T Suppressor Cells + CD8 Essential for the Differentiation of Human Micrornas by Ig-Like Transcript 3 Is Downregulation of Infla

Home , FOSB

Downregulation of Inflammatory MicroRNAs by Ig-like Transcript 3 Is Essential for the Differentiation of Human CD8 + T Suppressor Cells This information is current as of September 27, 2021. Chih-Chao Chang, Qing-Yin Zhang, Zhuoru Liu, Raphael A. Clynes, Nicole Suciu-Foca and George Vlad J Immunol 2012; 188:3042-3052; Prepublished online 2 March 2012; doi: 10.4049/jimmunol.1102899 Downloaded from http://www.jimmunol.org/content/188/7/3042

Supplementary http://www.jimmunol.org/content/suppl/2012/03/02/jimmunol.110289 http://www.jimmunol.org/ Material 9.DC1 References This article cites 46 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/188/7/3042.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 27, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

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

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Downregulation of Inﬂammatory MicroRNAs by Ig-like Transcript 3 Is Essential for the Differentiation of Human CD8+ T Suppressor Cells

Chih-Chao Chang, Qing-Yin Zhang, Zhuoru Liu, Raphael A. Clynes, Nicole Suciu-Foca, and George Vlad

We have investigated the mechanism underlying the immunoregulatory function of membrane Ig-like transcript 3 (ILT3) and soluble ILT3Fc. microRNA (miRNA) expression profile identified genes that were downregulated in ILT3-induced human CD8+ T suppressor cells (Ts) while upregulated in T cells primed in the absence of ILT3. We found that miR-21, miR-30b, and miR-155 target the 39-untranslated region of genes whose expression was strongly increased in ILT3Fc-induced Ts, such as dual specificity phosphatase

10, B cell CLL/lymphoma 6, and suppressor of cytokine signaling 1, respectively. Transfection of miRNA mimics or inhibitors and Downloaded from site-speciﬁc mutagenesis of their 39-untranslated region binding sites indicated that B cell CLL/lymphoma 6, dual speciﬁcity phosphatase 10, and suppressor of cytokine signaling 1 are direct targets of miR-30b, miR-21, and miR-155. Primed CD8+ T cells transfected with miR-21&30b, miR-21&155, or miR-21&30b&155 inhibitors displayed suppressor activity when added to autologous CD3-triggered CD4 T cells. Luciferase reporter assays of miR-21 and miR-155 indicated that their transcription is highly dependent on AP-1. Analysis of activated T cells showed that ILT3Fc inhibited the translocation to the nucleus of the AP-1 subunits,

FOSB and c-FOS, and the phosphorylation of ZAP70 and phospholipase C-g 1. In conclusion, ILT3Fc inhibits T cell activation and http://www.jimmunol.org/ induces the generation of Ts targeting multiple inﬂammatory miRNA pathways. The Journal of Immunology, 2012, 188: 3042–3052.

he induction of Ag-speciﬁc regulatory T cells (Treg) from erated by repeated allostimulation in MLC, or alternatively by primed lymphocytes is a complex process that limits the exposure to some cytokines such as IL-10, IFN-a,IFN-b (3–5), or T “collateral damage” resulting from protective immunity other agents, such as vitamin D receptor agonists (6, 7). Upregu- and inﬂammatory responses against self- and non–self-Ags. There lation of ILT3 expression or ILT3 transfection of DC results in is ample evidence that dendritic cells (DC) can prevent and inhibit inhibition of CD40 signaling and of NF-kB activation (3). Tol-

T cell-mediated effector responses acquiring tolerogenic property erogenic human DC are characterized by a high expression of by guest on September 27, 2021 and inducing anergy or instructing T cells to become suppressor/ ILT3/ILT4 on their membrane and by their capacity to induce regulatory cells (1). Th anergy and the differentiation of Treg/Ts (3, 4). In contrast, The human Ig-like transcript (ILT)3 and 4, also known as LIRB4/ knockdown (KD) of ILT3 from DC (ILT3 KD-DC) increases their LIR5/CD85k and LIRB2/LIR2/CD85D, belong to a family of innate TLR responsiveness, as reflected in synthesis and secretion of immune receptors that are expressed by DC and monocytes (2). proinflammatory cytokines (IL-1a and b, IL-6, and type I IFN) and These ILT receptors display a long cytoplasmic tail containing migration factors CXCL10 and CXCL11. ILT3KD-DC enhance ITIMs, which mediate inhibition of cell activation by recruiting T cell proliferation and secretion of IFN-g and IL-17 when pulsed tyrosine phosphatase Src homology region 2 domain-containing with CMV or used as allostimulators in MLC (8). phosphatase 1 (2). Upregulation of ILT3 and ILT4 on the mem- These data, in conjunction with the finding that CD8+ T cells brane of DC can be induced in a cytokine-independent manner by from rejection-free heart, kidney, or liver transplant recipients direct interaction with human CD8+ T suppressor cells (Ts), gen- induced the upregulation of ILT3/4 in donor APC, substantiate the importance of these inhibitory receptors for maintenance of im- Division of Immunogenetics and Cellular Immunology, Department of Pathology and munologic quiescence (1). Cell Biology, Columbia University, New York, NY 10032 The extracellular domain of ILT3 retains the T cell inhibitory Received for publication October 6, 2011. Accepted for publication January 29, function even upon deletion of the cytoplasmic, ITIM-containing 2012. tail, because DC transfected with a construct made up of only the This work was supported by research funding from Juvenile Diabetes Research extracellular portion were still capable to elicit the differentiation Foundation (1-2008-550) and the Interuniversitary Organ Transplantation Consor- + tium (Rome, Italy). of CD8 Ts (9). On the basis of this finding, we engineered a Address correspondence and reprint requests to Dr. Nicole Suciu-Foca, Department soluble form of ILT3, which was expressed as an ILT3Fc fusion of Pathology and Cell Biology, Columbia University, 630 West 168 Street, VC 15- protein and tested its immunomodulatory activity. This recombi- 204, New York, NY 10032. E-mail address: [email protected] nant protein inhibited primary and secondary T cell responses in The online version of this article contains supplemental material. MLC and blocked the differentiation of CD8+ cytotoxic T cells Abbreviations used in this article: BCL6, B cell CLL/lymphoma 6; DC, dendritic (CTL). Furthermore, it elicited the in vitro and in vivo differen- cell; DUSP, dual specificity phosphatase; FOS, FBJ murine osteosarcoma; hRluc, + luciferase (Renilla); ILT3, Ig-like transcript 3; KD, knockdown; luc, luciferase (Fire- tiation of CD8 Ts, which produced no cytokines, inhibited T cell fly); miRNA, microRNA; PLC, phospholipase C; SOCS1, suppressor of cytokine reactivity and induced the upregulation of ILT3 on priming APC signaling 1; TGFBR2, TGF-bR 2; Treg, T regulatory cell; TOB1, transducer of (9–11). In vivo studies showed that ILT3Fc induced tolerance to ERBB2; Ts, T suppressor cell; 39-UTR, 39-untranslated region. allogeneic human pancreatic islet cells transplanted in humanized Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 diabetic NOD/SCID mice (10, 11). www.jimmunol.org/cgi/doi/10.4049/jimmunol.1102899 The Journal of Immunology 3043

