Dendritic Cell Differentiation Induced − Critical Role of AZI2 in GM-CSF

Critical Role of AZI2 in GM-CSF−Induced Dendritic Cell Differentiation Masahiro Fukasaka, Daisuke Ori, Tatsukata Kawagoe, Satoshi Uematsu, Kenta Maruyama, Toshihiko Okazaki, This information is current as Tatsuya Kozaki, Tomoko Imamura, Sarang Tartey, Takashi of September 27, 2021. Mino, Takashi Satoh, Shizuo Akira and Osamu Takeuchi J Immunol published online 22 April 2013 http://www.jimmunol.org/content/early/2013/04/23/jimmun

ol.1203155 Downloaded from

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 27, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published April 24, 2013, doi:10.4049/jimmunol.1203155 The Journal of Immunology

Critical Role of AZI2 in GM-CSF–Induced Dendritic Cell Differentiation

Masahiro Fukasaka,*,† Daisuke Ori,*,‡,x Tatsukata Kawagoe,*,1 Satoshi Uematsu,*,2 Kenta Maruyama,* Toshihiko Okazaki,*,† Tatsuya Kozaki,* Tomoko Imamura,*,‡,x Sarang Tartey,*,‡,x Takashi Mino,‡,x Takashi Satoh,* Shizuo Akira,*,{ and Osamu Takeuchi*,‡,x

TNFR-associated factor family member–associated NF-kB activator (TANK)–binding kinase 1 (TBK1) is critical for the activation of IFN regulatory factor 3 and type I IFN production upon virus infection. A set of TBK1-binding proteins, 5-azacytidine–induced gene 2 (AZI2; also known as NAP1), TANK, and TBK1-binding protein 1 (TBKBP1), have also been implicated in the production

of type I IFNs. Among them, TANK was found to be dispensable for the responses against virus infection. However, physiological Downloaded from roles of AZI2 and TBKBP1 have yet to be clarified. In this study, we found that none of these TBK1-binding proteins is critical for type I IFN production in mice. In contrast, AZI2, but not TBKBP1, is critical for the differentiation of conventional dendritic cells (cDCs) from bone marrow cells in response to GM-CSF. AZI2 controls GM-CSF–induced cell cycling of bone marrow cells via TBK1. GM-CSF–derived DCs from AZI2-deficient mice show severe defects in cytokine production and T cell activation both in vitro and in vivo. Reciprocally, overexpression of AZI2 results in efficient generation of cDCs, and the cells show enhanced T cell

activation in response to Ag stimulation. Taken together, AZI2 expression is critical for the generation of cDCs by GM-CSF and http://www.jimmunol.org/ can potentially be used to increase the efﬁciency of immunization by cDCs. The Journal of Immunology, 2013, 190: 000–000.

endritic cells (DCs) play a central role in instructing DCs and macrophages sense the invasion of pathogens through a T cells to activate acquired immunity upon infection set of pattern-recognition receptors, including TLRs and retinoic D (1–4). They phagocytose microbes such as bacteria and acid inducible gene-I (RIG-I)–like receptors (RLRs) (5, 6). Whereas viruses, process them into Ags, and present the Ags on MHC TLRs are transmembrane proteins that recognize microbial com- molecules. Besides Ag presentation, innate immune cells such as ponents on the cell surface or in endosomes, RLRs sense viral dsRNA in the cytoplasm. The signaling pathways triggered by the receptors lead to the production of proinflammatory cytokines, in- by guest on September 27, 2021 *Laboratory of Host Defense, Immunology Frontier Research Center; †Nitto Denko cluding TNF, IL-6, IL-12p40, and type I IFNs (7). In addition, the Company, Ibaraki, Osaka 567-8680, Japan; ‡Laboratory of Infection and Prevention, x signaling pathways increase the surface expression of costimula- Institute for Virus Research, Kyoto University; Core Research for Evolutional Sci- ence and Technology, Japan Science and Technology Agency, Sakyo-ku, Kyoto 606- tory molecules, such as CD40, CD80, and CD86, on DCs (8). It is 8507, Japan; and {Research Institute for Microbial Diseases, Osaka University, Suita, well documented that costimulation of DCs is critical for mounting Osaka 565-0871, Japan efficient T cell activation and vaccination (9). 1 Current address: Department of Ophthalmology, Yokohama City University School DCs are classified into several different subsets, including CD82 of Medicine, Yokohama, Japan. and CD8+ conventional DCs (cDCs) and plasmacytoid DCs (pDCs) 2Current address: International Research and Development Center for Mucosal Vac- cines, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan. (10, 11). pDCs, which are characterized by surface expression of CD11c and B220, are known to produce vast amounts of IFN-a in Received for publication November 14, 2012. Accepted for publication March 26, 2013. response to virus infection via TLR7 and TLR9 (12). On the other This work was supported by Special Coordination Funds of the Japanese Ministry of hand, cDCs recognize pathogen infection via TLRs and RLRs for Education, Culture, Sports, Science and Technology and grants from the Ministry the production of cytokines. Among the cDCs, CD8+ DCs differ- of Health, Labour and Welfare in Japan and the Japan Society for the Promotion of entiate in a Batf3-dependent manner (13) and have a strong capacity Science through the Funding Program for World-Leading Innovative R&D on Sci- ence and Technology (FIRST Program). This work was also supported in part by for cross-presentation of foreign Ags to MHC class I molecules grants from the Takeda Science Foundation, the Naito Foundation, and the Mochida (11). In mice, both cDCs and pDCs can be generated from bone Foundation. marrow (BM) cells by culture with distinct cytokines. FLT3 ligand Address correspondence and reprint requests to Prof. Osamu Takeuchi, Laboratory of (FLT3L) is essential for pDC and a set of cDC differentiation from Infection and Prevention, Institute for Virus Research, Kyoto University, 53 Shogoin Kawara-cho, Sakyo-ku, Kyoto 606-8507, Japan. E-mail address: [email protected]. BM precursor cells (14). In contrast, culture of mouse BM cells or ac.jp human PBMCs in the presence of GM-CSF is widely used to induce Abbreviations used in this article: AZI2, 5-azacytidine–induced gene 2; BM, bone cDCs (15). GM-CSF induces the survival, proliferation, and dif- marrow; BMDC, bone marrow–derived dendritic cell; cDC, conventional dendritic ferentiation of hematopoietic cells including DCs (16). The GM- cell; DC, dendritic cell; DKO, double knockout; EMCV, encephalomyocarditis virus; FLT3L, FLT3 ligand; FLT3L-DC, FLT3L-derived BMDC; Flu, influenza virus; GM- CSF receptor is composed of an a subunit and a common b subunit DC, GM-CSF–derived BMDC; IKK-i,IkB kinase-i; IRF, IFN regulatory factor; KO, shared by IL-3Rs and IL-5Rs (17). After stimulation of GM-CSF knockout; MOI, multiplicity of infection; NDV, Newcastle disease virus; pDC, plasmacytoid DC; poly(I:C), polyinosinic-polycytidylic acid; RLR, retinoic acid–induc- receptors, signaling pathways are activated through the formation of ible gene-I–like receptor; TANK, TNFR-associated factor family member–associated a unique dodecameric receptor complex (18, 19). The transcription NF-kB activator; TBK1, TANK-binding kinase 1; TBKBP1, TBK1-binding protein factors STAT5 and IFN regulatory factor 4 (IRF4) (20, 21) are es- 1; TRAF, TNFR-associated factor; WT, wild-type. sential for GM-CSF–mediated signaling pathways, and MAPK and Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 AKT are also activated by GM-CSF (17).

