IFN-Stimulated Response Element Characterization of a New Subtype of Protein, and IFN Regulatory Factor-4

Home , Response element

IFN-Stimulated Gene 15 Is Synergistically Activated Through Interactions Between the Myelocyte/Lymphocyte-Specific Transcription Factors, PU.1, IFN Regulatory This information is current as Factor-8/IFN Consensus Sequence Binding of September 27, 2021. Protein, and IFN Regulatory Factor-4: Characterization of a New Subtype of IFN-Stimulated Response Element

David Meraro, Merav Gleit-Kielmanowicz, Hansjörg Hauser Downloaded from and Ben-Zion Levi J Immunol 2002; 168:6224-6231; ; doi: 10.4049/jimmunol.168.12.6224 http://www.jimmunol.org/content/168/12/6224 http://www.jimmunol.org/

References This article cites 40 articles, 27 of which you can access for free at: http://www.jimmunol.org/content/168/12/6224.full#ref-list-1

Why The JI? Submit online. by guest on September 27, 2021

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

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*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 © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

IFN-Stimulated Gene 15 Is Synergistically Activated Through Interactions Between the Myelocyte/Lymphocyte-Speciﬁc Transcription Factors, PU.1, IFN Regulatory Factor-8/IFN Consensus Sequence Binding Protein, and IFN Regulatory Factor-4: Characterization of a New Subtype of IFN-Stimulated Response Element1

David Meraro,2* Merav Gleit-Kielmanowicz,2* Hansjo¬rg Hauser,† and Ben-Zion Levi3*

Type I IFNs cause the induction of a subset of genes termed IFN-stimulated genes (ISGs), which harbor a speciﬁc DNA element, Downloaded from IFN-stimulated response element (ISRE). This ISRE confers the responsiveness to the IFN signal through the binding of a family of transcription factors designated IFN regulatory factors (IRFs). Some IRFs can bind to the DNA alone, such as IRF-1, which elicits transcriptional activation, or IRF-2, which leads to transcriptional repression. In addition, these factors associate with IRF-8/IFN consensus sequence binding protein (ICSBP), an immune cell-restricted IRF, and the assembled heterocomplexes lead to synergistic repression of ISRE elements. ISG15 is a prototype ISG that contains a well-characterized ISRE. Here we show that PU.1, an ETS member essential for myeloid/lymphoid cell differentiation, forms heterocomplexes with the immune-restricted http://www.jimmunol.org/ IRFs, IRF-8\ICSBP and IRF-4, which lead to transcriptional activation of ISG15. These data allowed the characterization of a subset of ISREs designated ETS/IRF response element (EIRE), which are differentially regulated in immune cells. EIREs are unique in their ability to recruit different factors to an assembled enhanceosomes. In nonimmune cells the factors will mainly include IRF members, while cell type-restricted factors, such as PU.1, IRF-8\ICSBP, and IRF-4, will be recruited in immune cells. IRF heterocomplex formation leads to transcriptional repression, and conversely, PU.1/IRFs heterocomplex formation leads to transcriptional activation. The fact that IRF-8\ICSBP is an IFN-␥-induced factor explains why some of the EIREs are also induced by type II IFN. Our results lay the molecular basis for the unique regulation of ISGs, harboring EIRE, in immune cells. The

Journal of Immunology, 2002, 168: 6224Ð6231. by guest on September 27, 2021

ype I IFNs (␣ or ␤) confer an antiviral state, affect cell the effects of IFN have been identified to date. Some of the IRFs, growth and differentiation, and modulate the immune sys- such as IRF-1, IRF-2, IRF-3, and IRF-7, bind directly to ISRE T tem. Type I IFNs initiate a signaling cascade that results elements, while others, such as IRF-4 and IRF-8/IFN consensus in the induction of a subset of genes termed IFN-stimulated genes sequence binding protein (ICSBP), do not bind effectively to the (ISGs).4 This induction is mediated by a specific DNA binding ISRE element alone (for review, see Ref. 2). However, IRF- elements designated IFN-stimulated response elements (ISREs), 8\ICSBP can associate with either IRF-1 or IRF-2, forming het- which are located at their promoters (1). The ISRE element serves erocomplexes that bind efficiently to the DNA (3, 4). Further, both as the binding site for a family of transcription factors termed IFN IRF-4 and IRF-8\ICSBP also associate with non-IRF members regulatory factors (IRFs). Nine cellular IRFs that execute part of such as PU.1 or E47 on a DNA composite element, half of which is an IRF binding site and half of which is the binding site of the *Department of Food Engineering and Biotechnology, Technion, Haifa, Israel; and interacting partner (5, 6). These protein-protein interactions are †Department of Gene Regulation and Differentiation, Gesellschaft fu¬r Biotechnolo- mediated through a conserved domain termed the IRF association gische Forschung, Braunschweig, Germany domain (IAD) (7). The IAD motif was found in all seven IRFs, Received for publication January 24, 2002. Accepted for publication April 17, 2002. excluding IRF-1 and IRF-2, which associate with IRF-8\ICSBP. In The costs of publication of this article were defrayed in part by the payment of page contrast, a domain rich in Pro, Glu, Ser, and Thr (PEST) mediates charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the interaction of PU.1,E47, IRF-1, and IRF-2 with IRF-8\ICSBP. 1 This work was supported by a grant from the German-Israeli Foundation for Sci- Further, the IAD and the PEST domains also mediate the interac- entific Research and Development (to B.-Z.L. and H.H.) and the Fund for the Pro- tions between IRF-4 and PU.1 or E47 (6Ð8). motion of Research at Technion (to B.-Z.L.). The expression of both IRF-4 and IRF-8\ICSBP is limited to 2 D.M. and M.G.-K. contributed equally to this work. immune cells (5, 9Ð11). Studies with knockout mice demonstrated 3 Address correspondence and reprint requests to Dr. Ben-Zion Levi, Department of that IRF-4 plays a crucial role in controlling the activation and Food Engineering and Biotechnology, Technion, Haifa 32000, Israel. E-mail address: [email protected] homeostasis of immune responses and is critical for proper matu- ration of B cells (12). IRF-8\ICSBP is essential for myeloid cell 4 Abbreviations used in this paper: ISG, IFN-stimulated gene; DBD, DNA binding domain; IRF, IFN regulatory factor; EICE, ETS/IRF composite element; EIRE, ETS/ differentiation toward mature macrophages (13). Both IRFs pos- IRF response element; IAD, IRF association domain; PEST, Pro, Glu, Ser, and Thr; sess a dual role in lymphoid cells; binding to ISRE elements leads ICSBP, IFN consensus sequence binding protein; ISRE, IFN-stimulated response element; IVT, in vitro transcription/translation; PKR, dsRNA-activated kinase; PRDI, to a repression of ISG expression, while interaction with lymphoid/ positive regulatory domain I. myeloid essential transcription factors (PU.1 and E47) leads to

