TNF-α Induces Tyrosine Phosphorylation and Recruitment of the Src Homology Protein-Tyrosine Phosphatase 2 to the gp130 Signal-Transducing Subunit of the IL-6 This information is current as Receptor Complex of October 1, 2021. Johannes G. Bode, Jens Schweigart, Jan Kehrmann, Christian Ehlting, Fred Schaper, Peter C. Heinrich and Dieter Häussinger

J Immunol 2003; 171:257-266; ; Downloaded from doi: 10.4049/jimmunol.171.1.257 http://www.jimmunol.org/content/171/1/257

References This article cites 54 articles, 32 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/171/1/257.full#ref-list-1

Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists by guest on October 1, 2021

• Fast Publication! 4 weeks from acceptance to publication

*average

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

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

TNF-␣ Induces Tyrosine Phosphorylation and Recruitment of the Src Homology Protein-Tyrosine Phosphatase 2 to the gp130 Signal-Transducing Subunit of the IL-6 Receptor Complex1

Johannes G. Bode,2* Jens Schweigart,* Jan Kehrmann,* Christian Ehlting,* Fred Schaper,† Peter C. Heinrich,† and Dieter Ha¨ussinger*

Recently, it has been demonstrated that TNF-␣ and LPS induce the expression of suppressor of cytokine signaling 3 (SOCS3) and inhibit IL-6-induced STAT3 activation in macrophages. Inhibitor studies suggested that both induction of SOCS3 and inhibition of IL-6-induced STAT3 activation depend on the activation of p38 mitogen-activated protein kinase. Since recruitment of the tyrosine phosphatase Src homology protein tyrosine phosphatase 2 (SHP2) to the signal-transducing receptor subunit gp130 attenuates IL-6-mediated STAT-activation, we were interested in whether TNF-␣ also induces the association of SHP2 to the gp130 receptor subunit. In this study we demonstrate that stimulation of macrophages and fibroblast cell lines with TNF-␣ causes Downloaded from the recruitment of SHP2 to the gp130 signal-transducing subunit and leads to tyrosine phosphorylation of SHP2 and gp130. In this context the cytoplasmic SHP2/SOCS3 recruitment site of gp130 tyrosine 759 is shown to be important for the inhibitory effects of TNF-␣, since mutation of this residue completely restores IL-6-stimulated activation of STAT3 and, consequently, of a STAT3- dependent promoter. In this respect murine fibroblasts lacking exon 3 of SHP2 are not sensitive to TNF-␣, indicating that functional SHP2 and its recruitment to gp130 are key events in inhibition of IL-6-dependent STAT activation by TNF-␣. Fur- thermore, activation of p38 mitogen-activated protein kinase is shown to be essential for the inhibitory effect of TNF-␣ on IL-6 http://www.jimmunol.org/ signaling and TNF-␣-dependent recruitment of SHP2 to gp130. The Journal of Immunology, 2003, 170: 257–266.

