Dendritic Cell Activation of NK Cells Innate Immune Response By

Lyn-Dependent Signaling Regulates the Innate Immune Response by Controlling Dendritic Cell Activation of NK Cells

This information is current as Danielle L. Krebs, Manreet K. Chehal, Alexander Sio, of September 25, 2021. Nicholas D. Huntington, Mei Lin Da, Pascal Ziltener, Melissa Inglese, Nicole Kountouri, John J. Priatel, Jessica Jones, David M. Tarlinton, Gary P. Anderson, Margaret L. Hibbs and Kenneth W. Harder

J Immunol 2012; 188:5094-5105; Prepublished online 9 Downloaded from April 2012; doi: 10.4049/jimmunol.1103395 http://www.jimmunol.org/content/188/10/5094 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2012/04/09/jimmunol.110339 Material 5.DC1 References This article cites 85 articles, 33 of which you can access for free at: http://www.jimmunol.org/content/188/10/5094.full#ref-list-1

Why The JI? Submit online. by guest on September 25, 2021

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

Lyn-Dependent Signaling Regulates the Innate Immune Response by Controlling Dendritic Cell Activation of NK Cells

Danielle L. Krebs,*,1 Manreet K. Chehal,*,1 Alexander Sio,* Nicholas D. Huntington,† Mei Lin Da,* Pascal Ziltener,‡ Melissa Inglese,‡ Nicole Kountouri,‡ John J. Priatel,x Jessica Jones,{,‖ David M. Tarlinton,† Gary P. Anderson,{,‖ Margaret L. Hibbs,‡,2,3 and Kenneth W. Harder*,2

The innate immune response is a first line of defense against invading pathogens; however, the magnitude of this response must be tightly regulated, as hyper- or suboptimal responses can be detrimental to the host. Systemic inflammation resulting from bacterial Downloaded from infection can lead to sepsis, which remains a serious problem with high mortality rates. Lyn tyrosine kinase plays a key role in adaptive immunity, although its role in innate immunity remains unclear. In this study, we show that Lyn gain-of-function (Lynup/up) mice display enhanced sensitivity to endotoxin and succumb to upregulated proinflammatory cytokine production at a dose well tolerated by control animals. Endotoxin sensitivity in Lynup/up mice depends on dendritic cells (DCs) and NK cells and occurs though a mechanism involving increased maturation and activation of the DC compartment, leading to elevated production of IFN-g by NK cells. We further show that modulation of endotoxin-induced signal transduction in DCs by Lyn involves the http://www.jimmunol.org/ phosphatases Src homology 2 domain-containing phosphatase-1 and SHIP-1. Collectively, we demonstrate that Lyn regulates DC physiology such that alterations in Lyn-dependent signaling have profound effects on the nature and magnitude of inflammatory responses. Our studies highlight how perturbations in signaling pathways controlling DC/NK cell-regulated responses to microbial products can profoundly affect the magnitude of innate immune responses. The Journal of Immunology, 2012, 188: 5094–5105.

he innate immune response serves as a first line of defense an innate immune response to PAMPs leads to leukocyte re- against invading pathogens (1). Pathogens are recognized cruitment to infected tissues and clearance of infection. Impor- T based on their expression of highly conserved pathogen- tantly, the innate immune response must be tightly regulated, as an associated molecular patterns (PAMPs). Innate immune cells of inefficient response could lead to an inability to clear infection, by guest on September 25, 2021 the mononuclear phagocyte system such as monocytes (Mos), whereas an enhanced or prolonged response could lead to auto- macrophages (Mf), and dendritic cells (DCs) recognize PAMPs immunity or shock and organ injury as occurs during sepsis (3). through germline-encoded pattern recognition receptors including Phagocyte recognition of microbial products is not only im- the TLR family (1, 2). LPS is a highly potent PAMP produced portant for initiating inflammatory responses but also for sculpting by Gram-negative bacteria such as Escherichia coli. LPS binds subsequent innate and adaptive responses. Crosstalk between DCs to TLR4 and activates a complex network of signal transduc- and NK cells is critical in both innate and T cell responses (4, 5). A tion cascades leading to cell activation and secretion of proin- large body of literature shows that NK cells require activation by flammatory mediators, including IL-1, IL-6, IL-12, type I IFNs, DCs to reach their full functional potential. In vitro, TLR- TNF-a, IFN-g, chemokines, and NO (3). Appropriate execution of activated DCs secrete IFN-a, IL-12, IL-15 (trans-presented on

*Department of Microbiology and Immunology, I3 Research Group, Life Sciences D.M.T. hold fellowships from the National Health and Medical Research Council of Centre, University of British Columbia, Vancouver, British Columbia V6T 1Z3, Australia. M.K.C. held studentships from the Michael Smith Foundation and the Canada; †Division of Immunology, The Walter and Eliza Hall Institute for Medical Canadian Institutes of Health Research. Research, Parkville, Victoria 3052, Australia; ‡Signal Transduction Laboratory, Lud- Address correspondence and reprint requests to Dr. Kenneth W. Harder, Life Sciences wig Institute for Cancer Research, Royal Melbourne Hospital, Parkville, Victoria x Centre, 2350 Health Sciences Mall, Vancouver, BC V6T 1Z3, Canada. E-mail ad- 3050, Australia; Child and Family Research Institute, Vancouver, British Columbia { dress: [email protected] V5Z 4H4, Canada; Department of Medicine, University of Melbourne, Parkville, Victoria 3010, Australia; and ‖Department of Pharmacology, University of Mel- The online version of this article contains supplemental material. bourne, Parkville, Victoria 3010, Australia Abbreviations used in this article: BALF, bronchoalveolar lavage ﬂuid; BM, bone 1D.L.K. and M.K.C. contributed equally to this work. marrow; BMDC, bone marrow-derived dendritic cell; BMMf, bone marrow-derived macrophage; DC, dendritic cell; DTx, diphtheria toxin; eGFP, enhanced GFP; 2M.L.H. and K.W.H. are cosenior authors. FLT3L-BMDC, dendritic cell derived from bone marrow in FLT3L; GM-BMDC, 3Current address: Department of Immunology, Monash University, Melbourne, Vic- dendritic cell derived from bone marrow in GM-CSF; iNOS, inducible NO synthase; toria, Australia. Mf, macrophage; MFI, mean ﬂuorescence intensity; MHC-I, MHC class I; MHC-II, MHC class II; Mo, monocyte; ODN, oligodeoxynucleotide; PAMP, pathogen- Received for publication November 29, 2011. Accepted for publication March 14, associated molecular pattern; pDC, plasmacytoid dendritic cell; qPCR, real-time 2012. quantitative PCR; RT, reverse transcription; SFK, Src family tyrosine kinase; SHP, This work was supported by grants from the Canadian Institutes of Health Research Src homology 2 domain-containing phosphatase; shRNA, short hairpin RNA; WT, (CIHMOP-86694 to K.W.H.), the Canadian Breast Cancer Foundation (to K.W.H. wild-type. and D.L.K.), the National Health and Medical Research Council of Australia (to K.W.H. and M.L.H.), and the Cooperative Research Centre for Cellular Growth Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 Factors (to M.L.H.). K.W.H. holds a Tier II Canada Research Chair and was supported by a fellowship from the National Cancer Institute of Canada. M.L.H. and www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103395 The Journal of Immunology 5095

