Stat3 and Recruiting Both IFN Regulatory Factor 1 Induces the Human IL-10 Gene by Α IFN

IFN-α Induces the Human IL-10 Gene by Recruiting Both IFN Regulatory Factor 1 and Stat3

This information is current as Löms Ziegler-Heitbrock, Mark Lötzerich, Annette Schaefer, of September 28, 2021. Thomas Werner, Marion Frankenberger and Elke Benkhart J Immunol 2003; 171:285-290; ; doi: 10.4049/jimmunol.171.1.285 http://www.jimmunol.org/content/171/1/285 Downloaded from

References This article cites 30 articles, 16 of which you can access for free at: http://www.jimmunol.org/content/171/1/285.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 28, 2021 *average

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

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

IFN-␣ Induces the Human IL-10 Gene by Recruiting Both IFN Regulatory Factor 1 and Stat31

Lo¨ms Ziegler-Heitbrock,2*†§ Mark Lo¨tzerich,* Annette Schaefer,† Thomas Werner,‡ Marion Frankenberger,§ and Elke Benkhart*

The anti-inflammatory cytokine IL-10 can be induced by type I IFNs, but the molecular mechanisms involved have remained elusive. With in silico analysis of the human IL-10 promoter we identified a module consisting of an IFN regulatory factor 1 (IRF-1) site and a Stat3 site. We demonstrate that IFN-␣ will induce the binding of IRF-1 and Stat3 to the respective motifs. Mutational analysis revealed that inactivation of the IRF-1 motif substantially reduces trans-activation from 5- to 2-fold and that inactivation of the Stat3 motif completely ablates trans-activation by IFN-␣. The dominant role of Stat3 in this module was confirmed with the blockade of trans-activation by a dominant negative Stat3. By contrast, Stat1 contributes a minor proportion to the DNA binding to the Stat site, and overexpression will counteract Stat3-mediated trans-activation. The data show that IFN-␣ Downloaded from induces the IL-10 gene via a module consisting of interdependent IRF-1 and Stat3 motifs. Of note, LPS-induced trans-activation does not target this module, since it is independent of the IRF-1 motif but completely depends on Stat3. The Journal of Immu- nology, 2003, 171: 285–290.

nterleukin-10 is a cytokine that down-regulates the immune start. Based on the conservation of the IRF and Stat motifs in the response and inflammation by suppressing the expression of mouse at a similar distance, the in silico analysis predicted these two I cytokines and by down-regulating important cell surface sites to form a module of interdependent elements (15). We show http://www.jimmunol.org/ molecules such as MHC class II molecules (1). IL-10 is expressed herein that IFN-␣ does, in fact, use the IRF-1 site within the human by monocytes, T cells, and B cells and also by some tumor cells IL-10 promoter, but on its own the IFN-␣-induced IRF-1 is unable to (2Ð4). The expression of IL-10 directed by the murine promoter trans-activate the IL-10 gene. Cooperation with the concomitantly was shown to depend on specific protein-1 (Sp1)3 (5, 6). For the induced Stat3 is absolutely required to induce the expression of IL-10. human IL-10 promoter, Sp1 was shown to be essential for expres- Thus, we experimentally confirm the in silico prediction of an IRF- sion in a monoblastic cell line (7), and a contribution of cAMP- 1/Stat3 module that is used by IFN-␣ (but not by LPS) in the human responsive elements to catecholamine-driven trans-activation has IL-10 gene.

been demonstrated (8). More recently, it was shown that catechol- by guest on September 28, 2021 amine action depends to a large extent on c/EBP-␣ (9). Materials and Methods In a previous study we have shown that expression of the human Cell culture IL-10 gene after LPS stimulation of a B cell line is controlled by the transcription factor Stat3, which binds to a single motif at The human RPMI 8226.1 B cell line (16) was grown in RPMI 1640 culture Ϫ medium supplemented with 2 mM L-glutamine (25030-024; Life Technol- 120 bp (10). ogies, Gaithersburg, MD), 200 U/ml penicillin, 200 ␮g/ml streptomycin IL-10 is also regulated by IFNs. Specifically, IFN-␥ will down- (15140-114; Life Technologies), 1ϫ nonessential amino acids (11140-035; regulate IL-10 (11), while the type I IFN, IFN-␣, will up-regulate Life Technologies), and 1% (v/v) OPI supplement (O-5003; Sigma- IL-10 expression (12). IFN-␣ exerts most of its activities via the Aldrich, Munich, Germany). This medium was passed through a Gambro ϩ ϩ U-2000 ultrafiltration column (Gambro Medizintechnik, Planegg-Martin- transcription factor complex ISGF3 (Stat1 Stat2 p48) and via sried, Germany) to deplete contaminating LPS, and this was followed by IFN regulatory factor-1 (IRF-1) (13). the addition of 10% v/v low LPS FCS (477U; Seromed, Berlin, Germany). When analyzing the human IL-10 promoter by MatInspector (14), RPMI 8226.1 cells were passaged in 75-cm2 tissue culture flasks (Costar, we noted an IRF-1 motif at Ϫ180 bp upstream of the transcription Bodenheim, Germany). HEK 293 cells (17) were grown in culture medium in six-well plates (no. 3506; Costar, Cambridge, MA). LPS from Salmonella minnesota was obtained from Sigma-Aldrich (L6261). IFN-␣2b (intron A) was provided *Institute for Immunology, University of Munich, Munich, Germany; †Division of by Essex Pharma (Munich, Germany). Immunology, University of Leicester, Leicester, United Kingdom; ‡Institute of Mam- malian Genetics, GSF National Research Center for Environment and Health, Neu- ELISA herberg and Genomatix Software, Munich, Germany; and ¤Clinical Cooperation Group Aerosols in Medicine, GSF National Research Center for Environment and IL-10 protein was determined in a sandwich ELISA using a biotinylated Health, Institute for Inhalation Biology and Asklepios Hospitals, Gauting, Germany second Ab and streptavidin-peroxidase for detection (M1910 Pelikine Received for publication January 13, 2003. Accepted for publication January Compact Plus and M1980 Pelikine Tool Set; CLB, Amsterdam, The 28, 2003. Netherlands). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance Sequence analysis with 18 U.S.C. Section 1734 solely to indicate this fact. Analysis of the sequences for transcription factor binding sites was con- 1 This work was supported by the Deutsche Forschungsgemeinschaft (Grant Zi 288/-1 ducted with the program MatInspector Professional (Genomatix Software, and Project A9 (SFB464)) and by MediSearch (Leicester, U.K.). Munich, Germany) based on the MatInspector program (14) using the se- 2 Address correspondence and reprint requests to Prof. L. Ziegler-Heitbrock, Division lected matrix library (vertebrate section) and optimized thresholds. Mod- of Immunology, University of Leicester, Leicester, U.K. LE1 9HN. E-mail address: ules were detected by comparative analysis of the human and mouse IL-10 [email protected] promoter sequences identifying modular conservation of binding sites in 3 Abbreviations used in this paper: Sp1, specific protein-1; IRF, IFN regulatory factor. analogy to the NF-␬B/IRF module in the HLA class I genes (15).