Gene profile analysis of CD8+ T cells primed for 7 d in MLC in CFSE-labeled CD4 Th responders was assessed at day 3 analyzing dye di- the presence or absence of ILT3Fc showed that several hundred lution by flow cytometry using a FACSCalibur instrument and CellQuest genes belonging to .28 gene ontology categories were modulated software (BD Biosciences). As controls, conditions were set up, in which no CD8 T cells, or no anti-CD3 mAb were added. CD8 T cells transfected with by ILT3Fc (12). There was a striking increase in the expression of control RNA (no miRNA inhibitors) were also assayed. transcription factors belonging to a class of zinc finger transcriptional repressors including B cell CLL/lymphoma 6 (BCL6), a Gene promoter and 39-UTR constructs known inhibitor of IFN-g (13), IL-17 (14, 15), IL-5 (16, 17), and Sequences corresponding to 21296 to +40 bp from the mRNA start of granzyme B (18) expression. BCL6 was found to be crucial to CD8 BCL6 and to 2748 to +52 from the RNA start of SOCS1 genes were Ts function because KD of BCL6 from unprimed human CD8 cloned from genomic DNA by PCR using high-fidelity Taq and gene- specific primers (Supplemental Table I). These gene promoters, previ- T cells prevented their differentiation into Ts, whereas transfection ously shown to be functional in response to various stimuli in reporter + of BCL6 in primed CD8 T cells resulted in the acquisition of Ts assays (22, 23), were subsequently cloned into HindIII or BglII and NheI function (11). High expression of BCL6 was also found in hu- sites of pGL3 basic Firefly-luciferase construct (Promega). Sequences manized mice tolerating islet allografts (11). Other genes strongly corresponding to 2410 to +40 bp from the mRNA start of miR-21 promoter were similarly constructed. This region includes two AP-1 sites, upregulated by ILT3Fc included the dual specificity phosphatase P D AP-1 (proximal) and AP-1 (distal), shown to be fully functional, PMA (DUSP)8, DUSP10, TGF-bR 2 (TGFBR2), transducer of ERBB2 responsive, and critical for expression of miR-21 in 293T cells (24). miR- (TOB1), CXCR4, adrenergic b-2R, and suppressor of cytokine 155 promoter construct and its AP-1 and NF-kB binding sites mutants signaling (SOCS)1 (11, 12). All of these genes are involved in were provided by Dr. E. Flemmington (Tulane Health Sciences Cen- functional differentiation of T cells, and are targeted by micro- ter, New Orleans, LA) (25). The NF-AT/AP-1 basic reporter construct pGL4.30, which contains an enhancer element from the human IL-2 gene RNAs (miRNA), some of which are discussed in the current study. promoter, was obtained from Promega; the AP-1(miR21P) basic reporter Downloaded from miRNA, a newly discovered class of small noncoding RNAs, are construct was prepared by replacing a 140-bp NF-AT/AP-1 sequence of key negative regulators of gene expression. As with conventional pGL4.30 with a triplet of the AP-1P site of the miR-21 gene promoter. Full-length 39-UTR of various genes (1.1 kb BCL6, 0.4 kb SOCS1, and protein-encoding RNA, miRNA are transcribed by RNA poly- + merase II, and their expression is controlled by transcriptional 1.0 kb DUSP10) were obtained from CD8 T cell cDNA libraries by PCRs using a high-fidelity Taq DNA polymerase (Invitrogen) and gene specific factors. The mature miRNAs inhibit target mRNA translation or primers (Supplemental Table I). PCR products were first cloned into promote their degradation by directly binding to the 39-untranslated pGEM-T Easy vector (Promega), excised from recombinant plasmids http://www.jimmunol.org/ region (39-UTR) of target genes (19). Increasing evidence sug- by NotI digestion, and subcloned into psiCHECK-2 luciferase reporter gests that the differentiation of T cells with diverse effector or (Promega). The psiCHECK-2 vector contains an SV40 early enchancer/ promoter, a synthetic Renilla luciferase (hRluc) gene flanked at the 39-end regulatory function is largely controlled by miRNAs, which in- by a NotI restriction site, and a synthetic poly(A) tail. All reporter con- tegrate cues from the microenvironment to either turn on or si- structs were completely sequenced from both ends. lence lineage-specific genes (20, 21). The aim of our present study was to identify the specific miRNAs miRNA mimics and hairpin inhibitors that are involved in the upregulation of BCL6 and other genes Stability-enhanced miR-30b RNA oligonucleotide (miRIDIAN Mimic- related to the differentiation of CD8 Ts primed in the presence of 30b) and miR-155 oligonucleotide (miRIDIAN Mimic-155), hairpin ILT3Fc. To our knowledge, our current findings show for the first miRNA inhibitors (meridian hairpin inhibitor miR-21, miR-30b, and miR- by guest on September 27, 2021 155), and the control nontargeting RNA oligonucleotide (miRIDIAN time how an immunomodulatory molecule, such as ILT3Fc, Mimic Negative Control number 1) were purchased from Dharmacon. inhibits T cell immune/inflammatory responses by downregulating miRNA mimics (0–10 nmol) were transfected into fresh, nonactivated CD8 the expression of inflammatory miRNA. T cells (3 3 106), and hairpin inhibitors (0–10 nmol) were transfected into polyclonally activated CD8+ T cells (3 3 106) using Amaxa’s Human T cell Nucleofector Kit (Lonza). Materials and Methods Mutagenesis Cell cultures Site-directed mutagenesis of AP-1 binding sites in miR-21 promoter Preparation of PBMC and negative isolation of CD3+ T cells have been + + 2 constructs and mutagenesis of miR-30b and miR-155 binding sites in 39- described previously (9). To generate allospecific CD8 Ts, CD3 CD25 UTR of BCL6 and SOCS1 was performed using QuickChangeII (Stra- 3 6 T cells (1 10 /ml) were cultured with irradiated CD2-depleted PBMC tagene) with mutated gene-specific primers and corresponding DNA 3 6 (0.5 10 /ml) in the presence of ILT3Fc (50 mg/ml) or in its absence, as templates (Supplemental Table I). control. Preparation of ILT3Fc protein, which shows .99% purity by SDS-PAGE, has been described previously (9, 11). On day 7, CD8+ T cells Transfections of reporter constructs and miRNA were negatively selected from these cultures and tested for suppressor oligonucleotides activity prior to use in mRNA or microarray assays. Allostimulated CD8+ T cells treated with ILT3Fc, which inhibited the MLC response by .80%, Polyclonally activated CD8+ T cells were generated by stimulation with were used as Ts. Non–ILT3-treated CD8+ T cells had no inhibitory activity anti-CD3/CD28 mAb in the presence or absence of ILT3Fc (50 mg/ml) and were used as controls. as described above. After 48 h, 3–5 3 106 cells were collected and Polyclonally activated CD8+ T cells were prepared from fresh PBMC by cotransfected with 3 mg pGL3-based (Promega) promoter constructs negative selection (using anti-CD4, -CD14, -CD20, -CD56–coated magnetic (containing either BCL6 or SOCS1 promoters and the Firefly luciferase, beads) followed by incubation in anti-CD3–coated T cell activation plates luc, gene) and 2 mg pGL4.70[hRluc] (Promega) (containing the constitu- (BD Biosciences) in the presence of 2 mg/ml anti-CD28 Abs. After 2 d of tively expressed Renilla luciferase, hRluc, gene), using Amaxa Human incubation, cells were collected, washed twice, and transfected with either T cell Nucleofector Kit (Lonza). Sixteen hours after transfection, cells reporter constructs or RNA oligomers (miRNA mimics or hairpin inhibitors) were lysed and assayed for both Firefly and Renilla luciferase activities in using the nucleofectin method. Putative CD8+ suppressor cells were generated a single tube luminometer (Turner BioSystems 20/20). Normalized pro- by transfection of polyclonally activated CD8+ T cells with individual hairpin moter activity was measured as units of Firefly luciferase activity divided inhibitors of miRNA or mixtures of two or three inhibitors. To test the ability by units of Renilla luciferase activity. of these putative Ts to inhibit autologous CD4 Th responses, transfected CD8 The 39-UTR reporter gene activities in polyclonally activated CD8 T cells (50 3 103/well) were coincubated with CFSE-labeled CD4+CD252 T cells were similarly assayed, using 2 mg psiCHECK-2 vectors. These Th responders, and CD14+ monocytes at a 1:1:1 ratio, in the presence of 1 mg/ vectors contained an SV40 promoter, the Renilla hRluc reporter gene, and ml anti-CD3 mAb (clone UCHT1; BD Biosciences). Magnetically sorted the 39-UTR of BCL6, SOCS1, or DUSP10 genes, in addition to a consti- CD4+CD252 responder T cells were labeled by incubation with 3 mM CFSE tutively expressed Firefly luc gene. For normalization, hRluc activity was (“CellTrace”; Invitrogen) for 10 min in PBS/0.1% BSA at room temperature. divided by internal Firefly luc activity. ILT3Fc treatment did not affect the The reaction was stopped by addition of ice-cold culture medium containing normalized hRluc/luc activity in activated Jurkat or CD8 T cells trans- 10% FCS for 5 min, followed by extensive washing. Proliferation of the fected with the unmodified psiCHECK-2 plasmid. 3044 miRNA SIGNATURE OF ILT3-Fc–INDUCED CD8+ Ts