transcriptional activation (5, 9Ð11). PU.1 is predominately ex- 3 ␮g. The expression plasmids coding for the various transcription factors pressed in cells of the hemopoietic lineage and is a critical regu- and reporter genes are indicated in the text. The cells were harvested 48 h lator of myeloid/lymphocyte differentiation. PU.1 can bind alone later and lysed using the lysis buffer with the Dual Luciferase assay kit (Promega), and luciferase activities were determined according to the man- to a core motif (A/GGAA) that is shared by most ETS proteins ufacturer’s instructions using a TD-20/20 luminometer (Turner Design; (14); however, additional flanking sequences define the ultimate Promega). Reporter gene activities were normalized for protein concentra- binding specificity: for example, the ETS/IRF composite element tion and transfection efficiencies as previously described (24). Western blot (EICE) to which IRF-4 and IRF-8\ICSBP bind following interac- analyses were performed with each transfected expression vector to ensure the expected level of ectopic protein expression (data not shown). Each set tion with PU.1. Such composite DNA motifs were identified in of transfection experiments was repeated at least three times, generating numerous genes essential for proper macrophage and B cells func- similar results. tion (15). U937 cells were diluted to 106 cells/ml 16Ð24 h before transfection. At The ISG15 gene, which contains ISRE element in its promoter, the time of transfection the cells were washed twice in PBS and suspended ϫ 7 is one of the most strongly induced proteins in cells following IFN at 2.5 10 cells/ml in RPMI medium lacking FCS and antibiotics. Part (0.4 ml) of the cell suspension was placed into a 0.4-cm electroporation type I treatment or viral infection (16Ð18). Deletion or mutations cuvette (Bio-Rad, Richmond, CA), and up to 30 ␮g plasmid DNA sus- in this ISRE element lead to the inability to respond to these sig- pended in Յ20 ␮l distilled water was added. Cells and plasmids DNA were nals (16). ISG15 is a 17-kDa ubiquitin-like protein produced as a incubated for 10 min at room temperature before electroporation at 500 ␮F precursor protein. Correct processing is essential for its conjuga- and 300 V (Bio-Rad gene pulser). Following the electric shock the cells were left at room temperature for an additional 15 min and then diluted into tion to cellular target and for its extracellular cytokine function 10 ml RPMI containing 10% FCS. Forty-eight hours after electroporation the (19, 20). ISG15 is secreted from monocytes, and the secreted form cells were harvested and analyzed as described above for NIH-3T3 cells. leads to NK proliferation and augmentation of non-MHC-re- Downloaded from stricted cytotoxicity. Thus, ISG15 may be considered a cytokine In vitro transcription and translation responsible for augmenting and amplifying the immunomodula- The assays were performed as described previously (7). Plasmids contain- tory effects of IFN-␣ or IFN-␤. ing the gene of interest under the T7 promoter were linearized downstream Here we show that ISG15 contains a unique ISRE element that of the coding region with the appropriate restriction enzyme. Five micro- grams of linearized plasmids were in vitro transcribed by T7 RNA poly- enables the binding not only of IRFs, but also of PU.1. This ISRE merase using a commercial kit (Stratagene, La Jolla, CA). Proteins were represents a subtype of ISREs that, in addition to IRFs, recruits translated in vitro using the rabbit reticulocyte lysate system (Promega) http://www.jimmunol.org/ PU.1. This composite ISRE element is synergistically activated by according to the manufacturer’s instructions. To monitor translation effi- 35 a heterocomplex formed between PU.1 and either IRF-8\ICSBP or ciency, small scale reactions containing [ S]methionine were performed each time, and the labeled proteins were separated on 10% SDS-PAGE and IRF-4. These interactions lead to specific regulation of ISG15 in subjected to autoradiography. immune cells, laying the molecular basis for its unique role in these cells. EMSA Nuclear extracts and gel-shift reactions were conducted as previously de- Materials and Methods scribed (24, 25). A typical reaction contained 1Ð5 ␮l IVT proteins or 5Ð8 ␮ Cell culture g nuclear extracts that were incubated in binding buffer (10 mM HEPES (pH 8), 5 mM MgCl2, 50 mM KCl, 0.025% bromophenol blue, 0.005% by guest on September 27, 2021 NIH-3T3, U937, Namalwa, and K562 cells were obtained from American xylene cyanole, 10% Ficoll, 3% glycerol, and 1 ␮g sonicated polyd(IC)) Type Culture Collection (Manassas, VA). NIH-3T3 were maintained in for 20 min. One microgram of sheared salmon sperm DNA was added only DMEM, while the other cell lines were maintained in RPMI 1640. All with IVT proteins. At least 50,000 cpm of the labeled probe (ISG15-ISRE) media were supplemented with 10% FCS, except for Namalwa cells, which was added for an additional 10 min on ice. The two synthetic oligonucle- were supplemented with 7.5% FCS. otides corresponding to the ISG15-ISRE (5Ј-GATCCTCGGGAAAGG GAAACCGAAACTGAAGCC-3Ј) were annealed and labeled with Kle- Plasmids now fragment. The samples were loaded on a pre-run 6.5% polyacrylamide gel. The dried gels were exposed to x-ray films. For supershift reaction, 1Ð2 Mammalian expression vectors, which are also suitable for in vitro tran- ␮l Abs was added, and the reactions were incubated for 60Ð90 min on ice scription (IVT), corresponding to IRF-1, IRF-4, PU.1, IRF-8\ICSBP, before addition of the labeled probe. Abs directed against PU.1 were pur- ICSBPL331P, and ICSBPS260A, were previously described (7, 21). ⌬ chased from BD PharMingen (San Diego, CA), and Abs against IRF-4 The plasmids pCB6PU.1 PEST and pCB6PU.1S148A were obtained were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti- from Dr. M. Atchison (University of Pennsylvania, Philadelphia, PA) (22). peptide Abs against IRF-8\ICSBP (anti-pepC) were obtained from Dr. K. Site-directed mutagenesis of tyrosine residues Y23 and Y48 within the Ozato. Antipeptide Abs against IRF-1 (pep300) and IRF-2 (pep311) were DNA binding domain (DBD) of IRF-8\ICSBP (ICSBPY23F and previously described (26). ICSBPY48F) to phenylalanine were generated using the plasmid pTarget- ICSBP (7) and the GeneEditor kit (Promega, Madison, WI) with the prim- ers Y23F (5Ј-GACAGTAGCATGTTTCCAGGCCTGATTTGGG-3Ј) and Results Y48F (5Ј-GGAAACACGCCGGCAAGCAAGATTTTAAT-3Ј). Both IRF-8\ICSBP and IRF-4 inhibit IRF-1-mediated

