and Immunity (2008) 9, 168–175 & 2008 Nature Publishing Group All rights reserved 1466-4879/08 $30.00 www.nature.com/gene

SHORT COMMUNICATION IRF-3-dependent and augmented target genes during viral infection

J Andersen1, S VanScoy2, T-F Cheng2, D Gomez2 and NC Reich2 1Department of Biochemistry and Cell Biology, Stony Brook University, New York, NY, USA and 2Department of Molecular & Microbiology, Stony Brook University, New York, NY, USA

Activation of the factor regulatory factor-3 (IRF-3) is an essential event in the innate immune response to viral infection. To understand the contribution of IRF-3 to host defense, we used a systems biology approach to analyze global expression dependent on IRF-3. Comparison of expression profiles in cells from IRF-3 knockout animals or wild-type siblings following viral infection revealed three sets of induced genes, those that are strictly dependent on IRF-3, augmented with IRF-3, or not responsive to IRF-3. Products of identified IRF-3 target genes are involved in innate or acquired immunity, or in the regulation of cell cycle, apoptosis and proliferation. These results reveal the global effects of one in the immune response and provide information to evaluate the integrated response to viral infection. Genes and Immunity (2008) 9, 168–175; doi:10.1038/sj.gene.6364449; published online 20 December 2007

Keywords: transcription factor; host defense; innate immunity; global profiling; ;

