Expression of the Immune Regulator Tripartite-Motif 21 Is Controlled by IFN Regulatory Factors

This information is current as Maria Sjöstrand, Aurélie Ambrosi, Susanna Brauner, Jenna of September 27, 2021. Sullivan, Stephen Malin, Vijay K. Kuchroo, Alexander Espinosa and Marie Wahren-Herlenius J Immunol 2013; 191:3753-3763; Prepublished online 23 August 2013;

doi: 10.4049/jimmunol.1202341 Downloaded from http://www.jimmunol.org/content/191/7/3753

References This article cites 63 articles, 24 of which you can access for free at: http://www.jimmunol.org/content/191/7/3753.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 by guest on September 27, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Expression of the Immune Regulator Tripartite-Motif 21 Is Controlled by IFN Regulatory Factors

Maria Sjo¨strand,*,1 Aure´lie Ambrosi,*,1 Susanna Brauner,* Jenna Sullivan,† Stephen Malin,‡ Vijay K. Kuchroo,† Alexander Espinosa,*,x and Marie Wahren-Herlenius*

Tripartite-motif 21 (TRIM21) is an E3 ligase that regulates innate immune responses by ubiquitinating IFN regulatory factors (IRFs). TRIM21 is mainly found in hematopoietic cells in which its expression is induced by IFNs during viral. infections and in systemic autoimmune diseases such as systemic erythematosus and Sjo¨gren’s syndrome. However, the exact molec- ular mechanism by which the expression of the Trim21 is regulated is unknown. In this study, we demonstrate that IFNs induce Trim21 expression in immune cells via IRFs and that IFN-a and IFN-b are the most potent inducers of Trim21.A functional IFN-stimulated response element but no conserved IFN-g–activated site was detected in the promoter of Trim21. IRF1 and IRF2 strongly induced Trim21 expression in an IFN-stimulated response element–dependent manner, whereas IRF4 and Downloaded from IRF8 strongly repressed the IRF1-mediated induction of Trim21. Consistent with this observation, baseline expression of Trim21 was elevated in Irf42/2 cells. TRIM21, IRF1,andIRF2 expressionwasincreasedinPBMCsfrompatientswithSjo¨gren’s syndrome compared with healthy controls. In contrast, IRF4 and IRF8 expression was not increased in PBMCs from patients. The IFN-g–mediated induction of Trim21 was completely abolished by inhibiting synthesis with cycloheximide, and Trim21 expression could not be induced by IFN-g in Irf12/2 cells, demonstrating that IFN-g induces Trim21 indirectly via IRF1 and not directly via STAT1 activation. Our data demonstrate that multiple IRFs tightly regulate expression of Trim21 in immune cells, http://www.jimmunol.org/ suggesting that a well-controlled expression of the E3 ligase TRIM21 is important for regulation of immune responses. The Journal of Immunology, 2013, 191: 3753–3763.

ripartite-motif 21 (TRIM21) is a member of the tripartite- regulators of innate immune responses or antiviral defense, in- motif protein family characterized by three N-terminal cluding TRIM5, TRIM22, and TRIM25 (6‑14). T domains: RING, B-box, and coiled-coil (1). The RING TRIM21 is an E3 ligase that ubiquitinates several IFN regulatory domain is critical for the E3 activity of TRIMs (2, factors (IRFs), a family of transcription factors activated by type I 3), whereas the B-box and coiled-coil domains are important for IFN or pathogen-associated molecular pattern induced signaling. homo- or heteromultimerization (4, 5). Many TRIM are TRIM21 has been reported to ubiquitinate IRF3, IRF5, IRF7, and by guest on September 27, 2021 IRF8 (9, 15–18), thus affecting the expression of IRF target such as IFNb, IL-6, and IL-12/IL-23 p40. Two independent studies *Unit of Rheumatology, Department of Medicine, Karolinska Institutet, SE-171 76 2/2 Stockholm, Sweden; †Center for Neurologic Diseases, Brigham and Women’s Hos- demonstrate that Trim21 cells overexpress proinflammatory 2 2 pital, Harvard Medical School, Boston, MA 02115; ‡Department of Medicine, Unit of after TLR activation in vitro (8, 9). In addition, Trim21 / Experimental Cardiovascular Research, Center for Molecular Medicine, Karolinska x mice are hypersensitive to innate immune activation (9), confirming Institutet, Stockholm, SE-171 76 Sweden; and Department of Medicine, Weill Cor- nell Medical College, New York, NY 10065 that TRIM21 is an important regulator of immune responses in vivo. 1M.S. and A.A. contributed equally. TRIM21 has recently also been described as an important negative Received for publication August 21, 2012. Accepted for publication July 29, 2013. regulator of the helicase DDX41 (19). DDX41 is a sensor of cyto- solic DNA and a critical trigger of STING-dependent type I IFN This work was supported by the Karolinska Institutet, the Swedish Research Council, the Go¨ran Gustafsson Foundation, the Torsten and Ragnar So¨derberg Foundation, the responses. TRIM21 targets DDX41 for proteasomal degradation via King Gustaf the Vth 80-Year Foundation, the Swedish Foundation for Strategic K48 ubiquitination; consequently, viral titers of HSV-1 are strongly Research, the Heart-Lung Foundation, the Tore Nilsson Foundation, the Magnus 2/2 Bergwall Foundation, the Lars Hiertas Minne Foundation, the Swedish Rheumatism reduced in HSV-1–infected Trim21 mice compared with wild- Association, the Center of Excellence for Research on Inflammation and Cardiovas- type mice because of the increased IFN-b production mediated by cular Disease, and Temacentrum fo¨r Inflammationssjukdomar, Karolinska Institutet. the DDX41/STING pathway (19). The microarray data presented in this article have been submitted to the Gene Ex- A large number of TRIM genes are induced by type I or type II pression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE48378. IFNs (e.g., during viral or bacterial infections) (20, 21). Interest- ingly, some TRIM genes are also overexpressed in systemic auto- Address correspondence and reprint requests to Dr. Alexander Espinosa, Department of Medicine, E9, Weill Cornell Medical College, 1300 York Avenue, New York, NY immune diseases such as systemic lupus erythematosus and 10065 or Unit of Experimental Rheumatology, Department of Medicine, Center for Sjo¨gren’s syndrome, suggesting that TRIM proteins participate in Molecular Medicine, CMM L8:04, Karolinska Institutet, SE-171 76 Stockholm, Sweden. E-mail addresses: [email protected] or [email protected] the pathogenesis of these diseases (22–24). Trim21 is induced by both type I and type II IFNs and is overexpressed in patients with Abbreviations used in this article: BMDM, bone marrow–derived macrophage; ChIP- Seq, chromatin immunoprecipitation sequencing; CHX, cycloheximide; GAF, g ac- systemic lupus erythematosus and Sjo¨gren’s syndrome compared tivated factor; GAS, IFN-g–activated site; IFNAR, IFN-a/b receptor; IFNGR, IFN-g with healthy controls (23, 24). The molecular mechanisms regu- receptor; IRF, IFN regulatory factor; IRF8DDBD, IFN regulatory factor 8 plasmid construct lacking the DNA-binding domain; ISG, IFN-stimulated gene; ISGF3, IFN- lating Trim21 expression are unknown, but because Trim21 is in- stimulated gene factor 3; ISRE, IFN-stimulated response element; qRT-PCR, quan- duced by IFNs we hypothesized that transcription factors activated titative RT-PCR; RLU, relative luminescence unit; TRIM21, tripartite-motif 21; TSS, by IFN signaling control Trim21 expression. Type I IFNs bind to transcription start site. the heterodimeric IFN-a/b receptor (IFNAR), leading to the acti- Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 vation of the JAK-STAT pathway and phosphorylation of STAT1 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1202341 3754 REGULATION OF TRIM21 EXPRESSION and STAT2, as well as other STATs (25, 26). pSTAT1 and pSTAT2 U/ml murine IFN-a4 (PBL IFN Source), IFN-b (PBL IFN Source), and IFN- form a complex with IRF9 (IFN-stimulated gene factor 3 [ISGF3]) g (Millipore) for 24 h before flow cytometric analysis. that subsequently activates the transcription of IFN-stimulated For experiments on the kinetics of Trim21 expression, spleens were harvested from C57BL/6J, followed by mechanical dissociation. After genes (ISGs) by binding to IFN-stimulated response elements RBC lysis, single-cell suspensions were obtained by passing cells through (ISRE) (27–29). Importantly, type I IFN signaling additionally a 40-mm cell strainer. Splenocytes were grown in 24-well plates (3 3 106 leads to formation of pSTAT1 homodimers, followed by a rapid cells/well) and stimulated with murine IFN-a4, IFN-b,orIFN-g at 500 induction of IRF1 (30, 31), which in turn binds to ISRE sites in U/ml for the indicated time points, followed by RNA extraction. For analysis of dose-dependent effects on Trim21 expression, EL-4 cells the promoters of ISGs (32, 33). IFN-g signaling also induces the were grown in 24-well plates (7.5 3 105 cells/well) and maintained in the phosphorylation of STAT1 and formation of pSTAT1 homodimers same media as the splenocytes. The cells were stimulated with murine (IFN-g–activated factor [GAF]) that bind to IFN-g–activated sites IFN-a4, IFN-b, and IFN-g at indicated concentrations for 4 h. For the (GAS) in promoters and transactivate IFNg-dependent genes (34– cycloheximide (CHX) experiments, EL-4 cells were grown in 6-well plates 3 6 36). Among these IFN-g-induced genes is IRF1, which can in turn (3 10 cells/well) and were pretreated with CHX (Sigma-Aldrich) at 2.5 or 20 mg/ml for 30 min before adding IFNs at 500 U/ml. Cells were activate the transcription of ISRE-containing genes and thereby harvested after 4 h for RNA extraction. extend the range of genes induced by IFN-g (32, 33). Bone marrow–derived macrophages (BMDMs) were generated using To identify transcription factors involved in regulating the expres- L929-conditioned DMEM (Sigma-Aldrich) containing 20% 0.20-mmfil- sion of Trim21, we used chromatin immunoprecipitation sequencing tered medium from a confluent L929 culture, 10% FCS (Sigma-Aldrich), 2mML-glutamine, and 100 U/ml penicillin–streptomycin. After 7 d, the (ChIP-seq) data and phylogenetic footprinting and identified a cells were plated in 6-well plates (5 3 105 cells/well) and stimulated the conserved ISRE site just upstream of the transcription start site of next day with 1000 U IFN-a4, IFN-b, or IFN-g for 6 h, followed by RNA