The plasmid p(ISRE)4-LUC, obtained from Dr. K. Ozato (National In- transcriptional activation of ISG15-IRES stitutes of Health, Bethesda, MD), is a reporter gene plasmid in which the luciferase gene (firefly) is driven by the thymidine kinase basal promoter ISGs are regulated by type I IFNs through the binding of IRFs to connected to four repeats of the ISG15-ISRE in the basic plasmid pBL-2 the ISRE located at their promoter region. IRF-1 is a prototype (Promega). In the plasmid p1.4Ld-LUC, 1.4 kb of the promoter region of IRF, which transcriptionally activates such promoters. We have d the L gene from the murine MHC class I is driving the luciferase (firefly) previously shown that IRF-8\ICSBP can associate with IRF-1, and reporter gene (obtained from Dr. K. Ozato, National Institutes of Health). To monitor transfection efficiency, a dual luciferase reporter assay was this association leads to repression of IRF-1 action on either the used with the plasmid pMDISRluc, in which the SV40 promoter drives the ISRE of MHC class I or the positive regulatory domain I (PRDI) expression of Renilla luciferase. element (a core sequence of ISRE) located on the promoter of IFN-␤ (24, 27). Here we show that the same effect of IRF-8\ICSBP DNA transfections and reporter gene analyses on IRF-1 also takes place on the ISRE element of ISG15. Fig. 1A NIH-3T3 cells were transfected by the calcium phosphate-DNA coprecipi- shows that IRF-1 can activate the expression of a reporter gene tation method as described previously (7, 23). Cells were plated in a six- driven by four ISRE repeats from the promoter of ISG15 (Fig. 1A, well dish and transfected with 400 ng to 1 ␮g of the various expression plasmids and reporter plasmids (firefly luciferase), 600 ng pMDISRluc lane 2). Cotransfection of IRF-8\ICSBP leads to repression (Fig. (SV40 promoter driving the expression of Renilla luciferase to monitor 1A, lane 8). Previously, we have shown that this repression is transfection efficiencies), and pUC19 serving as carrier DNA up to total of mediated through association between IRF-1 and the IAD module 6226 CHARACTERIZATION OF A NEW SUBTYPE OF ISRE

the repression activity of IRF-8\ICSBP (Fig. 1A, lanes 11 and 12, respectively). Fig. 1B shows that unlike IRF-8\ICSBP, IRF-4 elicits weak transcriptional activation on a reporter gene driven by ISG15-ISRE (Fig. 1B, lane 3). Cotransfection of IRF-1 and IRF-4 at equimolar levels does not lead to synergistic repression, but, rather, to a mid- level expression, which is probably a result of co-occupancy of the ISRE elements by both IRF-1 and IRF-4 (Fig. 1B, lane 4). Co- transfection of IRF-1 with IRF-8\ICSBP and IRF-4 leads to an additional decrease in the reporter gene level (Fig. 1B, lane 5). These results are further supported by EMSA, as shown in Fig. 1C. It is clear that unlike IRF-1, neither IRF-8\ICSBP nor IRF-4 binds to the ISRE element alone (Fig. 1C, lanes 2, 4, and 5, respectively). Yet IRF-1 and ICSBP form a heterocomplex (Fig. 1C, lane 3) that was supershifted with Abs directed against each of the interacting factors (data not shown). In contrast, no heterocomplex band was observed for IRF-1 and IRF-4 or IRF-8\ICSBP and IRF-4 (Fig. 1C, lanes 6 and 7, respectively). The same phenom- enon was also observed with IRF-2 (data not shown). These results Downloaded from are consistent with our previous data showing that only IRF- 8\ICSBP can interact with either IRF-1 or IRF-2, and this interaction leads to synergistic repression (7, 24). Such an effect was not detected with IRF-4. http://www.jimmunol.org/ PU.1 and either IRF-8\ICSBP or IRF-4 elicit synergistic activation of ISG15-ISRE The ISG15 expression is dysregulated in macrophages from IRF- 8ICSBP knockout mice, demonstrating its unique role in these cells (28). This prompted us to determine whether this dysregulated expression can be attributed to lack of heterocomplex formation between PU.1 and IRF-8\ICSBP, which are both essential for myeloid cell differentiation (14). To test this possibility, tran- sient transfection studies were performed with PU.1, IRF-8\ICSBP by guest on September 27, 2021 and their mutants (Fig. 2). Cotransfection of IRF-8\ICSBP and PU.1 leads to the activation of a reporter gene driven by the ISG15-ISRE (Fig. 2A, compare lanes 2 and 3 to lane 8). These FIGURE 1. The effects of IRF-1, IRF-8\ICSBP, and IRF-4 on the ex- results are in contrast to the results observed with IRF-1 (Fig. 1). pression of ISG15-ISRE-driven reporter plasmid and the formation of het- This activation is dependent on protein-protein interaction, be- erocomplexes. NIH-3T3 cells were transfected with the 600 ng p(ISRE)4- cause cotransfection of mutant IRF-8\ICSBP expression constructs LUC (A and B) or 400 ng IRF-1, IRF-8\ICSBP, IRF-4 expression plasmids, defective in the IAD (mutation of the conserved leucine 331 or the or various IRF-8\ICSBP mutant constructs as indicated (for details see nonconserved serine 260) leads to elimination of this transcrip- Materials and Methods). The luciferase activity of the reporter gene alone tional synergy (Fig. 2A, compare lane 8 to lanes 9 and 10). In was normalized to a value of 1. Relative luciferase activities of the various addition, this interaction is dependent on the presence of tyrosine transfections were calculated accordingly as fold activation relative to this residues 23 and 48 in IRF-8\ICSBP, which are conserved in IRF-4. value. This experiment was repeated four times with similar results. Error Mutating these residues to phenylalanine diminished the synergis- bars denote the SD. C, The ability of IVT IRF-1 and IRF-8\ICSBP to bind 32P-labeled ISG15-ISRE oligonucleotide and to form heterocomplexes was tic activity with PU.1 (Fig. 2A, lanes 11 and 12). It is possible that analyzed by EMSA. Arrowheads indicate the binding of IRF-1 or IRF-1/ the phosphorylation state of these residues is essential for the in- ICSBP heterocomplexes. teraction with only PU.1, because these mutations on tyrosine residues had no effect on the interaction with IRF-1 (Fig. 1) and IRF-2 (data not shown). This suggests that the interaction of PU.1 with of IRF-8\ICSBP. This was demonstrated by mutating the con- IRF-8\ICSBP is dependent upon its intact IAD as well as DBD. served leucine 331, which is located in the IAD in a predicted Synergistic activation of the ISG15-driven reporter gene by IRF- ␣-helix structure to proline. This point mutation interferes with this 8ICSBP and PU.1 was also observed with the promonocytic cell structured domain, leading to a mutant IRF-8\ICSBP incapable of line U937, which constitutively expresses these two factors (data interacting with other factors (7). It is clear from Fig. 1A that this not shown). To avoid endogenous background expression of these mutation also affects the ability of IRF-8\ICSBP to repress IRF-1 genes, which may interfere with the ectopic expression, additional on the ISG15-ISRE element (Fig. 1A, compare lane 9 to lane 2). experiments were performed with NIH-3T3 cells, which do not Further, we have also shown that serine 260 in IRF-8\ICSBP is express these genes. essential for protein-protein interaction (21). Similar results are Previous studies demonstrated that association of PU.1 with ei- demonstrated here; mutation of this serine residue within the IAD ther IRF-4 or IRF-8\ICSBP is mediated through a PEST domain, of IRF-8\ICSBP eliminates its ability to negate IRF-1 (Fig. 1A, and the phosphorylation of serine 148 is crucial for this association lane 10). In contrast, mutating tyrosine residues within the DBD of (6Ð8). The data in Fig. 2B clearly show that the synergistic effect IRF-8\ICSBP to phenylalanine (Y48F and Y23F) have no effect on of PU.1 and IRF-8\ICSBP on the ISRE of ISG15 is lost when The Journal of Immunology 6227