he inflammatory cascade underlying the immune response macrophages (5Ð7), astrocytes (9), and fibroblasts (10) IL-6 has of an organism toward pathogens is largely controlled by suppressive effects on their inflammatory response and represses T the actions of different mediators released under inflam- the expression of IL-12, IFN-␥, IL-1␤, TNF-␣, adhesion mole- matory conditions. Upon activation, blood monocytes and tissue cules, and proteases both in vitro and in vivo (5Ð10). In line with macrophages release a set of primary inflammatory mediators, this, IL-6 is able to induce the expression of IL-1␤R antagonist and ␤ ␣ ␤ ␣ such as IL-1 and TNF- , thereby inducing the synthesis and se- soluble TNF receptor p55, antagonizing IL-1 and TNF- activ- by guest on October 1, 2021 cretion of several secondary cytokines and chemokines, such as ities, respectively (11, 12). IL-6 and IL-8, by macrophages, monocytes, and local stromal Thus, IL-6 mediates both pro- and anti-inflammatory effects. cells. Recruitment of other immune effector cells by chemotaxis IL-6 is expressed throughout every stage of the inflammatory re- then rapidly augments the local inflammatory response to coun- sponse, and at least during the onset of inflammation several of its teract the inflammatory stimulus and to remove cellular debris as- anti-inflammatory properties would be inconvenient (1Ð5). There- sociated with tissue damage (reviewed in Refs. 1Ð5). fore, it is mandatory that molecular mechanisms modulating IL-6 In this context IL-6 has been considered a proinflammatory cy- action exist. tokine, since its expression is elevated in inflammatory diseases IL-6 mediates its biological activities through a receptor com- and is induced by inflammatory stimuli, such as IL-1␤ and TNF-␣ plex composed of the specific receptor subunit gp80 and a dimer of (1, 2, 5). Many of its proinflammatory and immune properties are the signal-transducing receptor subunit gp130 (2). After ligand due to the activation of B cells to produce Abs due to the stimu- binding and dimerization of gp130, tyrosine kinase of the Janus lation of T cells and the induction of chemokine and adhesion family (Jak),3 Jak1, Jak2, and Tyk2, constitutively associated with molecule expression in endothelial cells (2, 5Ð8). However, on gp130, becomes activated by autophosphorylation. The gp130 sub- sequently tyrosine-phosphorylated on its cytoplasmic tail recruits transcription factors of the STAT family (13, 14) and Src homol- *Klinik fu¬r Gastroenterologie, Hepatologie und Infektiologie, Medizinische Klinik der Heinrich Heine Universita¬t, Düsseldorf, Germany; and †Institut fu¬r Biochemie, ogy protein tyrosine phosphatase 2 (SHP2) (15) via specific phos- Universita¬tsklinikum der Rheinisch-Westfa¬lischen Technischen Hochschule Aachen, photyrosine-SH2 domain interactions involving the tyrosine 759 of Aachen, Germany the gp130 receptor (16, 17). In turn, these components also become Received for publication September 19, 2002. Accepted for publication April tyrosine phosphorylated. Activated STATs homo- or heterodimer- 18, 2003. ize (18), translocate to the nucleus, and bind to enhancer elements 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 of target genes (19). with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by the Deutsche Forschungsgemeinschaft (Bonn, Ger- many) through the collaborative research center SFB 575 Experimentelle Hepatologie (Düsseldorf, Germany). 3 Abbreviations used in this paper: Jak, Janus kinase; Erk, extracellular signal-regu- 2 Address correspondence and reprint requests to Dr. Johannes G. Bode, Medizinische lated kinase; Epo, erythropoietin; IRF, IFN-regulatory factor; MAPK, mitogen-acti- Universita¬tsklinik, Klinik fu¬r Gastroenterologie, Hepatologie und Infektiologie, Hei- vated protein kinase; PIAS, protein inhibitor of activated STATs; SHP, SH2-contain- nrich Heine Universita¬t, Moorenstrasse 5, D-40225 Düsseldorf, Germany. E-mail ing protein tyrosine phosphatase; SOCS, suppressor of cytokine signaling; wt, wild address: [email protected] type.

Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00 258 TNF-␣-INDUCED PREASSOCIATION OF SHP2 AND gp130