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203155 2 DC DIFFERENTIATION REQUIRES AZI2

The molecular mechanisms for how TLRs and RLRs transactivate deoxynucleotide 1668) were purchased from Invivogen. LPS from Escher- type I IFN genes and proinflammatory cytokines have been ex- ichia coli was purchased from Sigma. GM-CSF and FLT3L were purchased tensively studied. With the exception of TLR3, TLRs trigger an from PeproTech. MyD88-dependent signaling pathway that induces activation of Generation of GM-DCs and FLT3L-derived BMDCs the transcription factor NF-kB, thereby leading to the expression BM cells were harvested from the femurs and tibias of mice by flushing the of proinflammatory cytokine genes. In contrast, TLR3 and TLR4 marrow cavity with HBSS. GM-DC:BM cells were resuspended at the signal through another adaptor molecule, Toll/IL-1R domain– density of 1 3 106 cells/ml in complete RPMI 1640 (supplemented with containing adaptor protein inducing IFN-b, which results in the 10% FCS, 50 mM 2-ME, 1% penicillin [100 U/ml] and 1% streptomycin activation of a set of kinases called IkB kinase-i (IKK-i, also known [100 U/ml]) and 10 ng/ml mouse GM-CSF. The cell culture medium was ε replaced on days 2 and 4, and cells were grown for 6 d. FLT3L-derived as IKK ) and TNFR-associated factor (TRAF) family member–as- BMDC (Flt3L-DC):BM cells were lysed with RBC lysing buffer (Sigma) sociated NF-kB activator (TANK)–binding kinase 1 (TBK1). These and resuspended in complete RPMI 1640 containing 20 ng/ml Flt3L, and kinases phosphorylate IRF3 and IRF7, which induce transactivation cells were grown for 8 d. of type I IFN and IFN-inducible genes (6, 22). RLR signaling also Quantitative real-time PCR and RT-PCR activates IKK-i and TBK1 for the production of type I IFNs. IKK-i and TBK1 are known to associate with several adaptor molecules, Total RNA was extracted using TRIzol (Invitrogen), and cDNA was gen- including TANK, 5-azacytidine–induced gene 2 (AZI2; also known erated by using ReverTraAce (Toyobo) according to the manufacturer’s instructions. Quantitative real-time PCR was carried out with primers, asNAP1),andTBK1-bindingprotein1(TBKBP1;alsoknownas purchased from Applied Biosystems, specific for IL-6, TNF, IFN-b, and 18S SINTBAD). These three proteins harbor a coiled-coil motif and a rRNA, as an internal control, using a 7700 Sequence Detector (Applied TBK1-binding motif (23), and were reported to be required for the Biosystems) as described previously (31). RT-PCR analysis was conducted Downloaded from downstream signaling of TLRs and RLRs to activate TBK1 and on cDNA samples with the following primers: human AZI2 (forward: 59- IKK-i (23–25). However, the generation of TANK-deficient mice ggatgcactggtagaagatgatatctg-39,reverse:59-tcagctggaggagttctacttctttagat-39) and GAPDH (forward: 59-cttactccttggaggccatg-39,reverse:59-ttagcccccctggc- revealed that TANK is dispensable for TLR- and RLR-mediated caagg-39). The primer pairs and rTaq polymerase (Toyobo) were used for PCR. type I IFN production. In contrast, TANK acts as a negative regulator of MyD88-dependent signaling by suppressing the activation ELISA of TRAF6 (26). TANK-deficient mice experience development of Cell culture supernatants were assessed for the concentration of IL-6, IL- http://www.jimmunol.org/ fatal autoimmune glomerular nephritis and osteoporosis because of 12p40, TNF-a, IFN-b, and IFN-g according to manufacturers’ instructions enhanced activation of immune cells and osteoclasts (26, 27). (R&D Systems or PBL Interferon Source). In this study, we analyzed the functional roles of AZI2 and Flow cytometry TBKBP1 in vivo by generating mice that lack these genes. Sur- prisingly, both AZI2 and TBKBP1 were dispensable for the sig- All Abs used were purchased from BD Pharmingen or BioLegend. Cells were washed in flow cytometry buffer (2% [v/v] FCS and 2 mM EDTA in naling pathways emanating from TLRs and RLRs. We found that PBS, pH 7.5) and then, after treatment with each Ab, incubated for 20 min AZI2, but not TBKBP1, is essential for the differentiation of cDCs and washed twice with flow cytometry buffer. The cells were analyzed by in response to culture with GM-CSF by activating TBK1, and at the flow cytometry on a FACSCalibur instrument (BD Biosciences), followed same time, AZI2–/– GM-CSF–derived BM-derived DCs (BMDCs by data analysis using FlowJo v7.2.2 software (Tree Star). by guest on September 27, 2021 [GM-DCs]) showed severely impaired T cell activation both in vitro DC maturation and in vivo. Taken together, these data demonstrate a novel function Day 6 GM-DCs were harvested and stimulated at 1 3 106 cells/ml with LPS of AZI2 for the induction of cDCs and further suggest the potential (1 mg/ml) or CpG DNA (1 mM) to induce DC maturation. After 24 h, the of AZI2-targeting approaches for enhancement of DC vaccine cells were subjected to flow cytometric analysis and the culture supernatants potency. were frozen at 220°C for measurement of secreted cytokines by ELISA. CD4+ T cell proliferation assays