Gel shift analysis ng/ml resulted in strong expression of IL-10 protein in the super- Ϯ Nuclear extracts were isolated according to the method described by Dig- natant with 85,615 24,504 pg/ml after 18 h compared with nam et al. (18) in the presence of a protease inhibitor mixture (10 ␮g/ml 149 Ϯ 92 pg/ml in unstimulated cells (n ϭ 3). Stimulation of the aprotinin (A6279; Sigma-Aldrich), 1 mM PMSF (P7626; Sigma-Aldrich), cells with IFN-␣ at 100 U/ml also led to the production of IL-10 40 ␮g/ml leupeptin-propionyl (L3402; Sigma-Aldrich), 20 ␮g/ml leupep- protein, such that 1,665 Ϯ 502 pg/ml was detectable. After stim- tin-acetate (L2023; Sigma-Aldrich), 20 ␮g/ml antipain (A6191; Sigma- ␣ ␮ ␮ ulation with both LPS and IFN- , there was a slight increase com- Aldrich), 20 g/ml pepstatin A (P4265; Sigma-Aldrich), 400 M ALLN Ϯ (A6185; Sigma-Aldrich), and 2 mM DTT (11474; Merck, Rahway, NJ)). pared with the LPS alone (116,834 40,095 pg/ml). 32 Five micrograms of nuclear protein was then admixed with P-labeled ␣ double-stranded IRF oligonucleotide (GATGCAAAAATTGAAAACTA IFN- induced DNA binding to the IRF motif AGT) or with the LS4 oligonucleotide (ATCCTGTGCCGGGAAACC) in Previous studies have demonstrated that the transcription factor ␮ the presence of 0.5 g of poly(dI-dC) (US20539; Amersham Pharmacia Ϫ Biotech, Arlington Heights, IL) and 1 mg/ml BSA (A2153; Sigma-Aldrich) Stat3, binding to the TGCCGGGAA motif at position 120 of the per 20 ␮l. After 20 min of incubation at 21¡C, samples were electropho- human IL-10 gene, is responsible for the LPS-stimulated gene ex- resed on nondenaturing polyacrylamide gels in 0.25ϫ TBE buffer (22.5 pression in RPMI 8226 cells. When searching for additional po- mM Tris borate and 0.5 mM EDTA, pH 8.5). For supershift analysis nu- tential transcription factor binding sites by MatInspector analysis, clear extracts were first incubated with a 1/20 dilution of Ab for 30 min, we noted an IRF motif (AATTGAAA) ϳ50 bp upstream of the followed by incubation with the oligonucleotides. The following Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA): IRF-1 sc- Stat site among many other predicted binding sites (see sequence 497, IRF-2 sc-498, Stat1 p84/p91 sc-346, Stat2 sc-839, Stat3 sc-7179, Stat4 in Materials and Methods). This IRF motif was selected for further sc-485, Stat5 sc-1656, and Stat6 sc-621. For competition analysis, nuclear analysis because in silico analysis demonstrated that the IRF and 32 extracts were admixed with P-labeled double-stranded oligonucleotides Stat3 motifs appeared to be the only composite feature in the prox- Downloaded from in the presence of 100, 10, or 1 ng of unlabeled double-stranded oligonucleotide (Stat consensus, TCGTTCGATTCCGGGAATTGA (19); IRF- imal promoter conserved between mouse and man. Moreover, the motif, GATGCAAAAATTGAAAACTAAGT; LS4-Stat3 motif, ATC relative arrangement of the binding sites suggested the possibility CTGTGCCGGGAAACC). of a functional promoter module. Constructs Gel-shift analysis with nuclear extracts from IFN-␣-stimulated cells revealed an inducible band (arrowhead in Fig. 1A). This band The Ϫ195 bp IL-10 promoter fragment analyzed has the following sequence:GCTTACGATGCAAAAATTGAAAACTAAGTTTATTAGAGA was not affected by an admixture of unlabeled Stat motif, but it http://www.jimmunol.org/ GGTTAGAGAAGGAGGAGCTCTAAGGAGAAAAAATCCTGTGCCG was specifically competed by unlabeled IRF-motif (Fig. 1A). The GGAAACCTTGATTGTGGCTTTTTAATGAATGAAGAGGCCTCCCT nonspecific band marked by an asterisk in Fig. 1 was competed by GAGCTTACAATATAAAAGGGGGACAGAGAGGTGAAGGTCTACA both the specific IRF oligonucleotide and the nonspecific Stat oli- CATCAGGGGCTTGCTCTTGCAAAACCAAACCACAAGACAGACTT gonucleotide, indicating that this band cannot be involved in spe- GCAAAAGAAGGC.The IRF motif (AATTGAAA) and the Stat motif (TGCCGGGAA) are shown in bold letters, the TATAA box is underlined, cific binding to the IRF motif. When performing supershift anal- and the transcription start site ϩ1 is in bold and underlined. This sequence ysis, the gels were run for longer periods of time to better separate was generated by oligonucleotide synthesis of three subfragments with a the specific binding complex. Using