The Jak/STAT signal transduction pathway is under negative anti-phosphotyrosine mouse mAb 4G10 (Upstate Biotechnology, Lake control by several different mechanisms. The presence of a nuclear Placid, NY) was used. phosphatase leading to dephosphorylation and inactivation of ac- Cultivation and stimulation of cells tivated STATs in the nucleus has been proposed by Haspel et al. RAW 264.7 cells were cultivated in DMEM (1000 mg of glucose/liter) (20). On the other hand, Kim and Maniatis (21) demonstrated a with Glutamax supplemented with 10% heat-inactivated FCS, streptomy- proteasome-dependent loss of activated STAT1 in the nucleus. Re- cin (100 mg/liter), and penicillin (60 mg/L). NIH-3T3 cells were grown in cently, another group of inhibitors of the Jak/STAT pathway has DMEM (4500 mg of glucose/L) with Glutamax supplemented with 10% been described: STAT-binding proteins, known as protein inhibi- heat-inactivated FCS, streptomycin (100 mg/L), and penicillin (60 mg/L). tors of activated STATs (PIAS) (22, 23). Although the PIAS do 3T3 embryonal fibroblasts were from SHP2 wild-type mice (SHP2-wt) or SHP2 exon 3-deficient mice (SHP2-mut) and were grown in DMEM (4500 not contain phosphotyrosine binding domains such as SH2 or pro- mg of glucose/L) with Glutamax, 10% FCS, 100 mg/L streptomycin, and tein tyrosine binding domains, they associate with activated, ty- 60 mg/L penicillin (35). rosine-phosphorylated STATs, leading to a loss of STAT-DNA All experiments, except those for the analysis of protein phosphoryla- binding activity. The mechanism and regulation of this highly spe- tion, were performed in the respective culture medium supplemented with 10% FCS. For the analysis of STAT3 and Erk1/2 phosphorylation, medium cific interaction of protein inhibitor of activated STATs with ac- was changed 12 h before the experiments were performed, and incubation tivated STAT factors remain to be elucidated. Another new family was continued in the respective culture medium supplemented with 0.5% of inhibitors of cytokine signaling has recently been discovered in FCS to reduce background activity of Erk-type MAPKs. three different laboratories. These proteins are referred to as sup- Nuclear extracts were prepared as described by Andrews and Faller (36). pressors of cytokine signaling (SOCS) (24), Jak-binding proteins Protein concentration was determined by protein assay (Bio-Rad, Munich, Germany). (25), or STAT-induced STAT inhibitors (26). SOCS proteins can also be regarded as feedback inhibitors of cytokine signaling, since Transfection procedure and reporter gene assay Downloaded from they are partially induced by cytokines mediating their own signals For reporter gene assay NIH-3T3 or MEF SHP2-wt and MEF SHP2-mut via activation of the Jak/STAT cascade. Furthermore, the protein cells were transfected using Lipofectamine 2000 (Invitrogen, San Diego, tyrosine phosphatase SHP2 was found to inhibit IL-6 signal trans- CA). Briefly, cells were grown in DMEM with 4500 mg of glucose/L duction. Activation of the IL-6R complex leads to recruitment of supplemented with 10% FCS on 12-well plates toward 80Ð95% conflu- ence. Cells were preincubated with 1 ml of OptiMEM 1 h before trans- SHP2 to tyrosine 759 in gp130 and its subsequent tyrosine phos- fection. Lipofectamine 2000 (6 ␮l) and in total 1.5 ␮g of DNA were predi- phorylation (15). SHP2 activation is a crucial event for the induc- luted in 100 ␮l of OptiMEM each. Diluted reagent and DNA were mixed, http://www.jimmunol.org/ tion of the mitogen-activated protein kinase (MAPK) pathway incubated for 20 min at room temperature, and then added to the cells. upon IL-6 stimulation (27). Mutation of Tyr759 in gp130 results After 16 h medium was changed, and incubation was continued in DMEM with 4500 mg of glucose/liter supplemented with 10% FCS, streptomycin in enhanced and prolonged STAT1 and STAT3 activation by IL-6- (100 mg/L), and penicillin (60 mg/L). For determination of the transcrip- type cytokines and increased gene induction of STAT-dependent tional activation of Elk-1, medium was changed before stimulation to cul- genes (28Ð30). ture medium supplemented with 0.5% FCS to reduce background activity. In terms of cytokine cross-talk, all these mechanisms can like- Cells were stimulated as indicated. Cell lysis and luciferase assays were wise be used by other mediators to influence cytokine signaling via conducted using the dual luciferase kit (Promega, Madison, WI) as de- scribed by the manufacturer. Luciferase activity values were normalized to the Jak/STAT