Materials and Methods 2/2 2/2 2/2 Wild-type (WT) and AZI2 GM-DCs or FLT3L-DCs were stimulated Generation of AZI2 and TBKBP1 mice with LPS (1 mg/ml) and CpG DNA (1 mM) for 24 h, and irradiated (3000 + The AZI2 and TBKBP1 genes were isolated from genomic DNA extracted rad) to prevent cell division. For the MLR, CD4 T cells were isolated 3 6 + from embryonic stem cells by PCR. The targeting vector was constructed from BALB/c mice splenocytes. A total of 1 10 CD4 T cells was 2 3 5 1 3 5 3 5 by replacing fragments encoding the exon 3 (AZI2) or exons 4–6 (TBKBP1) mixed with (1/3) 10 , (1/3) 10 ,or1 10 GM-DCs or FLT3L- DCs in each well. After 72 h of coculture, for proliferation assays, cells with a neomycin-resistance gene cassette (neo), and an HSV thymidine ki- 3 nase driven by PGK promoter was inserted into the genomic fragment for were pulsed with 2 mCi [ H]thymidine for the last 18 h. Levels of cyto- negative selection (Fig. 1A, 1B). After the targeting vector was transfected kines in the culture supernatant was measured by ELISA. into embryonic stem cells, G418 and ganciclovir doubly resistant colonies were selected and screened by PCR, and further confirmed by Southern Proliferation and cell-cycle assay blotting. Homologous recombinants were microinjected into C57BL/6 fe- BM cells were isolated from WT or AZI22/2 mice and stimulated with GM- male mice, and heterozygous F1 progenies were intercrossed to obtain CSF (10 ng/ml) for 48 h. For cell-cycle analysis, cells were stained with knockout (KO) mice. Double KO (DKO) mice were generated by inter- propidium iodide buffer (0.4% sodium citrate, 0.03% Nonidet P-40, 0.05 crossing AZI2 hetero and TBKBP1 hetero mice. KO and littermate control mg/ml propidium iodide [Sigma], and 0.02 mg/ml RNaseA) for 15 min on mice were used throughout the experiments. All animal experiments were ice and assessed by flow cytometry. For cell proliferation assay, the BM carried out with the approval of the Animal Research Committee of the cells were pulsed with BrdU for the last 24 h, and BrdU incorporation was Research Institute for Microbial Diseases at Osaka University. analyzed using the BrdU Flow kit (BD Pharmingen) according to the manufacturer’s instructions or cells were pulsed with 2 mCi [3H]thymidine Cells, viruses, and reagents for the last 12 h. At 3 d after injection of 2 ml of 4% (w/v) thioglycollate medium (Sigma) Immunoblot analysis i.p., peritoneal exudate cells were isolated by washing with ice-cold HBSS as previously described (28). CD4+ T cells or CD90.22 cells were isolated BM cells were isolated from WT or AZI22/2 mice, and stimulated with from splenocyte single-cell suspensions by negative selection with CD4+ GM-CSF for 0–20 min and lysed with lysis buffer (20 mM Tris-HCl, pH T cell Isolation kit (Miltenyi Biotec) or CD90.2 Microbeads (Miltenyi 7.5, 150 mM NaCl, 1 mM EDTA, and 1% [v/v] Nonidet P-40) containing Biotec). Encephalomyocarditis virus (EMCV), HSV, Newcastle disease compete mini protease inhibitor mixture (Roche). Cell lysates were sep- virus, and influenza virus (Flu; PR8) have been described previously (29, arated by standard SDS-PAGE and analyzed by immunoblot. The fol- 30). Polyinosinic-polycytidylic acid [poly(I:C)] and CpG DNA (oligo- lowing Abs were used: phospho-STAT5 (Cell Signaling Technology), The Journal of Immunology 3 phospho-ERK1/2 (T202/Y204; Cell Signaling Technology), phospho-AKT AZI2 and TBKBP1 are dispensable for TLR- and RLR-induced (S473; Cell Signaling Technology), and b-actin (Santa Cruz Biotechnol- cytokine and type I IFN production in macrophages ogy). Because AZI2 and TBKBP1 are implicated in the activation of IRF3 Retroviral transduction by interacting with TBK1 and IKK-i (23, 24), we first examined BM cells were isolated from AZI22/2 mice and incubated with retroviral the responses of peritoneal macrophages to infection with NDV, an supernatant supplemented with 10 ng/ml GM-CSF and 10 ng/ml polybrene RNA virus recognized by RIG-I in macrophages and cDCs. As for 12 h on days 0 and 2. Virus was produced by PlatE packaging cells shown in Fig. 2A and 2B, the IL-6 and IFN-b mRNA expression transfected with indicated plasmids by using Fugene6 Transfection Re- levels were comparable between WT, AZI2–/–, TBKBP1–/–,and agent (Roche) 3 d before infection. Postinfection, the cells were cultivated in RPMI 1640 medium supplemented with 10% FCS, 10 ng/ml GM-CSF DKO macrophages, indicating that AZI2 and TBKBP1 are dis- until day 6, and 2 mg/ml puromycin was added for the last 2 d before pensable for the IFN responses resulting from IRF3 activation in harvesting floating cells. macrophages. Furthermore, the levels of IL-6 and TNF production in response to various TLR ligands, such as poly(I:C) (TLR3 li- Immunization gand), LPS (TLR4 ligand), and CpG DNA (TLR9 ligand), were 2 2 WT or AZI2 / derived GM-DCs were pulsed with OVA (100 mg/ml; comparable between WT, AZI2–/–, TBKBP1–/–, and DKO mac- Sigma) in the presence of CpG DNA (1 mM) for 18 h. For in vivo transfer rophages (Fig. 2C, 2D). These findings indicate that AZI2 and experiments, mice were injected i.p. with Ag-pulsed GM-DCs (1 3 105 cells). Two weeks after the injection, CD4+ T cells were purified from the TBKBP1 are dispensable for cytokine responses after RLR and spleen and stimulated with OVA (0–1000 mg/ml) in the presence of irra- TLR stimulation. + diated (3000 rad) CD90.2 depleted spleen cells of WT littermates for 72 h. 2/2 2/2 For proliferation assays, cells were pulsed with 2 mCi [3H]thymidine for AZI2 , but not TBKBP1 , GM-DCs showed impaired Downloaded from the last 18 h. Levels of cytokines in the culture supernatant were measured production of cytokines by ELISA. Next, we generated pDCs from mouse BM cells by culture in the Statistics presence of FLT3L. When FLT3L-DCs were infected with NDV, IFN-a production was not impaired in the absence of AZI2 and Statistical significance was calculated with the two-tailed Student t test. TBKBP1 (Fig. 3A). Furthermore, IL-6 production in response to

CpG-DNA stimulation and Flu infection was comparable between http://www.jimmunol.org/ Results AZI2–/–, TBKBP1–/–, and DKO FLT3L-DCs (Fig. 3B). These Generation of mice that lack AZI2 and TBKBP1 findings demonstrate that AZI2 and TBKBP1 are dispensable for To investigate the roles of AZI2 and TBKBP1 in vivo, we generated the production of type I IFNs and proinflammatory cytokines ac- AZI2–/– and TBKBP1–/– mice (Fig. 1A, 1B). Homologous recom- tivated via IRF3 and IRF7 signaling in macrophages and pDCs. bination of the AZI2 and TBKBP1 loci was confirmed by Southern Next, we generated GM-DCs by cultivating BM cells in the blotting (Fig. 1C, 1D). RT-PCR analysis with total RNA isolated presence of GM-CSF. Interestingly, the production of IL-6 and from spleens showed deletion of AZI2 or TBKBP1 expression in TNF in response to stimulation with various TLR ligands, including AZI2–/– and TBKBP1–/– mice (Fig. 1E, 1F). AZI2–/–, TBKBP1–/–, poly(I:C), LPS, and CpG DNA, as well as infection with RNA vi- and AZI2–/–TBKBP1–/– doubly deficient mice were born in a typi- ruses, including NDV, Flu, and EMCV, was severely impaired in the by guest on September 27, 2021 cal Mendelian pattern, appeared to grow normally, and were fer- absence of AZI2, but was unaffected by TBKBP1 deficiency (Fig. tile. In addition, the absence of AZI2 and TBKBP1 did not alter the 3D, 3E). In addition, IFN-a production in response to NDV in- 2 2 differentiation of T and B cells, macrophages, CD11c+B220 CD8 fection was impaired in the absence of AZI2 in GM-DCs (Fig. 3C). cDCs, CD8+ cDCs, and CD11c+B220+ pDCs in the spleen and The cytokine production levels in AZI2–/– and DKO GM-DCs lymph nodes (data not shown). were impaired to similar extents to WT cells, indicating that

FIGURE 1. Generation of AZI2- and TBKBP1-deﬁcient mice. (A and B) Structure of the murine AZI2 (A) and TBKBP1 (B) gene, their targeting construct, and disrupted gene. (C and D) Southern blot analysis of genomic DNA from offspring from the het- erozygote intercrosses in AZI2 (C)- or TBKBP1 (D)-deﬁcient mice. (C) Genomic DNAs were extracted, digested with BamHI, separated by elec- trophoresis, and hybridized with the radiolabeled probe indicated in (A). (D) Genomic DNAs were digested with XhoI and EcoRV, and hybridized with probe indicated in (B). (E and F) RT-PCR analysis of cDNA in AZI22/2 (E)orTBKBP12/2 (F)mice and their controls (WT). The expression of GAPDH gene was analyzed with the same RNA. 4 DC DIFFERENTIATION REQUIRES AZI2 Downloaded from http://www.jimmunol.org/

