Cloning and functional characterization of the 5′-regulatory region of the human CD86 gene

Cloning and Functional Characterization of the 5ⴕ-Regulatory Region of the Human CD86 Gene Jianfeng Li, Adriana I. Colovai, Raffaello Cortesini, and Nicole Suciu-Foca ABSTRACT: The induction of CD86 expression by IFN-␥ on the surface of various antigen presenting cells has been previously reported. In order to understand the mechanisms by which the expression of the CD86 gene is regulated by IFN-␥ at the transcriptional level, we have cloned and characterized the 5⬘-flanking region of the human CD86 gene. To functionally analyze the upstream regulatory region of the CD86 gene, a series of luciferase reporter gene constructs were prepared and used for transfection of cells from the monocytic line U937 and Raji B cell line. Under basal conditions, functional activity of these constructs was detected in Raji cells, which show high constitutive expression of the CD86 molecule, but not in U937 cells, which show low expression of CD86 in non-activated state. Induction of CD86 expression by stimulation of U937 cells with IFN-␥ revealed the presence of two functional GAS (gamma-interferon activation site) elements. Gel mobility shift assays showed that these

INTRODUCTION T cells require at least two types of signals from APC in order to become activated, secrete cytokines and attain their effector function [1, 2]. The first signal is delivered by the antigenic peptide/MHC complex, while the second signal is delivered by costimulatory molecules, such as B7 molecules. While the first signal is essential for the initiation of specific T cell activation, the second signal, although non-specific, is critical for the progression of the activation events which culminate with T cell proliferation and cytokine secretion. In the absence of the From the Department of Pathology (J.L., A.I.C., N.S-F.), College of Physicians and Surgeons of Columbia University, New York, NY, USA; and Department of Surgery (R.C.), University Degli Studi di Roma “La Sapienza”, Instituto di II Clinica Chirurgica Servizio Trapianti d’Organo, Rome, Italy. Address reprint requests to: Nicole Suciu-Foca, College of Physicians and Surgeons of Columbia University, 630 West 168th Street, P&S 14-401, New York, NY 10032, USA; Tel: (212) 305-6941; Fax: (212) 3053429. Received December 17, 1999; accepted January 10, 2000. Human Immunology 61, 486 – 498 (2000) © American Society for Histocompatibility and Immunogenetics, 2000 Published by Elsevier Science Inc.

two GAS elements specifically bind an IFN-␥-induced transcriptional complex. The DNA-protein complex was supershifted by antibody to Stat1 ␣ (signal transducer and activator of transcription), but not by antibodies to Stat 2, Stat 3 and Sp1, indicating that GAS elements interact with Stat1 ␣. Point mutations in the GAS elements prevented the formation of DNA-protein complex and significantly reduced the responsiveness of the reporter gene to IFN-␥. These findings suggest that two functional GAS elements within the human CD86 promoter play an important role in the induction of CD86 gene by binding to IFN-␥-induced Stat1 ␣. Human Immunology 61, 486 – 498 (2000). © American Society for Histocompatibility and Immunogenetics, 2000. Published by Elsevier Science Inc. KEYWORDS: costimulatory molecules; CD86 promoter; transcription factors; cytokines

second signal, T cells become unresponsive or anergic [3–5]. The B7 family of costimulatory molecules comprises B7.1 (CD80) and B7.2 (CD86) proteins, both of which can interact with two receptors, CD28 and CTLA-4, that are expressed by T cells [6, 7]. Although both CD80 and CD86 molecules are capable of costimulating T cell proliferation, increasing evidence indicates that they may not deliver identical signals to T cells, and that they may skew Th1 and Th2 phenotypes [8 –10]. Both CD80 and CD86 molecules are expressed on activated B and T cells, monocytes, activated and resting dendritic cells, yet are absent from resting B and T cells [11, 12]. However, resting monocytes express CD86, but not CD80. Furthermore, the kinetic of CD86 expression upon activation is different from that of CD80. Thus, the expression of CD86 on activated B cells is upregulated within 6 h of stimulation, while CD80 becomes detectable only after 24 h [11–13]. These findings indicate that regula0198-8859/00/$–see front matter PII S0198-8859(00)00099-9

Regulation of CD86 Expression

tion of CD80 and CD86 expression is cell-type specific, and that the CD86 molecule may play a distinct role during early phases of the interaction between T cells and APC. CD80 and CD86 molecules are members of the Ig supergene family and display similar overall structures, with an extracellular domain containing two Ig-like regions, a transmembrane domain, and a cytoplasmic tail [7, 14, 15]. The genomic organization of the CD86 gene is similar to that of CD80, with the exception of the 3⬘ end of the gene. Notably, the cytoplasmic tail of CD86 spans three exons (exons 6 to 8), while that of CD80 is entirely contained in the final exon, a finding which suggests potential signaling capacity of the CD86 molecule through alternate splicing [16]. The physiological importance of the CD80 and CD86 molecules has been revealed by studies using transgenic mice [17, 18]. CD86-deficient mice show lack of germinal center formation and immunoglobulin class switching after systemic administration of antigen. In contrast, CD80-deficient mice have only slightly reduced levels of antibodies following immunization. A profound defect in antibody formation and complete absence of germinal center formation have been found in mice lacking both costimulating ligands [18]. These findings point to a major role played by CD86 in activation and differentiation of B cells, which only partially overlaps with the role of the CD80 molecule. Induction of CD86 expression on APC appears to be a critical event in the initiation of immune responses against foreign antigens, promoting T helper activation and proliferation [19]. In contrast, CD86 expression on anergic self-reactive B cells is repressed, preventing T cell costimulation [20]. Hence, regulation of CD86 expression is important for the outcome of the immune response to self and foreign antigens. Several studies have shown that B7 molecules may also play a role in eliciting T cell-mediated immunity against tumors [21–25]. In mouse models, CD80 and CD86-transfected tumor cells have been shown to induce strong antitumor responses both in vitro and in vivo [23, 24]. Furthermore, human tumor cells which had been transfected with a B7 retroviral vector have been reported to elicit T cell reactivity [25]. These findings suggest that manipulation of B7 expression may provide a therapeutic means for controlling the immune response in autoimmunity and cancer. Because of the important role of the CD86 molecule as a costimulatory and, possibly, signaling molecule, we have cloned and functionally characterized the promoter of the human CD86 gene. Our data demonstrates that the presence of GAS elements in the promoter of this gene, which specifically bind Stat1 ␣ transcription factor, is essential for the regulation of transcriptional activity in response to IFN-␥.