Trim21. Using luciferase reporter assays, we found that IRF1 and extraction. Downloaded from IRF2 both strongly induced Trim21 expression and that IRF4 and 293T cells were maintained in DMEM (Sigma-Aldrich) supplemented with 10% FCS (Sigma-Aldrich), 2 mM L-glutamine, and 100 U/ml peni- IRF8 were potent repressors of Trim21 expression. We verified the 5 cillin–streptomycin and grown in 24-well plates (1.5 3 10 cells/well). physical binding of IRFs to the Trim21 ISRE site by DNA pulldown. The importance of IRFs in controlling Trim21 expression in vivo Patients and analysis 2/2 was confirmed using macrophages and T cells from Irf1 and PBMCs from 11 female Ro/SSA autoantibody-positive and untreated patients fl/fl fl/fl T cells from Irf4 ;CD4-Cre and Irf8 ;CD4-Cre mice. Further- with primary Sjo¨gren’s syndrome classified according to the revised Euro- http://www.jimmunol.org/ more, IRF1 and IRF2, but not IRF4 and IRF8, had an increased pean–American consensus criteria and 17 age- and gender-matched healthy expression together with TRIM21 in patients with Sjo¨gren’s syn- controls without any immunological disease were collected. Total RNA was prepared using the Qiagen RNeasy Plus Mini Kit, and a gene expression drome and an IFN signature compared with controls. In all, our data array was performed using the Human Exon 1.0 ST chip from Affymetrix. provide evidence that the expression of the immune regulatory gene CEL files were preprocessed using the robust multi-array average algorithm Trim21 is stringently regulated by IRFs. as implemented in Affymetrix Power Tools-1.12.0. The IFN score was cal- culated as described previously (37). Microarray data are deposited at Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) accession number Materials and Methods GSE48378. Briefly, using the five top IFN-regulated genes with the lowest p Animals value from the gene expression array IFIT3, EPSTI1, OAS3, PELI1, and IFI6, a standardized expression level of each gene was summarized for each by guest on September 27, 2021 +/GFP The generation of C57BL/6-Trim21 reporter mice has been described study subject, and a cutoff at 10 was used for defining presence or absence of 2/2 fl/fl fl/fl elsewhere (9). Irf1 , Irf4 , Irf8 , and CD4-Cre mice were purchased an IFN signature. from The Jackson Laboratory. The Irf4 and Irf8 allele were inactivated in The study was approved by the Regional Ethical Ethics Committee, fl/fl fl/fl T cells by crossing Irf4 and Irf8 mice to CD4-Cre mice generating Stockholm, Karolinska Institute, and all subjects gave informed written fl/fl fl/fl Irf4 ;CD4-Cre mice and Irf8 ;CD4-Cre mice, respectively. Mice were consent. kept in specific pathogen-free conditions at the animal facility of Center for Molecular Medicine at Karolinska Institutet or at the animal facility at Quantitative RT-PCR the Harvard Institutes of Medicine. The study was approved by the Ethical Review Committee North, Stockholm County and Standing Committee on RNA from BMDMs and sorted naive T cells was isolated using TRIzol Animals at Harvard Medical School. (Invitrogen) according to the manufacturer’s instructions. To enhance RNA precipitation, 10 mg glycogen (Fermentas) was added. RNA from other cell Plasmids types was purified using the RNeasy kit (Qiagen). The following primers were used for SYBR quantitative RT-PCR (qRT-PCR): Hprt,59-CTCATGGACT- Trim21 promoter fragments were amplified by PCR using the bacterial GATTATGGACAGGAC-39 (forward) and 59-GCAGGTCAGCAAAGA- artificial RP23-311N6 as a template (BACPAC Resources ACTTATAGCC-39 (reverse); Trim21,59-CTGCAGGAGCTGATCTCAG-39 Center, Oakland, CA). The PCR products were inserted into pCR-XL- and 59-TGTCCTCAGCATCTTCTTC-39 (reverse); and Irf2,59-CAGTT- TOPO (Invitrogen), sequenced, and transferred to the pGL4.14 luciferase GAGTCATCTTTGGGGC-39 (forward) and 59-TGGTCATCACTCTCAGT- plasmid (Promega). The coding cDNA sequences of IRFs and PU.1 were GG-39 (reverse). The DDCt method adjusted for amplification efficiency (Bio- amplified by RT-PCR from mouse splenocyte RNA and inserted into pCR- Rad) was used to calculate expression relative to Hprt. XL-TOPO (Invitrogen), sequenced, and transferred to the FLAG-CMV6c plasmid (Sigma-Aldrich). pTK-Renilla (Promega) was used as a normali- Reporter assays zation plasmid for reporter assays. ISRE-mutated luciferase reporter con- 293T cells were transfected with luciferase reporter plasmid, TK-Renilla structs were generated by PCR using long primers containing the mutations. normalization plasmid, and an IRF, PU.1, or E47 encoding plasmid us- pcDNA3-E47 was obtained from Addgene (plasmid 16059). Plasmid ing FugeneHD or X-treme Gene (Roche) and incubated for 24 h. The Dual- details and primer sequences are available upon request. Luciferase Reporter Assay System (Promega) was used to quantify lumi- nescence from transfected cells using a Modulus Single Tube Reader Cell culture and IFN stimulations (Promega). +/GFP For analysis of GFP fluorescence in Trim21 cells by flow cytometry, DNA pulldown spleens were collected from C57BL/6-Trim21+/GFP reporter mice, perfused with 400 U/ml collagenase D (Roche) in HBSS, and incubated for 45 min EL-4 cells were stimulated with 1000 U/ml IFN-g for 6 h, and nuclear ex- at 37˚C, followed by mechanical dissociation. Single-cell suspensions tracts were prepared with NE-PER reagent (Pierce), according to the man- were obtained after RBC lysis by passing cells through a 70-mm cell ufacturer’s instructions using protease and phosphatase inhibitors (Thermo strainer. Splenocytes were grown in 24-well plates (5 3 106 cells/well) Scientific). Complementary oligonucleotides spanning the Trim21 ISRE site in RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% FCS were synthesized by Eurofins MWG Operon. For each oligo pair, the sense (Sigma-Aldrich), 2 mM L-glutamine, 100 U/ml penicillin–streptomycin, oligo was biotinylated on the 59-end. To generate dsDNA probes, oligos 20 mM HEPES, and 50 mM 2-ME. The cells were stimulated with 1000 were annealed by incubating for 3 min at 95˚C, followed by decreasing The Journal of Immunology 3755 temperature: 95˚C→15˚C (D˚C = 21˚C/min). The following sequences Immunoblotting were as follows: ISRE-oligo sequence (sense), 59-Biotin-TGCCTCTGA- GAGGAAACTGAAAGTGAAATTTGGACTGCAG-39; mutated ISRE oligo Proteins were separated on a 10% SDS-PAGE gel and transferred onto sequence (sense), 59-Biotin-TGCCTCTGAGAGGGGACTGGGAGTGGGAT- a nitrocellulose membrane. The membranes were blocked in 5% nonfat milk TTGGACTGCAG-39; and scrambled oligo sequence (sense), 59-Biotin-ATA- in 0.05% TBS-Tween before incubating with a 1:200 dilution of anti-IRF1 ACGGAGGTGCGATATCAACGGCAATTCTGAATAGTGG-39. For the (Santa Cruz Biotechnology) or anti-STAT1 (Santa Cruz Biotechnology) binding reaction 3 mg biotinylated probe was mixed with 7 mgnuclearex- Ab, followed by an anti-rabbit HRP-conjugated Ab (DakoCytomation). tract, 3 mg salmon sperm DNA (Sigma-Aldrich), and 3 mgBSAinafinal For anti-FLAG immunoblotting membranes were blocked in 5% nonfat volume of 15 ml in binding buffer (12% glycerol, 12 mM HEPES, 4 mM Tris milk in 0.05% TBS-Tween and incubated with anti-FLAG (Sigma-Aldrich) [pH 8], 60 mM KCl, 1 mM EDTA, and 1 mM DTT). After incubating at diluted 1:3000. This was followed by incubation with anti-rabbit HRP- room temperature for 15 min, 30 ml streptavidin beads (1 mg/ml) resus- conjugated Ab (DakoCytomation) diluted 1:2,000 and with anti–b-actin- pended in 5 ml binding buffer was added (Dynabeads; Invitrogen), and the HRP (Sigma Aldrich) diluted 1:50,000. reaction was further incubated for 10 min. After three washing steps with Statistical analysis binding buffer, proteins bound to the probe were eluted by resuspending in 25 ml SDS-PAGE sample buffer containing 2-ME and incubating at Data were analyzed by Student t test using GraphPad Prism version 5.01 95˚C for 5 min. for Macintosh (GraphPad Software, San Diego, CA; www.graphpad.com). A p value , 0.05 was considered significant. ChIP-seq Previously published ChIP-Seq reads for CTCF and RAD21 (38) and Results H3K4me3, H3K4me2, and H3K4me1 (39) were displayed in the Integrated IFN-a and IFN-b are potent inducers of Trim21 in a variety of Genome Browser (40). immune cell populations Downloaded from Flow cytometry and cell sorting It has previously been reported that Trim21 expression is induced Splenocytes were incubated with mouse Fc block (BD Biosciences) for by both type I and type II IFNs in a variety of cells (20, 21, 41, 5 min at 4˚C and then stained for 20 min at 4˚C with the following 42). We confirmed and extended these data by using a mouse fluorescent Ab combinations: CD8-PE (BD Biosciences), CD4-PerCP, T cell line and observed that Trim21 expression is indeed induced NK1.1-allophycocyanin (BioLegend), and CD3-allophycocyanin-Cy7; by IFNs in a dose-dependent manner and that type I IFNs (espe- GR1-PE (BD Biosciences), CD11b-PerCP, and F4/80-allophycocyanin cially IFN-b) are more potent at inducing Trim21 transcription (BioLegend); or CD19-PE, B220-PerCP, and CD11c-allophycocyanin http://www.jimmunol.org/ (BD Biosciences). The different cell populations were defined as fol- than IFN-g (Fig. 1A). To verify that the induction of Trim21 ex- lows: T, CD3+;B,B220+CD19+;NK,CD32NK1.1+; dendritic cell, pression by IFNs is not an artifact because of the use of tumor cell CD11c+; M, CD11b+F4/80+; and GR, GR-1+. Irf12/2, Irf42/2 Irf82/2, + lines or extensive culture of primary cells in vitro, we treated and wild-type CD4 T cells were purified from pooled spleen and lymph freshly isolated mouse splenocytes with IFNs and analyzed Trim21 node cell suspensions using CD4-positive MACS selection (Miltenyi Biotec), and naive CD4+CD62LhighCD44low T cells were subsequently mRNA levels over time. Both type I and type II IFNs induced sorted by FACS. Trim21 expression in splenocytes. However, compared with IFN-g, by guest on September 27, 2021

