransduced IRF-8, TRAF6, and HA-Ub into RAW264.7 and CL-2 cells for 48 h and analyzed TRAF6 ubiquitination. In contrast to RAW264.7 cells, TRAF6 ubiquitination was significantly reduced in CL-2 cells. However, transduction of IRF-8 restored TRAF6 ubiquitination in both cell types to comparable levels (Fig. 7D). To investigate the relationship of the enhancement of TRAF-6 ubiquitination by IRF-8 with MAP kinase activation, we transduced CL-2 cells for 2 days with GFP or the IRF-8 mutant (1-390), which can enhance TRAF6 ubiquitination. Cells were activated with LPS (1 g/ml) at intervals, and ERK phosphorylation was analyzed as above. The results showed that transduction of FIGURE 5. TLR2 and TLR3-induced MAP kinase activation was impaired in IRF-8؊/؊ macrophages. A, PGN and poly(I:C)-induced ERK and JNK activation was impaired in IRF-8Ϫ/Ϫ macrophages. RAW264.7 cells and IRF-8Ϫ/Ϫ macrophages were activated with PGN (5 g/ml) or poly(I:C) (1 g/ml) for 10, 30, and 60 min, and phosphorylation of ERK and JNK was analyzed by immunoblotting with phosphorylation specific and control antibodies. B, retroviral transduction of IRF-8 into CL-2 cells rescued ERK and JNK activation. Cl-2 cells were transduced with IRF-8 or GFP for 48 h and the cells were activated with PGN (5 g/ml) or poly(I:C) (1 g/ml) for 10, 30, and 60 min, and phosphorylation of ERK and JNK was detected as in A. C, retroviral transduction of IRF-8 DNA binding domain mutant rescued ERK activation. CL-2 cells were transduced with IRF-8 DNA binding domain mutant or GFP for 48 h, and the cells were activated with LPS (1 g/ml) for 10 and 30 min, and phosphorylation of ERK was detected as in A.

DISCUSSION
In this study we investigate the role of IRF-8 in the cross-talk between IFN-␥ and TLR signaling. We have shown that the activation of TLR4 by LPS to induce the expression of proinflammatory factors, including IL-1, IL-6, IL-12, NO, and TNF-␣ is synergized by IFN-␥. Comparable results were observed using PGN (a TLR2 ligand) and poly(I:C) (a TLR3 ligand) and IFN-␥ to activate the IL-12 promoter. IFN-␥ augmented only LPS-induced ERK and JNK phosphorylation and had no effect on NF-B activation induced by LPS. Interestingly, in IRF-8Ϫ/Ϫ macrophages both the LPS-stimulated activation of ERK and JNK as well as expression of proinflammatory factors was inhibited. We found by coimmunoprecipitation experiments, that IRF-8 interacts physically with TRAF6 at a binding site between residues 305 and 356 of IRF-8. Furthermore, transfection of IRF-8 enhanced TRAF6 ubiquitination, which was dependent on the physical interaction between these two proteins. Taken together, the results suggest that the interaction of IRF-8 with TRAF6 modulates TLR signaling and may contribute to the cross-talk between IFN-␥ and TLR signaling. We initially hypothesized that IFN-␥ would enhance LPS-induced NF-B activation, because NF-B is an essential downstream target for TLR signaling (1,3). Unexpectedly, IFN-␥ had no effect on LPS-induced NF-B activation and had only a marginal effect on NF-B activation. The synergistic activation by IFN-␥ and LPS of ERK and JNK suggests that MAP kinases mediate IFN-␥ and TLR cross-talk.
