The Dual Functions of IL-1 -Associated Kinase 2 in TLR9-Mediated IFN and Proinflammatory Production This information is current as of September 26, 2021. Youzhong Wan, Tae Whan Kim, Minjia Yu, Hao Zhou, Michifumi Yamashita, Zizhen Kang, Weiguo Yin, Jian-an Wang, James Thomas, Ganes C. Sen, George R. Stark and Xiaoxia Li

J Immunol 2011; 186:3006-3014; Prepublished online 26 Downloaded from January 2011; doi: 10.4049/jimmunol.1003217 http://www.jimmunol.org/content/186/5/3006 http://www.jimmunol.org/

Supplementary http://www.jimmunol.org/content/suppl/2011/01/26/jimmunol.100321 Material 7.DC1 References This article cites 36 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/186/5/3006.full#ref-list-1 by guest on September 26, 2021 Why The JI? Submit online.

• 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

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 © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Dual Functions of IL-1 Receptor-Associated Kinase 2 in TLR9-Mediated IFN and Proinflammatory Cytokine Production

Youzhong Wan,*,† Tae Whan Kim,‡ Minjia Yu,‡,x Hao Zhou,‡ Michifumi Yamashita,* Zizhen Kang,‡ Weiguo Yin,‡ Jian-an Wang,x James Thomas,{,‖ Ganes C. Sen,* George R. Stark,*,† and Xiaoxia Li‡

Bone marrow-derived plasmacytoid dendritic cells (pDCs) from IL-1R–associated kinase (IRAK)2-deficient mice produced more IFNs than did wild-type pDCs upon stimulation with the TLR9 ligand CpG. Furthermore, in CpG-stimulated IRAK2-deficient pDCs there was increased nuclear translocation of IFN regulatory factor 7, the key transcription factor for IFN transcription in these cells. In IRAK2-deficient macrophages, enhanced NF-kB activation and increased expression of CpG-induced were Downloaded from detected within 2 h after treatment. However, at later times, NF-kB activation was decreased and, in contrast to the results with IFN, there was less secretion of other proinflammatory (such as TNF-a) and chemokines in CpG-stimulated IRAK2- deficient pDCs and macrophages. Therefore, although IRAK2 is a negative regulator of TLR9-mediated IFN production through its modulation of the transcriptional activity of IFN regulatory factor 7, it is also a positive regulator of TLR9-mediated proin- flammatory cytokine and chemokine production at some level subsequent to transcription. The Journal of Immunology, 2011, 186: 3006–3014. http://www.jimmunol.org/

oll-like receptors recognize pathogen-associated molecu- The IRAK1–TRAF6 complex then dissociates from the receptor lar patterns from invading microorganisms, and activate to interact with and activate the downstream kinases transforming T inflammatory responses through the induction of proin- growth factor b-activated kinase 1 (TAK1) and MEK kinase 3 flammatory cytokines and IFNs (1–6). Whereas TLR4 recognizes (17), which then activate NF-kB and MAPK, resulting in the LPS from bacterial cell walls (7), TLRs 1, 2, and 6 detect a variety production of proinflammatory cytokines and chemokines (18). of lipoproteins and peptidoglycans (8–11). TLR5 senses conserved Interestingly, some TLR-induced proinflammatory cytokines, such bacterial flagellin (12), and TLRs 3, 7, 8, and 9, located in the in- a к

