Complete Dependence on IRAK4 Kinase Activity in TLR2, but Not TLR4, Signaling Pathways Underlies Decreased Cytokine Production and Increased Susceptibility to This information is current as Streptococcus pneumoniae Infection in of October 2, 2021. IRAK4 Kinase−Inactive Mice Meghan E. Pennini, Darren J. Perkins, Andres M. Salazar, Michael Lipsky and Stefanie N. Vogel

J Immunol 2013; 190:307-316; Prepublished online 3 Downloaded from December 2012; doi: 10.4049/jimmunol.1201644 http://www.jimmunol.org/content/190/1/307 http://www.jimmunol.org/

References This article cites 40 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/190/1/307.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on October 2, 2021

• 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 © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Complete Dependence on IRAK4 Kinase Activity in TLR2, but Not TLR4, Signaling Pathways Underlies Decreased Cytokine Production and Increased Susceptibility to Streptococcus pneumoniae Infection in IRAK4 Kinase–Inactive Mice

Meghan E. Pennini,* Darren J. Perkins,* Andres M. Salazar,† Michael Lipsky,‡ and Stefanie N. Vogel*

IRAK4 is critical for MyD88-dependent TLR signaling, and patients with Irak4 mutations are extremely susceptible to recurrent KDKI bacterial infections. In these studies, mice homozygous for a mutant IRAK4 that lacks kinase activity (IRAK4 ) were used to Downloaded from address the role of IRAK4 in response to TLR agonists or bacterial infection. IRAK4KDKI macrophages exhibited diminished responsiveness to the TLR4 agonist LPS and little to no response to the TLR2 agonist Pam3Cys compared with wild-type macrophages as measured by cytokine mRNA, cytokine protein expression, and MAPK activation. Importantly, we identified two kinases downstream of the MAPKs, MNK1 and MSK1, whose phosphorylation is deficient in IRAK4KDKI macrophages stimulated through either TLR2 or TLR4, suggesting that IRAK4 contributes to TLR signaling beyond the initial phosphorylation KDKI of MAPKs. Additionally, IRAK4 macrophages produced minimal cytokine mRNA expression in response to the Gram- http://www.jimmunol.org/ positive bacteria Streptococcus pneumoniae and Staphylococcus aureus compared with WT cells, and IRAK4KDKI mice exhibited increased susceptibility and decreased cytokine production in vivo upon S. pneumoniae infection. Treatment of infected mice with a complex of polyinosinic-polycytidylic acid with poly-L-lysine and carboxymethyl cellulose (Hiltonol), a potent TLR3 agonist, significantly improved survival of both WT and IRAK4KDKI mice, thereby providing a potential treatment strategy in both normal and immunocompromised patients. The Journal of Immunology, 2013, 190: 307–316.

major component of the mammalian immune system been identified in genes that encode TLR pathway components (3–7).

relies on the recognition of certain microbial compo- by guest on October 2, 2021 A nents, which are absent in host cells, called pathogen- In humans, 10 TLRs have been identified. TLR2 heterodimerizes associated molecular patterns (PAMPs) (1, 2). Host cells recog- with TLR1 or TLR6, leading to the recognition of Gram-positive nize PAMPs through pattern recognition receptors that initiate bacterial components, such as lipoproteins (8). TLR4, the first a signal cascade that results in the upregulation of proinflam- human TLR cloned (9, 10), is engaged by LPS found on Gram- matory cytokines and the clearance of invading microbes. Of the negative bacteria, whereas TLR3, TLR5, TLR7, TLR8, and TLR9 pattern recognition receptors, TLRs play a critical role in con- recognize other bacterial and/or viral components, such as dsRNA trolling microbial infection in both mice and humans. Mutations (TLR3), flagellin (TLR5), ssRNA (TLR7 and TLR8), and CpG that occur in the TLR signaling pathway result in increased DNA (TLR9) (1). All TLRs activate NF-kB and MAPKs, but the susceptibility to viral, bacterial, and/or fungal infections, de- intermediate signaling molecules used by a particular TLR can pending on the specific signaling component affected. In the past vary. Although some proteins are common to all TLR signaling decade, an increasing number of innate immune deficiencies has pathways, others are unique to a particular TLR or are shared among only a subset of the TLRs. For example, cells deficient in the adapter protein MyD88 are completely refractory to signaling *Department of Microbiology and Immunology, School of Medicine, University through nearly all TLRs, with the notable exceptions of TLR3 of Maryland, Baltimore, MD 21201; †Oncovir, Inc., Washington, DC 2000; and and TLR4 (11, 12). TLR3 is completely independent of MyD88, ‡Department of Pathology, School of Medicine, University of Maryland, Baltimore, MD 21201 whereas TLR4 has two signaling pathways: one is MyD88 de- pendent, and the other is MyD88 independent. Upon TLR en- Received for publication June 14, 2012. Accepted for publication November 1, 2012. gagement, MyD88 is recruited to the cytosolic TIR domain of a This work was supported by National Institutes of Health Grant AI18797 (to S.N.V.). TLR and facilitates recruitment of the IRAK kinases, IRAK4, Address correspondence and reprint requests to Prof. Stefanie N. Vogel, Department of Microbiology and Immunology, School of Medicine, University of Maryland, 685 IRAK1, and IRAK2. Each of these proteins contributes to the West Baltimore Street, HSF1 Suite 380, Baltimore, MD 21201. E-mail address: assembly of a multiprotein structure, called the “myddosome,” [email protected] which is critical for TLR-mediated signal transduction (13). Abbreviations used in this article: HKPA, heat-killed Pseudomonas aeruginosa; IRAK1 and IRAK2 are thought to have redundant roles in myd- HKSA, heat-killed Staphylococcus aureus; i.n., intranasally; ODN, oligodeoxynu- cleotide; PAMP, pathogen-associated molecular pattern; poly I:C, polyinosinic- dosome formation and signal transduction, but the unique con- polycytidylic acid; poly IC:LC, a complex of polyinosinic-polycytidylic acid with tributions, if any, of IRAK2 have not been thoroughly addressed poly-L-lysine and carboxymethyl cellulose; qPCR, quantitative real-time PCR; WT, (14). Once IRAK4 complexes with MyD88, it phosphorylates wild-type. IRAK1, allowing IRAK1 to autophosphorylate and recruit the Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 ubiquitin ligase TRAF6. TRAF6 ubiquitinates both itself and www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201644 308 DIFFERENTIAL NECESSITY OF IRAK4 KINASE FOR TLR SIGNALING

IRAK1, enabling interaction of the complex with TAK1, the ki- C57BL/6J mice were purchased from The Jackson Laboratory (Bar Har- nase responsible for IkB phosphorylation and degradation, and bor, ME). Peritoneal macrophages were obtained by performing a perito- leading to the translocation of activated NF-kB to the nucleus. neal lavage 5 d after i.p. administration of sterile thioglycollate (Remel, Lenexa, KS). For bone marrow–derived macrophages, bone marrow was Another adapter protein, TIRAP (also called Mal), facilitates cultured in standard medium supplemented with 20–25% LADMAC cell– MyD88recruitmenttoTLR2.LikeTLR2,TLR4usesTIRAP/ conditioned medium (24). After 6 d, nonadherent cells were removed. MyD88 to initiate signaling through the MyD88-dependent path- Adherent cells were harvested using 0.25% Trypsin-EDTA (Life Tech- way; however, TLR4 also uses another “bridging adapter,” TRAM, nologies, Grand Island, NY) after 7–14 d in culture and replated for ex- perimental use in the absence of a CSF-1 source. All animal experiments to recruit TRIF to the receptor complex and initiate TRIF-depen- were conducted under the guidelines and approval of the Institutional dent signaling. TLR3 signals exclusively through the adapter Animal Care and Use Committee. All cellular experiments were conducted TRIF. using DMEM (Life Technologies) supplemented with 5% FCS, 2 mM L- Because of its role in propagating MyD88-dependent signaling, glutamine, and 1% penicillin–streptomycin (penicillin–streptomycin was IRAK4 is presumably a critical component in most TLR-signaling omitted in experiments using live or heat-killed bacteria) at 37˚C and 5% CO . pathways, with the exception of TLR3 (15–18). Mice deficient 2 2 2 Agonists for TLR2 (Pam3Cys) and TLR3 (polyinosinic-polycytidylic in IRAK4 (IRAK4 / ) show characteristics reminiscent of acid [poly I:C]) were purchased from InvivoGen (San Diego, CA). Poly MyD882/2 mice in that they are resistant to doses of certain TLR IC:LC (Hiltonol) was generously provided by Dr. Andres Salazar (Oncovir). agonists, such as LPS or CpG, which are lethal to wild-type (WT) Protein-free, phenol water–extracted Escherichia coli K235 LPS was prepared as described elsewhere (25). mice (19, 20). However, there are discrepancies in the literature with regard to whether the kinase activity of IRAK4 is required for Quantitative real-time PCR TLR signaling (19–23). Interestingly, cytokine expression induced 3 6

