Fas-Associated Death Domain-Containing -Mediated Antiviral Innate Immune Signaling Involves the Regulation of Irf7

This information is current as Siddharth Balachandran, Thiagarajan Venkataraman, Paul B. of September 24, 2021. Fisher and Glen N. Barber J Immunol 2007; 178:2429-2439; ; doi: 10.4049/jimmunol.178.4.2429 http://www.jimmunol.org/content/178/4/2429 Downloaded from

References This article cites 55 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/178/4/2429.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 24, 2021 *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 © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Fas-Associated Death Domain-Containing Protein-Mediated Antiviral Innate Immune Signaling Involves the Regulation of Irf71

Siddharth Balachandran,2* Thiagarajan Venkataraman,2* Paul B. Fisher,† and Glen N. Barber3*

The induction of type I (␣␤) IFN following virus infection is necessary for the stimulation of effective antiviral host defense. In fibroblasts, a subset of primary (including those encoding IFN-␤ and IFN-␣4) are induced directly by intracellular dsRNA generated by the virus during its replication. These primary type I IFNs induce expression of IFN regulatory factor (IRF)-7, required for production of a second cascade of IFN-␣ subtypes and the further establishment of a complete antiviral state.

Previously, we had reported on a role for Fas-associated death domain-containing protein (FADD) in the control of TLR-inde- Downloaded from pendent innate immune responses to virus infection. Our data in this study demonstrate that FADD is not only required for efficient primary induction, but is also essential for induction of Irf7 and effective expression of secondary IFN-␣s and other antiviral genes. Ectopic overexpression of IRF-7 partially rescued dsRNA responsiveness and IFN-␣ production, and a constitu- tively active variant of IRF-7 displayed normal activity in Fadd؊/؊ murine embryonic fibroblasts. MC159, a FADD-interacting viral protein encoded by the molluscum contagiosum poxvirus was found to inhibit dsRNA-activated signaling events upstream of IRF-7. These data indicate that FADD’s antiviral activity involves regulation of IRF-7-dependent production of IFN-␣ subtypes http://www.jimmunol.org/ and consequent induction of secondary antiviral genes. The Journal of Immunology, 2007, 178: 2429–2439.

he recognition of viral infection and subsequent triggering dsRNA-activated protein kinase R (PKR),4 which, following in- of antiviral innate immune responses has become the sub- teraction with viral dsRNA, predominantly function to inhibit pro- T ject of intense research over the past few years. Particular tein synthesis (8, 9). We have also recently demonstrated the attention has recently focused on the role of the TLRs, which have existence of a TLR-independent cytosolic dsRNA recognition emerged as key transmembrane responsible for recogniz- pathway involving Fas-associated death domain-containing pro- ing conserved components of pathogenic microorganisms (referred tein (FADD) that may be related to the immune deficiency (IMD) to as pathogen-associated molecular patterns) (1–3). In mamma- innate immune pathway in Drosophila (10, 11). In addition, other by guest on September 24, 2021 lian cells, there appear to be at least 10 TLR family members, each molecules critical for the detection of virally produced cytosolic of which respond to different pathogen-associated molecular pat- dsRNA include retinoic acid-inducible gene I (RIG-I) and mela- terns such as extracellular viral dsRNA (TLR3), ssRNA (TLR7), noma differentiation associated gene (MDA)-5 (5, 12–14). RIG-I is and unmethylated CpG DNA (TLR9) (1). Stimulation of these a DExD/H box RNA helicase with an amino terminal caspase re- TLRs by their cognate ligands culminates in the transcriptional cruitment domain (CARD) and significant homology to another activation of numerous genes, such as the type I IFNs, that directly RNA helicase MDA-5 (15, 16). Although RIG-I appears to activate induce antiviral responses as well as facilitate innate and adaptive type I IFN through tank binding kinase (TBK)1/I␬B kinase (IKK)␧, immune responses (1, 4–7). little is known regarding the intermediate steps in this pathway. Re- Although TLRs 3, 7, and 8 mediate recognition of extracellular cently, a CARD containing mitochondrial molecule called MAVS/ and/or endosomal viral RNA species, distinct pathways for the VISA/Cardif/IPS-1 (hereafter referred to as IFN-␤ promoter stimula- recognition of cytosolic dsRNA also exist. Such mechanisms in- tor (IPS)-1) with homology to the CARDs of MDA-5 and RIG-I has clude those initiated by 2Ј,5Ј-oligoadenylate synthase and the been shown to transduce signals generated following activation of these helicases (17–20). Although the mechanisms by which IPS-1 induces type I IFN remain to be clarified, data indicate that this mol- ecule may activate IFN-␤ via IFN regulatory factor (IRF)-3, NF-␬B, *Department of Microbiology and Immunology and Sylvester Comprehensive Cancer and TNFR-associated factor (TRAF)3 (17–21). † Center, University of Miami School of Medicine, Miami, FL 33136; and Department In this study, we confirm that FADD is required for optimal of Pathology, Columbia University College of Physicians and Surgeons, New York, NY 10032 antiviral signaling and RIG-I, MDA-5, and IPS-1 activity (10, 17, Received for publication August 30, 2006. Accepted for publication November 22, 23). Importantly, our results implicate FADD in the production 28, 2006.

The costs of publication of this article were defrayed in part by the payment of page 4 Abbreviations used in this paper: PKR, protein kinase R; FADD, Fas-associated charges. This article must therefore be hereby marked advertisement in accordance death domain-containing protein; IMD, immune deficiency; RIG-I, retinoic acid-in- with 18 U.S.C. Section 1734 solely to indicate this fact. ducible gene I; MDA, melanoma differentiation Ag; CARD, caspase recruitment do- 1 This work was supported by Defense Advanced Research Projects Agency and main; TBK, tank binding kinase; IKK, I␬B kinase; IPS, IFN-␤ promoter stimulator; National Institutes of Health Grant R01GM068448 (to P.B.F.). IRF, IFN regulatory factor; TRAF, TNFR-associated factor; MEF, murine embryonic ␤ 2 fibroblast; TRIF, Toll/IL-1R domain-containing adaptor-inducing IFN- ; SA, super- S.B. and T.V. contributed equally to this work. active; IRAK, IL-1R-associated kinase; VSV, vesicular stomatitis virus; SeV, Sendai 3 Address correspondence and reprint requests to Dr. Glen N. Barber, Room 511, virus; m.o.i., multiplicity of infection; HA, hemagglutinin. Papanicolaou Building, 1550 Northwest Tenth Avenue, University of Miami School of Medicine, Miami, FL 33136. E-mail address: [email protected] Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 www.jimmunol.org 2430 REGULATION OF Irf7 BY FADD of secondary IFN-␣ genes following dsRNA stimulation or virus the nuclear pellet was suspended in 75 ␮l of buffer B (20 mM HEPES (pH infection. In an effort to identify a mechanistic basis for this ob- 7.9), 400 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 1 mM DTT, 5% servation, we noticed that FaddϪ/Ϫ murine embryonic fibroblasts glycerol, and protease inhibitor mixture) and incubated for 30 min at 4°C with brief mixing. The mixture was centrifuged at 10,000 ϫ g for 5 min at (MEFs) are markedly impaired in the induction of Irf7 by virus 4°C. Protein concentrations were measured using the Bradford assay Ϫ/Ϫ infection or dsRNA stimulation. Indeed, reconstitution of Fadd (Bio-Rad). MEFs with IRF-7 was able to significantly rescue the IFN-␣ in- NF-␬B consensus oligonucleotides (Santa Cruz Biotechnology) were duction defect in these cells, and a constitutively active version of end-labeled with ATP using T4 polynucleotide kinase. Before the addition of oligonucleotide probe, 10 ␮g of nuclear protein was incubated with IRF-7 (but not RIG-I, MDA-5, or IPS-1) could activate a dsRNA- ␮ Ϫ/Ϫ binding buffer (2 g poly(dI-dC), 12% glycerol, 20 mM HEPES (pH 7.0), responsive promoter in Fadd MEFs. Finally, we show that 1 mM DTT, 1 mM EDTA, and 50 mM NaCl) for 10 min at room tem- MC159, a FADD-interacting viral protein encoded by the poxvirus perature. Radiolabeled oligonucleotide was incubated with reaction mix- molluscum contagiosum, can efficiently block RIG-I, MDA-5, and ture for 15 min at room temperature and subjected to 5% nondenaturing ϫ IPS-1-mediated activation of a dsRNA-responsive promoter up- PAGE in 0.5 Tris-borate-EDTA buffer. The gels were dried and analyzed by autoradiography. Supershift was performed by adding Abs to the incubation stream of IRF-7. Thus, FADD is required for robust RIG-I, mixture of nuclear extract and incubating for 15 min on ice before the addition MDA-5, and IPS-1-mediated dsRNA-signaling, an event involving of radiolabeled probe. the control of Irf7. Results Materials and Methods Requirement for FADD in RIG-I, MDA-5, and IPS-1-mediated Cells and reagents induction of IFN-␤ MEFs were obtained from the following sources: Faddϩ/Ϫ and FaddϪ/Ϫ, We and others have recently demonstrated the existence of a TLR- Downloaded from Tbk1ϩ/ϩ and Tbk1Ϫ/Ϫ, Traf2ϩ/ϩ, and Traf2Ϫ/Ϫ, Traf6ϩ/ϩ and Traf6Ϫ/Ϫ independent cytosolic dsRNA recognition pathway involving ϩ ϩ (provided by W.-C. Yeh, University of Toronto, Toronto, Canada); Trif / FADD (10, 17, 22, 23). It has been also been shown that the RNA and TrifϪ/Ϫ (provided by S. Akira, Osaka University, Osaka, Japan); and ϩ/ϩ Ϫ/Ϫ helicases RIG-I and MDA-5 associate with intracellular dsRNA to Stat1 and Stat1 MEFs (provided by J. Durbin, Columbus Children’s ␤ Research Institute, Columbus, OH). All other cell lines were obtained from activate IFN- (16, 23), whereas a CARD-less RIG-I/MDA-5 ho- the American Type Culture Collection. Poly(I:C) (Amersham) was recon- mologue, LGP2, may act as a dominant-negative inhibitor of stituted in PBS at 2 mg/ml, denatured at 55°C for 30 min, and allowed to RIG-I and MDA-5 signaling (23, 24). Recent data further show a http://www.jimmunol.org/ anneal to room temperature before use. MEFs were transfected with 6 ␮g role for IPS-1 in TBK1/IKK␧-mediated activation of IFN-␤ by of poly(I:C) in 8 ␮l of Lipofectamine 2000 per milliliter of medium to stimulate promoter activation and IFN production. Murine IFN-␣ and -␤ RIG-I and MDA-5 (25, 26). Given that FADD also governs ELISA Kits were acquired from PBL. NF-␬B p65 EZ-Detect transcription dsRNA-triggered activation of IFN-␤, we investigated the role of factor ELISA Kits were purchased from Pierce. Anti-hemagglutinin (HA), FADD in RIG-I/MDA-5/IPS-1-mediated dsRNA signaling. To anti-p50, and anti-p65 Abs were purchased from Santa Cruz Biotechnology. this end, Faddϩ/Ϫ or FaddϪ/Ϫ MEFs were transfected in the pres- All other reagents were from Sigma-Aldrich, unless mentioned otherwise. ence or absence of dsRNA with expression vectors encoding Plasmids RIG-I, MDA-5, and their respective CARD-containing modules, LGP2, IPS-1, and TRIF. As expected, we found that overexpressed