For miRNA promoter reporter assays, Jurkat T cells (American Type Induction of BCL6, SOCS1, and DUSP10 by ILT3Fc is 39-UTR Culture Collection) were pretreated with either human IgG or ILT3Fc for dependent 16 h. Cells were transfected using Lipofectamine 2000 (Invitrogen) with one of the following Firefly luc reporter plasmids: pGL3 containing miR-21 or Computer aid searches for putative targets of these miRNAs miR-155 promoters, pGL4.30 [luc2/NFAT-RE/Hygro] containing the NFAT/ showed that many of them were genes whose expression was AP-1 binding sites of the human IL-2 promoter, or pGL4.30 from which the upregulated (.3-fold) in CD8+ T cells allostimulated in the pres- NFAT/AP-1 sites were removed and replaced with three copies of the AP-1P site of the miR-21 promoter. The cells were cotransfected with pGL4.70 ence of ILT3Fc compared with CD8 T cells primed in cultures [hRluc] plasmid constitutively expressing Renilla luciferase (Promega) for without ILT3Fc (11, 12). Table I shows a partial list of these assessing transfection efficiency and normalization. The cells were then ILT3Fc-upregulated genes whose 39-UTRs contain target sites stimulated for 48 h with PMA (10 nM) or mAb anti-CD3/CD28, and re- for these ILT3Fc-modulated miRNAs. Genes induced by ILT3Fc porter gene readouts were performed and normalized as above. treatment, which are predicted to be targeted by miR-21, include miRNA array and real-time PCR dual-specific phosphatases DUSP8 and DUSP10, known to be inhibitors of the MAPK pathway and cytokine production (26). ILT3Fc-induced allospecific CD8+ Ts cells and primed (non-ILT3 treated) CD8+ T cell controls were generated in MLC as described above. Five Also included in this group are TGFBR2 and TOB1, involved in micrograms of total RNA prepared from these cells using TRIzol (Invi- regulation of T cell responses (27, 28). Genes upregulated by trogen) was annealed to oligonucleotide primer mix and hybridized to 132 ILT3Fc, which are predicted to be targeted by miR-30b and miR- miRNA oligonucleotide probes. Streptavidin–HRP chemiluminescence was 146a, are BCL6 and CXCR4, respectively. There is already evi- used for detection of miRNA expression (Signosis). Real-time PCR detec- dence that miR-146a controls expression of CXCR4 (29), miR-21 tion of miRNA was performed using TaqMan Small RNA assays (Applied Biosystems). Relative expression of miRNA was normalized by U6 miRNA. controls TGFBR2 (30), and miR-155 acts on SOCS1, a negative Downloaded from Relative expression or protein encoding genes was normalized by GAPDH. regulator of cytokine signaling through STAT1 (31–33). SOCS1 also contains a putative target site for miR-30b. Nuclear protein extraction and immunoblotting The upregulation of these target genes by ILT3Fc was further For signaling studies, Jurkat or primary CD8 T cells were stimulated with confirmed by real-time PCR analysis using three independent CD3/CD28 mAbs (2 mg/ml) and a cross-linking anti-mouse IgG Ab for the samples sets. ILT3Fc-induced inhibition of IFN-g expression was indicated times in the presence or absence of ILT3Fc (15 mg/ml). Total cell + found in all of our previous studies (9–12) and is shown as lysates (20 mg) prepared from CD8 T lymphocytes or Jurkat T cells were http://www.jimmunol.org/ loaded onto either 10 or 4–12% SDS-PAGE gels, transferred to a polyvi- a control for the ILT3Fc effect (Fig. 2A). The findings that ILT3Fc nylidene difluoride membrane, and probed with various Abs as described induces upregulation of these genes and downregulation of the previously (8). In some experiments, a commercial protein extract kit miRNA (which target them) prompted us to examine the under- (NucBuster; Novagen) was used to extract nuclear proteins. Polyclonal lying molecular mechanism. rabbit Abs recognizing various TCR and AP-1 signaling proteins (Cell Signaling Technology), SOCS1, BCL6 (Santa Cruz Biotechnology), and The upregulated expression of BCL6, SOCS1, and DUSP10 is + DUSP10 (Abcam) were used to probe the membranes. integral to the signature of ILT3Fc-induced CD8 Ts. These genes are known inhibitors of cytokine production and TCR signaling miRNA target prediction and are targeted by miRNAs, which are suppressed by ILT3Fc. We Several Web-accessible miRNA database searching programs were used for thus analyzed the relationship between miRNAs and gene ex- by guest on September 27, 2021 prediction of miRNA target sites. These include http://www.microrna.org, pression in the presence and absence of ILT3Fc. http://www.miRBase.org, and http://www.targetscan.org. Because target sites of miRNAs, with very few exceptions, are Statistical analysis located in the 39-UTR of messenger RNAs, we performed 39-UTR reporter assays. PsiCHECK-2 luciferase reporter constructs made Differences between control and experimental groups (each consisting of up of an SV40 promoter, the Renilla luciferase gene, and the 39- a minimum of three individual experiments) were estimated using the t test + for paired two-tailed distribution. Mean and SE were calculated for each UTRs of BCL6, SOCS1, or DUSP10 were transfected into CD8 group. Differences associated with p , 0.05 were considered significant. T cells, which had been stimulated for 2 d with anti-CD3/CD28 mAbs in the presence of absence of ILT3Fc. ILT3Fc induced the Results increase of luciferase reporter activity in primed CD8+ T cells ILT3 inhibits proinflammatory miRNAs expression in transfected with BCL6, SOCS1, or DUSP10 39-UTR reporters alloantigen-stimulated T cells (Fig. 2B), confirming the importance of the 39-UTR for the ILT3- induced upregulation of these genes. To investigate the role of the We first explored the miRNA transcriptional profiles of CD8+ Ts + respective promoters on the induction of these genes by ILT3Fc, generated in 7-d MLC containing ILT3Fc and in control CD8 we generated pGL3-based promoter reporter constructs consisting T cells allostimulated without ILT3Fc. A group of at least six of the Firefly luciferase gene under the transcriptional control of miRNAs (miR-21, miR-146a, miR-30b, miR-30c, miR-29, and the BCL6 or SOCS1 promoters. These constructs were transfected miR-155), which are expressed at negligible levels in unstimulated + into activated CD8 T cells which had been stimulated for 2 d with CD8 T cells, were consistently upregulated by allostimulation anti-CD3/CD28 mAbs in the presence of absence of ILT3Fc. in three independent experiments (Fig. 1A). When ILT3Fc was These experiments showed that ILT3Fc had no significant effect added to the cultures, the level of expression of these genes was on luciferase activity and therefore the upregulation of Ts signa- strongly downregulated, as confirmed by real-time PCR analysis ture genes by ILT3Fc is not mediated through induction of tran- (Fig. 1B). Similarly, membrane ILT3 also inhibits the expression scriptional activity at the BCL6 and SOCS1 promoters (Fig. 2C). of miR-21, -30b, -146a, and -155 as demonstrated in experiments in which T cells were allostimulated with DC transfected miR-30b, miR-155, and miR-21 target the 39-UTR of with ILT3 small interfering RNA (ILT3KD-DC) or empty vector ILT3Fc-inducible genes (control-DC). Upon priming with ILT3KD DC, CD8+ T cells, The dependence of ILT3Fc-induced upregulation of DUSP10, sorted after 16 h from these cultures, showed much higher ex- SOCS1, and BCL6 on their 39-UTR suggested that the miRNA pression of miR-21, miR-30b, miR-146a, and miR-155, as deter- which target them may play a role in regulation of gene expres- mined by RT-PCR, compared with CD8+ T cells stimulated with sion. We analyzed BCL6, SOCS1, and DUSP10 expression in ILT3+ control DC (Fig. 1C). Hence, both membrane and soluble CD8+ T cells by transient transfection of either gene-specific, ILT3 inhibit miRNAs expressed by MLC-stimulated CD8+ T cells. chemically stabilized, dsRNA oligomers, that mimic the func- The Journal of Immunology 3045 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. Identification of miRNA differentially expressed by CD8+ T cells primed in the presence or absence of ILT3. (A) miRNA array of unprimed CD8+ T cells and CD8+ T cells allostimulated in the presence or absence of ILT3Fc. Arrows indicate miRNA upregulated in alloactivated CD8+ T cells and suppressed by ILT3Fc. (B) RT-PCR analysis of miRNAs identified above. (C) Expression of inflammatory miRNAs in CD8+ T cells stimulated in MLC with ILT3-KD or control DC. Each experiment was independently repeated three times. Means values 6 SEM, *p , 0.05. tion of endogenous mature miRNA (miRNA mimic), or chemi- transfected with nontargeting RNA oligomers. Western blot analyses cally modified, single-stranded antisense oligomers that inhibit of BCL6 and SOCS1 proteins, using b-actin as an internal control, miRNA function (miRNA hairpin inhibitor). Transfection effici- showed that overexpression of miR-30b downregulated the expres- ency of either unstimulated or polyclonally stimulated CD8+ sion of BCL6 and SOCS1, whereas overexpession of miR-155 lead to T cells, monitored by FACS analysis of transfected GFP RNA downregulation SOCS1 (Fig. 3B) but not BCL6 (which has no target oligomer, was ∼75%. The predicted miRNA target sites in the 39- site for it). Inhibition of BCL6 and SOCS1 gene expression by these UTRs of BCL6, SOCS1, and DUSP10 are shown in Fig. 3A. miRNA mimics increased with the amount added to the culture. + We transfected unprimed CD8 T cells with miRNA mimics As a corollary, transfection of the polyclonally stimulated CD8+ specific for miR-30b and miR-155. As a control, the cells were T cells with gene-specific hairpin inhibitors to knock down the respective miRNAs induced the upregulation of the target genes (Fig. Table I. ILT3Fc-inducible genes contain potential binding sites for 3C). Hairpin inhibitors of miR-30b induced upregulation of BCL6 miRNAs, which are inhibited by ILT3Fc and SOCS1, whereas inhibitors of miR-155 upregulated only SOCS1. KD of miR-21 resulted in upregulation of DUSP-10 yet had no effect Gene Induced Induction by ILT3Fc (Fold) miRNA Sites in 39-UTR on BCL6, which lacks the target sequence for miR-21 (Fig. 3C). To determine whether physical binding of miRNA to their tar- DUSP8 11.23 miR-21 DUSP10 3.2 miR-21 get sites in 39-UTR of the corresponding genes is required for TGFBR2 3.3 miR-21 their downregulation, we generated a series of reporters in which TOB1 4.9 miR-21 the consensus regions of miR-21, miR-30b, or miR-155 sites ZNF331 8.2 miR-146a were mutated. Experiments in which these mutant reporters were BCL-6 3.1 miR-30b + CXCR4 6.6 miR-146a transfected in stimulated CD8 T cells from three different donors ADRB2 4.5 miR-21 and miR-30b showed that mutations of miR-30b sites in 39-UTR of BCL6 and SOCS1 3.7 miR-30b and miR-155 SOCS1, and mutation of miR-21 site in 39-UTR of DUSP10 ab- ADRB2, Adrenergic b-2R. rogated their responsiveness to miRNA suppression, inducing 3046 miRNA SIGNATURE OF ILT3-Fc–INDUCED CD8+ Ts Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 2. Activation of BCL6, SOCS1, and DUSP10 genes by ILT3Fc is 39-UTR dependent. (A) Real-time PCR analysis of selected genes expressed by CD8+ T cells allostimulated in the presence or absence of ILT3Fc in three independent sample sets. (B) Effect of ILT3Fc on BCL6, SOCS1, and DUSP10 39-UTR reporter assays. Polyclonally activated CD8+ T cells stimulated in the presence of absence of ILT3Fc were transfected with psiCHECK-2 vectors containing the Renilla hRluc gene linked to the 39-UTRs of BCL6, SOCS1, or DUSP10 and a constitutively expressed Firefly luc gene. The luciferase activity was assayed 16 h later, and the normalized hRluc/luc values are represented. (C) Effect of ILT3Fc on the BCL6 and SOCS1 promoter activities. Polyclonally activated CD8+ T cells stimulated in the presence of absence of ILT3Fc were cotransfected with pGL3 plasmids containing the Firefly luc gene under the control of BCL6 or SOCS1 promoters and pGL4 plasmids containing a constitutively expressed Renilla hRluc gene. The luciferase activity was assayed 16 h later, and the normalized luc/hRluc values are represented. Mean and SEM of three experiments and statistical significance (*p , 0.05) are indicated.