anti-IRF Abs under these con- BamHI sequence at the very 5Ј end and a XhoI site at the very 3Ј end. The ␤ ditions we could shift most of the binding protein with the anti- three fragments were annealed and ligated into the p TATA.luci reporter by guest on September 28, 2021 plasmid (provided by T. Wirth, Wurzburg, Germany) that had been cut IRF-1 Ab (Fig. 1B, lane 3). A weak band with somewhat slower with BamHI and XhoI, followed by isolation of the vector via gel mobility remained, and this was supershifted with the anti-IRF-2 electrophoresis. Ab (Fig. 1B, lane 4). The combination of both Abs removed both Mutations were obtained and controlled based on MatInspector predic- of the specific IRF bands (Fig. 1B, last lane). These data indicate tions. For the IRF site the sequence AATTGAAA was exchanged by GT that the specific complex consists of a minor IRF-2 band with TAGTAC, while for mutation of the Stat site the sequence TGCCGGGAA was replaced by AGACTTGAA. The mutant promoter constructs were lower mobility and a major IRF-1 band with higher mobility. again generated by having oligonucleotide subfragments synthesized, an- We next studied the time and dose dependence of IRF induction nealed, ligated, and cloned into p␤TATA.luci. by IFN-␣. As shown in Fig. 2A, a strong band was induced as early Expression plasmids for wild-type Stat3 (pCAGGS Neo HA STAT3) as 1 h after IFN-␣ stimulation, and this band gradually declined, dominant negative Stat3 (pCAGGS Neo HA STAT3F), wild-type Stat1 (pCAGGS Neo HA STAT1), and the control plasmid (pCAGGS) (20) were but was still detectable at 6 h. In a dose-response analysis 10 U of provided by Dr. T. Hirano (Osaka, Japan) via Drs. P. C. Heinrich and I. IFN-␣/ml led to some IRF binding, and a strong specific band (Fig. Behrmann (Aachen, Germany). 2B, black arrowhead) was seen with 100 U/ml. There was no fur- Transfection ther substantial increase at 1000 U/ml. Additional bands with higher and lower mobilities also changed with IFN-␣ treatment, RPMI 8226 cells were transfected according to the method described by Shakhov et al. (21) with 5 ␮g of reporter plasmid/107 cells in the presence but these where shown to be nonspecific, as demonstrated in Fig. 1. of DEAE-dextran (125 ␮g/ml; E1210; Promega, Madison, WI) for 90 min, While LPS acts on the IL-10 promoter via Stat3, we asked followed by the addition of 10% (v/v) DMSO (D58791; Sigma-Aldrich) whether this stimulus might also induce DNA binding of IRF. As 6 for 150 s. Cells were then cultured for 3 days in six-well plates (3.3 ϫ 10 shown in Fig. 3A, LPS could, in fact, rapidly activate IRF at 2 h. cells/well) and were stimulated for 6 h with LPS (S. minnesota; L-6261; Again, supershift analysis demonstrated that the LPS-induced Sigma-Aldrich) at 100 ng/ml. Cotransfection studies were performed with an admixture of reporter plasmids (5 ␮g) with empty control plasmids and binding protein consists of a major IRF-1 and a minor IRF-2 pro- expression plasmids for IRF-1, for wild-type Stat3, and dominant negative tein (data not shown). Stat3 (2 ␮g each). Given the ability of LPS to induce IRF, one might argue that the For measurement of luciferase activity cells were harvested and lysed, IFN-␣ effect in this system could be due at least in part to LPS and luciferase activity in cell lysates was determined using a model ␣ LB9501 luminometer (Berthold, Wildbad, Germany) and a luciferase assay contamination of the recombinant protein. Boiling IFN- for 10 kit (ELA100; Berthold). min at 95¡C completely ablated IRF induction, such that only a weak constitutive band remained. This indicates that no heat-re- Results sistant LPS was active in the IFN preparation (Fig. 3B). This was ␣ Induction of IL-10 protein by IFN- confirmed by the addition of polymyxin B for inactivation of LPS, For analysis of IL-10 gene expression we employed the human B which did not reduce the activity of IFN-␣. These data show that cell line RPMI 8226 clone 1. Consistent with earlier findings of the induction of IRF-1 by IFN-␣ is not due to LPS contamination, mRNA analysis (16), stimulation of these cells with LPS at 100 but is a genuine activity of IFN. The Journal of Immunology 287