The IFN-␥ signal transduction pathway is composed of primary and secondary phases. In the primary phase, after IFN-␥ binding to IFN-␥R, STAT1 mediates the primary signaling response. After ligand engagement, STAT1 is first phosphorylated by tyrosine kinases JAK1 and JAK2, then dimerizes, and subsequently translocates to the nucleus (29). In the nucleus, STAT1 binds to IFN-␥-activated sequence sequences and activates transcription of target genes including IRF-1 and IRF-8, which are important in both innate and adaptive immune responses (17,21). IRF-3, IRF-5, and IRF-7 have all been implicated in the TLR signal pathway. Several viruses activate the TLR3 signal pathway (30 -32), suggesting that TLR3 may play an important role in anti-viral host defense. IRF-5-deficient mice are severely impaired in TLR ligand stimulation of proinflammatory cytokine synthesis (24). Furthermore, IRF-5 interacts with and is activated by MyD88 and TRAF6. TLR activation results in the nuclear translocation of IRF-5 to activate cytokine gene expression. IRF-7 is also reported to play a role in TLR signaling. IRF-7 interacts with MyD88 and the TLR stimulated synthesis of IFN-␣ requires the formation of a complex consisting of MyD88, TRAF6, and IRF-7 (33).
IRF-8, an essential transcription factor in the induction of Th1 immune responses, is also involved in the TLR signaling pathway. CpG DNA-induced NF-B activation was impaired in IRF-8Ϫ/Ϫ dendritic cells, suggesting IRF-8 is involved in TLR9 signaling. However, the molecular mechanism was not elucidated (34). We report that macrophages from IRF-8Ϫ/Ϫ mice produced sharply decreased levels of IL-12, TNF-␣, IL-1, and IL-6 in response to LPS. Furthermore, we have found that LPS-induced ERK and JNK MAP kinase activation was severely impaired in peritoneal macrophages from IRF-8Ϫ/Ϫ mice. Unlike IRF-7, IRF-8 binds to TRAF6, but not to MyD88 (33). IRF-8 resides predominantly in the nucleus but, upon stimulation, moves to the cytoplasm. Because signal transmission occurs in a short time frame, it is reasonable to postulate that, upon TLR activation, IRF-8 interacts with TRAF6 in the cytosol. IRF-3 is produced constitutively in most cell types and is present in the cytoplasm in the absence of activation. Both IRF-5 and IRF-7 are expressed constitutively at very low levels mainly in lymphoid tissues, but their expression is enhanced by IFN-␣/␤ treatment (35)(36)(37). Unactivated macrophages express very low levels of IRF-8. However, IFN-␥ stimulation induces IRF-8 synthesis, which is further augmented by TLR ligands LPS, PGN, and poly(I:C).
The present study implicates TRAF6 as a target for IRF-8. TRAF6 is a ring domain-containing ubiquitin ligase, essential for the activation of NF-B and MAP kinases downstream of the TLR signaling pathway (3). TRAF6 forms a complex with an E2 ubiquitin-conjugating enzyme complex, consisting of Ubc13 and Uev1A, which catalyze the synthesis of lysine 63 polyubiquitin chains. Lysine 63 ubiquitination regulates protein-protein interaction rather than serving as a degradation signal. TRAF6, acting as an E3 ligase, activates signal transduction through association with downstream proteins, and TRAF6 itself is polyubiquitinated after oligomerization or dimerization. We find that IRF-8 interacts with TRAF6 and regulates TRAF6 ubiquitination. Because IRF-8 is not an E3 ligase, we suggest that IRF-8 either is part of the signaling complex, altering the activity of the UBC13-UEV1a complex, or is involved in recruitment of a different E3 ligase.
Our study provides an insight into the mechanism of the gene induction program downstream of TLR signaling. The observation that IRF-8, whose synthesis is stimulated by IFN-␥ and augmented by TLR ligands, is required for MAP kinase but not NF-B activation clarifies the complex regulation of the inflammatory responses. Thus, IRF-8 is not only involved in TLR signaling, but also contributes to the cross-talk between IFN-␥ and the TLR signaling pathways, resulting in the maximal production of proinflammatory cytokines.
The interaction of IRF-8 with the TLR signaling pathway is a secondary signal required to boost the immune response, especially during intracellular bacteria or viral infection. IRF-8, therefore, is a potential target for therapeutic intervention aimed at controlling harmful immune responses.