Protein inhibitor of activated signal transducer and activator of transcription (STAT)-1 (PIAS-1) regulates the IFN-γ response in macrophage cell lines

Cellular Signalling 14 (2002) 537 – 545 www.elsevier.com/locate/cellsig

Protein inhibitor of activated signal transducer and activator of transcription (STAT)-1 (PIAS-1) regulates the IFN-g response in macrophage cell lines Eliana M. Cocciaa,*, Emilia Stellaccib, Roberto Orsattib, Eleonora Benedettib, Elena Giacominia, Giovanna Marzialib, Benigno C. Valdezc, Angela Battistinib a Laboratory of Immunology, Viale Regina Elena 299, 00161 Rome, Italy Laboratory of Virology, Istituto Superiore di Sanita`, Viale Regina Elena 299, 00161 Rome, Italy c Department of Pharmacology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA b

Received 2 September 2001; accepted 13 November 2001

Abstract Macrophage cell lines exhibit different responses to IFN-g depending on their maturation stage. We investigated the mechanisms underlying the differential IFN-g responsiveness in the less mature P388.D1 and in mature RAW264.7 cells. A reduction in the binding activity of the signal transducer and activator of transcription-1 (STAT-1) to different STAT binding elements (SBEs) was observed in P388.D1. This reduced binding activity was not due to an impaired STAT-1 activation. Studies on the expression of a negative regulator of cytokine signalling, protein-inhibiting activated STAT-1 (PIAS-1), showed that this protein was expressed constitutively at high levels in P388.D1. Forced expression of a PIAS-1 homologue, the Gu binding protein (GBP), inhibited the STAT-1-mediated gene activation in RAW264.7 cells, whereas a construct expressing the 50 portion of GBP in the antisense orientation reverts the IFN-g-resistant phenotype of P388.D1. Thus, our results indicate that PIAS-1 may account for the differential IFN-g responsiveness in macrophage cell lines at different stages of maturation. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Negative regulator of cytokine signalling; Macrophage differentiation; IFN-g signalling; Transcription factors; PIAS-1

1. Introduction Innate immunity provides an immediate response against infections regulated at a variety of levels by cells of the monocyte/macrophage lineage. The antimicrobial mechanisms recruited by innate immune response usually are triggered by cytokines and microbial products that can control the dynamic function of macrophages. In this respect, IFN-g has long been recognized as a regulatory cytokine, converting macrophages from resting to activated

Abbreviations: EMSA, electrophoretic mobility shift assay; GBP, Gu binding protein; iNOS, inducible nitric oxide synthase; IRF, interferon regulatory factor; Jak, Janus kinases; JAB, Jak binding protein; PIAS-1, protein-inhibiting activated STAT-1; RPA, RNAse protection; SBE, STAT binding element; SOCS, suppressor of cytokine signalling; STAT, signal transducers and activators of transcription * Corresponding author. Tel.: +39-6-49902897; fax: +39-6-49387115. E-mail address: [email protected] (E.M. Coccia).

state [1]. This activation is acquired during macrophage differentiation and correlates with an increase of IFN-g transcriptional response [2 –4]. An important mechanism for the transcriptional activation of IFN-g-inducible genes acts through the binding of signal transducers and activators of transcription (STAT-1) to the promoter regulatory sequence termed STAT binding element (SBE) [5 – 8]. Binding of IFN-g to its receptor triggers the phosphorylation of STAT-1 on specific tyrosine residues by the Janus kinases Jak-1 and Jak-2 [9]. The Jak/STAT pathway is subjected to relatively rapid down-regulation through different negative feedback mechanisms: receptor-mediated endocytosis, expression of dominant negative isoforms of STAT proteins, activation of tyrosine phosphatases and expression of suppressor of cytokine signalling (SOCS) family members that bind the cytokine receptors or the Jak kinases [10,11]. In addition, a new family of negative regulators of signal transducers, the protein-inhibiting activated STAT (PIAS) proteins, has been

0898-6568/01/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved. PII: S 0 8 9 8 - 6 5 6 8 ( 0 1 ) 0 0 2 7 2 - 8