Macrophages were cultured in 12-well plates at a density of 2 10 cells/ Downloaded from by LPS in macrophages expressing a mutant IRAK4 that lacks well. After stimulation, RNA was extracted using Tripure reagent (Roche, kinase activity was diminished but not to the same extent as Indianapolis, IN), and 1 mg of total RNA was reversed transcribed (RT- IRAK42/2 cells (20). These data suggest that IRAK4 plays a role PCR) using iScript (Bio-Rad, Hercules, CA), according to the manu- facturer’s instructions. The resulting cDNA template was used for real- in propagating TLR signaling independent of its ability to phos- time PCR analysis with Power SYBR Green and the AB 7900HT phorylate IRAK1. In contrast, a separate study demonstrated Fast Time fluorescence detection system. The following primers were 2 2 diminished gene expression in both IRAK4 / and IRAK4 ki- used: IL-1b sense, 59-AAATACCTGTGGCCTTGGGC-39 and antisense, nase–inactive cells, thus demonstrating a need for further explo- 59-CTTGGGATCCACACTCTCCAG-39;IFN-b sense, 59-TTCTCCACCA- http://www.jimmunol.org/ 9 9 9 ration into the necessity of IRAK4 kinase activity in the propagation CAGCCCTCTC-3 and antisense, 5 -CTTTCCATTCAGCTGCTCCAG-3 ; IL-12p40 sense, 59-TCTTTGTTCGAATCCAGCGC-39 and antisense, 59- of TLR signaling (19). GGAACGCACCTTTCTGGTTACA-39; HPRT sense, 59-GCTGACCTGC- Patients with mutations in IRAK4 present with recurrent bac- TGATTACATTAA-39 and antisense, 59-TGATCATTACAGTAGCTCTTC- terial infections but show no impaired defense against viral AGTCTGA-39;KCsense,59-CCATGGCTGGGATTCACC-39 and antisense, 9 9 9 infections (presumably due to their retained ability to signal 5 -GACCATTCTTGAGTGTGGCTATGAC-3 ; IL-10 sense, 5 -ATTTGAA- TTCCCTGGGTGAGAAG-39 and antisense, 59-CACAGGGGAGAAATCG- through TLR3 and other non-TLR viral receptors). Streptococcus ATGACA-39;COX2sense,59-ACTGGGCCATGGAGTGGAC-39 and pneumoniae is the most common infection found in these patients, antisense, 59-AATGACCTGATATTTCAATTTTCCATC-39; PAI2 sense, followed by Staphylococcus aureus and Pseudomonas aeruginosa 59-TACAGGCACAAGCAGGAGATAAAA-39 and antisense, 59-AGCT- by guest on October 2, 2021 (4, 6). Although other investigators observed impaired gene ex- GAGAGAGGAGAAGGCTGA-39; IL-6 sense, 59-TGTCTATACCACTT- 9 9 pression in IRAK4-deficient cells stimulated with synthetic TLR CACAAGTCGGAG-3 and antisense, 5 -GCACAACTCTTTTCTCAT- TTCCAC-39; IP-10 sense, 59-GTGTTGAGATCATTGCCACGA-39 and agonists, there is a paucity of data to address how this deficiency antisense, 59-TTTTTGGCTAAACGCTTTCATTAA-39;andS. pneumoniae fares against infection with whole bacteria. In the studies pre- 16S sense, 59-GAGTCGCAAGCCGGTGACGG-39 and antisense, 59- sented in this article, we sought to clarify the literature with regard CGTATTCACCGCGGCGTGCT-39. Relative gene expression was cal- D to impaired TLR signaling in IRAK4KDKI macrophages and to culated using the Ct calculation method, in which all samples were normalized to the housekeeping gene HPRT (26). develop a mouse model of infection to address the importance of IRAK4 kinase activity in controlling bacterial infection. We found Cytokine protein measurements that TLR4 signaling is partially deficient in macrophages derived Macrophages were cultured in 12-well plates at a density of 2 3 106 cells/ KDKI from IRAK4 mice, whereas TLR2 signaling appears to be well. Six hours after stimulation, supernatants were collected and sent to fully dependent upon IRAK4 kinase activity for signaling. Mac- the University of Maryland Cytokine Core Laboratory (Baltimore, MD) for rophage responses were more impaired when exposed to bacteria cytokine analysis using the Luminex Multianalyte System (R&D Systems, Minneapolis, MN). For in vivo measurements, serum samples were col- known to be potent TLR2 simulators, whereas bacteria, such as lected from mice 48 h postinfection and processed as stated above. P. aeruginosa, behaved more like a TLR4 agonist in that signaling was only partially impaired. Finally, differential susceptibility to Cell lysate preparation and Western blot analysis KDKI S. pneumoniae was observed in WT versus IRAK4 mice, Macrophages (3 3 106) were plated in six-well plates and incubated with indicating the importance of IRAK4 during infection. Treatment the indicated concentrations of TLR agonists or bacteria. Cells were of infected mice with Hiltonol, a stabilized complex of poly- washed with ice-cold PBS and resuspended in lysis buffer (50 mM Tris [pH 7.5], 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.15 M inosinic-polycytidylic acid with poly-L-lysine and carboxymethyl NaCl, 2 mM EDTA with a protease inhibitor mixture [P8340; Sigma- cellulose (poly IC:LC) and potent TLR3 agonist, significantly Aldrich, St. Louis, MO] and a phosphatase inhibitor mixture [PhosStop; improved the survival rate of both WT and IRAK4KDKI mice. Roche]). Protein concentrations were determined (Bio-Rad), and lysates These data represent a potential treatment strategy for both normal were boiled for 5 min in reducing sample buffer (Thermo Scientific, and immunocompromised patients. Pittsburgh, PA). Equal quantities of total protein were electrophoresed on SDS polyacrylamide gels and transferred onto polyvinylidene difluoride membranes. Membranes were washed in TBST, incubated for 1 h at room Materials and Methods temperature in 5% BSA or 5% nonfat milk in TBST, incubated overnight Mice, macrophage isolation, and reagents at 4˚C with primary Ab, washed, incubated for 1 h at room temperature with HRP-labeled anti-rabbit or anti-mouse secondary Ab (Amersham IRAK4KDKI mice, generously provided by Dr. Kirk A. Staschke and Dr. Biosciences, Piscataway, NJ), and developed with the ECL detection kit Raymond Gilmour (Lilly Research Laboratories, Indianapolis, IN), were (Thermo Scientific). Anti–p-JNK, anti–p-ERK, anti–p-p38, anti-p38 total, derived on a C57BL/6J background and were bred homozygously at the anti-MSK1, and anti-MNK1 were purchased from Cell Signaling Tech- University of Maryland, School of Medicine. Age- and sex-matched nology (Danvers, MA); anti-H3ser10 and anti-H3 were purchased from The Journal of Immunology 309