IL-15 receptor a-chain), and IL-18, which induce NK cell IFN-g Staphylococcus aureus) were from InvivoGen. ELISA kits for TNF-a, production and cytotoxicity (6–9). Recently, a mouse model of IL-1a, IL-6, IFN-g, IL-12, and IL-10 were from eBioscience or BD inducible DC ablation was used to show that production of type I Biosciences; the ELISA kit for IL-1b was from R&D Systems; and the high ELISA kit for IFN-a was from PBL InterferonSource. Anti-NK1.1 Ab IFNs and IL-15 by CD11c DCs is required for NK cell priming (PK136) was from the Biomedical Research Centre (Vancouver, BC). in response to pathogens or their products (10). Lympholyte-M was from Cedarlane Laboratories. Ab-coupled magnetic Lyn is a member of the Src family of membrane-associated, beads were from Miltenyi Biotec. IL-15 was from Invitrogen. GM-CSF, nonreceptor protein tyrosine kinases (SFKs) and is expressed in FLT3L, and L cell-conditioned medium (a source of M-CSF) were produced and standardized in our laboratories. all leukocytes except T cells (11). Lyn is activated in response to a wide range of receptors and/or stimuli, including the BCR, CD40, Mice LPS, cytokines, and integrins. Interestingly, Lyn can act as either 2 2 Rag12/2 (33), Lyn / (16), Lynup/up (21), CD11c-DTR/GFP (34), SHIP- a positive or a negative regulator of signal transduction, depend- 12/2 (35), and MeV/MeV (36) mice have been previously described. Mice ing on the type of stimulus, developmental stage of the cell and were C57BL/6 generation .10 genetic background except SHIP-12/2, extracellular environment. Negative regulation occurs when Lyn which were mixed 129Ola 3 C57BL/6. Animal experimentation was phosphorylates tyrosine residues within ITIMs present in inhibitory performed in accordance with the University of British Columbia Animal proteins such as CD22, FcgRIIB1, PIR-B, and SIRPa that are Care Committee, the Canadian Council of Animal Care, and/or the Ludwig Institute for Cancer Research and the Department of Surgery Animal localized at the plasma membrane. ITIM phosphorylation gen- Ethics Committee following National Health and Medical Research erates docking sites for the SH2 domains of phosphatases such as Council of Australia guidelines. Mice were age-, sex-, and weight-matched Src homology 2 domain-containing phosphatase (SHP)-1, SHP-2, and were analyzed between 7 and 12 wk age. and SHIP-1, which are recruited to the plasma membrane where Downloaded from Cell culture they antagonize/modulate signaling. In mast cells, Lyn phosphorylates the membrane protein Cbp, leading to signal inhibition and Complete medium consisted of RPMI 1640 (Invitrogen) supplemented with reduction in degranulation (12). Alternatively, positive regulation 10% heat-inactivated FCS (Invitrogen), 100 U/ml penicillin G, 100 mg/ml streptomycin, 2 mM GlutaMAX (Invitrogen) and 10 mM 2-ME. GM-CSF occurs when Lyn phosphorylates ITAMs on membrane proteins BMDCs (GM-BMDCs) were derived from femoral cells as described (37) such as Iga/b and CD19, which recruit proteins such as Syk and and cultured in complete medium supplemented with GM-CSF (DC me- phospholipase Cg2 that amplify signaling (11, 13–15). dium). BM-derived macrophages (BMMfs) were derived in complete http://www.jimmunol.org/ Both Lyn loss- and gain-of-function mouse models (Lyn2/2 and medium supplemented with L cell-conditioned medium (a source of Lynup/up) have been instrumental in revealing the physiological M-CSF, produced in-house). FLT3 ligand-derived DCs (FLT3L-BMDCs) 2/2 were derived in complete medium supplemented with FLT3 ligand (pro- roles of Lyn. Lyn mice develop a progressive, lethal, lupus-like duced in-house). In all cases, cells were used after 8–10 d culture. NK cells disease characterized by accumulation of self-reactive Abs and were purified from spleens using biotinylated anti-NK1.1 mAb together nephritis (13, 16, 17) that is partially MyD88-dependent (18) and with anti-biotin magnetic beads and were cultured in complete medium involves overproduction of IL-6 (19) and BAFF (20). Lyn2/2 mice supplemented with 50 ng/ml IL-15 for 7 d. also exhibit age-dependent expansion of myeloid cells and their LPS-induced cytokine production and morbidity progenitors, leading to splenomegaly and myeloid neoplasia (21, 22). Lynup/up mice contain a gain-of-function mutation (Y508F) at Mice were injected i.p. with LPS as indicated. To deplete DCs, CD11c- by guest on September 25, 2021 DTR/GFP background mice were injected i.p. with 5 ng/g diphtheria the endogenous Lyn locus, resulting in a constitutively active toxin (DTx) 24 h (for morbidity experiments) or 48 h (for cytokine pro- version of Lyn that is expressed in a spatially and temporally duction experiments) before LPS injection. To deplete NK cells, mice were 2 2 appropriate manner (21). Together with Lyn / mice, these mice injected i.p. with 200 mg anti-NK1.1 Ab, both 48 and 24 h before LPS have demonstrated the critical role for Lyn in B cell biology and injection. Cytokines were determined by ELISA analysis of serum from have helped define Lyn-regulated signal transduction pathways. blood obtained by cardiac puncture. To determine NK cell or T cell IFN-g levels, splenocytes were stained extracellularly for NK1.1 and DX5 (NK LPS activates Lyn in DCs, Mos, and Mfs (23–26), and studies cells) or for Thy1 and CD4/CD8 (T cells). Splenocytes were then fixed using pan-SFK inhibitors have implicated SFKs in LPS-induced and permeablized with 2% paraformaldehyde/0.2% Tween-20 in PBS and responses in DCs (24, 27) and Mfs (28–30). We and others have stained intracellularly with anti–IFN-g or isotype control Abs. To assess shown that Lyn2/2 bone marrow-derived DCs (BMDCs) are im- morbidity, mice were carefully monitored hourly and sacrificed when moribund. Alternatively, core body temperature was measured using an paired with respect to LPS-induced maturation and IL-12 secre- ELAMS s.c. transponder system (BioMedic Data Systems). tion, suggesting a positive regulatory role for Lyn in LPS-induced DC activation (31, 32). However, to our knowledge, the role Transnasal LPS challenge and bronchoalveolar lavage fluid played by Lyn in endotoxin responsiveness in vivo has not been extraction 2/2 up/up directly investigated. In this study, we used Lyn and Lyn LPS (10 mg) was introduced transnasally into the lungs of anesthetized mouse models to show that Lyn activity dramatically lowered the mice (38). Cells and differential counts were assessed according to stan- threshold of endotoxin sensitivity in vivo, such that LPS-induced dard morphological criteria on Diff-Quick (Dade Behring)-stained cyto- hypothermia and morbidity occurred in Lynup/up mice at a dose spins of bronchoalveolar lavage fluid (BALF). Total and viable cell counts well tolerated in control wild-type (WT) mice. We dissected the were determined using ethidium bromide/acridine orange staining. up/up cell types contributing to LPS hypersensitivity in Lyn mice Cell stimulations and lysis and found that DCs and NK cells were required. Specifically, Lyn enhanced endotoxin-induced DC maturation and modified DC For real-time quantitative PCR (qPCR) and Western blotting experiments, cells were starved for 3 h in DC medium containing 0.5% serum, prior to cytokine expression profiles, leading to elevated NK cell IFN-g stimulation with 100 ng/ml LPS. For ELISA experiments, cells were not production. Taken together, these results show that Lyn is not only starved. For Western blots, cells were lysed on ice and insoluble material important in adaptive immunity, but is also critical in controlling was removed by centrifugation. the magnitude and nature of innate immune responses. NK cell coculture assays Materials and Methods The indicated number of in vitro-propagated NK cells were cocultured with 5 General reagents 5.0 3 10 GM-BMDCs, FLT3L-BMDCs (either sorted or unsorted populations), or BMMfs. Cells were stimulated with LPS (range, 1–100 ng/ LPS (from E. coli 0111:B4), diphtheria toxin, and polybrene were from ml) for 16 h, and cytokine levels in culture medium were determined by Sigma-Aldrich. CpG-oligodeoxynucleotide ODN and peptidoglycan (from ELISA. 5096 Lyn REGULATES THE INNATE IMMUNE RESPONSE qPCR Total RNA was isolated using the RNeasy kit (Qiagen). Reverse transcription (RT) was performed using the SuperScript first-strand synthesis kit (Invitrogen). qPCR was performed using SsoFast EvaGreen real-time PCR mix (Bio-Rad) on a Bio-Rad CFX96 instrument, and analyzed using CFX Manager software. Target genes were normalized to GAPDH. Flow cytometry Splenocytes were incubated with 2.4G2 mAb (Fc block) and stained with fluorochome-labeled Abs. Data were acquired on a FACSCalibur or LSRII instrument using CellQuest or FACSDiva software and analyzed using FlowJo software. Dead cells were excluded by propidium iodide uptake and red cells were excluded by size. Design and subcloning of short hairpin RNAs targeting SHP-1 and SHIP-1 Short hairpin RNAs (shRNAs) targeting SHP-1, SHIP-1, or control shRNAs were designed using an in-house algorithm based on characteristics of effective small interfering RNAs (39) (sequences available upon request). Three shRNAs (minimum) were designed for each target, and specificity was assessed by searching the National Center for Biotechnology Infor- Downloaded from mation Expressed Sequence Tags database. Annealed oligonucleotides were directionally subcloned into the Pme1/Mfe1-digested pQCXIP vector (BD Biosciences/Clontech) containing the mU6 promoter, puromycin resistance gene, and enhanced GFP (eGFP) reporter gene. To generate ret- 6 rovirus, 3.5 3 10 Phoenix cells were transfected using the CaPO4 method with 21 mg pCL-Eco and 21 mg pQCXIPmU6-shRNA-eGFP plasmids.

Retroviral infection of DCs http://www.jimmunol.org/ For infections, 2 3 105 BM cells were cultured in 1 ml DC medium for 2 d. One milliliter of fresh retrovirus-containing medium containing 8 mg/ml polybrene was added, and cells were centrifuged at 800 3 g for 2 h. A second infection was performed as above, a day later. Cells were replated in DC medium. For biochemical analyses, infected DCs (day 6) were treated with puromycin (1 mg/ml) for 48 h, and dead cells were removed. For NK cell coculture assays, infected DCs were sorted based on eGFP fluorescence. For flow cytometry, infected DCs were gated based on eGFP fluorescence and examined for LPS-induced maturation marker expression. FIGURE 1. A by guest on September 25, 2021 Flow cytometry Abs Lyn activity regulates endotoxin sensitivity in mice. ( ) Lyn+/+ and Lynup/up mice were injected i.p. with 25 mg LPS, were moni- Abs recognizing CD4, CD8, SIRPa, and CD45RA were from BD Bio- tored carefully for 24 h, and were sacrificed when moribund. (B)Lyn+/+ and sciences. Abs recognizing Thy1, IFN-g, DX5, NK1.1, CD24, CD11b, Lynup/up mice were injected i.p. with 50 mg LPS and core body temperature CD11c, MHC class II (MHC-II), and B220 were from eBioscience. was measured for 10 h. (C)Lyn+/+ and Lyn+/up mice were injected with PBS D Western blotting Abs (0) or 100 mg LPS, and core body temperature was measured after 10 h. ( ) Lyn+/+ and Lyn2/2 mice were treated as in (C). The LPS dose was increased Abs from Cell Signaling Technology include p-p38 (9215), p-Akt (Thr308; incrementally until a 8–10˚C temperature drop in 10 h was achieved. This 473 4056), p-Akt (Ser ; 4058), p-Erk1/2 (4377), p-Jnk (9251), p-Ikka/b required 4000 mgLPS.(E)Lyn+/+ and Lynup/up mice were challenged (2681), p-p65RelA (3033), p-IkB (9246), p-Gsk3a/b (9331), p-Bad transnasally with PBS (0) or 10 mg LPS. After 0, 2, or 24 h BALF was (9291), p-Tbk1 (5483), P-S6 (4856), total-IkBa (9242), total-p65 (3034), collected. Total and differential cell counts and TNF-a levels were total-Akt (9272), and total-Erk1/2 (4695). Anti–SHP-1 was from Dr. H.-C. A D E Cheng (University of Melbourne, Melbourne, VIC, Australia); anti–SHP-2 assessed. For ( )–( ), n $ 5 mice/group; for ( ), n = 8 mice/group. All was from Santa Cruz Biotechnologies. Anti–P-Tyr (clone 4G10) was from experiments were performed a minimum of three times. Error bars repre- Millipore; anti-inducible NO synthase (iNOS) was from BD Biosciences. sent SD (C and D) and SEM (E). *p # 0.05, **p # 0.01, ***p # 0.001. Statistical analysis All statistical analyses were performed using GraphPad Prism (GraphPad sepsis is associated with a drop in core body temperature in Software). Student t tests were used to analyze significance of data sets, rodents (41), we injected Lyn+/+ and Lynup/up mice with 50 mg which are represented as arithmetic means 6 SEM or SD as stated. The p values indicated are as follows: *p # 0.05, **p # 0.01, ***p # 0.001. LPS and monitored core body temperature over time using s.c. transponders. Body temperature in Lynup/up mice dropped pro- gressively, with a loss of 10–12˚C after 10 h exposure to LPS. In Results +/+ up/up contrast, Lyn mice exhibited no significant temperature change Lyn mice exhibit enhanced endotoxin-induced (Fig. 1B). Mice carrying one Lynup allele (Lyn+/up) were also inflammation and morbidity hyperresponsive to LPS when compared with Lyn+/+ mice, but Injection of LPS into mice mimics pathophysiological conse- they required a higher dose of LPS than did Lynup/up mice to quences of sepsis, including tissue necrosis, multiorgan failure, and stimulate a 10˚C drop in body temperature (Fig. 1C). Lynup/up ultimately death (40). We therefore used LPS injection to inves- mice were ∼80-fold more sensitive to LPS than Lyn+/+ or Lyn2/2 tigate whether Lyn regulates endotoxin sensitivity in vivo. Lyn+/+ mice, as Lyn+/+ and Lyn2/2 mice required 4000 mg LPS to induce and Lynup/up mice were injected with 1 mg/g LPS (25 mg/mouse) a 10˚C drop in body temperature, whereas only 50 mg LPS was and monitored for 24 h. Interestingly, whereas all Lyn+/+ mice required in Lynup/up mice (Fig. 1D). appeared grossly unaffected, .60% of Lynup/up mice became To further assess the role of Lyn in LPS-induced inflammation, moribund between 12 and 24 h postinjection (Fig. 1A). Given that we employed an acute lung inflammation model in which BALF is The Journal of Immunology 5097 extracted and analyzed following inhalation of LPS (42). Lynup/up gated how Lyn modulated IFN-g production in LPS-treated mice. mice exhibited enhanced cellularity in pulmonary exudates in NK cells are the primary producers of IFN-g in response to LPS; response to inhalation of 10 mg LPS, an effect that was detectable however, CD4+ and CD8+ T cells and NKT cells can also produce at 2 h postinhalation, but was more dramatic after 24 h. Indeed, at IFN-g (46). We found that the NK cell surface marker expression 24 h postinhalation, Lynup/up BALF contained ∼6-fold more Mfs, profile was similar in Lyn+/+ and Lynup/up spleens and livers (Table 2.5-fold more neutrophils, and 25-fold more lymphocytes than did I), as was the absolute number of NK cells (Lyn+/+, 1.3 3 106 6 Lyn+/+ BALF. Whereas eosinophils were virtually undetectable in 9 3 104;Lynup/up, 1.1 3 106 6 2 3 105). Absolute numbers of Lyn+/+ BALF, .12 3 104 were present in Lynup/up BALF fol- CD4+ and CD8+ T cells were also similar in Lyn+/+ and Lynup/up lowing LPS inhalation. Levels of TNF-a were also elevated ∼7.5- spleens, suggesting that an overabundance of T or NK cells was fold in Lynup/up BALF at 2 h postinhalation (Fig. 1E). not responsible for excessive IFN-g production in Lynup/up mice (not shown). The increased frequency of Lynup/up NK cells (Fig. Lyn promotes production of proinflammatory cytokines in 3A) reflected the diminution of B cells in these spleens (47). response to endotoxin We next assessed the source of IFN-g and found that whereas In humans and mice, endotoxin-induced mortality is associated Lynup/up CD4+ and CD8+ T cells showed little LPS-induced IFN- with excessive production of proinflammatory cytokines (40). To g, ∼85% of splenic NK cells in Lynup/up mice expressed IFN-g investigate whether Lyn regulated endotoxin-induced cytokine compared with 20% in Lyn+/+ mice. Lynup/up NK cells also ex- production, we injected Lyn+/+ and Lynup/up mice with LPS and pressed more IFN-g per cell as indicated by mean fluorescence assessed cytokines in serum over time by ELISA. Lynup/up serum intensity (MFI), suggesting that NK cells were the primary pro- contained significantly higher concentrations of the proinflam- ducers of IFN-g in LPS-injected Lynup/up mice (Fig. 3A). We Downloaded from matory cytokines IL-12, IFN-g, IL-1a, IL-1b, TNF-a, IFN-a, and further showed that B, T, or NKT cells did not contribute signif- IL-6 and lower concentrations of the anti-inflammatory cytokine icantly to LPS-induced morbidity or NK cell IFN-g production, as IL-10 (Fig. 2A). Cytokine arrays also revealed heightened LPS- Lynup/up mice on a Rag2/2 genetic background also exhibited induced production of proinflammatory cytokines and chemo- exaggerated LPS-induced NK cell IFN-g production and mor- kines, as well as modestly lower production of IL-4 and IL-10 in bidity (Fig. 3B, 3C). up/up