Introduction A critical cellular transcription factor that is activated both in response to viral infection and to toll-like An effective first line of defense to infection depends on receptors is the interferon regulatory factor-3 (IRF-3).1–11 the action of cells and mediators of innate immunity, IRF-3 is expressed constitutively in all cells of the body, and a specific and long-lasting defense depends on and animals with a targeted gene disruption are the collaboration of these components with cells and susceptible to viral infection.9,10 IRF-3 resides in a latent mediators of acquired immunity. Infections usually begin state primarily in the cytoplasm, but following serine in a single site or tissue, but the physiological reaction is phosphorylation by TANK-binding kinase or inhibitor of both local and disseminating. Responses include produc- NF-kB kinase-related IKKe, it accumulates in the nucleus tion of at the site that lead to recruitment of in strong association with the histone acetyl transferases phagocytes and lymphocytes, and in turn these cytokines CREB-binding or p300.12–14 and cellular responders disseminate to involve the entire Identification of genes targeted by IRF-3 is critical to immune system. For this reason analyzing the effects of understanding an effective defense response to viral infection on a single cell or tissue can provide knowledge infection. IRF-3 is known to regulate expression of the of a more global systemic response. type I interferon-beta gene (IFN-b), and a subset of Viral infection of cells causes the activation of a type I IFN-stimulated genes independent of the action number of transcription factors that, in turn, induce of IFN.3,11,15 IFNs are unique among the cytokines for genes whose products generate a biological response, be conferring cellular resistance to viral infections, and it survival of the infected cell or survival of the host with therefore a successful innate immune response depends the sacrifice of the infected cell. Each activated transcrip- on their production and action.16–19 IRF-3 contributes to tion factor may function distinctly to regulate expression the induction of the IFN-b gene, but it is not sufficient for of specific target genes, or they may function coopera- IFN-b gene induction and needs to cooperate in an tively to regulate expression of a common target gene. with NF-kB and ATF-2/c-jun transcrip- Cooperative gene induction can result from the con- tion factors.20 IFN is secreted by the infected cell and certed action of multiple transcription factors, although binds to cell surface receptors on the same cell or each transcription factor individually may not be adjacent cells stimulating a signal pathway that induces sufficient to induce the gene target. expression of IFN-stimulated genes. IFN- binding activates Janus tyrosine kinases that phosphory- late the DNA-binding transcription factors, signal trans ducers and activators of transcription (STAT).16–18 Correspondence: Dr NC Reich, Department of Molecular Genetics Since IRF-3 plays a significant role in viral defense, & Microbiology, Stony Brook University, Nicolls Rd, Life Sciences efforts have been made to identify genes regulated by Bldg, Stony Brook, New York 11794, USA. E-mail: [email protected] IRF-3. One approach evaluated the effects of a consti- 21 Received 7 September 2007; revised 8 November 2007; accepted 13 tutively active IRF-3 mutant. Serine amino acids November 2007; published online 20 December 2007 were replaced with aspartic acid to mimic serine IRF-3 target genes J Andersen et al 169 phosphorylation and the active IRF-3 conformation. Only a handful of genes were found to be induced by overexpression of the constitutive IRF-3 mutant and this group did not include IFN. The low number of induced genes may be due to the fact that IRF-3 often functions cooperatively with other activated transcription factors, or that the constitutive mutant does not fully simulate the active form of IRF-3. Another approach evaluated the response to viral infection of mutated cell lines that were selected to either overexpress IRF-3, or express IRF-3 below normal levels.22 Cells with reduced IRF-3 levels underexpressed some known IFN-stimulated genes, but there was no effect on expression of IFN genes. Interpretation of these results is limited since low levels of IRF-3 are still expressed and activation of other transcription factors may be sufficient to induce ISG expression independent of IRF-3. To provide knowledge of the global contribution of IRF-3 to host defense in viral infection, it is essential to evaluate the profile of genes regulated in the presence or absence of IRF-3. Genes induced after viral infection can thereby be compared using controlled parameters to Figure 1 Gene expression in virally infected cells in the presence or absence of IRF-3. (a) Northern blot analysis. MEF cultures were identify those genes that are independently or coordina- generated from the IRF-3 knockout (KO) and wild-type (WT) sibling tely regulated by IRF-3. In this study, we used a systems mice and used between passages 20–35.9 Cells were mock infected biology approach to identify the global set of cellular (À) or infected with 100 hemagglutination units (HAU) mlÀ1 genes influenced by IRF-3 during the course of viral Newcastle Disease Virus (NJ-LaSota-1946) ( þ ) for 6 or 12 h.23 infection by employing microarray technology and Neutralizing antibody to type I IFN (IFN Ab) (313 units mlÀ1; virally-infected murine embryo fibroblasts (MEFs) Biosource International Inc.) was added one hour prior to infection þ derived from IRF-3 knockout (irf-3À/À) or wild-type of half the 6 h samples as indicated ( ). RNA was isolated with þ / þ 9 RNeasy reagents (Qiagen, Valencia, CA, USA). DNA probes were (WT) (irf-3 ) sibling mice. We used Newcastle labeled with a-32P-dCTP using PrimeIT (Stratagene, Cedar Creek, Disease Virus as our model infectious agent, because TX, USA) with human b-actin 700 bp NotI/BamHI cDNA fragment, this negative strand paramyxovirus leads to efficient murine GAPDH cDNA (gift of Dr Jizu Zhi) and murine ISG54 activation of latent IRF-3, and viral replication is limited cDNA generated by RT-PCR with primers 50-AAACACCAGTGGG 0 to one round of infection in murine cells. GATGAAG and 5 -CGTCTCATACTGGGCCCACTT. The reverse image of ethidium bromide-stained 18S rRNA on the membrane is shown. (b) Western blot analysis. Cells were infected for 0, 6, or 12 h with Newcastle Disease Virus similarly to (a) and were lysed with Results and discussion 0.5% NP40 buffer.14 Molecular mass marker (m) or 70 mg protein from each sample were separated by SDS-PAGE. were Cells and used for microarray analysis transferred to membrane and reacted with anti-ISG54 antibody Cultures of IRF-3 knockout (KO) (irf-3À/À/MEF6) and WT (peptide 26–50: Genemed Synthesis Inc., South San Francisco, CA, þ / þ USA) and Alexa Fluor 680 goat anti-rabbit secondary antibody sibling (irf-3 /MEF7) MEFs used in this study were (Molecular Probes, Eugene, OR, USA) before visualization with an evaluated for mRNA expression prior to and following Odyssey infrared imaging system (Li-COR Biosciences, Lincoln, 6 or 12 h infection with Newcastle Disease Virus.9 NE, USA). Northern blot hybridization was used to ensure RNA integrity of samples used in the microarray assays (Figure 1a). Following infection there was a clear increase system, we evaluated the expression of the ISG54 protein in a known IRF-3 target gene, IFN-stimulated gene 54 in WT and IRF-3 KO cells following viral infection (ISG54), in the WT cells but not the IRF-3 KO cells. The (Figure 1b). Western blot analysis demonstrated that addition of exogenous neutralizing antibody to type I expression of the ISG54 protein accurately reflected the IFN in the culture resulted in approximately a twofold induction of ISG54 mRNA in WT cells and not in decrease in the ISG54 levels in the infected WT (lane 8). KO cells. This is not unexpected since autocrine IFN produced during the course of infection can induce ISG54 via the Experimental design for microarray analysis Janus tyrosine kinases/STAT pathway and thereby A comparison of the genes expressed in infected WT increase the levels. The complete lack of ISG54 expres- MEFs versus infected IRF-3 KO MEFs should identify sion in the IRF-3 KO MEFs supports the premise that differences due to the presence or absence of IRF-3. IRF-3 is required for its induction either directly or Therefore, we evaluated RNA samples from cells mock- indirectly. The RNAs represented on these northern blots infected or infected for 6 or 12 h, with or without were among similar pooled samples submitted for the neutralizing antibody or the translation inhibitor cyclo- microarray analyses. heximide (50 mgmlÀ1) to block the synthesis of autocrine Changes in gene expression following viral infection IFN or other secondary transcription factors. Genes and activation of IRF-3 can have a dramatic impact on induced in the presence of cycloheximide should host defense. The products of induced genes affect host represent those directly responsive to infection. The responses both locally and distant to the site of infection. parameters for these comparisons are listed in Table 1, To ensure that the induced mRNAs are translated in our and the profiles of genes in each treatment that changed

Genes and Immunity IRF-3 target genes J Andersen et al 170 Table 1 Parameters of seven primary comparisons and strategy diagram for secondary comparisons

1. WT MEFs NDV infected for 6 hours vs. WT MEFs mock infected

2. WT MEFs NDV infected for 6 hours plus cycloheximide vs. WT MEFs mock infected for 6 hours plus cycloheximide

IRF-3 3. WT MEFs NDV infected for 6 hours plus neutralizing IFN Ab Dependent vs. WT MEFs infected for 6 hours without Ab change in WT, not KO 4. WT MEFs NDV infected for 12 hours vs. WT MEFs mock infected IRF-3 Independent 5. IRF-3 KO MEFs NDV infected for 6 hours vs. IRF-3 KO MEFs mock infected change in WT and KO 6. IRF-3 KO MEFs NDV infected for 6 hours plus neutralizing IFN Ab vs. IRF-3 KO MEFs infected for 6 hours without Ab