FIGURE 2. Time and dose dependency of IRF induction by IFN-␣. RPMI 8226 cells were untreated or treated with IFN-␣ at 1000 U/ml for 1Ð6h(A), or they were treated for 2 h with IFN-␣ at 10, 100, or 1000 U/ml Downloaded from (B). Nuclear extracts were admixed with the radiolabeled IRF motif from .Nonspeciﬁc band; arrowhead, IRF ,ء .the human IL-10 promoter

mutant, indicating that for LPS-induced trans-activation the IRF motif is not required. http://www.jimmunol.org/ When we mutated the Stat motif, the LPS-induced trans-activation was markedly reduced, but surprisingly the IFN-␣ activity was completely ablated (Fig. 4, third panel). Mutation of both sites resulted in a loss of trans-activation by any of the stimuli (Fig. 4, fourth panel). The loss of IFN-␣-induced trans-activation after mutation of the Stat motif demonstrates that IRF alone is unable to drive the human IL-10 promoter. Rather, cooperation with the Stat3 site is by guest on September 28, 2021 essential to achieve trans-activation of this promoter. FIGURE 1. Induction of IRF by IFN-␣. Nuclear extracts from RPMI IFN-␣ induced DNA binding to the Stat motif ␣ 8226 cells with and without stimulation by IFN- at 1000 U/ml were To study the contribution of Stat to the IFN-␣ action, we per- admixed with the radiolabeled IRF motif from the human IL-10 promoter. formed gel-shift analysis with the Stat motif using nuclear extracts A, Induction and competition analysis. Nuclear extracts were tested directly ␣ ␣ or after the addition of decreasing amounts (100, 10, and 1 ng/lane) of from IFN- -stimulated cells. IFN- did, in fact, lead to a strong -Non- induction of a Stat binding protein (Fig. 5A, arrowhead). The spec ,ء .either cold IRF motif or cold Stat motif from the IL-10 promoter specific band; arrowhead ϭ specific band. B, Supershift analysis. Nuclear ificity of this band has been demonstrated previously (10). There extracts from IFN-␣-stimulated cells were admixed with Abs against IRF-1 Nonspecific bands; open arrows, IRF proteins; Ϫ,noAb ,ء .or IRF-2 added; c, nonimmune serum; 1, Abs against IRF-1; 2, Abs against IRF-2; 1,2, Abs against both IRF-1 and IRF-2.

Trans-activation of the human IL-10 promoter by IFN-␣ We then analyzed the effect of IFN-␣ on the human IL-10 promoter by studying a luciferase reporter construct with the Ϫ195 bp IL-10 promoter, which contains both the Stat and the IRF motif. IFN-␣ stimulation of RPMI 8226 cells transfected with this construct led to an average 5-fold trans-activation (Fig. 4, left panel, f) com- parable to what was observed with LPS stimulation ( ). The combination of both stimuli did not lead to a further increase (p). To demonstrate the importance of the IRF site for the IFN-␣- induced trans-activation, we mutated the respective motif. This FIGURE 3. Induction of IRF by LPS. A, LPS induces IRF. RPMI 8226 cells were either untreated, treated with LPS at 100 ng/ml for 1Ð6h,or led, in fact, to a clear reduction of IFN-␣-induced trans-activation treated with IFN-␣ at 100 U/ml for 1 h. All samples were run on the same (Fig. 4, second panel), but some activity (2-fold trans-activation) gel, but samples to the left and right of the black line were from separate ␣ remained. This indicates that IFN- can induce the human IL-10 sets of nuclear extracts. B, Induction of IRF by IFN-␣ is not due to LPS promoter most likely via another motif, in addition to the IRF contamination. Cells were stimulated with IFN-␣ (100 U/ml), with IFN-␣ motif that was mutated in this construct. By contrast, trans-acti- that was denatured by heating for 10 min at 95¡C, or with IFN-␣ admixed .Nonspeciﬁc bands; solid arrowhead, IRF ,ء .(vation induced by LPS was similar for the wild type and the IRF with polymyxin B (10 ␮g/ml 288 IRF/STAT MODULE IN IL-10 PROMOTER

to the Stat3 motif after IFN-␣ stimulation by supershift analysis, the gels were again run for a longer period of time for better res- olution of the bands (Fig. 5B). Here, anti-Stat1 Ab removed a minor high mobility band, while anti-Stat3 ablated the major low mobility band (lanes 3 and 5, respectively, in Fig. 5B). The combination of the Abs against Stat1 and Stat3 removed the entire speciﬁc complex (Fig. 5B, lane 6). The anti-Stat3 Ab induced the appearance of two supershifted bands (Fig. 5B, open arrows, lanes 5 and 6), while the anti-Stat1 Ab did not result in a visible band in the upper part of the gel. All other anti-Stat-Abs had no effect. These data suggest that the IFN-␣-induced complex, which binds to the Stat3 motif, consists mainly of Stat3 homodimers (strong low mobility band) plus a small amount of Stat1 homodimers (weak higher mobility band). FIGURE 4. Trans-activation of the human IL-10 promoter by IFN-␣. Again, heating of the IFN-␣ preparation abolished this effect, and RPMI 8226 cells were transfected with either the Ϫ195 IL-10 promoter polymyxin B had no effect (Fig. 5C), indicating that the Stat induction luciferase reporter constructs or constructs carrying mutations at the IRF is not due to LPS contamination but is a genuine effect of IFN-␣. motif, the Stat3 motif, or both, as indicated at the top of the ﬁgure. Cells