538

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

identified recently. PIAS seems to bind directly to STATs and to inhibit DNA binding: PIAS-1, originally identified as Gu/RH-11 binding protein [12], was characterized as a specific partner of tyrosine-phosphorylated STAT-1 dimers [13,14], whereas PIAS-3 specifically associates with STAT3 but not STAT-1 [15]. An interesting feature of PIAS proteins is that their association with STATs is cytokine dependent since PIAS proteins have higher binding affinity toward tyrosine-phosphorylated STATs [13]. A variety of mechanisms can modulate a cell’s response to cytokine and, in turn, control many biological functions. For instance, the differences in responsiveness of tumor cell lines to IFN remain unclear and events downstream from the IFN receptor, or deregulation of the negative feedback of the Jak/STAT pathway, could be involved in IFN resistance. Recently, it has been observed that dysregulated overexpression of Jak binding protein (JAB) can confer IFN resistance to M1, NIH 3T3 and DAUDI cells [16]. Other studies have reported defects in the levels of STAT-1 as being responsible for the different sensitivities to IFN in human melanoma cell lines [17,18]. Moreover, previous work has shown that murine macrophage cells at different stages of maturation [19,20] have distinct responses to IFN-g as measured by the analysis of STAT-1 phosphorylation status and of the transcriptional response of SBEcontaining genes to IFN-g [3,4,21 –23]. In the present report, we investigated the mechanism of differential responsiveness to IFN-g in the less mature macrophage P388.D1 cells and in the mature cell line RAW264.7 by studying the expression and activation of components involved in the IFN-g signal transduction pathway. Differences in STAT-1 binding activity in response to IFN-g treatment between these cell lines were observed on SBE sequences from genes encoding markers of macrophage activation/maturation, such as FcgRI and inducible nitric oxide synthase (iNOS), and from the promoter of the transcription factor, interferon regulatory factor-1 (IRF-1) [5,6,24,25]. We therefore hypothesized that an inhibitory mechanism might be responsible for the differential responsiveness to IFN-g. Differences were identified in the expression of PIAS-1: in the less mature P388.D1 cells, high levels of PIAS-1 were expressed constitutively compared to the low steady-state level present in RAW264.7 cells. Thus, the goal of the present study was to characterize the functional role of PIAS-1 as modulators of IFN-g signalling events in macrophage cell lines at different stages of maturation.

2. Materials and methods 2.1. Cell culture and reagents The macrophage-like cell lines RAW264.7 and P388.D1 were obtained from American Type Culture Collection (Rockville, MD). The cells were grown in RPMI-1640 supplemented with 10% FCS endotoxin free, penicillin

(100 IU/ml) and streptomycin (100 mg/ml). Geneticin G-418 sulfate (GIBCO-BRL, Grand Island, NY) was used at 100 mg/ml to select the stable transfected cells. Recombinant IFN-g (107 U/mg of protein) produced by Genentech (San Francisco, CA) was used at 100 U/ml. Murine IFN-b prepared and purified as previously reported [8] to 107 U/mg of protein was used at 200 U/ml. 2.2. Preparation of cell extracts Nuclear cell extracts were prepared as described by Schreiber et al. [26]. Briefly, five millions of cells were washed twice in cold PBS and then collected by centrifugation. The pellet was resuspended in 400 ml of 10-mM HEPES (N-2-hydroxyethylpiperazine-N0-2-ethanesulfonic acid, pH 7.9), 10-mM KCl, 0.1-mM EDTA, 0.1-mM EGTA, 1-mM DTT and 0.5-mM PMSF. After incubation for 15 min on ice, 25 ml of a 10% solution of NP-40 was added and the tube was vortexed for 10 s. After centrifugation of 30 s in an Eppendorf microfuge at full speed, the nuclear pellet was resuspended in 50 ml of 20-mM HEPES, pH 7.9, 0.4-M NaCl, 1-mM EDTA, 1-mM EGTA, 1-mM DTT and 1-mM PMSF and the tube was vigorously rocked at 4 C for 15 min on a shaking platform. The suspension was centrifuged for 5 min in an Eppendorf microfuge at full speed and the supernatants were stored at  80 C. Whole cell extracts were prepared as previously described [8]. Five millions of cells were lysed in 50 ml of 20-mM HEPES, pH 7.9, 50-mM NaCl, 10-mM EDTA, 2-mM EGTA, 0.5% NP-40, 1-mM DTT, 10-mM sodium molybdate, 10-mM sodium orthovanadate, 100-mM NaF, 0.5-mM PMSF and 10-mg/ml leupeptine. The suspension was centrifuged at 10,000 rpm for 10 min in an Eppendorf microfuge and the supernatants were stored at  80 C. 2.3. DNA electrophoretic mobility shift assay (EMSA) To measure the association of DNA binding proteins with different DNA sequences, synthetic double-stranded oligodeoxynucleotides were end labelled with [g-32P]ATP by T4 polynucleotide kinase (Boehringer-Mannheim, Mannheim, Germany). Binding reaction mixture (20 ml final volume) contained labelled oligodeoxynucleotide probes (20,000 cpm) in binding buffer [20-mM HEPES, pH 7.9, 1-mM EDTA, 1-mM DTT and 4-mg poly(dI)-poly(dC)]. Nuclear cell lysates (10 mg) were added and the reaction mixture was incubated for 30 min at room temperature. Glycerol was added to 13% (v/v) and samples were electrophoresed in 6% polyacrylamide gel in 0.5  TBE buffer for 1.30 h at 200 V at 18 C. For supershift analysis, cell extracts were incubated with 1 mg of anti-STAT-1 antibody (Transduction Lab., Lexington, KY) for 20 min at 4 C before the addition of binding buffer containing labelled oligodeoxynucleotide. The oligodeoxynucleotide probes used were SBE IRF-1 (50-GATCGATTTCCCCGAAATGA-30) [6], SBE FcgRI (50-TTCCTTTTCTGGGAAATA-