7. IRF-3 KO MEFs NDV infected for 12 hours vs. IRF-3 KO MEFs mock infected

RNAs from similar cell treatments were pooled and cDNA generated for hybridization to affymetrix MG U74A version 2 gene chip array. Chips were scanned and data analyzed using Microarray Suite v2.

using the Affymetrix murine microarray Chip MG following 12 h of viral infection is presented in Table 2. U74Av2 are provided in Supplementary Tables 1–7. Even They are grouped based on their roles in particular inductions less than twofold detected by Affymetrix biological responses of innate immunity, acquired im- microarray were found to be significant since most genes munity, proliferation and apoptosis. verified by quantitative-PCR proved to have greater fold At the top of the list of immunity genes that require inductions (subset in Supplementary Table 8). Analyses IRF-3 is a subset of the type I IFN genes, IFN-b, IFN-a4 were performed with a pooled population of RNA and IFN-a5. The IFN cytokines are critical for an effective obtained from three experiments. antiviral response and for integration of the innate and To reduce differences due to parameters of individual acquired immune systems. The expression of type I IFN experiments or genetic changes that occur during genes is tightly regulated, and IRF-3 is known to be passage of cells in tissue culture, the results from these involved in induction of IFN-b and IFN-a4 in response to primary comparisons were subjected to second-genera- virus.9,25 Induction of IFN-a5 was previously described tion comparisons following a logical strategy. The to be dependent on the IRF-7 transcription factor criterion of a second-generation comparison should induced by IFN.25 Our analysis now identifies IFN-a5 exclude gene expression changes that are dependent on as an IRF-3-dependent direct response gene. The IFNs new protein synthesis, and gene changes that did not are secreted from infected cells and thereby disseminate respond reproducibly in multiple experiments. Genes immune response signals to other cells of the body. were identified that changed in WT cells following a 6 h Another IRF-3-dependent direct response gene that leads infection in the presence of cycloheximide, and these to signal dissemination and recruitment of monocytes, were considered the direct response genes (Table 1). This mast cells and lymphocytes to the site of infection is set of direct response genes was then compared to the the chemokine CCL4 (MIP-1b).26 Binding of CCL4 to sets of genes that consistently changed in the IRF-3 KO the CCR5 receptor may also play a more specific role cells after infection for either 6 or 12 h (Table 1). Direct in inhibition of specific viral entry since CCR5 is a response genes that changed in WT but not in the IRF-3 coreceptor for human immunodeficiency virus-1.27 KO infected cells from these secondary comparisons Secreted cytokines and chemokines can recruit other represent strong candidates for IRF-3-dependent direct immune cells to the infection site, and cell adherence response genes. may be stimulated by the IRF-3-dependent production of the adhesion molecule CD166 (ALCAM) as well as T-cell IRF-3-dependent direct response genes proliferation.28 IRF-3 was also found to be necessary for Some of the IRF-3-dependent direct response genes have induction of non-classical major histocompatability class known functions, whereas the functions of others remain I genes Q-I, T-10 and T-22. The non-classical major to be determined, and still others remain as novel EST histocompatability class I proteins serve as for genes (Supplementary Table 9). We have focused on receptors on natural killer cells and g/d T cells.29,30 It is genes whose expression increased during infection clear from this small set of genes that the IRF-3 rather than decreased since it is known that RNA can transcription factor acts to elicit widespread immune be degraded during viral infection by activated enzymes responses. such as RNase L.24 These genes increased expression in The IRF-3-dependent direct response genes include a WT cells during viral infection despite the presence of few that are also induced by the Janus tyrosine kinases/ cycloheximide, and did not increase in the IRF-3 KO cells STAT pathway stimulated in response to type I IFN; even after 12 h of viral infection. A subset of IRF-3- IFP35, ISG20, I-8U and ISG54. IFP35 encodes a transcrip- dependent response genes that significantly changed tion factor and can thereby regulate a secondary set of

Genes and Immunity IRF-3 target genes J Andersen et al 171 Table 2 IRF-3-dependent direct response genes

Genbank WT 12 h WT 6 h CX Description Change Change

Innate/acquired immunity K00020 68.6 34.3 Interferon-b X01973 11.3 2.6 Interferon-a 4 X01971 5.7 2.1 Interferon-a 5 M35244 13.9 5.3 Class I major histocompatability 2, non-classical T-10 M35247 11.3 4.9 Class I major histocompatability 2, non-classical T-22 X01838 6.1 2.8 Class I major histocompatability, non-classical Q-1 X62502 42.2 24.3 CCL4 chemokine/MIP-1b L25274 4.6 1.6 CD166/Activated leukocyte adhesion molecule (ALCAM) AI846797 2.8 2.0 Tropomodulin 3