higher reporter activities than wild-type miRNA binding sequen- CD28 mAbs in the presence or absence of ILT3Fc. ILT3Fc treat- ces (Fig. 3D). Mutation of miR-155 in 39-UTR of SOCS1, as ment no longer affected the endogenous expression of BCL6, expected, also expressed higher reporter gene activities. DUSP10, and SOCS1 in cells transfected with these miRNA To further determine whether downregulation of these miRNAs mimics (Fig. 4). is essential for the ILT3-induced upregulation of genes involved Taken together, these data indicate that BCL6, SOCS1, and in the differentiation of Ts, we analyzed the response of the en- DUSP10 are direct targets of posttranscriptional regulation me- dogenous target genes in CD8 T cells that have been transfected diated by miR-30b, miR-155, and miR-21, respectively. Hence, with vectors containing constitutively expressed miR-30b, miR-21, downregulation of these miRNAs by ILT3Fc in primed CD8+ and miR-155 mimics and then stimulated for 2 d with anti-CD3/ T cells leads to upregulation of ILT3Fc-inducible gene expression. The Journal of Immunology 3047 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. Direct targeting of ILT3Fc-inducible genes BCL6, SOCS1, and DUSP10 by miR-21, miR-30b, and miR-155. (A) Predicted miRNA targets in the 39-UTRs of DUSP10, SOCS1, and BCL6. (B) Western blot analyses of unstimulated CD8+ T cells transfected with control, nontargeting RNA oligomers (ctl) or BCL6 and SOCS1 gene-specific mimic miRNA oligomers. Expression of b-actin was used as a loading control. One of three independent experiments is illustrated. (C) RT-PCR analysis of DUSP10, BCL6, and SOCS1 expression in polyclonally activated T cells transfected with increasing concentrations (1–10 nmol) of hairpin inhibitors of miR-21, miR-30b, miR-155, or nontargeting control (ctl). (D) Analysis of site-directed mutagenesis of miR-21, miR-30b, and miR-155 targets located in 39-UTR regions of DUSP10, BCL6, and SOCS1. Consensus nucleotides are indicated in uppercase; mutated nucleotides within the consensus are shown as lowercase. Normalized reporter activity shows that mutations in the 39-UTR abrogate the inhibitory activity of miR-21, miR-30b, and miR-155. The mean, SE, and statistical significance of the differences between control and transfected cells (*p , 0.05) in three independent experiments are shown in (B)–(D). NSB, Nonspecific band.