signal transduction cascade and therefore represent transfection efficiency monitored by the cotransfected Renilla expression by guest on October 1, 2021 a potential molecular switch modulating the cellular response to- vector (Promega). Error bars are the SD calculated from three independent ward IL-6, e.g., TNF-␣ and LPS act as inhibitors of IL-6 and experiments performed on the same day. The data shown are representative IFN-␥-mediated STAT-activation, at least in the case of TNF-␣ prob- of at least three different experiments with similar results. MAPK For higher transfection efficiency cells were transfected using Lipo- ably through p38 -dependent induction of SOCS3 (31, 32). fectamine 2000 and a modified transfection procedure. In brief, 6 ␮lof Since SHP2 recruitment to the gp130 receptor negatively regu- Lipofectamine 2000 were diluted in 250 ␮l of OptiMEM, and 4 ␮gofDNA lates IL-6-induced STAT activation (33), we asked whether SHP2 was diluted in 50 ␮l of OptiMEM. Diluted reagent and DNA were mixed might participate in TNF-␣-mediated modulation of IL-6 signal and incubated for 20 min at room temperature. Meanwhile a confluent flask 2 transduction. Using mouse peritoneal RAW 264.7 macrophages (75 cm ) of NIH-3T3 cells was trypsinized and, after centrifugation, re- suspended in 3.5 ml of DMEM with 4500 mg glucose/L supplemented with and NIH-3T3 fibroblasts, we observed association to gp130 and 10% FCS, then 300 ␮l of resuspended cells were added to the transfection tyrosine phosphorylation of SHP2 upon stimulation with TNF-␣. mixture. One milliliter of DMEM with 4500 mg of glucose/L supple- Finally, mutation of the SHP2 binding site tyrosine 759 of gp130 mented with 10% FCS was added, and cells were seeded on a 60-mm results in an insensitivity toward preincubation with TNF-␣. The culture dish. After a 12-h incubation at 37¡C, medium was replaced by ␣ DMEM with 4500 mg glucose/L supplemented with 10% FCS, streptomy- recruitment of SHP2 to gp130 as a key event for the TNF- -in- cin (100 mg/L), and penicillin (60 mg/L), and cell culture was continued duced inhibition of IL-6-dependent STAT activation is further sup- for another 24 h. Thereafter, experiments were performed as outlined in ported by studies in murine fibroblasts lacking functional SHP2. Results. DNA constructs Materials and Methods ␣ Ϫ Materials pGL3- 2M-215Luc contains the promoter region 215 to 18 of the rat ␣ 2-macroglobulin gene fused to the luciferase encoding sequence and was Polymerase was purchased from Roche (Mannheim, Germany); oligonu- described previously (30). cDNAs for dominant negative p38MAPK mutant cleotides were obtained from MWG-Biotech (Ebersberg, Germany); re- tagged with the flag epitope (37) were cloned into the KRSPA expression combinant erythropoietin (Epo) was a gift from Drs. J. Burg and K. H. vector as described by Flory et al. (38). The expression vector pRc/ Sellinger (Roche); DMEM, DMEM nutritional mix F-12, OpitMEM, and CMV-EG encoding the chimeric EpoR/gp130 receptor (pRc/CMV-EG FCS were obtained from Life Technologies (Eggstein, Germany). Recom- (YYYYYY)) and the mutants where tyrosine 759 in the cytoplasmic do- binant human IL-6 and soluble IL-6R sgp80 were prepared as previously main of gp130 was exchanged for phenylalanine (pRc/CMV-EG described (34). The following Abs were used: rabbit polyclonal Ab spe- (YFYYYY)) or all tyrosines of the cytoplasmic domain of gp130 were cifically raised against extracellular signal-regulated kinase 2 (Erk2), replaced by phenylalanine (pRc/CMV-EG (FFFFFF)) have been described SHP2, or gp130 (M20; Santa Cruz Biotechnology, Santa Cruz, CA); rabbit previously (39). The construction of pCBC1-SHP2-wt was described pre- polyclonal Ab specifically raised against STAT3 phosphorylated at ty- viously (30). For the investigation of Erk-type MAPK activation, the vec- rosine 705, and mouse mAb specifically recognizing phosphorylated (ac- tors pFR-Luc and pFA2 Elk1 of the PathDetect trans-reporting system tivated) p42/p44 were obtained from Cell Signal Transduction Technology (Stratagene, La Jolla, CA) were used. pFR-Luc encodes a sequence of five (Beverly, MA); mAb against STAT3 was purchased from BD Transduction GAL4 binding elements fused to the luciferase encoding sequence. pFA2- Laboratories (San Diego, CA). For detection of tyrosine phosphorylation, Elk1 represents a fusion trans-activator plasmid that expresses a fusion The Journal of Immunology 259 protein of the activation domain of the transcription factor Elk1 fused with the yeast GAL4 DNA binding domain.