Lyn serum (Supplemental Fig. 1). Western blot analysis http://www.jimmunol.org/ GM-BMDCs or FLT3L-BMDCs, but not BMMfs, activate NK showed that LPS-induced expression of the proinﬂammatory me- cells in response to LPS diator iNOS was also elevated in the kidney, spleen, lung, and liver of Lynup/up mice within 6 h of LPS injection (Fig. 2B). NK cells do not exist in the periphery in an activated state, but Consistent with elevated cytokine production, myeloid cells in instead must be stimulated by DCs or Mfs to acquire effector Lynup/up BM exhibited enhanced expression of MHC class I functions such as IFN-g production (10, 48). In light of this, we (MHC-I) and Mac-1 after LPS injection (Fig. 2C). tested whether Mfs or different DC subsets contributed to excessive NK cell IFN-g production in LPS-stimulated Lynup/up mice. NK cells are the primary source of IFN-g in LPS-challenged up/up Puriﬁed NK cells were cocultured with GM-BMDCs, FLT3L-

Lyn mice BMDCs, or BMMfs. DC/NK cell or Mf/NK cell cocultures by guest on September 25, 2021 IFN-g production in response to LPS is directly linked to the were then stimulated with increasing doses of LPS, after which magnitude of innate immune responses in vivo (43–45), and IFN-g levels in the culture medium were assessed. We found that blocking IFN-g with neutralizing anti–IFN-g Abs can prevent GM-BMDCs and FLT3L-BMDCs, but not BMMfs, were effective fatal LPS-induced septicemia in mice (46). We therefore investi- at inducing NK cells to secrete IFN-g in response to LPS (Fig. 4A).

FIGURE 2. Lynup/up mice exhibit enhanced inﬂammatory cytokine production in response to LPS. (A)Lyn+/+ and Lynup/up mice (n =5 mice/group) were injected i.p. with PBS (0) or 50 mg LPS. Serum cytokine concentrations were then determined by ELISA after 0, 1.5, 3, and 6 h. Error bars indicate SD. (B)At 0 and 6 h postinjection (+LPS), kidney, spleen, lung, and liver were removed and lysates were subjected to Western blot analysis for iNOS andErk1/2(loadingcontrol).(C)At6h postinjection BM was analyzed by ﬂow cytometry to assess upregulation of MHC-I and Mac1. The MFI (x- and y-coordinates) for the Mac1+MHC-I+ populations are shown. All experiments were performed a minimum of three times. *p # 0.05, **p # 0.01. 5098 Lyn REGULATES THE INNATE IMMUNE RESPONSE

2 Table I. Expression of NK cell markers on NK1.1+CD49b+ cells from CD24highCD4 CD8+CD11blow DCs, and CD45RA+ plasmacytoid +/+ up/up Lyn and Lyn spleen and liver DCs (pDCs) (Fig. 4B, boxes 1–3). Sorting of FLT3L-BMDCs into these subsets revealed that SIRPa+CD24lowCD42CD82CD11bhigh Spleen Liver DCs were the most potent activators of NK cell IFN-g production Surface Ag Lyn+/+ (%) Lynup/up (%) Lyn+/+ (%) Lynup/up (%) whereas pDCs were ineffective. DCs or NK cells did not produce IFN-g when cultured alone or without LPS (Fig. 4A, 4B). Ly49A 22.6 6 1.4 23.0 6 4.0 ND ND Ly49C+I 58.0 6 1 52.0 6 3.8 53 53 The DC lineage is required for enhanced NK cell IFN-g Ly49D 56.1 6 2.5 41.0 6 3.1 53 49 up/up Ly49G2 48.7 6 2.4 51.3 6 3.7 ND ND production in Lyn mice Ly49F 13.1 6 0.2 15.7 6 2.7 ND ND Our results suggested that DCs might be required for the excessive 6 6 FcRgII 82.4 0.1 79.8 3.9 ND ND NK cell IFN-g production in Lynup/up mice. To investigate this CD94 51.0 6 1.9 50.3 6 1.4 ND ND NKG2A/C/E 53.9 6 2.5 49.3 6 1.9 ND ND possibility, we took advantage of the CD11c-DTR/GFP mouse NKG2D 97.3 6 1.2 97.3 6 1.8 90 91 model, where a DTx receptor-GFP fusion protein is expressed under KLRG1 45.0 6 2.1 51.0 6 2.4 60 72 the control of the murine CD11c promoter (34). Injection of DTx 6 6 CD27 49.3 3.8 46.3 6.2 47 40 into CD11c-DTR/GFP mice leads to ablation of CD11c-DTR/GFP+ 2B4 98.3 6 0.9 98.0 6 0.8 ND ND up/up 6 6 cells (primarily DCs). We generated Lyn mice on the CD11c- MAC-1 96.1 0.6 91.0 2.0 96 97 +/+ up/up CD43 98.3 6 0.5 98.7 6 0.9 ND ND DTR/GFP background. Lyn - and Lyn -CD11c-DTR/GFP mice were then injected with DTx or PBS, and LPS-induced IFN- ND, Not detected. g production in both NK cells and serum was assessed. Following Downloaded from LPS injection without DC ablation, Lynup/up-CD11c-DTR/GFP mice FLT3L-BMDCs are a heterogeneous population and can be contained signiﬁcantly more IFN-g+ NK cells and elevated serum further subdivided into at least three other subpopulations, rep- +/+ + low 2 2 high 2 IFN-g levels compared with Lyn -CD11c-DTR/GFP mice. In resenting SIRPa CD24 CD4 CD8 CD11b DCs, SIRPa contrast, in both Lyn+/+- and Lynup/up-CD11c-DTR/GFP mice, DC- ablated mice injected with LPS exhibited a reduction in production