FIGURE 2. Production of IL-6, TNF, and type I IFN was not altered between WT, AZI22/2, TBKBP12/2, and DKO mouse macrophages. (A–D) Thioglycollate-elicited peritoneal macrophages were collected from WT, AZI22/2, TBKBP12/2, and DKO mice. (A and B) Quantitative PCR analysis showing expression of IL-6 (A) and IFN-b (B) mRNAs in peritoneal macrophages infected with NDV (multiplicity of infection [MOI] 1) for 0, 4, and 8 h. (C and D) Cells were stimulated for 24 h with poly(I:C) (10 or 100 mg/ml), LPS (10 or 100 ng/ml), or CpG DNA (0.1 or 1 mM), and IL-6 (C) or TNF (D) production of supernatants was measured by ELISA. Results are representative of two independent experiments (error bars indicate SD). by guest on September 27, 2021

AZI2, but not TBKBP1, is involved in the production of cytokines mice (Fig. 4B). Then we examined the T cell–stimulating activity of in GM-DCs. GM-DCs are known to recognize RNA virus infec- WT and AZI2–/– FLT3L-DCs. Allogenic CD4+ T cells proliferated tion via RLRs for the production of type I IFNs and proin- comparably after coculture with FLT3L-DCs from WT and AZI2–/– flammatory cytokines, indicating that AZI2 is required for the mice with or without stimulation with LPS and CpG DNA (Fig. cytokine production in response to stimulation of TLRs and RLRs 4C). These findings demonstrate that AZI2 is important for GM- in this cell type. In addition to RNA viruses, IL-6 and TNF pro- DCs, but not pDCs, to confer the ability to activate CD4+ Tcells. duction in response to HSV infection was also significantly re- AZI2 is critical for the differentiation of GM-DCs duced in AZI2–/– and DKO, but not TBKBP1–/–, GM-DCs compared with WT cells (Fig. 3D, 3E). The expression of IL-6 and TNF Because macrophages and GM-DCs transactivate cytokine gene mRNAs in response to LPS and CpG DNA stimulation was im- expression downstream of identical intracellular signaling path- paired in AZI2–/– GM-DCs (Fig. 3F, 3G), indicating that the im- ways, it is unlikely that AZI2 directly controls RLR- and TLR- paired cytokine production was controlled at the gene expression mediated signaling molecules. Thus, we examined the differentia- levels. Collectively, these findings demonstrate that AZI2 controls tion of BMDCs in response to culture in the presence of GM-CSF. the production of proinflammatory cytokines in a cell-type–specific As shown in Fig. 5A, culture of BM cells from WT mice with GM- + manner and is critical for the production of cytokines in GM-DCs. CSF increased the population of CD11c cells in a time-dependent manner, and ∼70% of cells were CD11c+ after 6 d of culture. In T cell–stimulating activity of GM-DCs is impaired in the contrast, CD11c expression was severely impaired in cells that lack absence of AZI2 AZI2 (Fig. 5A). Although WT cells increased during the course of GM-DCs harbor the ability to stimulate T cells, and the activity GM-CSF culture, AZI2–/– cells failed to proliferate (Fig. 5B). In is increased in response to TLR ligand stimulation. To examine contrast, the population of CD11c+ cells in TBKBP1–/– BM cells whether AZI2 is important for the T cell–stimulating activity of cultured with GM-CSF was comparable with that of WT cells (data GM-DCs, we stimulated WT and AZI2–/– GM-DCs with LPS (1 not shown). To investigate whether AZI2 controls GM-DC differ- mg/ml) or CpG DNA (1 mM) for 24 h and then examined them for entiation in a cell-intrinsic manner, we cocultured CD45.1+ WT and their allogeneic T cell–stimulating activities. As shown in Fig. 4A, CD45.2+ AZI2–/– BM cells in the presence of GM-CSF (Fig. 5C). the proliferation of allogeneic CD4+ T cells after coculture with The proportion of CD11c+ DCs generated from AZI2–/– BM cells GM-DCs was severely impaired in the absence of AZI2,irre- was severely decreased, even in the presence of WT cells, indicating spective of LPS and CpG DNA stimulation. The ability of GM- that AZI2 is required for GM-DC differentiation in a cell-autonomous DCs from AZI2–/– mice to support IFN-g production from WT manner. Furthermore, the number of cells expressing the costim- CD4+ T cells was severely impaired compared with cells from WT ulatory markers CD40, CD80, and CD86 on the surface was se- The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. AZI22/2, but not TBKBP12/2, GM-DCs impaired the production of cytokines after stimulation of various TLR ligands and viruses. (A and B)BM cells from WT, AZI22/2, TBKBP12/2, and DKO mice were cultured in the presence of FLT3L for 8 d. (A) The cells were infected with NDV (MOI 1), and the production of IFN-a in the culture supernatant was measured by ELISA. (B) FLT3L-DCs were stimulated with CpG DNA (1 mM) or Flu (PR8) (MOI 1), and the IL-6 production was measured. (C–E)GM-DCsfromWT,AZI22/2, TBKBP12/2, and DKO mice were stimulated with poly(I:C) (100 mg/ml), LPS (1 mg/ml), CpG DNA (1 mM), and infected with EMCV (MOI 1), HSV (MOI 1), NDV (MOI 1), and Flu (PR8) (MOI 1) for 24 h. Production of IFN-a (C), IL-6 (D), and TNF (E) in the culture supernatants was determined by ELISA. (F and G)GM-DCsfromWTandAZI22/2 mice were stimulated with LPS (1 mg/ml) and CpG DNA (1 mM) for 2 and 4 h. The total RNAs were prepared, and the IL-6 and TNF mRNA levels were determined by quantitative PCR. Similar experiments were taken at least three times and one representative experiment is shown (error bars indicate SD). *p , 0.05, **p , 0.01, ***p , 0.001, two-tailed Student t test. verely decreased in AZI2–/– GM-DCs even without TLR ligand contribution of AZI2 to cell-cycle progression in BM cells cultured stimulation, and even after stimulation with TLR ligands, a severe with GM-CSF for 48 h, we found that cells in S and G2/M phases reduction in the expression of these costimulatory molecules was were severely decreased in the absence of AZI2 (Fig. 6A), In ad- observed in the absence of AZI2 (Fig.5D).Next,wesortedCD11c+ dition, incorporation of BrdU as well as [3H]thymidine was im- and CD11c2 cells from WT and AZI2–/– GM-DCs, and examined the paired in AZI2–/– BM cells cultured with GM-CSF (Fig. 6B, 6C). production of IL-6 in response to LPS and CpG DNA stimulation. Collectively, AZI2 is critical for cell-cycle progression of BM cells IL-6 levels were much more abundant in CD11c+ cells compared in response to GM-CSF. with