Expression vectors (pcDNA3-Neo; Invitrogen Life Technologies) encod- by guest on September 24, 2021 RIG-I and MDA-5 robustly induced the IFN-␤ promoter in the pres- ing FLAG-tagged versions of human RIG-I, MDA-5, and LGP2, FADD, ϩ Ϫ IPS-1, TBK-1, Toll/IL-1R domain-containing adaptor-inducing IFN-␤ ence of dsRNA (Ͼ100-fold) in Fadd / MEFs (Fig. 1). The CARDs (TRIF), and their various mutants were generated by PCR. Other plasmids of RIG-I and MDA-5 alone was found to induce the IFN-␤ promoter ␤ were obtained as follows: IFN- -Luc (provided by J. Hiscott, McGill Uni- Ͼ100- to 300-fold compared with controls, even in the absence of versity, Montreal, Canada), PRD-III-I-Luc (provided by T. Maniatis, Har- vard University, Cambridge, MA), RL-1-Luc (provided by P. Pitha-Rowe, dsRNA. In contrast, overexpression of LGP2 inhibited activation Johns Hopkins University, Baltimore, MD), NF-␬B-Luc (Stratagene), ex- of the IFN-␤ promoter in response to dsRNA (Fig. 1). When RIG-I pression vectors for MC159 (provided by J. Shisler, University of Illinois, or MDA-5 or their corresponding CARDs alone were transfected Urbana, IL), TLR3, IL-1R-associated kinase (IRAK)1, TRAF6, IRF-7, and into FaddϪ/Ϫ MEFs, there was a significant reduction in activation IRF-7 (superactive; SA) (InvivoGen). of the IFN-␤ promoter following dsRNA stimulation (Fig. 1). Im- DNA microarray analysis portantly, these studies also demonstrated a clear role for FADD in optimal IPS-1 signaling. Whereas IPS-1 overexpression induced Total RNA was extracted from MEFs stimulated with or without poly(I:C) ␤ Ͼ ϩ/Ϫ (6 ␮g/ml in Lipofectamine 2000) or murine IFN-␤ (200 U/ml) at the in- the IFN- promoter 200-fold in Fadd MEFs, it caused only Ϫ Ϫ dicated time points. Preparation of cDNA and microarray analysis was a ϳ10-fold activation of this promoter in Fadd / MEFs (Fig. performed at the W.M. Keck Foundation Biotechnology Research Labo- 1A). TRIF signaling was unimpaired in FADD-deficient cells, indi- ratory DNA microarray facility at Yale University (New Haven, CT). The cating that, at least in MEFs, FADD is predominantly involved in Mouse Genome 430 2.0 Array (Affymetrix) was used. Data analysis was performed with Microarray Suite software (version 5.0; Affymetrix) and TLR-independent signaling, similar to its role in Drosophila innate GeneSpring software (Silicon Genetics). immune responses (27, 28) (Fig. 1). Real-time PCR Defective induction of primary antiviral genes in the absence of FADD Total RNA was isolated from cells using the RNeasy RNA extraction (Qiagen), and cDNA synthesis was performed using 1 ␮g of total RNA To confirm a role for FADD in dsRNA- and virus-triggered pro- (Roche). Fluorescence real-time PCR analysis was performed using a duction of type I IFN, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs were trans- LightCycler 2.0 instrument (Roche Molecular Biochemicals) and TaqMan Assays (Applied Biosystems). Relative amounts of fected with poly(I:C), or infected with vesicular stomatitis virus mRNA were normalized to the 18S ribosomal RNA levels in each sample. (VSV) or Sendai virus (SeV) at multiplicities of infection (m.o.i.) of 1 or 10, and examined for type I IFN protein production at 6, 12, EMSA or 24 h posttreatment. As before, and in agreement with previous Cells (5 ϫ 106) washed two times with cold PBS were suspended in 400 results (10, 22, 23), we found that FaddϪ/Ϫ MEFs exhibited a ␮l of hypotonic buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 1.5 mM significant defect in the production of IFN-␤ following poly(I:C) MgCl2, 1 mM DTT, 0.5 mM PMSF, and protease inhibitor mixture). After ␮ transfection or VSV infection (Fig. 2A, left and middle panels). 15 min of incubation on ice, 25 l of 10% Nonidet P-40 was added, and Ϫ/Ϫ the sample was vigorously vortexed. The nuclei were collected by centrif- Indeed, Fadd MEFs produced only 30–50% the amount of ϩ Ϫ ugation at 3000 ϫ g for 5 min at 4°C and washed once with buffer A, and IFN-␤ of similarly treated Fadd / MEFs. When these cells were The Journal of Immunology 2431

FIGURE 1. FADD is required for optimal RIG-I, MDA-5, and IPS-1 sig- naling. The indicated expression vec- tors, together with the IFN-␤-Luc fire- fly luciferase reporter plasmid were transfected into Faddϩ/Ϫ and FaddϪ/Ϫ MEFs. Twenty-four hours later, these cells were either left untreated, or were transfected with poly(I:C) for a further 6 h, and IFN-␤ promoter-driven firefly luciferase activity was measured at that time. Values are normalized to an inter- nal Renilla luciferase control. Error bars, ϮSD. Downloaded from infected with SeV, however, FaddϪ/Ϫ MEFs produced approxi- supernatants for IFN-␤ by ELISA. Overexpressed RIG-I or mately normal levels of IFN-␤, albeit with delayed kinetics, com- MDA-5 were unable to stimulate effective IFN-␤ production in the pared with controls (Fig. 2A, right panel). To determine whether absence of FADD, confirming a pivotal role for FADD down- ectopic overexpression of RIG-I and MDA-5 was able to over- stream of these helicases in the induction of IFN-␤ (Fig. 2B). To come the IFN-␤ production defects seen following transfection determine whether defective induction of IFN-␤ extended to other with poly(I:C), we transfected Faddϩ/ϩ and FaddϪ/Ϫ MEFs with primary antiviral genes, we compared the gene induction profile of expression vectors encoding full-length RIG-I and MDA-5, and Stat1-independent (i.e., “primary”) genes in FaddϪ/Ϫ MEFs to http://www.jimmunol.org/ stimulated such cells with poly(I:C) for 24 h before examining their heterozygous counterparts. For the purpose of this analysis, by guest on September 24, 2021

