Factor-3 Pathways by Activating Transcription Negative Regulation of ...

The Journal of Immunology

Negative Regulation of TLR-Signaling Pathways by Activating Transcription Factor-31 Mark M. Whitmore,2*† Amaya Iparraguirre,‡ Lindsey Kubelka,* Wolfgang Weninger,‡ Tsonwin Hai,§ and Bryan R. G. Williams3¶ Activating transcription factor-3 (ATF3) is rapidly induced by LPS in mouse macrophages and regulates TLR4 responses. We show that ATF3 is rapidly induced by various TLRs in mouse macrophages and plasmacytoid dendritic cells (DCs), as well as plasmacytoid and myeloid subsets of human DCs. In primary macrophages from mice with a targeted deletion of the atf3 gene (ATF3-knockout (KO)), TLR-stimulated levels of IL-12 and IL-6 were elevated relative to responses in wild-type macrophages. Similarly, targeted deletion of atf3 correlated with enhanced responsiveness of myeloid DCs to TLR activation as measured by IL-12 secretion. Ectopic expression of ATF3 antagonized TLR-stimulated IL-12p40 activation in a reporter assay. In vivo, CpGoligodeoxynucleotide, a TLR9 agonist, given i.p. to ATF3-KO mice resulted in enhanced cytokine production from splenocytes. Furthermore, while ATF3-KO mice challenged with a sublethal dose of PR8 influenza virus were delayed in body weight recovery in comparison to wild type, the ATF3-KO mice showed higher titers of serum neutralizing Ab against PR8 5 mo postinfection. Thus, ATF3 behaves as a negative regulatory transcription factor in TLR pathways and, accordingly, deficiency in atf3 alters responses to immunological challenges in vivo. ATF3 dysregulation merits further exploration in diseases such as type I diabetes and cancer, where altered innate immunity has been implicated in their pathogenesis. The Journal of Immunology, 2007, 179: 3622–3630.

T

he innate immune system of mammals controls the rapid spread of pathogens through antimicrobial activities and initiates the development of acquired immunity. Pattern recognition receptors of innate immune cells identify infection through binding distinctive macromolecular components of pathogens termed pathogen-associated molecular patterns. TLRs are a class of pattern recognition receptors expressed on the cell membrane or in the endosomal compartment of APCs such as macrophages and dendritic cells (DCs).4 Activation of TLRs triggers proinflammatory cytokine production, promotes innate effector activity, and stimulates Ag-presentation function (1). These functions work as a regulated network, ultimately eradicating infection by controlling the spread of pathogens in the early phases of in*Department of Cancer Biology, Lerner Research Institute, Cleveland Clinic, Cleveland, OH 44195; †Taussig Cancer Center, Cleveland Clinic, Cleveland, OH 44195; ‡ Immunology Program, The Wistar Institute, University of Pennsylvania, Philadelphia, PA 19104; §Department of Molecular and Cellular Biochemistry and Center for Molecular Neurobiology, Ohio State University, Columbus, OH 43210; and ¶Monash Institute of Medical Research, Monash University, Melbourne, Australia Received for publication January 31, 2007. Accepted for publication July 10, 2007. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by National Institutes of Health Grants (NIH) RO1 AI34039 and PO1 CA62220 (to B.R.G.W.) and NIH/National Cancer Institute T32 CA059366 (to M.M.W.). 2 Current address: Immunotherapy Research Group, Genzyme Corporation, 5 Mountain Road, Framingham, MA 01701. 3 Address correspondence and reprint requests to Prof. Bryan R. G. Williams, Monash Institute of Medical Research, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia. E-mail address: [email protected] 4 Abbreviations used in this paper: DC, dendritic cell; pIC, poly-IC; ATF3, activating transcription factor-3; IRAK, IL-1R-associated kinase; KO, knockout; ODN, oligodeoxynucleotide; HEK, human embryonic kidney; BMM, bone marrow-derived macrophage; BMDC, bone marrow-derived DC; mDC, myeloid DC; pDC, plasmacytoid DC; IRF, IFN regulatory factor; HA, hemagglutinin.; TRIF, toll/IL-1 receptor domain-containing adapter including IFN-␤.

Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 www.jimmunol.org

fection and supporting the development of acquired immunity (2). Moreover, TLR ligands can be effective antitumor agents as a result of their immunostimulatory capacity (3, 4). Indeed, a combination of TLR ligands, i.e., poly-IC (pIC) for TLR3 and CpG for TLR9, can stimulate enhanced antitumor activity against metastatic melanoma (5). Dysregulation of TLR signaling can have serious consequences, such as the cytokine storm associated with septic shock (6, 7), lethality from ineffective defenses against pathogens (8), or misdirected attacks against host cells resulting in autoimmunity (9 – 11). A number of proteins target components of TLR-signaling pathways as negative regulators, presumably to prevent the deleterious effects of an overactive or prolonged innate immune response (12). These include soluble forms of TLR2 and 4 (13, 14), a short form of MyD88 (15), IL-1R-associated kinase M (IRAK-M) (16, 17), A20 (18, 19), Tollip (20, 21), and SIGIRR (22, 23). Although all of these proteins can affect transcription of genes targeted by TLR pathways, none directly modify gene activity at the transcriptional level. Activating transcription factor-3 (ATF3) is a member of the ATF/CREB family of bZip transcription factors that bind to the consensus CRE (24). In contrast to the implication of its name, ATF3 as a homodimer functions as a transcriptional repressor, not an activator (25). This property sets ATF3 apart from many of the other ATF/CREB proteins, which function as transcriptional activators. Overwhelming evidence indicates that the expression of the atf3 gene is induced by a variety of stress signals (26). In macrophages, ATF3 has been demonstrated to be induced by the TLR4 ligand LPS and bacillus Calmette Gue´rin (27, 28) and by IFN (thus referred to as an IFN-stimulated gene) in human PBMC (29). Despite these observations, the role of ATF3 in innate immune responses has only recently been described. Gilchrist et al. (30) demonstrated that ATF3 is induced by TLR4 activation and is part of the negative feedback loop to modulate the TLR4-stimulated inflammatory response of macrophages. In this report,

The Journal of Immunology we demonstrate that ATF3 plays a much broader role in the innate immune response than previously described. First, it serves as a negative regulator for not only the TLR4-stimulated inflammatory response, but also that stimulated by TLR2/6 heterodimer, TLR3, TLR5, TLR7, and TLR9. Second, ATF3 regulates TLR-stimulated responses in DCs. Using transcript profiling and Western blot analyses, we show that ATF3 is induced following stimulation of primary mouse APCs and human DCs upon the activation of various TLRs. TLRstimulated levels of TNF-␣, IL-12, and IL-6 were elevated in primary macrophages from atf3 knockout (ATF3-KO) mice compared with wild-type mice, indicating that ATF3 functions as a negative regulator of their expression. TLR-stimulated levels of IL-12 were also increased in myeloid DCs from ATF3-KO mice relative to wild type. Transient transfection coupled with reporter assay indicated that ATF3 represses TLR-stimulated activation of the IL-12p40 promoter, which contains a consensus CRE. Finally, we investigated the effect of a targeted deletion of atf3 on cytokine production following TLR9 stimulation in vivo. We also found that atf3 accelerates recovery from influenza virus infection.