as TNF- , in turn activate NF- B through autocrine or paracrine by guest on September 26, 2021 tracellular endosomal-lysosomal compartment, detect nucleic acids pathways, which further promote the additional production of derived from viruses and bacteria (13, 14). In addition, TLR7 and proinflammatory cytokines and chemokines at late times after TLR8 also recognize synthetic imidazoquinoline-like molecules TLR ligand stimulation (19, 20). (13, 14). We and others recently reported that IRAK2 is essential for Upon stimulation with ligands, the adaptor molecule MyD88 is TLR-mediated proinflammatory cytokine and chemokine produc- recruited to all the TLRs, except TLR3 (15). MyD88 interacts in tion (21, 22). IRAK2-deficient macrophages showed normal early, turn with the IL-1R–associated kinases (IRAKs), including but reduced late activation of NF-kB in response to the TLR2 IRAK4 and IRAK1. IRAKs are activated through cross- or auto- ligand Malp-2 and the TLR7 ligand R848, suggesting that IRAK2 phosphorylation in the receptor-proximal complex, which medi- is critical for the late transcriptional response to the activation ates the recruitment of TNFR-associated factor 6 (TRAF6) (16). of TLR2 and TLR7 (21, 22). Importantly, a reduction of cytokine and chemokine mRNA stability and translation was observed in IRAK2-deficient (BM)-derived macrophages in re- *Department of Molecular Genetics, Cleveland Clinic Foundation, Cleveland, OH 44195; †Department of Genetics, Case Western Reserve University, Cleveland, OH sponse to LPS, demonstrating that IRAK2 is also involved in the 44106; ‡Department of Immunology, Cleveland Clinic Foundation, Cleveland, posttranscriptional control of TLR-mediated cytokine and che- OH 44195; xDepartment of Cardiology, Second Affiliated Hospital, School of Med- { mokine production (22). icine, Zhejiang University, Hangzhou, China; Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, TX 75390; and ‖Department of Mo- It is important to note that the activation of TLR-mediated sig- lecular Biology, University of Texas Southwestern Medical Center, Dallas, TX 75390 naling can also lead to the production of type I IFN, including Received for publication September 29, 2010. Accepted for publication December IFN-a and IFN-b (23). Plasmacytoid dendritic cells (pDCs) are the 28, 2010. major cell type that produces type I IFN and regulates antiviral The sequences presented in this article have been submitted to the Gene Expression immunity in vivo (24). pDCs express high levels of TLR7 and Omnibus Web site under accession number GSE24264. TLR9, which detect microbial-derived ssRNA and unmethylated Address correspondence and reprint requests to Dr. Xiaoxia Li, Cleveland Clinic Foundation, Cleveland, OH 44195. E-mail address: [email protected] CpG dinucleotides, respectively. Previous studies showed that The online version of this article contains supplemental material. TLR9 activation in pDCs leads to high levels of type I IFN through Abbreviations used in this article: BM, bone marrow; IRAK, IL-1R–associated ki- the formation of a MyD88–IRAK1–TRAF6 complex (25, 26) and nase; IRF, IFN regulatory factor; MHV68, murine herpesvirus 68; MOI, multiplicity the subsequent activation of IFN regulatory factor 7 (IRF7), of infection; pDC, plasmacytoid ; TAK1, transforming growth factor a master transcription factor for type I IFN expression (27). b-activated kinase 1; TRAF6, TNFR-associated factor 6. We now unexpectedly find that IRAK2 is actually a negative Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 regulator of CpG-induced IFN production at the transcriptional www.jimmunol.org/cgi/doi/10.4049/jimmunol.1003217 The Journal of Immunology 3007 level. IRAK2-deficient pDCs displayed increased nuclear trans- GGCATTAATCTG-39; suppressor of cytokine signaling 3, 59-GCACGGG- location of IRF7 and IFN production in response to stimulation GAGCCCCTTTGTA-39 and 59-CCGGATCCCGGGGAGGTAGT-39; IFN- a 9 9 9 with TLR9 ligands. In contrast, IRAK2 is still required for TLR9- 4, 5 -GACCTGCCTCACACTTATAA-3 and 5 -TCACTCCTTCTCCT- CACTCAG-39; IFN-b,59-CACAGCCCTCTCCATCAACT-39 and 59-TC- mediated posttranscriptional control of proinflammatory cytokine CCACGTCAATCTTTCCTC-39. and chemokine production. CpG-stimulated IRAK2-deficient pDCs and macrophages both produced smaller amounts of proin- Western analysis flammatory cytokines and chemokines than did wild-type cells, Cells were harvested, washed once with PBS, and lysed for 30 min at 4˚C in which coincided with reduced TNF-a secretion and decreased late 1.0% Nonidet P-40, 100 mM Tris hydrochloride (pH 8.0), 20% glycerol, NF-kB activation in IRAK2-deficient macrophages. Taken to- and 0.2 mM EDTA. Cellular debris was removed by centrifugation at 3 gether, the results demonstrate the dual functions of IRAK2 in 10,000 g for 5 min. For immunoblotting, cell extracts were fractionated by NaDodSO4-PAGE and transferred to Immobilon-P transfer membranes TLR9-mediated IFN and proinflammatory cytokine production. (Millipore), using a wet transfer apparatus (Bio-Rad Laboratories). Im- munoblot analysis was performed, and the bands were visualized with HRP-coupled goat anti-rabbit, goat anti-mouse, or donkey anti-goat Ig as Materials and Methods appropriate (Rockland), using the ECL chemiluminescence Western blot- Reagents ting detection system (GE Healthcare).

LPS (Escherichia coli 055:B5) was purchased from Sigma-Aldrich, R848 Murine herpesvirus 68 infection from GLSynthesis, and CpG A (ODN 1585) and CpG B (ODN1826) from InvivoGen. Abs against phospho-ERK, phospho-p38, and phospho-IkBa BM-derived pDCs were plated in 12-well plates at 5 3 105 cells/ml and were purchased from Cell Signaling Technology. Abs against IRAK1, infected with murine herpesvirus 68 (MHV68) (multiplicity of infection IkBa, and b-actin were from Santa Cruz Biotechnology. Abs against [MOI] = 5) for 24 h. IFN, proinflammatory cytokine, and chemokine Downloaded from IRAK2 were from Abcam. Allophycocyanin-conjugated TNF-a Ab, bre- concentrations in the supernatant media were measured by ELISA. To feldin A, and permeabilization buffer were from eBioscience. IRAK2 analyze the proliferation of virus in pDCs, GFP was measured by knockout and control mice were generated, as previously described (22). the Western method. TNF-a knockout and control mice were purchased from Taconic Farms. IL-1R, IL-18R, and IL-33R knockout and control mice were purchased ELISA from The Jackson Laboratory. m