FIGURE 1. Trim21 expression is induced by type I and type II IFNs. (A) EL-4 cells were treated with 100, 500, and 1000 U/ml IFNs for 4 h and Trim21 mRNA levels were determined by qRT-PCR and normalized to Hprt. The graph is representative of two independent experiments. (B) Freshly isolated mouse splenocytes from two mice were stimulated with 500 U/ml IFN in triplicates for indicated time points. Trim21 mRNA levels were determined by qRT-PCR and normalized to Hprt. The median values are shown, and error bars represent range (min-max). The graph is representative of two independent experiments. (C) Splenocytes from C57BL/6-Trim21+/GFP reporter mice were stimulated with 1000 U/ml IFN for 24 h, and the GFP fluorescence was determined by flow cytometry. Graph shows mean fluorescent intensity (MFI) values for GFP from three C57BL/6-Trim21+/GFP mice. 3756 REGULATION OF TRIM21 EXPRESSION stimulation with type I IFNs led to a more rapid and higher ex- different species, we found a highly conserved putative ISRE site pression of Trim21, with the strongest effect observed after IFN-b with three GAAA elements in the promoter of Trim21 (Fig. 2B, stimulation (Fig. 1B). The IFNAR has been reported to form 2C). To validate the functional relevance of this putative ISRE a more stable complex with IFN-b compared with IFN-a, which and to find additional important regulatory sites, we performed may explain the stronger effect of IFN-b on inducing and main- promoter-mapping experiments. Five genomic DNA fragments of taining the expression of Trim21 (43, 44). the Trim21 promoter region were cloned and inserted into a IFNs clearly induced the expression of Trim21 in splenocytes, luciferase reporter plasmid. Three of these fragments (2654 but it is possible that Trim21 expression differs between different to +1342, 21160 to +40, and 2461 to +40) included the ISRE splenic immune cell populations. To address this issue, we cul- site, whereas the other two did not (2790 to 2290 and 21160 tured splenocytes from Trim21+/GFP reporter mice, and measured to 2656). Only the three fragments containing the ISRE site were the basal expression level and induction of Trim21-GFP expres- transcriptionally active (Fig. 3A), strongly suggesting that the ISRE sion in response to IFNs by flow cytometry (Fig. 1C). Trim21 had is a critical regulatory element for the induction of Trim21 expres- a low basal expression in B cells and granulocytes compared with sion. Interestingly, we observed a higher baseline expression of T cells, NK cells, macrophages, and dendritic cells. Interestingly, the reporter construct 21160 to +40 compared with 2654 to +1342. the induction of Trim21 by IFNs also differed between these cell This suggests that a repressive transcription factor binds down- types, being lowest in granulocytes and B cells. stream of +40, or alternatively that the mRNA produced from the two reporters differ in stability. The ISRE site is a known binding The Trim21 promoter contains an ISRE site for the many IRFs that are activated or induced after IFNAR