either serine 148 is mutated to alanine or the whole PEST domain is deleted (Fig. 2B, compare lanes 6, 7, and 8, respectively). IRF-4 was the ﬁrst IRF member that demonstrated transcriptional synergy with PU.1 on EICE. Therefore, we tested whether such a synergistic effect also takes place on the classical ISRE element of ISG15. The data in Fig. 2C clearly show a synergistic effect of PU.1 and IRF-4 on the expression of an ISRE-driven reporter gene (Fig. 2C, lane 5). This synergistic effect is also dependent upon intact IRF-4-IAD, because mutation of the conserved leucine 386 to proline (corresponding to leucine 331 in IRF-8\ICSBP), which interferes with the predicted ␣-helix structure, also results in a lack of synergistic activity of this mutant IRF-4 with PU.1 (Fig. 2C, lane 6). To determine whether the synergistic effect between PU.1 and IRF-8\ICSBP or IRF-4 is unique to the ISRE of ISG15, we also tested the MHC class I promoter containing characterized ISRE and the PRDI element from the IFN-␤ promoter (10, 27). No synergistic effect between PU.1 and IRF-8\ICSBP was observed on a luciferase reporter gene driven by either MHC class I ISRE (Fig. Downloaded from 2D) or the PRDI element of IFN-␤ (data not shown). These results suggest that the observed synergistic activity of PU.1 is not common to all ISREs.

The ISRE of ISG15 supports association between PU.1 and

either IRF-8\ICSBP or IRF-4 http://www.jimmunol.org/ The above data demonstrate that the ISRE of ISG15 is synergistically activated by PU.1 and either IRF-8\ICSBP or IRF-4. To demonstrate the physical interaction between these factors, EMSA was performed. PU.1 as well as IRF-8\ICSBP and various mutants of IRF-8\ICSBP were in vitro-translated and reacted with the labeled ISRE element of ISG15. Neither PU.1 nor IRF-8\ICSBP binding to this ISRE element could be detected in this assay (Fig. 3A, lanes 2 and 3, respectively). However, when the two proteins by guest on September 27, 2021 were mixed together a new band corresponding to the heterocomplex band was observed (Fig. 3A, lane 4). This heterocomplex band was supershifted with Abs directed against IRF-8\ICSBP and was eliminated by Abs against PU.1 (data not shown). This interaction was dependent upon intact IRF-8\ICSBP-IAD, because the mutations in leucine 331 to proline and serine 260 to alanine lead to either loss of the heterocomplex band (Fig. 3A, lane 11, for ICSBPL331P) or a signiﬁcant reduction in its intensity (Fig. 3A, lane 10, for ICSBPS260A). The EMSA results with ICSBP-IAD are in agreement with the cotransfection studies, which demonstrated that these mutants lost their ability to synergize with PU.1 (Fig. 2). We also tested mutants of ICSBP in tyrosine residues within the DBD, which are conserved in IRF-4 (Y23F and Y48F). It is clear that mutation at tyrosine 48 lead to reduced interaction as demonstrated by a weaker heterocomplex band, while mutation at tyrosine 23 leads to increased interaction (Fig. 3A, lanes 8 and 6, respectively). Fig. 3B demonstrates that a similar heterocomplex band is formed between PU.1 and IRF-4 (lane 3) on the ISRE from the ISG15 promoter. This band was not detected with IRF-4L368P

FIGURE 2. The synergistic activation of ISG15-ISRE by PU.1, IRF- LUC was assayed. B, The ability of PU.1 and mutant PU.1 expression 8ICSBP, and IRF-4 is dependent upon intact IAD and PEST domains. constructs, defective in their PEST domains as indicted in the text, to NIH-3T3 cells were transfected as described in Fig. 1, A and B, with the synergize with IRF-8\ICSBP was tested by cotransfection studies with the indicted expression plasmids. The luciferase activity of the reporter gene reporter plasmid p(ISRE)4-LUC. C, The synergistic activity of PU.1 and alone was normalized to a value of 1, and relative luciferase activities of IRF-4 on p(ISRE)4-LUC reporter plasmid was compared with that of mu- the various transfections were calculated accordingly as fold activation tant IRF-4 L386P. D, The ability of IRF-1, PU.1 IRF-8\ICSBP, or their relative to this value. The experiments were repeated four times with sim- combination, as indicated, to affect the activity of a luciferase reporter gene ilar results. Error bars denote the SD. A, The ability of PU.1 and either driven by the MHC class I ISRE (pLd1.4-LUC) was analyzed as described

IRF-8\ICSBP or mutated constructs of IRF-8\ICSBP to activate p(ISRE)4- in Fig. 1, A and B. 6228 CHARACTERIZATION OF A NEW SUBTYPE OF ISRE Downloaded from