CD8 T cells transfected with miRNA inhibitors suppress CD4 hairpin inhibitors of miR-21, -30b, or -155, were added to CD4+ T cell responses T cells, there was no inhibition in T cell proliferation compared with We sought to determine whether ILT3Fc-induced differentiation of control cultures containing nontargeted oligonucleotides (Fig. 5). + Although individually none of these miRNA inhibitors were Ts is due to inhibition of these inflammatory miRNA. For this, CD8 + T cells from three different blood donors were activated by 2-d in- sufficient to induce Ts, activated CD8 T cells transfected with cubation with anti-CD3 and CD28 mAb and then transfected with mixtures of miR-21/30b, miR-21/155, or miR-21/30b/155 inhib- + individual or combinations of miRNA inhibitors. Sixteen hours after itors suppressed CD3-triggered proliferation of CD4 T cells by transfection, these CD8+ cells were added to autologous CFSE- .50% in five independent experiments. ILT3Fc added to CD3- labeled CD4+CD252 T cells in cultures containing 1 mg/ml anti- activated CD4 T cells inhibited proliferation by .80%. This sug- CD3 Abs and autologous APC. Cell proliferation assay showed that gests that miRNAs targeting DUSP10, BCL6, and SOCS1 act in when CD8+ T cells, which had been transfected with individual concert, playing a role in the differentiation of CD8+ Ts (Fig. 5). 3048 miRNA SIGNATURE OF ILT3-Fc–INDUCED CD8+ Ts

FIGURE 4. Enforced expression of miR-21, miR-30b, and miR-155 abolishes the effect of ILT3-Fc–induced upregulation of DUSP10, BCL6, and SOCS1. RT-PCR ampliﬁcation of BCL6, DUSP10, and SOCS1 using RNA from CD8+ T cells primed with anti-CD3/CD28 Abs for 48 h in the presence or absence of ILT3-Fc and transfected with mixtures of miR-21, miR-30b, and miR-155 or nontargeting control for additional 16 h. The average of three individual experiments is shown. Error bars represent SEM.

ILT3Fc regulates microRNA gene expression by inhibiting their rescinated ILT3Fc (.90%), lending itself as an ideal tool to study promoter activities signaling effects downstream of the ILT3 ligand. Having learned that ILT3Fc induces the upregulation of BCL6, To determine whether miRNA gene promoters can be activated

DUSP10, and SOCS1 via downregulation of miRs, which target by anti-CD3/CD28 mAbs in Jurkat cells, Firefly luciferase reporter Downloaded from their 39-UTR, we tried to understand the mechanism of miRNA gene plasmids (pGL3 based), in which the reporter was under the inhibition. For some of these experiments, we used the Jurkat transcriptional control of either miR-21 or miR-155 gene pro- T cell line whose miRNA expression profile is similar to that of moters, were cotransfected with a Renilla luciferase control re- activated human T lymphocytes and shows high binding of fluo- porter construct. After 48 h of incubation with CD3/CD28 mAbs, http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 5. Effect of CD8+ T cells transfected with miR inhibitors on T cell proliferation. Polyclonally activated CD8 T cells were transfected with individual or combinations of miR inhibitors (as indicated) and then added to CFSE-labeled, resting, and autologous CD4+CD252 T cells and APC. Anti- CD3 mAb were added to trigger CD4 T cell proliferation. CD4 T cell responses are depicted in various culture conditions in which the anti-CD3 mAb was omitted (negative control) (A), no CD8 T cells were added (uninhibited CD4 Th response) (B), ILT3Fc was added to inhibit proliferation (C), CD8 T cells transfected with nontargeted RNA [CD8(control)] were added (D), or putative CD8 Ts transfected with individual (E–G) or combinations (H–J) of miR inhibitors were added, as indicated. (K) shows the overlay between the proliferation of CD4 responders in cultures where no CD8 Ts were added (uninhibited response) and the CD4 Th response in the presence of CD8 T cells transfected with inhibitors of miR-21/miR-30b/miR-155. Each histogram represents 104 CD4-gated T cells. One of five independent experiments is illustrated. The Journal of Immunology 3049 normalized promoter activities (Firefly luciferase/Renilla lucifer- Because transcription factors binding to AP-1 site play an im- ase) showed a 2- to 3.5-fold increase above the level seen in portant role in TCR signaling, we investigated the possibility that nonstimulated Jurkat cells (Fig. 6A). Mutations at the AP-1 TCR/AP-1 signaling pathways are affected by ILT3Fc. Jurkat or binding sites in the miR-21 or miR-155 gene promoters com- primary CD8 T cells were preincubated with 15 mg/ml ILT3Fc for pletely abolished reporter gene activity in three independent ex- 16 h and then stimulated by adding anti-CD3/CD28 Abs to the periments. This indicates that the transcriptional activities of cultures over 3 h. We isolated nuclear fractions of cell extracts miR-21 and miR-155 depend on AP-1. obtained from these cultures at various time points. Nuclear pro- We next tested the promoter activities of miR-21 and miR-155 teins from these cells were probed with Abs recognizing various genes in CD3/CD28-triggered Jurkat cells pretreated with various subunits of AP-1 protein complexes. Western blot analyses showed doses of ILT3Fc. The results indicated that ILT3Fc at concen- that the FBJ murine osteosarcoma (FOS) proteins—c-FOS and two trations as low as 5 mg/ml strongly inhibited miR-21 and miR-155 forms of FOSB—were rapidly synthesized and translocated into miRNA gene promoter activities (Fig. 6B). Hence, the AP-1 bind- the nucleus of non–ILT3Fc-treated (control) Jurkat cells. After 3 h ing sites within the promoter regions of miR-21 or miR-155 were of stimulation, the level of c-FOS was already downregulated, crucial to the inhibitory effect of ILT3Fc. whereas the level of FOSB was still upregulated. In ILT3Fc-treated Jurkat cells, c-FOS and FOS-B proteins were detected in the nu- ILT3Fc inhibits basal TCR signaling and TCR-mediated cleus in lower amounts and at later times, indicating that ILT3Fc nuclear localization of AP-1 inhibited and delayed TCR signaling events (Fig. 7B). Primary Because AP-1 binding to the promoters of miR-21 and miR-155 is T cells, in contrast to Jurkat cells, display low levels of ILT3Fc essential for their activation (24, 25), we investigated the possi- binding (,5%) in a resting state but upregulate the ILT3 ligand Downloaded from bility that the inhibitory effect of ILT3Fc on these miRNAs is expression following activation (9). Thus, the downregulation of c- mediated by modulation of AP-1 activity. The proximal AP-1 site FOS and FOSB activity in primary CD8 T cells is observed at 3 h, of the miR-21 promoter and the distal NFAT/AP-1 site of the but not at earlier time points (Fig. 7B), likely because of the low human IL-2 promoter were engineered as three tandem copies and expression of ILT3 ligand on these cells at the time of activation. inserted into the pGL4.30 Firefly luciferase reporter construct. These results suggest that the inhibitory activity of ILT3Fc is

These constructs were transfected in Jurkat cells, which were ac- mediated through its effect on TCR signaling during very early http://www.jimmunol.org/ tivated with PMA (10 nM) in the presence or absence of different stages of T cell activation. To identify TCR signaling proteins, concentrations of ILT3Fc (1.5–15 mg/ml). Reporter genes con- which may be affected by ILT3Fc, we treated Jurkat cells and taining only AP-1 binding sites were strongly induced by PMA. primary CD8 T cells with increasing concentrations of ILT3Fc for ILT3Fc at 1.5–5 mg/ml consistently inhibited PMA-induced, 16 h and then probed cell lysates with Abs recognizing different AP-1–mediated reporter activities in a dose-dependent manner AP-1 pathway components. Normalization of the phosphorylated in three repeat experiments (Fig. 7A). fractions (p) of ZAP70 and phospholipase C (PLC)-g 1 by the by guest on September 27, 2021