EMSA EMSAs were performed as described previously (19). The protein-DNA complexes were separated on a 4.5% polyacrylamide gel containing 7.5% glycerol in 0.25-fold TBE (20 mM Tris base, 20 mM boric acid, and 0.5 mM EDTA, pH 8.0) at 20 V/cm for 4 h. Gels were fixed in 10% methanol, 10% acetic acid, and 80% water for 1 h; dried; and autoradiographed. The double-stranded 32P-labeled mutated m67SIE oligonucleotide from the c- Fos promoter (m67SIE, 5Ј-GATCCGGGAGGGATTTACGGGAAATGC TG-3Ј) (40) was used for EMSA. FIGURE 1. TNF-␣-dependent coimmunoprecipitation of gp130 and SHP2. RAW 264.7 cells were stimulated with TNF-␣ (10 ng/ml) for the Coimmunoprecipitation times indicated. Cells were solubilized, and immunoprecipitation was per- formed using 1 ␮g of a polyclonal Ab specific for the tyrosine phosphatase For immunoprecipitation cells grown in a 100-mm dish were stimulated with the respective cytokine at the concentrations indicated. Cells were SHP2 as described in Materials and Methods. Thereafter, precipitated pro- teins were separated on an SDS-polyacrylamide (7.5%) gel and blotted washed twice with PBS supplemented with 0.1 mM Na3VO4 and solubi- lized in 1 ml of lysis buffer (1% Triton, 20 mM Tris-HCl (pH 7.4), 136 mM onto a polyvinylidene difluoride membrane. Membranes were cut in two NaCl, 2 mM EDTA, 50 mM ␤-glycerophosphate, 20 mM sodium pyro- parts at the level of ϳ80 kDa, and the upper part was incubated with phosphate, 1 mM Na3VO4, 4 mM benzamidine, 0.2 mM Pefabloc (Roche, polyclonal Ab raised against gp130 (1/2500; upper panel), whereas the Mannheim, Germany), 5 ␮g/ml aprotinin, 5 ␮g/ml leupeptin, and 10% lower part was incubated with polyclonal Ab for detection of SHP2 (1/ glycerol) for 20 min at 4¡C. Insoluble material was removed by centrifu- 4000; lower panel). Downloaded from gation, and the cell lysate was incubated overnight with specific Abs and protein A-Sepharose 4 Fast Flow (Amersham Pharmacia, Freiburg, Ger- many) (8 mg/ml in lysis buffer) at 4¡C. After centrifugation, the Sepharose beads were washed twice with wash buffer (0.1% Triton, 20 mM Tris-HCl TNF-␣ modulates IL-6-induced association of SHP2 with gp130 (pH 7.4), 136 mM NaCl, 2 mM EDTA, 50 mM ␤-glycerophosphate, 20 and inhibits IL-6-mediated activation of STAT3, whereas mM sodium pyrophosphate, 1 mM Na3VO4, 4 mM benzamidine, 0.2 mM activation of Erk is not affected Pefabloc, 5 ␮g/ml aprotinin, 5 ␮g/ml leupeptin, and 10% glycerol). The samples were boiled in gel electrophoresis sample buffer, and the precip- To further analyze the cross-talk between TNF-␣ and IL-6 signal- http://www.jimmunol.org/ itated proteins were separated on an SDS-polyacrylamide (7.5%) gel. ing the effect of TNF-␣ on the IL-6-dependent recruitment of SHP2 to the cytoplasmic tail of gp130 was studied in RAW 264.7 Immunoblotting and immunodetection macrophages. As shown in Fig. 3A, SHP2 was recruited to gp130 The electrophoretically separated proteins were transferred onto polyvi- upon stimulation of RAW 264.7 macrophages with IL-6. Recruit- nylidene difluoride membranes by the semidry Western blotting method. ment of SHP2 to gp130 was most pronounced after ϳ20 min of Nonspecific binding was blocked with 3% nonfatty dry milk powder in TBS-T (20 mM Tris-HCl (pH 7.4), 137 mM NaCl, and 0.1% Tween) treatment with IL-6 and was paralleled by IL-6-induced tyrosine overnight at 4¡C. For analysis of protein phosphorylation of Erk-type phosphorylation of gp130 and SHP2. Pretreatment with TNF-␣ for MAPKs or STAT3, nonspecific binding was blocked with 5% BSA in 40 min led to maximal complex formation of gp130 and SHP2 by guest on October 1, 2021 TBS-T. The blots were incubated overnight at 4¡Corfor2hatroom after 5 min of IL-6 treatment; this declined thereafter to baseline temperature with primary Abs at the dilution indicated in TBS-T. After levels within 40 min. Since this time course (complex formation 40 extensive rinsing with TBS-T, blots were incubated with secondary Abs, ␣ goat anti-rabbit IgG, or goat anti-mouse IgG conjugated to HRP for 1.5 h. min post-TNF- ) resembles that shown in Fig. 1, it is likely that After further rinsing in TBS-T, the immunoblots were developed with the TNF-␣ prevents SHP2 recruitment in response to IL-6 treatment, ECL system following the manufacturer’s instructions (Amersham Phar- although the preceding TNF-␣-induced complex formation is macia Biotech, Arlington Heights, IL). unaffected. Additionally, we analyzed the influence of TNF-␣ preincubation Results on the two major IL-6 downstream signaling events, the activation ␣ TNF- induces recruitment of SHP2 to gp130 of STAT factors and of Erk-type MAPKs. It is conspicuous that The SHP2 recruitment site within gp130 is involved in negative pretreatment with TNF-␣ inhibits IL-6-dependent tyrosine phos- regulation of IL-6-induced STAT activation and gene induction phorylation of STAT3 and STAT3 DNA binding (Figs. 4A and (28Ð30). Thus, SHP2 might also play a role in the TNF-␣-medi- 3B), whereas activation of Erk-type MAPKs was not inhibited, but, ated modulation of IL-6 signal transduction. Therefore, we tested rather, was accelerated and enhanced (Fig. 4B), particularly during whether TNF-␣ affects SHP2 binding to gp130. Figs. 1 and 2A the first 10Ð20 min, which correlates to the SHP2 recruitment show SHP2 and gp130 coimmunoprecipitated by either a gp130- found in macrophages upon costimulation with TNF-␣ and IL-6 or an SHP2-specific Ab from total cell lysates of RAW 264.7 shown in Fig. 3A. These findings suggest that TNF-␣ not only mouse peritoneal macrophages stimulated with TNF-␣ for differ- inhibits IL-6 signaling, but modulates IL-6-induced signal ent time periods. Association of gp130 and SHP2 was already transduction. detectable after 5 min of stimulation and became pronounced after 20Ð40 min. This indicates that TNF-␣ is able to mediate recruit- Tyrosine 759 of gp130 mediates the inhibition of IL-6-dependent ␣ ment of SHP2 to gp130. Association of SHP2 and gp130 was STAT activation by TNF paralleled by TNF-␣-induced tyrosine phosphorylation of both Recruitment of SHP2 and SOCS3 to the gp130 receptor subunit proteins, as shown in Fig. 2, A and B. In contrast to IL-6, TNF-␣ depends on the phosphorylation of the tyrosine residue 759 of did not lead to activation of STAT1 and -3, as shown in Fig. 2C. gp130 (15, 39, 41). To understand the role of this SHP2/SOCS3 This indicates that TNF-␣ time-dependently induces association of recruitment site in TNF-␣-mediated inhibition of IL-6-induced SHP2 to the signal-transducing subunit gp130 paralleled by ty- STAT activation, signaling through a gp130 receptor mutant, rosine phosphorylation of SHP2 and gp130. The fact that upon where a phenylalanine residue was substituted for the tyrosine res- stimulation with TNF-␣ STAT3 activation was not detectable ar- idue 759 of the SHP2/SOCS3 recruitment site, was analyzed. To gues against TNF-␣-induced release of IL-6 leading to subsequent perform such studies NIH-3T3 cells were selected because these SHP2 recruitment to gp130. cells can easily be transfected. As shown in Fig. 5, A and B, TNF-␣ 260 TNF-␣-INDUCED PREASSOCIATION OF SHP2 AND gp130 Downloaded from http://www.jimmunol.org/