Materials and Methods Reagents and cell lines Mouse and human class B CpG oligodeoxynucleotides (published sequences ODN-1826 and ODN-2006, respectively) were purchased as reverse phase HPLC-purified with phosphorothioate linkages from the Midland Certified Reagent. pIC and LPS were purchased from Sigma-Aldrich. Zymosan was purchased from Invivogen. FliC was provided by Dr. J. DiDonato (Lerner Research Institute, Cleveland Clinic, Cleveland, OH) and purified as previously described (31). Endotoxin was removed from preparations using a polymyxin B resin and FliC activity was assessed using an NF-␬B band-shift assay on A549 cells. FliC did not have activity in RAW 264.7 cells that are negative for TLR5 (the FLiC receptor) and positive for TLR4 (the LPS receptor), and was thus confirmed to be free of LPS. CpG-ODN and zymosan were tested to be free of endotoxin by the chromogenic Limulus amebocyte lysate assay from BioWhittaker. dsRNA, similar to LPS, stimulates positive reactivity in Limulus coagulation assays; therefore, pIC was determined to be free of LPS contamination as RNase digestion completely abolished the ability of pIC to stimulate NO production from RAW 264.7 cells. RAW 264.7 cells were purchased from American Type Culture Collection. Human embryonic kidney (HEK)-293 cells and those stably transfected with human TLR3 and 9 (293-TLR3 and -TLR9), as developed in the laboratory of Dr. D. Golenbock (University of Massachusetts School of Medicine, Worcester, MA), were a gift from the laboratory of Dr. K. Fitzgerald (University of Massachusetts School of Medicine, Worcester, MA). Growth medium for all cell lines was DMEM supplemented with 10% FBS and penicillin/streptomycin antibiotics. All cells were maintained in a 37°C humidified incubator in an atmosphere of 5% CO2.

Mice and primary APCs Wild-type (C57BL/6J) mice were purchased from The Jackson Laboratory. Mice with a targeted deletion of atf3 on a background of C57BL/6 (ATF3KO) were derived as described (32). Mice were housed at the Biological Resources Unit of the Lerner Research Institute in compliance with federal and institutional care and use committee protocols and procedures. Bone marrow-derived macrophages (BMMs) were prepared as previously described (5), by differentiation of bone marrow flushed from the femur of the indicated mice and differentiated for 7–9 days in 15–20% L cell-conditioned medium. BMMs differentiated using 100 ng/ml rM-CSF (BD Biosciences) behaved identically to those differentiated with L cell-conditioned medium. For elicited peritoneal macrophages, cells were collected 3– 4 days following thioglycolate injection by peritoneal lavage and removed of RBC by ammonium chloride lysis. Cells were then plated for 3– 4 h in DMEM supplemented with 10% FBS and penicillin/streptomycin to allow for attachment. After three washes with PBS, adherent cells were stimulated as indicated in growth medium. Mouse bone marrow-derived DCs (BMDCs) were prepared by culture of mouse bone marrow cells in culture medium supplemented with rFlt3 ligand (100 –200 ng/ml; BD Biosciences) for 8 days. Culture medium consisted of RPMI 1640, 10% FBS, penicillin-streptomycin antibiotics, nonessential amino acids, and 2 mM 2-ME. Type I and type II DCs were

3623 purified using CD11b⫹ and B220⫹ magnetic microbead-positive selection (Miltenyi Biotec), respectively. Cytospin preparations of purified DCs stained with H&E indicated morphology consistent with the respective cell types in both wild-type and ATF3-KO mice. Cells from wild-type (C57BL/6) and ATF3-KO mice were plated at equal density (0.1 ⫻ 106 cells/well on 24-well plates). Further culture of purified type I DCs was done in the presence of rGM-CSF (10 ng/ml) and that of type II DCs was done in the presence of IL-3 (10 ng/ml). Type I BMDCs were stimulated with the indicated concentration of TLR ligands and release of IL-12 into culture supernatants was measured by ELISA.

Isolation and culture of human myeloid and plasmacytoid DCs from PBMC Human blood was collected in EDTA and PBMC were purified by Ficoll gradient. PBMC were depleted of B cells and monocytes by anti-CD19 and anti-CD14 magnetic bead separation (Miltenyi Biotec). Type I myeloid DCs (mDCs) were purified from flow-through by anti-CD1c magnetic bead-positive selection and plasmacytoid DCs (pDCs) were subsequently isolated from anti-CD1c flow-through by anti-BDCA-4 magnetic beadpositive selection (Miltenyi Biotec). mDCs and pDCs were plated at 0.8 – 1 ⫻ 106 cells/well on a 24-well plate in RPMI 1640 supplemented with 10% FBS and penicillin/streptomycin antibiotics. For overnight incubation, mDCs were supplemented with 10 ng/ml GM-CSF and pDCs were supplemented with IL-3 (100 ng/ml). After overnight culture, mDCs were stimulated for 4 h with either medium, LPS (100 ng/ml), or imiquimod (R387; 10 ␮g/ml). pDCs were stimulated for 4 h with either medium or imiquimod (10 ␮g/ml).

Isolation of splenocytes and cytometric bead assay C57BL/6 and ATF3-KO mice were i.p. injected with PBS or CpG-ODN 1826 (20 ␮g/mouse). Twenty-four hours later, single-cell suspensions of splenocytes were collected into RPMI 1640 by pressing excised spleens through a 70-␮M nylon mesh. RBC were removed by ammonium chloride lysis. Crude splenocytes were either stained for flow cytometry analysis as described below or cultured for 24 h at 1 ⫻ 106/ml/well in 24-well plates using RPMI 1640 supplemented with 10% FBS and penicillin/streptomycin antibiotics as culture medium to generate conditioned medium to be measured for cytokine as described below.

Mouse 15K cDNA microarray Bone marrow-derived macrophages isolated from wild-type (129SV) and type I IFN-␣/IFN-␤ receptor KO mice (isogenic on 129SV background) were stimulated with either medium, pIC (10 ␮g/ml), CpG-ODN 1826 (1 ␮g/ml), or a combination of pIC and CpG (10 and 1 ␮g/ml, respectively) for 6 h. Total RNA was extracted using TRIzol (Invitrogen Life Technologies) following the manufacturer’s instructions and provided to the Lerner Research Institute Genomics Core facility for analysis using the mouse 15K cDNA microarray from the National Institute of Aging. cDNA was prepared from RNA of medium- and TLR ligand-treated cells and differentially labeled with Cy3 and Cy5 fluorophores. Labeled cDNA from medium-treated cells served as controls and cDNA from TLR ligand-treated cells served as experimentals that were competitively hybridized to a mouse 15K cDNA glass slide array. Spot fluorescence was measured using a GenePix dual laser fluorescence scanner. GeneSpring software was used to produce a list of genes up-regulated at least 2-fold by TLR ligand(s). GeneSpring was then used to generate a Venn diagram of genes up-regulated 2-fold vs control for each TLR ligand treatment within the appropriate genotype.

Western blot Mouse and human APCs (i.e., BMM, mDC, or pDC), as well as HEK-293 cells, 293-TLR3, or 293-TLR9 cells, were treated with medium or TLR ligand(s) as indicated. After treatment, cells were washed twice in ice-cold PBS, and whole cell extracts were collected in Triton-X lysis buffer (50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 50 mM NaF, 10 mM ␤-glycerophosphate, 0.1 mM EDTA, 10% glycerol, 1% Triton X-100, 1 mM DTT, 2 mM inorganic sodium pyrophosphate, and 2 mM sodium orthovanadate) supplemented with a protease inhibitor mixture of PMSF (1 mM), leupeptin (2 ␮g/ml), aprotinin (2 ␮g/ml), and pepstatin A (2 ␮g/ml). Proteins from whole cell extracts were separated by 15% SDS-PAGE and transferred onto Imobinon-P (Millipore) membranes for immunoblotting. ATF3 was detected using rabbit anti-ATF3 (C-19) polyclonal Ab from Santa Cruz Biotechnology. Actin was detected using mouse anti-␤-actin mAb obtained from Sigma-Aldrich. Phosphorylated Stat1 was detected using anti-phospho-Stat1 (Tyr701) obtained from Cell Signaling Technology.