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

CATCTC-30) [24] and SBE iNOS (50-TGTTTGTTCCTTTTCCCCTAACACTG-30) [25]. 2.4. Flow cytometric analysis of IFN-g receptor Cells were incubated for 30 min at 4 C in the presence of a pretitrated dilution of rat GR20 mAb anti-IFN-gR achain [27]. Cells were washed twice with cold PBS solution then incubated for 30 min at 4 C with secondary FITCconjugated anti-rat antibody (Pharmingen, San Diego, CA). For isotype-matched negative controls, the cells were incubated with an irrelevant isotype rat mAb. Cells were washed twice with cold PBS solution, fixed in 1% paraformaldehyde and then analysed for fluorescence emission using a flow cytometer (FACScan, Becton-Dickinson, Mountain View, CA). Fluorescence data were analysed by the Lysis II software program (Becton-Dickinson). 2.5. Immunoprecipitation of Jak-1 and Jak-2 Ten millions of cells were washed twice in cold PBS and the pellet was then resuspended in 500 ml of lysis buffer containing 25-mM Tris –HCl, pH 7.4, 75-mM NaCl, 0.125% sodium deoxycholate, 0.5-mM EGTA, 0.5-mM EDTA, 0.5% Triton X-100, 0.25% NP-40, 1-mM PMSF, 2-mg/ml aprotinin, 1-mg/ml leupeptin, 1-mg/ml pepstatin, 1-mM sodium orthovanadate and 20-mM NaF. The lysate was incubated for 30 min on ice and insoluble debris was removed by centrifugation (12,000 rpm, 4 C, 3 min). The lysate was precleared for 1 h with a specific rabbit preimmune serum (2 ml) and then with 40 ml protein A-Sepharose (10 mg/ml) on a shaker at 4 C. Jak-1 and Jak-2 were immunoprecipitated with rabbit polyclonal anti-Jak-1 and anti-Jak-2 antibodies (Upstate Biotechnology, New York, NY) for 1 h at 4 C with agitation. The immunocomplex was collected by protein A-Sepharose and resolved by SDS-PAGE. Following transfer to nitrocellulose, blots were blocked with 3% BSA in ICN (10-mM Tris – HCl, pH 7.4, 100-mM NaCl, 1-mM EDTA, 0.1% Tween 20 and 1-mM sodium orthovanadate) and probed with anti-phosphotyrosine antibody 4G10 (Upstate Biotechnology, dilution 1:2000 in 1% BSA) and immunoreactive bands were visualized using the ECL Western blotting system (Amersham, Buckinghamshire, UK). To normalize equal protein loading, blots were stripped and reprobed with anti-Jak-1 (dilution 1:500) and anti-Jak-2 (dilution 1:1000) antibody. 2.6. Western blot assay Whole cell extracts (50 mg) were separated by 7.5% SDSPAGE gel and blotted onto nitrocellulose membranes. Blots were incubated with mouse monoclonal STAT-1 antibody (Transduction Lab.) and reacted with anti-mouse horseradish peroxidase-coupled secondary antibody (Amersham, Little Chalfont, UK) using an ECL system. Rabbit antiserum specific for tyrosine-phosphorylated STAT-1 was

539

purchased from UBI (Upstate Biotechnology) and used according to the manufacturer’s instructions. The protein level of PIAS-1 was evaluated using a rabbit polyclonal anti-PIAS-1 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). 2.7. Electroporation and enzymatic assays Ten millions of cells were washed twice with cold PBS and resuspended in 1 ml of cold 10% FCS-RPMI medium containing 20-mM HEPES, pH 7.4, and 20 mg of plasmid DNA. The suspension was allowed to sit on ice for 10 min with occasional mixing. About 250 ml of this cell-DNA suspension was then introduced into the electroporation chamber (4-mm space) and the following parameters were used for transfection: field strength 0.625 kV/cm, capacitance 960 mF and time constant  65 ms (Bio-Rad gene pulser transfection apparatus, Hercules, CA). After 10 min on ice, the electroporated cell suspensions were divided and treated as indicated. All the DNAs used were prepared according to ENDO Free Qiagen product (Valencia, CA) in order to obtain DNA preparation with minimal endotoxin content. The wt-SBE construct contains an oligodeoxynucleotide that includes the SBE sequence between  138 and  107 of the IRF-1 gene. About 15 mg of wt-SBE construct or 5 mg pFLAGCMV2-Gu binding protein (GBP) expression vector [12] were used in cotransfection experiments. Transfection efficiency was normalized according to the activity of the b-galactosidase expressed from 1 mg of cotransfected RSV-b-galactosidase plasmid [28]. Reagents from Promega were used to assay extracts for luciferase activity in a LUMAT LB9501 luminometer (EG&G Berthold Bad Wild Bad, D). Data are reported as fold of induction, which was calculated by dividing the relative light units of each stimulated culture with those of the corresponding unstimulated control culture. For stable transfection experiments, P388.D1 cells were electroporated as described using a construct (RcCMV-PBG) containing the 50 end of the human GBP cDNA (1 – 850) cloned in the antisense orientation into the HindIII-XbaI sites of the RcCMV eukaryotic expression vector (Invitrogen, San Diego, CA). As control, the cells were transfected with RcCMV empty vector. Cells were selected after 48 h with Geneticin G-418 sulfate (GIBCO-BRL) for 2 weeks. 2.8. RNA isolation and analysis Total RNA from RAW264.7 and P388.D1 cells was purified as previously described [8]. About 100 ng of DNase-digested RNA was reverse transcribed using Moloney murine leukemia virus reverse transcriptase (MLV RT, GIBCO-BRL, Gaithersburg, MD) and sequence specific oligodeoxynucleotides. PCR was performed using 0.5 mM of oligodeoxynucleotide primers (each), 200-mM dNTPs, 2.5-U Taq polymerase (Roche, Indianapolis, IN), 10-mM Tris – HCl, pH 8.3, 1.5-mM MgCl2 and 50-mM KCl in a