BM-derived pDCs were stimulated with MHV68 (MOI = 5) or CpG A (1 g http://www.jimmunol.org/ BM-derived macrophages and pDCs ml), and supernatant media were collected 24 h after stimulation. BM- derived macrophages were stimulated with LPS (1 mg/ml), R848 (1 mg/ BM cells were obtained from tibias and femurs by flushing with DMEM. m For BM-derived macrophages, the BM cells were cultured in DMEM with ml), or CpG B (1 g/ml) for 24 h, and supernatant media were collected 24 m h after stimulation. IL-6, KC, and TNF-a concentrations were measured by 100 U/ml penicillin, 100 g/ml streptomycin, 20% heat-inactivated FBS, a b and 30% L929 cell-conditioned medium for 5 d for differentiation. For ELISA kits from R&D Systems. IFN- 4 and IFN- concentrations in pDC BM-derived pDCs, the BM cells were plated at 1 3 106 cells/ml in RPMI culture media were measured by using ELISA kits from PBL Biomedical Laboratories. One million cells were used for ELISA. 1640 complete medium (10% FBS, 2 mM L-glutamine, 100 U/ml peni- cillin, 100 mg/ml streptomycin, and 50 mM 2-ME) containing 100 ng/ml Flt3L (PeproTech), and collected for experiments after 6 d. Preparation and fractionation of nuclear extracts

Quantitative real-time PCR BM-derived pDCs were stimulated with CpG A (1 mg ml) for the indicated by guest on September 26, 2021 times. Cell pellets were collected and washed with ice-cold PBS twice, Total RNA was isolated with the TRIzol reagent (Invitrogen). Real-time resuspended in 6 vol buffer A (10 mM HEPES [pH 7.9], 1.5 mM MgCl2, PCR was performed, as described previously (28). The specific primer 10 mM KCl, 0.5 mM DTT, 1 mM PMSF, 0.3 mM Na3VO4, and 5 mM sequences used for b-actin, TNF-a, IL-6, mouse KC, CCL4, CCL5, sup- NaF), and incubated on ice for 10 min. The cell suspension was transferred pressor of cytokine signaling 3, IFN-a4, and IFN-b are as follows: b-actin, to a Dounce homogenizer, and the cells were disrupted with 45 strokes. 59-GGTCATCACTATTGGCAACG-39 and 59-ACGGATGTCAACGTCA- After centrifugation at 500 3 g for 1 min at 4˚C, the supernatants were CACT-39; TNF-a,59-CAAAGGGAGAGTGGTCAGGT-39 and 59-ATTG- collected as cytoplasmic fractions. The nuclear pellets were washed with 1 CACCTCAGGGAAGAGT-39; IL-6, 59-GGACCAAGACCATCCAATTC- ml Nuclei EZ Lysis buffer (Sigma-Aldrich Nuclei EZ Prep ), followed 39 and 59-ACCACAGTGAGGAATGTCCA-39; KC, 59-TAGGGTGAGGA- by washing with 1 ml buffer A. Nuclear pellets were collected by centri- CATGTGTGG-39 and 59-AAATGTCCAAGGGAAGCGT-39; CCL4, 59-C- fugation at 500 3 g for 1 min at 4˚C, resuspended in 50 ml buffer B (20 TCCCACTTCCTGCTGTTTC-39 and 59-GCTCACTGGGGTTAGCACA- mM HEPES [pH 7.9], 500 mM NaCl, 1.5 mM MgCl2, 0.5 mM EDTA, 0.5 39; CCL5, 59-CTGCAATGAAGTGCAGGAGT-39 and 59-GTGTTGGCT- mM DTT, 1 mM PMSF, 0.3 mM Na3VO4, and 5 mM NaF), and incubated