3624

ATF3 ACTS AS A NEGATIVE REGULATOR OF TLR PATHWAYS

FIGURE 1. Up-regulation of ATF3 transcript and protein levels by TLR3 and TLR9 ligands in primary mouse macrophages. A, Venn diagram of transcripts up-regulated in the 15K mouse cDNA array (NIAD) ⬎2-fold up vs medium-treated controls in BMMs from wild-type (129SV) and type I IFNR-KO mice. Cells were stimulated with medium, pIC (10 ␮g/ml), CpG (1 ␮g/ml), or pIC and CpG (10 and 1 ␮g/ml) for 6 h. cDNA, prepared from total RNA, from medium controls and ligand-treated cells were differentially labeled with fluorophore and hybridized to the 15K mouse cDNA microarray as described in Materials and Methods. B, Fold increase of top 10 genes significantly up-regulated ⬎2-fold (p ⬍ 0.05) by pIC/CpG in wild-type BMMs that were not significantly increased above 2-fold (p ⬎ 0.05) by pIC or CpG alone in A. 䡺, CpG stimulation in wild-type BMMs; u, pIC stimulation in wild type; f, pIC/CpG stimulation in wild type; z, pIC/CpG stimulation in type I IFNR-KO. Gene symbols are official symbols listed in the National Center for Biotechnology Information Entrez Gene database for the Mus musculus genome. C, Western blot for ATF3 in elicited peritoneal macrophages stimulated as indicated using ligand concentrations similar to A. D, Western blot analysis for ATF3 in BMM from C57BL/6 mice treated as in A and B; med, medium. E, Kinetics of ATF3 induction and STAT1 phosphorylation as measured by Western blot in RAW 264.7 cells stimulated with LPS (100 ng/ml), pIC (10 ␮g/ml), CpG (1 ␮g/ml), or zymosan (Zym; 10 ␮g/ml).

Analysis of cytokines

Results

Conditioned medium from primary macrophages were collected 6 and 20 h after the indicated treatments (LPS, pIC, CpG, pIC/CpG, or zymosan). Culture supernatants were cleared of contaminating debris by brief centrifugation at 400 ⫻ g for 5 min at 4°C. Culture supernatants were measured by ELISA using Opt-EIA ELISA sets obtained from BD Biosciences/BD Pharmingen following the manufacturer’s instructions. Conditioned medium from crude splenocytes was measured for TNF-␣, IFN-␥, IL-2, IL-4, and IL-5 by cytometric bead assay using the mouse Th1/Th2 cytokine cytometric bead array (BD Biosciences) following the manufacturer’s instructions.

ATF3 regulates TLR-mediated activation of murine macrophages

IL-12p40 promoter-reporter assays HEK-293, 293-TLR3, or 293-TLR9 cells were cotransfected using Lipofectamine Plus (Invitrogen Life Technologies) with 25 ng of mouse IL12p40-firefly luciferase promoter-reporter plasmid (a gift of K. Ozato, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD), 25 ng of pRL-null vector (Promega), 25 ng of a plasmid encoding IFN regulatory factor 8 (IRF8) under the control of the CMV promoter (pIRF8; a gift of K. Ozato), and either 50 ng of pCGN-ATF3 (encoding hemagglutinin (HA)-tagged mouse ATF3 under the control of CMV promoter) or 50 ng of pcDNA-HA (empty vector for pCGN-ATF3). Twenty-four hours after transfection, cells were treated for 24 h with medium, LPS (100 ng/ml), pIC, or CpG-ODN 2006 as indicated. Firefly and Renilla luciferase activity was assessed using the Dual Luciferase Reporter Assay System (Promega) following the manufacturer’s instructions. Firefly IL-12p40 luciferase levels were normalized to Renilla luciferase and fold-induction by TLR ligand was calculated relative to similarly transfected medium controls.

Influenza infection of wild-type and ATF3-KO mice Mice (n ⫽ 5/group) were anesthetized and inoculated intranasally with 500 50% tissue-culture infective doses of influenza virus strain A/PR/8/34 (in 50 ␮l of PBS). Morbidity was quantified daily by measure of body weight. To determine PR8 virus neutralizing Abs, sera were collected ⬎5 mo postinfection, heat-inactivated, and tested by hemagglutination inhibition assay as previously described (33). For analysis of memory PR8-virusspecific T cells, splenocytes were lysed with ACK lysis buffer and stained with anti-CD8⫹ (BD Bioscience) and DbNP366 –374 tetramer (gift from Dr. E. J. Wherry, The Wistar Institute, Philadelphia, PA), which is recognized by endogenous CD8⫹ T cells specific for the epitope of the influenza virus nucleoprotein.

We previously showed that the TLR3 ligand, pIC, and the TLR9 ligand, CpG-ODN 1826 (CpG), displayed synergism for stimulating macrophage production of proinflammatory cytokines (TNF-␣, IL-12, and IL-6) and NO (5). To measure a broader spectrum of genes affected by the pIC/CpG combination, transcript profiles from BMMs stimulated with pIC and CpG either alone or in combination (pIC/CpG), in conditions known to elicit a synergistic cytokine response, were measured using the 15 K mouse cDNA microarray. Setting a threshold of ⱖ2-fold increase, pIC/CpG stimulated up-regulation of 116 genes that were not increased by pIC or CpG alone (Fig. 1A), suggesting that the combination induced a distinct set of transcripts. Because type I IFNs can play a role in controlling gene expression following TLR ligation, a parallel experiment using BMMs from knockout mice deficient in the type I IFN-␣/IFN-␤ receptor (IFNR-KO) indicated that up-regulation of a subset of transcripts by pIC alone or pIC/CpG were IFN dependent, because increases were not observed in the KO cells (Fig. 1, A and B). Not unexpectedly, a number of transcripts were up-regulated by the TLR ligands in the absence of the type I IFN␣/IFN-␤ receptor (Fig. 1, A and B). With a threshold set for at least 3-fold increase by pIC/CpG and not significantly up-regulated ⬎2-fold by pIC or CpG alone, the top 10 expressed sequence tags linked to experimentally defined genes (Fig. 1B) revealed genes with roles undefined in innate immunity (i.e., Cspg2, Slco3a1, Edg2, Sp100, Zfp36, and Lcn2), as well as genes with defined functions (i.e., Atf3, Jak2, and Ap3b1). For the top eight genes, the response to pIC/CpG was at least partially dependent on type I IFN-␣/IFN-␤ receptor expression, while the next two (Lcn2 and Ap3b1) increased independent of the receptor. This further supports previous findings that some genes stimulated by pIC/CpG are IFN dependent and others are IFN independent.


3625

FIGURE 2. Enhanced IL-12p40 and IL-6 production in ATF3-KO primary macrophages stimulated in culture with various TLR ligands. A, TLR ligand-stimulated release of IL-12p40 from BMMs of C57BL/6 (䡺) and ATF3-KO (f) mice. BMMs were stimulated with medium, LPS (0.1 ␮g/ml), pIC (10 ␮g/ml), CpG-ODN 1826 (1 ␮g/ml), zymosan (10 ␮g/ml) or pIC ⫹ CpG (10 ␮g/ml ⫹ 1 ␮g/ml) for 20 –24 h. Culture supernatants were collected and assessed for IL-12p40 levels by ELISA. B, IL-6 levels in culture supernatants of BMMs in A as measured by ELISA. C, TLR ligand-stimulated release of IL-12p40 from elicited peritoneal macrophages of C57BL/6 (䡺) and ATF3-KO (f) mice. Peritoneal macrophages were stimulated with medium, LPS (0.1 ␮g/ml), CpG-ODN 1826 (0.3 and 1 ␮g/ml), pIC (10 and 30 ␮g/ml), and pIC ⫹ CpG (10 ␮g/ml ⫹ 1 ␮g/ml) for 20 –24 h. Culture supernatants were collected and assessed for IL-12p40 levels by ELISA. D, TLR ligand-stimulated release of TNF-␣ from elicited peritoneal macrophages of C57BL/6 and ATF3-KO mice. Culture supernatants collected in B were measured for TNF-␣ levels by ELISA. Bars represent mean ⫾ SEM for triplicate wells. nd, not detected. Data are representative of at least two independent experiments.