487

MATERIALS AND METHODS Nucleic Acid Sequence Data The nucleotide sequence data reported in this article has been submitted to the EMBL/GenBank nucleotide sequence databases and has been assigned the accession number AF099105. Cloning and Sequencing of the 5ⴕ-Flanking Region of the Human CD86 Gene A bacteriophage library was constructed by cloning human lymphocyte genomic DNA into Lambda DASH 11 (Stratagene, La Jolla, CA, USA). The library was screened using a synthetic oligonucleotide probe (Ps-1) corresponding to nucleotide ⫹17 to ⫹101 of the human CD86 genomic DNA sequence [15]. Four positive plaques were obtained. In one of these plaques, Southern blot analysis revealed a 1.4 kB insert, which also hybridized to a second oligonucleotide probe (Ps-2), corresponding to ⫹2 to ⫹43 of the CD86 genomic DNA. This insert was subcloned into plasmid pBluescript II SK (⫹) (Stratagene) for DNA sequencing. Vector-specific T3 and T7 primers were used to obtain initial sequence information. Primer Extension Assay In order to map the transcription start site of the CD86 gene, a primer extension assay was performed as previously described [26]. Briefly, two synthetic oligonucleotide primers, Pe-1 corresponding to ⫺304 to ⫺283 of the 5⬘-flanking region of CD86 gene and Pe-2 corresponding to ⫹43 to ⫹61, were 32P-labeled, and then annealed to 5 ␮g of poly(A)⫹ RNA prepared from IFN␥-treated U937 cells. Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI, USA) was used for primer extension. Dideoxynucleotide sequencing of CD86 genomic DNA using the same primers as above was performed. Both the extension and DNAsequencing products were analyzed in a 6% polyacrylamide-urea gel and then visualized by autoradiography. Cell Culture U937 cells and Raji cells (ATCC, Rockville, MD, USA) were propagated in RPMI 1640 containing 10% fetal bovine serum (Gibco, Grand Island, NY, USA). Prior to analysis, cells were starved for 8 h in serum-free media, and then cultured in the presence of IFN-␥ (500 U/ml) or BSA (1 ␮g/ml) (Gibco) for the indicated time periods. Nuclear Run-on Transcription Assay Nuclear run-on transcription assay was performed as previously described [27]. Briefly, nuclei prepared from IFN-␥-treated U937 cells were incubated at 30°C for 1 h in the presence of ATP (0.5 mM), CTP (0.5 mM), GTP (0.5 mM), and 200 ␮Ci of [32P]UTP (3000 Ci/mmol).

488

FIGURE 1 Nucleotide sequence of the 5⬘-flanking region of the human CD86 gene. The transcription start site (⫹1), as determined by primer extension analysis, is indicated. The sequences corresponding to GAS elements are underlined.

Total RNA was then extracted and resuspended in hybridization buffer containing PIPES (50 mM, pH 6.8), EDTA (10 mM), NaCl (600 mM), SDS (0.2%), and denatured salmon test DNA (100 ␮g/ml). RNA was hybridized to denatured human CD86 cDNA and ␤-actin cDNA, which had been slot-blotted on nylon filters. After hybridization, the filters were washed, briefly airdried and exposed to X-ray film. To prepare the nylon filter containing CD86 and ␤-actin cDNA, cDNA was denatured by incubation with 0.3 N NaOH for 30 min at 65°C, spotted onto the nylon filter and cross-linked using a UV-cross-linker (Fisher, Pittsburgh, PA, USA). Flow Cytometry To determine cell surface expression of the CD86 molecule, U937 cells were starved in serum-free medium for 8 h and then stimulated with IFN-␥ (500 U/ml) for the indicated time. Cells were washed twice with PBS, stained with saturating amounts of FITC-conjugated isotype control mAb or FITC-conjugated CD86 mAb (Becton Dickinson, San Jose, CA, USA), and analyzed on a FACScan instrument.