Abcam (Cambridge, MA); and anti-actin was purchased from Santa Cruz induced by TLR2 or TLR4 signaling (Fig. 1A). IFN-b mRNA Biotechnology (Santa Cruz, CA). Anti-eIF4e and anti–p-eIF4e Abs were expression, which is induced by both LPS and poly I:C in a generously provided by Dr. Nahum Sonenberg (McGill University, Mon- MyD88-independent, TRIF-dependent fashion, was not inhibited treal, QC, Canada). in IRAK4KDKI macrophages stimulated with LPS or poly I:C (Fig. Nuclear extract preparation and oligodeoxynucleotide 1B). To examine whether these observations extend to additional coprecipitation genes, we performed qPCR analysis for the MyD88-dependent Macrophages (8 3 106) were plated in 10-cm dishes and treated with TLR genes that encode IL-10, COX-2, PAI-2, and IL-6, as well as the agonists. The cells were washed one time with PBS and pelleted by cen- MyD88-independent gene IP-10 (Fig. 1C). These data further trifugation, and nuclear extracts were prepared using the Nuclear Extract support the observation that TLR2-induced gene expression is Kit (Active Motif, Carlsbad, CA), according to the manufacturer’s proto- completely dependent on the kinase activity of IRAK4, whereas col. A biotinylated oligodeoxynucleotide (ODN) containing the NF-kB consensus sequence (59-AGTTGAGGGGACTTTCCCAGGC-39)was TLR4-induced gene expression is only partially dependent on the annealed to the complementary strand for 1 h at room temperature. This presence of this functional kinase. In contrast, LPS induction of procedure was also performed with a negative control ODN with a mutated IFN-b and IP-10 (Fig. 1B, 1C) are independent of IRAK4 kinase sequence (59-AGTTGAGGCGACATTCCCAGGC-39). Nuclear extract m activity. Additionally, we measured protein levels in the super- samples prepared from cells were diluted to 900 l with coprecipitation KDKI buffer (0.1% Triton X-100, 10 mM HEPES [pH 7.3], 2 mM EDTA, 1 mM natants of both WT and IRAK4 macrophages treated with EGTA, 10% glycerol, supplemented with a protease inhibitor mixture Pam3Cys, poly I:C, or LPS (Fig. 1D). In agreement with the [P8340; Sigma-Aldrich]) and precleared with 60 ml Salmon Sperm–Aga- mRNA data, Pam3Cys-treated IRAK4KDKI macrophages secreted rose beads (Millipore, Billerica, MA) for 30 min at 4˚C. Beads were re- very little IL-6 or KC, whereas LPS induced some IL-6 and KC moved by centrifugation, and the supernatant was incubated overnight secretion in these cells, but not to the same extent as in WT at 4˚C with 30 nM NF-kB–ODN. Streptavidin-Agarose beads (Sigma- Downloaded from Aldrich) were added for 1 h, and ODN–protein complexes were pel- macrophages. Although poly I:C was not a potent inducer of either leted, washed three times, and diluted in SDS-PAGE sample buffer. Boiled protein, it is interesting to note that KC protein expression was samples were centrifuged to remove beads, and the entire supernatant was enhanced in IRAK4KDKI macrophages compared with WT mac- loaded onto a 12% polyacrylamide gel. Western blots were performed as rophages (Fig. 1D, right panel, inset). This confirms, at both the described above. mRNA and protein levels, that the IRAK4KDKI macrophages are Mouse infections and bacteria capable of responding to some TLR agonists (i.e., TLR3, TLR4) http://www.jimmunol.org/ S. pneumoniae (ATCC 6303) was grown to log phase in BHI nutrient broth but not to others. (Teknova, Hollister, CA), supplemented with 10% sterile glycerol, and We also sought to determine whether IRAK4 kinase activity is stocks were stored at 280˚C. Mice were lightly anesthetized using iso- required to establish endotoxin tolerance in macrophages. As flurane (VetOne), and bacteria were administered intranasally (i.n.) at the expected from previous studies (27), LPS-treated macrophages appropriate concentration in 30 ml total saline. For studies using poly IC: LC, bacteria were administered i.n. in 50 ml saline or poly IC:LC solution from WT mice that were previously exposed to LPS demonstrated (100 mg/mouse). Bacterial concentration was verified for each experiment diminished induction of both MyD88-dependent (i.e., KC; Fig. by performing serial dilutions of the inoculum plated on Mueller–Hinton 1E, LPS/LPS) and MyD88-independent (i.e., IFN-b; Fig. 1E, LPS/ Agar supplemented with 5% sheep’s blood (Teknova). Mice were moni- LPS) genes compared with macrophages not previously exposed tored twice a day for survival studies. For mRNA analysis of cytokine to LPS (Fig. 1E, Med/LPS). Consistent with the data in Fig. 1A production, mice were sacrificed 24 or 48 h postinfection, bacterial load KDKI by guest on October 2, 2021 was assessed using serial plate dilutions, and a portion of the left lobe of and 1B, stimulation of medium-pretreated IRAK4 macro- lung was excised and homogenized in TriPure RNA extraction reagent phages with LPS resulted in reduced KC, but normal IFN-b, (Roche). RT-PCR and quantitative real-time PCR (qPCR) were performed mRNA levels (Fig. 1E; Med/LPS). However, like WT macro- as described above. phages, IRAK4KDKI macrophages that were previously exposed to LPS exhibited reduced expression of both KC and IFN-b mRNA Results compared with cells that were not. We conclude that, although IRAK4 kinase activity is required for optimal gene induction in IRAK4 kinase activity contributes to the MyD88-dependent response to TLR2 and TLR4, but not TLR3, agonists pathway, it is not required to establish endotoxin tolerance in IRAK4 was reported to be a critical kinase for propagating TLR- macrophages. Collectively, the data in Fig. 1 suggest that IRAK4 mediated signaling events leading to proinflammatory gene in- kinase activity is critical for the activation of MyD88-dependent duction (14). To assess the requirement for IRAK4 kinase activity gene expression in response to both TLR2 and TLR4 agonists, but in MyD88-dependent TLR signaling, macrophages from WT and it is not required for all macrophage responses because endotoxin IRAK4KDKI mice were isolated and treated with the TLR2 agonist tolerance and the response to TLR3 remain intact in the absence of Pam3Cys (MyD88 dependent) or the TLR4 agonist LPS (MyD88 kinase-active IRAK4. and TRIF dependent). mRNA expression for MyD88-dependent MAPK activation is severely deficient in TLR2-stimulated and MyD88-independent genes was measured 1 or 3 h poststim- KDKI ulation by qPCR. Induction of IL-1b and KC (Fig. 1A) mRNA in IRAK4 macrophages, whereas it is only partially deficient Pam3Cys-treated IRAK4KDKI macrophages was minimal com- in response to TLR4 stimulation pared with the robust induction induced in WT cells. In con- TLR signaling mediates activation of the MAPKs p38, ERK1/2, trast, LPS upregulated MyD88-dependent gene expression in and JNK that ultimately promote proinflammatory gene expres- IRAK4KDKI macrophages, although the levels of IL-1b and KC sion. Therefore, we assessed whether IRAK4 kinase activity is re- mRNA were significantly less compared with the full expression quired for activation of MAPKs. Macrophages from WT and observed in WT cells. Because the TLR3-signaling pathway is IRAK4KDKI mice were treated with Pam3Cys to activate TLR2 or MyD88 and IRAK4 independent, we also stimulated macrophages with LPS to activate TLR4, and the cells were lysed after 30 min. with the TLR3 agonist poly I:C to ensure that our IRAK4KDKI Activated MAPK proteins were measured by Western blot anal- macrophages were capable of normal gene expression in response ysis using Abs against phosphorylated (activated) proteins p38, to an IRAK4- and MyD88-independent stimulus. Very little dif- ERK1/2, and JNK. Although LPS-treated IRAK4KDKI macro- ference in mRNA induction in response to poly I:C was observed phages showed diminished activation of p38 and ERK1/2 com- in WT macrophages compared with IRAK4KDKI macrophages, pared with WT macrophages, activation of JNK was minimally although the magnitude of gene expression was less than that affected (Fig. 2A). In contrast, Pam3Cys treatment of IRAK4KDKI 310 DIFFERENTIAL NECESSITY OF IRAK4 KINASE FOR TLR SIGNALING Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 1. Cytokine gene expression in WT versus IRAK4KDKI macrophages. WT and IRAK4KDKI macrophages were treated with medium only (Med) or stimulated with Pam3Cys (P3C; 1 mg/ml), poly I:C (40 mg/ml), or LPS (100 ng/ml) for 1 or 3 h and analyzed for MyD88-dependent (A) and MyD88- independent (B) gene expression by qPCR. (C) Macrophages were treated as in (A) and (B) for 3 h, and additional gene expression was analyzed by qPCR. (D) Macrophages were treated as in (A)–(C) for 6 h, and supernatants were collected for cytokine measurement. (E) Macrophages isolated from WT and IRAK4KDKI mice were treated with medium only (Med/Med and Med/LPS) or LPS (100 ng/ml; LPS/LPS) for 18 h, washed three times, and restimulated with medium (Med/Med) or LPS (100 ng/ml; Med/LPS and LPS/LPS) for 5 h. MyD88-dependent (KC) and MyD88-independent (IFN-b) gene expression was analyzed by qPCR. All samples were normalized to HPRT. Data are representative of at least three separate experiments. macrophages yielded essentially no activation of any of the three observed in IRAK4KDKI macrophages (Fig. 2B), suggesting that MAPKs compared with WT cells (Fig. 2A). p38, ERK, and JNK the TLR4-signaling cascade is defective despite detectable, albeit were all activated in response to the TLR3 agonist poly I:C in both reduced, upstream activation of p38 and ERK1/2. As expected, WT and IRAK4KDKI macrophages, although to a lower extent than because of the lack of p38 and ERK activation observed in Fig. was detected with Pam3Cys or LPS (Fig. 2A, right panel). 2A, Pam3Cys stimulation in IRAK4KDKI macrophages also failed Interestingly, p38 and JNK activation were slightly augmented to activate MNK1 in IRAK4KDKI macrophages but not in WT cells in IRAK4KDKI cells compared with WT cells. Membranes were (Fig. 2B). Membranes were reprobed with anti-MNK1 to ensure reprobed with total p38 or actin to ensure equal loading. that the lack of activation is not attributable to a loss of protein MNK1 is a kinase that is activated by both p38 and ERK (28, expression in IRAK4KDKI cells. Interestingly, MNK1 appears 29). Nearly complete inhibition of MNK1 phosphorylation was to be differentially expressed in IRAK4KDKI macrophages, as The Journal of Immunology 311