Of the top 10 genes, we sought to further investigate the biological significance of induction of ATF3, because a transcription factor would be an excellent candidate as a “master regulator” modulating the transcriptional programs upon TLR activation by pIC/CpG. We first confirmed that ATF3 protein levels increased in elicited peritoneal macrophages and BMMs following TLR ligand stimulation (Fig. 1, C and D). Results from peritoneal macrophages (Fig. 1C) were consistent with the array findings, indicating levels of ATF3 were strongly enhanced by pIC/CpG at 6 h, while levels stimulated by pIC or CpG alone at 6 h were minimal. Interestingly, CpG alone stimulated an increase in ATF3 levels at 3 h that returned to baseline by 6 h. In BMMs, pIC/CpG increased ATF3 protein levels, but the increase was not significantly higher than that by pIC (Fig. 1D). Once again, albeit with different kinetics, ATF3 levels increased with CpG. The varied kinetics for CpG-stimulated ATF3 induction between peritoneal macrophages (Fig. 1, B and C) and BMMs (Fig. 1D) could have resulted from strain differences, because the peritoneal macrophages were from 129SV mice and the BMMs were from C57BL/6 mice. Alternatively, peritoneal macrophages and BMMs represent different subclasses of macrophages that have divergent responses to TLR9 stimulation (34 –36). Although the qualitative and quantitative differences between ATF3 induction in peritoneal macrophages and BMMs did not clarify the role of ATF3 in mediating pIC/CpG synergism, the data clearly indicated multiple TLR ligands stimulated accumulation of ATF3 in two different sources of murine macrophages. As shown in Fig. 1A, ATF3 was not induced 6 h following TLR ligation in type I IFNR-KO BMMs, suggesting the type I IFN-␣/ IFN-␤ receptor is necessary for the induction of ATF3 by TLR pathways. In contrast, CpG, a poor inducer of type I IFN in primary mouse macrophages, also stimulated accumulation of ATF3 protein (Fig. 1, C and D). Moreover, LPS is reported to induce ATF3 within 1 h, i.e., a time point before significant production of type I IFN. Thus, to clarify whether ATF3 induction is directly

downstream of TLR activation or a result of secondary IFN-␣/ IFN-␤ production, the kinetics of ATF3 accumulation was correlated with signaling events associated with the type I IFN-␣/IFN-␤ receptor activation, i.e., STAT phosphorylation, following stimulation of RAW 264.7 cells, a murine monocyte/macrophage cell line, with various TLR ligands (Fig. 1E). Ligands for TLR4, TLR3, TLR9, TLR3/9, and TLR2/6 (LPS, pIC, CpG, CpG/pIC, and zymosan, respectively) stimulated a rapid (within 1 h) increase in ATF3 levels before STAT1 phosphorylation, suggesting early induction occurred before type I IFN-␣/IFN-␤ receptor activation. Ligands for TLR3 and TLR4, pIC and LPS, which engage the TLR adapter toll/IL-1 receptor domain-containing adapter inducing IFN-␤ (TRIF) leading to strong type I IFN production, stimulated a biphasic and strong induction of ATF3 with the later phase corresponding with STAT1 phosphorylation. In contrast, TLR ligands for TLR2/6 and TLR9 that do not engage TRIF stimulating miniscule IFN accumulation, zymosan and CpG, stimulated a weaker monophasic induction over the time course with minimal STAT1 phosphorylation. Thus, early induction of ATF3 occurs independent of type I IFN-␣/IFN-␤, while the later component of biphasic induction was likely regulated by IFN. The induction of ATF3 by various TLR pathways suggested that it plays an important role in the inflammatory response. Therefore, we compared primary macrophages isolated from wild-type and ATF3-KO mice for their cytokine production upon stimulation with different TLR ligands. As shown in Fig. 2A, IL-12p40 secretion from BMMs derived from KO mice was higher than that from wild-type, ranging from 2- to 5-fold, depending on the TLR ligand. Similarly, IL-6 production was enhanced in BMMs from ATF3-KO mice (Fig. 2B). To rule out artifacts that may arise from differentiation of bone marrow precursors to macrophages in culture, cytokine levels were measured in peritoneal macrophages directly isolated from wild-type and KO mice. Consistent with BMMs, enhancement of IL-12 in ATF3-KO peritoneal macrophages, in comparison to wild-type, was observed (Fig. 2C). No

3626


FIGURE 3. TLR-stimulated induction of ATF3 in mouse and human DCs. A, atf3 levels in murine pDCs 4 h following treatment with CpG or infection with the PR8 strain of influenza measured by transcript profiling; med, medium. B, ATF3 protein levels in mouse pDCs stimulated as in A for two independent experiments, i.e., pDC preparations. C, Western blot for ATF3 levels in human pDCs or mDCs treated for 2 h with medium or the indicated TLR ligand. TLR ligands were imiquimod (R387; 10 ␮g/ml), CpG-ODN 2006 (CpG; 5 ␮g/ml), LPS (100 ng/ml), zymosan (Zym; 10 ␮g/ml), or FliC (1 ␮g/ml). D, TLR-stimulated IL-12 levels (left panel) or IL-10 levels (right panel) in mDCs isolated from C57BL/6 (䡺) and ATF3-KO (f) mice. Myeloid BMDCs were stimulated for 24 h with medium, zymosan (10 ␮g/ml), FliC (1 ␮g/ml), LPS (30 and 100 ng/ml), or pIC (10 and 30 ␮g/ml). Bars represent mean ⫾ SEM (n ⫽ 3). ⴱ, Value of p ⱕ 0.01 vs wild type. Numbers below blots in B and C represent ATF3 band intensity normalized to ␤-actin. nd, not detected.

enhancement was observed for TNF-␣ under most conditions, except for CpG treatment alone where the difference between wildtype and KO cells was significant (Fig. 2D). Taken together, these data indicate that ATF3 has a negative regulatory effect on IL12p40 and IL-6 production as well as TNF-␣ release under certain conditions. TLRs induce ATF3 in mouse and human DCs The most potent APCs of the mammalian innate immune system are DCs with two major subtypes, i.e., mDCs and high IFN-producing pDCs. Each subtype expresses a distinct pattern of TLRs that are vital for pathogen recognition and triggering the immune cascade. Thus, it was of interest to determine whether ATF3 was induced in these cells following TLR ligation or infection. Interestingly, transcript profiles in mouse pDCs treated with CpG-ODN or infected with the PR8 strain of influenza virus, which is recognized by TLR7 (37–39), indicated strong induction of ATF3 at the transcript level (Fig. 3A) that was confirmed at the protein level (Fig. 3B). TLR-stimulated levels of ATF3 were also examined in human mDCs and pDCs. Both subtypes responded to TLR ligand stimulation by increased levels of ATF3 protein (Fig. 3C). To determine whether targeted deletion of atf3 affected IL-12 production in mouse DCs, myeloid BMDCs from wild type (C57BL/6) and ATF3-KO were stimulated with zymosan, FliC, LPS, or poly-IC and measured for IL-12 secretion by ELISA (Fig. 3D). Similar to macrophages (Fig. 2), targeted deletion of atf3 resulted in enhanced levels of IL-12 released from mDCs treated with TLR ligands (Fig. 3D). Release of IL-10, which is capable of inhibiting IL-12 production, was not deficient in the ATF3-KO mDCs (Fig. 3D). Indeed, IL-10 levels from ATF3-KOs were either enhanced (TLRs 2/6, TLR3, and TLR4) or equivalent (TLR5) to those of wild type. Taken together, subsets of innate immune APCs in mice and humans rapidly up-regulated ATF3 after pathogen challenge or activation of TLR2/6, TLR3, TLR4, TLR5, TLR7, and TLR9. Furthermore, targeted deletion of atf3 resulted in enhanced cytokine release from macrophages and DCs stimulated via TLRs.