IFN-stimulated genes AA204579 415.9 64.0 CMV-induced gene 5 (CIG5)/Viperin U43085 42.2 64.0 IFN-stimulated gene 54 with tetratricopeptide repeat (ISG54) AW121732 27.9 6.1 IFN-induced 35 kDa with (IFP35) AW122677 16.0 3.0 IFN-stimulated gene 20, exoribonuclease (ISG20) AA960657 10.6 2.8 IFN-inducible p200 family transcription regulator (p202/Ifi 16) U19119 4.9 21.1 IFN-inducible GTPase (LRG-47) AW125390 4.0 4.3 IFN-inducible transmembrane protein (1-8U)

Stress-related Y12657 7.5 4.6 Cytochrome p450 AW046479 7.0 1.9 Ubiquitin-activating enzyme E1-like U60329 5.7 2.1 Proteasome 28-b/PA28b U43678 5.7 32.0 Ataxia telangiectasia mutated homolog/ATM AW123880 4.9 1.4 XBP-1 transcription factor X92665 4.6 1.7 Ubiquitin-conjugating ligase E1 AA822898 5.7 13.0 Sp-100 autoantigen, PML-associated AI851762 5.7 2.8 Ribonuclease T2 AB007136 4.3 1.3 Proteasome 28 subunit, a J03520 3.2 3.2 Tissue plasminogen activator (TPA)

Apoptosis X95504 137.2 4.0 finger regulator of apoptosis and cell cycle arrest (Zac1)/Plagl1 AF110520 5.7 2.3 Fas -associated protein/Daax M24377 4.3 1.3 Early gene response 2 (Erg2/Knox-20) AW123754 3.7 2.0 Scotin, -inducible proapoptotic protein

Proliferation L32973 111.4 9.2 Thymidylate kinase family X70472 8.6 2.6 Myeloblastosis-related oncogene (B-Myb) U15012 7.0 1.7 Growth AI836182 6.5 3.5 RNA polymerase II polypeptide G X69620 4.6 4.9 Inhibin-b A chain M17298 4.0 1.4 Nerve b (NGF) U35374 2.8 1.2 Purine nucleoside phosphorylase Y11666 2.6 2.5 Hexokinase 2

genes.31 ISG20 encodes a 30–50 exoribonuclease that there is an increase in expression of the Fas death exhibits antiviral activity against human immuno- domain-associated protein, Daax and a decrease in cyclin deficiency virus type I.32,33 The function of I-8U is A1. Conversely, there is an apparent counterbalance to unknown, but ISG54 may be involved in translation the promotion of apoptosis with the altered expression of control.34 Two of the IRF-3-dependent genes confer genes that stimulate cellular proliferation. These include resistance to specific infectious agents; cytomegalo- an increase in thymidylate kinase, B-Myb oncogene, virus-induced gene 5 (cig5)/viperin blocks late cyto- growth hormone receptor, RNA polymerase II subunit, megalovirus gene expression and virion production, and inhibin-b and nerve growth factor. In the context of viral the GTPase LRG-47 serves to protect macrophages infection, IRF-3 contributes to apparently opposing against infection by Mycobacterium tuberculosis indepen- effects of proliferative versus apoptotic gene expression. dent of inducible nitric oxide synthetase.35–37 Increased expression of stress-related genes was also IRF-3 action has been linked to apoptosis and for this dependent on the expression of IRF-3. Such genes reason target genes that affect cell death or survival were include the DNA damage response kinase, ataxia examined. Two of the IRF-3-dependent genes are known telaniectasia mutated homolog.40–42 transcription factors that stimulate apoptosis or growth The set of IRF-3-dependent genes include those in suppression, Zac1 and Egr2/Knox-20.38,39 In addition, which activated IRF-3 is sufficient to induce expression

Genes and Immunity IRF-3 target genes J Andersen et al 172 and those in which activated IRF-3 is required, but not results in recruitment of neutrophils, monocytes, sufficient to induce expression. Activated IRF-3 is lymphocytes, basophils and eosinophils to the site of sufficient to induce genes such as ISG54 and cig5/ infection. In addition, induction of genes encoding viperin and they were shown previously to be induced secreted cytokines such as interleukin-6 and inter- by a constitutively active mutation of IRF-3.21 In contrast, leukin-15 stimulates the proliferation/activation of IRF-3 is required but not sufficient to induce expression recruited cells. Other notable increases in gene expres- of the IFN-b gene.9,20,21 sion occur within families encoding both major histo- compatibility class I and class II molecules, which are IRF-3 independent and augmented direct response genes required for establishment of specific acquired immunity Viral infection also stimulated the expression of direct and long-term protection against viral infection.43 In response genes in the absence of IRF-3 (Supplementary addition, pro-apoptotic genes such as FasL receptor and Table 10). Although expression of this set of genes was Bid enhance suicide of the infected cell thereby limiting independent of IRF-3, evaluation of the comparative dissemination of virus. levels of gene expression in the WT or KO cells revealed A subset of genes in this group showed a notable two distinct classes. One class of response genes did not higher expression in the WT cells in comparison with the show significant differences in the WT or KO cells IRF-3 KO cells. Expression of this distinct class of following infection, and a subset of these genes is response genes was therefore augmented by the presence presented in Table 3. A concerted immune reaction of IRF-3, although not strictly dependent on IRF-3. to infection is apparent with the induction of three Therefore, the complete list of IRF-3-independent classes of chemokines, including CCL2 (MCP-1), CCL7 response genes (Supplementary Table 10) include those (MCP-3), CXCL10 (IP10), CXCL9, CXCL2 (MIP-2) and that can be induced by IRF-3, but can also be induced 26 CX3CL1. The collaborative action of these chemokines autonomously by another virally-activated transcription

