Francisella Novicida Mice Infected with Deficient

ASC Controls IFN-γ Levels in an IL-18− Dependent Manner in Caspase-1−Deficient Mice Infected with Francisella novicida

This information is current as Roberto Pierini, Magali Perret, Sophia Djebali, Carole Juruj, of September 27, 2021. Marie-Cécile Michallet, Irmgard Förster, Jacqueline Marvel, Thierry Walzer and Thomas Henry J Immunol 2013; 191:3847-3857; Prepublished online 23 August 2013;

doi: 10.4049/jimmunol.1203326 Downloaded from http://www.jimmunol.org/content/191/7/3847

Supplementary http://www.jimmunol.org/content/suppl/2013/08/23/jimmunol.120332 Material 6.DC1 http://www.jimmunol.org/ References This article cites 60 articles, 21 of which you can access for free at: http://www.jimmunol.org/content/191/7/3847.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 27, 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 © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

ASC Controls IFN-g Levels in an IL-18–Dependent Manner in Caspase-1–Deﬁcient Mice Infected with Francisella novicida

Roberto Pierini,*,†,‡,x Magali Perret,*,†,‡,x Sophia Djebali,*,†,‡,x Carole Juruj,*,†,‡,x Marie-Ce´cile Michallet,*,†,‡,x Irmgard Fo¨rster,{,1 Jacqueline Marvel,*,†,‡,x Thierry Walzer,*,†,‡,x and Thomas Henry*,†,‡,x

The inflammasome is a signaling platform that is central to the innate immune responses to bacterial infections. Francisella tularensis is a bacterium replicating within the host cytosol. During F. tularensis subspecies novicida infection, AIM2, an inflammasome receptor sensing cytosolic DNA, activates caspase-1 in an ASC-dependent manner, leading to both pyroptosis and release of the proinflammatory cytokines IL-1b and IL-18. Activation of this canonical inflammasome pathway is key to limit F. novicida Downloaded from infection. In this study, by comparing the immune responses of AIM2 knockout (KO), ASCKO, and Casp1KO mice in response to F. novicida infection, we observed that IFN-g levels in the serum of Casp1KO mice were much higher than the levels observed in AIM2KO and ASCKO mice. This difference in IFN-g production was due to a large production of IFN-g by NK cells in Casp1KO mice that was not observed in ASCKO mice. The deficit in IFN-g production observed in ASCKO mice was not due to a reduced Dock2 expression or to an intrinsic defect of ASCKO NK cells. We demonstrate that in infected Casp1KO mice, IFN-g production is due to an ASC-dependent caspase-1–independent pathway generating IL-18. Furthermore, we present in vitro data suggesting http://www.jimmunol.org/ that the recently described AIM2/ASC/caspase-8 noncanonical pathway is responsible for the caspase-1–independent IL-18 releasing activity. To our knowledge, this study is the first in vivo evidence of an alternative pathway able to generate in a caspase-1–independent pathway bioactive IL-18 to boost the production of IFN-g, a cytokine critical for the host antibacterial response. The Journal of Immunology, 2013, 191: 3847–3857.

he inflammasome (1) is an innate immune platform ac- masome receptor detecting DNA within the host cytosol (2–5). We tivated in response to danger signals and infections and and others have recently described that the AIM2 inflammasome T leading to caspase-1 activation. Caspase-1 is an inflam- is activated following infections with “cytosolic” bacteria such as by guest on September 27, 2021 matory caspase leading to the processing of pro–IL-1b and pro–IL- Listeria monocytogenes and Francisella tularensis (6–9). 18 and the release of the corresponding cytokines. Furthermore, F. tularensis is a highly infectious bacterium that causes tula- caspase-1 also triggers cell death in a process termed pyroptosis. remia in humans (10). Francisella novicida (also known as F. Caspase-1 activation takes place within a multiprotein complex, the tularensis subspecies novicida) is considered nonpathogenic for inflammasome, which includes a receptor, and an adaptor, ASC. humans and is used as a model to study highly virulent F. tular- Several inflammasomes have been described depending on the ensis infections. In mice, F. novicida is found mostly in myeloid detected ligand and on the receptor involved. AIM2 is an inflam- cells such as macrophages and neutrophils (11). Its ability to cause disease is tightly linked to its ability to rapidly escape from the phagosome into the host cytosol where it can replicate to very high *Centre International de Recherche en Infectiologie, Universite´ de Lyon, Lyon numbers (12, 13). 69007, France; †INSERM, Unite´ 851, Lyon 69007, France; ‡Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche 5308, Lyon 69007, France; xEcole Infection with wild-type (WT) F. novicida leads to the activation Normale Supe´rieure de Lyon, Lyon 69342, France; and {Department of Molecular of the AIM2 inflammasome (7–9). Mice deficient for AIM2, ASC, Immunology, Leibniz Research Institute for Environmental Medicine, 40225 Du¨ssel- or for both caspase-1 and caspase-11 (14) are very susceptible to F. dorf, Germany novicida infection. This high susceptibility of ASC knockout (KO) 1Current address: Immunology and Environment, Life and Medical Sciences Insti- tute, University of Bonn, Bonn, Germany. and Casp1/11 double KO mice is due to both caspase-1–mediated Received for publication December 4, 2012. Accepted for publication July 24, 2013. control of IL-1b and IL-18 and to caspase-1–mediated cell death (15–17). In contrast to other intracellular bacteria, detection of This work was supported by Marie Curie International Reintegration Grant PIRG07- GA-2010-268399 and by a Fondation Innovations en Infectiologie Young Inves- F. novicida by the inflammasome in murine cells seems to be ex- tigator grant. This work was performed within the framework of the Laboratoire clusively dependent on AIM2 (8). F. novicida infection is thus d’Excellence eco-evolutionary dynamics of infectious diseases (ECOFECT Grant ANR- 11-LABX-0042) of the Universite´ de Lyon, within the program “Investissements d’Avenir” emerging as a very good model to study both in vitro and in vivo this (Grant ANR-11-IDEX-0007) operated by the French National Research Agency. innate immune response (8, 9, 18). Caspase-11 is dispensable for Address correspondence and reprint requests to Dr. Thomas Henry, INSERM, Unite´ caspase-1 activation in response to F. novicida infection (14, 19). 851, 21 Avenue Tony Garnier, Lyon 69007, France. E-mail address: thomas.henry@ The canonical inflammasome pathway leads to caspase-1 acti- inserm.fr vation. However, we and others have shown that the inflammasome The online version of this article contains supplemental material. adaptor ASC could have caspase-1–independent activities. In vitro, Abbreviations used in this article: BMM, bone marrow macrophage; KO, knockout; ASC can recruit and activate apoptotic caspases leading to mac- PI, postinfection; WT, wild-type. rophage apoptosis (18, 20–23). In several noninfectious inflam- Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 matory models, ASC contributes to inflammation in a caspase-1– www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203326 3848 ASC/IL-18–DEPENDENT REGULATION OF IFN-g IN Casp1KO MICE dependent and caspase-1–independent manner (24–26). Similarly, in FACS buffer and Live/Dead fixable dead cell stain kit (Invitrogen) in during chronic Mycobacterium tuberculosis infection in mice, PBS containing 2.5 mM EDTA. Peripheral blood samples were collected ASC participates in granuloma formation mechanisms indepen- in EDTA-containing tubes and incubated for 30 min at 4˚C with mAbs cocktails. dently of caspase-1 (27). Recently, Dock2 expression was shown Phenotyping and analysis of NK cell maturation and function was to be deficient in several but not all ASC-deficient mouse lines, performed as previously described (32, 33). When applicable, spleen leading to a severe immune deficit in ASCKO mouse lines ex- single-cell suspensions were incubated for 4 h with IL-12 (20 ng/ml), IL- pressing low levels of Dock2 protein (28, 29). It is still unclear 12 plus IL-2 (3000 U/ml; PeproTech), IL-12 plus IL-18 (20 mg/ml; MBL International), or on coated Abs anti-Ly49 (4E5) or anti-NKp46 (29A1.4) whether some of the phenotypes described above relate to a deficit in the presence of monensin and anti-CD107a (Lamp-1) Ab (BD Bio- in Dock2. sciences). Cells were analyzed using a FACSCanto II flow cytometer In this study, using F. novicida infection as a model to engage (Becton Dickinson, San Jose, CA). Data were processed using FlowJo the AIM2/ASC inflammasome components, we compared the im- software (Tree Star, Asland, OR). mune responses of WT mice with the ones of mice deficient for RNA and quantitative real-time RT-PCR ASC (ASCKO) or for both caspase-1 and caspase-11 (termed Casp1KO). We found that ASC controls IFN-g secretion by NK Approximately a third of the spleen was ground in 1 ml TRIzol. RNA was extracted using chloroform followed by 70% ethanol-mediated precipitation. cells in a caspase-1–dependent and caspase-1–independent man- RNA was then isolated using the RNeasy Mini kit (Qiagen). Reverse tran- ner. This was independent of Dock2 because our ASCKO mouse scription and quantitative PCR were performed using Improm (Promega) line had normal Dock2 protein levels. In infected Casp1KO mice, and LightCycler 480 SYBR Green I master (Roche) kits, respectively, ASC regulated IL-18 at the posttranslational stage and neutrali- on a LightCycler 480 II (Roche). Gene-specific transcript levels were normalized to the amount of b-actin mRNA and expressed as fold increase