FIGURE 1. IRAK2-deficient pDCs are more resistant to the infection by MHV68. Wild-type and IRAK2-de- ficient BM-derived pDCs were infec- ted with MHV68 (MOI = 5) for 24 h. IFN-a and IFN-b (A) TNF-a and KC (B) concentrations in culture media were measured by ELISA. Results shown are the means and SD of trip- licate determinations. *p , 0.05. C, Cell lysates were analyzed by the Western method with Abs against GFP, IRAK2, and actin. 3008 THE FUNCTION OF IRAK2 IN TLR9 on ice for 30 min. After centrifugation at 14,000 3 g for 10 min at 4˚C, the technology Information (http://www.ncbi.nlm.nih.gov/geo/, accession num- supernatant nuclear fractions were collected. ber GSE24264). Illumina BeadChip microarray analysis Intracellular staining A total of 250 ng RNA was reverse transcribed into cRNA and labeled with biotin-UTP, using the TotalPrep RNA amplification kit (Ambion). cRNA BM-derived macrophages were stimulated with CpG B (1 mg ml) for in- was quantified using a nanodrop spectrophotometer, and the cRNA quality dicated times and fixed with formaldehyde (2% in PBS) on ice for 20 min. (size distribution) was further analyzed in a 1% agarose gel. cRNA was Two hours before fixing, brefeldin A was added to block protein secretion hybridized to the Illumina MouseRef8 v1.1 Expression BeadChip using and enrich intracellular TNF-a. The fixed cells were collected by centri- standard protocols (provided by Illumina). The microarray data were fuge at 1000 3 g for 5 min at 4˚C, washed with PBS twice, and blocked submitted to the Gene Expression Omnibus at National Center for Bio- with PBS containing 2% serum on ice for 30 min. The cell pellets were Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 2. Increased responses to CpG A stimulation in IRAK2-deficient pDCs. Wild-type and IRAK2-deficient BM-derived pDCs were stimulated with CpG A (1 mg/ml) for 24 h. IFN-a and IFN-b (A) TNF-a and KC (B) concentrations in culture media were measured by ELISA. Results shown are the means and SD of triplicate determinations. *p , 0.05. C, Wild-type and IRAK2-deficient BM-derived pDCs were treated with CpG A (1 mg/ml) for the indicated times, and total RNAs (2 mg) were analyzed by real-time PCR. *p , 0.05. D, Nuclear and cytoplasmic fractions prepared from wild-type or IRAK2-deficient pDCs, either untreated or treated with CpG A (1 mg/ml), were analyzed by the Western method with Abs against IRF7, histone H3, and GAPDH. E, Wild-type or IRAK2-deficient pDCs were treated with CpG A (1 mg/ml) for the indicated times, and cell lysates were analyzed by the Western method with Abs against IRAK2, phospho-IKBa, IKBa, phospho-ERK, phospho-JNK, phospho-p38, and actin. The protein levels were analyzed by ImageJ 1.43u and normalized to actin. The Journal of Immunology 3009 collected by centrifuge at 1000 3 g for 5 min, resuspended in perme- Results abilization buffer containing 1% serum, and incubated for 30 min at room IRAK2 is a negative regulator of TLR9-induced IFN production temperature. Allophycocyanin-conjugated TNF-a Ab (1 mg/ml) was added to the cell suspension, followed by incubation for 45 min. The cell pellets in pDCs 3 3 were collected by centrifuge at 1000 g for 5 min, washed with 1 To examine the role of IRAK2 in IFN production, BM-derived permeabilization buffer containing 1% serum twice, and resuspended in PBS. The TNF-a–stained cells were analyzed by flow cytometry by using wild-type and IRAK2-deficient pDCs were infected with a BD FACSCalibur instrument. MHV68 and examined for virus-induced IFN production by Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 3. Increased responses to CpG B stimulation in IRAK2-deficient macrophages. A, Wild-type and IRAK2-deficient BM-derived macrophages were treated with LPS (1 mg/ml), R848 (1 mg/ml), or CpG B (1 mg/ml) for the indicated times. Cell lysates were analyzed by the Western method with Abs against IRAK2, phospho-IKBa, IKBa, phospho-ERK, phospho-p38, and actin. B, Wild-type or IRAK2-deficient BM-derived macrophages were treated with R848 (1 mg/ml) or CpG B (1 mg/ml) for 2 h. A heat map is shown for some genes from Illumina microarray analyses that were induced less by R848, but more by CpG in IRAK2-deficient macrophages. C, Real-time PCR analysis of selected genes from microarray analysis. Wild-type and IRAK2-deficient BM-derived macrophages were treated with R848 or CpG B for the indicated times or were untreated. *p , 0.05. 3010 THE FUNCTION OF IRAK2 IN TLR9

ELISA. Interestingly, IRAK2-deficient pDCs produced more IFN TLR9-mediated IFN production by treating the cells with TLR9 (IFN-a4 and IFN-b; Fig. 1A), but less proinflammatory cytokines ligands. A group of synthetic TLR9 ligands, termed CpG A, are (TNF-a and KC; Fig. 1B), compared with wild-type pDCs in re- potent IFN inducers in pDCs (29). CpG A is characterized by sponse to infection. Consistent with the increased amounts of IFN, phosphodiester backbone CpG motifs and a phosphorothioate- virus-mediated GFP expression (an indication of MHV68 pro- modified poly G stretch at the 59 and 39 ends, and is capable of liferation) was decreased in IRAK2-deficient pDCs compared forming stable higher-order structures. In contrast, conventional with wild-type cells (Fig. 1B). Because MHV68 activates IFN phosphorothioate-modified CpG (CpG B) generally does not form production through TLR9, we examined the role of IRAK2 in higher-order structures (30, 31). Compared to CpG A, CpG B is Downloaded from http://www.jimmunol.org/