FIGURE 3. IRF-8\ICSBP or IRF-4 can form heterocomplexes with PU.1 on the ISRE of ISG15. EMSAs were conducted with IVT factors. A, FIGURE 4. Multimeric complexes containing PU.1, IRF-8\ICSBP, PU.1 and IRF-8\ICSBP or IRF-8\ICSBP mutant constructs or their com- IRF-4, and IRF-2 are detected in U937 and Namalwa, but not in K562, cell bination as indicated. B, PU.1, IRF-4, and mutant IRF-4L386 or their com- extracts. EMSAs were performed with indicated cell extracts using 32P- 32 bination as indicated. The ability of the various factors to bind P-labeled labeled ISG15-ISRE oligonucleotide as a probe. The presence of the indi- http://www.jimmunol.org/ ISG15-ISRE oligonucleotide and to form heterocomplexes was analyzed. cated factors (IRF-8\ICSBP, PU.1, IRF-4, IRF-2, and IRF-1) in some of the Arrowheads indicate the binding of PU.1/ICSBP or PU.1/IRF-4 hetero- observed bands was determined by supershift analysis with the indicated complexes. NS, Nonspecific band. Abs (for details see Materials and Methods). Arrowheads indicate the position of either PU.1 or IRF-2 and the position of the putative heterocomplex bands marked in roman numerals. Pre, Rabbit preimmune serum. mutant (Fig. 3B, lane 4). Further, no heterocomplex was detected when IRF-8\ICSBP and IRF-4 were reacted with the ISRE of ISG15 (Fig. 3B, lane 5). Interestingly, when all three proteins were shifted band, while Abs directed against PU.1 resulted in band elim- mixed together, a much stronger heterocomplex band was ob- ination (Fig. 4, upper panel, compare lanes 2, 4, and 5 to lane 3). In served (Fig. 3B, lane 6). This suggests that the heterocomplex contrast, Abs directed against IRF-1 did not lead to any change in the by guest on September 27, 2021 might be composed of all three components (PU.1, IRF-8\ICSBP, band pattern compared with that in the control (Fig. 4, upper panel, and IRF-4) in an unknown stoichiometry. compare lane 6 to lane 1), but, rather, led to a stronger signal. Similar No heterocomplex bands were observed with PU.1, IRF-8ICSBP, results were observed with Namalwa cell extract (Fig. 4, middle pan- and IRF-4 when the ISRE of MHC class I or the PRDI element of el). When K562 cell extract, which does not express IRF-8\ICSBP IFN-␤ was used as a probe in the EMSA (data not shown). and IRF-4, was used, there was no effect on the bands pattern with Abs against IRF-8\ICSBP, IRF-4, PU.1, and IRF-1. A supershifted Interaction between PU.1 and either IRF-8\ICSBP and/or IRF-4 band is clearly seen with Abs directed against IRF-2, resulting in a is also detected in hemopoietic cells significant reduction in the intensity of band II. Altogether these re- We next wanted to determine whether heterocomplexes between sults demonstrate that all the factors, PU.1, IRF-8\ICSBP, IRF-4, and PU.1 and either IRF-8\ICSBP and IRF-4 also occur in vivo. For IRF-2 (with the exception of IRF-1), interact with the ISRE of ISG15. that purpose, nuclear extracts were prepared from the promono- These interactions are dependent upon the composition of transcrip- cytic cell line U937 and the B lymphocyte cell line Namalwa, tion factors in the cells, because such interactions were not observed which both constitutively express all three factors. In addition, in K562 cell extract. nuclear extract was prepared from the erythroleukemia cell line Our data suggest that the classical ISRE element of ISG15 can K-562, which expresses PU.1, but does not express IRF-8\ICSBP be engaged not only through the direct binding of some IRF mem- or IRF-4. Five major bands (I, II, II, IRF-2, and PU.1) were ob- bers or through the interaction between IRF-8\ICSBP and either served in EMSA using labeled ISRE from ISG15 and U937 cell IRF-1 or IRF-2 (Fig. 1C), but also through interactions with non- extract (Fig. 4, upper panel). Abs against either PU.1 or IRF-2 IRF members, e.g., PU.1. The core motif for PU.1 DNA binding is enabled us to detect the bands corresponding to the binding of only GGAA (29), and the ISRE of ISG15 contains two GGAA motifs. these factors compared with cell extract reacted with preimmune To define the region essential for the interaction observed in U937 serum (Fig. 4, upper panel, lanes 3, 5, and 1, respectively). In cells, five mutant oligonucleotides (illustrated in Fig. 5A) were addition, these two Abs led to a reduction in the intensity of all the synthesized and used as competitors in the EMSA. It is clear that three upper bands (I, II, and III). However, the decrease in the inten- the M1 oligonucleotide, in which the 5Ј-positioned GGAA motif sity of band II was most prominent. This suggested that the band was mutated, competed for the binding as well as the wild-type marked as II might represent the major heterocomplex bands between oligonucleotide (Fig. 5B, compare lanes 2 and 3 to lane 1). In PU.1 and either IRF-8\ICSBP and IRF-4 or the heterocomplex band contrast, the M2 oligonucleotide, in which the 3Ј-positioned between IRF-2 and only IRF-8\ICSBP. Accordingly, the addition of GGAA motif was mutated, did not compete for the binding (Fig. Abs directed against IRF-8\ICSBP and IRF-4 mainly eliminated this 5B, lane 4). Like the M2 oligonucleotide, the M5 oligonucleotide, band (band II, Fig. 4, upper panel, lanes 2 and 4, respectively). Abs in which both putative PU.1 binding sites were mutated, did not against IRF-2, IRF-8\ICSBP, and IRF-4 led to the appearance of a elicit any competition. This suggests that the 3Ј-positioned GGAA The Journal of Immunology 6229

because its binding was not lost with any of the mutant oligonucleotides. Taken together our data suggest that the ISG15-ISRE allows not only the binding of IRF members, but also the binding of PU.1, and should be considered a DNA composite element.

Discussion Exposure of cells to IFN type I leads to the induction of many genes. Friedman and Stark (30) were the first to note that these genes are characterized by a specific DNA motif at their promoter region, which they termed IFN consensus sequence. This DNA motif was further characterized, and the core region was refined and renamed ISRE (31). The ISRE of ISG15 is considered a prototypic element that is regulated through the binding of various IRFs. The ability of IRF-1, IRF-3, and IRF-9/ISGF3␥ (in the context of ISGF3) to bind this element was demonstrated by EMSA, FIGURE 5. Competition effect of mutant ISG15-ISRE oligonucleotides and the effect on transcriptional regulation was shown in many cell on the binding pattern of ISG15-ISRE in EMSA with U937 cell extract. A, types and tissues (16, 32, 33). Further, while using this ISRE it was Various mutant ISRE (double-stranded) oligonucleotides were synthesized shown that both IRF-1 and IRF-2 can associate with IRF-8\ICSBP as illustrated. Mutant oligonucleotides M1 and M2 harbor different muta- to form heterocomplexes, and these complexes possess repression Downloaded from tions in the two putative PU.1 GGAA binding sites. The M5 mutant oli- activity in immune cells (3, 4). gonucleotide harbors mutations in both PU.1 binding sites. The M3 and M4 In this communication we provide evidence that the ISRE of mutant oligonucleotides harbor mutations at the putative two IRF binding ISG15 is synergistically activated by the myeloid/lymphocyte-re- sites as illustrated. B, EMSAs were performed with U937 cell extracts, stricted factors PU.1 and either IRF-8\ICSBP or IRF-4. This syn- which were reacted with a 50-fold molar excess of the various mutant oligonucleotides before the incubation with the 32P-labeled wild-type ergistic activity is further supported by the EMSA results showing ISG15-ISRE. Arrowheads indicate the positions of the putative heterocom- that heterocomplexes between these factors are detectable with in http://www.jimmunol.org/ plex bands marked in roman numerals. Dash indicates competition assay vitro-translated proteins or in nuclear extract from cell lines of with a nonrelevant oligonucleotide from the c-Fos promoter. monocytic and lymphocytic origin. These results demonstrate that this ISG15-ISRE element is able to recruit not only IRFs, but also a non-IRF member such as PU.1, which belong to the ETS family motif, which is closer to the IRF binding site, is essential for pro- of transcription factors. Interaction among IRF-8\ICSBP, IRF-4, tein-protein interaction. The M3 and the M4 oligonucleotides were and PU.1 was reported for numerous enhancer elements in many mutated in the two possible IRF binding sites, ACCGAAA and genes that are essential for proper biological activities of either ACTGAA. The data show that only partial competition was macrophages or B cells. These interactions occur on EICE, which achieved with each oligonucleotide, suggesting that both are es- were identified on enhancer elements of Ig L chain, CD20, IL-1␤, by guest on September 27, 2021 sential for IRF binding. Toll-like receptor 4, and more (for review, see Ref. 15). This is To complement these competition experiments, each mutant oli- consistent with the restricted expression of these factors in cells of gonucleotide was labeled and used as a probe in EMSA, and the myelocyte/lymphocyte origin. Our results clearly show that the band pattern was compared with the band pattern seen with the classical ISG15-ISRE can also recruit PU.1 and therefore can also labeled ISG15-ISRE oligonucleotide. The data summarized in Ta- be defined as EICE. Sequence alignment of numerous ISRE ele- ble I demonstrate that the M2 mutant had the strongest effect on ments displayed in Table II clearly shows that classical ISREs can PU.1 binding and the interaction with IRF-8\ICSBP and IRF-4 be divided into two categories. The first category (subtype) con- (similar to the pattern seen with M5). Further, when M4 was used sists of ISRE elements that recruit only IRFs (Table II, upper sec- as a probe, no heterocomplex band with IRF-8\ICSBP and IRF-4 tion). The lower part of Table II shows classical EICE elements in was observed, although the binding of IRF-2 was still detectable. which the PU.1 binding site (GGAA) is shaded, and the IRF bind- These experiments show that the 3Ј-positioned GGAA and the 3Ј IRF ing site is in the boxed area adjacent to the PU.1 element. The binding site (AACTGA) are essential for the formation of heterocom- middle part of Table II shows classical ISREs from many ISGs plexes between PU.1 and either IRF-8\ICSBP or IRF-4 on the ISRE that display sequence similarity to the ISG15-ISRE. The PU.1 element of ISG15. Both IRF binding sites were able to recruit IRF-2, binding site is shaded, and the IRF binding site is in the boxed area. The formation of heterocomplexes between IRF-8ICSBP and either IRF-1 or IRF-2 was reported for ISG15 such as ISREs a Table I. EMSA with mutated ISG15-ISRE oligonucleotides (ISG15 and ISG54). This may reflect the difference between classical EICEs, which are characterized by the formation of hetero- Heterocomplexes Heterocomplexes complexes with only PU.1 and either IRF-8\ICSBP or IRF-4. The Oligo IRF-2 PU.1 Containing ICSBP Containing IRF-4 major difference between these two subtypes is an extra adenosine ISG15 ϩϩϩϩ ϩϩϩϩ ϩϩϩϩ ϩϩϩ residue just downstream from the PU.1 binding site. The extra base M1 ϩϩϩϩ ϩϩ ϩϩ ϩ results in the formation of two IRF binding sites characterized by ϩϩ Ϫ Ϫ Ϫ M2 GAAANN. This explains the difference between EICEs, which M3 ϩϩϩ ϩϩϩ ϩ ϩ M4 ϩϩ ϩϩϩ Ϫ Ϫ have only one PU.1 binding site and one IRF binding site and M5 ϩϩϩϩ Ϫ Ϫ Ϫ ISG15 such as ISREs that contain, in addition to the PU.1 binding site, two putative ISREs. This fits with the fact that these elements a EMSAs were performed with U937 cell extracts, which were reacted with 32P- labeled ISG15-ISRE oligonucleotide or mutant oligonucleotide as indicated in the allow the formation of IRFs heterocomplexes as well as PU.1/IRF presence of Abs as described in Fig. 4. The intensities of the bands corresponding to heterocomplexes, thus designated here as ETS/IRF response ele- the binding of IRF-2, PU.1, or the heterocomplex bands containing IRF-8/ICSBP or IRF-4 and the super-shifted bands were evaluated. ϩϩϩϩ, Strong binding; ϩϩϩ, ment (EIRE). The data presented here support this duality of the moderate binding; ϩϩ, weak binding; ϩ, very weak binding; Ϫ, no binding. ISRE of ISG15 with respect to heterocomplex formation. Similar 6230 CHARACTERIZATION OF A NEW SUBTYPE OF ISRE