were then stimulated with either LPS at 100 ng/ml ( ), IFN-␣ at 100 U/ml Downloaded from (f), or both LPS and IFN-␣ (p), and luciferase activity was measured. The Involvement of Stat3 in trans-activation of the human IL-10 average Ϯ SD of three experiments are shown. promoter by IFN-␣ To conﬁrm that Stat3 is involved in the IFN-␣ action on the human IL-10 promoter, we overexpressed wild-type and dominant nega- was a clear signal after1hofstimulation, sometimes with a de- tive Stat3 together with the IL-10 promoter reporter construct in crease in binding activity at 2 h, as in this example. Later, the ␣ RPMI 8226 cells, followed by stimulation with IFN- at 100 U/ml. http://www.jimmunol.org/ signal was strong again (Fig. 5A). To identify the proteins bound In these experiments the dominant negative constructs strongly reduced constitutive and IFN-␣-induced trans-activation (Fig. 6). Fur- thermore, the wild-type Stat3 strongly increased trans-activation by IFN-␣ from 2-fold (pCAGGS) to 7-fold (Stat3) while leaving the constitutive promoter activity unaffected. These data indicate that IFN-␣ employs Stat3 to trans-activate the human IL-10 promoter. by guest on September 28, 2021

FIGURE 5. Induction of Stat3 by IFN-␣. A, IFN-␣ induces binding to the Stat3 motif. RPMI 8226 cells were stimulated with IFN-␣ at 1000 U/ml for 1Ð6 h or with LPS at 1000 ng/ml for 4 h. Nuclear extracts were then admixed Nonspecific ,ء .with the radiolabeled Stat3 motif of the human IL-10 promoter ␣ bands; arrowhead, specific Stat band. B, Supershift analysis of IFN-␣-induced FIGURE 6. Effect of transfected Stat3 on IFN- -induced trans-activation Stat-binding proteins. Nuclear extracts were first incubated for 30 min with of the human IL-10 promoter. Stat3 wild-type and dominant negative expres- 1/20 dilutions of different anti-Stat Abs as indicated at the top of the middle sion plasmids were cotransfected with the human Ϫ195 IL-10 promoter lu- panel, followed by incubation with the radiolabeled Stat oligonucleotide. ciferase reporter plasmid into RPMI 8226 cells, and after 3 days luciferase Nonspecific band; solid arrow, specific bands; open arrows, supershifted activity cells were stimulated with IFN-␣ at 1000 U/ml for 6 h, followed by ,ء bands. C, Induction of Stat binding by IFN-␣ is not due to LPS contamination. determination of luciferase activity. Shown are the average Ϯ SD of three Cells were stimulated with IFN-␣ (100 U/ml), with IFN-␣ that was denatured experiments. f, Untreated; Ⅺ, IFN-␣ treated; PCAGGS, empty vector; Stat3F, by heating for 10 min at 95¡C, or with IFN-␣ admixed with polymyxin B (10 expression plasmid containing Stat3 with tyrosine to phenylalanine mutation at .nonspecific bands; solid arrowhead, specific Stat band. position 705; Stat3, expression plasmid containing wild-type Stat3 ,ء .(␮g/ml The Journal of Immunology 289