540

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

final reaction volume of 50 ml. cDNAs were normalized with the b-actin gene expression using the following primers: sense strand b-actin 50-GTGGGCCGCTCTAGGCAC CAA30and antisense strand b-actin 50-CTCTTTGATGTC ACGCACGATTTC-30. Amplification cycles (20, 25 and 30) were performed (94 C denaturation, 45 s; 60 C annealing, 45 s; 72C extension, 1 min). Amplification of 10-ng RT-RNA within the linear range was achieved by 20 cycles. The primers used were sense strand PIAS-1 5 0-CTTAGAGTTTCTGAACTCCAAGTACTGTTGGGC TAC-3 0 and antisense strand PIAS-1 5 0-GACCTGCACTGTGAAGTCACATTTGGTCCCAGAAATATCCATGGAGGCTGCT-30, which amplify a region of 557 bp. To exclude the presence of contaminant DNA, negative controls lacking template or the reverse transcriptase were also included. Following amplification, an aliquot of PCR reaction was electrophoresed through a 2% agarose gel and analysed by Southern blot using specific oligodeoxynucleotide probes for both PIAS-1 and b-actin. The sequences of the probes used w e r e P I A S - 1 p r o b e 5 0- TA G T G G G G AT G AT G C AGGGTGGCCATCATAAGTGAGCTGTGGAATGGT-30 and b-actin probe 50-CCCAGATCATGTTTGAGACCTTCAACACCC-30. Signals were analysed quantitatively by electronic autoradiography using an Instant Imager Instrument (Canberra Packard, Meridien, CT). To analyse the expression of PBG construct in the stable transduced cells, RNAse protection (RPA) was performed on 5 mg of total RNA hybridized for 18 h to the RNA probes (3  105 cpm) as previously described [29]. To generate the 32P-labelled GBP probe, the human GBP cDNA was cloned in pBSKS and then the obtained BS-GBP construct was linearized with DdeI and transcribed with T7 RNA polymerase. A 18S RNA probe was used as a control for equal RNA loading.

3. Results 3.1. STAT-1 binding activity in P388.D1 and RAW264.7 macrophage cell lines at different stages of maturation To investigate the mechanisms underlying the differential response to IFN-g, the activation of STAT-1 was studied in P388.D1 and RAW264.7 macrophage cell lines at different stages of maturation according to the expression of cell surface markers [19,20]. EMSA experiments were performed by using oligodeoxynucleotides specific for the SBE element present within different IFN-g-inducible gene promoters, which are known to mediate the transcriptional induction by IFN-g through STAT-1. Nuclear extracts were obtained from P388.D1 and RAW264.7 cells treated for 30 min with IFN-g. As shown in Fig. 1, STAT-1 binding activities to oligodeoxynucleotides containing SBE sequences present in FcgRI, IRF-1 and iNOS genes were clearly reduced in the nuclear extracts from P388.D1 cells as

Fig. 1. STAT-1 binding activity in the P388.D1 and RAW264.7 cell lines. Nuclear cell extracts from P388.D1 and RAW264.7 cell lines treated with medium alone or with IFN-g (100 U/ml) for 30 min were incubated with the indicated SBE oligodeoxynucleotides and analysed by EMSA as described under Materials and methods. Where indicated, STAT-1 antibody was added to the reaction to supershift the IFN-g-induced complex.

compared to RAW264.7 cells after cytokine treatment, implying an impaired responsiveness of P388.D1 cells to IFN-g. The specificity of the IFN-g-induced complex was analysed by supershift experiment using specific antibodies to STAT-1. 3.2. Analysis of IFN-g signal transduction in P388.D1 cells The current model of IFN-g-mediated transcriptional induction involves activation of STAT-1 by Jak-mediated phosphorylation following ligand –receptor interaction. To shed light on the impaired STAT-1 activation observed in P388.D1 versus RAW264.7, we investigated the surface expression of the IFN-g receptor a-chain by flow cytometric analysis. Fig. 2A shows no differences between the two cell lines in the level of IFN-g receptor a-chain expression. We therefore analysed the activation of Jak proteins in response to IFN-g in both cell lines. Proteins from whole cell extracts were immunoprecipitated with specific anti-Jak antibodies and analysed by immunoblotting with the 4G10 anti-phosphotyrosine monoclonal antibody and membranes were sequentially reprobed with the corresponding anti-Jak antibodies (Fig. 2B). Within 15 min of treatment with IFN-g, tyrosine phosphorylations of Jak-1 and Jak-2 were induced both in RAW264.7 and P388.D1. Similarly, STAT-1 tyrosine phosphorylation was investigated following IFN-g treatment of the two cell lines. Western blot analysis was performed in total cell extracts from cells untreated or treated with IFN-g for 15 min. The blots were probed with antiserum recognizing specifically STAT-1 phosphorylated at Tyr-701 residue. As shown in Fig. 2B, Tyr-701 phosphorylation of STAT-1 induced by IFN-g was comparable to RAW264.7 and P388.D1 cells, indicating that other mechanisms are responsible for the impaired STAT-1 binding to SBE sequences in less mature macrophage cells.