of IFN-g in both NK cells and serum. Strikingly, LPS-induced IFN-g http://www.jimmunol.org/ production in Lynup/up-CD11c-DTR/GFP mice was almost com- pletely eliminated by DC ablation, suggesting that DCs are required for enhanced IFN-g production in Lynup/up mice (Fig. 4C, 4D). DCs and NK cells are required for LPS-induced morbidity in Lynup/up mice Given the importance of IFN-g in LPS-induced septicemia (46), we investigated whether DCs and NK cells might also be required for LPS-induced morbidity in Lynup/up mice. Interestingly, 75% of by guest on September 25, 2021 Lynup/up mice on the CD11c-DTR/GFP background were rescued from LPS-induced morbidity following depletion of DCs (Fig. 4E). Similarly, depletion of NK cells using anti-NK1.1 Abs rescued 100% of Lynup/up mice from LPS-induced morbidity (Fig. 4F). These results demonstrate that the DC/NK cell axis is essential for LPS hyperresponsiveness in Lynup/up mice. Lyn activity within DCs enhances endotoxin-induced IFN-g production in NK cells Having demonstrated the importance of DCs in heightened LPS-induced IFN-g responses in Lynup/up NK cells in vivo, we sought to dissect this phenomenon in vitro. To this end, we cocultured Lyn+/+,Lyn2/2,orLynup/up GM-BMDCs with WT NK cells and measured NK cell IFN-g production in response to LPS. We found a positive correlation between Lyn activity within DCs FIGURE 3. NK cells are the predominant source of IFN-g in LPS- 2/2 stimulated mice. (A)Lyn+/+ and Lynup/up mice were injected i.p. with and their ability to induce NK cell IFN-g production, with Lyn , +/+ , up/up sterile PBS (2LPS) or 50 mg LPS (+LPS). Six hours later spleens were Lyn Lyn DCs (Fig. 5A). The same trend was observed harvested and stained with mAbs defining NK cells (NK1.1+DX5+), CD4+ when whole E. coli or Listeria monocytogenes were used as T cells (Thy1+CD4+), or CD8+ T cells (Thy1+CD8+). Splenocytes were a stimulus instead of LPS, with higher activity of Lyn within DCs then fixed, permeablized, and stained with anti–IFN-g mAbs. Upper panel, associated with enhanced NK cell IFN-g production (Supple- DX5+ cells from Lyn+/+ and Lynup/up spleens were gated and analyzed for mental Fig. 2A). Importantly, Lyn2/2,Lyn+/+, and Lynup/up NK NK1.1 and IFN-g expression. Numbers indicate the percentage of IFN-g– cells produced similar amounts of IFN-g in response to stimula- expressing NK cells or the MFI of this population (6SD). Lower panel, tion with IL-15, IL-12, IL-18, and anti-NK1.1, suggesting that + up/up Thy1 cells from Lyn spleens were gated and analyzed for either CD4 up/up 2/2 2/2 up/up Lyn NK cells were not intrinsically hyperresponsive (Sup- or CD8, and IFN-g.(B) Rag and Rag Lyn mice were injected i. up/up 2 plemental Table I). Indeed, Lyn NK cells secreted less IFN-g p. with PBS ( LPS) or 25 mg LPS (+LPS). After 6 h splenocytes were +/+ A when cocultured with WT DCs than did Lyn NK cells, whereas analyzed by flow cytometry to assess NK cell IFN-g production as in ( ). 2/2 The percentage 6 SD and MFI for the NK1.1+IFN-g+ population is shown. Lyn NK cells showed a subtle enhancement of IFN-g pro- (C) Mice were injected i.p. with 10 mg LPS, were carefully monitored for duction (Fig. 5B). Taken together, our results suggest that Lyn 24 h, and were sacrificed when moribund. For all experiments, n $ 4 mice/ activity within DCs and not NK cells is important for excessive group and experiments were performed a minimum of three times. NK cell IFN-g production in LPS-injected Lynup/up mice. The Journal of Immunology 5099 Downloaded from

FIGURE 4. IFN-g production and hypersensitivity to LPS in Lynup/up mice depends on DCs and NK cells. (A) BMDCs derived in either GMCSF or FLT3L, or BMMfs, were cocultured with NK cells (5.0 3 105 DCs/Mfs with 2.0 3 105 NK cells). Cell mixtures were stimulated with 0, 1, 10, or 100 ng/ ml LPS for 16 h. As controls, DCs (2NK) and NK cells were cultured alone and stimulated with 100 ng/ml LPS. Culture medium was collected and IFN-g B levels were determined by ELISA. The p values indicate a comparison with BMMfs. ( ) FLT3L-BMDCs were sorted into three populations (upper right): http://www.jimmunol.org/ (SIRPa+CD24low [CD42CD82CD11bhigh]) (box 1), (SIRPa2CD24high [CD42CD8+CD11blow]) (box 2), and CD45RA+ pDCs (box 3). Populations were cocultured with NK cells and stimulated as in (A). (C)Lyn+/+ and Lynup/up mice on the CD11c-DTR/GFP genetic background were injected i.p. with PBS (2DTx) or diphtheria toxin (+DTx; 5 ng/g mouse). Forty-eight hours later, mice were injected i.p. with 25 mg LPS. After 6 h, splenocytes were analyzed by ﬂow cytometry to assess NK cell IFN-g production. The percentage and MFI of the NK1.1+IFN-g+ population 6 SD is shown. (D) Mice were treated as in (C), except that cardiac punctures were performed to collect blood and serum IFN-g levels were determined by ELISA. (E)Lyn+/+ and Lynup/up mice on the CD11c-DTR/GFP genetic background were injected i.p. with PBS or DTx (5 ng/g mouse). Twenty-four hours later, mice were injected with 10 mg LPS, were monitored carefully for 100 h, and were sacriﬁced when moribund. (F)Lyn+/+ and Lynup/up mice were injected twice over 48 h with an anti-NK1.1 mAb or an isotype control. Twenty-four hours after the second injection, mice were injected i.p. with 10 mg LPS and were monitored as in (E). Error bars indicate SD, and all experiments were performed a minimum of two times. For (C)–(F), n $ 4 mice/group. *p # 0.05, **p # 0.01, ***p # 0.001. by guest on September 25, 2021

Lyn activity within DCs modulates LPS-induced cytokine DCs to NK cells, DCs were cocultured with increasing numbers of production WT NK cells and stimulated with LPS. As was the case in the absence of NK cells, Lynup/up DCs produced more IL-12 and less DCs prime NK cells through ligation of activating receptors and 2/2 secretion of cytokines such as IL-12, IL-18, type I IFNs, and IL-15. IL-10, whereas Lyn DCs produced more IL-10 and less IL-12 DCs also modulate NK cell activation by secreting the inhibitory when compared with WT DCs. In all three DC genotypes, NK cytokines IL-10 and TGF-b (6, 10, 49). To explore whether Lyn cells further stimulated IL-12 production and inhibited IL-10 activity within DCs modulates their ability to produce IL-12 and/ production, suggesting that NK cells enhance the ability of Lyn or IL-10, we incubated Lyn2/2,Lyn+/+, and Lynup/up GM-BMDCs to polarize cytokine production (Fig. 5E). with increasing doses of LPS and assessed IL-12 and IL-10 levels Lyn activity stimulates DC maturation in the culture medium by ELISA. We found that Lynup/up DCs up/up exhibited enhanced LPS-induced IL-12 production and reduced Our ﬁnding that Lyn DCs exhibited enhanced proinﬂam- IL-10 production when compared with Lyn+/+ DCs. In contrast, matory cytokine production and NK cell activation suggested that Lyn2/2 DCs exhibited enhanced LPS-induced production of IL-10 Lyn might promote DC maturation. To assess DC maturation, we and, as previously reported, decreased IL-12 production (32) (Fig. stimulated GM-BMDCs with LPS (a TLR4 ligand), CpG-ODN (a 5C). The same trend was observed when whole E. coli or L. TLR9 ligand), or peptidoglycan (a TLR2/6 ligand) and assessed monocytogenes was used as a stimulus, with higher activity of Lyn expression of MHC-II. In both unstimulated and stimulated DCs, 2/2 associated with enhanced IL-12 and reduced IL-10 production a higher frequency of Lyn DCs and a lower frequency of up/up low (Supplemental Fig. 2B and not shown). We further investigated Lyn DCs fell into the MHC-II population when compared whether cytokine production was dysregulated in Lyn2/2 and with Lyn+/+ DCs. In contrast, a higher frequency of Lynup/up DCs 2 2 Lynup/up DCs by stimulating GM-BMDCs with LPS and moni- and a lower frequency of Lyn / DCs fell into the MHC-IIhigh toring the production of cytokine mRNAs over time by RT-qPCR. population. We observed the same trend when assessing CD40, We found that expression of mRNAs encoding IL-12p35, IL- CD80, and CD86, with Lynup/up DCs exhibiting enhanced matu- 2 2 12p40, IL-15, and IFN-b were increased in Lynup/up DCs and ration and Lyn / DCs exhibiting diminished maturation (not decreased in Lyn2/2 DCs when compared with Lyn+/+ DCs (Fig. shown). FLT3L-BMDCs did not show the same polarity, although 2 2 5D). The expression of IL-6 and TNF-a mRNA was similar in the Lyn / DCs expressed modestly lower MHC-II (Fig. 6A). three genotypes of DCs (not shown). In vivo, the absolute number of DCs was similar in Lyn+/+, 2 2 Coculture of NK cells with DCs can enhance DC activation, due Lyn / ,andLynup/up spleens (Lyn+/+,2.53 106 6 4.4 3 105; 2 2 to cell interactions and production of cytokines including IFN-g Lyn / ,2.03 106 6 5 3 105;Lynup/up,3.13 106 6 7.1 3 105)as and HMGB1 (50, 51). To test the role of Lyn in the response of was the frequency of CD11c+CCR9+ pDCs and CD11c+CD8+ 5100 Lyn REGULATES THE INNATE IMMUNE RESPONSE Downloaded from http://www.jimmunol.org/

2 2 FIGURE 5. Lyn polarizes DCs with respect to LPS-induced cytokine production and stimulation of NK cells to secrete IFN-g.(A)Lyn+/+,Lyn / ,or by guest on September 25, 2021 Lynup/up GM-BMDCs (5 3 105) were cocultured with 2 3 105 WT NK cells. Cocultures were stimulated with 0, 1, 10, or 100 ng/ml LPS. Alternatively, DCs or NK cells were cultured alone and stimulated with 100 ng/ml LPS. After 16 h, culture medium was collected and IFN-g levels were determined by ELISA. (B)Lyn+/+,Lyn2/2,orLynup/up NK cells (2 3 105) were cocultured with 5 3 105 WT GM-BMDCs. Cocultures were stimulated as in (A), and IFN- g levels were determined by ELISA. (C)Lyn+/+,Lyn2/2,orLynup/up GM-BMDCs (5.0 3 105) were stimulated with 0, 1, or 100 ng/ml LPS. After 16 h, IL- 12 and IL-10 levels in the culture medium were measured by ELISA. (D)Lyn+/+,Lyn2/2, and Lynup/up GM-BMDCs were stimulated with 100 ng/ml LPS for the times shown. At each time point RNA was harvested and subjected to RT-qPCR to assess synthesis of cytokine mRNAs (normalized to GAPDH mRNA). (E)Lyn+/+,Lyn2/2,orLynup/up GM-BMDCs (2.5 3 105) were cocultured with 0, 1 3 104,or53 104 WT NK cells and were stimulated with or without 100 ng/ml LPS for 16 h. IL-10 and IL-12 concentrations in the culture medium were determined by ELISA. The p values indicate a comparison with WT DCs. For all experiments, error bars indicate SD and experiments were performed a minimum of three times. *p # 0.05, **p # 0.01, ***p # 0.001. conventional DCs. The frequency of CD11c+CD4+ conventional ting using phospho-speciﬁc Abs. Within the MAPK pathway, DCs was similar in Lyn+/+ and Lynup/up mice but was consistently, activation of p38, Erk1/2, and Jnk was moderately increased in albeit variably, lower in Lyn2/2 mice (Fig. 6B). Lyn2/2 CD11c+ Lyn2/2 DCs and decreased in Lynup/up DCs. The same trend was splenocytes expressed lower levels of CD80, as previously described observed in the PI3K pathway (52), as phosphorylation of Akt (at (19). In keeping with our in vitro data, we observed that Lyn pro- both Thr308 and Ser473) was elevated in Lyn2/2 DCs and dimin- moted endotoxin-induced DC maturation in LPS-injected mice, ished in Lynup/up DCs. The phosphorylation of Bad, S6, and with Lyn2/2 DCs expressing less MHC-II (indicated by MFI) and Gsk3a/b can be mediated by Akt, and we found that phosphor- Lynup/up DCs expressing more MHC-II compared with Lyn+/+ DCs ylation of these proteins was also increased in Lyn2/2 DCs and (Fig. 6C). Taken together, our results demonstrate that Lyn acti- decreased in Lynup/up DCs (Fig. 7). vity can enhance spontaneous and inducible DC maturation, both Within the NF-kB family, p50/p65 (RelA) is predominantly in vitro and in vivo. used in TLR signaling (53). In the canonical pathway, p50/p65 NF-kB is sequestered in the cytoplasm through its association Lyn negatively regulates LPS-induced signal transduction in with the inhibitory protein ikB. LPS stimulation activates the ikB DCs kinase complex (Ikka/b), which phosphorylates ikBa, targeting it We next investigated the mechanism by which Lyn enhanced LPS- for degradation. Once released from inhibition by ikBa, NF-kB induced maturation of DCs. Ligation of TLR4 activates signal translocates to the nucleus and initiates gene transcription. Be- transduction pathways regulated by the MAPKs, PI3K, and NF-kB. cause the ikBa gene is itself regulated by NF-kB, ikBa is rapidly To assess whether Lyn regulated these pathways, Lyn+/+,Lyn2/2, resynthesized following NF-kB activation (1, 54). Compared to and Lynup/up GM-BMDCs were stimulated with LPS for various Lyn+/+ DCs, Ikka/b activation was moderately increased in periods of time, and cell lysates were subjected to Western blot- Lyn2/2 DCs and was signiﬁcantly decreased in Lynup/up DCs. The Journal of Immunology 5101