promoter (6) at Ϫ145 bp, i.e., it is located between the Stat3 and IRF-1 motifs studied in the present report. In our previous study of LPS induction, a linker-scanning series covering this area did not reveal any impact of this motif on induction in human B cells (10). However, we cannot exclude that this motif is involved in IFN- induced gene expression. When analyzing the human promoter after 24 h of LPS stimulation of the promonocytic cell line THP-1, Ma et al. (7) noted an Sp1 site at Ϫ636 bp, which appeared to be responsible for all inducible activity in this system. In our earlier report, in which we used an IL-10 promoter deletion series in a human B cell system, no such drop in LPS-inducible promoter activity was noted (10), suggesting that in B cells regulation of IL-10 may be different. Thus, it appears that Sp1 may be involved in regulation of the IL-10 gene dependent on the species and tissue investigated. For the human gene another study reported on the contribution of cAMP-responsive elements to catecholamine-mediated trans-activation, in that mutation of the three main cAMP- FIGURE 7. Stat1 suppresses Stat3-mediated trans-activation of the hu- responsive element sites reduced promoter activity by 50% (8). In man IL-10 promoter. Stat1 and Stat3 were cotransfected with the human addition, trans-activation by catecholamines appears to require the Downloaded from Ϫ195 IL-10 promoter luciferase reporter plasmid into RPMI 8226 cells, action of c/EBP (9). Furthermore, there is evidence for a role for and luciferase activity was determined after stimulation with IFN-␣ (100 c-Maf in IL-10 gene expression, but c-Maf may act indirectly by U/ml for 6 h). Cells received an identical total amount of expression plas- activating other transcription factors (25). mid (1 ␮g). In the third column pair (Stat1 alone), Stat1 expression plasmid We have demonstrated by systematic linker scanning analysis was at 0.3 ␮g. Stat3 was always at 0.3 ␮g, and in the combinations Stat1 that a site at Ϫ120 bp is crucial for LPS-stimulated IL-10 promoter was used at 0.6, 0.3, or 0.15 ␮g, with the remainder provided by pCAGGS. activity. This site was shown to bind LPS-induced Stat3 and, in fact, http://www.jimmunol.org/ ϭ The data given are the averages of five experiments, n 4 for the 1:1 and dominant negative Stat3 could strongly suppress trans-activation (10). 0.5:1 cotransfection. We now have analyzed the molecular mechanism that controls IFN- ␣-induced trans-activation of the human IL-10 promoter. Type I IFNs, including IFN-␣, act mainly via ISGF3 and IRF Stat1 counteracts IFN-␣/Stat3-mediated trans-activation of the (12), but IFN-␣ also has been shown to recruit Stat3 to IFN-␣ human IL-10 promoter receptor I, followed by phosphorylation and translocation into the In gel-shift analysis using the Stat motif we observed induction of nucleus (26, 27). ␣ a minor band that was identified as Stat1 (see Fig. 5). We therefore We show herein that IFN- recruits both IRF and Stat transcrip- by guest on September 28, 2021 asked what role Stat1 might play in trans-activation of the gene. tion factors to the human IL-10 promoter. The mobilization of ␣ For these studies RPMI 8226 cells were transfected with the Ϫ195 these factors is a genuine effect of IFN- and is not due to LPS fragment of the IL-10 promoter plus a total of 1 ␮g of expression contamination of the recombinant protein. This conclusion is based plasmid. The experiments in Fig. 7 confirmed induction by Stat3, but on the finding that transcription factor mobilization was ablated by ␣ Stat1 had no activity when added alone. When Stat1 was admixed heating the IFN- for 10 min to 95¡C, a procedure that denatures with Stat3 expression plasmid, Stat1 was able to counteract the Stat3- the protein, but leaves LPS unaffected. Also, neutralization of LPS ␣ mediated trans-activation at equal and 2/1 Stat1/Stat3 ratios (Fig. 7, by polymyxin B did not reduce the IFN- induction of DNA bind- ␣ right columns). These data demonstrate that Stat1 is able to counteract ing proteins. The same pattern of results was seen in IFN- -driven Stat3-mediated trans-activation of the human IL-10 promoter. reporter gene analyses (data not shown). When studying the IFN-␣-induced DNA-binding proteins that Discussion bind to the IRF motif, we noted the induction of a major IRF-1 While IFN-␥ can effectively down-regulate IL-10, the type I IFNs band, while the constitutive weak IRF-2 band was essentially un- have been shown to up-regulate IL-10 expression (12). This in- altered. Mutation of the IRF motif in the context of the IRF-Stat duction of the anti-inflammatory IL-10 may contribute to the ben- promoter module reduced the trans-activation induced by IFN-␣, eficial effects of IFN-␤ therapy in multiple sclerosis (22). We can but did not ablate it. This indicates that IFN-␣ invokes an addi- confirm this activity of IFN-␣ by the finding of a pronounced ex- tional independent DNA motif for trans-activation of the human pression of IL-10 protein in the supernatant in IFN-␣-treated IL-10 gene. This might well be the Stat3 site of the module. This RPMI 8226 cells. We used this model system to study the molec- could, in fact, be demonstrated, since trans-activation by IFN-␣ ular mechanisms and the transcription factors involved in IL-10 was reduced by mutation of the Stat3 motif. This implicates both up-regulation by type I IFNs. IRF and Stat in regulation of the IL-10 gene. It was surprising, A role of NF-␬B in IL-10 gene expression had been suggested however, that mutation of the Stat3 motif completely ablated IFN- in early studies (23), but our own in silico analysis of the impli- ␣-induced trans-activation. This indicates that IRF-1 on its own cated motifs (unpublished) and the finding that blockade of the cannot act on the IL-10 promoter. Rather, there is a complete de- NF-␬B pathway does not affect IL-10 gene expression (24) clearly pendence of the IRF action on Stat3. Hence, these functional data dismiss a role for this transcription factor. confirm the module nature of the two transcription factor binding For the murine IL-10 gene an important role of Sp1 in regulating sites that had been predicted based on FastM analysis (15) of the LPS-stimulated promoter activity was demonstrated (5). The re- human IL-10 promoter. This module is also conserved in the mu- spective motif, GAGGAGGAGC, was found within the first 100 rine promoter, which exhibits a 64-bp intervening sequence. A bp of the promoter, and its mutation reduced both the inducible and similar module was recently predicted and experimentally con- the constitutive promoter activity in the murine RAW 264 macro- firmed for the R(A) and R(B) elements within the human RANTES phage cell line (5, 6). A similar motif can be found in the human promoter, which bind Sp1 and NK-␬B (28). 290 IRF/STAT MODULE IN IL-10 PROMOTER