data suggest that, despite the ability of IRAK4KDKI macrophages to retain the capacity to elicit some early MAPK activation, down- stream activation is significantly impaired in response to both TLR2 and TLR4 stimuli. eIF4e, a protein involved in translational regulation (31), is a known substrate for MNK1 (32). Phospho-eIF4e was increased in response to Pam3Cys and LPS in WT macrophages, albeit with different kinetics, but neither Pam3Cys nor LPS induced signifi- cant activation of eIF4e in the IRAK4KDKI cells (Fig. 2C). Therefore, we conclude that a lack of MNK1 activation disrupts downstream signaling events that may ultimately contribute to reduced gene expression in IRAK4KDKI macrophages. MSK1 phosphorylation of histone H3 at serine 10 leads to upregulation of early immediate genes that promote cytokine transcriptional activation (i.e., c-jun, c-fos) (33–35). Therefore, we tested whether a lack of MSK1 activation in IRAK4KDKI macro- phages led to a decrease in H3ser10 phosphorylation. Indeed, H3ser10 phosphorylation was decreased in IRAK4KDKI macro- phages compared with WT cells in response to both Pam3Cys and LPS (Fig. 2C). Downloaded from Previous reports demonstrated IRAK4-independent NF-kB acti- vation in response to TLR2 and TLR4 agonists in mice (19, 20), whereas other studies showed reduced NF-kB activity in response to LPS or IL-1b stimulation in a human patient with IRAK4 deficiency (36). Therefore, we used both TLR2 and TLR4 agonists to explore the role of IRAK4 kinase activity in NF-kB activation in murine http://www.jimmunol.org/ macrophages. Because activation of NF-kB prompts its transloca- tion into the nucleus, nuclear lysates from macrophages treated with FIGURE 2. Activation of MAPKs is diminished in IRAK4KDKI mac- Pam3Cys or LPS for 1 h were analyzed by Western blot using anti- rophages in response to Pam3Cys and LPS. (A) WT and IRAK4KDKI p65 and anti-p50 Abs. We found no deficiency of p65 in the nucleus macrophages were treated with Pam3Cys (P3C; 1 mg/ml) or LPS (100 ng/ of IRAK4KDKI macrophages in response to LPS. In contrast, there m ml) for 30 min (left panel) or poly I:C (50 g/ml) for 30, 60, or 90 min was a slight p65 deficiency in the nucleus of IRAK4KDKI macro- (right panel), and cell lysates were analyzed by Western blot. Membranes phages compared with WT cells in response to Pam3Cys (Fig. 2D, were probed with anti–phospho-p38, anti–phospho-ERK1/2, and anti– left panel). NF-kB p50 expression in the nucleus was diminished in phospho-JNK Abs. Membranes were stripped and reprobed using anti-total KDKI by guest on October 2, 2021 p38 (left panel) or anti-actin (right panel) to ensure equal loading. (B)WT IRAK4 macrophages compared with WT cells in response to and IRAK4KDKI macrophages were stimulated with Pam3Cys (P3C; 1 mg/ both Pam3Cys and LPS (Fig. 2D, left panel). ml) or LPS (100 ng/ml) for 30 min, and cell lysates were analyzed by To determine whether NF-kB binding to its cognate DNA se- Western blot using anti–phospho-MNK1, anti–total MNK1, and anti– quence was impaired in IRAK4KDKI macrophages, we performed phospho-MSK1 Abs. Membranes were reprobed with anti-actin to ensure a coprecipitation assay using a biotinylated ODN containing the equal loading. (C) WT and IRAK4KDKI macrophages were treated with NF-kB consensus-binding sequence. The same nuclear extract Pam3Cys (P3C, 1 mg/ml) or LPS (100 ng/ml) for 30 and 60 min, analyzed preparations that were analyzed by Western blot in Fig. 2D were by Western blot using anti–phospho-eIF4e Ab and reprobed using anti– incubated with biotinylated NF-kB–specific ODN or a mutated total eIF4e Ab to ensure equal loading or analyzed by Western blot using sequence that served as a negative control (LPS-stimulated nu- anti–phospho-histone H3 serine 10 and reprobed using anti–total histone clear extracts were incubated with the mutated sequence to control H3 to ensure equal loading. (D) WT and IRAK4KDKI macrophages were treated as in (A) for 1 h. Nuclear extracts were prepared and analyzed by for nonspecific binding). Streptavidin-conjugated beads were used Western blot using anti-p65, anti-p50, and anti-actin (loading control) Abs to precipitate the biotinylated ODNs, and coprecipitated proteins (left panel). The same lysates were also incubated overnight with a syn- were resolved by Western blot and probed with anti-p65 and anti- thetic biotinylated ODN containing the NF-kB consensus sequence or with p50 Abs. There was no decrease in p65 binding in IRAK4KDKI a mutated sequence as a negative control (NC). Biotinylated ODNs were macrophages in response to Pam3Cys or LPS (Fig. 2D, right precipitated using streptavidin-conjugated agarose beads, and protein was panel). Similarly, binding of NF-kB p50 was not impaired in eluted with reducing sample buffer and analyzed by Western blot using Pam3Cys- or LPS-treated IRAK4KDKI cells, despite the dimin- anti-p65 or anti-p50 Abs (right panel). All results are representative of at ished quantity of p50 in the lysates used. Interestingly, binding of least three independent experiments. both p65 and p50 was slightly augmented in IRAK4KDKI macro- phages in response to LPS compared with WT cells. This may demonstrated by the presence of lower molecular weight bands. reflect augmented activation of p65 and p50 (independent of its These bands may represent degradation products of the protein, translocation) and/or the activation of other transcription factors because treatment with the proteasome inhibitor MG-132 reduced that enhance their DNA binding in IRAK4KDKI cells. Importantly, their expression (data not shown). these data demonstrate that, unlike the complete attenuation We also sought to determine whether our TLR agonists activate of MAPK signaling in response to the MyD88-dependent TLR2 MSK1, a protein kinase that is similarly activated by ERK1/2 (30). stimulus, Pam3Cys, NF-kB binding remains intact in IRAK4KDKI As demonstrated by Western blot analysis, MSK1 is slightly ac- macrophages. Altogether, these data suggest that, although TLR2- tivated by the TLR2 agonist Pam3Cys and is robustly activated by induced MAPK activation is completely dependent on IRAK4 LPS in WT macrophages, but phosphorylation of MSK1 was not kinase activity, TLR2-induced NF-kB activation is only partially detectable in IRAK4KDKI macrophages (Fig. 2B). Together, these dependent on the presence of a functional IRAK4 kinase (as 312 DIFFERENTIAL NECESSITY OF IRAK4 KINASE FOR TLR SIGNALING demonstrated by decreased translocation of p65). In contrast, Pam3Cys and LPS cotreatment elicits a more robust response than TLR4 activation of MAPKs is only partially dependent on IRAK4 either agonist alone, we treated WT macrophages with Pam3Cys, kinase activity, yet the activation of NF-kB is independent of LPS, or a combination of the two for 3 h and measured gene a functional IRAK4 kinase. To examine whether the NF-kB ac- expression by qPCR (Fig. 3B). Macrophages treated with both tivation in IRAK4KDKI macrophages is a result of activation of an Pam3Cys and LPS exhibited slightly higher levels of gene ex- alternative pathway, such as activation of PI3K, we repeated our pression but did not synergize. Therefore, the extent to which ODN precipitation in the presence of the PI3K inhibitor responses to whole heat-killed Gram-positive and Gram-negative LY294002. We found that the inhibitor did not decrease NF-kB bacteria are IRAK4 dependent appears to reflect the predominant activation at a dose that inhibited activation of PI3K (data not TLR agonist in the bacteria itself, because HKPA expresses both shown). Thus, the mechanism by which NF-kB achieves full ac- TLR2 and TLR4 PAMPs, whereas HKSA expresses predomi- tivation, despite the lack of IRAK4 kinase activity, remains un- nantly TLR2 agonists. The relative amounts of these agonists are clear but is independent of PI3K. likely to play a role in the magnitude of response. S. pneumoniae is the leading killer of patients with deficiencies IRAK4 kinase activity is required for complete gene induction in IRAK4 (4, 6). Therefore, we sought to determine what re- in macrophages in response to heat-killed and live bacteria sponse, if any, was induced in IRAK4KDKI macrophages infected Although previous work largely examined the role of IRAK4 with increasing numbers of S. pneumoniae. Strikingly, there was kinase activity in response to purified TLR agonists, few studies virtually no mRNA induction of the proinflammatory cytokines addressed the importance of this kinase in response to whole IL1-b, KC, or IL-12 p40 in IRAK4KDKI macrophages in response bacteria. Therefore, we cultured macrophages from WT and to infection with live bacteria, despite robust responses observed