KO Downloaded from zation of IL-18 greatly reduced IFN-g in infected Casp1 mice. over the normalized level of the corresponding transcript in WT uninfected In vitro experiments suggest that ASC-dependent caspase-1–in- sample. IFN-b, IFN-g, and b-actin primers have been published before dependent release of IL-18 is dependent on caspase-8. To our (34). Primers for pro–IL-1b, IL-10, and pro–IL-18 were: IL-1b, forward, knowledge, this is the ﬁrst report of an in vivo role for ASC in 59-GCCCATCCTCTGTGACTCAT-39, reverse, 59-AGGCCACAGGTATT- generating bioactive IL-18 independently of caspase-1. TTGTCG-39; IL-10, forward, 59-ATTTGAATTCCCTGGGTGAGAAG-39, reverse, 59-CACAGGGGAGAAATCGATGACA-39; IL-18, forward, 59- ACGTGTTCCAGGACACAACA-39, reverse, 59-ACAAACCCTCCCCAC- Materials and Methods CTAAC-39. http://www.jimmunol.org/ Mice and in vivo infections ELISA and multiplex immunoassay This study was carried out in strict accordance with the French recom- Spleen lysates were cleared by centrifugation (15 min, 20,000 3 g at 4˚C). mendations in the Guide for the Ethical Evaluation of Experiments Using ELISA was performed on lysates diluted 10 times and serums diluted up to Laboratory Animals (http://gircor.net/qui/ethicalEvaluationGuide4Labor- 10 times. All ELISAs with the exception of the IL-18 ELISA (MBL In- atoryAnimals.pdf) and the European guidelines 86/609/CEE. All experi- ternational) were from R&D Systems. When indicated, IFN-g, IL-2, IL-10, mental studies were approved by the bioethic committee (Comite´ ethique IL-12p70, IL-15, and IL-18 in serum were determined by multiplex im- pour l’expe´rimentation animal protocols ENS_2009_020, ENS_2011_006, munoassay (Bio-Plex Pro mouse cytokines; Bio-Rad) at the Plateau de and ENS_2012_033). All mouse strains were of the C57BL/6 background. Biologie Expe´rimental de la Souris genotyping and phenotyping platform. KO

WT mice were purchased from The Charles River Laboratory. ASC (30) by guest on September 27, 2021 and AIM2KO mice were obtained from the Vishva Dixit Laboratory Cell culture (Genentech, South San Francisco, CA). Casp1KO mice were obtained from the Denise Monack Laboratory (Stanford University). ASCKO and Preparation, culture, and infection of bone marrow macrophages (BMMs) Casp1KO mice have been backcrossed at least 10 times to C57BL/6 mice were performed as previously described (35). 293T cells were cultured in (8). AIM2KO mice were generated in C57BL/6 embryonic stem cells (8). DMEM medium supplemented with 10% FCS, 1 mM glutamine, 1 mM AIM2KO, ASCKO, and Casp1KO mice were bred at the Plateau de Biologie pyruvate, and 100 U/ml penicillin and streptomycin. Splenocytes were m Expe´rimental de la Souris animal facility using exhaust-ventilated racks. cultured in RPMI1640 medium supplemented with 10% FCS and 10 g/ml 3 F. novicida For infection, mice were injected intradermally with 5 3 10 CFU F. gentamicin. For the coculture experiments, –infected macrophages were washed at 1 h PI and incubated for an extra hour in 10 mg/ml novicida strain U112 in 50 ml PBS. When applicable, 500 mg IL-18 KO 3 neutralizing Ab (31) or its isotype control (mouse IgG1-MOPC-21; Bio gentamicin to eliminate all extracellular bacteria. ASC splenocytes (5 5 F. novicida X Cell, West Lebanon, NH) in 500 ml PBS were injected i.p. at 24 h 10 )wereaddedto –infected BMMs (ratio 5:1) at 2 h PI. postinfection (PI). Mice were euthanized at 48 h PI. Spleen was harvested Supernatants were collected at 12 h PI. When applicable, rIL-18 (200 pg/ and when required split in three equal pieces for RNA extraction, FACS ml), anti–IL-18 neutralizing Ab, or an IgG1 isotype control (400 ng/ml) analysis, or homogenization in PBS for CFU and cytokine level determi- was added. nation. To determine the number of CFU, serial dilutions of the spleen Bacterial strains and growth conditions homogenate in PBS were plated on supplemented Mueller–Hinton plates. Blood was collected by intracardiac puncture. F. tularensis subspecies novicida strain Utah (U112) was grown in tryptic soy broth supplemented with 0.1% (w/v) cysteine. Bone marrow transplants Immunoblotting analysis Recipient mice (B6.SJL-Ptprca Pepcb/BoyJ) that harbor the CD45.1/Ptprca marker were irradiated twice with 450 rads and reconstituted with 2 3 106 Protein extracts were obtained by incubating cells with lysis buffer (10 mM bone marrow cells consisting of a 1:1 mixture of B6.SJL-Ptprca Pepcb/ HEPES/KOH, 2 mM EDTA, 0.1% CHAPS, 250 mM sucrose, 5 mM DTT, BoyJ cells and either C57BL/6, Casp1KO, or ASCKO cells each harboring complete protease inhibitor mixture [Roche]). Samples were clarified by the CD45.2/Ptprcb marker. Eight weeks after transplant, mice were infected centrifugation at 4˚C, 13,000 3 g for 15 min. Protein concentration was as described above. determined using the Bradford method (Bio-Rad). Equal amounts of protein lysates (20 mg/sample) were loaded in 4–12% Bis/Tris gel (Invitrogen) Phenotypic staining and flow cytometry and run in MOPS buffer (Invitrogen). Abs anti-ASC (sc-22514R; Santa 3 KO Cruz Biotechnology; 3 3 10 dilution), anti-actin (A3853; Sigma-Aldrich; Spleens and bone marrow from WT or ASC mice were collected 3 3 3 3 aseptically and single-cell suspensions were prepared in DMEM medium 5 10 dilution), and anti-Dock (09-454; Millipore; 5 10 dilution) (Invitrogen) containing 2 mM glutamine, 100 mg/ml gentamicin, and 6% were used for immunodetection. FCS. For intracellular staining following infection, splenocytes were in- Reconstitution of an inflammasome-like pathway in 293T cells cubated for 4 h in the presence of monensin. Splenocytes were stained with anti-CD45 (30-F11), CD3ε (2c11), CD4 (RM4.5), CD8 (53-6.7), NK1.1 DNA sequences encoding for caspase-1, -2, -3, -8, and -9 and pro–IL-18 (PK136), CD25 (PC61), B220 (RA36B2), Ly6G (1A8), CD62L (Mel14), were generated by RT-PCR using RNA extracted from BMMs and the CD44 (IM7.8), CD69 (H1), TCRb (H57), CD45.1 (A20), CD45.2 (104), following primers: caspase-1, forward, 59-CGCCACCATGGCTGACAA- CD107a (1D4B), IFN-g (XMG1.2), and rat IgG1k isotype control (RTK2071) GATCCTGAGG-39, reverse, 59-TTAATGTCCCGGGAAGAGGTAGAA- The Journal of Immunology 3849