FIGURE 2. FADD participates in the regulation of a subset of primary dsRNA-responsive genes. A, ELISA was performed to detect levels of IFN-␤ ,Not detectable. B ,ء .in the supernatants of Faddϩ/Ϫ and FaddϪ/Ϫ MEFs treated with poly(I:C), VSV, or SeV at the m.o.i. and times indicated. Error bars, ϮSD Expression vectors encoding RIG-I or MDA-5 were transfected into Faddϩ/Ϫ and FaddϪ/Ϫ MEFs. Twenty-four hours posttransfection, these cells were treated with poly(I:C) for a further 24 h, and levels of IFN-␤ were measured in the supernatants at that time. Error bars, ϮSD. C, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs in tandem with Tbk1ϩ/ϩ, Tbk1Ϫ/Ϫ, Stat1ϩ/ϩ, and Stat1Ϫ/Ϫ MEFs, were either left untreated or were transfected with poly(I:C) for 3 h, and DNA microarray analysis of dsRNA-induced primary gene expression was performed as described in Materials and Methods and Results. 2432 REGULATION OF Irf7 BY FADD Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 3. Global defects in secondary gene induction by dsRNA in the absence of FADD. A, ELISA was performed to detect levels of IFN-␣ in the Not detectable. B, Faddϩ/Ϫ ,ء .supernatants of Faddϩ/Ϫ and FaddϪ/Ϫ MEFs treated with poly(I:C), VSV, or SeV at the m.o.i. and times indicated. Error bars, ϮSD and FaddϪ/Ϫ MEFs together with Tbk1ϩ/ϩ, Tbk1Ϫ/Ϫ, Stat1ϩ/ϩ, and Stat1Ϫ/Ϫ MEFs, were either left untreated or were transfected with poly(I:C) for 12 h, and DNA microarray analysis of dsRNA-induced secondary gene expression was performed as described in Materials and Methods and Results. all genes that were induced by dsRNA treatment at least 2-fold in and FaddϪ/Ϫ MEFs following dsRNA stimulation or virus infec- ϩ ϩ Stat1 / MEFs and were also induced at least 2-fold or more in tion. Accordingly, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs were either trans- Ϫ Ϫ the Stat1 / MEFs were classified as primary dsRNA-responsive fected with poly(I:C) or infected with VSV or SeV at different antiviral genes. After ensuring that these subsets were also induced m.o.i. (1 and 10), and supernatants from these cells were examined ϩ/ϩ ϩ/Ϫ at least 2-fold in the Tbk1 and Fadd MEFs, their behavior for secreted IFN-␣ production. As we had reported previously, Ϫ/Ϫ Ϫ/Ϫ was examined in Tbk1 and Fadd MEFs. This analysis re- FaddϪ/Ϫ MEFs produced profoundly reduced levels of IFN-␣ vealed that, whereas the majority of primary genes were induced in compared with Faddϩ/Ϫ MEFs following dsRNA treatment (Fig. FaddϪ/Ϫ MEFs to the same or greater extent than they were in ϩ/Ϫ 3A, left panel) (10). More importantly, we noticed that virus in- Fadd MEFs, the induction of a subset of primary genes (such Ϫ Ϫ fection of Fadd / MEFs also resulted in significantly reduced as Ifit1, Ifit2, and Icam1) was found to be significantly impaired in IFN-␣ production. In fact, IFN-␣ was undetectable in the super- the absence of FADD (Fig. 2C). Thus, FADD appears to be re- Ϫ/Ϫ quired for the optimal induction of some, but not all, STAT1- natant of VSV-infected Fadd MEFs at all m.o.i.s and time independent genes. points tested (Fig. 3A, middle panel). Notably, and unlike the case with IFN-␤, IFN-␣ production was also significantly defective (re- Global defects in secondary dsRNA-responsive gene induction duced by ϳ80%, compared with controls) in FaddϪ/Ϫ MEFs after in the absence of FADD SeV infection (Fig. 3A, right panel). Along with our previous ob- To determine whether defective IFN-␤ secretion in the absence of servation that IFN-␣ production is important for protection against FADD resulted in compromised induction of secondary (IFN-␤- virus infection (10), these results provide an explanation for the Ϫ Ϫ dependent) gene induction, including that of the other type I IFNs remarkable susceptibility of early passage Fadd / MEFs to viral (IFN-␣s), we next examined the production of IFN-␣ in Faddϩ/Ϫ infection. The Journal of Immunology 2433

FIGURE 4. Defective Irf7 induction by dsRNA in FaddϪ/Ϫ MEFs. A, Tbk1ϩ/ϩ, Tbk1Ϫ/Ϫ, Faddϩ/Ϫ, and FaddϪ/Ϫ MEFs were either left untreated or were trans- fected with poly(I:C) for 3 h, and DNA microarray anal- ysis of Irf gene expression was performed as described in Materials and Methods. B, RNA extracted from Faddϩ/Ϫ and FaddϪ/Ϫ MEFs treated with or without poly(I:C) for the indicated times was examined from Irf7 induction by real-time PCR. C, Faddϩ/Ϫ, FaddϪ/Ϫ, Stat1ϩ/ϩ, Stat1Ϫ/Ϫ, Tbk1ϩ/ϩ, and Tbk1Ϫ/Ϫ MEFs were either left untreated or were transfected with poly(I:C) for 12 h, and DNA microarray analysis of Irf7 gene expression was performed as described in Materials and Methods. D, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs were either left untreated or were treated with murine IFN-␤ (500 U/ml) for 8 h, and DNA microarray analysis of Irf gene expression was performed as described in Materials and Downloaded from Methods.

Given that type I IFN induction is severely impaired in the ab- following dsRNA or IFN stimulation, and were therefore not eval- http://www.jimmunol.org/ sence of FADD, we next directly addressed the expression profile uated further (data not shown). Irf3 was present in equivalent lev- of secondary (type I IFN/STAT1-dependent) antiviral genes in els in both Faddϩ/Ϫ and FaddϪ/Ϫ MEFs and was not inducible by Faddϩ/Ϫ and FaddϪ/Ϫ MEFs. For this experiment, all genes that dsRNA treatment (Fig. 4A). Following dsRNA stimulation, we were induced at least 2-fold in Stat1ϩ/ϩ MEFs, but were not in- noticed that Irf1, Irf2, and Irf9 were induced normally in FaddϪ/Ϫ duced (i.e., 1-fold or lower) in Stat1Ϫ/Ϫ MEFs, were classified as MEFs. Importantly, however, we observed a significant and secondary dsRNA-responsive genes (Fig. 2B). The induction pro- marked impairment in the induction of Irf7 (and, to a lesser extent, files of such genes were then examined in Tbk1Ϫ/Ϫ and FaddϪ/Ϫ Irf5) by dsRNA in FaddϪ/Ϫ MEFs 3 h after dsRNA stimulation MEFs, after ensuring that they were induced at least 2-fold in (Fig. 4A). Both IRF-5 and IRF-7 are virus-activated transcription ϩ ϩ ϩ Ϫ Tbk1 / and Fadd / MEFs. This study revealed a significant, factors implicated in antiviral host defense. However, IRF-7 has by guest on September 24, 2021 global impairment of the induction of a spectrum of secondary recently emerged as the master regulator of type I IFN-production responsive genes in the absence of FADD (Fig. 2A). For example, following virus infection, whereas IRF-5 appears to participate pri- genes such as Casp7, Tlr3, eif2ak2 (PKR), Rnasel, Ddx58 (RIG-I), marily in inflammatory production (32, 33). Furthermore, and Ube11 were poorly activated in FaddϪ/Ϫ MEFs (Fig. 3B). induction by dsRNA of Irf7 was more severely compromised than This analysis also confirmed a severe impairment in the induction Irf5 in FaddϪ/Ϫ MEFs. For these reasons, we examined the reg- of the genes encoding secondary IFNs (such as Ifnab, Ifna2, Ifna5, ulation of Irf7 by FADD in greater detail. Defective induction of Ifna6, Ifna7, and Ifna9) 12 h after dsRNA treatment of FaddϪ/Ϫ Irf7 by dsRNA in the absence of FADD was confirmed by real- MEFs. Interestingly, whereas inducibility of the majority of sec- time PCR analysis and persisted for at least 12 h poststimulation ondary dsRNA-responsive genes was compromised in FaddϪ/Ϫ (Fig. 4, B and C). It is noteworthy that induction of Irf7 (or any of MEFs, a subset of such genes (including Myd88 and B2m) were the other inducible IRFs) by type I IFN treatment per se is not induced normally in these cells (Fig. 3B). These results indicate defective in the absence of FADD, indicating that the FADD is that, whereas FADD appears essential for robust secondary gene required for the activation of these genes at a step before type I induction following dsRNA treatment, FADD-independent mech- IFN-triggered Jak/STAT signaling (Fig. 4D). In support of this anisms for the induction of these also likely exist. Nevertheless, fact, DNA microarray analysis of Irf7 in Stat1ϩ/ϩ and Stat1Ϫ/Ϫ taken together with the fact that RIG-I and MDA-5-dependent MEFs showed that Irf7 cannot be induced by dsRNA in the IFN-␣ and IFN-␤ production is significantly compromised in the absence of STAT1 and, by extension, type I IFN signaling (Fig. absence of FADD (Fig. 3B), our findings suggest an indispensable 4C). Taken together, these data indicate that the poor inducibility role for FADD in autocrine production of type I IFN by cytosolic of Irf7 seen in FaddϪ/Ϫ MEFs is not the result of defective IFN dsRNA. Jak/STAT signaling but rather suggests a role for FADD in the direct dsRNA-stimulated production of IRF-7 itself. Control of Irf7 by FADD Secondary IFN-␣ gene expression requires members of the IRF Ectopic overexpression of IRF-7 partially rescues IFN family of transcription factors for their expression (29–31). During Ϫ/Ϫ our earlier analyses, we noticed that Irf7 was among those genes production defects in FADD MEFs whose induction by dsRNA was defective in the absence of FADD Next, we performed experiments to examine whether reconstitu- (10). We therefore further evaluated IRF expression levels in tion of FaddϪ/Ϫ MEFs with wild-type IRF-7 can rescue defective Faddϩ/Ϫ and FaddϪ/Ϫ MEFs before and after dsRNA treatment. dsRNA signaling seen in these cells. To this end, Faddϩ/Ϫ and Irf1, Irf2, Irf3, Irf5, Irf7, and Irf9 were present in detectable levels FaddϪ/Ϫ MEFs were transfected with expression vectors encoding in unstimulated wild-type cells (Fig. 4A). In contrast, expression of either wild-type IRF-7 or a constitutively SA mutant of IRF-7 Irf4, Irf6, and Irf8 were undetectable in fibroblasts either basally or (IRF-7 (SA)) together with the IFN-␤-Luc reporter plasmid. IRF-7 2434 REGULATION OF Irf7 BY FADD