CD11c2 cells, and the production of IL-6 was comparable The GM-CSF receptor consists of a and b subunits (18), and between CD11c+ cells sorted from WT and AZI22/2 GM-DCs (Fig. treatment with GM-CSF initiates the formation of a unique 5E). Thus, impaired cytokine production to TLR ligands in the ab- dodecameric receptor complex that leads to activation of the sence of AZI2 is due to impaired generation of CD11c+ DCs in JAK2/STAT5, ERK1/2, and PI3K/AKT pathways (17). Thus, we AZI22/2 GM-DCs. Collectively, AZI2 deficiency impairs GM-DC assessed whether AZI2 deficiency had any effects on GM-CSF differentiation, causing defects in costimulatory molecule expression receptor signaling. STAT5, ERK, and AKT were phosphorylated in response to various stimuli. Although AZI2 is required for proper to comparable levels in WT and AZI2–/– BM cells after 5–20 min development of DCs in response to GM-CSF, we failed to observe of GM-CSF stimulation (Fig. 6D), indicating that AZI2 is not a defect in lung histology in AZI22/2 mice (data not shown). required for triggering of the initial GM-CSF receptor signaling. We then retrovirally expressed human AZI2 in AZI2–/– BM cells, AZI2 is required for the proliferation of GM-DCs via TBK1 followed by culture in the presence of GM-CSF (Fig. 6F). Ex- Next, we investigated the mechanisms underlying how AZI2 con- pression of full-length AZI2 (1–392) greatly increased the CD11c+ trols GM-CSF–induced DC differentiation. When we examined the cell population (Fig. 6G). AZI2 is composed of N-terminal coiled- 6 DC DIFFERENTIATION REQUIRES AZI2 Downloaded from http://www.jimmunol.org/ FIGURE 4. AZI22/2 GM-DCs show impaired activity to stimulate T cells. (A and B) WT and AZI22/2 GM-DCs were stimulated with LPS (1 mg/ml) and CpG DNA (1 mM) or cultivated without stimulation for 24 h. Then the cells were irradiated (3000 rad) and cocultured with splenic CD4+ T cells isolated from BALB/c mice. (A) Proliferation of CD4+ T cells was determined by [3H]thymidine incorporation 72 h after cultivation. (B) Levels of IFN-g in the culture supernatants (1 3 105 cells/well GM-DCs) were measured by ELISA 72 h after coculture. (C) WT and AZI22/2 FLT3L-DCs were stimulated with LPS (1 mg/ml) and CpG DNA (1 mM) or cultivated without stimulation for 24 h. Then the cells were irradiated (3000 rad) and cocultured with splenic CD4+ T cells isolated from BALB/c mice. Proliferation of CD4+ T cells was determined by [3H]thymidine incorporation 72 h after cultivation. All results are representative of at least three (A and B) or two (C) independent experiments (error bars indicate SD). *p , 0.05, **p , 0.01, ***p , 0.001, two-tailed Student t test. by guest on September 27, 2021 coil domains and a following TBK1-binding domain (Fig. 6E) (32, pulsed with OVA and CpG DNA for 18 h, and the Ag-loaded cells 33). Although expression of the coiled-coil domains of AZI2 (1– were i.p. injected into C57BL/6 mice. At 2 wk after the injection, 158) failed to rescue the defect in DC differentiation in AZI2–/– CD4+ T cells were collected from splenocytes and cocultured with BM cells, expression of AZI2 (1–270) containing the TBK1- OVA and CD90.22 cells to assess the DC vaccination-induced Ag- binding domain greatly increased the proportion of CD11c+ cells specific responses. Mice immunized with AZI2–/– GM-DCs showed (Fig. 6G). Furthermore, expression of AZI2 (1–270) and full-length severely impaired T cell proliferation compared with WT cells (Fig. AZI2 (1–392), but not AZI2 (1–158), in AZI2–/– BM cells cultured 7A). The IFN-g production in response to OVA was also signifi- with GM-CSF increased the production of IL-6 and IL-12p40 in cantly reduced in T cells immunized with AZI2–/– DCs (Fig. 7B). response to LPS and CpG DNA stimulation (Fig. 6H, 6I). In ad- These findings indicate that AZI2 is required for efficient immuni- dition, an AZI2 mutant lacking the TBK1-binding domain (D158– zation and T cell activation by GM-DCs in vivo. 270) also failed to rescue the generation of CD11c+ GM-DCs (Fig. Overexpression of AZI2 enhances the T cell–stimulating 6J). These findings indicate that expression of AZI2 containing the activity of GM-DCs TBK1-binding domain rescued the generation of GM-DCs and their ability to produce cytokines in response to TLR ligands. These The earlier observations prompted us to hypothesize that increased observations suggest that AZI2 controls the generation of GM-DCs expression of AZI2 potentiates the efficiency of GM-DC generation by binding to TBK1, IKK-i, or both. and T cell activation. To evaluate the potential use of AZI2 in Besides acting as IRF3/7 kinases, TBK1 and IKK-i have been modified DC vaccines, we retrovirally overexpressed human AZI2 implicated in cell proliferation and oncogenesis (34, 35). To in- in WT BM cells, followed by DC generation in the presence of GM- vestigate the contribution of TBK1 to GM-DC differentiation, we CSF. Interestingly, overexpression of human AZI2 in BM cells + assessed the effect of human TBK1 overexpression in AZI2–/– BM significantly increased the efficiency of CD11c DC generation cells on the differentiation of GM-DCs. Interestingly, overexpression (Fig. 7C). Furthermore, the levels of IL-6 and IL-12p40 produc- of TBK1 in AZI2–/– BM cells increased the CD11c+ cell population tion in response to LPS and CpG DNA were significantly higher in (Fig. 6K), suggesting that TBK1 acts downstream of AZI2 to fa- cells overexpressing AZI2 (Fig. 7D, 7E). These data demonstrate cilitate the differentiation of GM-DCs. that increased expression of AZI2 facilitates the generation of GM-DCs. Finally, we assessed the ability of AZI2-overexpressing AZI2 is critical for the ability of GM-DCs to stimulate DCs to stimulate T cells in vivo. AZI2-overexpressing GM-DCs Ag-specific T cell activation in vivo were loaded with OVA and stimulated with CpG DNA before To investigate the role of AZI2 in inducing the T cell–stimulating vaccination. The proliferation of CD4+ T cells in response to OVA activity of GM-DCs in vivo, we immunized WT mice with Ag- was significantly enhanced in mice vaccinated with GM-DCs loaded GM-DCs from WT and AZI2–/– mice. GM-DCs were overexpressing AZI2 (Fig. 7F). Thus, AZI2 overexpression in The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/