full TLR-induced DC maturation and IL-12 secretion (57, 58). To test whether Lyn-mediated activation of SHIP-1, and perhaps SHP-1, could stimulate LPS-induced maturation of DCs, we used shRNAs to reduce SHP-1 or SHIP-1 expression in Lyn+/+,Lyn2/2, and Lynup/up GM-BMDCs (Fig. 8C) and assessed LPS-induced expression of MHC-II. Whereas SHP-1 knockdown did not significantly affect Lyn+/+ or Lyn2/2 DCs (Table II), it impaired both the spontaneous and LPS-induced maturation of Lynup/up DCs (Fig. 8D). However, we could not detect dysregulated signal transduction immediately after LPS stimulation following SHP-1 knockdown (Fig. 8E). SHIP-1 knockdown impaired the spontaneous and LPS-induced maturation of all three DC genotypes, although Lynup/up DCs were most significantly affected (Fig. 8D, Table II). Correspondingly, LPS-induced activation of the PI3K pathway was enhanced by SHIP-1 knockdown in Lynup/up DCs (Fig. 8E). Importantly, SHIP- 1 knockdown in Lynup/up DCs significantly impaired their ability to activate NK cells to produce IFN-g (Fig. 8F), highlighting the

correlation between the role of Lyn as a regulator of SHIP-1 ac- Downloaded from tivity and Lyn-dependent control of DC maturation and NK cell activation. Interestingly, GM-BMDCs derived from Mev/Mev (lacking SHP-1 activity) or SHIP-12/2 BM exhibited impaired spontaneous and LPS-induced MHC-II expression (Fig. 8G), and similar

trends were observed when other activation/maturation markers http://www.jimmunol.org/ were examined (not shown). Collectively, these data support the FIGURE 6. Lyn controls spontaneous and TLR-dependent DC matura- hypothesis that Lyn-induced activation of SHIP-1, and to a lesser tion. (A) DCs were derived from BM in either GM-CSF or FLT3L. Cells extent SHP-1, contributes to spontaneous and LPS-stimulated DC were left untreated or exposed to LPS, CpG-ODN, or peptidoglycan maturation and NK cell activation potential. (PGN). Sixteen hours later, DCs were stained for CD11c and MHC-II. + MHC-II expression on CD11c DCs fell into three regions: low, inter- Discussion mediate, and high. Percentages of cells in each region are shown. For SFKs display redundant functions, with multiple members often FLT3L-BMDCs the MFI for MHC-II is shown. (B) Splenocytes from Lyn+/+, Lyn2/2, and Lynup/up mice were analyzed by flow cytometry to assess the coexpressed within one cell type; however, SFKs can also play by guest on September 25, 2021 frequency of CD11c+CCR9+ pDCs and CD11c+CD4+ or CD11c+CD8+ indispensible roles (11). In this study, we addressed redundancy by up/up conventional DCs. Percentages of cells in each region 6 SD are shown. (C) utilizing a Lyn gain-of-function mouse model (Lyn mice), in The expression of CD86, CD80, and PDL1 on CD11c+ splenic DCs was which a constitutively active version of Lyn is expressed by the assessed as in (B)(upper panel). Lyn+/+,Lyn2/2, and Lynup/up mice were endogenous promoter (21). This model enabled further explora- injected with 50 mg LPS or sterile PBS (control). After 1.5, 3, and 6 h, tion of the role of Lyn in cells of the innate immune system, where spleens were analyzed by flow cytometry for MHC-II expression on other SFKs are often coexpressed and might compensate for the + CD11c cells (lower panel). The MFI for MHC-II is shown 6 SD. For all loss of Lyn. We found that Lynup/up mice were hypersensitive to time points, n $ 4 mice/group. All experiments were performed a mini- systemic LPS challenge, becoming hypothermic and moribund in mum of three times. **p # 0.01, comparison with Lyn+/+ cells. 2 2 response to a dose of LPS well tolerated by WT and Lyn / mice. LPS sensitivity in Lynup/up mice was associated with dramatically IkBa phosphorylation occurred more rapidly and was of a greater elevated levels of serum IL-1a, IL-1b, IFN-g, IL-12, TNF-a, IFN- magnitude in Lyn2/2 DCs, whereas Lynup/up DCs exhibited de- a, and iNOS and lower IL-10 levels. An enhanced innate immune up/up layed and diminished phosphorylation of ikBa. Correspondingly, response in Lyn mice was also evident when an airway model up/up ikBa steady-state levels reached a minimum in a rapid and pro- of acute inflammation was employed, with Lyn mice dem- longed manner in Lyn2/2 DCs, but showed a less significant onstrating enhanced pulmonary leukocyte recruitment and TNF-a 2/2 change in Lynup/up DCs. Phosphorylation of p65/RelA and Tbk1 production following inhalation of LPS. Interestingly, Lyn were also decreased in Lynup/up DCs (Fig. 7). mice were not resistant to LPS-induced morbidity, perhaps due to compensation by other Src family members such as Hck and Fgr, Lyn interacts functionally with SHIP-1 and SHP-1 in DCs which also contribute to endotoxin responsiveness (59, 60). Ad- To further elucidate Lyn-regulated signaling pathways, we con- ditionally, although we analyzed young mice (6–8 wk age), Lyn2/2 ducted anti-phosphotyrosine Western blots on lysates from mice exhibit modest myelomonocytic expansion at this age, GM-BMDCs stimulated with and without LPS. Proteins of m.w. complicating analyses of TLR responses. Interestingly, untreated ∼145,000 and ∼70,000 were constitutively hyperphosphorylated Lynup/up mice did not demonstrate elevated serum cytokine levels in Lynup/up DCs and hypophosphorylated in Lyn2/2 DCs (Fig. or evidence of inflammatory disease, suggesting that the Lynup 8A). Immunoprecipitation together with anti-phosphotyrosine mutation does not lead to chronic inflammation, as in other models Western blotting revealed that these phosphoproteins included of endotoxin hypersensitivity such as SOCS12/2 or SIGIRR2/2 SHP-1 (68 kDa) and SHIP-1 (145 kDa) (Fig. 8B). mice (61–63). Lyn can engage inhibitory signaling by tyrosine phosphorylating Given that IFN-g plays a key role in sensitizing mice to LPS ITIMs on membrane proteins, creating docking sites for the SH2 (64) and is linked to endotoxin-induced lethality (43–45, 65–67), domains of SHP-1 and SHIP-1 (14, 21, 32, 55, 56). Recent re- we focused on identifying mechanistically how Lyn regulates search using SHIP-12/2 DCs revealed that SHIP-1 is required for IFN-g production. We found that NK cells were the primary 5102 Lyn REGULATES THE INNATE IMMUNE RESPONSE Downloaded from http://www.jimmunol.org/

FIGURE 7. Lyn regulates LPS-induced signaling pathways in GM-BMDCs. Lyn+/+,Lyn2/2, and Lynup/up GM-BMDCs were stimulated with or without 100 ng/ml LPS for the times shown (minutes). Lysates were prepared and Western blotting was carried out to assess activation of signaling pathways

regulated by the MAPKs, PI3K, and NF-kB. Relative band intensities are shown below each band (determined using ImageJ software). All experiments by guest on September 25, 2021 were performed a minimum of three times. source of IFN-g production in LPS-injected WT and Lynup/up viously reported (31), with Lyn2/2 DCs exhibiting Th2-polarized mice, as reported previously for WT mice (46, 68). Furthermore, T cell responses in vitro and in vivo. depletion of NK cells rescued Lynup/up mice from LPS-induced Analysis of splenic DC maturation following systemic LPS morbidity, underscoring the requirement of NK cells for LPS challenge revealed that Lyn activity in DCs correlated with LPS- hypersensitivity in these mice. induced DC maturation potential. This characteristic was consis- Activation of NK cells in response to PAMPs is primarily evoked tent with our analyses of GM-BMDCs but not FLT3L-BMDCs, by myeloid cells. We therefore sought to identify the phagocyte which are considered more representative of DC populations re- population(s) responsible for enhanced NK activation in Lynup/up siding in the spleen. A more thorough analysis of the LPS responses mice. We found that populations of DCs, but not Mfs, were able of speciﬁc subsets of DCs within the FLT3L-BMDC cultures under to stimulate NK cells to produce IFN-g. Moreover, depletion of optimized culture conditions may be required to understand the role CD11c-expressing cells (predominantly DCs) in Lynup/up mice of Lyn in FLT3L-BMDCs. Lyn did not impact DC development in blocked the ability of LPS to induce NK cell IFN-g production, the spleen, as total DC numbers and frequencies of pDCs and CD8+ and it protected mice from LPS-induced morbidity. Taken to- DCs were similar in Lyn+/+,Lyn2/2,and Lynup/up mice. We did, gether, our results indicate that DCs are primarily responsible for however, detect a small but reproducible decrease in the splenic enhanced LPS-induced NK cell IFN-g production in Lynup/up CD4+ DC subset in Lyn2/2 mice. mice. These ﬁndings are in agreement with previous studies We found that LPS-induced signaling and cytokine production showing that DCs are required for TLR- and CD40-induced NK was relatively unaltered in Lyn2/2 and Lynup/up BMMfs, with cell IFN-g production and tumor cytotoxicity in WT mice, and Lyn+/+,Lyn2/2, and Lynup/up BMMfs secreting little IL-12 in that NK cell priming does not require Mf subpopulations or TLR response to LPS, and similar levels of TNF-a and IL-6. However, signaling in NK cells (10, 48). Lyn2/2 BMMfs secreted more IL-10 (not shown). Interestingly, In vitro, Lyn2/2 BMDCs matured poorly in response to LPS whereas a recent report showed that Lyn2/2 BMMfs produce or microbes and were poor stimulators of NK cell IFN-g produc- elevated levels of IL-6 and TNF-a in response to LPS (23), pre- tion, whereas Lynup/up DCs showed enhanced LPS-induced mat- vious work showed that WT and Hck2/2Fgr2/2Lyn2/2 BMMfs uration and were better able to stimulate NK cells to produce IFN- were similar with respect to signal transduction and cytokine se- g. Additionally, Lyn2/2 BMDCs produced less IL-12, IL-15, and cretion (69). These discrepancies could possibly be attributed to IFN-b, but more IL-10, in response to LPS or bacteria, whereas the different methods used to generate BMMfs. In any event, all Lynup/up BMDCs showed the opposite trend. Impaired LPS- data point to Lyn playing a distinct role in LPS-stimulated DCs induced maturation and IL-12 production in Lyn2/2 DCs was pre- and Mfs. The Journal of Immunology 5103