KDKI Downloaded from IRAK4 mice and stimulated them with heat-killed S. aureus in WT cells (Fig. 3C). (HKSA; a Gram-positive bacterium) or heat-killed P. aeruginosa KDKI (HKPA; a Gram-negative bacterium) for 1, 3 or 5 h. Gene ex- IRAK4 mice are more susceptible to S. pneumoniae pression was measured by qPCR. As positive controls, we also infection than WT mice treated these cells with the purified TLR agonists, Pam3Cys or To assess whether IRAK4 kinase activity plays a role in con- LPS. We observed differences in the macrophage response to trolling S. pneumoniae infection, WT and IRAK4KDKI mice were bacteria that were species dependent. Fig. 3A shows that IL-1b, infected i.n. with S. pneumoniae, and survival was assessed over http://www.jimmunol.org/ KC, and IL-6 mRNA expression in response to HKSA was com- time. Although the majority of both WT and IRAK4KDKI mice pletely dependent on IRAK4 kinase activity (similar to Pam3Cys) succumbed to S. pneumoniae infection within 7 d, there was and was weaker than responses induced by LPS or HKPA. The a significant increase in the rate of death of IRAK4KDKI mice responses to HKPA were much more robust than even that in- compared with their WT counterparts (Fig. 4, p = 0.022, log-rank duced by LPS and were only partially dependent on the presence test). To determine whether this increased susceptibility was due of a functional IRAK4 kinase (Fig. 3A). To determine whether to a deficiency in cytokine production, lungs were isolated from by guest on October 2, 2021

FIGURE 3. Bacteria-induced cytokine gene expres- sion is impaired in IRAK4KDKI macrophages. (A)WT and IRAK4KDKI macrophages were treated with Pam3- Cys (P3C; 1 mg/ml), LPS (100 ng/ml), HKSA (5 3 107 bacteria), or HKPA (5 3 107 bacteria) for 1, 3, and 5 h. Samples were then analyzed for gene expression by qPCR. All samples were normalized to the house- keeping gene HPRT. Samples were tested in duplicate; and each graph represents three separate experiments. (B) WT and IRAK4KDKI macrophages were treated with Pam3Cys (P3C; 1 mg/ml), LPS (100 ng/ml), or both for 3 h. Samples were then analyzed for gene expression by qPCR. (C) WT and IRAK4KDKI macrophages were treated with Pam3Cys (P3C; 1 mg/ml) or various concentrations of live S. pneumoniae (SP; multiplicity of infection 1:1, 0.2:1, and 0.04:1). Samples were then analyzed for gene expression by qPCR. Each sample was normalized to HPRT, and each graph represents two separate experiments. The Journal of Immunology 313

(Fig. 5C) 48 h postinfection. IRAK4KDKI mice produced con- sistently lower levels of IL-6 and KC protein while demonstrating significantly higher bacterial burdens in their lung tissue. Inter- estingly, we did not observe any striking differences in tissue histology between WT and IRAK4KDKI mice when sacrificed at 48 or 72 h postinfection (data not shown). Both groups of mice showed low levels of inflammatory infiltration at both 48 and 72 h (a time point when 40% of the IRAK4KDKI mice had already expired because of infection). Together, these data suggest that the absence of a functional IRAK4 kinase decreases proin- FIGURE 4. IRAK4KDKI mice are more susceptible than WT mice to S. flammatory cytokine production in response to S. pneumoniae pneumoniae infection. WT or IRAK4KDKI mice were infected i.n. with 1 3 infection while facilitating bacterial replication and, thereby, in- 5 10 S. pneumoniae (S.p.) in a total volume of 30 ml or mock infected with creasing the susceptibility of IRAK4KDKI mice to pneumococcal an equal volume of saline. Survival was assessed over the course of 14 d. infection. Graph shows the combined data of four separate experiments with a total of 19 mice/mouse strain (S.p. infected) or 4 mice/mouse strain (mock Poly IC:LC (Hiltonol) improves survival of both WT and KDKI infected) in each group. *p = 0.022, log-rank test. IRAK4 mice infected with S. pneumoniae Our in vitro data demonstrate that TLR3 signaling remains intact S. pneumoniae–infected mice 48 h postinfection, RNA was in IRAK4KDKI macrophages. Therefore, we sought to determine extracted, and qPCR was used to quantify gene induction. As whether administration of the TLR3 agonist poly IC:LC (Hilto- Downloaded from observed in vitro using isolated macrophages, there was a signifi- nol), a stabilized poly I:C, could improve survival in IRAK4KDKI cant decrease in cytokine mRNA production in the IRAK4KDKI mice infected with a lethal dose of S. pneumoniae.Inthese lungs compared with WT lungs in response to S. pneumoniae studies, vehicle only or 100 mg of poly IC:LC was administered infection (Fig. 5A). Interestingly, in contrast to stimulation of i.n. at the time of infection in both WT and IRAK4KDKI mice. There macrophages with Pam3Cys, which does not induce IFN-b,in- was a significant increase in the survival of infected mice treated