Table II. Alignment of various ISRE and EICE elementsa 10 is induced and secreted by many cell types. Costimulation of macrophages by LPS and IFN-␥ leads to its expression and subsequent effect on Th1 cells harboring the target receptor, CXCR3 (37). The 9 Ð27 gene is also a prototypic ISG that is expressed in many cell types as a cell surface protein. However, it has a unique role in B cells, being a component of a larger B cell CD81 complex involved in signal transduction and cell adhesion (38). The fact that these EIRE elements can recruit IRF-8\ICSBP or IRF-4 suggests that these elements will respond not only to IFN type I, but also to IFN-␥, which leads to transcriptional induction of IRF-8\ICSBP and to antigenic stimulation resulting in the induction of both IRF-8\ICSBP and IRF-4. Hence, PKR, IFN-inducible protein 10, 9Ð27, CYBB, and NCF2 are also induced by IFN-␥ or antigenic stimulation or both (34, 36Ð38). We show that the interaction of PU.1 with either IRF-8\ICSBP or IRF-4 is dependent upon intact interaction modules, e.g., PEST and IAD, respectively. Our data also show that serine 260, which is unique to the IAD of IRF-8\ICSBP, is essential for this interaction, suggesting that the phosphorylation state of this residue is Downloaded from essential (for more details see Ref. 21). Tyrosine phosphorylation is also essential for proper activity of IRF-8\ICSBP. Speciﬁc tyrosine phosphorylation prevents ICSBP from binding alone to target DNA; nevertheless, this phosphorylation is essential for efﬁ- cient protein-protein interaction (24). Here we show that mainly