Table II. Percentage of MHC-IIhigh, MHC-IIint, and MHC-IIlow DCs following knockdown of SHIP-1 or SHP-1 in Lyn+/+ and Lyn2/2 GM-BMDCs

Control shRNA SHIP-1 shRNA SHP-1 shRNA

Region 2LPS +LPS 2LPS +LPS 2LPS +LPS Lyn+/+ 1 15.3 41.6 15.2 24.4 16.3 39.6 2 40.9 32.2 31.6 35.4 41.7 32.2 3 36.5 17.3 47.9 34.5 34.2 19.9 Lyn2/2 1 10.9 22.5 7.3 12.6 7.0 20.5 2 38.4 38.7 30.0 41.0 35.5 38.4 3 44.5 32.2 56.9 41.1 47.4 31.9 All values represent percentages. Regions are as follows: 1, MHC-IIhigh; 2, MHC- IIint; 3, MHC-IIlow.

The mechanism by which murine DCs activate NK cells requires TLR or CD40-dependent production of type I IFNs by DCs, which act back on DCs to induce trans-presentation of IL-15 to NK cells.

DC secretion of IL-12 and IL-18 has also been implicated in NK Downloaded from cell activation (6, 10, 49). We favor a model whereby Lyn regulates the nature of DC TLR responses by enhancing the production of IL-12, IL-15, and IFN-b, which together stimulate DC priming and activation of NK cells to produce IFN-g. In turn, NK cells stimulate DCs and other LPS-responsive cells, leading to an

amplified response to PAMPs. http://www.jimmunol.org/ Lyn is unique, performing both positive and negative regulatory roles in signal transduction (11, 14, 15). We found that in DCs, Lyn acted primarily as a negative regulator of LPS-induced activation of the MAPKs, NF-kB, and PI3K, acting most profoundly on the PI3K pathway. How can the ability of Lyn to enhance DC maturation and proinflammatory cytokine synthesis be reconciled with its ability to attenuate LPS-induced signal transduction? One possible explanation is that the PI3K pathway inhibits TLR- induced responses in myeloid cells (13, 70, 71). DCs lacking by guest on September 25, 2021 the p85a subunit of PI3K, or DCs treated with PI3K inhibitors, produce more IL-12 in response to LPS or other TLR ago- nists (70–72). Similarly, PI3K inhibition results in decreased LPS-induced IL-10 and increased IL-12 production in Mos and PBMCs (73). Correspondingly, DCs lacking the PI3K pathway antagonist SHIP-1 exhibit reduced LPS- or CpG-induced maturation and IL- 12 secretion (58). We confirmed that DCs lacking SHIP-1 exhibit enhanced PI3K signal transduction and impaired LPS-induced maturation. Lynup/up DCs were more profoundly affected by the 2/2 A +/+ lack of SHIP-1 than were WT or Lyn DCs. Interestingly, FIGURE 8. Lyn regulates SHIP-1 and SHP-1 in GM-BMDCs. ( )Lyn , v v Lyn2/2, and Lynup/up GM-BMDCs were stimulated with 100 ng/ml LPS Me /Me DCs also showed impaired spontaneous and LPS- for 0 or 15 min and lysates were subjected to Western blotting using anti- induced maturation, and although SHP-1 knockdown did not +/+ 2/2 phosphotyrosine (P-Tyr) Abs. Arrows indicate phosphoproteins that were significantly affect maturation of Lyn or Lyn DCs, it im- hyperphosphorylated in Lynup/up DCs and hypophosphorylated in Lyn2/2 paired spontaneous and LPS-induced maturation of Lynup/up DCs. DCs. (B)Lyn+/+,Lyn2/2, and Lynup/up GM-BMDCs were stimulated as in (A). SHP-1 and SHIP-1 were immunoprecipitated, followed by anti–P-Tyr Western blotting. Blots were reprobed with anti–SHIP-1 and anti–SHP-1 Abs to confirm equal precipitation. (C) WT GM-BMDCs were infected to Western blotting with Abs recognizing p-Ikka/b, p-Akt, p-Erk1/2, and with retrovirus producing shRNAs targeting SHP-1, SHIP-1, or control p-S6. The blot was reprobed with anti-total Akt (T-Akt) Abs (loading shRNA (Ctr). Two shRNAs per target are shown. Knockdown was con- control). (F)Lynup/up GM-BMDCs were infected with control or SHIP-1 firmed by Western blotting with anti–SHIP-1 and anti–SHP-1 Abs. Blots shRNA retrovirus. Infected DCs were then sorted based on eGFP ex- were reprobed with anti–SHP-2 and Erk1/2 Abs to confirm equal loading. pression and cocultured with NK cells (1.0 3 105 DCs with 4.0 3 104 NK (D)Lynup/up GM-BMDCs were infected with retrovirus as in (C). DCs cells). Cell mixtures were stimulated with 0, 1, 10, or 100 ng/ml LPS for were stimulated with or without LPS for 16 h and expression of MHC-II on 16 h. Culture medium was collected and IFN-g levels were determined by CD11c+ DCs was assessed by flow cytometry. MHC-II expression fell into ELISA. Error bars represent SEM. (G) GM-BMDCs were derived from three regions: low (box 3), intermediate (box 2), and high (box 1). Per- Lyn+/+,Lyn2/2,Mev/Mev (lacking SHP-1 activity), or SHIP-12/2 mice. 2 2 centages of cells in each region are shown. Lyn+/+ and Lyn / DCs were Cells were stimulated with or without 100 ng/ml LPS for 16 h and were 2 2 analyzed in the same way (see Table II for Lyn+/+ and Lyn / data). (E) analyzed by flow cytometry to assess expression of MHC-II on CD11c+ Lynup/up GM-BMDCs were infected with retrovirus as in (C). DCs were DCs. MHC-II expression fell into three regions, as in (D). The percentage stimulated with 100 ng/ml LPS for 0 or 15 min, and lysates were subjected of cells in each region is shown. *p # 0.05, ***p # 0.001. 5104 Lyn REGULATES THE INNATE IMMUNE RESPONSE

In keeping with a role for SHP-1 in DC maturation, DCs lacking by dendritic cells (DCs) requires the formation of a synapse leading to IL-12 polarization in DCs. Blood 104: 3267–3275. PIR-B, a SHP-1 binding ITIM-containing Lyn substrate (32, 74), 9. Krug, A., A. R. French, W. Barchet, J. A. Fischer, A. Dzionek, J. T. Pingel, exhibit impaired maturation and IL-12 production (75). Taken M. M. Orihuela, S. Akira, W. M. Yokoyama, and M. Colonna. 2004. TLR9- together, our data suggest that Lyn acts by tyrosine-phosphorylat- dependent recognition of MCMV by IPC and DC generates coordinated cytokine responses that activate antiviral NK cell function. Immunity 21: 107–119. ing ITIM-containing membrane proteins such as PIR-B, FcgRIIB, 10. Lucas, M., W. Schachterle, K. Oberle, P. Aichele, and A. Diefenbach. 2007. and/or SIRPa, creating binding sites for SH2 domains of SHP-1 Dendritic cells prime natural killer cells by trans-presenting interleukin 15. and SHIP-1, which then downregulate signal transduction path- Immunity 26: 503–517. 11. Lowell, C. A. 2004. Src-family kinases: rheostats of immune cell signaling. Mol. ways by dephosphorylating signaling proteins and/or phospholi- Immunol. 41: 631–643. pids (14, 21, 32, 55, 56, 76, 77). In this way, SHIP-1, and to 12. Odom, S., G. Gomez, M. Kovarova, Y. Furumoto, J. J. Ryan, H. V. Wright, a lesser extent SHP-1, can be mobilized and/or activated by Lyn to C. Gonzalez-Espinosa, M. L. Hibbs, K. W. Harder, and J. Rivera. 2004. Negative regulation of immunoglobulin E-dependent allergic responses by Lyn kinase. J. regulate DC maturation and cytokine production. Exp. Med. 199: 1491–1502. The ability of Lyn to attenuate NF-kB signaling in DCs might 13. Xu, Y., K. W. Harder, N. D. Huntington, M. L. Hibbs, and D. M. Tarlinton. 2005. up/up Lyn tyrosine kinase: accentuating the positive and the negative. Immunity 22: also contribute to enhanced LPS sensitivity of Lyn mice, as 9–18. mice lacking myeloid Ikkb (78), IKKa (79), p50, and p50 plus p65 14. Scapini, P., S. Pereira, H. Zhang, and C. A. Lowell. 2009. Multiple roles of Lyn (80) all resemble Lynup/up mice, exhibiting increased sensitivity to kinase in myeloid cell signaling and function. Immunol. Rev. 228: 23–40. 15. Hibbs, M. L., and K. W. Harder. 2006. The duplicitous nature of the Lyn tyrosine endotoxin-induced shock associated with elevated production of kinase in growth factor signaling. Growth Factors 24: 137–149. up/up proinﬂammatory cytokines, including IL-1b. Interestingly, Lyn 16. Hibbs, M. L., D. M. Tarlinton, J. Armes, D. Grail, G. Hodgson, R. Maglitto, DCs exhibited weak TLR4-induced NF-kB activation and reduced S. A. Stacker, and A. R. Dunn. 1995. Multiple defects in the immune system of Lyn-deﬁcient mice, culminating in autoimmune disease. Cell 83: 301–311.