ACG-39; caspase-2, forward, 59-ATATATATCTCGAGCGCCACCATG- in vitro roles for ASC that were independent of caspase-1. Fur- GCGGCGCCGAGCGGGAG-39, reverse, 59-GCGGCCGCTCACGTGG- thermore, using a multiplex immunoassay, we have recently ob- GTGGGTAGCCTG-39; caspase-3, forward, 59-CTCGAGCGCCACCAT- served that during F. novicida infection of Casp1KO animals, ASC GGAGAACAACAAAACCTCAGTGGATTC-39, reverse, 59-GCGGCC- GCCTAGTGATAAAAGTACAGTTCTTTCGTG-39; caspase-8, forward, controls IFN-g serum levels (18). This result was confirmed by 59-GCTAGCATGGATTTCCAGAGTTGTCTTTATG-39, reverse, 59-CTC- ELISA (Fig. 1A). Indeed, the concentration of IFN-g in the serum GAGTTAGGGAGGGAAGAAGAGCTTC-39; caspase-9, forward, 59-AT- of infected ASCKO and Casp1KO mice was 28- and 4-fold lower, ATATATCTCGAGCGCCACCATGGACGAGGCGGACCGGCAGC-39,re- respectively, than the one observed in the serum of infected WT verse, 59-GCGGCCGCTCATGAAGTTTTAAAAAACAGCTTTTTCCGGA- GGAAG-39; pro–IL-18, forward, 59-CTCGAGCGCCACCATGGCTGCCAT- mice. Similarly, deficiency in caspase-1 and ASC led to a 5-fold GTCAGAAGAC-39, reverse, 59-GCGGCCGCCTAACTTTGATGTAAGT- and 25-fold reduction, respectively, in IFN-g levels in the spleen TAGTGAGAG-39. PCR products were cloned into pEGFP-N1 vector (Clon- compared with WT mice (Fig. 1B). These results indicated that tech), from which eGFP-encoding sequence was removed. 293T cells (4 3 104) KO 2 ASC in Casp1 animals was as important as caspase-1 in WT were seeded in 0.3-cm wells. Cells were transfected with the plasmids mice to generate IFN-g in response to F. novicida. Furthermore, expressing the following proteins: pro–IL-18 (50 ng/well), ASC (20 ng/well), AIM2 (5 ng/well), and either caspase protein (40 pg/well). Supernatants were we observed a similar difference at the mRNA levels (Fig. 1C) collected 18 h after transfection and IL-18 levels were determined by ELISA. indicating that the ASC-dependent caspase-1–independent regulation of IFN-g is independent of posttranslational mechanisms. Statistical analysis The difference in IFN-g was not due to a difference in bacterial The statistical significances for in vitro data were determined by an unpaired t test burden because at the low dose injected (5 3 103 CFU intrader- analysis. In vivo data were analyzed by Mann–Whitney analysis. Two-tailed p mally), we did not observe any difference in bacterial burden at values were determined. A p value ,0.05 was considered statistically significant. 48 h PI between ASCKO and Casp1KO animals (Fig. 1D). As pre- Downloaded from viously described (16), WT mice displayed much less bacterial Results KO counts in the spleen as compared with inflammasome-deficient ASC specifically controls IFN-g levels in Casp1 animals mice. The ASC-mediated caspase-1–independent regulation of infected with F. novicida IFN-g was highly specific because we did not observe any dif- Although the major role for the inflammasome adaptor ASC is ferences in the levels of IL-2 or IL-10 cytokines in the serum (see to control activation of caspase-1, we and others have identified below), or IL-10 or IFN-b (Fig. 1E, 1F) transcripts in the spleen http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. IFN-g levels in Casp1KO are specifically higher than in ASCKO animals upon infection with F. novicida. Mice were infected intradermally with 5 3 103 F. novicida CFU for 48 h. One point represents the value observed for one mouse, and the geometric mean is shown. Inverted triangles, squares, and circles represent WT, Casp1KO and ASCKO mice, respectively. Closed and open circles represent noninfected and infected mice, respectively. IFN-g level in the serum (A) or in spleen lysates (B) was determined by ELISA. IFN-g (C), IL-10 (E), and IFN-b (F) transcripts levels in the spleen were determined by quantitative RT-PCR normalized to b-actin transcript level and expressed as fold increase over WT uninfected. Bacterial burden in the spleen was determined by a CFU assay (D). One experiment representative of three independent experiments (B), or results from two experiments representative of three independent experiments (C, E, F) or from three independent experiments (A, D) are shown. N.D., not detectable. *p , 0.05, **p , 0.01, ***p , 0.0001. 3850 ASC/IL-18–DEPENDENT REGULATION OF IFN-g IN Casp1KO MICE between Casp1KO and ASCKO animals. Collectively, these results in this study was not due to an impact on Dock2 expression or to suggested that in vivo during infection, ASC might specifically a global immune dysfunction of this mouse line. control IFN-g transcription in caspase-1–deficient animals. AIM2 is the inflammasome receptor detecting F. novicida within ASC controls IFN-g secretion in caspase-1–deficient NK cells the host cytosol (7–9) and triggering ASC-dependent pyroptosis during F. novicida infection and apoptosis in vitro (18, 23). IFN-g levels in the serum of To better understand the AIM2/ASC-dependent, caspase-1–inde- infected AIM2KO mice were significantly lower than in infected pendent mechanisms controlling IFN-g secretion, we determined Casp1KO mice (Fig. 2A). As previously described (8, 9), the bac- the cells responsible for secretion of this cytokine in infected terial burden in the spleen of AIM2KO mice was higher than in Casp1KO mice. Splenocytes were harvested at 48 h PI and were WT mice and similar to the bacterial burden observed in the stained for intracellular IFN-g following a 4-h ex vivo incubation spleen of Casp1KO and ASCKO mice (Fig. 2B). These data dem- in the presence of monensin. As presented in Fig. 4A, most IFN- onstrated that the AIM2/ASC pathway is largely responsible for g–producing cells in the spleen of infected animals were NK cells 2 2 controlling IFN-g level in vivo in Casp1KO mice during F. nov- (NK1.1+, TCRab cells). A smaller number of NK1.1 CD3ε+ icida infection. Interestingly, IFN-g level in the serum of AIM2KO TCRab+ cells was also producing IFN-g (Fig. 4A). The lower mice was not as low as IFN-g level in ASCKO mice. Another level of IFN-g detected in ASCKO mice as compared with inflammasome receptor might thus be activated in vivo during F. Casp1KO mice was not due to a reduction in splenic NK cell novicida infection, although less efficiently than AIM2, and might number because we observed a small but significant increase in signal through ASC to trigger IFN-g secretion. NK cells in the spleens of infected ASCKO mice compared with KO + infected Casp1 mice (Fig. 4B). The proportion of IFN-g NK Downloaded from The ASC-deficient mice used in this study are not defective for cells in ASCKO mice was significantly lower than in Casp1KO Dock2 expression and do not present any immune defect mice (Fig. 4C, 4D). Furthermore, the mean fluorescence inten- in lymphocyte population at steady-state sity of IFN-g staining was significantly lower in ASCKO than in A subset of ASC-deficient mice has been described as presenting Casp1KO NK cells (Fig. 4E). Finally, expression of the early ac- a deficit in Dock2 expression, which results in a drastic modifi- tivation marker CD69 (36) on NK cells showed a similar pattern KO cation of splenic lymphocyte numbers even in absence of any with an increased expression on NK cells from Casp1 compared http://www.jimmunol.org/ exogenous stimuli (28, 29). We assessed whether the deficiency in with NK cells from ASCKO mice (data not shown). Overall, these IFN-g production during F. novicida expression could be due to data indicated that ASC controlled the activation and the pro- Dock2 deficiency. We did not observe any modification of Dock2 duction of IFN-g by NK cells during infection in both caspase-1– protein expression level in bone marrow–derived macrophages or dependent and caspase-1–independent manners. As expected and in the spleen of our ASC-deficient line (Fig. 3A). Furthermore, we despite a much lower bacterial burden (Fig. 1D), a higher pro- did not observe any major differences in B, CD4 T, CD8 T, and portion of NK cells produced IFN-g in the spleen of WT animals KO regulatory T cell numbers in the spleen (Fig. 3B) of ASC mice (Fig. 4D). Similarly, the mean fluorescence intensity of IFN-g in compared with WT mice. Additionally, the proportions of these WT NK cells was higher than the one observed in inflammasome- subsets (Fig. 3C) and of monocytes and neutrophils (Fig. 3D) were deficient animals (Fig. 4E). by guest on September 27, 2021 similar in the blood of WT and ASCKO mice. Minor yet statistically significant differences were observed in splenic CD8 T cells NK cells from inflammasome-deficient animals have no and regulatory T cells and in blood B cell levels. The relevance of intrinsic defects in differentiation, degranulation, or these differences is still unclear. Collectively, these data indicated IFN-g secretion that the ASC-dependent, caspase-1–independent regulation observed ASC is ubiquitously expressed (Immunological Genome database) (37), and we thus wondered whether ASC could directly control IFN-g secretion in NK cells by affecting either their cell number, their maturation, or their responses. We undertook a thorough phenotyping of NK cells from WT, Casp1KO, and ASCKO animals. NK cells mature through four different stages characterized by expression of CD27 and CD11b (32). The four different subsets (CD272CD11b2,CD27+CD11b2,CD27+CD11b+,CD272CD11b+) were present in the same proportions in each mouse line (Fig. 5A, 5B), indicating that the absence of ASC or caspase-1 did not affect NK cell maturation. We next assessed the ability of NK cells from WT or inflammasome-deficient mice to degranulate as revealed by Lamp-1 (CD107) exposure at the plasma membrane. Similar degranulation was observed in response to cross-linking of the Ly49 (KLRA1) or the NKp46 (NCR1) receptors in the splenocytes from the different mice (Fig. 5C). Then, we checked whether ASCKO and Casp1KO NK cells had intrinsic defects in FIGURE 2. AIM2 and ASC control IFN-g levels in Casp1KO animals IFN-g production in response to either engagement of membrane 3 infected with F. novicida. Mice were infected intradermally with 5 3 10 receptors or to cytokines (IL-12, IL-12 plus IL-2, IL-12 plus IL- F. novicida CFU for 48 h. One point represents the value observed for one 18). As presented in Fig. 5D and 5E, no significant differences mouse, and the geometric mean is shown. Inverted triangles, squares, were observed for the in vitro ability of NK cells from inflam- circles, and triangles represent WT, Casp1KO, ASCKO, and AIM2KO mice, respectively. Closed and open circles represent noninfected and infected masome-deficient mice to respond to these signals by trigger- mice, respectively. IFN-g level in the serum was determined by ELISA ing IFN-g production. This phenotyping protocol performed on (A). Bacterial burden in the spleen was determined by a CFU assay (B). in vitro–stimulated NK cells suggested that the deficit presented KO One experiment representative of two independent experiments is shown. by ASC in secreting IFN-g during F. novicida infection was not *p , 0.05, **p , 0.01. due to an intrinsic NK cell defect. Similarly, when i.p. injection of The Journal of Immunology 3851