Table 3 IRF-3 independent direct response genes

Genbank WT 12 h WT 6 h KO 12 h KO 6 h Description Change Change Change Change

Innate/acquired immunity X70058 90.5 5.7 18.4 12.1 CCL7 chemokine/MCP-3 M34815 55.7 11.3 CXCL9 chemokine/platelet factor 4 M19681 39.4 5.7 97.0 19.7 CCL2 chemokine/MCP-1 U92565 6.5 2.8 8.6 3.5 CX3Cl1 chemokine/fractalkine X53798 3.2 À1.4 73.5 48.5 CXCL2 chemokine/MIP2 M90551 294.1 548.7 29.9 ICAM1/intercellular adhesion molecule U12884 11.3 2.5 4.0 6.5 Vascular molecule 1 (VCAM1) M27134 78.8 34.3 Class I histocompatability gene Q10 AI117211 55.7 18.4 17.1 Class I histocompatability gene H2-Kd a X00246 55.7 52 Class I histocompatability gene D1 X00496 26.0 13.9 2.8 Class II histocompatibility-associated Ia invariant chain X52490 24.3 3.7 8 2.0 Class I histocompatibility gene H2-D X16202 19.7 3.2 13.0 2.1 Class I histocompatability gene Q4 X58609 13.9 3.0 4.9 1.5 Class I histocompatability gene H2-Q2 M18837 11.3 4.0 7.0 2.1 Class I histocompatability b-2-microglobulin AI836367 17.1 4.9 2.5 1.7 Tapasin/TAP binding protein U06924 17.1 14.9 3.5 Signal transducer and activator of transcription 1 (STAT1) M21065 13.9 9.2 32.0 7.0 IFN regulatory factor (IRF-1) U51992 6.1 8.6 2.8 2.6 IFN regulatory factor 9 (IRF-9) L16956 4.9 1.6 3.0 Janus (JAK2) U59864 4.0 6.1 1.3 TRAF family member-associated NF-kB activator (TANK)

IFN-stimulated genes M74123 55.7 13.9 7.5 1.7 IFN-inducible p200 family protein (p203) M31419 52.0 14.9 24.3 1.9 IFN-inducible p200 family protein (p204) J03368 36.8 78.8 17.1 Myxovirus resistance 2 (MX2)

Stress related/apoptosis AW047653 891.4 831.7 119.4 39.4 Ubiquitin-specific protease 18 X51829 39.4 3.2 4.9 2.8 Growth arrest DNA damage-inducible phosphatase1 (GADD34 g) M83649 3.2 2.0 13.9 3.5 FASL receptor/tumor necrosis factor receptor member 6 U75506 3.0 1.4 12.1 1.6 Bid/BH3-interacting domain death agonist AF030896 59.7 39.4 7.0 NF-kB inhibitor epsilon (IkBe) AI642048 11.3 3.7 12.1 9.8 Nuclear factor of kappa B cell inhibitor, a (IkBa)

Proliferation X54542 42.2 6.5 42.2 Interleukin 6 U14332 5.3 3.5 Interleukin 15 M57999 2.8 1.3 7.0 1.5 Nuclear factor of kappa light chain in B-cells (NF-kB p105) U88909 1.6 1.5 4.3 Inhibitor of apoptosis protein 2 (c-IAP2)

Genes and Immunity IRF-3 target genes J Andersen et al 173 Table 4 IRF-3 augmented direct response genes

Genbank WT 12 h KO 12 h WT 6 h KO 6 h Description Change Change Change Change

Innate/acquired immunity AF065947 2195.0 64.0 675.6 42.2 CCL5 chemokine/RANTES M33266 337.8 13.9 78.8 19.7 CXCL10 chemokine/IP10 M69069 84.4 10.6 3.2 2.3 Class I histocompatability gene H2-Kb-a

IFN-stimulated genes AJ007970 955.4 16.0 238.9 18.4 Guanylate nucleotide-binding protein 2 (GBP2) X56602 675.6 8.0 157.6 2.6 IFN-stimulated gene 15 (ISG15) ubiquitin cross-reactive protein U43084 294.1 59.7 1024.0 84.4 IFN-stimulated gene 56 (ISG56) with tetratricopeptide repeats U43086 181.0 2.6 97.0 5.3 IFN-stimulated gene 60 (ISG60) with tetratricopeptide repeats M74123 55.7 7.5 13.9 1.7 IFN-inducible p200 family protein (p203) U53219 26.0 3.2 39.4 IFN-induced GTPase

factor. Several IRF-3 augmented genes are listed separately in Table 4. Included in this class is the gene encoding the CCL5 chemokine (RANTES). CCL5 is known to be induced directly by activated IRF-3, and also independently by the activated NF-kB transcription factor.44 Other augmented genes include a number of IFN-stimulated genes such as ISG15, in which either activated IRF-3 or activated STAT complexes are suffi- cient to induce expression.21,23 It is not uncommon for a gene to be regulated independently by distinct trans- cription factors. By comparative analysis of RNA expression in the presence or absence of IRF-3 we have identified IRF-3 augmented direct response genes.