FIGURE 6. ILT3Fc inhibits miRNA promoter activities. (A) miRNA gene promoters were cotransfected with Fireﬂy luciferase reporter plasmids containing miR-21 or miR-155 gene promoters and a control Renilla luciferase reporter construct into Jurkat cells. Mutation sites are indicated by “X.” Sites that are distal or proximal to the RNA start are indicated by “D” and “P,” respectively. Normalized luciferase activities in transfected Jurkat cells stimulated with CD3/CD28 mAbs or left untreated are shown. (B) ILT3Fc inhibits miR-21 and miR-155 promoter reporter activity. The promoter activity was tested in CD3/CD28-triggered Jurkat cells that had been pretreated with 5 mg/ml ILT3Fc. Error bars represent SE of three independent experiments. Differences between ILT3Fc-treated and nontreated cells are statistically signiﬁcant (p , 0.05). 3050 miRNA SIGNATURE OF ILT3-Fc–INDUCED CD8+ Ts Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 7. ILT3Fc inhibits basal TCR signaling and TCR-mediated nuclear localization of AP-1. (A) ILT3Fc inhibits PMA-induced AP-1–dependent transcriptional activity. PGL4.30 reporter constructs containing a minimal promoter and either the proximal AP-1 site of the miR-21 promoter or the distal AP-1 site of the human IL-2 promoter were transfected in Jurkat cells, which were activated with PMA in the presence or absence of different concentrations of ILT3Fc. Cotransfected PGL4.70 Renilla hRluc activity was used for normalization. Mean and SEM of three independent experiments are shown. *p , 0.05. (B) Western blot analysis of the AP-1 components FOSB and c-FOS in nuclear fractions of Jurkat, and primary CD8+ T cells stimulated with anti-CD3/CD28 mAbs in the presence or absence of ILT3Fc for the indicated time. (C) Dose-dependent inhibition of basal phosphorylation of Zap70 and PLC-g1 by ILT3Fc. SDS-PAGE analysis of Jurkat and primary CD8+ T cell lysates probed with Abs recognizing various phosphorylated (p) and total (T) fractions of TCR signaling proteins and DUSP10. (B) and (C) are representative of three independent experiments. respective total amounts (T) of these proteins shows that ILT3Fc We have demonstrated that similar to membrane, soluble ILT3 is inhibited the activation of these signaling molecules in a dose- a strong inducer of Ag specific CD8+ Ts (9–11). BCL6 is one of the dependent manner (Fig. 7C). We found that the dual-specific genes, which play a crucial role in the differentiation of ILT3- phosphatase DUSP10 was 2-fold higher in ILT3Fc-treated Jurkat induced Ts, because KD of this gene prevents Ts differentiation, T cells, similar to the results obtained in ILT3Fc-induced Ts. The whereas overexpression of BCL6 in primed CD8+ T cells endows finding that dephosphorylation of these TCR signaling proteins is them with suppressor activity (11). associated with the increased expression of DUSP10 suggests that Recently, the role of various miRNA in the differentiation of this dual-specific phosphatase modulates cell signaling pathways effector and Treg has been described previously (33–35). The aim in ILT3Fc-treated T cells. of our study was to identify miRNAs, which regulate the expression of signature genes induced by ILT3 in human CD8+ Ts cells. Discussion Our studies have demonstrated that miR-21, -30b, -146a, and -155 It is well established that inhibitory molecules expressed by APCs were downregulated in both membrane and soluble ILT3-induced have important roles in the induction of Treg/Ts (1). The ex- CD8+ Ts in conjunction with the upregulation of the genes whose pression of ILT3, a prototype of such inhibitory molecules, is 39-UTR they target (DUSP10, BCL6, CXCR4, and SOCS1). characteristically increased on the membrane of human tolero- Consistent with our study showing that the downregulation of genic DC and induces the differentiation of human CD8 and CD4 miR-21, and miR-155 results in the upregulation of genes essen- Treg. Adaptive Ts/Treg induce the upregulation of ILT3 on DC tial to Ts differentiation, such as DUSP10, and SOCS1, are recent which in turn trigger the differentiation of new waves of Ag- reports on the role of these two miRNAs in promoting autoimmune specific Ts/Treg (3, 4). inflammation and enhancing inflammatory T cell development (36, The Journal of Immunology 3051