2/2 2/2 FIGURE 5. Impaired differentiation of GM-DCs in AZI2 mice. (A and B) WT and AZI2 BM cells were cultured with GM-CSF. At days 2, 4, and 6 by guest on September 27, 2021 of culture, cells were harvested and examined by flow cytometry (A) and the total cell numbers were counted (B). (C) BM cells derived from WT CD45.1 mice and WT or AZI22/2 CD45.2 mice were harvested and were cocultured (1:1 ratio) with GM-CSF to generate GM-DCs. At day 6, cells were assessed by flow cytometry. (D) WT and AZI22/2 GM-DCs were stimulated with LPS (1 mg/ml) or CpG DNA (1 mM) for 24 h. The maturation of GM-DCs was assessed by measuring the expression levels of CD40, CD80, and CD86 by flow cytometry. (E)CD11c+ and CD11c2 cells were sorted from WT and AZI22/2 GM-DCs, and stimulated with LPS (1 mg/ml) or CpG DNA (1 mM) for 24 h. Then the production of IL-6 in the culture supernatant was measured by ELISA. Results are representative of over 10 (A), 3 (B and D), or 2 (C and E) independent experiments (error bars indicate SD). **p , 0.01, two-tailed Student t test.

GM-DCs increases the efficiency of immunization for activa- response to TLR ligands and virus infection, the defects appear to tion of CD4+ T cells in vivo. be caused by impaired development of DCs in response to GM- CSF. Therefore, it is likely that AZI2 and TBKBP1 are not directly Discussion involved in the signaling pathways that activate IRF3 and type I In this study, we discovered a novel function of AZI2, but not IFNs. We previously showed that TANK is also dispensable for TBKBP1, in controlling the differentiation of GM-DCs in mice. the production of type I IFNs in response to RNA virus infection In contrast, AZI2 and TBKBP1 are dispensable for the TLR- or and TLR stimulation (26). Therefore, this study clearly demon- RLR-mediated signaling pathways leading to IRF3 activation and strates that a set of proteins possessing a TBK1-binding domain, type I IFN production. Furthermore, the production of cytokines in namely, AZI2, TBKBP1, and TANK, are dispensable for the response to TLR stimulation was unaltered in macrophages and regulation of IRF3 downstream of TBK1. TANK deficiency in pDCs that lack AZI2 and TBKBP1. mice leads to overproduction of cytokines in response to TLR AZI2 (also known as NAP1) comprises an N-terminal coiled-coil stimulation and the development of autoimmune glomerular ne- domain and a central TBK1-binding domain (36). It forms a family phritis (26). In contrast, deficiencies in AZI2 and TBKBP1 did with proteins harboring coiled-coil domains and a TBK1-binding not enhance the proinflammatory cytokine responses in mac- domain, namely, TANK and TBKBP1 (23). AZI2 interacts with rophages or lead to the development of autoimmune diseases TBK1 and IKK-i, and was shown to be involved in IFN-b in- (M. Fukasaka and O. Takeuchi, unpublished observations), sug- duction in response to TLR and RLR stimulation (32, 33). AZI2 gesting that the functions of AZI2 and TBKBP1 in vivo are dis- was also found to be involved in the activation of IRF3 in response tinct from that of TANK, despite the molecular similarity. Our data to poly(I:C) (25). Nevertheless, this study revealed that deficien- clearly show that this set of TBK1-binding proteins is not essential cies in AZI2 and TBKBP1 do not alter the poly(I:C)- and RNA for the activation of IRF3 downstream of RLRs and TLRs. Given virus-mediated responses in macrophages and pDCs. Although that TRAF3 has been shown to interact directly with TBK1 and GM-DCs lacking AZI2 showed impaired cytokine production in IKK-i (37), it is possible that IFN-b promoter stimulator-1/ 8 DC DIFFERENTIATION REQUIRES AZI2 Downloaded from http://www.jimmunol.org/

FIGURE 6. The AZI2-TBK1 pathway is crucial for the generation of GM-DCs by controlling cell-cycle progression. (A–C) BM cells derived from WT 2/2 A and AZI2 mice were cultured with GM-CSF (10 ng/ml) for 48 h. ( ) The cells were then stained with propidium iodide and analyzed by flow cytometry. by guest on September 27, 2021 (B) The cells were pulsed with BrdU for last 24 h of culture, and BrdU+ proliferating cells were assessed by flow cytometry. (C) The cells were pulsed with [3H]thymidine for last 12 h of culture and [3H]thymidine incorporation to the cells was measured. (D) BM cells were incubated with GM-CSF for indicated periods, and cell lysates were subjected to immunoblot analysis using Abs against p-STAT5, p-ERK, and p-AKT, and b actin as a loading control. (E) Schematic representation of human AZI2 domain structure. (F–I) GM-DCs were generated from AZI22/2 BM cells infected with retroviruses expressing human AZI2 (hAZI2) or its deletion mutants (aa residues 1–158, 1–270, or 1–392). (F) Semiquantitative RT-PCR showing expression of hAZI2. GAPDH was amplified as a control. (G) Levels of CD11c expression in GM-DCs expressing indicating constructs were determined by flow cytometry. (H and I) Production of IL-6 (H) and IL-12p40 (I) in cells expressing hAZI2 and its mutants in response to LPS and CpG DNA stimulation was measured by ELISA. (J) GM-DCs were generated from AZI22/2 BM cells infected with retroviruses expressing hAZI2 or a mutant lacking the TBK1 binding domain (D158– 270). (J) Levels of CD11c expression in GM-DCs expressing indicating constructs were determined by flow cytometry. (K) GM-DCs were generated from AZI22/2 BM cells infected with retroviruses expressing human TBK1, and the levels of CD11c expression on the cell surface were analyzed by flow cytometry. Results are representative of two (A, H–J), three (B–D, K), or four (F, G) independent experiments (error bars indicate SD). *p , 0.05, **p , 0.01, ***p , 0.001, two-tailed Student t test.

TRAF3 proteins may directly activate TBK1 for phosphorylation Although GM-CSF has been widely used to generate cDCs from of IRF3. BM cells (15), GM-CSF deficiency does not affect the generation TBK1-deficient mice show embryonic lethality at embryonic day of tissue-resident cDCs in mice (45). In contrast, GM-CSF defi- 15.5 because of liver apoptosis (38). Although IKK-i–deficient mice ciency leads to the development of pulmonary alveolar proteinosis are viable, deficiency in both TBK1 and IKK-i leads to earlier death through macrophage dysfunction. Human pulmonary alveolar during embryonic development (39). In contrast, mice that lack both proteinosis can be explained by depletion of GM-CSF by anti– IRF3 and IRF7 are viable and do not show obvious developmental GM-CSF Abs, suggesting that the functional roles of GM-CSF defects (40, 41). Therefore, it is apparent that the functions of TBK1 in vivo are similar between humans and mice. We found that and IKK-i are not limited to the phosphorylation of IRF proteins AZI2 is dispensable for preventing pulmonary alveolar proteinosis. and subsequent production of type I IFNs. It has been shown that Because the initial activation of signaling pathways in response to TBK1 and IKK-i are critical for oncogenic transformation (35, 42). GM-CSF stimulation was not impaired, even in the absence of Although the mechanism is not yet fully understood, it was reported AZI2, the differentiation of alveolar macrophages seems to be that this occurs through activation of AKT kinase (43, 44). In this controlled independently of AZI2. study, we have shown that AZI2 is critical for the cell cycling of Although GM-CSF is dispensable for the development of splenic BM cells in response to GM-CSF stimulation. Ectopic expression DCs, this growth factor has been used to generate cDCs ex vivo in of TBK1 in AZI22/2 BM cells rescued the defects in the generation mice (46). In addition, GM-CSF is now widely used to generate of GM-DCs, suggesting that AZI2 controls the cell cycling of BM cDCs for human DC therapies that target cancer. We found that cells by activating TBK1 in response to GM-CSF. AZI2 was required for the activation of T cell responses by The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/

FIGURE 7. AZI2 is crucial for T cell activation by GM-DCs in vivo. (A and B) Ag-pulsed GM-DCs require AZI2 to stimulate T cells in vivo. WT and by guest on September 27, 2021 AZI22/2 GM-DCs pulsed with OVA (100 mg/ml) and CpG DNA (1 mM) for 18 h were injected into the peritoneal cavities of C57BL/6 mice. Two weeks after the immunization, splenic CD4+ T cells were collected from the mice and stimulated with indicated concentrations of OVA in the presence of CD90.22 splenocytes for 72 h. (A) CD4+ T cell proliferation was measured by [3H]thymidine incorporation, and (B) levels of IFN-g in supernatants (1000 mg/ml OVA) were determined by ELISA. (C–F) GM-DCs were generated from WT BM cells infected with a retrovirus expressing full-length hAZI2 (aa residues 1–392). (C) Cells expressing CD11c+ were determined by ﬂow cytometry. (D and E) Production of IL-6 (D) and IL-12p40 (E) in AZI2 overexpressing cells in response to stimulation with LPS and CpG DNA was measured by ELISA. (F) C57BL/6 mice were treated with hAZI2-overexpressing GM-DCs pulsed with OVA and CpG DNA. Two weeks later, proliferation of Ag-speciﬁc CD4+ T cells in the spleen was determined by [3H]thymidine incorporation. Results are representative of three (A, B, D, E), four (C), or two (F) independent experiments (error bars indicate SD). *p , 0.05, **p , 0.01, ***p , 0.001, two-tailed Student t test.