Profiling IRF-3 target genes Figure 2 Schematic illustration of the activation of cellular transcription factors such as IRF-3 and NF-kB in response to viral IRF-3 plays a significant role in the survival of viral infection and their subsequent direct regulation of gene subsets. infections, and to understand its action we have Direct response genes were identified during infection of WT and identified direct target genes by analyzing IRF-3-knock- KO IRF-3 cells, and secondary comparisons identified IRF-3- out cells with microarray technology. This approach dependent or IRF-3 independent genes. The IRF-3-dependent genes allows for global profiling of thousands of genes at once include those that respond directly to IRF-3 (sufficient) and those rather than analysis of one individual gene. Since the that require IRF-3 in concert with other transcription factors for expression (cooperative). The IRF-3 independent genes include microarray is not as sensitive as RT-PCR, the lack of those that are not regulated by IRF-3, and those that are regulated signal for any given gene by microarray analysis should either by IRF-3 or by other activated transcription factors not be confirmation of a negative result until other (augmented). The murine ISG54 gene exemplifies an IRF-3- methods are used to confirm the analysis. The power dependent gene. Activated IRF-3 is sufficient to induce ISG54 and of performing primary and secondary comparisons of binds to an IFN-stimulated response element (ISRE) in the promoter 45 genes that are regulated in response to virus infection is of the gene (À88 to À114). Expression of IFN-b is also dependent on IRF-3, but IRF-3 is not sufficient for its induction. Multiple DNA- the ability to verify expression changes as well as to binding factors cooperate in an enhanceosome to induce the murine distinguish subsets of genes regulated via distinct IFN-b gene, including ATF-2/c-Jun, IRF-3 and NF-kB(À51 to mechanisms. À96).46 Activated IRF-3 is sufficient to induce expression of the On the basis of information that is available for a CCL5/RANTES gene by binding to the ISRE within its promoter few IRF-3-responsive genes, it is clear that IRF-3 can (À133 to À150).47 However, CCL5 can also be induced indepen- 47 contribute to gene expression in different ways (Figure 2). dently during infection by activated NF-kB. For this reason CCL5 is considered an IRF-3 independent but augmented gene. An Following activation by serine phosphorylation IRF-3 example of an IRF-3 independent gene that cannot be induced by associates with CREB-binding protein/p300 and this activated IRF-3 is IRF-1. The promoter of murine IRF-1 has a complex can directly bind a DNA sequence leading to well-characterized IFN-g activated site (GAS) (À150 to À160) and induction of the responsive gene. ISG54 is an example of several putative NF-kB sites (consensus À297 to À314). a gene in which IRF-3 activation is sufficient to regulate gene expression. Activated IRF-3/CREB-binding protein can directly bind to the promoter of ISG54, and the IFN-b expression, but IRF-3 is not sufficient for its constitutively active IRF-3 mutant is sufficient to induce transcriptional induction. the ISG54 gene.48 This is distinct from the contribution of The IRF-3 augmented genes induced in response to IRF-3 to the expression of the IFN-b gene. The promoter virus include those whose expression is enhanced by region of IFN-b has been shown to bind several DNA- IRF-3 activation. From our knowledge of the promoter of binding transcription factors that cooperate in an one of these genes, CCL5, we propose that these genes enhanceosome complex including IRF-3, NF-kB and can be induced by IRF-3, but they can also be induced AP-1 (c-jun:ATF2).49–51 In this case, IRF-3 is required for independently by other virally activated transcription