To identify the molecular mechanisms involved in the regulation of signaling (e.g., IRF1, IRF2, and IRF9) or pattern recognition re- Downloaded from Trim21 expression, we looked for histone modifications and ge- ceptor signaling (e.g., IRF3 and IRF5). The presence of an ISRE nomic regions important for regulating the expression of Trim21. site in the Trim21 promoter thus indicates that the induction of We used ChIP-seq data from mouse pro-B cells to investigate Trim21 expression following IFN stimulation is most likely histone methylation marks around the Trim21 gene and found mediated by IRFs. H3K4 mono-, di-, and trimethylation peaks (H3K4me1, H3K4me2, Trim21 expression is induced by IRF1 and IRF2 and repressed

and H3K4me3) at the predicted transcription start site (TSS) (38, http://www.jimmunol.org/ 39), suggesting that Trim21 is a transcriptionally active gene in by IRF4 and IRF8 B cells and that the Trim21 core promoter is just upstream of the To test whether IRFs can initiate Trim21 transcription via binding predicted TSS (Fig. 2A). In addition, a putative enhancer element to the ISRE site in the Trim21 promoter, we cotransfected an IRF1 (H3K4me1 and H3K4me2 positive) was present in the 39-region of encoding plasmid with each of the five promoter constructs de- the Trim21 locus. Interestingly, CTCF insulating proteins associ- scribed in Fig. 3A. This experiment clearly showed that IRF1 ating with chromosomal organizing protein RAD21 were found could induce expression only of promoter fragments containing surrounding the Trim21 gene, suggesting that regulation of Trim21 the ISRE (Fig. 3B). Because all members of the IRF family can expression may be independent of its close ortholog Trim68. To potentially bind to ISRE sites, we investigated the ability of other identify Trim21 promoter elements, we analyzed the region be- IRFs to induce or repress Trim21 expression using luciferase re- by guest on September 27, 2021 tween 23 kb and +0.5 kb (+1 = TSS) using phylogenetic foot- porter assays (Fig. 3C). In addition to IRF1, we found that IRF2, printing. By comparing sequences of the Trim21 promoter between but no other IRFs, could transactivate the Trim21-Luc reporter

FIGURE 2. The Trim21 promoter contains an ISRE. (A) ChIP-seq data from primary naive mouse B cells revealed genomic regions regulating Trim21 expression. H3K4me2 and H3K4me3 histone marks typical of transcriptionally active regions were identified immediately upstream of exon 1 indicating an active promoter. (B and C) The region 23 kb to +0.5 kb (TSS = +1) was analyzed by phy- logenetic footprinting and a highly conserved ISRE site was identified in the promoter region of Trim21.No putative GAS site was identified. The Journal of Immunology 3757 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. IRF1 and IRF2 are potent inducers of Trim21 expression. (A) Different genomic fragments of the Trim21 promoter region were used to drive the expression of a luciferase reporter in 293T cells (n $ 3). (B) Trim21 promoter fragments were cotransfected with or without IRF1 in 293T cells leading to strong induction of the luciferase reporter only with fragments containing the ISRE (ISRE containing fragments are underlined) (n = 3). (C) IRFs were investigated for their ability to activate the ISRE site. The Trim21 promoter fragment 2654/+1342 was cotransfected with different IRFs, and expression of the luciferase reporter was measured (n = 4). (D) IRF1 and IRF2 were cotransfected with the Trim21 promoter fragment 2654/+1342 (n = 3). (E) Im- munoblotting to confirm similar expression levels of transfected IRFs in 293T cells. Equal amounts of cell extract (40 mg) were loaded in each well, and expression levels of IRFs were compared with that of b-actin on the same membrane. Because of the short half-life of IRF1 an additional well of IRF1 with 60 mg cell extract was also loaded. Statistics were calculated with Student t test and all graphs are representatives of at least three independent experiments. Error bars represent SEM. RLU, Relative luciferase unit. under these conditions. Depending on the cell type, IRF2 can this study, we clearly demonstrate that IRF1 and IRF2 do not either act as an antagonist of IRF1 or synergize with IRF1 in counteract each other in transactivating Trim21 expression (Fig. activating expression of ISRE-dependent promoters (45, 46). In 3D). Notably, the Trim21 promoter contains three repeats of the 3758 REGULATION OF TRIM21 EXPRESSION

GAAA element, whereas previous studies analyzing the simulta- To confirm the physical binding of IRF1 to the Trim21 ISRE site, neous effect of IRF1 and IRF2 used reporter constructs containing we performed DNA pull-down experiments. We treated EL-4 cells only one or two GAAA elements, suggesting that IRF1 and IRF2 with IFN-g for 6 h and prepared nuclear extract to enrich for can simultaneously bind separate single GAAA elements of the active nuclear IRF1. Equal fractions of the nuclear extract were Trim21 ISRE to facilitate the formation of active transcriptional then incubated with biotinylated probes, containing an intact complexes. By contrast, we observed that IRF4 and IRF8 signif- Trim21 ISRE site, a mutated ISRE site, or a scrambled ISRE site. icantly repressed the basal expression of Trim21 (Fig. 3C). This IRF1 was only pulled down by the intact ISRE probe and not by basal expression of Trim21 is likely due to binding of the ISRE the mutated or scrambled probes, demonstrating that IRF1 can site by endogenously expressed IRF1 and IRF2. IRF4 and IRF8 physically bind the ISRE in the Trim21 promoter (Fig. 4D). may thus repress basal expression from the Trim21 promoter by IFN-g induces Trim21 expression indirectly via IRF1, not competing with endogenous IRF1 and IRF2 for binding to the directly via STAT1 ISRE site. To confirm that the different IRF constructs were ex- pressed at a similar level after transfection, we performed immuno- IFN-g is a proinflammatory that induces a large set of blotting (Fig. 3E). genes involved in immune responses, including many TRIM genes (20). IFN-g signals through the IFN-g receptor (IFNGR) and IRF1 and IRF2 activate Trim21 expression via the triple GAAA induces the formation of pSTAT1 homodimers that enter the nu- element in the ISRE site of Trim21 cleus and activate transcription of target genes by binding to GAS To prove that the identified ISRE in the Trim21 promoter is a bona sites. Although Trim21 expression was induced by IFN-g,we

fide ISRE, we generated Trim21 reporter constructs with muta- could not find any putative GAS site in the Trim21 promoter. We Downloaded from tions in the ISRE sequence (Fig. 4A). Mutations of single GAAA hypothesized that IFN-g indirectly stimulates Trim21 expression elements did not affect the ability of IRF1 to induce expression via induction of IRF1 and subsequent binding of IRF1 to the (Fig. 4B), suggesting that the different GAAA elements are re- Trim21 ISRE site and not via direct binding of pSTAT1 homo- dundant and that IRF1 has the ability to bind to each of the in- dimers to the Trim21 promoter. To test this, we pretreated EL-4 dividual elements. By contrast, mutation of all three GAAA cells with CHX to block protein synthesis, stimulated the cells

elements of the ISRE abolished IRF1-induced expression (Fig. 4B). with type I and type II IFNs, and subsequently analyzed IRF1, http://www.jimmunol.org/ All three combinations of two simultaneously mutated GAAA STAT1, and Trim21 expression. CHX treatment completely blocked elements still allowed for IRF1-dependent expression from the the induction of nuclear IRF1 by type I and type II IFNs but had Trim21 ISRE (data not shown). Similar results were observed for only a minor effect on the levels of nuclear STAT1 (Fig. 5A). IRF2, demonstrating that the ISRE is a bona fide binding site for Importantly, addition of CHX dramatically reduced the induction IRF1 and IRF2 (Fig. 4C). of Trim21 by IFNs, suggesting that optimal induction of Trim21 is by guest on September 27, 2021