J. Li et al.

DNase I Footprint Assay DNase I footprint assay was performed using the assay system provided by Promega. The 3⬘-labeled DNA fragment, corresponding to position ⫺688 to ⫺1247 upstream of the transcription start site of CD86 gene, was incubated on ice in 0.05 ml of 1⫻ binding buffer for 40 min with 7 ␮g of nuclear extract prepared from U937 cells treated with either IFN-␥ or BSA. After digestion for 1 min at room temperature with DNase I, reactions were terminated by addition of the stop reagent. After extraction with phenol/chloroform and precipitation with ethanol, samples were dissolved in loading buffer and analyzed on a 6% polyacrylamide-urea gel. Dideoxynucleotide sequencing of the same 5⬘-flanking region of the CD86 gene was performed in parallel. Electrophoretic Mobility Shift Assays (EMSA) The double-stranded oligonucleotide probes corresponding to the CD86 GAS elements were prepared using the following synthetic oligonucleotides: 5⬘-GCATTTTG GTCTAAACTAA-3⬘ (CD86-GAS 3), and 5⬘-GTTTA ACTTGCTTTAAAGCT-3⬘(CD86-GAS 4). The labeled probes were incubated for 30 min with 10 ␮g of nuclear extract prepared from U937 cells which had been cultured for 4 h with either IFN-␥ (500 U/ml) or BSA (1 ␮g/ml). For competition assays, the nuclear extract was incubated for 15 min at room temperature with

Regulation of CD86 Expression

489

nology, Santa Cruz, CA, USA) and then incubated with the labeled-CD86 GAS probe. Samples were then analyzed on a 6% polyacrylamide gel and exposed to X-ray films. Generation of Reporter Constructs Luciferase reporter constructs carrying 5⬘ deletions of the CD86 promoter were generated by cloning into the pGL3-Basic vector (Promega) different 5⬘-flanking regions of the CD86 gene, which had been amplified by PCR. Site-directed mutations within the CD86-GAS were introduced by PCR using a sense primer containing the mutated CD86-GAS sites. All reporter constructs, which were used in transfection assays, were confirmed by DNA-sequencing of both strands. Transfection Experiments Transfection experiments were performed using DMRIE C, according to the manufacturer’s recommendations (Gibco). Briefly, 1 ⫻ 106 cells were co-transfected using 5 ␮g of the testing constructs, 1 ␮g of pSV ␤-galactosidase plasmid (Promega), which served as internal control, and 8 ␮l of DMRIE C. Twelve hours after transfection, cells were treated with IFN-␥ (500 U/ml) or BSA (1 ␮g/ml) for 24 h, and then harvested and tested for luciferase and ␤-galactosidase activity. Luciferase activity was normalized to ␤-galactosidase activity. The activity measured for the testing constructs was compared to that obtained for the negative control.

FIGURE 2 Identification of the transcription start site by primer extension analysis. Two oligonucleotide primers (Pe-1 and Pe-2) were derived from the 5⬘-flanking region or the first exon of CD86 gene (as indicated in “Material and Methods”). The primer extended product obtained with primer Pe-2 is indicated by the arrow. No extended product was obtained using Pe-1 (data not shown). Shown on the left side of the figure is the schematic diagram of the transcription start site mapped in this study.

100-fold excess of unlabeled CD86-GAS probe, GAS consensus sequence or Sp1 consensus sequence prior to the addition of the 32P-labeled CD86-GAS probe. For supershift assay, the nuclear extract was first incubated with antibody to Stat1-␣, Stat 2 or Stat 3 (Santa Cruz Biotech-

RESULTS Isolation and Characterization of the Human CD86 Promoter To investigate the regulatory elements which control the expression of human CD86 gene, we have cloned a 1323 bp portion of the 5⬘-flanking region of CD86 gene from a human genomic DNA library. This 1323 bp insert was subcloned into pBluescript II SK (⫹) vector and DNA sequenced using vector-specific T3 and T7 primers. Comparison of this sequence with the reported DNA sequence of the human CD86 gene [14, 15] indicated that the 1323 bp fragment contained at the 3⬘-end a segment of 76 bp that matched the first exon of CD86 gene (Fig. 1). The remaining 1247 bp of the DNA sequence extended into the 5⬘-flanking region of the CD86 gene. Primer extension assay was performed to map the transcription start site of the CD86 gene. Two primers, derived from sequences located within the 1.3 kb sequence, were designed. The first oligonucleotide, Pe-1, corresponds to the fragment ⫺304 to ⫺283 of the 5⬘-flanking region of the CD86 gene, while the second oligonucleotide, Pe-2, covered the region from ⫹43 to ⫹61 (Fig. 1). An extension product was found using

490

FIGURE 3 Functional analysis of the CD86 promoter under basal conditions by transfection assay. The characteristics of the CD86 promoter-luciferase reporter constructs used for transfection The position of each construct relative to the transcription start site (⫹1) and the position of GAS elements within pGL.Luc./1247 construct are indicated. pGL.Luc./ 1247R construct, containing an identical DNA fragment found in pGL.Luc./1247, but in opposite orientation, served as negative control. The constructs were transiently transfected into Raji cells and U937 cells. Twenty-four hours after transfection, cells were harvested for luciferase activity/␤-galactosidase activity assay. Luciferase activity values obtained for the testing constructs are relative to the activity obtained from negative control. Data represent the mean ⫾ SE of three independent experiments.

primer Pe-2, but not with primer Pe-1 (Fig. 2). These results indicate that the human CD86 gene has one major transcription start site, adenosine, designated as “⫹1” (Fig. 1). Computer analysis was used to match potential transcription factor binding sites in the 5⬘-flanking sequence of CD86 gene by comparison with known consensus sequences of transcription factors. The search using Transfast program revealed that the 5⬘-flanking sequence of CD86 gene lacks a consensus TATA or CAAT box, but comprises multiple potential transcription factor binding sequences, including four putative GAS elements, that are located in the region from ⫺1114 to ⫺1223 upstream of the transcription start site of CD86 gene, namely CD86-GAS1, CD86-GAS2, CD86-GAS3, and CD86-GAS4 (Fig.1). Functional Analysis of the 5ⴕ-Flanking Region of the CD86 Gene In order to determine the promoting function of the 5⬘-flanking region of CD86 gene identified in this study,