37). miR-30 was also found to induce the downregulation of BCL6 Acknowledgments in B lymphocytes and lymphoma cells (38). The role of miR-146a is We thank Dr. Riccardo Dalla-Favera for excellent suggestions and Dr. Erik not as clearly defined, because miR-146a is upregulated in immune Flemmington for providing miR-155 promoter plasmids. cells in response to activation or maturation. although it is also + + involved in the differentiation of murine CD4 Foxp3 Treg (33). Disclosures Variations in the expression of miRNAs in natural and adaptive, The authors have no financial conflicts of interest. human and mouse Treg are expected to occur because gene profile analysis have unraveled numerous differences including the expression in human, but not mice, of KIR and ILT genes and of References FOXP3 not only in Treg but also in Th cells (39, 40). There is also 1. Vlad, G., R. Cortesini, and N. Suciu-Foca. 2005. License to heal: bidirectional poor concordance between humans and mice for lymphocyte interaction of antigen-specific regulatory T cells and tolerogenic APC. J. miRNA signature (41). Immunol. 174: 5907–5914. + 2. Cella, M., C. Do¨hring, J. Samaridis, M. Dessing, M. Brockhaus, A. Lanzavecchia, The miRNAs inhibited by ILT3 in primed CD8 T cells target and M. Colonna. 1997. A novel inhibitory receptor (ILT3) expressed on mono- BCL6, which is crucial to the differentiation of Ts and some other cytes, macrophages, and dendritic cells involved in antigen processing. J. Exp. Med. 185: 1743–1751. genes with known immunoinhibitory functions such as SOCS1, 3. Chang, C. C., R. Ciubotariu, J. S. Manavalan, J. Yuan, A. I. Colovai, F. Piazza, DUSP1, DUSP8, DUSP10, and CXCR4 (11, 12). Although SOCS1 S. Lederman, M. Colonna, R. Cortesini, R. Dalla-Favera, and N. Suciu-Foca. is a negative regulator of cytokine signaling (31, 33), DUSP10 acts 2002. Tolerization of dendritic cells by T(S) cells: the crucial role of inhibitory receptors ILT3 and ILT4. Nat. Immunol. 3: 237–243. on the MAPK pathways inhibiting TCR signaling (42). CXCR4, a 4. Manavalan, J. S., P. C. Rossi, G. Vlad, F. Piazza, A. Yarilina, R. Cortesini, receptor for the trafficking ligand CXCL12, was shown to con- D. Mancini, and N. Suciu-Foca. 2003. High expression of ILT3 and ILT4 is Downloaded from tribute to the induction of T cell immunotolerance in mice (43). a general feature of tolerogenic dendritic cells. Transpl. Immunol. 11: 245–258. 5. Vlad, G., C. C. Chang, A. I. Colovai, P. Berloco, R. Cortesini, and N. Suciu- CD8 T cells transfected with mixtures of miR inhibitors sup- Foca. 2009. Immunoglobulin-like transcript 3: a crucial regulator of dendritic pressed CD4 Th proliferation in response to CD3 triggering, cell function. Hum. Immunol. 70: 340–344. 6. Penna, G., A. Roncari, S. Amuchastegui, K. C. Daniel, E. Berti, M. Colonna, and suggesting that miRNAs targeting BCL6, SOCS1, and DUSP10 L. Adorini. 2005. Expression of the inhibitory receptor ILT3 on dendritic cells is must act in concert to induce the differentiation of Ts. dispensable for induction of CD4+Foxp3+ regulatory T cells by 1,25-dihydroxy- vitamin D3. Blood 106: 3490–3497. ILT3Fc inhibits miR-21 and miR-155 at the transcriptional level http://www.jimmunol.org/ 7. Adorini, L., and G. Penna. 2009. Induction of tolerogenic dendritic cells by in activated human T cells by downregulating AP-1 (24). ILT3Fc vitamin D receptor agonists. Handb. Exp. Pharmacol. 251–273. impaired the phosphorylation of Zap70 and PLC-g and increased 8. Chang, C. C., Z. Liu, G. Vlad, H. Qin, X. Qiao, D. M. Mancini, C. C. Marboe, the expression of DUSP10. This dual-specific phosphatase is R. Cortesini, and N. Suciu-Foca. 2009. Ig-like transcript 3 regulates expression of proinflammatory cytokines and migration of activated T cells. J. Immunol. likely to modulate signaling pathways in primed T cells, which 182: 5208–5216. acquire suppressor activity upon exposure to ILT3. This finding is 9. Kim-Schulze, S., L. Scotto, G. Vlad, F. Piazza, H. Lin, Z. Liu, R. Cortesini, and N. Suciu-Foca. 2006. Recombinant Ig-like transcript 3-Fc modulates T cell consistent with other studies, which demonstrated that DUSP10 responses via induction of Th anergy and differentiation of CD8+ T suppressor dephosphorylates MAPK and JNK, attenuating the effect of TCR cells. J. Immunol. 176: 2790–2798. ligation and reducing the production of proinflammatory cytokines 10. Vlad, G., V. D. D’Agati, Q. Y. Zhang, Z. Liu, E. K. Ho, T. Mohanakumar, M. A. Hardy, R. Cortesini, and N. Suciu-Foca. 2008. Immunoglobulin-like by guest on September 27, 2021 (44). Knockout of DUSP10 in mice leads to the overproduction of transcript 3-Fc suppresses T-cell responses to allogeneic human islet trans- inflammatory cytokines and accelerated immune-mediated death plants in hu-NOD/SCID mice. Diabetes 57: 1878–1886. after rechallenge with lymphocytic choriomeningitis virus (42). 11. Chang, C. C., G. Vlad, V. D. D’Agati, Z. Liu, Q. Y. Zhang, P. Witkowski, A. A. Torkamani, M. B. Stokes, E. K. Ho, R. Cortesini, and N. Suciu-Foca. 2010. The finding that ILT3Fc regulates miRNA transcriptionally by BCL6 is required for differentiation of Ig-like transcript 3–Fc-induced CD8+ T targeting TCR/AP-1 signaling pathways suggests that this agent suppressor cells. J. Immunol. 185: 5714–5722. 12. Vlad, G., J. King, C. C. Chang, Z. Liu, R. A. Friedman, A. A. Torkamani, and raises the threshold of TCR triggering. N. Suciu-Foca. 2011. Gene profile analysis of CD8+ ILT3-Fc induced T sup- A potential advantage of using ILT3Fc as an immunosuppres- pressor cells. Hum. Immunol. 72: 107–114. sive agent emerges from its capacity to target multiple inflam- 13. Yu, D., S. Rao, L. M. Tsai, S. K. Lee, Y. He, E. L. Sutcliffe, M. Srivastava, M. Linterman, L. Zheng, N. Simpson, et al. 2009. The transcriptional repressor matory miRNA pathways. Although each single miRNA acts Bcl-6 directs T follicular helper cell lineage commitment. Immunity 31: 457–468. simultaneously on hundreds of target genes, we found that inhi- 14. Hsu, H. C., P. Yang, J. Wang, Q. Wu, R. Myers, J. Chen, J. Yi, T. Guentert, bition of several inflammatory miRNAs is required to suppress A. Tousson, A. L. Stanus, et al. 2008. Interleukin 17-producing T helper cells and interleukin 17 orchestrate autoreactive germinal center development in au- T cell proliferation. Via simultaneously targeting the expression of toimmune BXD2 mice. Nat. Immunol. 9: 166–175. several inflammatory miRNAs, ILT3Fc seems ideally suited to 15. Kassiotis, G., and A. O’Garra. 2009. Establishing the follicular helper identity. Immunity 31: 450–452. induce a cascade of inhibitory signals resulting in T cell anergy. 16. Arima, M., H. Toyama, H. Ichii, S. Kojima, S. Okada, M. Hatano, G. Cheng, The strong immunosuppressive activity of ILT3Fc has been M. Kubo, T. Fukuda, and T. Tokuhisa. 2002. A putative silencer element in the documented by its ability to suppress allograft rejection (10, 11), a IL-5 gene recognized by Bcl6. J. Immunol. 169: 829–836. 17. Kusam, S., L. M. Toney, H. Sato, and A. L. Dent. 2003. Inhibition of Th2 dif- phenomenon in which myriads of MHC-bound peptides are re- ferentiation and GATA-3 expression by BCL-6. J. Immunol. 170: 2435–2441. cognized via the direct and indirect pathways, triggering rejection 18. Yoshida, K., A. Sakamoto, K. Yamashita, E. Arguni, S. Horigome, M. Arima, M. Hatano, N. Seki, T. Ichikawa, and T. Tokuhisa. 2006. Bcl6 controls granzyme (45). Therefore, it is most likely that recognition of a limited number + B expression in effector CD8 T cells. Eur. J. Immunol. 36: 3146–3156. of autoantigenic, tissue-specific peptides can be inhibited by ILT3Fc. 19. Guo, H., N. T. Ingolia, J. S. Weissman, and D. P. Bartel. 2010. Mammalian Information gained from studying the effect of ILT3Fc in the microRNAs predominantly act to decrease target mRNA levels. Nature 466: 835–840. differentiation of Ts may readily be translated into clinical bene- 20. Liston, A., L. F. Lu, D. O’Carroll, A. Tarakhovsky, and A. Y. Rudensky. 2008. fits for treatment of autoimmune diseases. Because exosomes can Dicer-dependent microRNA pathway safeguards regulatory T cell function. affect immune cells if they deliver a specific mRNA or miRNA that J. Exp. Med. 205: 1993–2004. 21. Zhou, X., L. T. Jeker, B. T. Fife, S. Zhu, M. S. Anderson, M. T. McManus, and can modify gene expression in a recipient cell (46, 47), it may J. A. Bluestone. 2008. Selective miRNA disruption in T reg cells leads to un- become possible to deliver Ts signature genes or inhibitors of controlled autoimmunity. J. Exp. Med. 205: 1983–1991. miR-21, -155, and -30b alone or in combination to generate CD8+ 22. Ye, B. H., S. Chaganti, C. C. Chang, H. Niu, P. Corradini, R. S. Chaganti, and R. Dalla-Favera. 1995. Chromosomal translocations cause deregulated BCL6 Ts. Alternatively, immune deficiencies may be treated by inter- expression by promoter substitution in B cell lymphoma. EMBO J. 14: 6209– fering with the generation and function of Ts, inhibiting their 6217. 23. Travagli, J., M. Letourneur, J. Bertoglio, and J. Pierre. 2004. STAT6 and Ets-1 signature genes by targeted delivery of miRNA or by blockade of form a stable complex that modulates Socs-1 expression by interleukin-4 in the ILT3 interaction with its T cell ligand. keratinocytes. J. Biol. Chem. 279: 35183–35192. 3052 miRNA SIGNATURE OF ILT3-Fc–INDUCED CD8+ Ts