fection with S. pneumoniae in vivo induced IFN-b in both WT with poly IC:LC compared with those treated with vehicle only http://www.jimmunol.org/ and IRAK4KDKI mice, suggesting that there are non-TLR2, (Fig. 6; WT, p = 0.0097, log-rank test; IRAK4KDKI, p =0.0001, IRAK4-independent ligands present on S. pneumoniae that can log-rank test). Thus, stimulation of TLR3, a MyD88- and IRAK4- induce IFN-b mRNA, an observation that is consistent with other independent innate immune pathway, provides a possible thera- recent studies (37, 38). To characterize further differences in the peutic strategy to boost immunity against bacterial pathogens in course of infection between WT and IRAK4KDKI mice, we also patients lacking key signaling components of MyD88- and/or measured serum cytokine levels (Fig. 5B) and bacterial burden IRAK4-dependent signaling pathways. by guest on October 2, 2021

FIGURE 5. Gene induction is impaired in the lungs of S. pneumoniae–infected IRAK4KDKI mice. (A)WT or IRAK4KDKI mice were infected i.n. with 5 3 105 S. pneumoniae (SP) in a total volume of 30 ml or mock infected with an equal volume of saline. After 48 h, mice were sacrificed, lungs were harvested, RNA was extracted, and gene expression was measured by qPCR. Each sample was normalized to HPRT. Data shown are combined from at least three separate experiments using a total of four mice (saline) or at least nine mice (SP infected) per group. (B and C)WT or IRAK4KDKI mice were infected as in (A). After 48 h, serum cytokines (B) or bacterial rRNA (C) was mea- sured. Data are combined from two separate experi- ments using a total of nine (B) or six (C) mice/group. *p # 0.05, **p , 0.01, Student t test. 314 DIFFERENTIAL NECESSITY OF IRAK4 KINASE FOR TLR SIGNALING

FIGURE 6. Poly IC:LC (Hiltonol) treatment im- proves survival of mice infected with S. pneumoniae. WT (A) or IRAK4KDKI (B) mice were infected i.n. with 1 3 105 S. pneumoniae in a total volume of 50 mlof poly IC:LC (100 mg/mouse, s) or saline (vehicle, N). Survival was assessed over the course of 14 d. The data are combined from three separate experiments with a total of 11 mice/group (10 mice for WT vehicle group). *p = 0.0097, **p = 0.0001, log-rank test.