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

541

Fig. 2. Expression of the components of IFN-g signal transduction pathway. (A) Flow cytometric analysis of IFN-g receptor on RAW264.7 and P388.D1 cell membrane. Unstimulated cells were incubated with either rat GR20 mAb anti-IFN-gR a-chain (filled line) or an isotype control mAb (dotted line). Fluorescence emission was analysed using a flow cytometer as described under Materials and methods. A representative of three independent experiments is shown in each panel. (B) Analysis of Jak and STAT-1 tyrosine phosphorylation following IFN-g treatment. RAW264.7 and P388.D1 cells were treated for 15 min with IFN-g (100 U/ml). Jak proteins were immunoprecipitated from whole cell extracts with the specific antibody and analysed by immunoblotting with the 4G10 antiphosphotyrosine monoclonal antibody. The same cell extracts were also probed with anti-phospho-STAT-1 (STAT-1 pY701) antibody. Filters were stripped and reprobed with anti-Jak or STAT-1 antibodies (lower panels).

3.3. Analysis of PIAS-1 expression in P388.D1 and RAW264.7 cells Deregulation of the negative feedback of the Jak/STAT pathway has been recently proposed to contribute to the cellular resistance to IFNs [16]. Among the proteins that have been described to down-regulate the cytokine signal transduction, we focused our attention on PIAS-1 since it is a specific inhibitor of the STAT-1-mediated gene activation interacting with the Tyr-701 of STAT-1 [13]. To investigate the potential involvement of PIAS-1 expression in the IFN-resistant phenotype, we compared the PIAS-1 expression in P388.D1 versus RAW264.7 cells. We initially performed immunoblot analysis using whole cell extracts from P388.D1 and RAW264.7 cells. A faint band of 78 kDa was detected in P388.D1 cells independently of IFN-g treatment (Fig. 3A). Conversely, in RAW264.7 cells, only IFN-g was able to induce the PIAS-1 protein, indicating that a different pattern of PIAS-1 could be present in the two cell lines. To confirm these results, PIAS-1 expression was investigated also at the mRNA level. Total RNA was prepared from P388.D1 and RAW264.7 and analysed for the expres-

sion of PIAS-1 gene by RT-PCR (Fig. 3B). Quantitation of mRNA levels was performed using as internal control b-actin gene expression (Fig. 3C). RT-PCR analysis revealed that the basal level of PIAS-1 mRNA was 2-fold higher in P388.D1 versus RAW264.7 cell line. Moreover, a different pattern of PIAS-1 gene induction after IFN-g treatment was observed in the two cell lines. Indeed, PIAS-1 was induced only in the mature RAW264.7 cells in response to IFN-g treatment. These results suggest that the constitutive overexpression of PIAS-1 could be a mechanism responsible for the IFN-g resistance in P388.D1 cells. Conversely, in RAW264.7 cells, the IFN-g-induced PIAS-1 expression may account for the desensitization to IFN-g signalling. 3.4. PIAS-1 expression impairs STAT-1-dependent IFN signalling STAT-1 binding activity was evaluated using the SBE element of IRF-1 promoter in RAW264.7 at different time of IFN-g treatment. As shown in Fig. 4A, a clear reduction of STAT-1 complex was observed after 24 h of IFN-g stimulation compared to that present in the cells treated for 30 min with IFN-g, indicating that when the expression

542

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

3.5. Modulation of IFN-g response by ectopic expression of PIAS-1 and PIAS-1 antisense

Fig. 3. Expression of PIAS-1. RAW264.7 and P388.D1 cells were cultured in medium alone or with IFN-g (100 U/ml) for 24 h. (A) Whole cell extracts (50g) were analysed by immunoblotting with anti-PIAS-1 antibody. A representative of three experiments performed is shown. (B) Total RNA was purified and analysed by RT-PCR as described under Materials and methods. b-actin mRNA levels were used as internal control. (C) The results were quantified by scanning autoradiographs from three separate experiments. The values, obtained after normalization with respect to the b-actin signals and expressed in arbitrary units, are shown in the graph.

of PIAS-1 is high, STAT-1-dependent signalling pathway is impaired. These effects were independent from the phosphorylation state of STAT-1, which was similar in all conditions examined by Western blot using antiserum that recognizes specifically STAT-1 phosphorylated at Tyr-701 residue (Fig. 4B). Accordingly, the signal transduction pathway of IFN-b that also requires STAT-1 binding activity is also impaired in P388.D1 cells in spite of the Tyr-701 phosphorylation of STAT-1. Therefore, the negative regulatory mechanism of the IFN transduction pathway observed in this study does not seem to involve the endocytosis of Jak/receptor complexes or the recruitment of tyrosine phosphatase to the receptor complexes but rather can be explained by the high expression of PIAS-1. These results are in accord with the previously described mechanism of action of PIAS-1 that inhibits STAT-1 binding activity but not Tyr-701 phosphorylation [13].