ATF3 regulation of the IL-12p40 promoter The mouse IL-12p40 promoter contains a CRE-like site upstream of a functional IFN-␥-activated sequence (GAS) and downstream of functional NF-␬B and IFN-stimulated response element (ISRE) consensus sites (Fig. 4A). Because ATF3 is a transcriptional repressor and the CRE-like site is a potential binding site for ATF3, it is possible that ATF3 represses the IL-12p40 promoter, providing an explanation for the negative effect of ATF3 on IL-12p40 production described above. To test this possibility, we transfected HEK-293 cells stably expressing TLR3 (293-TLR3) with a luciferase reporter construct driven by the IL-12p40 promoter and an expression construct for ATF3. Because cotransfection of IRF8 enhanced the responsiveness of the reporter to pIC (data not shown), a plasmid encoding IRF8 was included in all transfections. Accordingly, 293 and 293-TLR3 cells were cotransfected with a control plasmid expressing only HA (pcDNA-HA) or an expression plasmid that expressed ATF3 with an N-terminal HA tag (pCGN-ATF3). Twenty-four hours after transfection, the cells were treated with pIC or LPS and relative luciferase levels were measured 24 h after treatment. As shown in Fig. 4B, pIC activated the IL-12p40 promoter in 293-TLR3. This activation was significantly reduced in parental 293 cells and was not observed with LPS, indicating specificity for the pIC-TLR3 interaction. Importantly, ectopic expression of ATF3 reduced the pIC-stimulated reporter activity (87 and 90% inhibition). To determine whether ATF3 antagonizes the TLR9-stimulated IL-12p40 promoter activity, we repeated the experiments using 293 cells stably expressing TLR9 (293-TLR9) and used CpG to activate the TLR9 pathway. As shown in Fig. 4C, ectopic expression of ATF3 inhibited (by 67%) a small, but statistically significant, increase in promoter activity stimulated by CpG through TLR9 ( p ⬍ 0.05). We note that the IL-12 promoter activity in this reporter assay was relatively weak, even in the absence of the ectopically expressed ATF3. We suspected that this was due to the expression of the endogenous ATF3. As shown in Figs. 1 and 3, ATF3 expression is


3627

FIGURE 4. Negative regulation of the IL-12p40 promoter by ATF3. A, Location of a near consensus CRE in the mouse IL-12p40 promoter. Sequence shown is ⫺510 to ⫹12 relative to the transcription start site (arrow). Bases in bold represent the indicated transcriptional regulatory element. B, Activation of the IL-12p40 promoter by pIC and LPS in 293-TLR3 cells (left panel) and parental 293s (right panel) as measured by an IL-12p40 promoter-reporter dual luciferase assay. HEK-293 cells or 293-TLR3 cells were cotransfected with plasmids for an IL-12p40 promoter-reporter dual luciferase assay (25 ng of pIL-12p40-firefly luciferase ⫹ 25 ng of pRenilla-null), a plasmid-encoding IRF8 (50 ng pIRF-8), and either an empty vector for the C-terminal HA tag (50 ng of pcDNA-HA) or HA-tagged ATF3 (50 ng of pCGN-ATF3). Twenty-four hours following cotransfection, cells were treated with either medium, pIC (30 and 60 ␮g/ml) or LPS (100 ng/ml), for 20 h, and cell lysates were measured for luciferase activity by the dual luciferase assay. C, Activation of the IL-12p40 promoter by CpG-ODN (5 ␮g/ml) and LPS (100 ng/ml) in 293-TLR9 cells as measured by an IL-12p40 promoter-reporter dual luciferase assay. HEK-293-TLR9 cells were cotransfected as described in B. Twenty-four hours following cotransfection, cells were treated with either medium, CpG-ODN (5 ␮g/ml) or LPS (100 ng/ml) for 20 h, and cell lysates were measured for luciferase activity by the dual luciferase assay. Bars represent mean ⫾ SEM (n ⫽ 3). ⴱ, Value of p ⬍ 0.05 vs pcDNA-HA transfected cells treated with CpG. Data are representative of at least two independent experiments. D, Induction of endogenous ATF3 in 293-TLR3 (left panel) and 293-TLR9 (right panel) cells as measured by Western blot in 20 h lysates obtained from pcDNA-HA cotransfected cells from A and B. Note: Levels of ATF3 in lysates from pCGN-ATF3 transfected 293-TLR3 and -TLR9 cells were at least 50-fold greater than pIC- or CpG-stimulated ATF3 for all treatments (data not shown), thus confirming overexpression by pCGN-ATF3 cotransfection.

induced by pIC and CpG in macrophages. Western analysis confirmed that endogenous ATF3 was strongly and specifically induced by pIC and CpG in 293 cells stably expressing the appropriate TLR (Fig. 4D). This provides an explanation for the relatively weak IL-12 promoter activity in our reporter assays. The level of exogenous ATF3 (derived from pCGN-ATF3) was at least 10-fold higher than that of the endogenous ATF3 induced by TLR activation (data not shown), thus providing further repression of the IL-12p40 promoter observed in Fig. 4, B and C. Targeted deletion of atf3 affects immune responses in vivo To measure whether targeted deletion of atf3 affected immune responses in vivo, wild-type, and congenic ATF3-KO mice were injected i.p. with CpG-ODN. To measure immune activity, crude splenocytes were collected 24 h after injection and cultured overnight to measure the release of TNF-␣, IL-4, IL-10, and IFN-␥ by cytometric bead array. Release of TNF-␣ from crude splenocytes of ATF3-KO mice was significantly increased in comparison to

wild types (Fig. 5A), while release of IL-10, IL-4, IL-5, and IFN-␥ was not detected for either genotype (data not shown). Although the enhancement of TNF-␣ production was consistent with in vitro observations (Fig. 2D), it did not predict whether ATF3 would have an impact on infection in vivo. The influence of targeted deletion of atf3 on infection in vivo was assessed using a clinically relevant murine model of influenza A infection. Wild-type and ATF3-KO mice were challenged with a sublethal dose of the PR8 strain of influenza and weight was monitored over time to measure the severity and duration of infection. Although the time of onset and the magnitude of weight loss were identical in both wild-type and ATF3-KO mice, the recovery to initial weight was delayed by at least 10 days for ATF3KOs in comparison to wild types (Fig. 5B). Interestingly, serumneutralizing Ab titers appeared increased 5 mo following infection in ATF3-KO mice (Fig. 5C, left panel), however. the increase did not achieve statistical significance ( p ⫽ 0.09). The number of Ag-specific memory CD8⫹ cells vs PR8 influenza was identical 5

3628


FIGURE 5. Targeted deletion of atf3 affects responses to in vivo CpGODN treatment (A) and PR8 influenza infection (B and C). A, Enhanced TNF-␣ release in the spleen of ATF3-KO mice, relative to wild types, 24 h following i.p. injection of CpG-ODN 1826 (20 ␮g/mouse). Wild-type (C57BL/6) and ATF3-KO mice were injected i.p. with either 200 ␮l of PBS or 20 ␮g of CpG-ODN 1826 in 200 ␮l of PBS. Twenty-four hours following injection, crude splenocytes were plated in quadruplicate wells onto 96-well plates. After overnight incubation, culture supernatants were collected and assessed for various cytokines (TNF-␣, IL-4, IL-5, and IFN-␥) by cytometric bead assay. Among the cytokines measured significant levels of only TNF-␣ were detected. B, The effect of atf3 gene deletion on weight loss triggered by PR8 influenza infection. Wild-type (C57BL/6) and ATF3-KO mice were inoculated with a sublethal dose of influenza A/PR8 intranasally on day 0. Body weight was measured on the indicated days postinfection and calculated as a percent of day 0 weights for each mouse. C, PR8 influenza neutralizing Ab titer (left panel) and CD8 memory cells (right panel) in mice 5 mo following infection. Serum was collected from mice and assessed for titers of PR8 influenza neutralizing Abs as described in Materials and Methods. The total number of PR8 Ag-specific CD8 memory cells was measured in spleens using tetramer staining for NP366 on CD8-positive cells as described in Materials and Methods. Bars represent mean ⫾ SEM for n ⫽ 5 mice.

mo following infection (Fig. 5C, right panel). Thus, the ATF3-KO mice showed no immunodeficiency for generating acquired humoral and memory immune responses.