24. Fujita, S., T. Ito, T. Mizutani, S. Minoguchi, N. Yamamichi, K. Sakurai, and 36. O’Connell, R. M., D. Kahn, W. S. Gibson, J. L. Round, R. L. Scholz, H. Iba. 2008. miR-21 gene expression triggered by AP-1 is sustained through A. A. Chaudhuri, M. E. Kahn, D. S. Rao, and D. Baltimore. 2010. MicroRNA- a double-negative feedback mechanism. J. Mol. Biol. 378: 492–504. 155 promotes autoimmune inflammation by enhancing inflammatory T cell 25. Yin, Q., X. Wang, J. McBride, C. Fewell, and E. Flemington. 2008. B-cell re- development. Immunity 33: 607–619. ceptor activation induces BIC/miR-155 expression through a conserved AP-1 37. Stagakis, E., G. Bertsias, P. Verginis, M. Nakou, M. Hatziapostolou, H. Kritikos, element. J. Biol. Chem. 283: 2654–2662. D. Iliopoulos, and D. T. Boumpas. 2011. Identification of novel microRNA 26. Lang, R., M. Hammer, and J. Mages. 2006. DUSP meet immunology: dual signatures linked to human lupus disease activity and pathogenesis: miR-21 specificity MAPK phosphatases in control of the inflammatory response. J. regulates aberrant T cell responses through regulation of PDCD4 expression. Immunol. 177: 7497–7504. Ann. Rheum. Dis. 70: 1496–1506. 27. Tzachanis, D., G. J. Freeman, N. Hirano, A. A. van Puijenbroek, M. W. Delfs, 38. Lin, J., T. Lwin, J. J. Zhao, W. Tam, Y. S. Choi, L. C. Moscinski, W. S. Dalton, A. Berezovskaya, L. M. Nadler, and V. A. Boussiotis. 2001. Tob is a negative E. M. Sotomayor, K. L. Wright, and J. Tao. 2011. Follicular dendritic cell- regulator of activation that is expressed in anergic and quiescent T cells. Nat. induced microRNA-mediated upregulation of PRDM1 and downregulation of Immunol. 2: 1174–1182. BCL-6 in non-Hodgkin’s B-cell lymphomas. Leukemia 25: 145–152. 28. Tzachanis, D., L. Li, E. M. Lafuente, A. Berezovskaya, G. J. Freeman, and 39. Vilches, C., and P. Parham. 2002. KIR: diverse, rapidly evolving receptors of V. A. Boussiotis. 2007. Twisted gastrulation (Tsg) is regulated by Tob and en- innate and adaptive immunity. Annu. Rev. Immunol. 20: 217–251. hances TGF-b signaling in activated T lymphocytes. Blood 109: 2944–2952. 40. Ziegler, S. F. 2006. FOXP3: of mice and men. Annu. Rev. Immunol. 24: 209–226. 29. Labbaye, C., I. Spinello, M. T. Quaranta, E. Pelosi, L. Pasquini, E. Petrucci, 41. Rossi, R. L., G. Rossetti, L. Wenandy, S. Curti, A. Ripamonti, R. J. Bonnal, M. Biffoni, E. R. Nuzzolo, M. Billi, R. Foa`, et al. 2008. A three-step pathway R. S. Birolo, M. Moro, M. C. Crosti, P. Gruarin, et al. 2011. Distinct microRNA comprising PLZF/miR-146a/CXCR4 controls megakaryopoiesis. Nat. Cell Biol. signatures in human lymphocyte subsets and enforcement of the naive state in 10: 788–801. CD4+ T cells by the microRNA miR-125b. Nat. Immunol. 12: 796–803. 30. Kim, Y. J., S. J. Hwang, Y. C. Bae, and J. S. Jung. 2009. miR-21 regulates 42. Zhang, Y., J. N. Blattman, N. J. Kennedy, J. Duong, T. Nguyen, Y. Wang, adipogenic differentiation through the modulation of TGF-b signaling in mes- R. J. Davis, P. D. Greenberg, R. A. Flavell, and C. Dong. 2004. Regulation of enchymal stem cells derived from human adipose tissue. Stem Cells 27: 3093– innate and adaptive immune responses by MAP kinase phosphatase 5. Nature 3102. 430: 793–797. 31. Palmer, D. C., and N. P. Restifo. 2009. Suppressors of cytokine signaling (SOCS) 43. Du¨rr, C., D. Pfeifer, R. Claus, A. Schmitt-Graeff, U. V. Gerlach, R. Graeser, in T cell differentiation, maturation, and function. Trends Immunol. 30: 592–602. S. Kru¨ger, A. Gerbitz, R. S. Negrin, J. Finke, and R. Zeiser. 2010. CXCL12 Downloaded from 32. Lu, C., X. Huang, X. Zhang, K. Roensch, Q. Cao, K. I. Nakayama, B. R. Blazar, mediates immunosuppression in the lymphoma microenvironment after alloge- Y. Zeng, and X. Zhou. 2011. miR-221 and miR-155 regulate human dendritic neic transplantation of hematopoietic cells. Cancer Res. 70: 10170–10181. cell development, apoptosis, and IL-12 production through targeting of p27kip1, 44. Owens, D. M., and S. M. Keyse. 2007. Differential regulation of MAP kinase KPC1, and SOCS-1. Blood 117: 4293–4303. signalling by dual-specificity protein phosphatases. Oncogene 26: 3203–3213. 33. Lu, L. F., M. P. Boldin, A. Chaudhry, L. L. Lin, K. D. Taganov, T. Hanada, 45. Liu, Z., A. I. Colovai, S. Tugulea, E. F. Reed, P. E. Fisher, D. Mancini, A. Yoshimura, D. Baltimore, and A. Y. Rudensky. 2010. Function of miR-146a in E. A. Rose, R. Cortesini, R. E. Michler, and N. Suciu-Foca. 1996. Indirect controlling Treg cell-mediated regulation of Th1 responses. Cell 142: 914–929. recognition of donor HLA-DR peptides in organ allograft rejection. J. Clin.

34. Rouas, R., H. Fayyad-Kazan, N. El Zein, P. Lewalle, F. Rothe´, A. Simion, Invest. 98: 1150–1157. http://www.jimmunol.org/ H. Akl, M. Mourtada, M. El Rifai, A. Burny, et al. 2009. Human natural Treg 46. Valadi, H., K. Ekstro¨m, A. Bossios, M. Sjo¨strand, J. J. Lee, and J. O. Lo¨tvall. microRNA signature: role of microRNA-31 and microRNA-21 in FOXP3 ex- 2007. Exosome-mediated transfer of mRNAs and microRNAs is a novel pression. Eur. J. Immunol. 39: 1608–1618. mechanism of genetic exchange between cells. Nat. Cell Biol. 9: 654–659. 35. Divekar, A. A., S. Dubey, P. R. Gangalum, and R. R. Singh. 2011. Dicer in- 47. Garchow, B. G., O. B. Encinas, Y. T. Leung, P. Y. Tsao, R. A. Eisenberg, sufﬁciency and microRNA-155 overexpression in lupus regulatory T cells: an R. Caricchio, S. Obad, A. Petri, S. Kauppinen, and M. Kiriakidou. 2011. Si- apparent paradox in the setting of an inﬂammatory milieu. J. Immunol. 186: 924– lencing of microR6-21 in vivo ameliorates autoimmune splenomegaly in lupus 930. mice. EMBO Mol. Med. 3: 605–615. by guest on September 27, 2021 Supplementary Table I.

A. Primers used for generation of reporter constructs

3’UTR reporter constructs BCL6: Forward: 5’-TGAAGCATGGAGTGTTGATGC-3’ Reverse: 5’-GCGGTAATGCAGTTTAGACACA-3’ SOCS1: Forward: 5’-GAGCTCTTCCCCTTCCAGATT-3’ Reverse: 5’-AAAATATAAAATAGGATTCTGCACAGC-3’ DUSP10: Forward: 5’-ATGCTCGAGTGACAATGGTCTGGATGGAA-3’ Reverse: 5’-CACAATCAACAGAAACACACCAAGA-3’

Promoter constructs: (underlines denote restriction site added for cloning into pGL3) miR-21: Forward: 5’-GCAGCTAGCTTTTCTAAGTTGCCCCAAGC-3’ Reverse: 5’-GCAAAGCTTTCCTCAGAGTAAGGTCAGCTC-3’ BCL6: Forward: 5’-TTTGCTAGCGTCGCTTGAAGGACTCTCATAGC-3’ Reverse: 5’-TCCAGATCTGCTAAATGCACAAAAGGGAGCG-3’ SOCS1: Forward: 5’-CGATGCTAGCAGTTTCTTCCGCAGCCGGGTAG -3’ Reverse: 5’-ACGTAAGCTTGCGCATGCTCCGGGGCCAGG-3’

B. Mutagenesis (underlines denote positions of mutant sequences)

AP-1 sites in the miR-21 promoter Site 1: Forward: 5’- CATTCTTTTTGGATAAGGATAACACCCAGATTGTCC-3’ Site 1: Reverse: 5’-GTCCTTATTAGGACAATCTGGGTGTTATCCTTATCC-3’ Site 2: Forward: 5’-TAGGGATGACACAAGCATAAACCCTTTCCTTATTAATTG-3’ Site 2: Reverse: 5’-GGTTTGAACCAATTAATAAGGAAAGGGTTTATGCTTG-3’ miR-30b recognition site in BCL6’s 3’UTR Forward: 5’-GTATTTTTTTGCAAGTGAAGGCCCACAATTTACAAAGTG-3’ Reverse: 5’-TTTTAATACACACTTTGTAAATTGTGGGCCTTCACTTGC-3’ miR-30b recognition site in SOCS1’s 3’UTR Forward: 5’-TCGCACCTCCTACCTCTTCATGGGGACATATACCCAG-3’ Reverse: 5’-GTGCAAAGATACTGGGTATATGTCCCCATGAAGAGG-3’ miR-155 recognition site in SOCS1’s 3’UTR Forward: 5’GCAGCGCCCGCCGTGCACGCAGCGGGAACTGGGATGCCG-3’ Reverse: 5’-CAAAATAACACGGCATCCCAGTTCCCGCTGCGTGCACG-3’