FIGURE 4. Impaired activation of NF-kB at late times after CpG B stimulation in IRAK2-deficient macrophages. A, Wild-type and IRAK2-deficient BM-derived macrophages were treated with LPS (1 mg/ml), R848 (1 mg/ml), or CpG B (1 mg/ml) for 24 h. TNF-a, KC, and IL-6 concentrations in by guest on September 26, 2021 culture media were measured by ELISA. Results shown are the means and SD of triplicate deter- minations. *p , 0.05. B, Wild-type and IRAK2- deficient BM-derived macrophages were treated with CpG B (1 mg/ml) or LPS (1 mg/ml) for the indicated times, and total RNAs (2 mg) were ana- lyzed by real-time PCR to determine TNF-a, KC, and IL-6 expression levels. *p , 0.05. C, Wild-type and IRAK2-deficient BM-derived mac- rophages were treated with CpG B (1 mg/ml) for the indicated times. Cell lysates were analyzed by the Western method with Abs against IRAK2, IRAK1, phospho-IKBa, IKBa, and actin. The Journal of Immunology 3011 a much weaker inducer of IFNs in pDCs, but a much stronger IRAK2 is a negative regulator of TLR9-mediated signaling in inducer of proinflammatory cytokines in macrophages and con- macrophages at the transcriptional level at early times after ventional DCs (32). Multimeric CpG A tends to be retained in CpG stimulation early endosomes, whereas monomeric CpG B typically resides in Wild-type and IRAK2-deficient BM-derived macrophages were late endosomes. Although it has been suggested that the endo- stimulated with TLR ligands, followed by analyses for the acti- somal localization of CpGs is critical for their functions, the vation of NF-kB and MAPKs. CpG B (instead of CpG A) was mechanisms for the differential properties of CpG A and CpG B used because it is a more potent activator of TLR9 signaling in are still not clear (33). Consistent with the results of virus in- macrophages. As previously reported (32), whereas IRAK2 de- fection, more IFN was detected in IRAK2-deficient pDCs than ficiency did not have much impact on TLR4- or TLR7-mediated that in wild-type pDCs upon CpG A stimulation (Fig. 2A). Fur- phosphorylation of IkBa at early times, TLR4- and TLR7-induced thermore, IRAK2-deficient pDCs showed markedly increased IFN early p38 and ERK activation were reduced in IRAK2-deficient mRNA and IRF7 nuclear translocation in response to CpG A macrophages compared with wild-type cells, within 30 min after stimulation (Fig. 2C,2D), suggesting that IRAK2 probably in- LPS and R848 stimulation (Fig. 3A). Interestingly, IRAK2- hibits the TLR9-induced transcription of IFN genes. Moreover, deficient macrophages showed higher levels of early phosphory- TLR9-induced phosphorylation of IkBa, p38, ERK, and JNK was lation of IkBa, p38, and ERK than did wild-type macrophages also increased in IRAK2-deficient pDCs compared with wild-type in response to CpG B, within 30 min (Fig. 3A), suggesting that cells, demonstrating increased NF-kB and MAPK signaling (Fig. IRAK2 specifically suppresses TLR9-mediated early signaling in 2E). Taken together, these results indicate that IRAK2 is a nega- macrophages. To further define the differential role of IRAK2 in tive regulator of TLR9-mediated signaling in pDCs, functioning at TLR7- versus TLR9-mediated signaling, we used Illumina micro- Downloaded from a receptor-proximal level to inhibit the activation of all branches arrays to examine the gene expression profiles of wild-type and of downstream signaling. IRAK2-deficient macrophages in response to stimulation with http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 5. Impaired activation of NF-kB at late times after CpG B stimulation in TNF-a–deficient macrophages. A, Wild-type and TNF-a–deficient BM-derived macrophages were treated with LPS (1 mg/ml), R848 (1 mg/ml), or CpG B (1 mg/ml) for 24 h. TNF-a, KC, and IL-6 concentrations were measured by ELISA. Results shown are the means and SD of triplicate determinations. *p , 0.05. B, Wild-type or TNF-a–deficient BM-derived mac- rophages were treated with CpG B (1 mg/ml) for the indicated times, and total RNAs (2 mg) were analyzed by real-time PCR. *p , 0.05. C, Wild-type or TNF-a–deficient BM-derived macrophages were treated with CpG B (1 mg/ml) for the indicated times. Cell lysates were analyzed by the Western method with Abs against IRAK2, IRAK1, phospho-IKBa, IKBa, and actin. 3012 THE FUNCTION OF IRAK2 IN TLR9

R848 and CpG B. A total of 334 genes was induced .2-fold in wild-type macrophages after 2 h of stimulation with CpG B. The majority (213, 63%) of the genes induced by CpG B were ex- pressed at higher levels in IRAK2-deficient macrophages than in wild-type cells. In contrast, the majority of the R848-induced genes (594 of 716; 83%) were expressed at lower levels in IRAK2-deficient macrophages. A large group of genes was in- duced by both R848 and CpG B, most of which showed increased expression in response to CpG B, but decreased expression in response to R848 in IRAK2-deficient macrophages (Fig. 3B). The expression patterns of several genes shown in Fig. 3B were con- firmed by real-time PCR (Fig. 3C). IRAK2 is required for sustained CpG B-initiated NF-kB activation at late times FIGURE 6. Impaired CpG B-mediated TNF-a secretion in IRAK2-de- ficient macrophages. A, Wild-type or IRAK2-deficient BM-derived mac- We next examined the overall effect of IRAK2 on TLR9-mediated rophages were treated with CpG B (1 mg/ml) for indicated times. TNF-a cytokine and chemokine production. Despite the suppressive effect was assayed by intracellular staining, followed by flow cytometry analysis. of IRAK2 in TLR9-mediated gene transcription at early times of The mean fluorescence intensities of TNF-a are shown, which is a repre-

CpG B stimulation (Fig. 3B), IRAK2-deficient macrophages still sentation of three independent experiments. B, The percentages of TNF-a– Downloaded from produced much less TNF-a, KC, and IL-6 than the wild-type positive cells are shown. Results are the means and SD of triplicate macrophages (Fig. 4A). IRAK2-deficient macrophages showed determinations. *p , 0.05. C, Wild-type and IRAK2-deficient BM-derived m decreased KC and IL-6 mRNA expression at late times of LPS macrophages were treated with CpG B (1 g/ml) or were untreated. TNF- a stimulation (Fig. 4B), which is consistent with our previous re- concentrations in the supernatant media were measured by ELISA. Results shown are the means and SD of triplicate determinations. *p , port (22). Interestingly, although IRAK2-deficient macrophages 0.05. a showed increased TNF- and KC mRNA expression at early times http://www.jimmunol.org/ after CpG stimulation (0.5, 1, or 2 h), TNF-a and KC mRNA a expression were dramatically decreased at late times (4 or 8 h; Fig. IRAK2 deficiency reduces the efficiency of TNF- secretion 4B). Moreover, IRAK2-deficient macrophages also showed de- Although previous studies have shown an important role of IRAK2 creased phosphorylation of IkBa at late times (4, 8, or 16 h; Fig. at late times of NF-kB activation in TLR2- and TLR7-dependent 4C), indicating that the activation of NF-kB is greatly decreased. signaling, it was unclear how IRAK2 is involved in sustaining NF- Taken together, these results demonstrate that IRAK2 is essential kB activation for up to 16 h. Many TLR ligand-induced cytokines, for sustaining CpG B-initiated NF-kB activation at late times. such as IL-1 and TNF-a, can potentially activate NF-kB through by guest on September 26, 2021