Discussion affected by the loss of IRAK4 kinase activity. Furthermore, the TLR signaling is a critical activator of immune defense during loss in gene expression cannot be attributed to a loss of early infection. In this study, we clarified the importance of IRAK4 signaling events, such as activation of p38, ERK1/2, or JNK, as is kinase activity in generating full responses to Pam3Cys and LPS, the case for TLR2 signaling. Our results show LPS-induced ac- agonists for TLR2 and TLR4, respectively. Importantly, cytokine tivation of MAPK in the absence of IRAK4 kinase activity, a re- production in IRAK4KDKI macrophages was also impaired in re- sult that agrees with previous studies showing MAPK activation 2 2 sponse to both heat-killed and live bacteria. We observed the most in both IRAK4KDKI and IRAK4 / macrophages (20, 39, 40). severe deficiency in IRAK4KDKI macrophages when these cells Together, these data support a model in which LPS can induce Downloaded from were infected with S. pneumoniae, reminiscent of the phenotype MAPK activation completely independently of IRAK4 protein. observed in patients with IRAK4 mutations (4, 6, 36). Although This may occur via the LPS-induced MyD88-independent/TRIF- early signaling events, such as phosphorylation of p38, ERK1/2, dependent pathway, because MAPK activation by poly I:C, and JNK, were completely abrogated in response to Pam3Cys, a TLR3 agonist that also signals in a MyD88-independent/TRIF- IRAK4KDKI macrophages retained some ability to activate these dependent fashion, is not affected in cells lacking IRAK4 kinase 2 2 proteins in response to LPS. However, activation of MNK1 and activity (Fig. 3A). Indeed, even MyD88 / macrophages dem- http://www.jimmunol.org/ MSK1 in response to LPS was diminished in IRAK4KDKI mac- onstrate MAPK activation in response to LPS (39, 41, 42). Im- rophages, suggesting that IRAK4 also plays a role downstream portantly, both MNK1 and MSK1, kinases that are activated of p38, ERK1/2, and JNK activation in response to TLR4 ago- downstream of activated MAPKs, were deficient in IRAK4KDKI nists. Proinflammatory cytokine mRNA and protein expression macrophages in response to LPS. To confirm a functional conse- remained partially intact in IRAK4KDKI macrophages in response quence for this loss of kinase activation, we demonstrated di- to LPS, suggesting that early MAPK signaling is sufficient for minished activation of their downstream substrates as well upregulation of gene expression, despite diminished downstream (i.e., eIF4e for MNK1 and histone H3 serine 10 for MSK1). Of activity. However, our data also indicate that LPS activation of particular note, the observation that LPS-induced tolerance was by guest on October 2, 2021 MAPKs is at least partially IRAK4 dependent, because we ob- not affected in IRAK4KDKI macrophages, despite the observed served diminished activation of their downstream substrates (such decrease in MyD88-dependent signaling, suggests either that the as MNK1 and MSK1, Fig. 2B). partial response to LPS observed in such macrophages achieves Our data support the hypothesis that IRAK4 kinase activity is a threshold sufficient to elicit the chromatin remodeling suggested most essential for MyD88-dependent TLR signaling. TLR2 is to underlie tolerance (43) or that tolerance induction is indepen- entirely MyD88 dependent, and we show a nearly complete loss of dent of IRAK4 activity. Studies are ongoing to determine the signaling in response to the TLR2 agonist Pam3Cys, as demon- extent to which IRAK4 contributes to endotoxin tolerance. strated by decreased cytokine mRNA expression (Figs. 1, 3) and Kim et al. (20) suggested that, although LPS-induced mRNA loss of MAPK activation (Fig. 2). Similar results were reported by expression was not greatly affected in IRAKKDKI macrophages, Kawagoe et al. (19) using the TLR2 agonist MALP2, a TLR2/6 protein levels were severely deficient as the result of a loss of agonist (Pam3Cys is a TLR1/2 agonist). However, NF-kB acti- mRNA stability and, thus, a reduction in translation of these vation in response to a TLR2 agonist was not diminished in cytokines. Although our data clearly demonstrate diminished IRAK4KDKI macrophages compared with WT cells (Fig. 2D), mRNA expression, our observation that eIF4e phosphorylation is suggesting that IRAK4 kinase–dependent activation of MAPKs is also severely diminished in response to LPS (Fig. 2C) suggests essential for upregulation of proinflammatory cytokine gene ex- that there may also be a secondary loss of translational activity pression. We also report that TLR3 signaling, which is completely for LPS-induced, MyD88-dependent genes. eIF4e plays a role in independent of MyD88, is not affected by the loss of IRAK4 ki- maintaining mRNA stability, leading to efficient mRNA transla- nase activity, as demonstrated by equal cytokine mRNA expres- tion (31). If eIF4e activation were lost, as we found in response to sion in both WT and IRAK4KDKI macrophages in response to the both TLR2 and TLR4 agonists (Fig. 2C), then it would be ex- TLR3 agonist poly I:C, a result that agrees with previous studies pected that translational activity would also be reduced. Interest- (19, 20). However, currently published data do not agree on the ingly, a similar pattern of MAPK activation (i.e., no defect in role of IRAK4 kinase activity in response to the TLR4 agonist activation of p38 or ERK with downstream impairment of MNK1 LPS. Because LPS can propagate both MyD88-dependent and and eIF4e activation) was demonstratedinIRAK2-deficient MyD88-independent (TRIF-dependent) signaling pathways, it is macrophages in response to LPS (44). This suggests that acti- of critical importance that induction of genes that are pathway vation of MNK1 and its downstream target, eIF4e, requires the dependent be considered separately. Our data confirm that LPS presence of IRAK2, as well as the kinase activity of IRAK4 signaling is dependent on IRAK4 kinase activity, because (potentially to phosphorylate/activate IRAK2). In addition, we IRAK4KDKI macrophages exhibit diminished MyD88-dependent demonstrate little activation of MSK1 in response to LPS in cytokine production in response to LPS (Fig. 1). However, IRAKKDKI macrophages. MSK1 phosphorylates histone H3 serine MyD88-independent genes, such as IFN-b and IP-10, are not 10, leading to increased gene expression, most notably of early The Journal of Immunology 315 immediate genes (33–35). In this study, we show that MSK1 9. 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. activation and subsequent phosphorylation of histone H3 serine Nature 388: 394–397. KDKI 10 are both diminished in LPS-stimulated IRAK4 macro- 10. Poltorak,A.,X.He,I.Smirnova,M.Y.Liu,C.VanHuffel,X.Du,D.Birdwell, phages, thus providing evidence for the potential role of MSK1 E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/ HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282: 2085– in the partial defect in the MAPK-signaling cascade seen in re- 2088. sponse to LPS. 11. Jiang, Z., M. Zamanian-Daryoush, H. Nie, A. M. Silva, B. R. Williams, and Patients with IRAK4 deficiency are most susceptible to S. X. Li. 2003. Poly(I-C)-induced Toll-like receptor 3 (TLR3)-mediated activation of NFkappa B and MAP kinase is through an interleukin-1 receptor-associated pneumoniae infection. Similarly, we observed increased suscep- kinase (IRAK)-independent pathway employing the signaling components tibility to S. pneumonia infection in IRAK4KDKI mice. These mice TLR3-TRAF6-TAK1-TAB2-PKR. J. Biol. Chem. 278: 16713–16719. 12. Watters, T. M., E. F. Kenny, and L. A. O’Neill. 2007. Structure, function and died faster than WT mice; they exhibited decreased cytokine gene regulation of the Toll/IL-1 receptor adaptor proteins. Immunol. Cell Biol. 85: expression, decreased protein levels, and increased bacterial bur- 411–419. den in their lungs compared with WT mice (Figs. 4, 5). Combined 13. Lin, S. C., Y. C. Lo, and H. Wu. 2010. Helical assembly in the MyD88-IRAK4- IRAK2 complex in TLR/IL-1R signalling. Nature 465: 885–890. with our in vitro data, this presents a feasible experimental model 14. Flannery, S., and A. G. Bowie. 2010. The interleukin-1 receptor-associated with which to explore therapeutic interventions that may com- kinases: critical regulators of innate immune signalling. Biochem. Pharmacol. pensate for the loss of IRAK4 signaling. To this end, our data 80: 1981–1991. 15. Burns, K., S. Janssens, B. Brissoni, N. Olivos, R. Beyaert, and J. Tschopp. 2003. show the IRAK4-independent, TLR3 agonist poly IC:LC signifi- Inhibition of interleukin 1 receptor/Toll-like receptor signaling through the al- cantly increased the survival of S. pneumoniae–infected mice. ternatively spliced, short form of MyD88 is due to its failure to recruit IRAK-4. J. Exp. Med. 197: 263–268. These data are in contrast with those of Nakamura et al. (45) who 16. Suzuki, N., and T. Saito. 2006. IRAK-4—a shared NF-kappaB activator in innate found increased S. pneumoniae colonization after administration and acquired immunity. Trends Immunol. 27: 566–572. of repeated doses of poly IC:LC. However, their study did not 17. Suzuki, N., S. Suzuki, G. S. Duncan, D. G. Millar, T. Wada, C. Mirtsos, Downloaded from H. Takada, A. Wakeham, A. Itie, S. Li, et al. 2002. Severe impairment of determine the survival rate of mice receiving poly IC:LC treat- interleukin-1 and Toll-like receptor signalling in mice lacking IRAK-4. ment, and there was no significant increase in colonization when Nature 416: 750–756. poly IC:LC was administered less frequently. Interestingly, the 18. Suzuki, N., S. Suzuki, and W. C. Yeh. 2002. IRAK-4 as the central TIR signaling mediator in innate immunity. Trends Immunol. 23: 503–506. increased S. pneumoniae colonization was dependent on poly IC: 19. Kawagoe, T., S. Sato, A. Jung, M. Yamamoto, K. Matsui, H. Kato, S. Uematsu, LC–induced IFN-b. Because poly IC:LC is a potent inducer of O. Takeuchi, and S. Akira. 2007. Essential role of IRAK-4 protein and its kinase b b activity in Toll-like receptor-mediated immune responses but not in TCR sig- http://www.jimmunol.org/ IFN- in vivo (46), we are investigating the requirement of IFN- naling. J. Exp. Med. 204: 1013–1024. to confer poly IC:LC protection in our model. However, our data 20. Kim, T. W., K. Staschke, K. Bulek, J. Yao, K. Peters, K. H. Oh, Y. Vandenburg, showing that IFN-b production in response to S. pneumoniae in- H. Xiao, W. Qian, T. Hamilton, et al. 2007. A critical role for IRAK4 kinase KDKI activity in Toll-like receptor-mediated innate immunity. J. Exp. Med. 204: 1025– fection is not inhibited in IRAK4 mice (Fig. 5) suggests that 1036. the mechanism may be independent of IFN-b induction. With 21. Fraczek, J., T. W. Kim, H. Xiao, J. Yao, Q. Wen, Y. Li, J. L. Casanova, J. Pryjma, further studies, such intervention could prove useful in patients and X. Li. 2008. The kinase activity of IL-1 receptor-associated kinase 4 is required for interleukin-1 receptor/toll-like receptor-induced TAK1-dependent with mutations in the MyD88/IRAK4-dependent pathway, be- NFkappaB activation. J. Biol. Chem. 283: 31697–31705. cause our in vitro data suggest that activation of MAPKs with 22. Lye, E., S. Dhanji, T. Calzascia, A. R. Elford, and P. S. Ohashi. 2008. IRAK-4 stimuli that are TLR/MyD88/IRAK4 independent might compen- kinase activity is required for IRAK-4-dependent innate and adaptive immune responses. Eur. J. Immunol. 38: 870–876. by guest on October 2, 2021 sate for deficiencies in early signaling proteins of the TLR 23. Song, K. W., F. X. Talamas, R. T. Suttmann, P. S. Olson, J. W. Barnett, S. W. Lee, cascade. With NF-kB signaling relatively intact in TLR2- or TLR4- K. D. Thompson, S. Jin, M. Hekmat-Nejad, T. Z. Cai, et al. 2009. The kinase KDKI activities of interleukin-1 receptor associated kinase (IRAK)-1 and 4 are re- stimulated IRAK4 macrophages, our data support the concept dundant in the control of inflammatory cytokine expression in human cells. Mol. of increasing activation of the MAPK pathway and targeting its Immunol. 46: 1458–1466. complete restoration in the absence or mitigation of IRAK4 kinase 24. Sklar, M. D., A. Tereba, B. D. Chen, and W. S. Walker. 1985. Transformation of mouse bone marrow cells by transfection with a human oncogene related to activity. c-myc is associated with the endogenous production of macrophage colony stimulating factor 1. J. Cell. Physiol. 125: 403–412. 25. McIntire, F. C., H. W. Sievert, G. H. Barlow, R. A. Finley, and A. Y. Lee. 1967. Disclosures Chemical, physical, biological properties of a lipopolysaccharide from Esch- A.S. works for Oncovir, which manufactures poly IC:LC, a reagent used erichia coli K-235. Biochemistry 6: 2363–2372. in this study. The other authors have no financial conflicts of interest. 26. Schmittgen, T. D., and K. J. Livak. 2008. Analyzing real-time PCR data by the comparative C(T) method. Nat. Protoc. 3: 1101–1108. 27. Dobrovolskaia, M. A. and S. N. Vogel. 2002. Toll receptors, CD14, and mac- rophage activation and deactivation by LPS. Microbes Infect. 4: 903–914. References 28. Fukunaga, R., and T. Hunter. 1997. MNK1, a new MAP kinase-activated protein 1. Kumar, H., T. Kawai, and S. Akira. 2011. Pathogen recognition by the innate kinase, isolated by a novel expression screening method for identifying protein immune system. Int. Rev. Immunol. 30: 16–34. kinase substrates. EMBO J. 16: 1921–1933. 2. Medzhitov, R. 2007. Recognition of microorganisms and activation of the im- 29. Waskiewicz, A. J. 1997. Mitogen-activated protein kinases activate the serine/ mune response. Nature 449: 819–826. threonine kinases Mnk1 and Mnk2. EMBO J. 16: 1909–1920. 3. Ku, C. L., C. Picard, M. Erdo¨s, A. Jeurissen, J. Bustamante, A. Puel, H. von 30. McCoy, C. E., D. G. Campbell, M. Deak, G. B. Bloomberg, and J. S. C. Arthur. Bernuth, O. Filipe-Santos, H. H. Chang, T. Lawrence, et al. 2007. IRAK4 and 2005. MSK1 activity is controlled by multiple phosphorylation sites. Biochem. J. NEMO mutations in otherwise healthy children with recurrent invasive pneu- 387: 507–517. mococcal disease. J. Med. Genet. 44: 16–23. 31. Rhoads, R. E. 2009. eIF4E: new family members, new binding partners, new 4. Ku, C. L., H. von Bernuth, C. Picard, S. Y. Zhang, H. H. Chang, K. Yang, roles. J. Biol. Chem. 284: 16711–16715. M. Chrabieh, A. C. Issekutz, C. K. Cunningham, J. Gallin, et al. 2007. Selective 32. Waskiewicz, A. J., J. C. Johnson, B. Penn, M. Mahalingam, S. R. Kimball, and predisposition to bacterial infections in IRAK-4-deficient children: IRAK-4- J. A. Cooper. 2012. Phosphorylation of the cap-binding protein eukaryotic dependent TLRs are otherwise redundant in protective immunity. J. Exp. Med. translation initiation factor 4E by protein kinase Mnk1 in vivo. Mol. Cell Biol. 204: 2407–2422. 19: 1871–1880. 5. Picard, C., J. L. Casanova, and A. Puel. 2011. Infectious diseases in patients with 33. Clayton, A. L., S. Rose, M. J. Barratt, and L. C. Mahadevan. 2000. Phosphoa- IRAK-4, MyD88, NEMO, or IkBa deficiency. Clin. Microbiol. Rev. 24: 490–497. cetylation of histone H3 on c-fos- and c-jun-associated nucleosomes upon gene 6. Picard, C., H. von Bernuth, P. Ghandil, M. Chrabieh, O. Levy, P. D. Arkwright, activation. EMBO J. 19: 3714–3726. D. McDonald, R. S. Geha, H. Takada, J. C. Krause, et al. 2010. Clinical features 34. Lau, P. N., and P. Cheung. 2011. Histone code pathway involving H3 S28 and outcome of patients with IRAK-4 and MyD88 deficiency. Medicine (Bal- phosphorylation and K27 acetylation activates transcription and antagonizes timore) 89: 403–425. polycomb silencing. Proc. Natl. Acad. Sci. USA 108: 2801–2806. 7. Suhir, H., and A. Etzioni. 2010. The role of Toll-like receptor signaling in human 35. Thomson, S., A. L. Clayton, C. A. Hazzalin, S. Rose, M. J. Barratt, and immunodeficiencies. Clin. Rev. Allergy Immunol. 38: 11–19. L. C. Mahadevan. 1999. The nucleosomal response associated with immediate- 8. Oliveira-Nascimento, L., P. Massari, and L. M. Wetzler. 2012. The Role of TLR2 early gene induction is mediated via alternative MAP kinase cascades: MSK1 as in Infection and Immunity. Front. Immunol. 3: 79. a potential histone H3/HMG-14 kinase. EMBO J. 18: 4779–4793. 316 DIFFERENTIAL NECESSITY OF IRAK4 KINASE FOR TLR SIGNALING