FIGURE 5. Ectopic overexpression of IRF-7 in FADDϪ/Ϫ MEFs partially rescues dsRNA signaling defects. A, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs were transfected with either wild-type IRF-7 or a constitutively active version of IRF-7 (IRF-7 (SA)), together with the IFN-␤-Luc firefly reporter plasmid for 24 h, and subsequently treated with or without poly(I:C) for a further 6 h. IFN-␤ promoter-driven firefly luciferase activity was then measured and ϩ/Ϫ normalized to an internal Renilla luciferase control. Error bars, ϮSD. B, An expression vector encoding wild-type IRF-7 was transfected into Fadd and Downloaded from FaddϪ/Ϫ MEFs. Twenty-four hours posttransfection, these cells were treated with poly(I:C) for a further 24 h, and levels of IFN-␤ or (C) IFN-␣ were .p Ͻ 0.0001 ,ءءء ;p Ͻ 0.001 ,ءء .measured in the supernatant by ELISA. Error bars, ϮSD

(SA) differs from wild-type human IRF-7 in that it contains a de- tivity 6 h poststimulation. Faddϩ/Ϫ MEFs showed robust activa- letion of a negative regulatory region (aa 238–410) that allows it tion of the IFN-␤ promoter after dsRNA stimulation, whereas to activate IFN-␤ without a requirement for prior phosphorylation FaddϪ/Ϫ MEFs showed almost no responsiveness to the same http://www.jimmunol.org/ by TBK1/IKK␧ (34). Twenty-four hours posttransfection, these stimulus (ϳ25-fold activation in Faddϩ/Ϫ MEFs, compared with cells were stimulated with dsRNA and examined for reporter ac- ϳ1.5-fold in FaddϪ/Ϫ MEFs; Fig. 5A). Remarkably, however, by guest on September 24, 2021

FIGURE 6. Partial requirement for FADD in activation of NF-␬B by dsRNA. A, Expression vectors encoding FADD, IPS-1, or TRIF were transfected into 293T cells, together with the indicated firefly luciferase reporter plasmids. Forty-eight hours posttransfection, firefly luciferase activity was measured, and normalized to an internal Renilla luciferase control. Error bars, ϮSD. B, Nuclear extracts prepared from Faddϩ/Ϫ and FaddϪ/Ϫ MEFs stimulated for the indicated times with transfected dsRNA were analyzed by EMSA for NF-␬B activity as described in Materials and Methods. C, Nuclear extracts prepared from Faddϩ/Ϫ and FaddϪ/Ϫ MEFs stimulated for the indicated times with transfected dsRNA were analyzed by p65-specific ELISA for NF-␬B activity per the manufacturer’s instructions (EZ-Detectp65 Kit; Pierce). Competition with mutant and wild-type NF-␬B consensus sequences shows specificity. TNF-␣-treated HeLa nuclear lysate was used as a positive control. The x-ray exposure of duplicate samples is shown. D,An expression vector encoding FADD was transfected into 293T cells, together with a firefly luciferase reporter plasmid driven by the full-length human Irf7 promoter (RL-1-Luc). Firefly luciferase activity was measured 48 h posttransfection and normalized to an internal Renilla luciferase control. Error bars, ϮSD. E, Wild-type MEFs were transfected with an expression vector encoding FADD. Forty-eight hours posttransfection, RNA was extracted from these cells and examined for endogenous Irf7 gene induction by real-time PCR. The Journal of Immunology 2435 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021 FIGURE 7. Role of TRAF2, TRAF6, and TRIF in RIG-I/MDA-5-dependent innate immune antiviral responses. A, Traf2ϩ/ϩ (wild-type; WT), Traf2Ϫ/Ϫ, Traf6Ϫ/Ϫ,orTrifϪ/Ϫ MEFs were infected with VSV (m.o.i. ϭ 10), with and without pretreatment with IFN-␣␤ (100 U/ml) or IFN-␥ (5 ng/ml). Photomi- crographs were taken 36 h postinfection. B, VSV progeny yield was determined from supernatants of cells treated as in A by standard plaque assay. C, IFN-␤-Luc was transfected into TRAF2, TRAF6, and TRIFϩ/ϩ and Ϫ/Ϫ MEFs. After 24 h, they were left unstimulated, or stimulated with poly(I:C) with and without transfection reagent (Lipofectamine; LF). Luciferase activity was measured after 6 h. D, Empty vector, TLR3, IRAK1, TRAF6, RIG-I (1–200), MDA-5 (1–220), TRIF, and IPS-1 were expressed along with the IFN-␤-Luc construct in Traf2ϩ/ϩ and Traf2Ϫ/Ϫ MEFs. Luciferase values were measured after 24 h.

FaddϪ/Ϫ MEFs transfected with wild-type IRF-7 were now able to Impaired activation of NF-␬B in the absence of FADD ϳ respond to dsRNA, albeit to a lesser extent ( 50%) than the lit- In Drosophila, FADD participates in an innate immune signaling ϩ/Ϫ termate control-derived Fadd fibroblasts. In agreement with cascade that culminates in NF-␬B-mediated activation of antimi- ␤ this finding, IRF-7 (SA) was able to activate the IFN- promoter crobial genes (11). Because mammalian FADD is also capable of ϳ equally well in the presence or absence of FADD ( 200-fold in activating NF-␬B (17, 35), and because several important antiviral both cases; Fig. 5A). We next complemented these studies by di- genes (including IFN-␤ and Irf7) have functional NF-␬B binding rectly examining whether IRF-7 overexpression can boost type I sites, we examined the role of FADD in activation of NF-␬B fol- Ϫ/Ϫ IFN production in Fadd MEFs following dsRNA stimulation. lowing dsRNA stimulation (36, 37). We initially overexpressed ϩ/Ϫ Ϫ/Ϫ Accordingly, Fadd and Fadd were transfected with an ex- FADD (in tandem with IPS-1 and TRIF as controls) together with pression vector encoding IRF-7 and subsequently stimulated with either an IFN-␤ promoter reporter plasmid (IFN-␤-Luc), or re- dsRNA for 24 h. Supernatants from cells treated in this manner porter plasmids responsive to either IRF3/7 (PRD III-I-Luc) or were examined by ELISA for either IFN-␣ or IFN-␤. Transfection NF-␬B (NF-␬B-Luc). We found that FADD was clearly capable of Ϫ Ϫ of a plasmid encoding IRF-7 increased the ability of Fadd / robust NF-␬B activation (150-fold), in agreement with previous MEFs to produce IFN-␣ by ϳ4-fold (from ϳ200 pg/ml to ϳ800 results (Fig. 6A) (17, 35). However, we also observed that FADD pg/ml after dsRNA treatment; p Ͼ 0.0001; Fig. 5B). IFN-␤ could modestly stimulate IFN-␤-Luc and the IRF3/7 responsive production was also increased by ϳ2-fold, from ϳ2500 pg/ml PRD III-I promoter (ϳ8-fold and ϳ5-fold, respectively; Fig. 6A). to ϳ5000 pg/ml after IRF-7 overexpression (Fig. 5C; p Ͼ Because FADD can robustly activate NF-␬B, we next directly 0.001). These results demonstrate that IRF-7 can at least par- tested whether dsRNA-induced NF-␬B activation was defective in tially rescue the defects in induction of type I IFN seen in FaddϪ/Ϫ MEFs. Accordingly, Faddϩ/Ϫ and FaddϪ/Ϫ MEFs were FaddϪ/Ϫ MEFs. transfected with dsRNA, and nuclear extracts prepared from these 2436 REGULATION OF Irf7 BY FADD