J. Li et al.

we generated a series of luciferase reporter constructs representing 5⬘ deletions of the CD86 promoter (Fig. 3). The constructs comprised DNA fragments starting from position ⫺212 (pGL.Luc./212), ⫺305 (pGL.Luc./305), ⫺502 (pGL.Luc./502), ⫺721 (pGL.Luc./721), or ⫺1247 (pGL.Luc./1247) upstream of the transcription start site to position ⫹45 of the CD86 gene. The construct for the negative control contained the identical DNA fragment found in pGL.Luc./1247 in an opposite orientation (pGL.Luc./1247R). The pGL-3 vector (Promega), containing the luciferase gene under the control of SV40 promoter, was used as positive control. These constructs were transiently transfected into U937 cells, which show low expression of cell surface CD86 molecule under basal conditions (Fig. 4A), and Raji cells, which express constitutively high levels of CD86 (data not shown). Thirtytwo hours after transfection, cells were harvested and tested for luciferase/␤-galactosidase activity. The luciferase activity was normalized to ␤-galactosidase activity and compared to the negative control. As shown in Fig. 3, all constructs showed transcriptional activity in Raji cells when compared to the negative control (p ⬍ 0.01 for each case). pGL.Luc./212, pGL.Luc./305, pGL.Luc./ 502, pGL.Luc/721, and pGL.Luc./1247 showed a 3.1fold increase (p ⬍ 0.05), 3.2-fold increase (p ⬍ 0.01), 4.7-fold increase (p ⬍ 0.01), 4.9-fold increase (p ⬍ 0.01), and 5.2-fold increase (p ⬍ 0.01), respectively, compared to the negative control. These data clearly indicate that the 5⬘-flanking region starting at ⫺1247 has promoter activity in cells which show constitutive expression of the CD86 molecule. However, no significant luciferase activity was detected in transfected U937 cells, when compared with the negative control (p ⬎ 0.5 for each construct, Fig. 3). The results suggest that the

Regulation of CD86 Expression

FIGURE 4 Analysis of expression of CD86 gene in response to IFN-␥ in U937 cells. A. Flow cytometry. U937 cells were incubated in medium containing either IFN-␥ (bold lines) or BSA (thin lines) for the indicated time. Cells were stained with FITC-conjugated CD86 mAb, and analyzed on a Becton Dickinson FACScan. Histograms obtained by using a control isotype-matched mAb (dotted lines) are also depicted. The percent and mean channel fluorescence of the cell population expressing the CD86 protein are indicated. The data are representative of three independent experiments. B. Nuclear run-on transcription assay. Nuclei were prepared from cells exposed for 8 h to either IFN-␥ (500U/ml) or BSA (1 ␮g/ml). 32P-labeled transcription products were hybridized to the human CD86 cDNA or ␤-actin cDNA slot-blotted onto nylon filters, and exposed to X-ray films. Densitometric analysis is shown in the inset.

491

492

FIGURE 5 Functional analysis of the CD86 promoter under IFN-␥-stimulation by transfection assay. The CD86 promoter-luciferase reporter constructs were transiently transfected into U937 cells. After exposure to IFN-␥ (500 U/ml) for 24 h, the transfected cells were harvested for luciferase/␤-gal activity assay. Values represent the average of at least three independent experiments.

transcriptional control of the CD86 gene is cell-type specific, and are consistent with the differential pattern of CD86 expression observed in vivo. Requirement of GAS for IFN-␥-Induced CD86 Expression in U937 Cells To study the regulation of CD86 expression by IFN-␥, U937 cells were cultured in the presence of this cytokine for 0 – 48 h and analyzed by flow cytometry. Expression of CD86 increased significantly after 24 h of stimulation, and continued to raise at 48 h (Fig. 4A). To determine whether IFN-␥ regulates the expression of CD86 in U937 cells at the level of transcription, a nuclear run-on transcription assay was performed. As shown in Fig. 4B,