In addition to the induction of Stat3, we noted that IFN-␣ also 5. Brightbill, H. D., S. E. Plevy, R. L. Modlin, and S. T. Smale. 2000. A prominent induced Stat1, which contributed a minor portion to the total DNA role for Sp1 during lipopolysaccharide-mediated induction of the IL-10 promoter in macrophages. J. Immunol. 164:1940. binding activity. With overexpression of Stat1 we found a suppression 6. Tone, M., M. J. Powell, Y. Tone, S. A. Thompson, and H. Waldmann. 2000. of IFN-␣/Stat3-mediated trans-activation. Hence, it appears that Stat1 IL-10 gene expression is controlled by the transcription factors Sp1 and Sp3. can have a negative effect on IL-10 gene expression. It is unclear why J. Immunol. 165:286. 7. Ma, W., W. Lim, K. Gee, S. Aucoin, D. Nandan, M. Kozlowski, F. Diaz-Mitoma, transfection of Stat1 on its own did not change IFN-␣-induced trans- and A. Kumar. 2001. The p38 mitogen-activated kinase pathway regulates the activation (Fig. 7). A possible explanation is that Stat1 is a weak human interleukin-10 promoter via the activation of Sp1 transcription factor in lipopolysaccharide-stimulated human macrophages. J. Biol. Chem. 276:13664. agonist in this system and will only blunt the activity of a strong 8. Platzer, C., E. Fritsch, T. Elsner, M. H. Lehmann, H.-D. Volk, and S. Prosch. agonist such as Stat3. Analysis of the molecular interaction of these 1999. Cyclic adenosine monophosphate-responsive elements are involved in the Stat proteins and their associated costimulators and corepressors may transcriptional activation of the human IL-10 gene in monocytic cells. Eur. J. Im- munol. 29:3098. be able to resolve this question. 9. Brenner, S., S. Prosch, K. Schenke-Layland, U. Riese, U. Gausmann, and Of note, LPS-induced trans-activation does not appear to de- C. Platzer. 2003. cAMP-induced interleukin-10 promoter activation depends on pend on the IRF-1/Stat3 module. LPS induces IRF binding to the CCAAT/enhancer-binding protein expression and monocytic differentiation. J. Biol. Chem. 278:5597. respective motif, albeit with a lower intensity (Fig. 3A). The mu- 10. Benkhart, E. M., M. Siedlar, A. Wedel, T. Werner, and H. W. Ziegler-Heitbrock. tation analysis does, however, indicate that this transcription factor 2000. Role of Stat3 in lipopolysaccharide-induced IL-10 gene expression. J. Im- is not required for LPS-induced trans-activation, since a mutation munol. 165:1612. 11. Chomarat, P., M. C. Rissoan, J. Banchereau, and P. Miossec. 1993. Interferon ␥ of the IRF site does not affect luciferase activity induced by LPS inhibits interleukin 10 production by monocytes. J. Exp. Med. 177:523. (Fig. 4). It is unclear why there is LPS-induced IRF binding to this 12. Aman, M. J., T. Tretter, I. Eisenbeis, G. Bug, T. Decker, W. E. Aulitzky, H. Tilg, ␣ site, but this does not impact on IL-10 promoter activity. A pos- C. Huber, and C. Peschel. 1996. Interferon- stimulates production of interleu- Downloaded from kin-10 in activated CD4ϩ T cells and monocytes. Blood 87:4731. sible explanation is that additional factors involved in cooperativ- 13. Darnell, J. E., Jr., I. M. Kerr, and G. R. Stark. 1994. Jak-STAT pathways and ity of the IRF and Stat binding proteins are not invoked in LPS- transcriptional activation in response to IFNs and other extracellular signaling stimulated cells. Taken together, LPS acts on the IL-10 gene via proteins. Science 264:1415. ␣ 14. Quandt, K., K. Frech, H. Karas, E. Wingender, and T. Werner. 1995. MatInd and Stat3 only while IFN- trans-activation requires the IRF-1/Stat3 MatInspector: new fast and versatile tools for detection of consensus matches in module. This indicates that the effects of IFN-␣ are more tightly nucleotide sequence data. Nucleic Acids Res. 23:4878. 15. Klingenhoff, A., K. Frech, K. Quandt, and T. Werner. 1999. Functional promoter controlled, requiring more complex interactions for the induction http://www.jimmunol.org/ modules can be detected by formal models independent of overall nucleotide of IL-10 expression. Here negative control mechanisms at two sequence similarity. Bioinformatics 15:180. sites can prevent the gene induction. For LPS the sole dependence 16. Ziegler-Heitbrock, H. W., H. Pechumer, I. Petersmann, J. J. Durieux, N. Vita, on Stat3 allows for a brisk response, which may be essential to M. O. Labeta, and M. Strobel. 1994. CD14 is expressed and functional in human B cells. Eur. J. Immunol. 24:1937. counteract the otherwise damaging effects of the proinﬂammatory 17. Graham, F. L., J. Smiley, W. C. Russell, and R. Nairn. 1977. Characteristics of cytokines that this bacterial product induces concomitantly. When a human cell line transformed by DNA from human adenovirus type 5. J. Gen. comparing promoter activity and IL-10 protein production, it is evi- Virol. 36:59. 18. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983. Accurate transcription dent that LPS and IFN-␣ show similar induction of the luciferase initiation by RNA polymerase II in a soluble extract from isolated mammalian reporter gene, but LPS induces much higher levels of IL-10 protein. nuclei. Nucleic Acids Res. 11:1475.