transferring GM-DCs. Furthermore, overexpression of AZI2 led A DC vaccine is defined as DCs loaded with Ags, especially to efficient generation of cDCs in response to GM-CSF and en- on a tumor-associated Ag (47, 48). Recently, strategies have been hanced the T cell–stimulating activity of GM-CSF–derived BM cells. developed to generate large-scale production of DCs, and many Thus, it is possible that overexpression of AZI2 in human cells can different protocols have been designed to load Ags onto DCs. The reinforce the efficiency of DC generation required for DC thera- associated findings made it possible to start clinical studies as pies. Because AZI2 deficiency does not affect the development of cancer immunotherapies, but the limited success indicated that it mouse tissues, it is possible that manipulation of AZI2 can en- was necessary to improve and enhance the DC vaccine potency. hance cDC generation without affecting other important cellular For instance, nanoparticle–Ag complexes are efficiently taken up by processes. AZI2 has been identified as a gene induced in response DCs (49). Downregulation of SRA/CD204, a pattern recognition to 5-azacytidine treatment. Therefore, AZI2 expression can be an scavenger receptor, strongly enhances DC-mediated antitumor im- efficient target of therapeutic strategies to enhance DC therapies munity (50). We have shown that increased AZI2 expression in GM- against cancer and other diseases. DCs significantly enhanced the efficacy to activate CD4+ Tcells In contrast with AZI2, lack of TBKBP1 did not cause any defects in vivo. Therefore, it may be possible to enhance the efficiency of in mouse development, type I IFN responses, or DC generation. DC vaccination by manipulating the levels of AZI2 expression, We also found no differences in the differentiation and activation although further studies are required to control AZI2 expression. of immune cells between WT and TBKBP1–/– mice. It is possible In summary, this study has clearly demonstrated that AZI2 is that TBKBP1, AZI2, and TANK act redundantly in unknown bio- critical for the development of DCs in response to GM-CSF stim- logical processes, although AZI2 and TANK have distinct func- ulation. Given that GM-CSF is useful for developing DC therapies, tions. Further studies are required to clarify the functional role of future studies involving manipulation of AZI2 expression will be TBKBP1 in vivo. beneficial for improving the methods for developing DCs ex vivo. 10 DC DIFFERENTIATION REQUIRES AZI2

Acknowledgments is critical for the prevention of autoimmune nephritis. Nat. Immunol. 10: 965– 972. We thank Y. Asahira, E. Kamada and M. Kageyama for secretarial assis- 27. Maruyama, K., T. Kawagoe, T. Kondo, S. Akira, and O. Takeuchi. 2012. TRAF tance and Y. Fujiwara, M. Kumagai, and N. Umano for technical assistance. family member-associated NF-kB activator (TANK) is a negative regulator of osteoclastogenesis and bone formation. J. Biol. Chem. 287: 29114–29124. 28. Satoh, T., O. Takeuchi, A. Vandenbon, K. Yasuda, Y. Tanaka, Y. Kumagai, Disclosures T. Miyake, K. Matsushita, T. Okazaki, T. Saitoh, et al. 2010. The Jmjd3-Irf4 axis The authors have no financial conflicts of interest. regulates M2 macrophage polarization and host responses against helminth infection. Nat. Immunol. 11: 936–944. 29. Kato, H., O. Takeuchi, S. Sato, M. Yoneyama, M. Yamamoto, K. Matsui, S. Uematsu, A. Jung, T. Kawai, K. J. Ishii, et al. 2006. Differential roles of References MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441: 101– 1. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of 105. immunity. Nature 392: 245–252. 30. Saitoh, T., T. Satoh, N. Yamamoto, S. Uematsu, O. Takeuchi, T. Kawai, and 2. Mellman, I., and R. M. Steinman. 2001. Dendritic cells: specialized and regu- S. Akira. 2011. Antiviral protein Viperin promotes Toll-like receptor 7- and Toll- lated antigen processing machines. Cell 106: 255–258. like receptor 9-mediated type I interferon production in plasmacytoid dendritic 3. Kapsenberg, M. L. 2003. Dendritic-cell control of pathogen-driven T-cell po- cells. Immunity 34: 352–363. larization. Nat. Rev. Immunol. 3: 984–993. 31. Fujimoto, K., T. Karuppuchamy, N. Takemura, M. Shimohigoshi, T. Machida, 4. Iwasaki, A., and R. Medzhitov. 2010. Regulation of adaptive immunity by the Y. Haseda, T. Aoshi, K. J. Ishii, S. Akira, and S. Uematsu. 2011. A new subset of innate immune system. Science 327: 291–295. CD103+CD8alpha+ dendritic cells in the small intestine expresses TLR3, TLR7, 5. Yoneyama, M., and T. Fujita. 2009. RNA recognition and signal transduction by and TLR9 and induces Th1 response and CTL activity. J. Immunol. 186: 6287– RIG-I-like receptors. Immunol. Rev. 227: 54–65. 6295. 6. Takeuchi, O., and S. Akira. 2010. Pattern recognition receptors and inflamma- 32. Goncalves, A., T. Bu¨rckstu¨mmer, E. Dixit, R. Scheicher, M. W. Go´rna, tion. Cell 140: 805–820. E. Karayel, C. Sugar, A. Stukalov, T. Berg, R. Kralovics, et al. 2011. Functional 7. Medzhitov, R., and T. Horng. 2009. Transcriptional control of the inflammatory dissection of the TBK1 molecular network. PLoS ONE 6: e23971. response. Nat. Rev. Immunol. 9: 692–703. 33. Sasai, M., M. Shingai, K. Funami, M. Yoneyama, T. Fujita, M. Matsumoto, and Downloaded from 8. Kaisho, T., O. Takeuchi, T. Kawai, K. Hoshino, and S. Akira. 2001. Endotoxin- T. Seya. 2006. NAK-associated protein 1 participates in both the TLR3 and the induced maturation of MyD88-deficient dendritic cells. J. Immunol. 166: 5688– cytoplasmic pathways in type I IFN induction. J. Immunol. 177: 8676–8683. 5694. 34. Chien, Y., S. Kim, R. Bumeister, Y. M. Loo, S. W. Kwon, C. L. Johnson, 9. Pulendran, B., and R. Ahmed. 2006. Translating innate immunity into immu- M. G. Balakireva, Y. Romeo, L. Kopelovich, M. Gale, Jr., et al. 2006. RalB nological memory: implications for vaccine development. Cell 124: 849–863. GTPase-mediated activation of the IkappaB family kinase TBK1 couples innate 10. Geissmann, F., M. G. Manz, S. Jung, M. H. Sieweke, M. Merad, and K. Ley. immune signaling to tumor cell survival. Cell 127: 157–170. 2010. Development of monocytes, macrophages, and dendritic cells. Science 35. Ou, Y. H., M. Torres, R. Ram, E. Formstecher, C. Roland, T. Cheng, R. Brekken,