FIGURE 3. ASCKO mice do not present any defect in Dock2 expression and display normal splenic and blood cell populations. BMMs (left panel) and spleen extracts (right panel) from WT, Casp1KO,orASCKO mice were analyzed by Western blot for the expression of Dock2 (top panel), ASC (middle panel), and b-actin (bottom panel). One experiment representative of three independent experiments is shown (A). Expressions of ASC (middle panel)andb-actin (lower panel)areshown.Lymphocyte(B, C), monocyte, and neutrophil (D) proportions in the spleen (B) or the blood (C, D) of uninfected WT (n =6)orASCKO mice (n = 3) were determined by flow cytometry. Cells were defined as follow: B cells (B220+CD32), CD4T cells (CD3+CD4+B2202), CD8 T cells (CD3+CD8+B2202), regulatory T cells (CD3+CD4+CD25+), monocytes (CD115+), and neutrophils (Ly6G+). Means 6 SEM from one experiment are shown. *p , 0.05. Downloaded from polyinosinic-polycytidylic acid, a TLR3 agonist, was used to ac- marrow from Casp1KO (CD45.2+) and WT CD45.1+ mice. Eight tivate NK cells in vivo, we could not observe any differences weeks after reconstitution, mice were infected and the production in the percentage of IFN-g–producing NK cells between WT, of IFN-g by NK cells was investigated. As shown in Fig. 5F and Casp1KO, or ASCKO mice (data not shown). These results indicated 5G, within the same animal, NK cells derived from WT CD45.1+ that NK cells from WT, Casp1KO, and ASCKO mice had a similar and ASCKO (CD45.2+) progenitors or from WT CD45.1+ and intrinsic ability to produce IFN-g. To confirm this result, we per- Casp1KO (CD45.2+) progenitors had the same ability to produce http://www.jimmunol.org/ formed a mixed chimera experiment. Irradiated WT CD45.1+ mice IFN-g both in terms of percentage of IFN-g+ cells and level of IFN-g were reconstituted with a 1:1 mixture of bone marrow from ASCKO production as assessed by IFN-g staining intensity (Fig. 5F and not (CD45.2+) and WT CD45.1+ mice or with a 1:1 mixture of bone shown). Taken together, these results indicated that the very low pro- by guest on September 27, 2021

FIGURE 4. NK cells produce IFN-g in an ASC-dependent manner in infected Casp1KO mice. Splenocytes from mice uninfected or infected for 48 h with 5 3 103 F. novicida were isolated and analyzed by ﬂow cytometry following intracellular staining. FACS plots show cells following gating to exclude doublets (A, left panel), gating on live IFN-g+ cells (A, right panel), or gating on live CD3ε2TCRa2NK1.1+ cells (C). Concatenate plots (n = 2–5 mice, as indicated) from one experiment representative of three independent experiments are shown. The numbers indicate the percentage of total cells in each gate. (B) Splenic NK cell (live CD3ε2TCRa2NK1.1+) number was determined in three independent experiments. The percentage of IFN-g+ NK cells (D) and the mean ﬂuorescence intensity (MFI) (E) of IFN-g staining in this population was determined in two independent experiments. IFN-g MFI is presented as percentage of the average MFI in IFN-g+ NK cells from infected WT mice. *p , 0.05, **p , 0.01, ***p , 0.0001. 3852 ASC/IL-18–DEPENDENT REGULATION OF IFN-g IN Casp1KO MICE Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 5. NK cells from ASCKO mice do not have intrinsic defects in maturation, degranulation, or IFN-g production. Splenocytes were isolated from uninfected mice. The four stages of NK cells (defined based on forward scatter [FSC]-A, side scatter [SSC]-A, CD3ε2NK1.1+) maturation were analyzed by flow cytometryusingCD11bandCD27markers(A, B). Concatenates from three mice per genotype are shown (A). Splenocytes were stimulated ex vivo using an isotype control, anti-Ly49, or anti-NKp46 cross-linking Abs (C, D) or cytokines (E)during4hinthepresenceofanti-CD107(Lamp1)-FITCAbandmonensin.The percentage of Lamp1+ NK cells (based on FITC+, FSC-A, SSC-A, NK1.1+ and CD3ε2 staining) and IFN-g+ NK cells was determined. Means 6 SD (B–E) from three mice per genotype are shown. WT CD45.1+ mice were irradiated and reconstituted with a 1:1 mixture of WT CD45.1+ and Casp1KO (CD45.2+)orASCKO (CD45.2+) progenitors. Eight weeks after reconstitution, mice were infected with 5 3 103 F. novicida CFU. Forty-eight hours PI, splenocytes were collected and incubated for 4 h in the presence of monensin before staining. NK cells were analyzed by flow cytometry using the following gating steps (live, FSC-A, SSC-A, NK.1.1+CD3ε2). Concatenates (F) and quantification (G) of the ratio of CD45.1 (originating from WT progenitors) over CD45.2 (originating from KO progenitors) cells from four (WT-Casp1KO)tosix(WT-ASCKO) chimeric mice are shown (one experiment representative of two independent experiments). duction of IFN-g in infected ASCKO mice was not due to an intrinsic termined the concentration of those cytokines in the serum of inability of ASC-deficient NK cells to produce IFN-g. In agreement infected mice at 48 h PI (Fig. 6A–E). Among the cytokines tested, with the absence of NK cell–specific defects in ASCKO animals, we IL-18 showed a distinctive pattern with a significantly lower level also observed that upon F. novicida infection, fewer CD3ε+NK1.12 in the serum of ASCKO compared with Casp1KO mice. None of the TCRab+ cells produced IFN-g in ASCKO animals compared with other cytokines statistically differed between ASCKO and Casp1KO Casp1KO animals (data not shown). This latter result suggested that, mice. IL-18 secretion is regulated in WT mice by caspase-1 in irrespective of the IFN-g–producing cell type, the cytokine milieu in a posttranslational manner. We next assessed whether the dif- infected ASCKO mice might be suboptimal to induce IFN-g production. ferences in serum IL-18 levels observed between Casp1KO and KO KO ASC mice were due to regulation occurring at the transcrip- ASC controls IL-18 serum levels in Casp1 animals during tional or at the posttranslational stages. Although we noticed a F. novicida infection slight reduction in pro–IL-18 mRNA levels in ASCKO compared IFN-g secretion by NK cells during infection is controlled by with Casp1KO mice (Fig. 6F), this difference was not statistically cytokines such as IL-2, IL-12, IL-15, IL-10, and IL-18. We de- significant. These results suggested that in Casp1KO animals, ASC The Journal of Immunology 3853 Downloaded from http://www.jimmunol.org/