36. Medvedev, A. E., A. Lentschat, D. B. Kuhns, J. C. Blanco, C. Salkowski, 41. Kawai, T., O. Adachi, T. Ogawa, K. Takeda, and S. Akira. 1999. Unrespon- S. Zhang, M. Arditi, J. I. Gallin, and S. N. Vogel. 2003. Distinct mutations in siveness of MyD88-deficient mice to endotoxin. Immunity 11: 115–122. IRAK-4 confer hyporesponsiveness to lipopolysaccharide and interleukin-1 in 42. Kim, L., B. A. Butcher, C. W. Lee, S. Uematsu, S. Akira, and E. Y. Denkers. a patient with recurrent bacterial infections. J. Exp. Med. 198: 521–531. 2006. Toxoplasma gondii genotype determines MyD88-dependent signaling in 37. Koppe, U., K. Ho¨gner, J. M. Doehn, H. C. Mu¨ller, M. Witzenrath, B. Gutbier, infected macrophages. J. Immunol. 177: 2584–2591. S. Bauer, T. Pribyl, S. Hammerschmidt, J. Lohmeyer, et al. 2012. Streptococcus 43. Foster, S. L., D. C. Hargreaves, and R. Medzhitov. 2007. Gene-specific control pneumoniae stimulates a STING- and IFN regulatory factor 3-dependent type I of inflammation by TLR-induced chromatin modifications. Nature 447: 972– IFN production in macrophages, which regulates RANTES production in mac- 978. rophages, cocultured alveolar epithelial cells, and mouse lungs. J. Immunol. 188: 44. Wan, Y., H. Xiao, J. Affolter, T. W. Kim, K. Bulek, S. Chaudhuri, D. Carlson, 811–817. T. Hamilton, B. Mazumder, G. R. Stark, et al. 2009. Interleukin-1 receptor- 38. Parker, D., F. J. Martin, G. Soong, B. S. Harfenist, J. L. Aguilar, A. J. Ratner, associated kinase 2 is critical for lipopolysaccharide-mediated post-transcriptional K. A. Fitzgerald, C. Schindler, and A. Prince. 2011. Streptococcus pneumoniae DNA control. J. Biol. Chem. 284: 10367–10375. initiates type I interferon signaling in the respiratory tract. MBio. 2: e00016-11. 45. Nakamura, S., K. M. Davis, and J. N. Weiser. 2011. Synergistic stimulation of 39. Oliveira, F. S., N. B. Carvalho, A. P. Branda˜o, M. T. Gomes, L. A. de Almeida, type I interferons during influenza virus coinfection promotes Streptococcus and S. C. Oliveira. 2011. Interleukin-1 receptor-associated kinase 4 is essential for pneumoniae colonization in mice. J. Clin. Invest. 121: 3657–3665. initial host control of Brucella abortus infection. Infect. Immun. 79: 4688–4695. 46. Caskey, M., F. Lefebvre, A. Filali-Mouhim, M. J. Cameron, J. P. Goulet, 40. Koziczak-Holbro, M., A. Glu¨ck, C. Tschopp, J. C. Mathison, and H. Gram. 2008. E. K. Haddad, G. Breton, C. Trumpfheller, S. Pollak, I. Shimeliovich, et al. 2011. IRAK-4 kinase activity-dependent and -independent regulation of lipopolysaccharide- Synthetic double-stranded RNA induces innate immune responses similar to inducible genes. Eur. J. Immunol. 38: 788–796. a live viral vaccine in humans. J. Exp. Med. 208: 2357–2366. Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021