tyrosine residue 48, which is shared only with IRF-4, affects the http://www.jimmunol.org/ ability of IRF-8\ICSBP to synergize and form an efficient heterocomplex with PU.1. Tyrosine residue 23, which is also conserved in IRF-4, IRF-5, and IRF-7, was not essential for the formation of heterocomplex with PU.1. These mutations of tyrosine residues in the DBD affect only the interaction with PU.1 and not that with IRF-1. These results support our published data demonstrating that the presence of target DNA is necessary for an efficient interaction of IRF-8\ICSBP with PU.1, but not with IRF-1 (7). In addition, tyrosine residues 110, within the DBD, and tyrosine residue 211, by guest on September 27, 2021 within the IAD, are conserved in all IRFs, and mutating it to phenylalanine leads to a defective IRF-8\ICSBP that does not interact effectively with either IRF-1/2 and PU.1 (D. Meraro and B. Z. Levi, unpublished observations) (39). These results suggest that a The various ISREs, EIREs and EICEs were aligned. The PU.1 binding site GGAA is shaded. The IRF binding site(s) are boxed. TLR4, Toll-like receptor 4; modulation of the phosphorylation state of some of the tyrosine ␤ ␤ IP-10, IFN-inducible protein 10; 2m, 2 microglobulin; GBP, guanylate binding residues in both the DBD and the IAD might have a role in the protein; OAS, oligo-adenylate synthetase. ability of the modified factor to form DNA heterocomplexes, thus affecting gene regulation. Accordingly, SHP1 protein tyrosine phosphatase inhibits the ability of IRF-8\ICSBP to interact with PU.1 and IRF-1, leading to reduced expression of the myeloid- results were obtained with regulatory elements from CYBB and specific genes CYBB (gp91phox) and NCF2 (p67phox) (39). NCF2 genes, which encode for the phagocyte respiratory burst The multicomplexes formed on the EIRE elements of CYBB oxidase subunits gp91phox and p67phox, respectively. These results and NCF2 are composed of PU.1, IRF-8\ICSBP, and IRF-1 and suggest that the CYBB and NCF2 enhancer elements are EIRE and also recruit the cAMP response element binding protein (34). The as such can be engaged by IRFs heterocomplexes as well as latter harbors histone acetyltransferase activity (40). Here we show PU.1IRFs heterocomplexes as reported recently (34). that the ISRE element of ISG15 recruits PU.1, IRF-8\ICSBP, The ability of the EIRE DNA motifs to recruit not only IRFs, but IRF-4, and IRF-2. However, in all these studies the stoichiometry also PU.1/IRF heterocomplexes, suggests that some of the genes of the multisubunit complexes is not clear. In vitro studies showed harboring such elements might be regulated differentially in im- that these EIRE can engage both IRFs heterocomplexes (e.g., IRF- mune cells. Consequently, some of the genes listed in Table II as 8ICSBP/IRF-2 or IRF-8\ICSBP/IRF-1) and EtsIRFs heterocom- harboring EIREs also possess a unique role in immune cells. For plexes (e.g., PU.1/IRF-8\ICSBP or PU.1/IRF-4). EMSA per- example, the induced expression of ISG15 by type I IFNs in many formed with nuclear extracts prepared from U937 cells or cell types and tissues is well documented. However, ISG15 is also Namalwa cells did not allow us to distinguish between these two secreted from monocytes and lymphocytes and modulates the im- options. When in vitro-translated proteins were reacted in EMSA mune system. Its ability to activate NK cells is dependent on cocul- it was clear that the interaction with the DNA was much stronger tivation with T cells (19). Similarly, the induced expression of the when PU.1, IRF-8ICSBP, and IRF-4 were incorporated (Fig. 3B, IFN-induced and dsRNA-activated kinase (PKR) in many cell types lane 6). This suggests that such multicomplexes are formed and tissues, resulting in antiviral and antiproliferative effects, is well through interaction between IADs located on both IRF-8\ICSBP established (35). Yet, recent studies also implicated PKR in the and IRF-4 and PEST domains that were identified on PU.1, IRF-1, growth control of mature T lymphocytes (36). IFN-inducible protein and IRF-2. Our results presented here demonstrate that both The Journal of Immunology 6231 modules (IAD and PEST) were essential for association and tran- 14. Lloberas, J., C. Soler, and A. Celada. 1999. The key role of PU.1/SPI-1 in B cells, scriptional synergy. Recently, a direct role for IRF-8\ICSBP in the myeloid cells and macrophages. Immunol. Today 20:184. 15. Marecki, S., and M. J. Fenton. 2000. PU.1/interferon regulatory factor interac- regulation of ISG15 was demonstrated in IRF-8\ICSBP knockout tions: mechanisms of transcriptional regulation. Cell Biochem. Biophys. 33:127. mice (28). It was demonstrated that the heterocomplex formed 16. Reich, N., B. Evans, D. Levy, D. Fahey, E. Knight, and J. E. Darnell, Jr. 1987. Interferon-induced transcription of a gene encoding a 15 kDa protein depends on between IRF-2, IRF-8ICSBP, and IRF-4 was detected in extracts an upstream enhancer element. Proc. Natl. Acad. Sci. USA 84:6394. from B cells, while a heterocomplex between IRF-4 and IRF- 17. Der, S. D., A. Zhou, B. R. Williams, and R. H. Silverman. 1998. Identification of 8\ICSBP was detected in cell extracts of macrophages. However, genes differentially regulated by interferon ␣, ␤,or␥ using oligonucleotide ar- rays. Proc. Natl. Acad. Sci. USA 95:15623. the presence of PU.1 in the heterocomplexes was not considered. 18. Farrell, P. J., R. J. Broeze, and P. Lengyel. 1979. Accumulation of an mRNA and In this communication we provide evidence showing that PU.1 is protein in interferon-treated Ehrlich ascites tumour cells. Nature 279:523. also a major component in the multicomplex formed on the ISRE 19. D’Cunha, J., E. J. Knight, A. L. Haas, R. L. Truitt, and E. C. Borden. 1996. Immunoregulatory properties of ISG15, an interferon-induced cytokine. Proc. of ISG15, which is essential for transcriptional activation. Natl. Acad. Sci. USA 93:211. In conclusion, our data indicate that EIREs are unique in their 20. Potter, J. L., J. Narasimhan, L. Mende-Mueller, and A. L. Haas. 1999. Precursor ability to recruit different factors in an assembled enhanceosomes. processing of pro-ISG15/UCRP, an interferon-␤-induced ubiquitin-like protein. J. Biol. Chem. 274:25061. In nonimmune cells the factors will mainly include IRF members. 21. Cohen, H., A. Azriel, T. Cohen, D. Meraro, S. Hashmueli, D. Bech-Otschir, In immune cells these EIRE elements will also recruit non-IRFs, R. Kraft, W. Dubiel, and B. Z. Levi. 2000. Interaction between interferon consensus sequence-binding protein and COP9/signalosome subunit CSN2 (Trip15): such as PU.1, to the assembled enhanceosome. Thus, EIREs may a possible link between interferon regulatory factor signaling and the COP9/ affect the transcriptional capacity of different enhanceosomes in a signalosome. J. Biol. Chem. 275:39081. cell type-restricted manner that is dependent upon the milieu of 22. Fisher, R. C., M. C. Olson, J. M. Pongubala, J. M. Perkel, M. L. Atchison,

E. W. Scott, and M. C. Simon. 1998. Normal myeloid development requires both the Downloaded from transcription factors. This leads to fine-tuning of the gene expres- glutamine-rich transactivation domain and the PEST region of transcription factor sion that is regulated in a spatial, temporal, and restricted manner. PU.1 but not the potent acidic transactivation domain. Mol. Cell. Biol. 18:4347. It is intriguing to speculate that myeloid cell genes that are spe- 23. Chen, C., and H. Okayama. 1987. High efficiency transformation of mammalian ␥ cells by plasmid DNA. Mol. Cell. Biol. 7:2745. cifically activated by IFN- or inflammation are characterized by 24. Sharf, R., D. Meraro, A. Azriel, A. M. Thornton, K. Ozato, E. F. Petricoin, enhanceosomes with similar characteristics. In this context, the A. C. Larner, F. Schaper, H. Hauser, and B. Z. Levi. 1997. Phosphorylation events PEST domains of PU.1, IRF-1, and IRF-2 and the IADs of IRF- modulate the ability of interferon consensus sequence binding protein to interact with interferon regulatory factors and to bind DNA. J. Biol. Chem. 272:9785. 8ICSBP or IRF-4 may be used to bridge all these elements together 25. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983. Accurate transcription http://www.jimmunol.org/ to the proximity of the RNA polymerase II holoenzyme complex. initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11:1475. 26. Schaper, F., S. Kirchhoff, G. Posern, M. Koster, A. Oumard, R. Sharf, B. Z. Levi, Acknowledgments and H. Hauser. 1998. Functional domains of interferon regulatory factor I (IRF- We are grateful to Aviva Azriel and Eyal Scheinman for technical support. 1). Biochem. J. 335:147. We thank Suzie Rosenthal for scientific editing of the manuscript. 27. Weisz, A., S. Kirchhoff, and B. Z. Levi. 1994. IFN consensus sequence binding protein (ICSBP) is a conditional repressor of IFN inducible promoters. Int. Im- munol. 6:1125. References 28. Rosenbauer, F., J. F. Waring, J. Foerster, M. Wietstruk, D. Philipp, and I. Horak. 1. Kalvakolanu, D. V., and E. C. Borden. 1996. An overview of the interferon 1999. Interferon consensus sequence binding protein and interferon regulatory factor-4/Pip form a complex that represses the expression of the interferon-stim-