Genes and Immunity IRF-3 target genes J Andersen et al 174 factors. IRF-3 and distinct transcription factors may act 12 Fitzgerald KA, McWhirter SM, Faia KL, Rowe DC, Latz E, additively or collaboratively to induce expression of Golenbock DT et al. IKKepsilon and TBK1 are essential augmented genes. It is not unexpected to find redun- components of the IRF3 signaling pathway. Nat Immunol dancy and collaboration in pathways that activate critical 2003; 4: 491–496. defense responses to viral infection. The approach of 13 Sharma S, tenOever BR, Grandvaux N, Zhou GP, Lin R, global gene profiling in virally infected cells that lack Hiscott J. Triggering the interferon antiviral response through IRF-3 in comparison to cells that express IRF-3 has an IKK-related pathway. Science 2003; 300: 1148–1151. revealed distinct modes of gene regulation otherwise not 14 Kumar KP, McBride KM, Weaver BK, Dingwall C, Reich NC. Regulated nuclear-cytoplasmic localization of interferon apparent, and in addition has provided a better under- regulatory factor 3, a subunit of double-stranded RNA- standing of the role of IRF-3 in innate immunity. activated factor 1. Mol Cell Biol 2000; 20: 4159–4168. 15 Daly C, Reich NC. Double-stranded RNA activates novel factors that bind to the interferon-stimulated response element. Mol Cell Biol 1993; 13: 3756–3764. Acknowledgements 16 Stark GR, Kerr IM, Williams BR, Silverman RH, Schreiber RD. How cells respond to ? Annu Rev Biochem 1998; 67: We thank all the members of the laboratory for support 227–264. and discussions. We gratefully acknowledge Dr Tadasugu 17 Levy DE, Darnell Jr JE. Stats: transcriptional control and Taniguchi (University of Tokyo) for providing us with biological impact. Nat Rev Mol Cell Biol 2002; 3: 651–662. cells from WT and IRF-3 KO animals. We also thank John 18 Sen GC. Viruses and interferons. Annu Rev Microbiol 2001; 55: Schwedes at the Affymetrix core facility for his role in 255–281. processing and analyzing the microarrays, and Michelle 19 Stetson DB, Medzhitov R. Type I interferons in host defense. Immunity 2006; 25: 373–381. Massoth for help with Genesifter Software (www.gene 20 Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, sifter.net). We are grateful to Dr Herb Lewis and Robert Thanos D. Ordered recruitment of modifying and Andersen for writing a Visual Basic code used for general transcription factors to the IFN-beta promoter. Cell secondary comparisons. Human IFN-aA was a kind 2000; 103: 667–678. gift of Hoffman-LaRoche, NJ, USA). These studies were 21 Grandvaux N, Servant MJ, tenOever B, Sen GC, Balachandran supported by grants from NIH (R21AI067885 and S, Barber GN et al. Transcriptional profiling of interferon PO1AI0555621). regulatory factor 3 target genes: direct involvement in the regulation of interferon-stimulated genes. J Virol 2002; 76: 5532–5539. 22 Elco CP, Guenther JM, Williams BR, Sen GC. Analysis of genes References induced by Sendai virus infection of mutant cell lines reveals essential roles of interferon regulatory factor 3, NF-kappaB, 1 Honda K, Taniguchi T. IRFs: master regulators of signalling by and interferon but not toll-like receptor 3. J Virol 2005; 79: Toll-like receptors and cytosolic pattern-recognition receptors. 3920–3929. Nat Rev 2006; 6: 644–658. 23 Daly C, Reich NC. Characterization of specific DNA-binding 2 Kawai T, Akira S. Innate immune recognition of viral factors activated by double-stranded RNA as positive infection. Nat Immunol 2006; 7: 131–137. regulators of interferon alpha/beta-stimulated genes. J Biol 3 Hiscott J. Triggering the innate antiviral response through Chem 1995; 270: 23739–23746. IRF-3 activation. J Biol Chem 2007; 282: 15325–15329. 24 Terenzi F, deVeer MJ, Ying H, Restifo NP, Williams BR, 4 Weaver BK, Kumar KP, Reich NC. Interferon regulatory factor Silverman RH. The antiviral enzymes PKR and RNase L 3 and CREB-binding protein/p300 are subunits of double- suppress gene expression from viral and non-viral based stranded RNA-activated transcription factor DRAF1. Mol Cell vectors. Nucleic Acids Res 1999; 27: 4369–4375. Biol 1998; 18: 1359–1368. 25 Marie I, Durbin JE, Levy DE. Differential viral induction of 5 Yoneyama M, Suhara W, Fukuhara Y, Fukuda M, Nishida E, distinct interferon-alpha genes by positive feedback through Fujita T. Direct triggering of the type I interferon system by interferon regulatory factor-7. EMBO J 1998; 17: 6660–6669. virus infection: activation of a transcription factor complex 26 Moser B, Wolf M, Walz A, Loetscher P. Chemokines: multiple containing IRF-3 and CBP/p300. EMBO J 1998; 17: 1087–1095. levels of leukocyte migration control. Trends Immunol 2004; 25: 6 Wathelet MG, Lin CH, Parekh BS, Ronco LV, Howley PM, 75–84. Maniatis T. Virus infection induces the assembly of coordi- 27 Arenzana-Seisdedos F, Virelizier JL, Rousset D, Clark-Lewis I, nately activated transcription factors on the IFN-beta Loetscher P, Moser B et al. HIV blocked by chemokine in vivo. Mol Cell 1998; 1: 507–518. antagonist. Nature 1996; 383: 400. 7 Sato M, Tanaka N, Hata N, Oda E, Taniguchi T. Involvement of 28 Starling GC, Whitney GS, Siadak AW, Llewellyn MB, Bowen the IRF family transcription factor IRF-3 in virus-induced MA, Farr AG et al. Characterization of mouse CD6 with novel activation of the IFN-beta gene. FEBS Lett 1998; 425: 112–116. monoclonal antibodies, which enhance the allogeneic mixed 8 Lin R, Heylbroeck C, Pitha PM, Hiscott J. Virus-dependent leukocyte reaction. Eur J Immunol 1996; 26: 738–746. phosphorylation of the IRF-3 transcription factor regulates 29 Chien YH, Jores R, Crowley MP. Recognition by gamma/delta nuclear translocation, transactivation potential, and protea- T cells. Annu Rev Immunol 1996; 14: 511–532. some-mediated degradation. Mol Cell Biol 1998; 18: 2986–2996. 30 Kambayashi T, Kraft-Leavy JR, Dauner JG, Sullivan BA, 9 Sato M, Suemori H, Hata N, Asagiri M, Ogasawara K, Laur O, Jensen PE. The nonclassical MHC class I molecule Nakao K et al. Distinct and essential roles of transcription Qa-1 forms unstable peptide complexes. J Immunol 2004; 172: factors IRF-3 and IRF-7 in response to viruses for IFN-alpha/ 1661–1669. beta gene induction. Immunity 2000; 13: 539–548. 31 Wang X, Johansen LM, Tae HJ, Taparowsky EJ. IFP 35 forms 10 Honda K, Yanai H, Negishi H, Asagiri M, Sato M, Mizutani T complexes with B-ATF, a member of the AP1 family of et al. IRF-7 is the master regulator of type-I interferon- transcription factors. Biochem Biophys Res Commun 1996; 229: dependent immune responses. Nature 2005; 434: 772–777. 316–322. 11 Honda K, Takaoka A, Taniguchi T. Type I interferon 32 Weichenhan D, Kunze B, Zacker S, Traut W, Winking H. [corrected] gene induction by the interferon regulatory factor Structure and expression of the murine Sp100 nuclear dot family of transcription factors. Immunity 2006; 25: 349–360. gene. Genomics 1997; 43: 298–306.