536 Downloaded from phosphorylation of p65/RelA on Ser , a pattern of NF-kB acti- 17. Nishizumi, H., I. Taniuchi, Y. Yamanashi, D. Kitamura, D. Ilic, S. Mori, vation associated with increased TLR-induced IL-12 production in T. Watanabe, and T. Yamamoto. 1995. Impaired proliferation of peripheral DCs (81). B cells and indication of autoimmune disease in lyn-deficient mice. Immunity 3: 549–560. Although the relationship between DC maturation and function 18. Silver, K. L., T. L. Crockford, T. Bouriez-Jones, S. Milling, T. Lambe, and is a subject of debate (82), we favor a model in which pathogen R. J. Cornall. 2007. MyD88-dependent autoimmune disease in Lyn-deficient mice. Eur. J. Immunol. 37: 2734–2743. recognition by immature versus mature DCs has distinct out- 19. Tsantikos, E., S. A. Oracki, C. Quilici, G. P. Anderson, D. M. Tarlinton, and comes. Depending on the type of DC, stimulus, and extracellular M. L. Hibbs. 2010. Autoimmune disease in Lyn-deficient mice is dependent on http://www.jimmunol.org/ milieu, DCs may become immunogenic, secreting high levels of an inflammatory environment established by IL-6. J. Immunol. 184: 1348–1360. 20. Scapini, P., Y. Hu, C. L. Chu, T. S. Migone, A. L. Defranco, M. A. Cassatella, IL-12, IL-15, and type I IFNs with a greater propensity for NK and C. A. Lowell. 2010. Myeloid cells, BAFF, and IFN-g establish an inflam- activation and Th1 T cell development, or alternatively may se- matory loop that exacerbates autoimmunity in Lyn-deficient mice. J. Exp. Med. crete IL-10 and reduced IL-12, IL-15, and type I IFNs leading to 207: 1757–1773. 21. Harder, K. W., L. M. Parsons, J. Armes, N. Evans, N. Kountouri, R. Clark, Th2, Th17, or regulatory T cell development (83–85). Distinct 2 2 C. Quilici, D. Grail, G. S. Hodgson, A. R. Dunn, and M. L. Hibbs. 2001. Gain- responses of Lyn / and Lynup/up DCs to LPS might therefore and loss-of-function Lyn mutant mice define a critical inhibitory role for Lyn in reflect global differences in gene expression correlating with mat- the myeloid lineage. Immunity 15: 603–615. 22. Harder, K. W., C. Quilici, E. Naik, M. Inglese, N. Kountouri, A. Turner, uration. A better understanding of Lyn-dependent signaling path- K. Zlatic, D. M. Tarlinton, and M. L. Hibbs. 2004. Perturbed myelo/eryth- ways in DCs may provide avenues to better modulate DC matu- ropoiesis in Lyn-deficient mice is similar to that in mice lacking the inhibitory by guest on September 25, 2021 ration and activation to enhance vaccines or control the nature phosphatases SHP-1 and SHIP-1. Blood 104: 3901–3910. 23. Keck, S., M. Freudenberg, and M. Huber. 2010. Activation of murine macro- of immune responses to microbes. phages via TLR2 and TLR4 is negatively regulated by a Lyn/PI3K module and promoted by SHIP1. J. Immunol. 184: 5809–5818. 24. Napolitani, G., N. Bortoletto, L. Racioppi, A. Lanzavecchia, and U. D’Oro. Acknowledgments 2003. Activation of src-family tyrosine kinases by LPS regulates cytokine production in dendritic cells by controlling AP-1 formation. Eur. J. Immunol. 33: We thank Prof. Ashley Dunn (Ludwig Institute for Cancer Research, Park- 2832–2841. ville, VIC, Australia) for advice and support, and Natalie Evans, Dianne 25. Stefanova´, I., M. L. Corcoran, E. M. Horak, L. M. Wahl, J. B. Bolen, and Grail, Sami Lay, Kristina Zlatic (Ludwig Institute for Cancer Research), I. D. Horak. 1993. Lipopolysaccharide induces activation of CD14-associated Joshua Jiang, and Dianne Fang (University of British Columbia, Vancou- protein tyrosine kinase p53/56lyn. J. Biol. Chem. 268: 20725–20728. ver, BC, Canada) for excellent technical assistance. 26. Herrera-Velit, P., and N. E. Reiner. 1996. Bacterial lipopolysaccharide induces the association and coordinate activation of p53/56lyn and phosphatidylinositol 3-kinase in human monocytes. J. Immunol. 156: 1157–1165. 27. Galgani, M., V. De Rosa, S. De Simone, A. Leonardi, U. D’Oro, G. Napolitani, Disclosures A. M. Masci, S. Zappacosta, and L. Racioppi. 2004. Cyclic AMP modulates the The authors have no financial conflicts of interest. functional plasticity of immature dendritic cells by inhibiting Src-like kinases through protein kinase A-mediated signaling. J. Biol. Chem. 279: 32507– 32514. 28. Orlicek, S. L., J. H. Hanke, and B. K. English. 1999. The src family-selective References tyrosine kinase inhibitor PP1 blocks LPS and IFN-g-mediated TNF and iNOS 1. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate production in murine macrophages. Shock 12: 350–354. immunity. Cell 124: 783–801. 29. Fan, H., G. Teti, S. Ashton, K. Guyton, G. E. Tempel, P. V. Halushka, and 2. Akira, S. 2006. TLR signaling. Curr. Top. Microbiol. Immunol. 311: 1–16. J. A. Cook. 2003. Involvement of Gi proteins and Src tyrosine kinase in TNFa 3. Sriskandan, S., and D. M. Altmann. 2008. The immunology of sepsis. J. Pathol. production induced by lipopolysaccharide, group B Streptococci and Staphylo- 214: 211–223. coccus aureus. Cytokine 22: 126–133. 4. Moretta, A., E. Marcenaro, S. Parolini, G. Ferlazzo, and L. Moretta. 2008. NK 30. Ferlito, M., O. G. Romanenko, K. Guyton, S. Ashton, F. Squadrito, cells at the interface between innate and adaptive immunity. Cell Death Differ. P. V. Halushka, and J. A. Cook. 2002. Implication of Gai proteins and Src ty- 15: 226–233. rosine kinases in endotoxin-induced signal transduction events and mediator 5. Ferlazzo, G., and C. Mu¨nz. 2009. Dendritic cell interactions with NK cells from production. J. Endotoxin Res. 8: 427–435. different tissues. J. Clin. Immunol. 29: 265–273. 31. Chu, C. L., and C. A. Lowell. 2005. The Lyn tyrosine kinase differentially 6. Andoniou, C. E., S. L. van Dommelen, V. Voigt, D. M. Andrews, G. Brizard, regulates dendritic cell generation and maturation. J. Immunol. 175: 2880–2889. C. Asselin-Paturel, T. Delale, K. J. Stacey, G. Trinchieri, and M. A. Degli- 32. Beavitt, S. J., K. W. Harder, J. M. Kemp, J. Jones, C. Quilici, F. Casagranda, Esposti. 2005. Interaction between conventional dendritic cells and natural killer E. Lam, D. Turner, S. Brennan, P. D. Sly, et al. 2005. Lyn-deficient mice develop cells is integral to the activation of effective antiviral immunity. Nat. Immunol. 6: severe, persistent asthma: Lyn is a critical negative regulator of Th2 immunity. J. 1011–1019. Immunol. 175: 1867–1875. 7. Koka, R., P. Burkett, M. Chien, S. Chai, D. L. Boone, and A. Ma. 2004. Cutting 33. Mombaerts, P., J. Iacomini, R. S. Johnson, K. Herrup, S. Tonegawa, and edge: murine dendritic cells require IL-15Ra to prime NK cells. J. Immunol. V. E. Papaioannou. 1992. RAG-1-deficient mice have no mature B and 173: 3594–3598. T lymphocytes. Cell 68: 869–877. 8. Borg, C., A. Jalil, D. Laderach, K. Maruyama, H. Wakasugi, S. Charrier, 34. Jung, S., D. Unutmaz, P. Wong, G. Sano, K. De los Santos, T. Sparwasser, S. Wu, B. Ryffel, A. Cambi, C. Figdor, W. Vainchenker, et al. 2004. NK cell activation S. Vuthoori, K. Ko, F. Zavala, et al. 2002. In vivo depletion of CD11c+ dendritic The Journal of Immunology 5105