FIGURE 7. Model of the dual functions of IRAK2 in TLR9-mediated signaling. Upon stimulation of TLR9, unmodified IRAK2 inhibits the activation of downstream signaling, probably by interfering with the activation of IRAK1. Meanwhile, IRAK2 is also required for the secretion of some cytokines, such as TNF-a, which is essential for the activation of NF-kB at late times. The Journal of Immunology 3013 autocrine or paracrine pathways, thereby promoting proin- in wild-type cells (Fig. 4C). Because the modification and deg- flammatory cytokine and chemokine production through positive radation of IRAK1 are essential for activation of TLR9-mediated feedback loops. Especially cytokines of the IL-1 family, including signaling (34), it is conceivable that IRAK2 might compete with IL-1, IL-18, and IL-33, activate NF-kB through MyD88- IRAK1 for an upstream kinase, such as IRAK4, and inhibit the dependent pathways that use IRAKs. We thus investigated the activation of IRAK1. It is important to point out that the inhibitory possibility that the reduced cytokine and chemokine production effect of IRAK2 is specific for TLR9-mediated signaling. In- in IRAK2-deficient macrophages is due to decreased responses to terestingly, whereas IRAK2 was modified in response to several the IL-1 family members. However, deficiencies in the IL-1R, IL- TLR ligands, including LPS and R848, IRAK2 modification was 18R, or IL-33R had little impact on cytokine and chemokine not detected at early times after stimulation with TLR9 ligands production in response to LPS, R848, or CpG B, suggesting that (Figs. 2D,3A). In addition, we only observed ligand-induced in- these cytokines are not essential for regulating TLR ligand-in- teraction of IRAK2 with MyD88 and TRAF6 in response to R848 duced cytokine and chemokine production (Supplemental Fig. 1). stimulation, but not in response to CpG B stimulation (Supple- In contrast, previous studies demonstrated that TNF-a is in- mental Fig. 2). Further study is required to map the modification volved in TLR-mediated NF-kB activation at late times through sites of IRAK2 and to show how such modification affects its a positive feedback mechanism (19, 20). To investigate the in- functions. volvement of TNF-a in CpG B-mediated responses, we made Many TLR-induced proinflammatory cytokines function not BM-derived macrophages from TNF-a–deficient and wild-type only as effectors in immune responses, but also as amplifiers of NF- mice. Compared with wild-type macrophages, TNF-a–deficient kB signaling, inducing proinflammatory cytokine and chemokine macrophages indeed produced less IL-6 and KC upon CpG B production through positive feedback mechanisms (19, 20). Our Downloaded from stimulation (Fig. 5A), indicating that TNF-a is involved in CpG B- study demonstrates the importance of the production of a second induced cytokine and chemokine production. Specifically, TNF-a wave of the cytokine TNF-a for sustaining CpG-initiated NF-kB deficiency resulted in decreased KC and IL-6 mRNA expression activation. Consistent with the critical role of IRAK2 in TLR4- and IkBa phosphorylation at late times (Fig. 5B,5C), indicating induced cytokine production at posttranscriptional levels (22), our that TNF-a participates in a second wave of CpG-induced studies show that IRAK2 deficiency affected TLR9-induced TNF-