Genes and Immunity IRF-3 target genes J Andersen et al 175 33 Espert L, Degols G, Lin YL, Vincent T, Benkirane M, Mechti N. 42 Shiloh Y. ATM and related protein kinases: safeguarding Interferon-induced exonuclease ISG20 exhibits an antiviral genome integrity. Nat Rev Cancer 2003; 3: 155–168. activity against human immunodeficiency virus type 1. J Gen 43 Monaco JJ, Nandi D. The genetics of proteasomes and antigen Virol 2005; 86 (Part 8): 2221–2229. processing. Annu Rev Genet 1995; 29: 729–754. 34 Terenzi F, Pal S, Sen GC. Induction and mode of action of the 44 Genin P, Algarte M, Roof P, Lin R, Hiscott J. Regulation of viral stress-inducible murine proteins, P56 and P54. Virology RANTES chemokine gene expression requires cooperativity 2005; 340: 116–124. between NF-kappa B and IFN-regulatory factor transcription 35 Chin KC, Cresswell P. Viperin (cig5), an IFN-inducible factors. J Immunol 2000; 164: 5352–5361. antiviral protein directly induced by human cytomegalovirus. 45 Bluyssen HA, Vlietstra RJ, Faber PW, Smit EM, Hagemeijer A, Proc Natl Acad Sci USA 2001; 98: 15125–15130. Trapman J. Structure, localization, and regu- 36 MacMicking JD, Taylor GA, McKinney JD. Immune control of lation of expression of the interferon-regulated mouse Ifi54/ tuberculosis by IFN-gamma-inducible LRG-47. Science 2003; Ifi56 gene family. Genomics 1994; 24: 137–148. 302: 654–659. 46 Honda K, Yanai H, Takaoka A, Taniguchi T. Regulation of the 37 Zhu H, Cong JP, Shenk T. Use of differential display analysis type I IFN induction: a current view. Int Immunol 2005; 17: to assess the effect of human cytomegalovirus infection on the 1367–1378. accumulation of cellular RNAs: induction of inter- 47 Lee AH, Hong JH, Seo YS. Tumour necrosis factor-alpha feron-responsive RNAs. Proc Natl Acad Sci USA 1997; 94: and interferon-gamma synergistically activate the RANTES 13985–13990. promoter through nuclear factor kappaB and interferon 38 Spengler D, Villalba M, Hoffmann A, Pantaloni C, Houssami regulatory factor 1 (IRF-1) transcription factors. Biochem S, Bockaert J et al. Regulation of apoptosis and cell cycle arrest J 2000; 350 (Part 1): 131–138. by Zac1, a novel protein expressed in the pituitary 48 Weaver BK, Ando O, Kumar KP, Reich NC. Apoptosis is gland and the brain. EMBO J 1997; 16: 2814–2825. promoted by the dsRNA-activated factor (DRAF1) during 39 Unoki M, Nakamura Y. EGR2 induces apoptosis in various viral infection independent of the action of interferon or p53. cancer cell lines by direct transactivation of BNIP3L and BAK. FASEB J 2001; 15: 501–515. Oncogene 2003; 22: 2172–2185. 49 Taniguchi T, Takaoka A. The interferon-alpha/beta system in 40 Calfon M, Zeng H, Urano F, Till JH, Hubbard SR, Harding HP antiviral responses: a multimodal machinery of gene regu- et al. IRE1 couples endoplasmic reticulum load to secre- lation by the IRF family of transcription factors. Curr Opin tory capacity by processing the XBP-1 mRNA. Nature 2002; 415: Immunol 2002; 14: 111–116. 92–96. 50 Lomvardas S, Thanos D. Modifying gene expression programs 41 Trubia M, Sessa L, Taramelli R. Mammalian Rh/T2/ by altering core promoter chromatin architecture. Cell 2002; S-glycoprotein ribonuclease family genes: cloning of a human 110: 261–271. member located in a region of chromosome 6 (6q27) 51 Maniatis T, Falvo JV, Kim TH, Kim TK, Lin CH, Parekh BS frequently deleted in human malignancies. Genomics 1997; et al. Structure and function of the interferon-beta enhanceo- 42: 342–344. some. Cold Spring Harb Symp Quant Biol 1998; 63: 609–620.

Supplementary Information accompanies the paper on Genes and Immunity website (http://www.nature.com/gene)

Genes and Immunity