J. Li et al.

there was a 5.1-fold increase in the transcriptional activity of CD86 gene in IFN-␥ treated U937 cells over the level seen in untreated cells. These data demonstrate that the regulation of the CD86 expression in U937 cells by IFN-␥ is, at least in part, controlled at the level of transcription. In order to test the functional activity of the 5⬘flanking region of the CD86 gene, we transfected U937 cells with each of the 5⬘-deletion CD86 promoter-Luc. reporter constructs described above. Transfected cells were incubated in the presence of IFN-␥ for the indicated time and then harvested for luciferase assay. As shown in Fig. 5, the relative luciferase activity in pGL.Luc./1247transfected cells increased significantly compared to untreated cells (4.3-fold, p ⬍ 0.01) or to the negative control (5.2-fold, p ⬍ 0.01). However, no significant luciferase activity was identified in cells transfected with pGL.Luc./212, pGL.Luc./305, pGL.Luc./502, or pGL.Luc./721, when compared to either untreated cells or negative control (p ⬎ 0.5). These results indicate that the 5⬘-flanking region from ⫺1247 to ⫺721 upstream of the transcription start site of CD86 gene contains IFN-␥ responsive elements. To determine the precise location of the IFN-␥-responsive motifs within the human CD86 promoter, DNase I footprinting assay was performed using the 32 P-labeled DNA fragment encompassing ⫺688 to ⫺1247 upstream of the transcription site of CD86 gene. Two protected regions, corresponding to GAS3 and GAS4, were observed when the DNA fragment was incubated with nuclear extract from IFN-␥-stimulated U937 cells (Fig. 6). In contrast, no protected region for either CD86-GAS1 or CD86-GAS2 was noted. The core nucleotide sequences in the protected regions matched the consensus sequence of GAS (TTNNNNNAA). These data indicate that the putative CD86-GAS 3 and 4 are capable of binding to nuclear complexes induced by IFN-␥ and are, therefore, potentially involved in upregulation of CD86 molecule by IFN-␥ in U937 cells. In order to test the functional role of CD86-GAS 3 and 4, electrophoresis mobility shift assays (EMSA) were performed using the CD86-GAS3 and CD86-GAS4 probes indicated in Fig. 7A. Compared to BSA-treatment (Fig. 7B, lane 1, 3 and 5), treatment of U937 cells with IFN-␥ (lane 2, 4, and 6) induced a significant increase in the intensities of the bands corresponding to retarded CD86-GAS probes (CD86-GAS3: lane 1 and 2, CD86-GAS4: lane 3 and 4 or GAS consensus probe: lane 5 and 6). In the competitive assay (Fig. 7C), the retarded bands of the CD86-GAS3 (lane 1, 2, 3, 4) and CD86GAS4 (lane 5, 6, 7, 8) probes were eliminated by preincubation of the nuclear extract from IFN-␥-treated U937 cells with 100-fold excess of non-labeled CD86GAS3 (lane 9 and 12), CD86-GAS4 (lane 10 and 13), or

Regulation of CD86 Expression

493

FIGURE 6 Identification of GAS elements in the 5⬘-flanking region of the human CD86 gene by DNase I footprinting assay. 32P-labeled DNA fragment from ⫺1345 to ⫺734 upstream of the transcription start site of CD86 gene was incubated with nuclear extract from U937 cells treated with either IFN-␥ or BSA. After digestion with DNase I, the products were analyzed on a 6% denatured polyacrylamide gel and exposed to X-film. Comparison of the core nucleotide sequences in the protected areas with GAS consensus sequence is shown.

GAS consensus sequence (lane 11 and 14). In contrast, preincubation of the nuclear extract with non-labeled GAS oligonucleotides, carrying point mutations as indicated in Fig. 7A, showed no such effect (CD86-GAS3m: lane 1 and 5, CD86-GAS4m: lane 2 and 6). Similarly, SP1 (lane 3 and 7) or AP2 (lane 4 and 8) consensus oligonucleotides did not compete with labeled probes for binding to nuclear extracts. This data indicates that the CD86-GAS elements bind specifically IFN-␥-activated transcription factors and further suggest that the CD86GAS elements are required for upregulation of CD86 expression. In order to identify the transcription factors that

respond to IFN-␥ and bind to the CD86-GAS elements, supershift assays were performed using antibodies to Sp1, Stat1 ␣, Stat2, and Stat3. As indicated in Fig. 7D, only anti-Stat1 ␣ antibody recognized the CD86-GAS3/protein complex and led to supershift. Similar results were obtained for CD86-GAS4 (data not shown). This finding indicates that transcription factor Stat1 ␣ mediates IFN␥-induced upregulation of CD86 expression by binding GAS elements within the human CD86 promoter. Although the GAS elements within the CD86 promoter seem to regulate activation of the CD86 gene induced by IFN-␥, it is possible that other elements contained within the 5⬘-flanking region play a role in

494

FIGURE 7 Specific binding of CD86-GAS to IFN-␥-induced Stat1␣ by EMSA. A. The indicated probes were employed in the experiments described below. The core nucleotide sequences of wild CD86-GAS elements are bolded, and the mutated nucleotides in CD86-GAS elements are underlined. B. Binding of CD86-GAS probes to nuclear proteins. The nuclear extract prepared from U937 cells (10 ␮g) treated with either IFN-␥ or BSA was incubated with 32P-labeled CD86-GAS3 (lanes 1 and 2), CD86-GAS4 (lanes 3 and 4), or GAS consensus (lanes 5 and 6). The products were analyzed in 6% polyacrylamide gel. C. Competition assay. The nuclear

J. Li et al.

extract prepared from IFN-␥-treated U937 cells was incubated with each of the indicated competitors prior to incubation with 32P-labeled CD86-GAS3 (lanes 1, 2, 3, 4, 9, 10, and 11) or CD86-GAS4 (lanes 5, 6, 7, 8, 12, 13, and 14). D. Supershift assay. The nuclear extract prepared from IFN-␥-treated U937 cells was incubated in the absence of antibodies (lane 1) or in the presence of anti-Sp1 (lane 2), anti-Stat1 ␣ (lane 3), antiStat 2 (lane 4), or anti-Stat 3 (lane 5) antibody prior to addition of 32P-labeled CD86-GAS3. The shifted bands are indicated by the arrow.