FIGURE 4. The GAAA elements in the ISRE site are redundant and are bound by IRFs. (A) Sequences of the ISRE and mutated ISRE used in (B) and (C) (mutations are indicated by stars). (B and C) IRF1 (B) and IRF2 (C) were cotransfected with either an intact Trim21 ISRE or a mutated ISRE and induction of the luciferase reporter was measured. (D) Nuclear extract was generated from EL-4 cells after 6 h of 1000 U/ml IFN-g treatment. Equal amounts (7 mg) of the nuclear extract were in- cubated with biotinylated ISRE probes (ISRE, mutated ISRE, scrambled ISRE, or no probe), followed by streptavidin bead–mediated DNA pulldown and detec- tion of IRF1 by Western blotting. All statistics were calculated with Student t test, and error bars represent SEM. All graphs are representative of at least two in- dependent experiments. RLU, Relative luciferase unit. The Journal of Immunology 3759 Downloaded from http://www.jimmunol.org/ FIGURE 5. IRF1 is necessary for optimal Trim21 expression. (A) EL-4 cells were pretreated with CHX for 30 min and stimulated by IFNs for 4 h. Nuclear extracts were prepared, and the IRF1 and STAT1 molecules were detected by Western blot analysis. Both the STAT1a (91 kDa) and STAT1b (84 kDa) isoforms were detected. (Two nonspecific bands between 120 and 150 kDa were also detected.) (B) EL-4 cells were pretreated with CHX for 30 min and stimulated by IFNs for 4 h before RNA extraction. Trim21 mRNA levels were determined by qRT-PCR and normalized to Hprt (FC, fold change over non–IFN-stimulated cells). The graph is representative of two independent experiments. (C) Wild-type and Irf12/2 naive (CD4+CD62LhighCD44low) T cells were isolated, and Trim21 mRNA levels were determined by qRT-PCR and normalized to Hprt (n = 5). (D) BMDMs from Irf1+/+ and Irf12/2 mice were stimulated with IFNs for 6 h, and Trim21 expression levels were determined by qRT-PCR. One representative experiment of two is shown, n = technical triplicates of individual wells. (E) BMDMs from wild-type mice were stimulated with IFNs for 6 h, and Irf2 expression levels were determined by qRT- PCR. One representative experiment of two is shown, n = technical triplicates. All statistical calculations were performed with Student t test, and all error by guest on September 27, 2021 bars represent SEM. independent of STAT1 but requires the presence of IRF1 (Fig. IRF8 would be able to induce Trim21 expression when combined 5B). To first confirm the critical role of IRF1 in Trim21 basal ex- with IRF1 in the presence of PU.1. However, cotransfection of pression, we sorted naive CD4+ T cells from wild-type and IRF1- IRF8 with IRF1 and PU.1 could not rescue the IRF8-mediated deficient mice. Expression of Trim21 was significantly lower in inhibition of IRF1-induced Trim21 expression(Fig.6A).Be- Irf12/2 cells compared with wild-type cells, demonstrating the cause both IRF4 and IRF8 can form activating complexes with necessary role for IRF1 in maintaining the basal expression level PU.1 and E47 (47), we also evaluated the transcriptional activating of Trim21 (Fig. 5C). To subsequently verify that IRF1 is necessary capacity of IRF4 and IRF8 in the presence of these two factors. for IFN-g–induced Trim21 expression, we treated IRF1-deficient Neither PU.1 nor E47 could rescue the IRF4- or IRF8-mediated BMDMs with IFNs and observed that IFN-g could not induce inhibition of Trim21 expression (data not shown), suggesting that Trim21 expression in the absence of IRF1, whereas the induction IRF4 and IRF8 are bona fide inhibitors of the Trim21 gene. To test of Trim21 by IFN-a and IFN-b was almost unchanged (Fig. 5D). whether the repressive effect of IRF8 was due to competitive Interestingly, IFN-g, although a potent inducer of IRF1 expres- binding to the ISRE, we generated a mutated IRF8 plasmid con- sion, was unable to induce IRF2 expression, whereas IFN-a and struct lacking the DNA-binding domain (IRF8DDBD). IRF8DDBD IFN-b strongly induced both IRF1 and IRF2 (Fig. 5A, 5E). In all, was unable to fully antagonize IRF1-mediated induction of Trim21 these data suggest that Trim21 induction by type II IFN relies expression when transfected in a 1:1 ratio with IRF1, strongly critically on IRF1, whereas Trim21 induction by type I IFN is suggesting that IRF8 blocks Trim21 expression by binding to the mediated by both IRF1 and IRF2. ISRE in the Trim21 promoter (Fig. 6B). To verify the effect of IRF4 and IRF8 on Trim21 expression in primary cells, we quantified IRF4 and IRF8 block the IRF1-mediated induction of Trim21 Trim21 expression in naive Irf42/2 or Irf82/2 CD4+ T cells by expression qRT-PCR. Trim21 expression was strongly increased in Irf42/2 Because IRF4 and IRF8 could inhibit the basal activity of the T cells compared with wild-type T cells but was unaffected in Trim21 promoter (Fig. 3C), we asked whether they could also Irf82/2 T cells (Fig. 6C), suggesting that the low levels of IRF4, but block the IRF1-mediated induction of Trim21. To answer this, we not of IRF8, in naive T cells are sufficient to repress Trim21 ex- cotransfected 293T cells with IRF1 and IRF4 or IRF8 and mea- pression. sured the induction of the Trim21 luciferase reporter. Strikingly, both IRF4 and IRF8 completely abolished the IRF1-mediated IRF and TRIM21 expression in Sjogren’s¨ syndrome induction of Trim21 expression (Fig. 6A). Because IRF8 can Sjo¨gren’s syndrome is an autoimmune disease with increased form complexes with IRF1 and PU.1, the possibility remained that levels of type 1 IFN and overexpression of type I IFN–stimulated 3760 REGULATION OF TRIM21 EXPRESSION

FIGURE 6. IRF4 and IRF8 are negative regulators of Trim21 ex- pression. (A) The Trim21 promoter fragment 2654/+1342 was cotrans- fected with IRF1 alone or in combi- nation with IRF4, IRF8, or IRF8, and PU.1 and expression of the luciferase reporter was measured. One represen- tative experiment of three is shown. (B)TheTrim21 promoter fragment 2654/+1342 was cotransfected with IRF1, IRF8, or IRF8DDBD alone and IRF1 in combination with IRF8 or IRF8DDBD, and luciferase expres- sion was measured. One representa- Downloaded from tive experiment out of three is shown. (C) Naive (CD4+CD62LhighCD44low) T cells lacking IRF4 (Irf42/2)orIRF8 (Irf82/2) were isolated and Trim21 mRNA levels were determined by qRT-PCR (n =4).(D) TRIM21, IRF1,

IRF2, IRF4,andIRF8 expression lev- http://www.jimmunol.org/ els were determined by a gene ex- pression array in PBMCs from 14 female patients with primary Sjo¨g- ren’s syndrome displaying an IFN signature (pSS) and 18 age- and gender-matched healthy controls (HC) without an IFN signature. All statis- tical calculations were performed with the Student t test, and all error bars represent SEM. by guest on September 27, 2021

genes, the so-called IFN signature (48–50). Using PBMCs from 12/IL-23 (8, 9, 16). Lack of TRIM21 leads to enhanced production patients with a verified IFN signature, we found that TRIM21 is of type I IFN and inflammatory cytokines after TLR stimuli (8, 9). indeed expressed at higher levels compared with healthy controls The expression of Trim21 itself is strongly induced by IFNs as (Fig. 6D). Consistent with the increased TRIM21 expression, both a negative feedback loop, but the molecular details of this have so IRF1 and IRF2 were also expressed at significantly higher levels far not been described previously. In this study, we characterize in patients with Sjo¨gren’s syndrome when compared with the the promoter of Trim21, and identify IRFs as transcription factors matched controls. In contrast, IRF4 and IRF8 expression did not that both induce and repress the expression of Trim21.More differ between patients and healthy controls. Taken together with precisely, we show that a functional ISRE is present in the Trim21 our mechanistic data, the data suggest that the increased levels promoter and that IRF1 and IRF2 activate Trim21 transcription, of IRF1 and IRF2 in patients with Sjo¨gren’s syndrome may lead whereas IRF4 and IRF8 repress it. to overexpression of TRIM21 and confirm the relevance of our We initially observed that Trim21 was induced by both type I observations also in the context of a human autoimmune disease. and type II IFNs, which suggested that Trim21 transcription may be induced not only by IRFs, but also directly by pSTAT1 homo- Discussion dimers that form downstream of both IFNAR and IFNGR sig- IFNs are critical for an effective immune defense against infections naling (32, 33). However, we observed that Trim21 was not and cancer. However, an uncontrolled production of IFNs can lead induced by IFN-g in Irf12/2 cells, demonstrating that IFN-g– to autoimmune disease; therefore, IFN signaling and IFN pro- mediated induction of Trim21 is strictly IRF1 dependent. A mo- duction is tightly controlled by several factors including suppressor lecular basis for this result comes from the observation that there of cytokine signaling 1–3, PIAS1–4, and Src homology region 2 is no GAS element in the Trim21 gene. Furthermore, ectopic ex- domain-containing phosphatase 2 (PTPN11) reviewed in Ref. 51. pression of either IRF1 or IRF2 strongly induced Trim21 ex- TRIM21 is a novel regulator of innate immune responses, acting pression, indicating that these factors are individually sufficient to by ubiquitinating IRF3, IRF5, IRF7, and IRF8, thus controlling induce Trim21. Interestingly, endogenous IRF2 was induced by the production of, for example, IFN-b and the p40 subunit of IL- type I IFNs but not by IFN-g, explaining why Trim21 induction by The Journal of Immunology 3761