FIGURE 6. IL-18 level is specifically regulated in an ASC-dependent manner in infected Casp1KO mice. Mice were infected with 5 3 103 F. novicida CFU. At 48 h PI, the serum was analyzed by multiplex immunoassay to determine IL-2 (A), IL-12p70 (B), IL-15 (C), IL-10 (D), and IL-18 (E) levels. Pro– IL-18 (F) and pro–IL-1b (G) transcript levels in the spleen were analyzed by quantitative RT-PCR, normalized to b-actin, and expressed as fold increase over WT uninfected level. Results from three independent experiments are shown (A–G). *p , 0.05, **p , 0.01, ***p , 0.0001. regulated IL-18 production by a posttranslational mechanism dur- (Fig. 7A). Similar results were observed by determining IFN-g by guest on September 27, 2021 ing F. novicida infection. transcript levels in the spleen of infected animals (Fig. 7B). These IL-1b can modulate IFN-g production by NK cells (38–40). As results clearly indicated that the ASC-dependent regulation of IL- expected, a strong increase in pro–IL-1b transcript levels in the 18 levels in Casp1KO mice contributed to IFN-g secretion. Sur- spleen was observed upon infection. The increase in pro–IL-1b prisingly, we observed a significant decrease in the IFN-g serum transcript was similar in the different mouse lines (Fig. 6G). As levels in ASCKO mice upon IL-18 neutralization, indicating that previously described (16), IL-1b levels in the serum of infected even in ASCKO mice, bioactive IL-18 was present and participated mice were below the limit of detection of the ELISA/immunoassay in IFN-g production. The single injection of IL-18 neutralizing Ab (data not shown). To investigate a possible role of IL-1 in regu- at 24 h PI in WT mice led to a 10-fold increase in the bacterial lating IFN-g secretion, IL-1 signaling was blocked by injecting burden, but the reduction in IFN-g level was not statistically sig- Casp1KO mice with IL-1R antagonist (Anakinra). Although nificant (Supplemental Fig. 2). Further experiments are required Anakinra efficiently blocked the ability of F. novicida–infected to determine whether IFN-g production in F. no vic id a–infected macrophages to induce KC production by fibroblasts (Supple- WT mice is partially independent of IL-18. mental Fig. 1A), no significant difference was observed in IFN-g Caspase-8 triggers IL-18 release from 293T cells reconstituted levels in the serum (Supplemental Fig. 1B) and in the bacterial with inflammasome-like complexes burden in IL-1R antagonist–injected Casp1KO mice compared with PBS-injected mice (Supplemental Fig. 1C). These results Macrophages are a key source of IL-18 (37). We observed that KO ruled out a major role of IL-1b in regulating IFN-g levels in upon F. novicida infection of BMMs, Casp1 macrophages re- Casp1KO mice. leased low amounts of IL-18 that were too close to the limit of detection of the IL-18 ELISA to be accurately quantified (,50 pg/ Bioactive IL-18 is responsible for the ASC-dependent IFN-g KO ml, data not shown). To test the functionality of the IL-18 protein production in Casp1 mice released by Casp1KO macrophage, we took advantage of the Bioactive IL-18 is produced and secreted by caspase-1–dependent ability of splenocytes to produce IFN-g to set up an IL-18 bio- mechanisms (41, 42). We thus wondered whether the IL-18 assay. WT, Casp1KO, or ASCKO BMMs were infected with F. detected in large quantities in the serum of Casp1KO animals in novicida for 1 h, extensively washed, and incubated with genta- an ASC-dependent manner was bioactive and could account for micin for an additional hour to eliminate any remaining extra- the higher level of IFN-g secretion observed in Casp1KO animals cellular bacteria. At 2 h PI, splenocytes were added onto BMMs compared with ASCKO animals. At 24 h PI, we injected an IL-18 and IFN-g production in the supernatant was measured at 12 h PI. neutralizing Ab or its isotype control to Casp1KO or ASCKO mice. We did not observe any IFN-g production in the absence of Neutralization of IL-18 completely abolished the difference in splenocytes (data not shown) or in the absence of infected mac- IFN-g levels in the serum between ASCKO and Casp1KO mice rophages (Fig. 8A). Infected WT and Casp1KO BMM triggered 3854 ASC/IL-18–DEPENDENT REGULATION OF IFN-g IN Casp1KO MICE

bated with infected Casp1KO BMMs. The caspase-8 inhibitor zIETD-FMK was one of the most effective inhibitors to reduce IFN-g secretion. However, other z-peptide-FMK inhibitors, in- cluding zYVAD-FMK, efficiently reduced IFN-g secretion (Sup- plemental Fig. 3). This result was consistent with the described poor selectivity of FMK inhibitors (43) and indicated that peptide inhibitors could not be used to discriminate which caspase controlled IFN-g secretion in this assay. To assess more specifically the role of the different caspases in pro–IL-18 cleavage, we reconstituted inflammasome-like complexes in 293T cells (Fig. 9A) by coexpressing AIM2, ASC, pro–IL-18, and either one of the caspases that are activated upon F. novicida infection of Casp1KO macrophages (18). Expression of ASC together with caspase-1 or caspase-8 resulted in an increase in IL-18 levels in the cell supernatant as detected by ELISA. In contrast, the coexpression of FIGURE 7. Bioactive IL-18 is responsible for the ASC-dependent IFN-g ASC with the other caspases did not increase IL-18 levels in this KO 3 3 production in infected Casp1 mice. Mice were infected with 5 10 assay (Fig. 9B). Furthermore, this increase in IL-18 levels in the F. novicida CFU. At 24 h PI, mice were injected i.p. with 500 mg anti–IL- supernatant of cells expressing AIM2, ASC, and caspase-8 cor- 18 neutralizing Ab or an IgG1 isotype control. At 48 h PI, mice were Downloaded from euthanized. IFN-g in the serum was quantified by ELISA (A). IFN-g related with an increase in IL-18 cleavage as detected by Western transcript level in the spleen was determined by quantitative RT-PCR, blot analysis (Supplemental Fig. 4). This result suggested that in normalized to b-actin, and expressed as fold increase over WT uninfected the absence of caspase-1, the specific activation of caspase-8 by level (B). Results from two independent experiments are shown. **p , the AIM2/ASC pyroptosome triggers pro–IL-18 cleavage and IL- 0.01, ***p , 0.0001. 18 secretion although at much lower levels than what is observed in caspase-1–proficient cells. KO IFN-g production by splenocytes whereas ASC BMMs did not http://www.jimmunol.org/ (Fig. 8A). The ability of infected WT and Casp1KO BMM to ac- Discussion tivate splenocytes to produce IFN-g was due to the production of Although most of the attention on inflammasome-dependent bioactive IL-18 because IL-18 neutralization completely abolished cytokines has focused on IL-1b, IL-18 is also playing a critical IFN-g production by splenocytes. Conversely, addition of rIL-18 role in various inflammatory situations (44). IL-18 is long known abolished the difference in IFN-g release between splenocytes for its ability to trigger IFN-g production by T and NK cells in coincubated with Casp1KO BMMs and splenocytes coincubated synergy with IL-12 (45). Furthermore, immunoregulatory func- with ASCKO BMMs (Fig. 8B). These in vitro assays thus indi- tions of IL-18 have recently emerged (33, 46–50) highlighting the cated that bioactive IL-18 can be secreted in an ASC-dependent, importance of precisely understanding the molecular mechanisms caspase-1–independent manner in F. novicida–infected Casp1KO regulating its production and secretion in homeostatic and in- by guest on September 27, 2021 BMMs. flammatory conditions. In contrast to pro–IL-1b, pro–IL-18 is We and others have recently reported that caspase-8 is activated constitutively expressed (51). Caspase-1 is clearly established as in an AIM2/ASC-dependent, caspase-1–independent manner in the primary enzyme responsible for cleavage of pro–IL-18 into Casp1KO macrophages (18, 23). Therefore, we tested whether bioactive IL-18 (41, 42). In agreement with Mariathasan et al. caspase-8 could trigger IL-18 secretion following its activation (16), we could also verify a key role for caspase-1 in generating within the AIM2 pyroptosome. We first tested whether caspase bioactive IL-18. Indeed, during F. novicida infection, we observed inhibitors could block IFN-g production by splenocytes coincu- a large increase in IL-18 and IFN-g in caspase-1–proficient

FIGURE 8. Bioactive IL-18 released by F. novicida–infected Casp1KO BMMs is responsible for IFN-g production by splenocytes in vitro. WT, Casp1KO, or ASCKO BMMs (105) were infected with F. novicida for 1 h, extensively washed, and incubated in gentamicin for an additional hour to eliminate extracellular bacteria. At 2 h PI, 5 3 105 splenocytes were added onto BMMs. (A) Culture medium was supplemented with anti–IL-18 neutralizing Ab or an IgG1 isotype control (400 ng/ml). (B) Cells were cultured in medium supplemented (or not) with rIL-18 (200 pg/ml). (A and B) Supernatants were collected at 12 h PI. IFN-g concentration in culture medium was determined by ELISA. One experiment representative of three independent experiments is shown. **p , 0.01. The Journal of Immunology 3855