To examine in vivo the effects of PIAS-1 on the IFN-ginduced gene transcription, cotransfection experiments were performed in RAW264.7 cells using PIAS-1/GBP expression plasmid with a luciferase reporter construct (wtSBE) containing a copy of STAT-1 binding site from the IRF-1 gene promoter (Fig. 5A). GBP is a homologue of PIAS-1 [12,13]. The transfected cells were treated with IFN-g for 16 h and the luciferase activity was measured (Fig. 5B). IFN-g caused 6.5-fold induction of luciferase activity, whereas in the presence of PIAS-1/GBP the luciferase expression in response to IFN-g was strongly reduced, demonstrating that PIAS-1/GBP can inhibit the STAT-1-mediated gene activation. Since the unresponsiveness of the P388.D1 cells to IFN-g seems to be mediated by an increased expression of PIAS-1, we asked whether the neutralization of PIAS-1 activity through the expression of PIAS-1 antisense could overcome the inhibitory effect of PIAS-1 and restore the gene inducibility after IFN-g treatment. P388.D1 cells were transfected with a construct expressing the PIAS-1 antisense (PBG). In order to obtain stable transfectants, the transduced cells were selected for neomycin resistance and then analysed for their responsiveness to IFN-g. Mixed populations were chosen to avoid clonal variability. After selection, expression of transgene was assessed by RPA (insert in Fig. 5C). No expression of PIAS-1 antisense (PBG) was detectable in cells transfected with the RcCMV empty vector, whereas a clear expression of PBG was present in cells transfected with RcCMV-PBG. The results of lucifer-

Fig. 4. Inhibition of IFNs signalling by PIAS-1. Total cell extracts were prepared from RAW264.7 and P388.D1 cells treated at the indicated times with 100 U/ml of IFN-g or 200 U/ml of IFN-b. (A) EMSA was performed using a radiolabelled oligodeoxynucleotide corresponding to the IRF-1 SBE regulatory sequence. (B) Cell extracts were also analysed by Western blot using anti-phospho-(STAT-1 pY701) or anti-STAT-1 (STAT-1) antibodies.

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

543

Fig. 5. Effects of PIAS-1 on the transcriptional activity of IRF-1 gene promoter. (A) Schematic structure of the wt-SBE construct used in transient transfection experiments. (B) Where indicated, the wt-SBE construct (15 mg) was cotransfected with 5 mg of GBP/PIAS-1 expression vector in RAW264.7. RSV-bgalactosidase plasmid (1 mg) was always cotransfected to normalize transfection efficiency. Cells were treated with control medium or with medium containing IFN-g (100 U/ml) for 16 h and luciferase activity was evaluated as described under Materials and methods. Each column and bar represent the mean ± S.E. relative luciferase activities in three independent experiments. (C) The P388.D1 cells stable transduced with RcCMV-PBG construct (GBP antisense) or with the empty vector (RcCMV) were used to perform transient transfection experiments with the wt-SBE construct (15 mg). Cells were treated with IFN-g as indicated, and the luciferase activity was evaluated as described under Materials and methods. The insert shows a RPA to detect PBG overexpression in P388.D1-transduced cells. 18S was used as internal control to establish the relative amount of RNA loaded.

ase activity shown in Fig. 5C indicate that the inhibition of PIAS-1 expression through the antisense restored the IFN-g signalling and the STAT-1-mediated gene activation in the less mature P388.D1 cells.

4. Discussion Binding of cytokine to its receptor triggers cytoplasmic signal transduction pathways that initiate changes in transcription. The duration and intensity of a cell’s response to cytokines appear to be determined by the

net effect of positive and negative regulatory mechanisms. This balance is however lost in certain chronic inflammatory and proliferative diseases and cancers, with the potential to induce abnormal production of cytokines in the organism. Specific evidence of the importance of the negative feedback mechanisms of the Jak/STAT system have been obtained from studies in cell lines and in knockout mice [10,11]. The biological role of SOCS proteins on IFN-mediated Jak/STAT signalling pathways has been elucidated in human cancer cell lines expressing SOCS proteins [30]. These studies demonstrated that SOCS-1 and SOCS-3 display strong inhibitory effects