Regulation of CD86 Expression

FIGURE 8 Effect of mutations within CD86-GAS elements. The position of the substitutions introduced by PCRdirected mutagenesis within GAS3 and GAS4 elements is indicated in Fig. 7A. Luciferase activity was measured in IFN-␥- and BSA-treated U937 cells, which had been transfected with pGL.Luc./1247, pGL.Luc./GASm (containing mutated GAS3 and GAS4 elements) or pGL.Luc./1247R (negative control) construct. Luciferase activity was normalized to ␤-galactosidase activity. Data represent the mean ⫾ SE of three independent experiments.

regulating the activity of the CD86 gene. In order to evaluate the importance of the GAS elements in the response to IFN-␥, point mutations in the CD86-GAS elements were introduced as indicated in Fig. 7A. These constructs were then transfected into U937 cells. Under basal conditions, mutational inactivation of GAS elements had no effect on the relative luciferase activity compared with wild-type pGL.Luc./1247 (p ⬎ 0.5 for each case, Fig. 8). This data indicates that GAS elements are not key factors for maintaining basal levels of the CD86 promoter function, and may not be the elements responsible for the tissue-specific control of CD86 expression. Mutations of CD86-GAS elements, however, caused a 65% decrease in responsiveness to IFN-␥ compared to the wild-type construct (p ⬍ 0.01). Hence, CD86-GAS elements play an important role in mediating IFN-␥-induced expression of the CD86 gene. The CD86 promoter also appears to contain elements distinct

495

of GAS, which may contribute to the regulation of gene activation induced by IFN-␥. DISCUSSION The expression of the costimulatory molecule CD86 is regulated through the engagement of distinct cell surface molecules, such as the B cell receptor, CD40 and several cytokine receptors [13, 28 –32]. Among the cytokines which modulate the expression of the CD86 molecule, IFN-␥ and IL-4 enhance CD86 expression on B cells and monocytes, whereas IL-10 acts as an inhibitor [31–34]. The finding that different cytokines can act as activators or inhibitors of CD86 expression suggests the existence of both positive and negative regulatory elements which control the transcription of this gene. Such regulatory elements are likely to be responsible for maintaining basal levels and inducing or repressing the expression of the CD86 gene. Recent studies on a number of IFN-␥-inducible genes showed that treatment of cells with IFN-␥ leads to the phosphorylation of cell surface receptors, which then phosphorylate and activate the receptor-associated kinases JAK-1 and JAK-2 [35–37]. These kinases in turn phosphorylate the cytoplasmic STAT-1 ␣, which is dimerized and translocated to the nucleus [38, 39]. The DNA binding sequence of Stat1 ␣ has been identified as GAS. This consensus sequence, TTNNNNNAA, binds Stat1 and Stat 3 hetero- and homodimers and plays an

496

important role in the regulation of genes that respond to IFN-␥ [40 – 42]. In addition, the DNA sequences described as acute phase response elements and sis-inducible elements have also been recognized as GAS-like seqences [43]. Genes for such proteins as c-fos, IFN-␥ regulatory factor (IRF-1), and ␣2-macroglobulin contain GAS-like sequences in their promoters, and bind both acute phase response and ␥-activating factors [43, 44]. In the present study, we have explored the role of the 5⬘-flanking region of the human CD86 gene in regulating expression of this gene. The cloning and functional analysis of a 1.3 kB fragment upstream of the transcription site of the CD86 gene indicated that this region has promoter activity and comprises two functional GAS elements, which mediate IFN-␥-induced expression of the CD86 gene. Under basal conditions, promoter activity was detected in Raji cells, a tumor B cell line that shows constitutive expression of CD86, but not in U937 cells, a tumor myeloid cell line that shows inducible expression of CD86. However, stimulation with IFN-␥induced promoter activity in both Raji and U937 cells. The search for IFN-␥-responsive motifs within the CD86 promoter, using reporter-gene constructs that encompass different regions of the promoter, revealed that the region from ⫺721 to ⫺1247 of the 5⬘-flanking sequence of the CD86 gene comprises IFN-␥-responsive elements. Point mutations introduced in two GAS elements located within this region, upstream of the transcription start site, reduced significantly the responsiveness of the CD86 promoter to IFN-␥, indicating that these two GAS elements have functional activity. Furthermore, the transcription factor Stat1-␣ binds GAS elements within this region indicating that Stat1-␣ mediates IFN-␥induced upregulation of CD86 gene expression. In a previous study, we have characterized an NF-␬B consensus binding site located at ⫺612 upstream of the transcription starting site of the CD86 gene [45]. This site specifically binds the NF-␬B p50/p65 heterodimer, and plays an important role in the transcriptional activation of CD86 expression in B cells pre-exposed to activated Th cells. Many genes containing NF-␬B binding elements have been identified [46, 47]. Such genes often contain binding sites for other transcription factors, including IFN-␥-activated factors. Our finding that the CD86 promoter contains NF-␬B binding sites as well as GAS elements suggests that the interaction of distinct transcription factors with the corresponding control regions are used for optimal regulation of CD86 gene expression. Taken together, the data presented in this study indicate that the transcriptional regulation of CD86 expression is strictly controlled under homeostatic conditions as well as under stimulatory conditions. Furthermore, the 5⬘flanking region of the CD86 gene contains two functional