FIGURE 9. Caspase-8 triggers the release of IL-18 in an ASC-dependent manner upon ectopic expression in 293T cells. (A) Diagram of inflammasome- like pathway reconstitution in 293T cells. (B) The different procaspases (procaspase-1, -2, -3,-8, -9) were screened for a functional interaction with the AIM2/ASC complex leading to IL-18 secretion. The inflammasome-like pathway was reconstituted by cotransfecting 293T cells with plasmids expressing AIM2, ASC, pro–IL-18, and either of procaspases. Interaction of ASC with procaspase-1 and procaspase-8 led to an increase in IL-18 release. Analysis of IL-18 concentration in culture medium of 293T cells was determined by ELISA. Results are presented as IL-18 concentration fold increase between Downloaded from 293T cells cotransfected with AIM2, ASC, pro–IL-18, and either of the procaspase versus cells undergoing the same cotransfection except for the absence of ASC-encoding plasmid. Raw values were typically 50–300 pg/ml IL-18 in the absence of ASC and reached 6000 pg/ml upon ASC addition. Data are shown as means 6 SEM (n = 5). One experiment representative of three independent experiments is shown. **p , 0.01. compared with inflammasome-deficient mice (Figs. 1, 6). How- also reduced IFN-g level in infected ASCKO mice to the level

ever, we observed a similar decrease in IFN-g level between observed in uninfected mice. This result suggests that at least three http://www.jimmunol.org/ infected WT and Casp1KO mice than between infected Casp1KO bioactive IL-18–generating pathways can occur in F. novicida– and ASCKO mice or Casp1KO and AIM2KO mice. This observation infected mice: 1) the canonical ASC-dependent caspase-1–depen- indicates the presence of a robust alternative pathway able to dent pathway, 2) a noncanonical ASC-dependent caspase-1–inde- generate IFN-g in absence of caspase-1–mediated pro–IL-18 pendent pathway, and 3) a yet to be characterized ASC-indepen- cleavage. Several alternative processing pathways have been de- dent pathway. scribed in vitro for the processing of IL-18 (52). Indeed, granzyme The importance of such pathways in caspase-1–proficient ani- B (53), elastase, cathepsin G (54), proteinase 3 (55), caspase-4 mals is still unclear. Indeed, in our experimental settings, caspase-1 (and its murine homolog also known as caspase-11) (41, 42), is playing a major role in the IL-18/IFN-g cascade. However, caspase-5 (42), and caspase-8 (56) have the in vitro ability to there are increasing connections between apoptotic and inflam- by guest on September 27, 2021 cleave pro–IL-18 and to generate an IFN-g–inducing mature cy- masome pathways (21, 56–59) suggesting that in certain situations tokine. However, to our knowledge, the presence and the relevance this pathway might be important. Furthermore, the importance of of the caspase-1–independent pathway remain to be established redundancy in innate immune pathways is likely to be very im- in vivo. Our results using caspase-1–deficient animals clearly portant with regard to the ability of pathogens to inhibit certain demonstrate that in vivo bioactive IL-18 can be generated in an innate immune responses and with regard to the frequency of host ASC-dependent, caspase-1–independent manner. Furthermore, mutations (60). Indeed, caspase-1 alleles encoding inactive en- because caspase-1–deficient mice are also deficient for caspase-11 zyme have been recently identified in humans (61) highlighting (caspase-4) (14) and mice do not posses a caspase-5 homolog, we the possible importance of caspase-1 alternative pathways. conclude that bioactive IL-18 can be generated in vivo in an ASC- dependent manner but independently of inflammatory caspases. Acknowledgments Recently, we have identified that ASC activates caspase-8 in KO We acknowledge the contribution of the Plateau de Biologie Expe´rimental Casp1 BMMs and bone marrow–derived dendritic cells leading de la Souris (D. Gallouche, G. Claveau, and S. Blanc) platform of the KO to apoptosis of F. novicida–infected Casp1 cells. In this study, Structure Fe´de´rative de Recherche BioSciences Gerland–Lyon Sud we observed that when reconstituting an inflammasome-like (UMS344/US8), of the Celphedia AniRA phenotyping platform. We are complex in 293T cells, caspase-8 cleaved pro–IL-18 and trig- grateful to D. Kaiserlian and B. Dubois for the B6.SJL-Ptprca Pepcb/BoyJ gered the release of IL-18 in an ASC-dependent manner, sug- mice. We thank Vishva Dixit (Genentech, South San Francisco, CA) and KO gesting that caspase-8 might be involved in generating IL-18 Giuseppe Teti (University of Messina, Messina, Italy) for AIM2 mice in vivo in Casp1KO animals. In agreement with our study, Latz and Beńe´dicte Py for critical reading of the manuscript. This work benefit- and collaborators (56) recently identified caspase-8 as the enzyme ted from data assembled by the Immunological Genome Consortium. responsible for IL-18 maturation in response to Fas engagement. In response to murine CMV infection, Rathinam et al. (7) have Disclosures observed that an AIM2/ASC-dependent pathway is controlling The authors have no financial conflicts of interest. both IL-18 serum level and IFN-g production by NK cells. Al- though the phenotype of caspase-1–deficient animals was not in- References vestigated in response to murine CMV, it is tempting to speculate 1. Martinon, F., K. Burns, and J. Tschopp. 2002. The inflammasome: a molecular that both caspase-1–dependent and caspase-1–independent path- platform triggering activation of inflammatory caspases and processing of proIL- ways might regulate IFN-g production by NK cells during viral b. Mol. Cell 10: 417–426. infection. 2. Roberts, T. L., A. Idris, J. A. Dunn, G. M. Kelly, C. M. Burnton, S. Hodgson, L. L. Hardy, V. Garceau, M. J. Sweet, I. L. Ross, et al. 2009. HIN-200 proteins Interestingly, although the neutralizing Ab to IL-18 drastically regulate caspase activation in response to foreign cytoplasmic DNA. Science KO reduced IFN-g in the serum and the spleen of Casp1 mice, it 323: 1057–1060. 3856 ASC/IL-18–DEPENDENT REGULATION OF IFN-g IN Casp1KO MICE

3. Hornung, V., A. Ablasser, M. Charrel-Dennis, F. Bauernfeind, G. Horvath, development of experimental autoimmune encephalomyelitis. J. Immunol. 184: D. R. Caffrey, E. Latz, and K. A. Fitzgerald. 2009. AIM2 recognizes cytosolic 4610–4614. dsDNA and forms a caspase-1-activating inflammasome with ASC. Nature 458: 27. McElvania Tekippe, E., I. C. Allen, P. D. Hulseberg, J. T. Sullivan, 514–518. J. R. McCann, M. Sandor, M. Braunstein, and J. P. Ting. 2010. Granuloma 4. Fernandes-Alnemri, T., J. W. Yu, P. Datta, J. Wu, and E. S. Alnemri. 2009. AIM2 formation and host defense in chronic Mycobacterium tuberculosis infection activates the inflammasome and cell death in response to cytoplasmic DNA. requires PYCARD/ASC but not NLRP3 or caspase-1. PLoS ONE 5: e12320. Nature 458: 509–513. 28. Ippagunta, S. K., R. K. Malireddi, P. J. Shaw, G. A. Neale, L. V. Walle, Y. Fukui, 5. Bu¨rckstu¨mmer, T., C. Baumann, S. Blu¨ml, E. Dixit, G. Du¨rnberger, H. Jahn, D. R. Green, M. Lamkanfi, and T. D. Kanneganti. 2012. Addendum: defective M. Planyavsky, M. Bilban, J. Colinge, K. L. Bennett, and G. Superti-Furga. Dock2 expression in a subset of ASC-deficient mouse lines. Nat. Immunol. 13: 2009. An orthogonal proteomic-genomic screen identifies AIM2 as a cytoplas- 701–702. mic DNA sensor for the inflammasome. Nat. Immunol. 10: 266–272. 29. Ippagunta, S. K., R. K. Malireddi, P. J. Shaw, G. A. Neale, L. V. Walle, 6. Sauer, J. D., C. E. Witte, J. Zemansky, B. Hanson, P. Lauer, and D. A. Portnoy. D. R. Green, Y. Fukui, M. Lamkanfi, and T. D. Kanneganti. 2011. The inflam- 2010. Listeria monocytogenes triggers AIM2-mediated pyroptosis upon infre- masome adaptor ASC regulates the function of adaptive immune cells by con- quent bacteriolysis in the macrophage cytosol. Cell Host Microbe 7: 412–419. trolling Dock2-mediated Rac activation and actin polymerization. Nat. Immunol. 7. Rathinam, V. A., Z. Jiang, S. N. Waggoner, S. Sharma, L. E. Cole, L. Waggoner, 12: 1010–1016. S. K. Vanaja, B. G. Monks, S. Ganesan, E. Latz, et al. 2010. The AIM2 30. Mariathasan, S., K. Newton, D. M. Monack, D. Vucic, D. M. French, W. P. Lee, inflammasome is essential for host defense against cytosolic bacteria and DNA M. Roose-Girma, S. Erickson, and V. M. Dixit. 2004. Differential activation of viruses. Nat. Immunol. 11: 395–402. the inflammasome by caspase-1 adaptors ASC and Ipaf. Nature 430: 213–218. 8. Jones, J. W., N. Kayagaki, P. Broz, T. Henry, K. Newton, K. O’Rourke, S. Chan, 31. Lochner, M., and I. Fo¨rster. 2002-2003. Anti-interleukin-18 therapy in murine J. Dong, Y. Qu, M. Roose-Girma, et al. 2010. Absent in melanoma 2 is required models of inflammatory bowel disease. Pathobiology 70: 164–169. for innate immune recognition of Francisella tularensis. Proc. Natl. Acad. Sci. 32. Chiossone, L., J. Chaix, N. Fuseri, C. Roth, E. Vivier, and T. Walzer. 2009. USA 107: 9771–9776. Maturation of mouse NK cells is a 4-stage developmental program. Blood 113: 9. Fernandes-Alnemri, T., J. W. Yu, C. Juliana, L. Solorzano, S. Kang, J. Wu, 5488–5496. P. Datta, M. McCormick, L. Huang, E. McDermott, et al. 2010. The AIM2 33. Chaix, J., M. S. Tessmer, K. Hoebe, N. Fuse´ri, B. Ryffel, M. Dalod,