544

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

towards the phosphorylation of STAT-1 and in turn towards the IFN-mediated antiproliferative and antiviral activities. Negative regulation of cytokine signalling downstream of STAT activation can also occur through other families of modulators. An important negative regulatory mechanism of Jak/STAT pathway involves the recruitment of tyrosine phosphatases containing tandem SH2 domains to receptor complexes [10]. In addition, a recently described class of proteins, PIAS, can bind and neutralize activated STATs [10]. In particular, PIAS-1 inhibits IFN signalling binding to STAT-1. Conversely, PIAS-3 specifically associates with STAT-3 and switches off the IL-6, OMS, CNTF and PRL-induced gene transcription [10,31]. All together, these observations support the hypothesis that the regulatory mechanisms mediated by SOCS and PIAS activity might play a more general role in regulating cellular responsiveness to cytokines besides their ability to switch off cytokine signalling. The aim of this study was to elucidate the mechanism of differential IFN-g sensitivity in macrophage-like cells at different stages of maturation. We showed that in less mature P388.D1 macrophages treated with IFN-g, the binding affinity of STAT-1 to SBE consensus sequences within the IRF-1, iNOS and FcgRI promoters was reduced compared to the STAT-1 activation present in the mature RAW264.7 cells (Fig. 1). Similar results were obtained in another immature macrophage cell line, WEHI-3, which exhibits a specific impaired response to IFN-g stimulation [4]. In the present report, we demonstrated that the reduced responsiveness of P388.D1 cells towards treatment with IFN-g is neither due to a lack of IFN-g receptor expression nor to a defect in IFN-g signal transduction via Jakmediated protein phosphorylation (Fig. 2) but rather can be explained by the overexpression of the negative regulator PIAS-1 that impairs STAT-1 activity and therefore IFN-g response (Fig. 3). PIAS-1 has been initially identified as human RNA helicase II/GBP [12] that shares sequence identity with androgen receptor interacting protein [32]. Recently, PIAS1 has been also recognized as a specific interaction partner for STAT-1 by yeast two-hybrid screening [13]. The phosphorylation on Tyr-701 and the dimerization of STAT-1 are required for its association with PIAS-1, indicating that this interaction is cytokine dependent [13,14]. Analysis of PIAS-1 expression showed a higher steady-state level in P388.D1 cells as compared to the IFN-g-sensitive RAW264.7 cells (Fig. 3), confirming that the dominant resistant phenotype to IFN-g could be in part attributed to PIAS-1 overexpression. Moreover, high levels of PIAS-1 can also inhibit the IFN-b signalling as shown in Fig. 4. Interestingly, we also observed an induction of PIAS-1 mRNA in RAW264.7 treated with IFN-g after 24 h, which correlates with a later downmodulation of IFN-g signalling. The ectopic expression of PIAS-1/GBP was indeed able to inhibit the IFN-g induction of gene transcription in RAW264.7 cells (Fig. 5B). These results

are in line with the data published by Liu et al. [13] showing that the overexpression of PIAS-1 in human 293 fibroblast cells inhibits STAT-1-mediated gene activation. Moreover, when the inhibitory effect of PIAS-1 was neutralized through the expression of the antisense RNA for PIAS-1 (PBG), the IFN-g response was restored in P388.D1 cells (Fig. 5C). All together, these studies suggest that the modulation of STAT-1 activity by PIAS-1 expression may regulate the IFN-mediated biological activities during the differentiation/maturation of macrophages. In line with this hypothesis, we have recently shown an increased expression and activity of STAT-1 in human monocytes maturing to macrophages [33] that accounts for the achievement of a full expression of a particular set of genes primarily involved in the cytostatic and/or cytotoxic activity of mature macrophages. IFN-g-mediated activation of STAT-1 induces also the transcription of genes coding for new transcription factors, such as IRF-1, which in turn stimulate the expression of other IFN-stimulated genes. For instance, the expression of iNOS is regulated at the transcriptional level by the activity of both STAT-1 and IRF-1 [25,27,34]. Therefore, only in mature macrophages iNOS promoter shows maximal transcriptional activity after stimulation with inflammatory cytokines. In fact, the full activation of STAT-1 drives both iNOS and IRF-1 gene transcription. The limited activation of these critical transcription factors by PIAS-1 may prevent significantly the IFN-g-activated gene transcription in immature macrophages. Conversely, in mature macrophages, the differentiation process is accompanied by a decreased basal expression of PIAS-1 and a concomitant increase in the transcriptional responses to IFNs that lead to immunological activation. In addition, longer IFN-g treatment in mature macrophages induces the expression of PIAS-1 that in turn switches off the IFN-g signalling. Therefore, PIAS-1 is one of the major elements that controls the balance of cytokine signals that utilizes STAT-1 in their pathways. However, we can not exclude that alternative regulatory mechanisms can affect the IFNg responsiveness in other immature macrophage cell lines. For instance, in WEHI-3, IFN-g stimulation does not result in STAT-1 activation despite the presence of intact components of IFN-g signal transduction pathway [4] suggesting that members of SOCS family might play a role in regulating IFN-g response in these cells. In conclusion, it appears that reduction in the expression of PIAS-1 is part of a developmental program that leads to terminal macrophage differentiation characterized by a stronger response to IFN-g.