system: signal transduction and mechanisms of action. Cancer Invest. 14:25. by guest on September 27, 2021 2. Mamane, Y., C. Heylbroeck, P. Genin, M. Algarte, M. J. Servant, C. LePage, ulated gene-15 in macrophages. Blood 94:4274. C. DeLuca, H. Kwon, R. Lin, and J. Hiscott. 1999. Interferon regulatory factors: 29. Kodandapani, R., F. Pio, C. Z. Ni, G. Piccialli, M. Klemsz, S. McKercher, the next generation. Gene 237:1. R. A. Maki, and K. R. Ely. 1996. A new pattern for helix-turn-helix recognition 3. Bovolenta, C., P. H. Driggers, M. S. Marks, J. A. Medin, A. D. Politis, revealed by the PU.1 ETS-domain-DNA complex. Nature 380:456. S. N. Vogel, D. E. Levy, K. Sakaguchi, E. Appella, J. E. Coligan, et al. 1994. 30. Friedman, R. L., and G. R. Stark. 1985. ␣-Interferon-induced transcription of Molecular interactions between interferon consensus sequence binding protein HLA and metallothionein genes containing homologous upstream sequences. Na- and members of the interferon regulatory factor family. Proc. Natl. Acad. Sci. ture 314:637. USA 91:5046. 31. Levy, D. E., D. S. Kessler, R. Pine, N. Reich, and J. E. Darnell. 1988. Interferon- 4. Sharf, R., A. Azriel, F. Lejbkowicz, S. S. Winograd, R. Ehrlich, and B.-Z. Levi. induced nuclear factor that bind a shared promoter element correlate with positive 1995. Functional domain analysis of interferon consensus sequence binding pro- and negative transcriptional control. Genes Dev. 2:383. tein (ICSBP) and its association with interferon regulatory factors. J. Biol. Chem. 32. Nakaya, T., M. Sato, N. Hata, M. Asagiri, H. Suemori, S. Noguchi, N. Tanaka, 270:13063. and T. Taniguchi. 2001. Gene induction pathways mediated by distinct IRFs 5. Eisenbeis, C. F., H. Singh, and U. Storb. 1995. PIP, a novel IRF family member, is during viral infection. Biochem. Biophys. Res. Commun. 283:1150. a lymphoid-specific, PU.1-dependent transcriptional activator. Genes Dev. 9:1377. 33. Au, W. C., P. A. Moore, W. Lowther, Y. T. Juang, and P. M. Pitha. 1995. 6. Nagulapalli, S., and M. L. Atchison. 1998. Transcription factor Pip can enhance Identification of a member of the interferon regulatory factor family that binds to DNA binding by E47, leading to transcriptional synergy involving multiple pro- the interferon-stimulated response element and activates expression of interferon- tein domains. Mol. Cell. Biol. 18:4639. induced genes. Proc. Natl. Acad. Sci. USA 92:11657. 7. Meraro, D., S. Hashmueli, B. Koren, A. Azriel, A. Oumard, S. Kirchhoff, 34. Eklund, E. A., and R. Kakar. 1999. Recruitment of CREB-binding protein by PU.1, H. Hauser, S. Nagulapalli, M. L. Atchison, and B. Z. Levi. 1999. Protein-protein IFN-regulatory factor-1, and the IFN consensus sequence-binding protein is neces- and DNA-protein interactions affect the activity of lymphoid-specific IFN regu- sary for IFN-␥-induced p67phox and gp91phox expression. J. Immunol. 163:6095. latory factors. J. Immunol. 163:6468. 35. Williams, B. R. 1997. Role of the double-stranded RNA-activated protein kinase 8. Perkel, J. M., and M. L. Atchison. 1998. A two-step mechanism for recruitment (PKR) in cell regulation. Biochem. Soc. Trans. 25:509. of Pip by PU.1. J. Immunol. 160:241. 36. Kadereit, S., H. Xu, T. M. Engeman, Y. L. Yang, R. L. Fairchild, and 9. Yamagata, T., J. Nishida, T. Tanaka, R. Sakai, K. Mitani, M. Yoshida, ϩ T. Taniguchi, Y. Yazaki, and H. Hirai. 1996. A novel interferon regulatory factor B. R. Williams. 2000. Negative regulation of CD8 T cell function by the IFN- family transcription factor, ICSAT/Pip/LSIRF, that negatively regulates the ac- induced and double-stranded RNA-activated kinase PKR. J. Immunol. 165:6896. tivity of interferon-regulated genes. Mol. Biol. Cell 16:1283. 37. Widney, D. P., Y. R. Xia, A. J. Lusis, and J. B. Smith. 2000. The murine che- ␣ ␥ 10. Driggers, P. H., D. L. Ennist, S. L. Gleason, W. H. Mak, M. S. Marks, B. Z. Levi, mokine CXCL11 (IFN-inducible T cell chemoattractant) is an IFN- - and J. R. Flanagan, E. Appella, and K. Ozato. 1990. An interferon ␥-regulated protein lipopolysaccharide-inducible glucocorticoid-attenuated response gene expressed that binds the interferon-inducible enhancer element of major histocompatibility in lung and other tissues during endotoxemia. J. Immunol. 164:6322. complex class I genes. Proc. Natl. Acad. Sci. USA 87:3743. 38. Levy, S., S. C. Todd, and H. T. Maecker. 1998. CD81 (TAPA-1): a molecule 11. Weisz, A., P. Marx, R. Sharf, E. Appella, P. H. Driggers, K. Ozato, and B.-Z. involved in signal transduction and cell adhesion in the immune system. Annu. Levi. 1992. The human interferon consensus sequence binding protein (H- Rev. Immunol. 16:89. ICSBP) is a negative regulator of enhancer elements common to interferon in- 39. Kautz, B., R. Kakar, E. David, and E. A. Eklund. 2001. SHP1 protein-tyrosine ducible genes. J. Biol. Chem. 267:25589. phosphatase inhibits gp91phox and p67phox expression by inhibiting interaction of 12. Mittrucker, H. W., T. Matsuyama, A. Grossman, T. M. Kundig, J. Potter, PU.1, IRF1, interferon consensus sequence-binding protein, and CREB-binding A. Shahinian, A. Wakeham, B. Patterson, P. S. Ohashi, and T. W. Mak. 1997. protein with homologous cis elements in the CYBB and NCF2 genes. J. Biol. Requirement for the transcription factor LSIRF/IRF4 for mature B and T lym- Chem. 276:37868. phocyte function. Science 275:540. 40. Korzus, E., J. Torchia, D. W. Rose, L. Xu, R. Kurokawa, E. M. McInerney, 13. Tamura, T., T. Nagamura-Inoue, Z. Shmeltzer, T. Kuwata, and K. Ozato. 2000. T. M. Mullen, C. K. Glass, and M. G. Rosenfeld. 1998. Transcription factor- ICSBP directs bipotential myeloid progenitor cells to differentiate into mature specific requirements for coactivators and their acetyltransferase functions. Sci- macrophages. Immunity 13:155. ence 279:703.