inflammasome is critical for innate immunity to Francisella tularensis. Nat. L. Alexopoulou, B. Beutler, L. Brossay, E. Vivier, and T. Walzer. 2008. Cutting Downloaded from Immunol. 11: 385–393. edge: priming of NK cells by IL-18. J. Immunol. 181: 1627–1631. 10. Keim, P. S., A. Johansson, and D. M. Wagner. 2007. Molecular epidemiology, 34. Henry, T., G. S. Kirimanjeswara, T. Ruby, J. W. Jones, K. Peng, M. Perret, L. Ho, evolution, and ecology of Francisella. Ann. N. Y. Acad. Sci. 1105: 30–66. J. D. Sauer, Y. Iwakura, D. W. Metzger, and D. M. Monack. 2010. Type I IFN 11. Hall, J. D., M. D. Woolard, B. M. Gunn, R. R. Craven, S. Taft-Benz, signaling constrains IL-17A/F secretion by gd T cells during bacterial infections. J. A. Frelinger, and T. H. Kawula. 2008. Infected-host-cell repertoire and cellular J. Immunol. 184: 3755–3767. response in the lung following inhalation of Francisella tularensis Schu S4, LVS, 35. Henry, T., A. Brotcke, D. S. Weiss, L. J. Thompson, and D. M. Monack. 2007. or U112. Infect. Immun. 76: 5843–5852. Type I interferon signaling is required for activation of the inflammasome during 12. Golovliov, I., V. Baranov, Z. Krocova, H. Kovarova, and A. Sjo¨stedt. 2003. An Francisella infection. J. Exp. Med. 204: 987–994. attenuated strain of the facultative intracellular bacterium Francisella tularensis 36. Lanier, L. L., D. W. Buck, L. Rhodes, A. Ding, E. Evans, C. Barney, and http://www.jimmunol.org/ can escape the phagosome of monocytic cells. Infect. Immun. 71: 5940–5950. J. H. Phillips. 1988. Interleukin 2 activation of natural killer cells rapidly induces 13. Clemens, D. L., B. Y. Lee, and M. A. Horwitz. 2004. Virulent and avirulent the expression and phosphorylation of the Leu-23 activation antigen. J. Exp. strains of Francisella tularensis prevent acidification and maturation of their Med. 167: 1572–1585. phagosomes and escape into the cytoplasm in human macrophages. Infect. 37. Heng, T. S., M. W. Painter; Immunological Genome Project Consortium. 2008. Immun. 72: 3204–3217. The Immunological Genome Project: networks of gene expression in immune 14. Kayagaki, N., S. Warming, M. Lamkanfi, L. Vande Walle, S. Louie, J. Dong, cells. Nat. Immunol. 9: 1091–1094. K. Newton, Y. Qu, J. Liu, S. Heldens, et al. 2011. Non-canonical inflammasome 38. van de Wetering, D., R. A. de Paus, J. T. van Dissel, and E. van de Vosse. 2009. activation targets caspase-11. Nature 479: 117–121. Salmonella induced IL-23 and IL-1b allow for IL-12 production by monocytes 15. Henry, T., and D. M. Monack. 2007. Activation of the inflammasome upon and Mphi1 through induction of IFN-gamma in CD56 NK/NK-like T cells. PLoS Francisella tularensis infection: interplay of innate immune pathways and vir- ONE 4: e8396.

ulence factors. Cell. Microbiol. 9: 2543–2551. 39. Hunter, C. A., R. Chizzonite, and J. S. Remington. 1995. IL-1b is required for by guest on September 27, 2021 16. Mariathasan, S., D. S. Weiss, V. M. Dixit, and D. M. Monack. 2005. Innate IL-12 to induce production of IFN-g by NK cells. A role for IL-1b in the T cell- immunity against Francisella tularensis is dependent on the ASC/caspase-1 axis. independent mechanism of resistance against intracellular pathogens. J. Immu- J. Exp. Med. 202: 1043–1049. nol. 155: 4347–4354. 17.Miao,E.A.,I.A.Leaf,P.M.Treuting,D.P.Mao,M.Dors,A.Sarkar, 40. Hughes, T., B. Becknell, A. G. Freud, S. McClory, E. Briercheck, J. Yu, C. Mao, S. E. Warren, M. D. Wewers, and A. Aderem. 2010. Caspase-1-induced pyroptosis C. Giovenzana, G. Nuovo, L. Wei, et al. 2010. Interleukin-1b selectively expands is an innate immune effector mechanism against intracellular bacteria. Nat. and sustains interleukin-22+ immature human natural killer cells in secondary Immunol. 11: 1136–1142. lymphoid tissue. Immunity 32: 803–814. 18. Pierini, R., C. Juruj, M. Perret, C. L. Jones, P. Mangeot, D. S. Weiss, and 41. Gu, Y., K. Kuida, H. Tsutsui, G. Ku, K. Hsiao, M. A. Fleming, N. Hayashi, T. Henry. 2012. AIM2/ASC triggers caspase-8-dependent apoptosis in Franci- K. Higashino, H. Okamura, K. Nakanishi, et al. 1997. Activation of interferon-g sella-infected caspase-1-deficient macrophages. Cell Death Differ. 19: 1709– inducing factor mediated by interleukin-1b converting enzyme. Science 275: 1721. 206–209. 19. Akhter, A., K. Caution, A. Abu Khweek, M. Tazi, B. A. Abdulrahman, 42. Ghayur, T., S. Banerjee, M. Hugunin, D. Butler, L. Herzog, A. Carter, L. Quintal, D. H. Abdelaziz, O. H. Voss, A. I. Doseff, H. Hassan, A. K. Azad, et al. 2012. L. Sekut, R. Talanian, M. Paskind, et al. 1997. Caspase-1 processes IFN-g-in- Caspase-11 promotes the fusion of phagosomes harboring pathogenic bacteria ducing factor and regulates LPS-induced IFN-g production. Nature 386: 619– with lysosomes by modulating actin polymerization. Immunity 37: 35–47. 623. 20. Puri, A. W., P. Broz, A. Shen, D. M. Monack, and M. Bogyo. 2012. Caspase-1 43. Pereira, N. A., and Z. Song. 2008. Some commonly used caspase substrates and activity is required to bypass macrophage apoptosis upon Salmonella infection. inhibitors lack the specificity required to monitor individual caspase activity. Nat. Chem. Biol. 8: 745–747. Biochem. Biophys. Res. Commun. 377: 873–877. 21. Gringhuis, S. I., T. M. Kaptein, B. A. Wevers, B. Theelen, M. van der Vlist, 44. Dinarello, C. A. 2009. Immunological and inflammatory functions of the T. Boekhout, and T. B. Geijtenbeek. 2012. Dectin-1 is an extracellular pathogen interleukin-1 family. Annu. Rev. Immunol. 27: 519–550. sensor for the induction and processing of IL-1b via a noncanonical caspase-8 45. Okamura, H., H. Tsutsi, T. Komatsu, M. Yutsudo, A. Hakura, T. Tanimoto, inflammasome. Nat. Immunol. 13: 246–254. K. Torigoe, T. Okura, Y. Nukada, K. Hattori, et al. 1995. Cloning of a new 22. Abdelaziz, D. H., M. A. Gavrilin, A. Akhter, K. Caution, S. Kotrange, cytokine that induces IFN-g production by T cells. Nature 378: 88–91. A. A. Khweek, B. A. Abdulrahman, Z. A. Hassan, F. Z. El-Sharkawi, S. S. Bedi, 46. Oertli, M., M. Sundquist, I. Hitzler, D. B. Engler, I. C. Arnold, S. Reuter, et al. 2011. Asc-dependent and independent mechanisms contribute to restriction J. Maxeiner, M. Hansson, C. Taube, M. Quiding-Ja¨rbrink, and A. Mu¨ller. 2012. of Legionella pneumophila infection in murine macrophages. Front. Microbiol. DC-derived IL-18 drives Treg differentiation, murine Helicobacter pylori-spe- 2: 18. cific immune tolerance, and asthma protection. J. Clin. Invest. 122: 1082–1096. 23. Sagulenko, V., S. J. Thygesen, D. P. Sester, A. Idris, J. A. Cridland, P. R. Vajjhala, 47. Terme, M., E. Ullrich, L. Aymeric, K. Meinhardt, J. D. Coudert, M. Desbois, T. L. Roberts, K. Schroder, J. E. Vince, J. M. Hill, J. Silke, and K. J. Stacey. 2013 F. Ghiringhelli, S. Viaud, B. Ryffel, H. Yagita, et al. 2012. Cancer-induced AIM2 and NLRP3 inflammasomes activate both apoptotic and pyroptotic death immunosuppression: IL-18-elicited immunoablative NK cells. Cancer Res. 72: pathways via ASC. Cell Death Differ. DOI: 10.1038/cdd.2013.37 2757–2767. 24. Ellebedy, A. H., C. Lupfer, H. E. Ghoneim, J. DeBeauchamp, T. D. Kanneganti, 48. Dupaul-Chicoine, J., G. Yeretssian, K. Doiron, K. S. Bergstrom, C. R. McIntire, and R. J. Webby. 2011. Inflammasome-independent role of the apoptosis- P. M. LeBlanc, C. Meunier, C. Turbide, P. Gros, N. Beauchemin, et al. 2010. associated speck-like protein containing CARD (ASC) in the adjuvant effect Control of intestinal homeostasis, colitis, and colitis-associated colorectal cancer of MF59. Proc. Natl. Acad. Sci. USA 108: 2927–2932. by the inflammatory caspases. Immunity 32: 367–378. 25. Kolly, L., M. Karababa, L. A. Joosten, S. Narayan, R. Salvi, V. Pe´trilli, 49. Zaki, M. H., K. L. Boyd, P. Vogel, M. B. Kastan, M. Lamkanfi, and J. Tschopp, W. B. van den Berg, A. K. So, and N. Busso. 2009. Inflammatory T. D. Kanneganti. 2010. The NLRP3 inflammasome protects against loss of role of ASC in antigen-induced arthritis is independent of caspase-1, NALP-3, epithelial integrity and mortality during experimental colitis. Immunity 32: 379– and IPAF. J. Immunol. 183: 4003–4012. 391. 26. Shaw, P. J., J. R. Lukens, S. Burns, H. Chi, M. A. McGargill, and 50. Hitzler, I., A. Sayi, E. Kohler, D. B. Engler, K. N. Koch, W. D. Hardt, and T. D. Kanneganti. 2010. Cutting edge: critical role for PYCARD/ASC in the A. Mu¨ller. 2012. Caspase-1 has both proinflammatory and regulatory properties The Journal of Immunology 3857