Acknowledgments This work was supported by grants from the ‘‘Special Project AIDS’’ and ‘‘1% Project’’ of the Istituto Superiore di Sanita`. The authors want to thank Dr. R. Pine (Cornell

E.M. Coccia et al. / Cellular Signalling 14 (2002) 537–545

University, New York) for critical reading of the manuscripts and for the generous gift of wt-SBE plasmid. Moreover, we thank Dr. S. Hemmi (University of Zurich) for GR20 mAb anti-IFN-gR a-chain. We are grateful to Romina Tomasetto and Sabrina Tocchio for editorial assistance and Eugenio Morassi for preparing drawings.

References [1] Nathan CF, Prendergast TJ, Wiebe ME, Stanley ER, Platzer E, Remold HG, Welte K, Rubin BY, Murray HW. J Exp Med 1984;160:600 – 5. [2] Eilers A, Georgellis D, Klose B, Schindler C, Ziemiecki A, Harpur AG, Wilks AF, Decker T. Mol Cell Biol 1995;15:3579 – 86. [3] McDowell MA, Lucas DM, Nicolet CM, Paulnock DM. J Immunol 1995;155:4933 – 8. [4] Lucas DM, Lokuta MA, McDowell MA, Doan JE, Paulnock DM. J Immunol 1998;160:4337 – 42. [5] Decker T, Kovarik P, Meinke A. J Interferon Cytokine Res 1997;17: 121 – 34. [6] Pine R, Canova A, Schindler C. EMBO J 1994;13:158 – 67. [7] Boehm U, Klamp T, Groot M, Howard JC. Annu Rev Immunol 1997; 15:749 – 95. [8] Coccia EM, Marziali G, Stellacci E, Perrotti E, Ilari R, Orsatti R, Battistini A. Virology 1995;211:113 – 22. [9] Schindler C, Darnell JE. Annu Rev Biochem 1995;64:621 – 51. [10] Starr R, Hilton D. J Bioassays 1999;21:47 – 52. [11] Naka T, Fujimoto M, Kishimoto T. Trends Biochem Sci 1999;24: 394 – 8. [12] Valdez BC, Henning D, Perlaky L, Busch RK, Busch H. Biochem Biophys Res Commun 1997;234:335 – 40. [13] Liu B, Liao J, Rao X, Kushner SA, Chung CD, Chang DD, Shuai K. Proc Natl Acad Sci USA 1998;95:10626 – 31. [14] Liao J, Fu Y, Shuai K. Proc Natl Acad Sci USA 2000;97:5267 – 72.

545

[15] Chung CD, Liao J, Liu B, Rao X, Jay P, Berta P, Shuai K. Science 1997;278:1803 – 5. [16] Sakamoto H, Yasukawa H, Masuhara M, Tanimura S, Sasaki A, Yuge K, Ohtsubo M, Ohtsuka A, Fujita T, Ohta T, Furukawa Y, Iwase S, Yamada H, Yoshimura A. Blood 1998;92:1668 – 76. [17] Sun WH, Pabon C, Alsayed Y, Huang PP, Jandeska S, Uddin S, Platanias LC, Rosen ST. Blood 1998;91:570 – 6. [18] Wong LH, Krauer KG, Hatzinisiriou I, Estcourt MJ, Hersey P, Tam ND, Edmondson S, Devenish RJ, Ralph SJ. J Biol Chem 1997;272: 28779 – 85. [19] Nibbering PH, Furth RV. Immunobiology 1988;176:432 – 9. [20] Leenen PJM, Jansen AMAC, van Ewijk W. Differentiation 1986;32: 157 – 64. [21] Nicolet CM, Paulnock DM. J Immunol 1994;152:153 – 62. [22] Eilers A, Seegert D, Schindler C, Baccarini M, Decker T. Mol Cell Biol 1993;13:3245 – 54. [23] Finn PW, Kara CJ, Douhan J, Van TT, Folsom V, Glimcher LH. Proc Natl Acad Sci USA 1990;87:914 – 8. [24] Perez C, Wietzerbin J, Benech PD. Mol Cell Biol 1993;13:2182 – 92. [25] Gao J, Morrison DC, Parmely TJ, Russell SW, Murphy WJ. J Biol Chem 1997;272:1226 – 30. [26] Schreiber E, Matthias P, Mu¨ller MM, Schaffner W. Nucleic Acids Res 1989;17:6419. [27] Basu M, Pace JL, Pinson DM, Russell SW. J Interferon Res 1989; 9:551 – 62. [28] Bonnerot C, Rocancourt D, Briand P, Grimber G, Nicolas JF. Proc Natl Acad Sci USA 1987;84:6795 – 9. [29] Coccia EM, Stellacci E, Marziali G, Weiss G, Battistini A. Int Immunol 2000;12:977 – 85. [30] Song MM, Shuai K. J Biol Chem 1998;273:35056 – 62. [31] Russell DL, Richards JS. Mol Endocrinol 1999;13:2049 – 64. [32] Tan J, Hall SH, Hamil KG, Grossman G, Petrusz P, Liao J, Shuai K, French FS. Mol Endocrinol 2000;14:14 – 26. [33] Coccia EM, Del Russo N, Stellacci E, Testa U, Marziali G, Battistini A. Int Immunol 1999;11:1075 – 83. [34] Martin E, Nathan C, Xie Q-W. J Exp Med 1994;180:977 – 84.