Discussion Although induction of ATF3 by LPS and bacillus Calmette Gue´rin was described over a decade ago (27, 28), more recent reports have indicated ATF3 is induced in innate immune cells following viral infection (40, 41). This suggested ATF3 likely functions in several TLR-stimulated pathways. Indeed, we have identified ATF3 as a negative regulator of multiple TLR pathways. Different TLR ligands (i.e., zymosan for TLR2/3, pIC for TLR3, LPS for TLR4, and CpG-ODN for TLR9) stimulated rapid induction of ATF3 in cultured mouse macrophages. ATF3 induction was also observed in mouse pDCs following CpG stimulation and infection with the PR8 strain of influenza, which activates pDCs via TLR7 (38, 39). Extending to human cells, TLR-dependent up-regulation of endogenous ATF3 was observed in HEK-293s stably expressing TLR transgenes when stimulated with the appropriate ligand. Most importantly, human DCs that are vital for pathogen recognition and vaccine efficacy were shown to increase ATF3 upon TLR activation. Indeed, LPS, zymosan, and the TLR5 ligand, FliC, stimulated increased ATF3 levels in mDCs. The TLR7 ligand imiquimod and CpG-ODN stimulated increased levels in pDCs. Thus, stimulation of TLR2/6, TLR3, TLR4, TLR5, TLR7 and TLR9 all resulted in up-regulation of ATF3 in mouse and human APCs.

The role of ATF3 in regulating innate immunity was not clear until a recent report from Aderem and colleagues (30), which appeared during preparation of our work, provided evidence of ATF3 functioning in the negative feedback loop of TLR4-stimulated macrophages. Their work demonstrated ATF3 binding to the IL-6 and IL-12 promoter by chromatin-immunoprecipitation assays correlating with enhanced inflammatory responses to TLR4 activation in ATF3-KO mice. This was consistent with ATF3 acting as a transcriptional repressor because, in other systems of cellular stress, ATF3 acts as a transcriptional repressor binding at consensus CREs (42, 43) similar to those present in the promoters of the mouse IL12b and IL6 genes. Our data extend and confirm the report by Aderem and colleagues (30). In this study, a role for ATF3 in regulating responses to multiple TLRs was uncovered using mice with a targeted deletion of atf3 (32) and by promoter-reporter assays. Production of IL-12 and IL-6 was enhanced in primary macrophages from ATF3-KO mice, in comparison to wild type, for all TLR ligands stimulating up-regulation of ATF3 in wild-type cells. TNF-␣ production was selectively enhanced by CpG treatment, but not with other TLR ligands. Enhanced secretion of IL-12 was also associated with targeted deletion of atf3 in cultured mDCs stimulated via TLRs. Moreover, mDCs from ATF3-KO mice produced equivalent or enhanced amounts of IL-10, a well-established autocrine/ paracrine inhibitor of IL-12, relative to wild types. Thus, enhanced IL-12 levels are not a result of IL-10 deficiency in ATF3-KO mice. In accord with this, overexpression of ATF3 in promoter-reporter assays antagonized activation of the IL-12p40 promoter triggered by TLR3 and TLR9, directly implicating ATF3 as transcriptional repressor. Thus, ATF3 is induced by TLR stimulation in both mouse and human APCs and functions as a negative regulatory transcription factor for cytokine production. The identification of an ATF3-dependent feedback regulatory loop for TLR4, 2/6, 3, 5, 7, and 9 indicates a broad role for ATF3 in innate immunity. Considering that zymosan, a TLR2/6 ligand, is a cell wall component of yeast and TLR3, 7, and 9 are involved in viral defense (1) where ATF3 has been reported as a host response gene (40, 41), a role for protection from LPS toxicity (30) could be expanded to similar protective roles associated with yeast or viral infection. Because many viruses manipulate host response genes to gain advantages for entry and replication, it would be of interest to determine whether viruses manipulate the negative regulatory effect of ATF3 for survival advantage. Moreover, because synthetic compounds that mimic the natural ligands for TLR3, 7, and 9 are at various stages of clinical development as vaccine adjuvants, cancer immunotherapeutics, and immune response modifiers (5, 44, 45), ATF3 may influence the efficacy or toxicity of pharmaceutical agents. Because overactive or prolonged innate immune responses can have deleterious effects on the host, negative regulators target various steps of TLR signaling (12). Following TLR ligation, adapter molecules such as MyD88 and TRIF are recruited to the cytoplasmic tail via mutual interaction of TLR, IL-1R interaction domains. Protein-protein interactions lead to recruitment and activation of IRAK-1, resulting in activation of TNFR-associated factor-6 via ubiquitinylation. These activities ultimately lead to activation of transcription factors c-jun, ATF-2, and NF-␬B that act as positive regulators of transcription of genes containing AP-1 and NF-␬B sites. Soluble TLRs, orphan receptors and truncated adapter proteins short-circuit signaling resulting in negative regulation (12). Tollip, a negative regulatory TLR adapter, and IRAK-M both act by antagonizing IRAK-1 phosphorylation (20, 46). Most of these changes occur at late stages of innate immune activation and function as feedback inhibitors. Conversely, A20 rapidly induces and

The Journal of Immunology blocks TNFR-associated factor-6 ubiquitinylation and subsequent activation, thereby controlling the amplitude of the initial response to TLR ligation (19, 47). Similar to A20, ATF3 was rapidly induced, suggesting that ATF3 may control the amplitude of IL-12 and IL-6 responses following TLR detection of pathogen-associated molecular patterns. We demonstrated enhancement of cytokine release from splenocytes in ATF3-KO mice treated with a TLR9 agonist in vivo. Aderem and colleagues (30) similarly reported enhanced serum levels of IL-12 in a LPS challenge model at a time when wild-type mice resolve IL-12 production to near baseline, i.e.12 h postinjection. They concluded that ATF3 was functioning in the negative feedback loop following TLR4 stimulation and our data with CpGODN are in accord. When challenged with a sublethal dose of the PR8 strain of influenza, the ATF3-KO mice showed a delay in the recovery phase, yet no deficiency in generating PR8-specific neutralizing Abs or CD8 memory cells (Fig. 5, B and C). Because cytokine production following influenza infection is triggered via TLR7 (38, 39), it is plausible that the delayed recovery in body weight for ATF3-KOs derives from a lack of negative feedback for TLR7-stimulated cytokine production and concomitant toxicity. Thus, the negative feedback role of ATF3 for controlling cytokine toxicity would extend to a model of infection with high clinical relevance. It cannot be ruled out, however, that viral clearance is slowed as a result of atf3 deficiency. The ATF3 promoter contains several TLR-responsive regulatory elements such as AP-1 and NF-␬B sites (48), thus the early phase of ATF3 induction is likely a primary response to TLR signaling. Indeed, ATF3 was strongly induced before (within 1 h) production of significant amounts of cytokines or STAT1 phosphorylation that can have secondary roles in gene induction after TLR stimulation. In contrast, TLRs that stimulate type I IFN production resulted in a second wave of ATF3 induction between 4 and 6 h poststimulation (Fig. 1D) and all tested TLR ligands stimulated up-regulation of ATF3 at 20 –24 h poststimulation (Fig. 4D). Whether subsequent waves of ATF3 induction were dependent on secondary cytokine production remains to be determined. Indeed, Taylor et al. (29) have identified ATF3 as an IFN-stimulated gene. Consistent with this, ATF3 was not a transcript upregulated by 6 h treatment with pIC/CpG in BMMs from type I IFNR-KO mice (Fig. 1A). Whether targeted deletion of atf3 mimics biological or clinical circumstances where ATF3 is dysregulated needs to be further ascertained. Pancreatic ␤ cells from ATF3-KO mice were partially protected from apoptosis induced by proinflammatory cytokines and NO, suggesting ATF3-dependent apoptosis may have a role in type I diabetes (32). It is unlikely that this is impacting the enhanced cytokine response in ATF3-KO mice, because significant apoptosis in wild-type APC has not been observed (as measured by morphological changes, cell viability, or poly(ADP-ribose) polymerase cleavage; data not shown). Indeed, macrophages and DCs must survive and function in an inflammatory environment, thus it would be of interest to determine whether ATF3 provided a protective effect against apoptosis in APCs in contrast to ␤ cells. In cancer, enhanced up-regulation of negative regulatory immune functions may provide a growth advantage for neoplastic cells in the tumor microenvironment (49, 50). Interestingly, ectopic expression of ATF3 enhanced the metastatic potential B16-F1 melanoma cell line in mice and expression of ATF3 is increased in the highly metastatic B16-F10 cells (51). ATF3 was also an IFN-stimulated gene in WM-9 human melanoma cells (unpublished observations as communicated by M. Whitmore and E. Borden/Taussig Cancer Center, Cleveland Clinic, Cleveland, OH). Thus, it is reasonable to hypothesize that the negative regulatory effects of ATF3

3629 on innate immunity provide a growth advantage in melanoma. Furthermore, overexpression of ATF3 is correlated with metastatic tumors in human colon cancer (52). The role of ATF3 in immune regulation and connections to inflammatory diseases warrant further investigation.