FIGURE 8. Inhibition of cytosolic dsRNA signaling by the poxviral protein MC159. A, HeLa cells were transfected with an expression vector encoding MC159 together with the IFN-␤-Luc firefly reporter plasmid. Twenty-four hours posttransfection, such cells were further transfected with poly(I:C), or were left untreated. Firefly luciferase activity was measured 6 h posttreatment and normalized to an internal Renilla control. Error bars, ϮSD. B, Empty vector (lane 1) or HA-tagged MC159 (250 ng-lane 2; 500 ng-lane 3) was transfected into 293T cells and expression was confirmed by Western blot analysis using an anti-HA Ab. ␤-actin (lower panel) was used as a control to ensure equal loading. C, 293T cells were transfected with the indicated expression vectors (500 ng/sample), along with increasing amounts of the expression vector encoding MC159 (250 ng/sample, 500 ng/sample) and the IFN-␤-Luc firefly luciferase reporter plasmid. Total DNA concentrations were kept constant by adjustment with a control vector. Firefly luciferase activity was measured 48 h posttransfection and normalized to an internal Renilla control. Error bars, ϮSD. D, 293T cells were transfected with an expression vector encoding FADD (500 ng/sample), together with a firefly luciferase reporter plasmid driven by a NF-␬B-responsive promoter, and increasing amounts of an expression vector Downloaded from encoding MC159 (250 ng/sample, 500 ng/sample). Firefly luciferase activity was measured 48 h posttransfection, and normalized to an internal Renilla luciferase control.

cells were examined by EMSA for binding activity to a radiola- however, neither Traf2Ϫ/Ϫ, Traf6Ϫ/Ϫ,orTrifϪ/Ϫ MEFs appeared ␬ beled NF- B consensus sequence. As shown in Fig. 6B, both basal to be overtly sensitive to either VSV-induced cytopathic effect http://www.jimmunol.org/ and dsRNA-induced NF-␬B DNA-binding activity was found to (Fig. 7A) or replication (Fig. 7B) compared with control cells, in be partially defective in early passage FaddϪ/Ϫ MEFs compared the presence or absence of IFN pretreatment. Indeed, intracellular with Faddϩ/Ϫ MEFs. Supershift Ab analyses indicated that the dsRNA-mediated activation of the IFN-␤ promoter was found to NF-␬B p65 subunit was predominantly involved in DNA complex be normal in cells deficient in either Traf6, Traf2,orTrif (Fig. 7C). formation in response to transfected dsRNA in MEFs (Fig. 6B). Next, Traf2ϩ/ϩ and Traf2Ϫ/Ϫ MEFs were transiently transfected We therefore next examined direct p65 binding to its cognate con- with expression plasmids encoding TLR3, IRAK1, TRAF6, RIG-I sensus DNA sequence by ELISA. Nuclear extracts prepared as (1–200), MDA-5 (1–220), TRIF or MAVS/IPS-1/VISA/Cardif, to- described above were analyzed for p65 DNA-binding activity us- gether with the IFN-␤ luciferase reporter construct. Measurement ing a p65 transcription factor ELISA kit. As shown in Fig. 6C,we of luciferase activity 24 h posttransfection revealed no significant by guest on September 24, 2021 again observed that FaddϪ/Ϫ cells possessed significantly less differences in activation of the luciferase promoter by any of these basal as well as dsRNA-inducible p65-dependent NF-␬B activity. polypeptides in the absence of TRAF2, implying a nonessential or These results demonstrate that FADD participates in the activation redundant role for this molecule in dsRNA-triggered signaling of NF-␬B following dsRNA stimulation in fibroblasts. events (Fig. 7D). Thus, FADD, but not TRAF2, TRAF6, or TRIF We next tested whether FADD could directly activate the Irf7 are required for IFN induction by intracellular dsRNA. promoter. To this end, we cotransfected 293T cells with expression vectors encoding FADD along with a luciferase reporter plasmid un- Regulation of dsRNA signaling by the FADD-interacting der control of a 1.6-kb region of the human Irf7 promoter (RL-1-Luc) poxviral FLIP MC159 (36). Luciferase activity measured 24 h posttransfection revealed that A number of viral proteins have been shown to inhibit TLR-inde- FADD can also modestly stimulate the Irf7 promoter (ϳ5-fold; Fig. pendent dsRNA-induced antiviral signaling (for example, see 6D). To extend these findings, we next overexpressed FADD in wild- Refs. 18, 39–42). Viral FLIPs are a family of death effector do- type MEFs and examined endogenous murine Irf7 gene induction by main-containing proteins encoded by gammaherpesviruses and real-time PCR 24 h posttransfection. In agreement with the luciferase poxviruses that have previously been shown to inhibit FADD- and reporter assay results, we found that overexpression of FADD could caspase-8-dependent apoptotic signaling (43, 44). It is noteworthy also activate Irf7 directly (ϳ4-fold) (Fig. 6E). that, of the various viral FLIPs, MC159, encoded by the mollscum contagiosum virus of the poxviridae family interacts specifically Role of TRAF2 in RIG-I/MDA-5-dependent antiviral responses and strongly with FADD (45, 46). We hypothesized that, in addi- Our findings indicate that FADD can potently activate NF-␬B and tion to its role as an inhibitor of FADD-dependent apoptosis, appears to be at least partially required for NF-␬B activation in MC159 could inhibit innate immune signaling by blocking FADD. response to cytoplasmic dsRNA. Reports based on recent data First, we examined whether MC159 inhibits dsRNA-mediated speculated a role for TRAF2 in signaling events upstream of activation or IFN-␤. Hela cells were cotransfected with IFN-␤-Luc FADD, specifically those involving NF-␬B activation or triggered and either an empty vector control or with a vector encoding by IPS-1 (20, 38). To examine whether TRAF2 was indeed re- MC159. Such cells were stimulated with dsRNA 24 h posttrans- quired for TLR3-dependent and -independent signaling, early pas- fection, and luciferase activity was measured 6 h later. As shown sage Traf2ϩ/ϩ and Traf2Ϫ/Ϫ MEFs, in tandem with MEFs lacking in Fig. 8A, we noticed that MC159 expression potently inhibited Traf6,orTrif, were treated with either IFN-␣␤ or IFN-␥ and sub- dsRNA-induced activation of the IFN-␤ promoter. In empty vec- sequently infected with VSV at a m.o.i. of 10. In this assay, early tor-transfected cells, dsRNA stimulation resulted in a Ͼ250-fold passage primary MEFs deficient in Fadd, Ripk1, Tbk1,orIrf3 suc- activation of IFN-␤-Luc (Fig. 8A). In contrast, cells expressing cumb to VSV-induced replication and cytolysis even in the pres- MC159 manifested only ϳ20-fold activation of this reporter (Fig. ence of IFN pretreatment (10). Unlike the case with these cells, 8A). To better elucidate the mechanism by which MC159 inhibits The Journal of Immunology 2437

FIGURE 9. Schematic of cytosolic dsRNA-triggered signaling events. Cytosolic viral dsRNA is detected by PKR and 2Ј,5Ј-oligoadenylate synthase leading to translational shutdown to restrict viral growth, or by the RNA helicases RIG-I and MDA-5, which initiate a signaling cascade leading to the activation of an antiviral program. RIG-I and MDA-5 require IPS-1 to active IFN-␤, and may transduce their signal via complex formation with molecules Downloaded from such as FADD, RIP1, and TRAF3. Stimulation of IPS-1 ostensibly leads to activation of TBK-1 and IKK, consequent nuclear translocation of IRF-3, IRF-7 and NF-␬B homo- and heterodimers, and activation of primary antiviral genes, including IFN-␤. IFN-␤ binds to IFNAR and induces Irf7 through the JAK/STAT pathway leading to a secondary antiviral response, including the induction of the CARD-less RIG-I/MDA-5 homologue, LGP2, which may dampen dsRNA signaling in a negative feedback loop. Viruses have evolved various mechanisms to disrupt cytosolic dsRNA-activated innate immune signaling. For example, the paramyxovirus V proteins inhibit MDA-5, hepatitis C virus NS3/4A inhibits IPS-1, molluscum contagiosum poxviral FLIP MC159 inhibits FADD and TBK-1, and the Borna disease virus P protein blocks TBK-1. http://www.jimmunol.org/