cells abrogates priming of CD8+ T cells by exogenous cell-associated antigens. 60. Lowell, C. A., and G. Berton. 1998. Resistance to endotoxic shock and reduced Immunity 17: 211–220. neutrophil migration in mice deficient for the Src-family kinases Hck and Fgr. 35. Liu, Q., T. Sasaki, I. Kozieradzki, A. Wakeham, A. Itie, D. J. Dumont, and Proc. Natl. Acad. Sci. USA 95: 7580–7584. J. M. Penninger. 1999. SHIP is a negative regulator of growth factor receptor- 61. Starr, R., D. Metcalf, A. G. Elefanty, M. Brysha, T. A. Willson, N. A. Nicola, mediated PKB/Akt activation and myeloid cell survival. Genes Dev. 13: 786–791. D. J. Hilton, and W. S. Alexander. 1998. Liver degeneration and lymphoid 36. Tsui, H. W., K. A. Siminovitch, L. de Souza, and F. W. Tsui. 1993. Motheaten deficiencies in mice lacking suppressor of cytokine signaling-1. Proc. Natl. and viable motheaten mice have mutations in the haematopoietic cell phos- Acad. Sci. USA 95: 14395–14399. phatase gene. Nat. Genet. 4: 124–129. 62. Nakagawa, R., T. Naka, H. Tsutsui, M. Fujimoto, A. Kimura, T. Abe, E. Seki, 37. Lutz, M. B., N. Kukutsch, A. L. Ogilvie, S. Ro¨ssner, F. Koch, N. Romani, and S. Sato, O. Takeuchi, K. Takeda, et al. 2002. SOCS-1 participates in negative G. Schuler. 1999. An advanced culture method for generating large quantities of regulation of LPS responses. Immunity 17: 677–687. highly pure dendritic cells from mouse bone marrow. J. Immunol. Methods 223: 63. Xiao, H., M. F. Gulen, J. Qin, J. Yao, K. Bulek, D. Kish, C. Z. Altuntas, D. Wald, 77–92. C. Ma, H. Zhou, et al. 2007. The Toll-interleukin-1 receptor member SIGIRR 38. Coyle, A. J., G. Le Gros, C. Bertrand, S. Tsuyuki, C. H. Heusser, M. Kopf, and regulates colonic epithelial homeostasis, inflammation, and tumorigenesis. Im- G. P. Anderson. 1995. Interleukin-4 is required for the induction of lung Th2 munity 26: 461–475. mucosal immunity. Am. J. Respir. Cell Mol. Biol. 13: 54–59. 64. Ozmen, L., M. Pericin, J. Hakimi, R. A. Chizzonite, M. Wysocka, G. Trinchieri, 39. Khvorova, A., A. Reynolds, and S. D. Jayasena. 2003. Functional siRNAs and M. Gately, and G. Garotta. 1994. Interleukin 12, interferon gamma, and tumor miRNAs exhibit strand bias. Cell 115: 209–216. necrosis factor a are the key cytokines of the generalized Shwartzman reaction. 40. Rittirsch, D., M. A. Flierl, and P. A. Ward. 2008. Harmful molecular mechanisms J. Exp. Med. 180: 907–915. in sepsis. Nat. Rev. Immunol. 8: 776–787. 65. Schroder, K., P. J. Hertzog, T. Ravasi, and D. A. Hume. 2004. Interferon-g:an 41. Steiner, A. A., S. Chakravarty, J. R. Robbins, A. S. Dragic, J. Pan, M. Herkenham, overview of signals, mechanisms and functions. J. Leukoc. Biol. 75: 163–189. and A. A. Romanovsky. 2005. Thermoregulatory responses of rats to conventional 66. Heremans, H., J. Van Damme, C. Dillen, R. Dijkmans, and A. Billiau. 1990. preparations of lipopolysaccharide are caused by lipopolysaccharide per se—not Interferon g, a mediator of lethal lipopolysaccharide-induced Shwartzman-like by lipoprotein contaminants. Am. J. Physiol. Regul. Integr. Comp. Physiol. 289: shock reactions in mice. J. Exp. Med. 171: 1853–1869. R348–R352. 67. Doherty, G. M., J. R. Lange, H. N. Langstein, H. R. Alexander, C. M. Buresh, 42. Xing, Z., J. Gauldie, G. Cox, H. Baumann, M. Jordana, X. F. Lei, and and J. A. Norton. 1992. Evidence for IFN-g as a mediator of the lethality of M. K. Achong. 1998. IL-6 is an antiinflammatory cytokine required for con- endotoxin and tumor necrosis factor-a. J. Immunol. 149: 1666–1670. Downloaded from trolling local or systemic acute inflammatory responses. J. Clin. Invest. 101: 68. Emoto, M., M. Miyamoto, I. Yoshizawa, Y. Emoto, U. E. Schaible, E. Kita, and 311–320. S. H. Kaufmann. 2002. Critical role of NK cells rather than Va14+NKT cells 43. Heinzel, F. P. 1990. The role of IFN-g in the pathology of experimental endo- in lipopolysaccharide-induced lethal shock in mice. J. Immunol. 169: 1426– toxemia. J. Immunol. 145: 2920–2924. 1432. 44. Jurkovich, G. J., W. J. Mileski, R. V. Maier, R. K. Winn, and C. L. Rice. 1991. 69. Meng, F., and C. A. Lowell. 1997. Lipopolysaccharide (LPS)-induced macro- Interferon g increases sensitivity to endotoxin. J. Surg. Res. 51: 197–203. phage activation and signal transduction in the absence of Src-family kinases 45. Schoenborn, J. R., and C. B. Wilson. 2007. Regulation of interferon-g during Hck, Fgr, and Lyn. J. Exp. Med. 185: 1661–1670.

innate and adaptive immune responses. Adv. Immunol. 96: 41–101. 70. Fukao, T., and S. Koyasu. 2003. PI3K and negative regulation of TLR signaling. http://www.jimmunol.org/ 46. Heremans, H., C. Dillen, J. van Damme, and A. Billiau. 1994. Essential role for Trends Immunol. 24: 358–363. natural killer cells in the lethal lipopolysaccharide-induced Shwartzman-like 71. Hazeki, K., K. Nigorikawa, and O. Hazeki. 2007. Role of phosphoinositide reaction in mice. Eur. J. Immunol. 24: 1155–1160. 3-kinase in innate immunity. Biol. Pharm. Bull. 30: 1617–1623. 47. Hibbs, M. L., K. W. Harder, J. Armes, N. Kountouri, C. Quilici, F. Casagranda, 72. Fukao, T., M. Tanabe, Y. Terauchi, T. Ota, S. Matsuda, T. Asano, T. Kadowaki, A. R. Dunn, and D. M. Tarlinton. 2002. Sustained activation of Lyn tyrosine T. Takeuchi, and S. Koyasu. 2002. PI3K-mediated negative feedback regulation kinase in vivo leads to autoimmunity. J. Exp. Med. 196: 1593–1604. of IL-12 production in DCs. Nat. Immunol. 3: 875–881. 48. Fernandez, N. C., A. Lozier, C. Flament, P. Ricciardi-Castagnoli, D. Bellet, 73. Martin, M., K. Rehani, R. S. Jope, and S. M. Michalek. 2005. Toll-like receptor- M. Suter, M. Perricaudet, T. Tursz, E. Maraskovsky, and L. Zitvogel. 1999. mediated cytokine production is differentially regulated by glycogen synthase Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate kinase 3. Nat. Immunol. 6: 777–784. anti-tumor immune responses in vivo. Nat. Med. 5: 405–411. 74. Ho, L. H., T. Uehara, C. C. Chen, H. Kubagawa, and M. D. Cooper. 1999. 49. Ferlazzo, G., M. Pack, D. Thomas, C. Paludan, D. Schmid, T. Strowig, Constitutive tyrosine phosphorylation of the inhibitory paired Ig-like receptor

G. Bougras, W. A. Muller, L. Moretta, and C. Mu¨nz. 2004. Distinct roles of PIR-B. Proc. Natl. Acad. Sci. USA 96: 15086–15090. by guest on September 25, 2021 IL-12 and IL-15 in human natural killer cell activation by dendritic cells from 75. Ujike, A., K. Takeda, A. Nakamura, S. Ebihara, K. Akiyama, and T. Takai. 2002. 2/2 secondary lymphoid organs. Proc. Natl. Acad. Sci. USA 101: 16606–16611. Impaired dendritic cell maturation and increased TH2 responses in PIR-B 50. Piccioli, D., S. Sbrana, E. Melandri, and N. M. Valiante. 2002. Contact- mice. Nat. Immunol. 3: 542–548. dependent stimulation and inhibition of dendritic cells by natural killer cells. 76. Zhang, J., A. K. Somani, and K. A. Siminovitch. 2000. Roles of the SHP-1 J. Exp. Med. 195: 335–341. tyrosine phosphatase in the negative regulation of cell signalling. Semin. Im- 51. Gerosa, F., B. Baldani-Guerra, C. Nisii, V. Marchesini, G. Carra, and munol. 12: 361–378. G. Trinchieri. 2002. Reciprocal activating interaction between natural killer cells 77. Erbacher, A., F. Gieseke, R. Handgretinger, and I. Mu¨ller. 2009. Dendritic cells: and dendritic cells. J. Exp. Med. 195: 327–333. functional aspects of glycosylation and lectins. Hum. Immunol. 70: 308–312. 52. Manning, B. D., and L. C. Cantley. 2007. AKT/PKB signaling: navigating 78. Greten, F. R., M. C. Arkan, J. Bollrath, L. C. Hsu, J. Goode, C. Miething, downstream. Cell 129: 1261–1274. S. I. Go¨ktuna, M. Neuenhahn, J. Fierer, S. Paxian, et al. 2007. NF-kB is a neg- 53. Hayden, M. S., A. P. West, and S. Ghosh. 2006. NF-kB and the immune re- ative regulator of IL-1b secretion as revealed by genetic and pharmacological sponse. Oncogene 25: 6758–6780. inhibition of IKKb. Cell 130: 918–931. 54. Hayden, M. S., and S. Ghosh. 2004. Signaling to NF-kB. Genes Dev. 18: 2195– 79. Lawrence, T., M. Bebien, G. Y. Liu, V. Nizet, and M. Karin. 2005. IKKa limits 2224. macrophage NF-kB activation and contributes to the resolution of inﬂammation. 55. Smith, K. G., D. M. Tarlinton, G. M. Doody, M. L. Hibbs, and D. T. Fearon. 1998. Nature 434: 1138–1143. Inhibition of the B cell by CD22: a requirement for Lyn. J. Exp. Med. 187: 807– 80. Gadjeva, M., M. F. Tomczak, M. Zhang, Y. Y. Wang, K. Dull, A. B. Rogers, 811. S. E. Erdman, J. G. Fox, M. Carroll, and B. H. Horwitz. 2004. A role for NF-kB 56. Cornall, R. J., J. G. Cyster, M. L. Hibbs, A. R. Dunn, K. L. Otipoby, E. A. Clark, subunits p50 and p65 in the inhibition of lipopolysaccharide-induced shock. J. and C. C. Goodnow. 1998. Polygenic autoimmune traits: Lyn, CD22, and SHP-1 Immunol. 173: 5786–5793. are limiting elements of a biochemical pathway regulating BCR signaling and 81. Bode, K. A., F. Schmitz, L. Vargas, K. Heeg, and A. H. Dalpke. 2009. Kinetic of selection. Immunity 8: 497–508. RelA activation controls magnitude of TLR-mediated IL-12p40 induction. J. 57. Antignano, F., M. Ibaraki, J. Ruschmann, J. Jagdeo, and G. Krystal. 2010. SHIP Immunol. 182: 2176–2184. negatively regulates Flt3L-derived dendritic cell generation and positively reg- 82. Reis e Sousa, C. 2006. Dendritic cells in a mature age. Nat. Rev. Immunol. 6: ulates MyD88-independent TLR-induced maturation. J. Leukoc. Biol. 88: 925– 476–483. 935. 83. Walzer, T., M. Dalod, S. H. Robbins, L. Zitvogel, and E. Vivier. 2005. Natural- 58. Antignano, F., M. Ibaraki, C. Kim, J. Ruschmann, A. Zhang, C. D. Helgason, and killer cells and dendritic cells: “l’union fait la force”. Blood 106: 2252–2258. G. Krystal. 2010. SHIP is required for dendritic cell maturation. J. Immunol. 184: 84. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of 2805–2813. immunity. Nature 392: 245–252. 59. Ernst, M., M. Inglese, G. M. Scholz, K. W. Harder, F. J. Clay, S. Bozinovski, 85. De Smedt, T., E. Butz, J. Smith, R. Maldonado-Lo´pez, B. Pajak, M. Moser, and P. Waring, R. Darwiche, T. Kay, P. Sly, et al. 2002. Constitutive activation of the C. Maliszewski. 2001. CD8a2 and CD8a+ subclasses of dendritic cells undergo SRC family kinase Hck results in spontaneous pulmonary inﬂammation and an phenotypic and functional maturation in vitro and in vivo. J. Leukoc. Biol. 69: enhanced innate immune response. J. Exp. Med. 196: 589–604. 951–958.