J. Li et al.

GAS elements, which bind Stat1 ␣ transcription factor and play an important role in the induction of CD86 expression by IFN-␥. REFERENCES 1. Harding FA, McArthur JG, Gross JA, Raulet DH, Allison JP: CD28-mediated signalling costimulates murine T cells and prevents induction of anergy in T cell clones. Nature 356:607, 1992. 2. Linsley PS, Brady W, Grosmarie L, Aruffo A, Damle NK, Ledbetter JA: Binding of the B cell activation antigen B7 to CD28 costimulates T cell proliferation and interleukin 2 mRNA accumulation. J Exp Med 173:721, 1991. 3. Boussiotis VA, Freeman GJ, Gribben JG, Nadler LM: The role of B7-1/B7-2:CD28/CTLA-4 pathways in the prevention of anergy, induction of productive immunity and down-regulation of the immune response. Immunol Rev 153:5, 1996. 4. Jenkins MK, Schwartz RH: Antigen presentation by chemically modified splenocytes induces antigen-specific T cell unresponsiveness in vitro and in vivo. J Exp Med 165:302, 1987. 5. Gimmi CD, Freeman GJ, Gribben JG, Gray G, Nadler LM: Human T cell clonal anergy is induced by antigen presentation in the absence of B7 costimulation. Proc Natl Acad Sci USA 90:6586, 1993. 6. Lenschow DJ, Walunas TL, Bluestone JA: CD28/B7 system of T cell costimulation. Ann Rev Immunol 14:233, 1996. 7. Azuma M, Ito D, Yagita H, Okumura K, Phillips JH, Lanier LL, Somoza C: B7 antigen is a second ligand for CTLA-4 and CD28. Nature 366:76, 1993. 8. Slavick JM, Huthcroft JE, Bierer BE: CD80 and CD86 are not equivalent in their ability to induce tyrosine phosphorylation of CD28. J Biol Chem 274:3116, 1999. 9. Freeman GJ, Boussiotis VA, Anumanthan A, Bernstein GM, Ke XY, Rennert PD, Gray GS, Gribben JG, Nadler LM: B7.1 and B7.2 do not deliver identical costimulatory signals, since B7.2 but not B7.1 preferentially costimulates the initial production of IL-4. Immunity 2:253, 1995. 10. Kuchroo VK, Das MP, Brown JA, Ranger AM, Zamvil SS, Sobel RA, Weiner HL, Nabavi N, Glimcher LH: B7-1 and B7-2 costimulatory molecule activate differentially the Th1/Th2 developmental pathways: application to antoimmune disease therapy. Cell 90:707, 1995. 11. Hathcock LS, Laszlo G, Pucillo C, Linsley P, Hodes RJ: Comparative analysis of B7-1 and B7-2 costimulatory ligands: expression and function. J Exp Med 180:631, 1994. 12. Azuma M, Ysell H, Phillips JH, Spits H, Lanier LL: Functional expression of B7/BB1 on activated T lymphocytes. J Exp Med 177:845, 1993. 13. Lenschow DJ, Sperling AI, Cooke MP, Freeman G, Rhee L, Decker DC, Gray G, Nadler LM, Goodnow CC, Bluestone JA: Differential upregulation of the B7-1 and B7-2

Regulation of CD86 Expression

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

costimulatory molecules following immunoglobulin receptor engagement by antigen. J Immunol 153:1990, 1994. Freeman GJ, Gribben GJ, Boussiotis VA, Hg JW, Restivo VA Jr, Lombard LA, Gray GS, Nadler LM: Cloning of B7-2: a CTLA-4 counter-receptor that stimulates human T cell proliferation. Science 262:909, 1993. Jellis CL, Wang SS, Rennert P, Borriello F, Sharpe AH, Green NR, Gray GS: Genomic organization of the gene coding for the costimulatory human B-lymphocyte antigen B7-2 (CD86). Immunogenetics 42:85, 1995. Boriello F, Oliveros J, Freeman GJ, Nadler LM, Sharpe AH: Differential expression of alternated mB7.2 transcripts. J Immunol 155:5490, 1995. Borriello F, Sethna MP, Boyd SD, Schweitzer AN, Tivol EA, Jacoby D, Strom TB, Simpson EM, Freeman GJ, Sharpe, AH: B7-1 and B7-2 have overlapping, critical roles in immunoglobulin class switching and germinal center formation. Immunity 6:303, 1997. Schweitzer AN, Sharpe AH: Studies using antigen-presenting cells lacking expression of both B7-1 (CD80) and B7-2 (CD86) show distinct requirements for B7 molecules during priming versus restimulation of Th2 but not Th1 cytokine production. J Immunol 61:2762, 1998. Chen C, Nabavi N: In vitro induction of T cell anergy by blocking B7 and early T cell costimulatory molecule ETC-1/B7-2. Immunity 1:147, 1994. Rathmell JC, Fournier S, Weintraub BC, Allison JP, Goodnow CC: Repression of B7.2 on self-reactive B cells is essential to prevent proliferation and allow fas-mediated deletion by CD4⫹ T cells. J Exp Med 188:651, 1998. Li J, Yang Y, Inoue H, Mori M, Akiyoshi T: The expression of costimulatory molecules CD80 and CD86 in human carcinoma cell lines: its regulation by interferon-␥ and interleukin-10. Cancer Immunol Immunother 43: 213, 1996. Zheng Z, Takahashi M, Aoki S, Toba K, Liu A, Osman Y, Takahashi H, Tsukada N, Suzuki N, Nikkuni K, Furukawa T, Koike T, Aizawa Y: Expression patterns of costimulatory molecules on cells derived from human hematological malignancies. J Exp Clin Res 17:251, 1998. Chen L, Brady WA, Hellstrom I, Hellstrom KE, Ledbetter JA, McGowan P, Linsley PS: Costimulation of antitumor immunity by the B7 counterreceptor for the T lymphocytes molecules CD28 and CTLA-4. Cell 71:1093, 1992. Towsend SE, Allison, JP: Tumor rejection after direct costimulation of CD8⫹ T cells by B7-transfected melanoma cells. Science 259:358, 1993. Dohring C, Angman L, Spagnoli G, Lanzavecchia A: T-helper- and accessory-cell–independent cytotoxic responses to human tumor cells transfected with a B7 retroviral vector. Int J Cancer 57:754, 1994. Li J, Schmidt AM: Characterization and functional anal-