IFN-g was completely abolished in Irf12/2 cells compared with tion. This suggests that IRF4 has a stronger affinity than IRF1 wild-type cells, whereas Trim21 induction by type I IFN was only for the GAAA elements and that IRF4 therefore can outcompete marginally affected. Furthermore, these data might also explain IRF1 in binding to the ISRE site of the Trim21 promoter when why type I IFNs induce significantly more expression of Trim21 present at a 1:1 ratio. However, IFN stimulation with increasing than IFN-g does. levels of IRF1 leads to release of the repression and induction Although ectopic expression of IRF1 or IRF2 was sufficient to of Trim21 expression. Genes containing ISRE sites with triple induce Trim21 expression, we could not detect induction of Trim21 GAAA elements, or alternatively tandem ISRE sites with double expression after transfection with IRF3, IRF5, IRF7, and IRF9 GAAA elements (e.g., Isg15, Oas2, Mx1, Ifit1, and Trim21) are plasmids. This could be because these factors are dependent on more strongly induced by IFNs than genes containing ISRE sites posttranslational modifications or complex formation for their with only one or two GAAA elements (63). We suggest that for transcriptional activity (52, 53). By contrast, ectopic expression strongly induced ISGs with multiple GAAA elements the negative of IRF1 and IRF2 is sufficient for induction of Trim21 because regulationbyIRF4andIRF8mayplayanimportantrolein functional binding to ISRE sites by these factors does not require controlling their induction and the subsequent IFN response. This posttranslational modifications (54–56). It remains possible that might be achieved for example by a higher affinity of IRF4 and IRF3, IRF5, IRF7, and IRF9 can activate Trim21 expression after IRF8 for ISRE sites in such ISGs, leading to an efficient block of proper posttranslational modifications, for example, phosphory- ISG expression. Thus, multiple GAAA elements may not only lation of IRF3 by TANK-binding kinase 1 after activation of TLR3 be an important feature for strong induction during immune re- or formation of ISGF3 complexes after type I IFN stimulation. sponses, but also for tight regulation of IFN dependent genes

Pretreatment with CHX dramatically reduced the induction of by IRFs. Downloaded from Trim21 by IFNs, even though the levels of nuclear STAT1 were In all, the data presented in this study show for the first time, to almost unchanged, indicating that Trim21 induction is indepen- our knowledge, that TRIM21, an E3 ubiquitin ligase involved in the dent of pSTAT1 homodimers and ISGF3. In contrast, because of control of IRF-mediated immune responses, is itself regulated by its short half-life, nuclear IRF1 levels were completely abolished IRFs at the transcriptional level. The induction of Trim21 expres- by pretreatment with CHX. The very short half-life of IRF1 (30– sion observed in type I IFN responses, for example, during antiviral

60 min) compared with IRF2 (∼8 h) (57, 58), and the dramatic defense and systemic autoimmune diseases, is mainly mediated http://www.jimmunol.org/ effect of CHX on Trim21 expression, indicates that IRF1 is the by IRF1 and IRF2. In contrast, induction of IRF4 and IRF8 is most important factor for Trim21 induction after stimulation with a mechanism for turning Trim21 expression off such that effector IFNs. and memory B and T cells are generated effectively. Taken to- IRF4 is a transcription factor important mainly in the adaptive gether, these data help us mechanistically connect the elevated immune system where it regulates the activation and differentiation expression of TRIM21 via IRF1 and IRF2 with the IFN signature of B and T cells (59). IRF4 is induced by T cell activation via TCR observed in systemic autoimmune diseases. signaling, and the dramatic effects of IRF4 on Trim21 expression suggests that TRIM21 plays a role in T cell biology as previously Acknowledgments suggested by Ishii et al. (60). The higher expression level of IRF4 We thank A˚ se Elfving for excellent technical assistance. We also thank by guest on September 27, 2021 in B cells compared with T cells (http://www.immgen.org/data- Drs. Ulf Klein and Keiko Ozato for generously providing Irf4fl/fl and Irf8fl/fl browser/index.html) could explain the lower basal expression and mice, respectively. weaker induction of Trim21 in B cells following IFN stimulation, because IRF4 can repress both the basal and IRF1-induced ex- Disclosures pression of Trim21. It is also possible that the low induction of The authors have no financial conflicts of interest. Trim21 in B cells and granulocytes could be due to cell-specific epigenetic modifications (e.g., H3K9me2 histone modifications) in the Trim21 locus in these two cell types (61). References Expression analysis in patients with Sjogren’s syndrome and an 1. Reymond, A., G. Meroni, A. Fantozzi, G. Merla, S. Cairo, L. Luzi, D. Riganelli, ¨ E. Zanaria, S. Messali, S. Cainarca, et al. 2001. The tripartite motif family IFN signature verified increased expression of TRIM21, IRF1, and identifies cell compartments. EMBO J. 20: 2140–2151. IRF2. In contrast to IRF1 and IRF2, the expression of IRF4 and 2. Zheng, N., P. Wang, P. D. Jeffrey, and N. P. Pavletich. 2000. Structure of a c-Cbl- UbcH7 complex: RING domain function in ubiquitin-protein ligases. Cell 102: IRF8 was not increased in PBMCs of patients with Sjo¨gren’s 533–539. syndrome. IRF1 and 2 are induced by type I IFNs but IRF4 and 8 3. Joazeiro, C. A., S. S. Wing, H. Huang, J. D. Leverson, T. Hunter, and Y. C. Liu. are not, and it is therefore possible that IRF1 and 2 increase 1999. The tyrosine kinase negative regulator c-Cbl as a RING-type, E2- dependent ubiquitin-protein ligase. Science 286: 309–312. TRIM21 expression in response to type I IFN in systemic autoim- 4. Peng, H., G. E. Begg, D. C. Schultz, J. R. Friedman, D. E. Jensen, mune disease, whereas IRF4 and 8, which are induced by other D. W. Speicher, and F. J. Rauscher, III. 2000. Reconstitution of the KRAB-KAP- factors, repress TRIM21 expression in other situations, for ex- 1 repressor complex: a model system for defining the molecular anatomy of RING-B box-coiled-coil domain-mediated protein-protein interactions. J. Mol. ample in cells stimulated with Ag or pathogen-associated mo- Biol. 295: 1139–1162. lecular patterns. Our data alsoconfirmTRIM21aspartofthe 5. Napolitano, L. M., and G. Meroni. 2012. TRIM family: pleiotropy and diver- IFN-signature, which relates to disease activity in systemic sification through homomultimer and heteromultimer formation. IUBMB Life 64: 64–71. autoimmunity. 6. Pertel, T., S. Hausmann, D. Morger, S. Zu¨ger, J. Guerra, J. Lascano, C. Reinhard, It has previously been suggested that ISGs with ISRE sites F. A. Santoni, P. D. Uchil, L. Chatel, et al. 2011. TRIM5 is an innate immune containing multiple GAAA elements are regulated differently than sensor for the retrovirus capsid lattice. Nature 472: 361–365. 7. Stremlau, M., C. M. Owens, M. J. Perron, M. Kiessling, P. Autissier, and ISGs with ISRE sites containing only one or two GAAA elements J. Sodroski. 2004. The cytoplasmic body component TRIM5a restricts HIV-1 (62). Indeed, as shown by our mutational analysis of the Trim21 infection in Old World monkeys. Nature 427: 848–853. 8. Yoshimi, R., T. H. Chang, H. Wang, T. Atsumi, H. C. Morse, III, and K. Ozato. ISRE site that contains three GAAA elements, the destruction of 2009. Gene disruption study reveals a nonredundant role for TRIM21/Ro52 in only one or two GAAA elements still leaves at least one GAAA NF-kB‑dependent cytokine expression in fibroblasts. J. Immunol. 182: 7527– element available for IRF binding and allows for induction of 7538. 9. Espinosa, A., V. Dardalhon, S. Brauner, A. Ambrosi, R. Higgs, F. J. Quintana, Trim21 expression. Cotransfection with IRF1 and IRF4 in a 1:1 M. Sjo¨strand, M. L. Eloranta, J. Nı´ Gabhann, O. Winqvist, et al. 2009. Loss of ratio resulted in a complete loss of IRF1-mediated Trim21 induc- the lupus autoantigen Ro52/Trim21 induces tissue inflammation and systemic 3762 REGULATION OF TRIM21 EXPRESSION