in Helicobacter infections, which are differentially mediated by its substrates IL- et al. 2012. Cutting edge: FAS (CD95) mediates noncanonical IL-1b and IL-18 1b and IL-18. J. Immunol. 188: 3594–3602. maturation via caspase-8 in an RIP3-independent manner. J. Immunol. 189: 51. Puren, A. J., G. Fantuzzi, and C. A. Dinarello. 1999. Gene expression, synthesis, 5508–5512. and secretion of interleukin 18 and interleukin 1b are differentially regulated in 57. Miwa, K., M. Asano, R. Horai, Y. Iwakura, S. Nagata, and T. Suda. 1998. human blood mononuclear cells and mouse spleen cells. Proc. Natl. Acad. Sci. Caspase 1-independent IL-1b release and inflammation induced by the apoptosis USA 96: 2256–2261. inducer Fas ligand. Nat. Med. 4: 1287–1292. 52. van de Veerdonk, F. L., M. G. Netea, C. A. Dinarello, and L. A. Joosten. 2011. 58. Labbe´, K., C. R. McIntire, K. Doiron, P. M. Leblanc, and M. Saleh. 2011. Inflammasome activation and IL-1b and IL-18 processing during infection. Cellular inhibitors of apoptosis proteins cIAP1 and cIAP2 are required for ef- Trends Immunol. 32: 110–116. ficient caspase-1 activation by the inflammasome. Immunity 35: 897–907. 53. Omoto, Y., K. Yamanaka, K. Tokime, S. Kitano, M. Kakeda, T. Akeda, 59. Shimada, K., T. R. Crother, J. Karlin, J. Dagvadorj, N. Chiba, S. Chen, I. Kurokawa, E. C. Gabazza, H. Tsutsui, N. Katayama, et al. 2010. Granzyme B V. K. Ramanujan, A. J. Wolf, L. Vergnes, D. M. Ojcius, et al. 2012. Oxidized is a novel interleukin-18 converting enzyme. J. Dermatol. Sci. 59: 129–135. Im- 54. Robertson, S. E., J. D. Young, S. Kitson, A. Pitt, J. Evans, J. Roes, D. Karaoglu, mitochondrial DNA activates the NLRP3 inflammasome during apoptosis. L. Santora, T. Ghayur, F. Y. Liew, et al. 2006. Expression and alternative pro- munity 36: 401–414. cessing of IL-18 in human neutrophils. Eur. J. Immunol. 36: 722–731. 60. Kondrashov, A. 2012. Genetics: the rate of human mutation. Nature 488: 467– 55. Sugawara, S., A. Uehara, T. Nochi, T. Yamaguchi, H. Ueda, A. Sugiyama, 468. K. Hanzawa, K. Kumagai, H. Okamura, and H. Takada. 2001. Neutrophil pro- 61. Luksch, H., M. J. Romanowski, O. Chara, V. Tungler, E. R. Caffarena, teinase 3-mediated induction of bioactive IL-18 secretion by human oral epi- M. C. Heymann, P. Lohse, I. Aksentijevich, E. F. Remmers, S. Flecks, et al. thelial cells. J. Immunol. 167: 6568–6575. 2013. Naturally occurring genetic variants of human caspase-1 differ consid- 56. Bossaller, L., P. I. Chiang, C. Schmidt-Lauber, S. Ganesan, W. J. Kaiser, erably in structure and the ability to activate interleukin-1b. Hum. Mutat. 34: V. A. Rathinam, E. S. Mocarski, D. Subramanian, D. R. Green, N. Silverman, 122–131. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021 Supplemental figures

Figure S1: IL-1β plays no major role in regulating IFN-γ levels in Casp1KO animals infected with F. novicida

(A) IL-1 Receptor antagonist (IL-1Ra, Kineret®, SOBI) blocked IL-1 signaling in a co-culture of

F.novicida-infected WT BMM and 3T3 cells (a mouse embryonic fibroblast cell line that releases KC in response to IL-1 stimulation). 105 WT BMM were infected with F. novicida at a MOI of 100:1 for 1 h, extensively washed and incubated with gentamicin. Supernatants from infected BMM were collected at 7 h PI and transferred onto 3T3 cells (5x104 cells in 0.3 cm2 wells). Supernatants were collected 9 h later. KC levels were determined by ELISA (Duoset R&D systems).

(B and C) Casp1KO mice were infected intradermally with 5x103 F. novicida cfu for 48h.

Intraperitoneal injection of IL-1Ra was performed 8 h, 24 h and 32 h post-infection (50 µg per mouse per injection). One point represents the value observed for one mouse, geometric mean is shown.

Triangles and squares represent PBS- and IL-1Ra-injected mice, respectively. IFN-γ level in the serum was determined by ELISA (B). Bacterial burden in the spleen was determined by a colony forming unit assay (C).

Figure S2: IL-18 neutralizing antibody increases bacterial burden and reduces in a non-statistical manner IFN-γ levels in the serum of F.novicida-infected WT mice.

Mice were infected with 5x103 F. novicida cfu. At 24 h PI, mice were injected intraperitoneally with

500 µg anti-IL-18 neutralizing antibody or an IgG1 isotype control. Mice were euthanized at 48 h PI.

Bacterial burden in the spleen was determined by a colony forming unit assay (A). IFN-γ in the serum was quantified by ELISA (B).

Figure S3: The low specificity of FMK-peptide caspase inhibitors precludes identification of the caspase involved in the IL-18/IFN-γ cascade.

105 Casp1KO BMM were infected with F. novicida for 1h, extensively washed and incubated in gentamicin for an additional hour to eliminate extracellular bacteria. At 2 h PI, 5 x 105 ASCKO splenocytes were added onto BMM in presence on not of either caspase inhibitor (zIETD-FMK, zDEVD-FMK, zYVAD-FMK, zLEHD-FMK, zVDVAD-FMK, Bachem, 50 µM). Supernatants were collected at 12 h PI. IFN-γ concentration in culture medium was determined by ELISA. The described primary caspase target is indicated for each inhibitor. One experiment representative of three independent experiments is shown.

Figure S4: caspase-8 cleaves proIL-18 in an ASC-dependent manner upon ectopic expression in 293T cells. Procaspases-1, -8 or-9 were screened for a functional interaction with the AIM2/ASC complex leading to proIL-18 cleavage. Inflammasome-like pathway was reconstituted by cotransfecting 293T cells with plasmids expressing AIM2, ASC, pro-IL-18 and either of procaspases. Interaction of ASC with procaspase-1 or procaspase-8 led to proIL-18 cleavage, as determined by western blot (anti IL-18 antibody from MBL, ref. D046-3). Band intensity for active IL-18 was normalized to the proIL-18 band. Ratios of results obtained in presence and absence of ASC are shown. β-actin western blot was performed as loading control (antibody anti-β-actin from Sigma, ref A3853).