Acknowledgments We thank Dr. Katherine Fitzgerald (University of Massachusetts, Worcester, MA) for providing HEK-293-TLR3 and HEK-293-TLR9 cell lines, Dr. Keiko Ozato (National Institutes of Health, Bethesda, MD) for providing the IL-12p40-firefly luciferase promoter-reporter construct and Dr. E. John Wherry (The Wistar Institute, Philadelphia, PA) for providing the DbNP366 –374 tetramer. We also thank Dr. Pierre Triozzi and Wayne Aldrich (Taussig Cancer Center, Cleveland Clinic, Cleveland, OH) for technical assistance in preparing primary mouse and human DC cultures. Finally, we are indebted to Dr. Ernest Borden (Taussig Cancer Center) for helpful comments and to Dr. Frances Cribbin (Monash Institute of Medical Research, Monash University, Melbourne, Australia) for technical assistance in drafting the manuscript.

Disclosures The authors have no financial conflict of interest.

References 1. Kaisho, T., and S. Akira. 2006. Toll-like receptor function and signaling. J. Allergy Clin. Immunol. 117: 979 –987; quiz 988. 2. Garcia-Sastre, A., and C. A. Biron. 2006. Type 1 interferons and the virus-host relationship: a lesson in detente. Science 312: 879 – 882. 3. Prins, R. M., N. Craft, K. W. Bruhn, H. Khan-Farooqi, R. C. Koya, R. Stripecke, J. F. Miller, and L. M. Liau. 2006. The TLR-7 agonist, imiquimod, enhances dendritic cell survival and promotes tumor antigen-specific T cell priming: relation to central nervous system antitumor immunity. J. Immunol. 176: 157–164. 4. Whitmore, M. M., S. Li, L. Falo, Jr., and L. Huang. 2001. Systemic administration of LPD prepared with CpG oligonucleotides inhibits the growth of established pulmonary metastases by stimulating innate and acquired antitumor immune responses. Cancer Immunol. Immunother. 50: 503–514. 5. Whitmore, M. M., M. J. DeVeer, A. Edling, R. K. Oates, B. Simons, D. Lindner, and B. R. Williams. 2004. Synergistic activation of innate immunity by doublestranded RNA and CpG DNA promotes enhanced antitumor activity. Cancer Res. 64: 5850 –5860. 6. Zacharowski, K., P. A. Zacharowski, A. Koch, A. Baban, N. Tran, R. Berkels, C. Papewalis, K. Schulze-Osthoff, P. Knuefermann, U. Zahringer, et al. 2006. Toll-like receptor 4 plays a crucial role in the immune-adrenal response to systemic inflammatory response syndrome. Proc. Natl. Acad. Sci. USA 103: 6392– 6397. 7. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282: 2085–2088. 8. Stack, J., I. R. Haga, M. Schroder, N. W. Bartlett, G. Maloney, P. C. Reading, K. A. Fitzgerald, G. L. Smith, and A. G. Bowie. 2005. Vaccinia virus protein A46R targets multiple Toll-like-interleukin-1 receptor adaptors and contributes to virulence. J. Exp. Med. 201: 1007–1018. 9. Roelofs, M. F., L. A. Joosten, S. Abdollahi-Roodsaz, A. W. van Lieshout, T. Sprong, F. H. van den Hoogen, W. B. van den Berg, and T. R. Radstake. 2005. The expression of Toll-like receptors 3 and 7 in rheumatoid arthritis synovium is increased and costimulation of Toll-like receptors 3, 4, and 7/8 results in synergistic cytokine production by dendritic cells. Arthritis Rheum. 52: 2313–2322. 10. Leadbetter, E. A., I. R. Rifkin, A. M. Hohlbaum, B. C. Beaudette, M. J. Shlomchik, and A. Marshak-Rothstein. 2002. Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors. Nature 416: 603– 607. 11. Prinz, M., F. Garbe, H. Schmidt, A. Mildner, I. Gutcher, K. Wolter, M. Piesche, R. Schroers, E. Weiss, C. J. Kirschning, et al. 2006. Innate immunity mediated by TLR9 modulates pathogenicity in an animal model of multiple sclerosis. J. Clin. Invest. 116: 456 – 464. 12. Liew, F. Y., D. Xu, E. K. Brint, and L. A. O’Neill. 2005. Negative regulation of Toll-like receptor-mediated immune responses. Nat. Rev. Immunol. 5: 446 – 458. 13. Iwami, K. I., T. Matsuguchi, A. Masuda, T. Kikuchi, T. Musikacharoen, and Y. Yoshikai. 2000. Cutting edge: naturally occurring soluble form of mouse Toll-like receptor 4 inhibits lipopolysaccharide signaling. J. Immunol. 165: 6682– 6686. 14. LeBouder, E., J. E. Rey-Nores, N. K. Rushmere, M. Grigorov, S. D. Lawn, M. Affolter, G. E. Griffin, P. Ferrara, E. J. Schiffrin, B. P. Morgan, and M. O. Labeta. 2003. Soluble forms of Toll-like receptor (TLR)2 capable of modulating TLR2 signaling are present in human plasma and breast milk. J. Immunol. 171: 6680 – 6689. 15. Janssens, S., K. Burns, E. Vercammen, J. Tschopp, and R. Beyaert. 2003. MyD88S, a splice variant of MyD88, differentially modulates NF-␬B- and AP1-dependent gene expression. FEBS Lett. 548: 103–107.