proinflammatory cytokine and chemokine production. We then a production through reduced TNF-a secretion. TNF-a is pro- http://www.jimmunol.org/ carefully examined TNF-a expression and production in wild-type duced as a membrane-bound form, and its secretion requires the and IRAK2-deficient macrophages. Importantly, consistent with activation of TNF-a–converting enzyme, which cleaves TNF-a to the negative regulatory role of IRAK2 in TLR9 signaling, IRAK2- release it (35). The activation of this enzyme is suggested to be deficient macrophages actually showed a higher percentage of mediated by activation of the MAPKs ERK and p38 (36). How- TNF-a–positive cells than did wild-type macrophages in response ever, the activation of ERK and p38 is not impaired in CpG- to CpG B (Fig. 6B). However, the amounts of secreted TNF-a stimulated IRAK2-deficient pDCs or macrophages (Figs. 2D, were lower in the supernatant media from IRAK2-deficient mac- 3A), suggesting that IRAK2 may mediate TNF-a secretion rophages than in medium from wild-type macrophages despite through a different mechanism. the fact that wild-type and IRAK2-deficient TNF-a–positive cells Although the innate immune response is crucial to suppress by guest on September 26, 2021 showed similar fluorescence intensities for TNF-a staining (Fig. infection, aberrant immune responses to host Ags are frequently 6A,6C). These results suggest that IRAK2 deficiency probably observed in many autoimmune diseases, such as systemic lupus has an impact on the efficiency of TNF-a secretion. erythematosus. Therefore, it is vital to keep immune responses at appropriate levels. Our study has revealed a new mechanism that Discussion regulates TLR9-mediated immune responses, in which IRAK2 We report two novel functions of IRAK2 in TLR9-mediated sig- plays multiple roles. IRAK2 suppresses TLR9-mediated signaling naling. Although IRAK2 is a negative regulator of TLR9-mediated at early times after stimulation, thus increasing the threshold of transcription, it is also involved in the regulation of cytokine pro- immune responses induced by TLR9 ligands, and therefore po- duction at the posttranscriptional level (Fig. 7). IFN production tentially preventing overreaction to low-level interactions between seems mainly to be affected by the negative effect of IRAK2 at the TLR9 and host DNA. However, when TLR9-mediated proinflam- transcriptional level, and therefore IRAK2-deficient pDCs produce matory cytokines are induced, IRAK2 promotes their secretion and more IFN upon viral infection or CpG A stimulation (Figs. 1A, increases cytokine production through positive feedback mech- 2A). For proinflammatory cytokine and chemokine production in anisms, leading to an efficient immune response. macrophages, we observed both increased CpG B-induced tran- scription (Fig. 3B) and reduced TNF-a secretion in IRAK2- Disclosures deficient macrophages at early times (Fig. 6). Decreased proin- The authors have no financial conflicts of interest. flammatory cytokine and chemokine production in IRAK2- deficient macrophages 24 h after CpG B stimulation (Fig. 4A) demonstrates that the effect of reduced TNF-a secretion is dom- References inant at late times. Interestingly, IRAK2-deficient pDCs also pro- 1. Medzhitov, R., P. Preston-Hurlburt, and C. A. Janeway, Jr. 1997. A human ho- mologue of the Drosophila Toll protein signals activation of adaptive immunity. duced less proinflammatory cytokines and chemokines than did Nature 388: 394–397. wild-type pDCs after virus infection or CpG A stimulation (Figs. 2. Rock, F. L., G. Hardiman, J. C. Timans, R. A. Kastelein, and J. F. Bazan. 1998. A family of human receptors structurally related to Drosophila Toll. Proc. Natl. 1B,2B), suggesting that IRAK2 might play a similar role in pDCs Acad. Sci. USA 95: 588–593. in regulating cytokine and chemokine production. 3. Takeuchi, O., T. Kawai, H. Sanjo, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, IRAK2-deficient pDCs and macrophages show increased acti- K. Takeda, and S. Akira. 1999. TLR6: a novel member of an expanding Toll-like k receptor family. Gene 231: 59–65. vation of both NF- B and MAPK signaling cascades at early 4. Chuang, T. H., and R. J. Ulevitch. 2000. Cloning and characterization of a sub- times after activation of TLR9-mediated signaling (Figs. 2B,3A), family of human Toll-like receptors: hTLR7, hTLR8 and hTLR9. Eur. Cytokine demonstrating that the inhibitory effect of IRAK2 is at the Netw. 11: 372–378. 5. Hemmi, H., O. Takeuchi, T. Kawai, T. Kaisho, S. Sato, H. Sanjo, M. Matsumoto, receptor-proximal level. It is intriguing that IRAK1 is degraded K. Hoshino, H. Wagner, K. Takeda, and S. Akira. 2000. A Toll-like receptor faster and more completely in IRAK2-deficient macrophages than recognizes bacterial DNA. Nature 408: 740–745. 3014 THE FUNCTION OF IRAK2 IN TLR9