19. Horvath, C. M., Z. Wen, and J. E. Darnell, Jr. 1995. A STAT protein domain that by guest on September 28, 2021 This indicates that LPS uses additional transcriptional and post-tran- determines DNA sequence recognition suggests a novel DNA-binding domain. scriptional elements to induce high levels of IL-10 protein. Genes Dev. 9:984. The mode of interaction of the two transcription factors after 20. Nakajima, K., Y. Yamanaka, K. Nakae, H. Kojima, M. Ichiba, N. Kiuchi, ␣ T. Kitaoka, T. Fukada, M. Hibi, and T. Hirano. 1996. A central role for Stat3 in IFN- stimulation may be dependent on their cognate binding to IL-6-induced regulation of growth and differentiation in M1 leukemia cells. DNA, as has been shown for the NF-AT interaction with AP-1 EMBO J. 15:3651. (29), or it may be a direct protein-protein interaction, as has been 21. Shakhov, A. N., M. A. Collart, P. Vassalli, S. A. Nedospasov, and C. V. Jongeneel. 1990. ␬B-type enhancers are involved in lipopolysaccharide-mediated transcriptional shown for Stat3 and c-Jun (30). These authors have defined two activation of the tumor necrosis factor ␣ gene in primary macrophages. J. Exp. Med. Stat3 domains that are involved in the interaction with c-Jun, and 171:35. these could also be involved in the Stat3 interaction with IRF-1. 22. Ozenci, V., M. Kouwenhoven, Y. M. Huang, P. Kivisakk, and H. Link. 2000. Multiple sclerosis is associated with an imbalance between tumour necrosis fac- Alternatively, additional adaptor molecules could come into play tor-␣ (TNF-␣)- and IL-10-secreting blood cells that is corrected by interferon-␤ to form an enhanceosome complex that ultimately controls the (IFN-␤) treatment. Clin. Exp. Immunol. 120:147. expression of the IL-10 gene. 23. Mori, N., and D. Prager. 1997. Activation of the interleukin-10 gene in the human T lymphoma line HuT 78: identification and characterization of NF-␬B binding sites in Taken together the unique finding in our study is the discovery the regulatory region of the interleukin-10 gene. Eur. J. Haematol. 59:162. in the human IL-10 promoter of an IRF-1/Stat3 module that is 24. Bondeson, J., K. A. Browne, F. M. Brennan, B. M. Foxwell, and M. Feldmann. stringently dependent on Stat3 for trans-activation by IFN-␣. 1999. Selective regulation of cytokine induction by adenoviral gene transfer of I␬B␣ into human macrophages: lipopolysaccharide-induced, but not zymosan- induced, proinflammatory cytokines are inhibited, but IL-10 is nuclear factor-␬B Acknowledgments independent. J. Immunol. 162:2939. We gratefully acknowledge helpful discussions with E. Weiss (Munich, 25. Cao, S., J. Liu, M. Chesi, P. L. Bergsagel, I. C. Ho, R. P. Donnelly, and X. Ma. Germany); the provision of Stat3 expression plasmids by T. Hirano (Osaka, 2002. Differential regulation of IL-12 and IL-10 gene expression in macrophages by the basic leucine zipper transcription factor c-Maf fibrosarcoma. J. Immunol. Japan) that were kindly forwarded by I. Behrmann and P. C. Heinrich 169:5715. (Aachen, Germany), and B. W. Kiernan for efficient data management. 26. Yang, C. H., W. Shi, L. Basu, A. Murti, S. N. Constantinescu, L. Blatt, E. Croze, J. E. Mullersman, and L. M. Pfeffer. 1996. Direct association of STAT3 with the References IFNAR-1 chain of the human type I interferon receptor. J. Biol. Chem. 271:8057. 1. de Waal Malefyt, R., J. Abrams, B. Bennett, C. G. Figdor, and J. E. de Vries. 27. Chung, C. D., J. Liao, B. Liu, X. Rao, P. Jay, P. Berta, and K. Shuai. 1997. 1991. Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: an Specific inhibition of Stat3 signal transduction by PIAS3. Science 278:1803. autoregulatory role of IL-10 produced by monocytes. J. Exp. Med. 174:1209. 28. Fessele, S., S. Boehlk, A. Mojaat, N. G. Miyamoto, T. Werner, E. L. Nelson, 2. de Waal Malefyt, R., H. Yssel, M. G. Roncarolo, H. Spits, and J. E. de Vries. D. Schlondorff, and P. J. Nelson. 2001. Molecular and in silico characterization 1992. Interleukin-10. Curr. Opin. Immunol. 4:314. of a promoter module and C/EBP element that mediate LPS-induced RANTES/ 3. Gastl, G. A., J. S. Abrams, D. M. Nanus, R. Oosterkamp, J. Silver, F. Liu, CCL5 expression in monocytic cells. FASEB J. 15:577. M. Chen, A. P. Albino, and N. H. Bander. 1993. Interleukin-10 production by 29. Chen, L., J. N. Glover, P. G. Hogan, A. Rao, and S. C. Harrison. 1998. Structure human carcinoma cell lines and its relationship to interleukin-6 expression. Int. J. of the DNA-binding domains from NFAT, Fos and Jun bound specifically to Cancer 55:96. DNA. Nature 392:42. 4. Dummer, W., B. C. Bastian, N. Ernst, C. Schanzle, A. Schwaaf, and E. B. Brocker. 30. Zhang, X., M. H. Wrzeszczynska, C. M. Horvath, and J. E. Darnell, Jr. 1999. 1996. Interleukin-10 production in malignant melanoma: preferential detection of Interacting regions in Stat3 and c-Jun that participate in cooperative transcrip- IL-10-secreting tumor cells in metastatic lesions. Int. J. Cancer 66:607. tional activation. Mol. Cell. Biol. 19:7138.