IFN-␤ activation by dsRNA, 293T cells were cotransfected with induction of IFN-␤, although the underlying molecular mecha- the IFN-␤-Luc reporter plasmid, together with expression vectors nisms remain unclear (17–20). One report suggest that IPS-1 can encoding RIG-I (1–284), IPS-1, FADD, and IRF-7 (SA), along interact directly with RIG-I and MDA-5 and may signal via FADD with increasing amounts of an expression vector encoding MC159. to activate the NF-␬B arm of dsRNA signaling (17), whereas an- Expression of MC159 was confirmed by immunoblot analysis other group reports that IPS-1 does not bind RIG-I or MDA-5 to (Fig. 8B). As shown in Fig. 8C, MC159 was able to markedly any significant extent (19). A further study indicates that IPS-1 may inhibit signaling by RIG-I, IPS-1, FADD, and TBK-1 (by ϳ60, bind to RIG-I, but not MDA-5, and can directly associate with mem- by guest on September 24, 2021 ϳ95, ϳ70, and ϳ80%, respectively). Importantly, MC159 had no bers of the IKK family to trigger IFN-␤ induction (18). Nevertheless, inhibitory effect on IRF-7 (SA) (Fig. 8C). Furthermore, MC159 these reports and subsequent in vivo studies do indicate that IPS-1 is also inhibited FADD-mediated NF-␬B activation by ϳ50% (Fig. indeed required for optimal RIG-I and MDA-5 function (13, 14). 8D). These results describe a novel role for the FADD-interacting Our data presented in this study confirm a role for FADD down- protein MC159 in specific inhibition of RIG-I-dependent host an- stream of RIG-I, MDA-5, and IPS-I in activation of IFN-␤ by tiviral responses upstream of IRF-7 and reinforce the importance dsRNA. FADD may affect signaling downstream of IPS-1 by ei- of FADD in innate antiviral signaling. ther direct interaction (17) or by participation in a signaling com- Collectively, our data demonstrate that FADD is required for op- plex initiated by RIG-I/MDA-5 activation. For example, our pre- timal primary innate immune responses triggered by activation of liminary data show that the FADD-interacting protein RIP1, RIG-I, MDA-5, and IPS-1. In addition, secondary type I IFN-depen- homologous to the Imd gene product of Drosophila can interact dent responses are significantly compromised, explaining the severe with the CARDs of RIG-I and MDA-5, and associates with Ϫ Ϫ loss of antiviral activity observed in Fadd / MEFs. Our observa- dsRNA-containing complexes in an IFN-dependent manner (data Ϫ Ϫ tions of defective type I IFN production in Fadd / MEFs likely not shown). Given that FADD interacts strongly with RIP1, and involve an impairment in dsRNA-activated Irf7 transcription. that Ripk1Ϫ/Ϫ MEFs also display defective type I IFN (especially IFN-␣) gene induction following virus infection or dsRNA stim- Discussion ulation (10), these data suggest possible physiologically relevant dsRNA produced during viral replication is considered a primary links between RIG-I, MDA-5, RIP1, and FADD. trigger of antiviral cytokine production and other innate and adap- Although a lack of FADD does not severely block virus-medi- tive immune responses (3). In an emerging picture, extracellular ated activation of IFN-␤, loss of FADD does appear to have a dsRNA appears to signal via TLR3 and the adaptor molecule TRIF detrimental effect on the kinetics of IFN-␤ induction and on the to activate type I IFN production, whereas intracellular dsRNA activation of a subset of primary dsRNA-responsive genes. An initiates a distinct signal transduction program mediated by RNA analysis of the molecular determinants underlying this defect re- helicases RIG-I and MDA-5 (5). Although these two helicases vealed that FADD appears to play a role in regulating the levels of share ϳ75% homology and can activate similar pathways, they the IRF-7 transcription factor following virus infection. Recent seem to manifest distinct affinities for dsRNA and respond to dif- mouse knockout studies have revealed that IRF-7 is the master ferent viruses. RIG-I appears important for IFN production after virus-responsive regulator of type I IFN production (32). In the paramyxovirus, VSV, and influenza virus infection, whereas absence of virus infection, IRF-7 in MEFs is found at very low MDA-5 is necessary for protection against encephalomyocarditis levels. The levels of IRF-7, however, increase with passage num- virus and response to synthetic dsRNA in vivo (12–14). A number ber, underscoring the importance of performing experiments in of studies now demonstrate that IPS-1 links these helicases to the early passage, nonimmortalized MEFs (47). Basal levels of IRF-7 2438 REGULATION OF Irf7 BY FADD seem to be maintained by low constitutive IFN production and the recent demonstration of a critical role for FADD in dsRNA- signaling, and these basal levels of IRF-7 are important for the mediated proliferative responses in B cells, indicating that, whereas initial induction of primary genes in response to dsRNA (31). In a FADD seems dispensable for TLR3-mediated IFN-␤ activation in current model, dsRNA produced by the virus activates RIG-I or fibroblasts, it is nevertheless required for TLR3-dependent signaling MDA-5, culminating in TBK1/IKK-␧-mediated formation of ei- in other more physiologically relevant cell types (51). Whether FADD ther IRF-7 homodimers or IRF-3/IRF-7 heterodimers and conse- plays a broader role in TLR signaling events, for example by partic- quent induction of primary genes, including IFN-␤. Autocrine type ipating in TLR9-subfamily mediated IRF7-dependent induction of I IFN activity then results in the production of greater amounts of IFN-␣ in plasmacytoid DCs, remains to be seen. IRF-7, its further activation by TBK1/IKK-␧, and full induction of Numerous viruses encode products that inhibit dsRNA-initiated the entire range of antiviral genes, including secondary IFN-␣s, in signaling events. For example, paramyxovirus V proteins have a positive feedback loop (4). Our data indicate that FADD partic- been shown to directly interact with MDA-5 and inhibit its activity ipates in the critical initial phase of primary type I IFN induction (39). Several groups have also shown that the hepatitis C virus- by dsRNA and virus. In the absence of FADD, we notice a sig- encoded NS3/4a protease cleaves and inactivates IPS-1, thereby nificant delay in the production of IFN-␤ by either dsRNA stim- neutralizing RIG-I and MDA-5 signaling (18, 40, 52–54). Addi- ulation or virus infection. Because robust early induction of Irf7 is tionally, Borna disease virus P protein binds to and inhibits TBK1 critical for the establishment of the positive feedback expression of (55). In this study, we examined the effects of molluscum conta- IFN-␣ (and other secondary response genes), we propose that a giosum poxvirus-encoded v-FLIP MC159 on FADD-dependent delay in IFN-␤ production prevents efficient establishment of this dsRNA signaling. MC159 binds strongly to FADD via its death positive feedback loop and results in poor induction of Irf7 and effector domains, and inhibits FADD-dependent apoptotic events. Downloaded from consequent loss of secondary gene activation. Continual produc- We show in this study that MC159 is also a potent inhibitor of tion of autocrine IFN during the course of a viral infection is crit- dsRNA-activated RIG-I and IPS-1 signaling, adding it to a grow- ical for the protection of fibroblasts, despite pretreatment with ex- ing list of virally encoded inhibitors of innate immune signaling. ogenously supplied IFN, as we have shown previously (10). In summary, our studies here indicate that FADD is required for Alternatively, our preliminary data also suggest that FADD may RIG-I, MDA-5, and IPS-1-mediated production of type I IFN. Al-