3630


16. Kobayashi, K., L. D. Hernandez, J. E. Galan, C. A. Janeway, Jr., R. Medzhitov, and R. A. Flavell. 2002. IRAK-M is a negative regulator of Toll-like receptor signaling. Cell 110: 191–202. 17. Pathak, S. K., S. Basu, A. Bhattacharyya, S. Pathak, M. Kundu, and J. Basu. 2005. Mycobacterium tuberculosis lipoarabinomannan-mediated IRAK-M induction negatively regulates Toll-like receptor-dependent interleukin-12 p40 production in macrophages. J. Biol. Chem. 280: 42794 – 42800. 18. O’Reilly, S. M., and P. N. Moynagh. 2003. Regulation of Toll-like receptor 4 signalling by A20 zinc finger protein. Biochem. Biophys. Res. Commun. 303: 586 –593. 19. Boone, D. L., E. E. Turer, E. G. Lee, R. C. Ahmad, M. T. Wheeler, C. Tsui, P. Hurley, M. Chien, S. Chai, O. Hitotsumatsu, et al. 2004. The ubiquitin-modifying enzyme A20 is required for termination of Toll-like receptor responses. Nat. Immunol. 5: 1052–1060. 20. Zhang, G., and S. Ghosh. 2002. Negative regulation of Toll-like receptor-mediated signaling by Tollip. J. Biol. Chem. 277: 7059 –7065. 21. Burns, K., J. Clatworthy, L. Martin, F. Martinon, C. Plumpton, B. Maschera, A. Lewis, K. Ray, J. Tschopp, and F. Volpe. 2000. Tollip, a new component of the IL-1RI pathway, links IRAK to the IL-1 receptor. Nat. Cell Biol. 2: 346 –351. 22. Qin, J., Y. Qian, J. Yao, C. Grace, and X. Li. 2005. SIGIRR inhibits interleukin-1 receptor- and Toll-like receptor 4-mediated signaling through different mechanisms. J. Biol. Chem. 280: 25233–25241. 23. Wald, D., J. Qin, Z. Zhao, Y. Qian, M. Naramura, L. Tian, J. Towne, J. E. Sims, G. R. Stark, and X. Li. 2003. SIGIRR, a negative regulator of Toll-like receptorinterleukin 1 receptor signaling. Nat. Immunol. 4: 920 –927. 24. Hai, T., and M. G. Hartman. 2001. The molecular biology and nomenclature of the activating transcription factor/cAMP responsive element binding family of transcription factors: activating transcription factor proteins and homeostasis. Gene 273: 1–11. 25. Chen, B. P., G. Liang, J. Whelan, and T. Hai. 1994. ATF3 and ATF3 ⌬ Zip: transcriptional repression versus activation by alternatively spliced isoforms. J. Biol. Chem. 269: 15819 –15826. 26. Hai, T., C. D. Wolfgang, D. K. Marsee, A. E. Allen, and U. Sivaprasad. 1999. ATF3 and stress responses. Gene Exp. 7: 321–335. 27. Drysdale, B. E., D. L. Howard, and R. J. Johnson. 1996. Identification of a lipopolysaccharide inducible transcription factor in murine macrophages. Mol. Immunol. 33: 989 –998. 28. Farber, J. M. 1992. A collection of mRNA species that are inducible in the RAW 264.7 mouse macrophage cell line by gamma interferon and other agents. Mol. Cell. Biol. 12: 1535–1545. 29. Taylor, M. W., W. M. Grosse, J. E. Schaley, C. Sanda, X. Wu, S. C. Chien, F. Smith, T. G. Wu, M. Stephens, M. W. Ferris, et al. 2004. Global effect of PEG-IFN-␣ and ribavirin on gene expression in PBMC in vitro. J. Interferon Cytokine Res. 24: 107–118. 30. Gilchrist, M., V. Thorsson, B. Li, A. G. Rust, M. Korb, K. Kennedy, T. Hai, H. Bolouri, and A. Aderem. 2006. Systems biology approaches identify ATF3 as a negative regulator of Toll-like receptor 4. Nature 441: 173–178. 31. Tallant, T., A. Deb, N. Kar, J. Lupica, M. J. de Veer, and J. A. DiDonato. 2004. Flagellin acting via TLR5 is the major activator of key signaling pathways leading to NF-␬B and proinflammatory gene program activation in intestinal epithelial cells. BMC Microbiol. 4: 33. 32. Hartman, M. G., D. Lu, M. L. Kim, G. J. Kociba, T. Shukri, J. Buteau, X. Wang, W. L. Frankel, D. Guttridge, M. Prentki, et al. 2004. Role for activating transcription factor 3 in stress-induced ␤-cell apoptosis. Mol. Cell. Biol. 24: 5721–5732. 33. Scherle, P. A., G. Palladino, and W. Gerhard. 1992. Mice can recover from pulmonary influenza virus infection in the absence of class I-restricted cytotoxic T cells. J. Immunol. 148: 212–217.

34. Hume, D. A., I. L. Ross, S. R. Himes, R. T. Sasmono, C. A. Wells, and T. Ravasi. 2002. The mononuclear phagocyte system revisited. J. Leukocyte Biol. 72: 621– 627. 35. Gordon, S., and P. R. Taylor. 2005. Monocyte and macrophage heterogeneity. Nat. Rev. Immunol. 5: 953–964. 36. Irvine, K. M., C. J. Burns, A. F. Wilks, S. Su, D. A. Hume, and M. J. Sweet. 2006. A CSF-1 receptor kinase inhibitor targets effector functions and inhibits proinflammatory cytokine production from murine macrophage populations. FASEB J. 20: 1921–1923. 37. Chaperot, L., A. Blum, O. Manches, G. Lui, J. Angel, J. P. Molens, and J. Plumas. 2006. Virus or TLR agonists induce TRAIL-mediated cytotoxic activity of plasmacytoid dendritic cells. J. Immunol. 176: 248 –255. 38. Lund, J. M., L. Alexopoulou, A. Sato, M. Karow, N. C. Adams, N. W. Gale, A. Iwasaki, and R. A. Flavell. 2004. Recognition of single-stranded RNA viruses by Toll-like receptor 7. Proc. Natl. Acad. Sci. USA 101: 5598 –5603. 39. Diebold, S. S., T. Kaisho, H. Hemmi, S. Akira, and C. Reis e Sousa. 2004. Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 303: 1529 –1531. 40. Zhao, H., F. Granberg, L. Elfineh, U. Pettersson, and C. Svensson. 2003. Strategic attack on host cell gene expression during adenovirus infection. J. Virol. 77: 11006 –11015. 41. Barnabas, S., and O. M. Andrisani. 2000. Different regions of hepatitis B virus X protein are required for enhancement of bZip-mediated transactivation versus transrepression. J. Virol. 74: 83–90. 42. Kim, H. B., M. Kong, T. M. Kim, Y. H. Suh, W. H. Kim, J. H. Lim, J. H. Song, and M. H. Jung. 2006. NFATc4 and ATF3 negatively regulate adiponectin gene expression in 3T3-L1 adipocytes. Diabetes 55: 1342–1352. 43. Wolfgang, C. D., B. P. Chen, J. L. Martindale, N. J. Holbrook, and T. Hai. 1997. gadd153/Chop10, a potential target gene of the transcriptional repressor ATF3. Mol. Cell. Biol. 17: 6700 – 6707. 44. Urosevic, M., R. Dummer, C. Conrad, M. Beyeler, E. Laine, G. Burg, and M. Gilliet. 2005. Disease-independent skin recruitment and activation of plasmacytoid predendritic cells following imiquimod treatment. J. Natl. Cancer Inst. 97: 1143–1153. 45. Furumoto, K., L. Soares, E. G. Engleman, and M. Merad. 2004. Induction of potent antitumor immunity by in situ targeting of intratumoral DCs. J. Clin. Invest. 113: 774 –783. 46. Nakayama, K., S. Okugawa, S. Yanagimoto, T. Kitazawa, K. Tsukada, M. Kawada, S. Kimura, K. Hirai, Y. Takagaki, and Y. Ota. 2004. Involvement of IRAK-M in peptidoglycan-induced tolerance in macrophages. J. Biol. Chem. 279: 6629 – 6634. 47. Gon, Y., Y. Asai, S. Hashimoto, K. Mizumura, I. Jibiki, T. Machino, C. Ra, and T. Horie. 2004. A20 inhibits Toll-like receptor 2- and 4-mediated interleukin-8 synthesis in airway epithelial cells. Am. J. Respir. Cell Mol. Biol. 31: 330 –336. 48. Liang, G., C. D. Wolfgang, B. P. Chen, T. H. Chen, and T. Hai. 1996. ATF3 gene: genomic organization, promoter, and regulation. J. Biol. Chem. 271: 1695–1701. 49. Smyth, M. J., M. W. Teng, J. Swann, K. Kyparissoudis, D. I. Godfrey, and Y. Hayakawa. 2006. CD4⫹CD25⫹ T regulatory cells suppress NK cell-mediated immunotherapy of cancer. J. Immunol. 176: 1582–1587. 50. Chen, M. L., M. J. Pittet, L. Gorelik, R. A. Flavell, R. Weissleder, H. von Boehmer, and K. Khazaie. 2005. Regulatory T cells suppress tumorspecific CD8 T cell cytotoxicity through TGF-␤ signals in vivo. Proc. Natl. Acad. Sci. USA 102: 419 – 424. 51. Ishiguro, T., M. Nakajima, M. Naito, T. Muto, and T. Tsuruo. 1996. Identification of genes differentially expressed in B16 murine melanoma sublines with different metastatic potentials. Cancer Res. 56: 875– 879. 52. Ishiguro, T., and H. Nagawa. 2000. Expression of the ATF3 gene on cell lines and surgically excised specimens. Oncol. Res. 12: 181–183.