497

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

ysis of the promoter of RAGE, the receptor for advanced glycation end products. J Biol Chem 272:16498, 1997. Li J, Qu XQ, Schmidt AM: Sp1-binding element in the promoter of RAGE are essential for amphoterin-mediated gene expression in cultured neuroblastoma cells. J Biol Chem 276:30870, 1998. Boussiotis VA, Freeman GJ, Gribben JG, Nadler LM: The role of B7-1/B7-2:CD28/CTLA-4 pathways in the prevention of anergy, induction of productive immunity and down-regulation of the immune response. Immunol Rev 153:5, 1996. Ranheim EA, Kipps TJ: Activated T cells induce expression of B7/BB1 on normal or leukemic B cells through a CD40-dependent signal. J Exp Med 177:925, 1993. Valle A, Aubry JP, Durand I, Banchereau J: IL-4 and IL-2 upregulate the expression of antigen B7, the B cell counterstructure to T cell CD28: an amplification mechanism for T-B cell interactions. Int Immunol 3:229, 1991. Stack RM, Lenschow DJ, Gray GS, Bluestone JA, Fitch FW: IL4-treatment of small splenic B cells induces costimulatory molecules B7-1 and B7-2. J Immunol 152: 5723, 1994. Freedman AS, Freeman GJ, Rhynhart K, Nadler LM: Selective induction of B7/BB-1 on interferon-gamma stimulated monocytes: a potential mechanism for amplification of T cell activation through the CD28 pathway. Cell Immunol 137:429, 1991. Bulens C, Willems F, Delvaux A, Pierard G, Delville JP, Velu T, Goldman M: Interleukin-10 differentially regulates B7-1 (CD80) and B7-2 (CD86) expression of human peripheral blood dendritic cells. Eur J Immunol 25:2421, 1995. Willems F, Marchant A, Delville JP, Gerard C, Delvaux A, Velu T, de Boer M, Goldman M: Interleukin-10 inhibits B7 and intracellular adhesion molecule-1 expression on human monocytes. Eur J Immunol 24:1007, 1994. Velazquez L, Fellous M, Stark GR, Pellegrini S: A protein tyrosine kinase in the interferon alpha/beta signaling pathway. Cell 70:313, 1992. Igarashi K, Garotta G, Ozmen L, Ziemiecki A, Wilks AF, Harp AG, Larner AC, Finbloom DS: Interferon-gamma induces tyrosine phosphorylation of interferon-gamma receptor and regulated association of protein tyrosine kinases, Jak 1 and Jak 2, with its receptor. J Biol Chem 269:143, 1994. Ihle JN: The Janus kinase family and signaling through members of the cytokine receptor superfamily. Proc Soc Exp Biol Med 206:268, 1994. Lew DJ, Decker T, Strehlow I, Darnell JE: Overlapping elements in the guanylate-binding protein gene promoter mediate transcriptional induction by alpha and gamma interferons. Mol Cell Biol 11:182, 1991. Wright TM, Farber M: 5⬘ regulatory region of a novel cytokine gene mediates selective activation by interferon gamma. J Exp Med 173:417, 1991.

498

40. Khan KD, Shuai K, Lindwall G, Maher SE, Darnell JE, Bothwell AL: Induction of the Ly-6A/E gene by interferon alpha/beta and gamma requires a DNA element to which a tyrosine-phosphorylated 91-kDa protein binds. Proc Natl Acad Sci USA 90:6806, 1993. 41. Wegenka U, Lutticken C, Buschmann J, Yuan J, Lottspeich F, Muller-Esterl W, Schindler C, Roeb E, Heinrich P, Horn F: The interleukin-6-activated acute-phase response factor is antigenically and functionally related to members of the signal transducer and activator of transcription (STAT) family. Mol Cell Biol 14:3186, 1994. 42. Wagner BJ, Hayes TE, Hoban CJ, Cochran BH: The SIF binding element confers sis/PDGF inducibility onto the c-fos promoter. EMBO J 9:4477, 1990. 43. Wegenka U, Buschmann J, Lutticken C, Heinrich P, Horn F: Acute-phase response factor, a nuclear factor binding to acute-phase responsive elements, is rapidly

J. Li et al.

44.

45.

46.

47.

activated by interleukin-6 at the posttranslational level. Mol Cell Biol 13:276, 1993. Tanaka N, Kawakami T, Taniguchi T: Recognition DNA sequences of interferon regulatory factor1 (IRF-1) and IRF-2, regulators of cells growth and the interferon system. Mol Cell Biol 13:4531, 1993. Li J, Liu Z, Jiang S, Cortesini R, Lederman S, Suciu-Foca N: T suppressor lymphocytes inhibit NF-␬B-mediated transcription of CD86 gene in APC. J Immunol 163: 6386, 1999. Ouaaz F, Li M, Beg AA: A critical role for the RelA subunit of nuclear factor ␬B in regulation of multiple immune response genes and in fas-induced cell death. J Exp Med 189:999, 1999. Thanos D, Maniatis T: NF-␬B: a lesson in family values. Cell 80:529, 1995.