directly impinge upon the Irf7 promoter following dsRNA stimulation though the exact mechanism(s) by which FADD participates in http://www.jimmunol.org/ or virus activation. IFN-independent activation of Irf7 is not without these signaling events remains unclear, it appears that FADD may precedent. For example, studies on the Irf7 promoter show that it can lie downstream of IPS-1 and upstream of IRF-7 in the “classical” be activated by TNF-␣ and TPA in a manner dependent on NF-␬B (i.e., TLR-independent) antiviral gene induction pathway required (36). In addition, EBV-encoded latent membrane protein-1 can also for the production of type I IFNs, especially the IFN-␣s. In the induce Irf7 by mechanisms that also may depend on NF-␬B signaling initial stages of a viral infection, a lack of FADD impairs estab- (48). It is thus conceivable that FADD, which can potently stimulate lishment of a positive feedback loop required for production of NF-␬B activity, may regulate the Irf7 promoter though similar mech- IFN-␣ and the robust induction of an antiviral state. Taken together anisms following dsRNA stimulation or virus infection. In agreement with the fact that IFN-␣s are crucial components of antiviral host with such a role for FADD, we show that NF-␬B p65 binding to a defense, these findings help explain the remarkable susceptibility by guest on September 24, 2021 consensus NF-␬B binding site following dsRNA stimulation is sig- of FaddϪ/Ϫ fibroblasts to virus infection. nificantly impaired in the absence of FADD. FADD-mediated NF- ␬B-dependent activity would represent a more direct parallel to IMD Acknowledgments signaling in Drosophila, where a FADD-dependent innate-immune We are grateful to Drs. Shizuo Akira (Osaka University, Osaka, Japan), cascade uses NF-␬B-like signaling to directly transactivate innate im- Joan Durbin (Columbus Children’s Research Institute, Columbus, OH), mune genes. In the IMD pathway, the RIP1 homologue IMD binds Wen-Chen Yeh (University of Toronto, Toronto, Canada), John Hiscott dFADD to activate innate immune responses (11). (McGill University, Montreal, Canada), Paula Pitha-Rowe (Johns Hopkins Although FADD is required for optimal induction of IFN-␤ fol- University, Baltimore, MD), Tom Maniatis (Harvard University, Cam- lowing dsRNA stimulation, its requirement for the production of bridge, MA), and Joanna Shisler (University of Illinois, Urbana, IL) for various cells and plasmids. IFN-␤ after infection with either VSV, SeV, or Newcastle disease virus appears less critical (23). Virally infected cells thus appear to Disclosures possess a FADD-independent mechanism(s) of IFN-␤ gene induc- The authors have no financial conflict of interest. tion that is at least partially distinct from synthetic dsRNA-acti- Ϫ/Ϫ vated signaling, and requires TBK1, because Tbk1 MEFs are References almost completely defective in IFN-␤ induction by either virus or 1. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate dsRNA (49, 50). Also of interest is the fact that TRIF (as well as immunity. Cell 124: 783–801. other members of the TLR signaling pathway) can activate IFN-␤ 2. Medzhitov, R., and C. A. Janeway, Jr. 1997. Innate immunity: the virtues of a normally in FaddϪ/Ϫ MEFs, whereas RIG-I, MDA-5, and IPS-1 nonclonal system of recognition. Cell 91: 295–298. 3. Akira, S., and K. Takeda. 2004. Toll-like signalling. Nat. Rev. Immunol. cannot. Recent studies have begun to highlight significant differ- 4: 499–511. ences between RIG-I-mediated and TLR-3-, 7-, or 9-mediated in- 4. Honda, K., H. Yanai, A. Takaoka, and T. Taniguchi. 2005. Regulation of the type ␣␤ I IFN induction: a current view. Int. Immunol. 17: 1367–1378. duction of type I IFN (5). For example, TLR3 activates IFN- 5. Meylan, E., and J. Tschopp. 2006. Toll-like receptors and RNA helicases: two gene expression via a TRIF-NAP1-TBK1-dependent pathway, parallel ways to trigger antiviral responses. Mol. Cell 22: 561–569. whereas the TLR9 subfamily appears to directly recruit and phos- 6. Pasare, C., and R. Medzhitov. 2005. Toll-like receptors: linking innate and adap- tive immunity. Adv. Exp. Med. Biol. 560: 11–18. phorylate IRF-7 via MyD88 and the IRAK family of kinases to 7. Kawai, T., and S. Akira. 2006. Innate immune recognition of viral infection. Nat. induce type I IFN gene expression (4, 5). Although these examples Immunol. 7: 131–137. provide possible explanations for FADD-independent TLR signal- 8. Balachandran, S., P. C. Roberts, L. E. Brown, H. Truong, A. K. Pattnaik, D. R. Archer, and G. N. Barber. 2000. Essential role for the dsRNA-dependent protein ing in fibroblasts, the fact that these cells express undetectable kinase PKR in innate immunity to viral infect. Immunity 13: 129–141. endogenous levels of TLR3 and are essentially unresponsive to 9. Williams, B. R. 1999. PKR: a sentinel kinase for cellular stress. Oncogene 18: 6112–6120. untransfected dsRNA make them poor cells of choice to examine 10. Balachandran, S., E. Thomas, and G. N. Barber. 2004. A FADD-dependent innate FADD’s role in TLR3-dependent signaling. This is emphasized by immune mechanism in mammalian cells. Nature 432: 401–405. The Journal of Immunology 2439

11. Hoffmann, J. A. 2003. The immune response of Drosophila. Nature 426: 33–38. 34. Marie, I., E. Smith, A. Prakash, and D. E. Levy. 2000. Phosphorylation-induced 12. Kato, H., S. Sato, M. Yoneyama, M. Yamamoto, S. Uematsu, K. Matsui, T. dimerization of regulatory factor 7 unmasks DNA binding and a bi- Tsujimura, K. Takeda, T. Fujita, O. Takeuchi, and S. Akira. 2005. Cell type- partite transactivation domain. Mol. Cell. Biol. 20: 8803–8814. specific involvement of RIG-I in antiviral response. Immunity 23: 19–28. 35. Matsuda, A., Y. Suzuki, G. Honda, S. Muramatsu, O. Matsuzaki, Y. Nagano, 13. Kato, H., O. Takeuchi, S. Sato, M. Yoneyama, M. Yamamoto, K. Matsui, S. T. Doi, K. Shimotohno, T. Harada, E. Nishida, et al. 2003. Large-scale identifi- Uematsu, A. Jung, T. Kawai, K. J. Ishii, et al. 2006. Differential roles of MDA5 cation and characterization of human genes that activate NF-␬B and MAPK and RIG-I helicases in the recognition of RNA viruses. Nature 441: 101–105. signaling pathways. Oncogene 22: 3307–3318. 14. Gitlin, L., W. Barchet, S. Gilfillan, M. Cella, B. Beutler, R. A. Flavell, M. S. 36. Lu, R., P. A. Moore, and P. M. Pitha. 2002. Stimulation of IRF-7 gene expression ␣ ␬ Diamond, and M. Colonna. 2006. Essential role of mda-5 in type I IFN responses by : requirement for NF B transcription factor and gene to polyriboinosinic:polyribocytidylic acid and encephalomyocarditis picornavi- accessibility. J. Biol. Chem. 277: 16592–16598. ␤ rus. Proc. Natl. Acad. Sci. USA 103: 8459–8464. 37. Thanos, D., and T. Maniatis. 1995. Virus induction of human IFN gene ex- 15. Kang, D. C., R. V. Gopalkrishnan, Q. Wu, E. Jankowsky, A. M. Pyle, and P. B. pression requires the assembly of an enhanceosome. Cell 83: 1091–1100. Fisher. 2002. Mda-5: an interferon-inducible putative RNA helicase with double- 38. McWhirter, S. M., B. R. Tenoever, and T. Maniatis. 2005. Connecting mitochon- stranded RNA-dependent ATPase activity and melanoma growth-suppressive dria and innate immunity. Cell 122: 645–647. properties. Proc. Natl. Acad. Sci. USA 99: 637–642. 39. Andrejeva, J., K. S. Childs, D. F. Young, T. S. Carlos, N. Stock, S. Goodbourn, and R. E. Randall. 2004. The V proteins of paramyxoviruses bind the IFN-in- 16. Yoneyama, M., M. Kikuchi, T. Natsukawa, N. Shinobu, T. Imaizumi, M. ducible RNA helicase, mda-5, and inhibit its activation of the IFN-␤ promoter. Miyagishi, K. Taira, S. Akira, and T. Fujita. 2004. The RNA helicase RIG-I has Proc. Natl. Acad. Sci. USA 101: 17264–17269. an essential function in double-stranded RNA-induced innate antiviral responses. 40. Li, X. D., L. Sun, R. B. Seth, G. Pineda, and Z. J. Chen. 2005. Hepatitis C virus Nat. Immunol. 5: 730–737. protease NS3/4A cleaves mitochondrial antiviral signaling protein off the 17. Kawai, T., K. Takahashi, S. Sato, C. Coban, H. Kumar, H. Kato, K. J. Ishii, mitochondria to evade innate immunity. Proc. Natl. Acad. Sci. USA 102: O. Takeuchi, and S. Akira. 2005. IPS-1, an adaptor triggering RIG-I- and Mda5- 17717–17722. mediated type I interferon induction. Nat. Immunol. 6: 981–988. 41. Cardenas, W. B., Y. M. Loo, M. Gale, Jr., A. L. Hartman, C. R. Kimberlin, 18. Meylan, E., J. Curran, K. Hofmann, D. Moradpour, M. Binder, R. Bartenschlager, L. Martinez-Sobrido, E. O. Saphire, and C. F. Basler. 2006. virus VP35 and J. Tschopp. 2005. Cardif is an adaptor protein in the RIG-I antiviral pathway protein binds double-stranded RNA and inhibits ␣␤ interferon production in- and is targeted by hepatitis C virus. Nature 437: 1167–1172. duced by RIG-I signaling. J. Virol. 80: 5168–5178. Downloaded from 19. Seth, R. B., L. Sun, C. K. Ea, and Z. J. Chen. 2005. Identification and charac- 42. Loo, Y. M., D. M. Owen, K. Li, A. K. Erickson, C. L. Johnson, P. M. Fish, terization of MAVS, a mitochondrial antiviral signaling protein that activates D. S. Carney, T. Wang, H. Ishida, M. Yoneyama, et al. 2006. Viral and thera- ␬ NF- B and IRF 3. Cell 122: 669–682. peutic control of IFN-␤ promoter stimulator 1 during hepatitis C virus infection. 20. Xu, L. G., Y. Y. Wang, K. J. Han, L. Y. Li, Z. Zhai, and H. B. Shu. 2005. VISA Proc. Natl. Acad. Sci. USA 103: 6001–6006. is an adapter protein required for virus-triggered IFN-␤ signaling. Mol. Cell 19: 43. Thome, M., P. Schneider, K. Hofmann, H. Fickenscher, E. Meinl, F. Neipel, 727–740. C. Mattmann, K. Burns, J. L. Bodmer, M. Schroter, et al. 1997. Viral FLICE- 21. Saha, S. K., E. M. Pietras, J. Q. He, J. R. Kang, S. Y. Liu, G. Oganesyan, inhibitory proteins (FLIPs) prevent apoptosis induced by death receptors. Nature