327: 656–661. R. Wurz, A. Tasker, T. Polverino, et al. 2011. TBK1 directly engages Akt/PKB http://www.jimmunol.org/ 11. Heath, W. R., and F. R. Carbone. 2009. Dendritic cell subsets in primary and survival signaling to support oncogenic transformation. Mol. Cell 41: 458–470. secondary T cell responses at body surfaces. Nat. Immunol. 10: 1237–1244. 36. Fujita, F., Y. Taniguchi, T. Kato, Y. Narita, A. Furuya, T. Ogawa, H. Sakurai, 12. Gilliet, M., W. Cao, and Y. J. Liu. 2008. Plasmacytoid dendritic cells: sensing T. Joh, M. Itoh, M. Delhase, et al. 2003. Identification of NAP1, a regulatory nucleic acids in viral infection and autoimmune diseases. Nat. Rev. Immunol. 8: subunit of IkappaB kinase-related kinases that potentiates NF-kappaB signaling. 594–606. Mol. Cell. Biol. 23: 7780–7793. 13. Hildner, K., B. T. Edelson, W. E. Purtha, M. Diamond, H. Matsushita, 37. Oganesyan, G., S. K. Saha, B. Guo, J. Q. He, A. Shahangian, B. Zarnegar, M. Kohyama, B. Calderon, B. U. Schraml, E. R. Unanue, M. S. Diamond, et al. A. Perry, and G. Cheng. 2006. Critical role of TRAF3 in the Toll-like receptor- 2008. Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in dependent and -independent antiviral response. Nature 439: 208–211. cytotoxic T cell immunity. Science 322: 1097–1100. 38. Bonnard, M., C. Mirtsos, S. Suzuki, K. Graham, J. Huang, M. Ng, A. Itie´, 14. Naik, S. H., A. I. Proietto, N. S. Wilson, A. Dakic, P. Schnorrer, M. Fuchsberger, A. Wakeham, A. Shahinian, W. J. Henzel, et al. 2000. Deficiency of T2K leads to M. H. Lahoud, M. O’Keeffe, Q. X. Shao, W. F. Chen, et al. 2005. Cutting edge: apoptotic liver degeneration and impaired NF-kappaB-dependent gene tran-

generation of splenic CD8+ and CD8- dendritic cell equivalents in Fms-like scription. EMBO J. 19: 4976–4985. by guest on September 27, 2021 tyrosine kinase 3 ligand bone marrow cultures. J. Immunol. 174: 6592–6597. 39. Matsui, K., Y. Kumagai, H. Kato, S. Sato, T. Kawagoe, S. Uematsu, O. Takeuchi, 15. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, and S. Akira. 2006. Cutting edge: Role of TANK-binding kinase 1 and inducible and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from IkappaB kinase in IFN responses against viruses in innate immune cells. J. mouse bone marrow cultures supplemented with granulocyte/macrophage colony- Immunol. 177: 5785–5789. stimulating factor. J. Exp. Med. 176: 1693–1702. 40. Sato, M., H. Suemori, N. Hata, M. Asagiri, K. Ogasawara, K. Nakao, T. Nakaya, 16. Metcalf, D. 2008. Hematopoietic cytokines. Blood 111: 485–491. M. Katsuki, S. Noguchi, N. Tanaka, and T. Taniguchi. 2000. Distinct and es- 17. Guthridge, M. A., F. C. Stomski, D. Thomas, J. M. Woodcock, C. J. Bagley, sential roles of transcription factors IRF-3 and IRF-7 in response to viruses for M. C. Berndt, and A. F. Lopez. 1998. Mechanism of activation of the GM-CSF, IFN-alpha/beta gene induction. Immunity 13: 539–548. IL-3, and IL-5 family of receptors. Stem Cells 16: 301–313. 41. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, 18. Hansen, G., T. R. Hercus, B. J. McClure, F. C. Stomski, M. Dottore, J. Powell, Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master H. Ramshaw, J. M. Woodcock, Y. Xu, M. Guthridge, et al. 2008. The structure of regulator of type-I interferon-dependent immune responses. Nature 434: 772– the GM-CSF receptor complex reveals a distinct mode of cytokine receptor 777. activation. Cell 134: 496–507. 42. Boehm, J. S., J. J. Zhao, J. Yao, S. Y. Kim, R. Firestein, I. F. Dunn, 19. Sakamaki, K., I. Miyajima, T. Kitamura, and A. Miyajima. 1992. Critical cy- S. K. Sjostrom, L. A. Garraway, S. Weremowicz, A. L. Richardson, et al. 2007. toplasmic domains of the common beta subunit of the human GM-CSF, IL-3 and Integrative genomic approaches identify IKBKE as a breast cancer oncogene. IL-5 receptors for growth signal transduction and tyrosine phosphorylation. Cell 129: 1065–1079. EMBO J. 11: 3541–3549. 43. Joung, S. M., Z. Y. Park, S. Rani, O. Takeuchi, S. Akira, and J. Y. Lee. 2011. Akt 20. Sebastiań, C., M. Serra, A. Yeramian, N. Serrat, J. Lloberas, and A. Celada. contributes to activation of the TRIF-dependent signaling pathways of TLRs by 2008. Deacetylase activity is required for STAT5-dependent GM-CSF functional interacting with TANK-binding kinase 1. J. Immunol. 186: 499–507. activity in macrophages and differentiation to dendritic cells. J. Immunol. 180: 44. Xie, X., D. Zhang, B. Zhao, M. K. Lu, M. You, G. Condorelli, C. Y. Wang, and 5898–5906. K. L. Guan. 2011. IkappaB kinase epsilon and TANK-binding kinase 1 activate 21. Tamura, T., P. Tailor, K. Yamaoka, H. J. Kong, H. Tsujimura, J. J. O’Shea, AKT by direct phosphorylation. Proc. Natl. Acad. Sci. USA 108: 6474–6479. H. Singh, and K. Ozato. 2005. IFN regulatory factor-4 and -8 govern dendritic 45. Vremec, D., G. J. Lieschke, A. R. Dunn, L. Robb, D. Metcalf, and K. Shortman. cell subset development and their functional diversity. J. Immunol. 174: 2573– 1997. The influence of granulocyte/macrophage colony-stimulating factor on 2581. dendritic cell levels in mouse lymphoid organs. Eur. J. Immunol. 27: 40–44. 22. Honda, K., A. Takaoka, and T. Taniguchi. 2006. Type I interferon [corrected] 46. Palucka, K., and J. Banchereau. 2012. Cancer immunotherapy via dendritic cells. gene induction by the interferon regulatory factor family of transcription factors. Nat. Rev. Cancer 12: 265–277. Immunity 25: 349–360. 47. Turnis, M. E., and C. M. Rooney. 2010. Enhancement of dendritic cells as 23. Ryzhakov, G., and F. Randow. 2007. SINTBAD, a novel component of innate vaccines for cancer. Immunotherapy 2: 847–862. antiviral immunity, shares a TBK1-binding domain with NAP1 and TANK. 48. Palucka, K., H. Ueno, and J. Banchereau. 2011. Recent developments in cancer EMBO J. 26: 3180–3190. vaccines. J. Immunol. 186: 1325–1331. 24. Guo, B., and G. Cheng. 2007. Modulation of the interferon antiviral response by 49. Cho, N. H., T. C. Cheong, J. H. Min, J. H. Wu, S. J. Lee, D. Kim, J. S. Yang, the TBK1/IKKi adaptor protein TANK. J. Biol. Chem. 282: 11817–11826. S. Kim, Y. K. Kim, and S. Y. Seong. 2011. A multifunctional core-shell nano- 25. Sasai, M., H. Oshiumi, M. Matsumoto, N. Inoue, F. Fujita, M. Nakanishi, and particle for dendritic cell-based cancer immunotherapy. Nat. Nanotechnol. 6: T. Seya. 2005. Cutting edge: NF-kappaB-activating kinase-associated protein 1 675–682. participates in TLR3/Toll-IL-1 homology domain-containing adapter molecule- 50. Yi, H., C. Guo, X. Yu, P. Gao, J. Qian, D. Zuo, M. H. Manjili, P. B. Fisher, 1-mediated IFN regulatory factor 3 activation. J. Immunol. 174: 27–30. J. R. Subjeck, and X. Y. Wang. 2011. Targeting the immunoregulator SRA/ 26. Kawagoe, T., O. Takeuchi, Y. Takabatake, H. Kato, Y. Isaka, T. Tsujimura, and CD204 potentiates specific dendritic cell vaccine-induced T-cell response and S. Akira. 2009. TANK is a negative regulator of Toll-like receptor signaling and antitumor immunity. Cancer Res. 71: 6611–6620.