autoimmunity by disregulating the IL-23‑Th17 pathway. J. Exp. Med. 206: 35. Lew, D. J., T. Decker, and J. E. Darnell, Jr. 1989. Alpha interferon and g in- 1661–1671. terferon stimulate transcription of a single gene through different signal trans- 10. Barr, S. D., J. R. Smiley, and F. D. Bushman. 2008. The interferon response duction pathways. Mol. Cell. Biol. 9: 5404–5411. inhibits HIV particle production by induction of TRIM22. PLoS Pathog. 4: 36. Shuai, K., C. Schindler, V. R. Prezioso, and J. E. Darnell, Jr. 1992. Activation of e1000007. transcription by IFN-g: tyrosine phosphorylation of a 91-kD DNA binding 11. Gack, M. U., R. A. Albrecht, T. Urano, K. S. Inn, I. C. Huang, E. Carnero, protein. Science 258: 1808–1812. M. Farzan, S. Inoue, J. U. Jung, and A. Garcı´a-Sastre. 2009. Influenza A 37.Feng,X.,H.Wu,J.M.Grossman,P.Hanvivadhanakul, J. D. FitzGerald, NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral G.S.Park,X.Dong,W.Chen,M.H.Kim,H.H.Weng,etal.2006.Asso- RNA sensor RIG-I. Cell Host Microbe 5: 439–449. ciation of increased interferon-inducible gene expression with disease activity 12. Gack, M. U., Y. C. Shin, C. H. Joo, T. Urano, C. Liang, L. Sun, O. Takeuchi, and lupus nephritis in patients with systemic lupus erythematosus. Arthritis S. Akira, Z. Chen, S. Inoue, and J. U. Jung. 2007. TRIM25 RING-finger E3 Rheum. 54: 2951–2962. ‑ ubiquitin ligase is essential for RIG-I mediated antiviral activity. Nature 446: 38. Ebert, A., S. McManus, H. Tagoh, J. Medvedovic, G. Salvagiotto, 916–920. M. Novatchkova, I. Tamir, A. Sommer, M. Jaritz, and M. Busslinger. 2011. 13. Ozato, K., D. M. Shin, T. H. Chang, and H. C. Morse, III. 2008. TRIM The distal V(H) gene cluster of the Igh locus contains distinct regulatory elements family proteins and their emerging roles in innate immunity. Nat. Rev. with Pax5 transcription factor-dependent activity in pro-B cells. Immunity 34: Immunol. 8: 849–860. 175–187. 14. Nisole, S., J. P. Stoye, and A. Saı¨b. 2005. TRIM family proteins: retroviral re- 39. Lin, Y. C., S. Jhunjhunwala, C. Benner, S. Heinz, E. Welinder, R. Mansson, striction and antiviral defence. Nat. Rev. Microbiol. 3: 799–808. M. Sigvardsson, J. Hagman, C. A. Espinoza, J. Dutkowski, et al. 2010. A global 15. Kong, H. J., D. E. Anderson, C. H. Lee, M. K. Jang, T. Tamura, P. Tailor, network of transcription factors, involving E2A, EBF1 and Foxo1, that orches- H. K. Cho, J. Cheong, H. Xiong, H. C. Morse, III, and K. Ozato. 2007. Cutting trates B cell fate. Nat. Immunol. 11: 635–643. edge: autoantigen Ro52 is an interferon inducible E3 ligase that ubiquitinates 40. Nicol, J. W., G. A. Helt, S. G. Blanchard, Jr., A. Raja, and A. E. Loraine. 2009. IRF-8 and enhances cytokine expression in macrophages. J. Immunol. 179: 26– The Integrated Genome Browser: free software for distribution and exploration 30. 16. Higgs, R., J. Nı´ Gabhann, N. Ben Larbi, E. P. Breen, K. A. Fitzgerald, and of genome-scale datasets. Bioinformatics 25: 2730–2731. C. A. Jefferies. 2008. The E3 ubiquitin ligase Ro52 negatively regulates IFN-b 41.Rhodes,D.A.,G.Ihrke,A.T.Reinicke,G.Malcherek,M.Towey, production post-pathogen recognition by polyubiquitin-mediated degradation of D. A. Isenberg, and J. Trowsdale. 2002. The 52000 MW Ro/SS-A autoantigen Downloaded from IRF3. J. Immunol. 181: 1780–1786. in Sjo¨gren’s syndrome/systemic lupus erythematosus (Ro52) is an interferon- 17. Higgs, R., E. Lazzari, C. Wynne, J. Nı´ Gabhann, A. Espinosa, M. Wahren- gamma inducible tripartite motif protein associated with membrane proximal Herlenius, and C. A. Jefferies. 2010. Self protection from anti-viral responses— structures. Immunology 106: 246–256. Ro52 promotes degradation of the transcription factor IRF7 downstream of the 42. Strandberg, L., A. Ambrosi, A. Espinosa, L. Ottosson, M. L. Eloranta, W. Zhou, viral Toll-Like receptors. PLoS One 5: e11776. A. Elfving, E. Greenfield, V. K. Kuchroo, and M. Wahren-Herlenius. 2008. In- 18. Yang, K., H. X. Shi, X. Y. Liu, Y. F. Shan, B. Wei, S. Chen, and C. Wang. 2009. terferon-a induces up-regulation and nuclear translocation of the Ro52 auto- TRIM21 is essential to sustain IFN regulatory factor 3 activation during antiviral antigen as detected by a panel of novel Ro52-specific monoclonal antibodies. J.