A. Shahangian, B. Zarnegar, T. L. Shiba, Y. Wang, and G. Cheng. 2006. Reg- 386: 517–521. http://www.jimmunol.org/ ulation of antiviral responses by a direct and specific interaction between TRAF3 44. Bertin, J., R. C. Armstrong, S. Ottilie, D. A. Martin, Y. Wang, S. Banks, G. H. and Cardif. EMBO J. 25: 3257–3263. Wang, T. G. Senkevich, E. S. Alnemri, B. Moss, et al. 1997. Death effector 22. Takahashi, K., T. Kawai, H. Kumar, S. Sato, S. Yonehara, and S. Akira. 2006. domain-containing herpesvirus and poxvirus proteins inhibit both Fas- and Roles of caspase-8 and caspase-10 in innate immune responses to double- TNFR1-induced apoptosis. Proc. Natl. Acad. Sci. USA 94: 1172–1176. stranded RNA. J. Immunol. 176: 4520–4524. 45. Hu, S., C. Vincenz, M. Buller, and V. M. Dixit. 1997. A novel family of viral 23. Yoneyama, M., M. Kikuchi, K. Matsumoto, T. Imaizumi, M. Miyagishi, K. Taira, death effector domain-containing molecules that inhibit both CD-95- and tumor E. Foy, Y. M. Loo, M. Gale, Jr., S. Akira, et al. 2005. Shared and unique func- necrosis factor receptor-1-induced apoptosis. J. Biol. Chem. 272: 9621–9624. tions of the DExD/H-box helicases RIG-I, MDA5, and LGP2 in antiviral innate 46. Garvey, T. L., J. Bertin, R. M. Siegel, G. H. Wang, M. J. Lenardo, and J. I. immunity. J. Immunol. 175: 2851–2858. Cohen. 2002. Binding of FADD and caspase-8 to molluscum contagiosum virus 24. Rothenfusser, S., N. Goutagny, G. DiPerna, M. Gong, B. G. Monks, A. MC159 v-FLIP is not sufficient for its antiapoptotic function. J. Virol. 76: Schoenemeyer, M. Yamamoto, S. Akira, and K. A. Fitzgerald. 2005. The RNA 697–706.

helicase Lgp2 inhibits TLR-independent sensing of viral replication by retinoic 47. Hata, N., M. Sato, A. Takaoka, M. Asagiri, N. Tanaka, and T. Taniguchi. 2001. by guest on September 24, 2021 acid-inducible gene-I. J. Immunol. 175: 5260–5268. Constitutive IFN-␣␤ signal for efficient IFN-␣␤ gene induction by virus. Bio- 25. Sun, Q., L. Sun, H. H. Liu, X. Chen, R. B. Seth, J. Forman, and Z. J. Chen. 2006. chem. Biophys. Res. Commun. 285: 518–525. The specific and essential role of MAVS in antiviral innate immune responses. 48. Zhang, L., L. Wu, K. Hong, and J. S. Pagano. 2001. Intracellular signaling mol- Immunity 24: 633–642. ecules activated by Epstein-Barr virus for induction of interferon regulatory fac- 26. Kumar, H., T. Kawai, H. Kato, S. Sato, K. Takahashi, C. Coban, M. Yamamoto, tor 7. J. Virol. 75: 12393–12401. S. Uematsu, K. J. Ishii, O. Takeuchi, and S. Akira. 2006. Essential role of IPS-1 49. Hemmi, H., O. Takeuchi, S. Sato, M. Yamamoto, T. Kaisho, H. Sanjo, T. Kawai, ␬ in innate immune responses against RNA viruses. J. Exp. Med. 203: 1795–1803. K. Hoshino, K. Takeda, and S. Akira. 2004. The roles of two I B kinase-related kinases in lipopolysaccharide and double stranded RNA signaling and viral in- 27. Leulier, F., S. Vidal, K. Saigo, R. Ueda, and B. Lemaitre. 2002. Inducible ex- fection. J. Exp. Med. 199: 1641–1650. pression of double-stranded RNA reveals a role for dFADD in the regulation of 50. Perry, A. K., E. K. Chow, J. B. Goodnough, W. C. Yeh, and G. Cheng. 2004. the antibacterial response in Drosophila adults. Curr. Biol. 12: 996–1000. Differential requirement for TANK-binding kinase-1 in type I interferon re- 28. Naitza, S., C. Rosse, C. Kappler, P. Georgel, M. Belvin, D. Gubb, J. Camonis, sponses to Toll-like receptor activation and viral infection. J. Exp. Med. 199: J. A. Hoffmann, and J. M. Reichhart. 2002. The Drosophila immune defense 1651–1658. against gram-negative infection requires the death protein dFADD. Immunity 17: 51. Imtiyaz, H. Z., S. Rosenberg, Y. Zhang, Z. S. Rahman, Y. J. Hou, T. Manser, and 575–581. J. Zhang. 2006. The Fas-associated death domain protein is required in apoptosis 29. Barnes, B., B. Lubyova, and P. M. Pitha. 2002. On the role of IRF in host defense. and TLR-induced proliferative responses in B cells. J. Immunol. 176: 6852–6861. J. Interferon Cytokine Res. 22: 59–71. 52. Foy, E., K. Li, C. Wang, R. Sumpter, Jr., M. Ikeda, S. M. Lemon, and M. Gale, 30. Nguyen, H., J. Hiscott, and P. M. Pitha. 1997. The growing family of interferon Jr. 2003. Regulation of interferon regulatory factor-3 by the hepatitis C virus regulatory factors. Cytokine Rev. 8: 293–312. serine protease. Science 300: 1145–1148. 31. Taniguchi, T., and A. Takaoka. 2002. The interferon-␣␤ system in antiviral re- 53. Lin, R., J. Lacoste, P. Nakhaei, Q. Sun, L. Yang, S. Paz, P. Wilkinson, I. sponses: a multimodal machinery of gene regulation by the IRF family of tran- Julkunen, D. Vitour, E. Meurs, and J. Hiscott. 2006. Dissociation of a MAVS/ scription factors. Curr. Opin. Immunol. 14: 111–116. IPS-1/VISA/Cardif-IKK␧ molecular complex from the mitochondrial outer mem- 32. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, brane by hepatitis C virus NS3-4A proteolytic cleavage. J. Virol. 80: 6072–6083. Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master 54. Cheng, G., J. Zhong, and F. V. Chisari. 2006. Inhibition of dsRNA-induced regulator of type-I interferon-dependent immune responses. Nature 434: signaling in hepatitis C virus-infected cells by NS3 protease-dependent and -in- 772–777. dependent mechanisms. Proc. Natl. Acad. Sci. USA 103: 8499–8504. 33. Takaoka, A., H. Yanai, S. Kondo, G. Duncan, H. Negishi, T. Mizutani, S. Kano, 55. Unterstab, G., S. Ludwig, A. Anton, O. Planz, B. Dauber, D. Krappmann, G. K. Honda, Y. Ohba, T. W. Mak, and T. Taniguchi. 2005. Integral role of IRF-5 Heins, C. Ehrhardt, and T. Wolff. 2005. Viral targeting of the interferon-␤-in- in the gene induction programme activated by Toll-like receptors. Nature 434: ducing Traf family member-associated NF-␬B activator (TANK)-binding ki- 243–249. nase-1. Proc. Natl. Acad. Sci. USA 102: 13640–13645.