6. Zhang, D., G. Zhang, M. S. Hayden, M. B. Greenblatt, C. Bussey, R. A. Flavell, associated kinase 2 is critical for lipopolysaccharide-mediated post- and S. Ghosh. 2004. A Toll-like receptor that prevents infection by uropatho- transcriptional control. J. Biol. Chem. 284: 10367–10375. genic bacteria. Science 303: 1522–1526. 23. Honda, K., A. Takaoka, and T. Taniguchi. 2006. Type I interferon [corrected] 7. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, gene induction by the interferon regulatory factor family of transcription factors. E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ Immunity 25: 349–360. and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282: 2085–2088. 24. Gilliet, M., W. Cao, and Y. J. Liu. 2008. Plasmacytoid dendritic cells: sensing 8. Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda, nucleic acids in viral infection and autoimmune diseases. Nat. Rev. Immunol. 8: and S. Akira. 1999. Differential roles of TLR2 and TLR4 in recognition of 594–606. Gram-negative and Gram-positive bacterial cell wall components. Immunity 11: 25. Kawai, T., S. Sato, K. J. Ishii, C. Coban, H. Hemmi, M. Yamamoto, K. Terai, 443–451. M. Matsuda, J. Inoue, S. Uematsu, et al. 2004. Interferon-alpha induction 9. Takeuchi, O., K. Hoshino, and S. Akira. 2000. Cutting edge: TLR2-deficient and through Toll-like receptors involves a direct interaction of IRF7 with MyD88 and MyD88-deficient mice are highly susceptible to Staphylococcus aureus in- TRAF6. Nat. Immunol. 5: 1061–1068. fection. J. Immunol. 165: 5392–5396. 26. Uematsu, S., S. Sato, M. Yamamoto, T. Hirotani, H. Kato, F. Takeshita, 10. Underhill, D. M., A. Ozinsky, A. M. Hajjar, A. Stevens, C. B. Wilson, M. Matsuda, C. Coban, K. J. Ishii, T. Kawai, et al. 2005. Interleukin-1 receptor- M. Bassetti, and A. Aderem. 1999. The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401: associated kinase-1 plays an essential role for Toll-like receptor (TLR)7- and 811–815. TLR9-mediated interferon-alpha induction. J. Exp. Med. 201: 915–923. 11. Underhill, D. M., A. Ozinsky, K. D. Smith, and A. Aderem. 1999. Toll-like 27. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, receptor-2 mediates mycobacteria-induced proinflammatory signaling in mac- Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master rophages. Proc. Natl. Acad. Sci. USA 96: 14459–14463. regulator of type-I interferon-dependent immune responses. Nature 434: 772– 12. Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R. Goodlett, 777. J. K. Eng, S. Akira, D. M. Underhill, and A. Aderem. 2001. The innate immune 28. Kim, T. W., K. Staschke, K. Bulek, J. Yao, K. Peters, K. H. Oh, Y. Vandenburg, response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410: H. Xiao, W. Qian, T. Hamilton, et al. 2007. A critical role for IRAK4 kinase 1099–1103. activity in Toll-like receptor-mediated innate immunity. J. Exp. Med. 204: 1025– 13. Heil, F., H. Hemmi, H. Hochrein, F. Ampenberger, C. Kirschning, S. Akira, 1036. G. Lipford, H. Wagner, and S. Bauer. 2004. Species-specific recognition of 29. Verthelyi, D., K. J. Ishii, M. Gursel, F. Takeshita, and D. M. Klinman. 2001. Downloaded from single-stranded RNA via Toll-like receptor 7 and 8. Science 303: 1526–1529. Human peripheral blood cells differentially recognize and respond to two dis- 14. Diebold, S. S., T. Kaisho, H. Hemmi, S. Akira, and C. Reis e Sousa. 2004. Innate tinct CPG motifs. J. Immunol. 166: 2372–2377. antiviral responses by means of TLR7-mediated recognition of single-stranded 30. Krieg, A. M. 2006. Therapeutic potential of Toll-like receptor 9 activation. Nat. RNA. Science 303: 1529–1531. Rev. Drug Discov. 5: 471–484. 15. Akira, S., K. Takeda, and T. Kaisho. 2001. Toll-like receptors: critical 31. Ishii, K. J., and S. Akira. 2006. Innate immune recognition of, and regulation by, linking innate and acquired immunity. Nat. Immunol. 2: 675–680. DNA. Trends Immunol. 27: 525–532. 16. Cao, Z., J. Xiong, M. Takeuchi, T. Kurama, and D. V. Goeddel. 1996. TRAF6 is 32. Hemmi, H., T. Kaisho, K. Takeda, and S. Akira. 2003. The roles of Toll-like

a signal transducer for interleukin-1. Nature 383: 443–446. receptor 9, MyD88, and DNA-dependent protein kinase catalytic subunit in the http://www.jimmunol.org/ 17. Deng, L., C. Wang, E. Spencer, L. Yang, A. Braun, J. You, C. Slaughter, effects of two distinct CpG DNAs on dendritic cell subsets. J. Immunol. 170: C. Pickart, and Z. J. Chen. 2000. Activation of the IkappaB kinase complex by 3059–3064. TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and a unique 33. Guiducci, C., G. Ott, J. H. Chan, E. Damon, C. Calacsan, T. Matray, K. D. Lee, Cell polyubiquitin chain. 103: 351–361. R. L. Coffman, and F. J. Barrat. 2006. Properties regulating the nature of the 18. Ninomiya-Tsuji, J., K. Kishimoto, A. Hiyama, J. Inoue, Z. Cao, and plasmacytoid dendritic cell response to Toll-like receptor 9 activation. J. Exp. K. Matsumoto. 1999. The kinase TAK1 can activate the NIK-I kappaB as well as Med. 203: 1999–2008. the MAP kinase cascade in the IL-1 signalling pathway. Nature 398: 252–256. 34. Xiao, H., W. Qian, K. Staschke, Y. Qian, G. Cui, L. Deng, M. Ehsani, X. Wang, 19. Covert, M. W., T. H. Leung, J. E. Gaston, and D. Baltimore. 2005. Achieving stability of lipopolysaccharide-induced NF-kappaB activation. Science 309: Y. W. Qian, Z. J. Chen, et al. 2008. Pellino 3b negatively regulates interleukin-1- 1854–1857. induced TAK1-dependent NF kappaB activation. J. Biol. Chem. 283: 14654– 20. Werner, S. L., D. Barken, and A. Hoffmann. 2005. Stimulus specificity of gene 14664. 35. Black, R. A., C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack, expression programs determined by temporal control of IKK activity. Science by guest on September 26, 2021 309: 1857–1861. M. F. Wolfson, B. J. Castner, K. L. Stocking, P. Reddy, S. Srinivasan, et al. 1997. 21. Kawagoe, T., S. Sato, K. Matsushita, H. Kato, K. Matsui, Y. Kumagai, T. Saitoh, A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from T. Kawai, O. Takeuchi, and S. Akira. 2008. Sequential control of Toll-like cells. Nature 385: 729–733. receptor-dependent responses by IRAK1 and IRAK2. Nat. Immunol. 9: 684–691. 36. Fan, H., and R. Derynck. 1999. Ectodomain shedding of TGF-alpha and other 22. Wan, Y., H. Xiao, J. Affolter, T. W. Kim, K. Bulek, S. Chaudhuri, D. Carlson, transmembrane proteins is induced by receptor tyrosine kinase activation and T. Hamilton, B. Mazumder, G. R. Stark, et al. 2009. Interleukin-1 receptor- MAP kinase signaling cascades. EMBO J. 18: 6962–6972.