response. J. Immunol. 182: 3782–3792. Clin. Immunol. 28: 220–231. http://www.jimmunol.org/ 19. Zhang, Z., M. Bao, N. Lu, L. Weng, B. Yuan, and Y. J. Liu. 2013. The E3 43. Abramovich, C., L. M. Shulman, E. Ratovitski, S. Harroch, M. Tovey, P. Eid, and ubiquitin ligase TRIM21 negatively regulates the innate immune response to M. Revel. 1994. Differential tyrosine phosphorylation of the IFNAR chain of the intracellular double-stranded DNA. Nat. Immunol. 14: 172–178. type I interferon receptor and of an associated surface protein in response to IFN- 20. Carthagena, L., A. Bergamaschi, J. M. Luna, A. David, P. D. Uchil, F. Margottin- a and IFN-b. EMBO J. 13: 5871–5877. Goguet, W. Mothes, U. Hazan, C. Transy, G. Pancino, and S. Nisole. 2009. 44. Croze, E., D. Russell-Harde, T. C. Wagner, H. Pu, L. M. Pfeffer, and H. D. Perez. Human TRIM gene expression in response to interferons. PLoS One 4: e4894. 1996. The human type I interferon receptor: identification of the interferon 21. Rajsbaum, R., J. P. Stoye, and A. O’Garra. 2008. Type I interferon-dependent b-specific receptor-associated phosphoprotein. J. Biol. Chem. 271: 33165–33168. and -independent expression of tripartite motif proteins in immune cells. Eur. J. 45. Harada, H., T. Fujita, M. Miyamoto, Y. Kimura, M. Maruyama, A. Furia, Immunol. 38: 619–630. T. Miyata, and T. Taniguchi. 1989. Structurally similar but functionally distinct 22. Lyons, P. A., E. F. McKinney, T. F. Rayner, A. Hatton, H. B. Woffendin, factors, IRF-1 and IRF-2, bind to the same regulatory elements of IFN and IFN- M. Koukoulaki, T. C. Freeman, D. R. Jayne, A. N. Chaudhry, and K. G. Smith. inducible genes. Cell 58: 729–739. 2010. Novel expression signatures identified by transcriptional analysis of sep- 46. Oshima, S., T. Nakamura, S. Namiki, E. Okada, K. Tsuchiya, R. Okamoto, by guest on September 27, 2021 arated leucocyte subsets in systemic lupus erythematosus and vasculitis. Ann. M. Yamazaki, T. Yokota, M. Aida, Y. Yamaguchi, et al. 2004. Interferon regu- Rheum. Dis. 69: 1208–1213. latory factor 1 (IRF-1) and IRF-2 distinctively up-regulate gene expression and 23. Oke, V., I. Vassilaki, A. Espinosa, L. Strandberg, V. K. Kuchroo, F. Nyberg, and production of interleukin-7 in human intestinal epithelial cells. Mol. Cell. Biol. M. Wahren-Herlenius. 2009. High Ro52 expression in spontaneous and UV- 24: 6298–6310. ‑ induced cutaneous inflammation. J. Invest. Dermatol. 129: 2000 2010. 47. Meraro, D., S. Hashmueli, B. Koren, A. Azriel, A. Oumard, S. Kirchhoff, 24. Espinosa, A., W. Zhou, M. Ek, M. Hedlund, S. Brauner, K. Popovic, L. Horvath, H. Hauser, S. Nagulapalli, M. L. Atchison, and B. Z. Levi. 1999. Protein-protein T. Wallerskog, M. Oukka, F. Nyberg, et al. 2006. The Sjogren’s syndrome- and DNA-protein interactions affect the activity of lymphoid-specific IFN reg- associated autoantigen Ro52 is an E3 ligase that regulates proliferation and ulatory factors. J. Immunol. 163: 6468–6478. cell death. J. Immunol. 176: 6277–6285. 48. Baechler, E. C., F. M. Batliwalla, G. Karypis, P. M. Gaffney, W. A. Ortmann, 25. van Boxel-Dezaire, A. H., M. R. Rani, and G. R. Stark. 2006. Complex mod- K. J. Espe, K. B. Shark, W. J. Grande, K. M. Hughes, V. Kapur, et al. 2003. Immunity ulation of cell type-specific signaling in response to type I interferons. Interferon-inducible gene expression signature in peripheral blood cells of 25: 361–372. patients with severe lupus. Proc. Natl. Acad. Sci. USA 100: 2610–2615. 26. Hervas-Stubbs, S., J. L. Perez-Gracia, A. Rouzaut, M. F. Sanmamed, A. Le Bon, 49. Bennett, L., A. K. Palucka, E. Arce, V. Cantrell, J. Borvak, J. Banchereau, and and I. Melero. 2011. Direct effects of type I interferons on cells of the immune V. Pascual. 2003. Interferon and granulopoiesis signatures in systemic lupus system. Clin. Cancer Res. 17: 2619–2627. erythematosus blood. J. Exp. Med. 197: 711–723. 27. Porter, A. C., Y.Chernajovsky, T. C. Dale, C. S. Gilbert, G. R. Stark, and I. M. Kerr. 50. Mavragani, C. P., and M. K. Crow. 2010. Activation of the type I interferon 1988. Interferon response element of the human gene 6-16. EMBO J. 7: 85–92. J. Autoimmun. 28. Friedman, R. L., and G. R. Stark. 1985. a-Interferon‑induced transcription of pathway in primary Sjogren’s syndrome. 35: 225–231. 51. Starr, R., and D. J. Hilton. 1999. Negative regulation of the JAK/STAT pathway. HLA and metallothionein genes containing homologous upstream sequences. Nature 314: 637–639. Bioessays 21: 47–52. 29. Pine, R., and J. E. Darnell, Jr. 1989. In vivo evidence of interaction between 52. Tamura, T., H. Yanai, D. Savitsky, and T. Taniguchi. 2008. The IRF family interferon-stimulated gene factors and the interferon-stimulated response ele- transcription factors in immunity and oncogenesis. Annu. Rev. Immunol. 26: ment. Mol. Cell. Biol. 9: 3533–3537. 535–584. 30. Yuan, J., U. M. Wegenka, C. Lu¨tticken, J. Buschmann, T. Decker, C. Schindler, 53. Honda, K., and T. Taniguchi. 2006. Toll-like receptor signaling and IRF tran- P. C. Heinrich, and F. Horn. 1994. The signalling pathways of interleukin-6 and g scription factors. IUBMB Life 58: 290–295. interferon converge by the activation of different transcription factors which bind 54. Nakagawa, K., and H. Yokosawa. 2000. Degradation of transcription factor IRF- to common responsive DNA elements. Mol. Cell. Biol. 14: 1657–1668. 1 by the ubiquitin- pathway: the C-terminal region governs the 31. Rein, T., M. Mu¨ller, and H. Zorbas. 1994. In vivo footprinting of the IRF-1 protein stability. Eur. J. Biochem. 267: 1680–1686. promoter: inducible occupation of a GAS element next to a persistent struc- 55. Han, K. J., L. Jiang, and H. B. Shu. 2008. Regulation of IRF2 transcriptional tural alteration of the DNA. Nucleic Acids Res. 22: 3033–3037. activity by its sumoylation. Biochem. Biophys. Res. Commun. 372: 772–778. 32. Miyamoto, M., T. Fujita, Y. Kimura, M. Maruyama, H. Harada, Y. Sudo, 56. Nakagawa, K., and H. Yokosawa. 2002. PIAS3 induces SUMO-1 modification T. Miyata, and T. Taniguchi. 1988. Regulated expression of a gene encoding and transcriptional repression of IRF-1. FEBS Lett. 530: 204–208. a nuclear factor, IRF-1, that specifically binds to IFN-b gene regulatory ele- 57. Stevens, A. M., and L. Y. Yu-Lee. 1992. The transcription factor interferon ments. Cell 54: 903–913. regulatory factor-1 is expressed during both early G1 and the G1/S transition 33. Pine, R., T. Decker, D. S. Kessler, D. E. Levy, and J. E. Darnell, Jr. 1990. Purification in the prolactin-induced lymphocyte cell cycle. Mol. Endocrinol. 6: 2236– and cloning of interferon-stimulated gene factor 2 (ISGF2): ISGF2 (IRF-1) can bind 2243. to the promoters of both b interferon- and interferon-stimulated genes but is not 58. Watanabe, N., J. Sakakibara, A. G. Hovanessian, T. Taniguchi, and T. Fujita. a primary transcriptional activator of either. Mol. Cell. Biol. 10: 2448–2457. 1991. Activation of IFN-b element by IRF-1 requires a posttranslational event in 34. Decker, T., D. J. Lew, J. Mirkovitch, and J. E. Darnell, Jr. 1991. Cytoplasmic addition to IRF-1 synthesis. Nucleic Acids Res. 19: 4421–4428. activation of GAF, an IFN-g‑regulated DNA-binding factor. EMBO J. 10: 927– 59. Pernis, A. B. 2002. The role of IRF-4 in B and T cell activation and differen- 932. tiation. J. Interferon Cytokine Res. 22: 111–120. The Journal of Immunology 3763

60. Ishii,T.,K.Ohnuma,A.Murakami,N.Takasawa,T.Yamochi,S.Iwata,M.Uchiyama, 62. Fujii, Y., T. Shimizu, M. Kusumoto, Y. Kyogoku, T. Taniguchi, and T. Hakoshima. N. H. Dang, H. Tanaka, and C. Morimoto. 2003. SS-A/Ro52, an autoantigen involved 1999. Crystal structure of an IRF-DNA complex reveals novel DNA recogni- in CD28-mediated IL-2 production. J. Immunol. 170: 3653–3661. tion and cooperative binding to a tandem repeat of core sequences. EMBO J. 61. Fang, T. C., U. Schaefer, I. Mecklenbrauker, A. Stienen, S. Dewell, M. S. Chen, 18: 5028–5041. I. Rioja, V. Parravicini, R. K. Prinjha, R. Chandwani, et al. 2012. Histone H3 63. Der, S. D., A. Zhou, B. R. Williams, and R. H. Silverman. 1998. Identification of lysine 9 di-methylation as an epigenetic signature of the interferon response. J. genes differentially regulated by interferon a, b,org using oligonucleotide Exp. Med. 209: 661–669. arrays. Proc. Natl. Acad. Sci. USA 95: 15623–15628. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021