FIGURE 3. Influence of TNF-␣ pretreatment on IL-6-mediated SHP2 recruitment to gp130 and on IL-6-induced STAT3-activation. After a pre- incubation period of 40 min with or without TNF-␣ (10 ng/ml) RAW 264.7 cells were stimulated with IL-6 (200 U/ml) for the times indicated. For A, cells were solubilized, and immunoprecipitation was performed using a polyclonal Ab specific for SHP2 as described in Materials and Methods. For analyses of tyrosine phosphorylation (first and third panels), mem- branes were incubated using a phosphotyrosine-specific monoclonal Ab by guest on October 1, 2021 (4G10; 1/1000). Blots were stripped, and membranes were cut into two parts at the level of ϳ80 kDa. The upper part was incubated with poly- clonal Ab specific for gp130 (1/2500; second panel), whereas the lower part was incubated with polyclonal Ab raised against SHP2 (1/4000; lower panel). For B, determination of STAT DNA binding from identically treated cells was conducted as described in Fig. 1.

FIGURE 2. TNF-␣-dependent tyrosine phosphorylation of SHP2 and gp130. RAW 264.7 cells were stimulated with TNF-␣ (10 ng/ml) for the exerts the same effects on IL-6 signal transduction in NIH-3T3 time indicated. As described in Materials and Methods, cell lysates were cells as those described for RAW 264.7 macrophages. Preincuba- prepared, and immunoprecipitation and subsequent immunoblotting were tion with TNF-␣ inhibits IL-6-induced STAT3 activation (Fig. performed using 1 ␮g of a polyclonal Ab specific for the receptor subunit 5A), and TNF-␣ again leads to recruitment of SHP2 to gp130 (Fig. gp130 (A) or the tyrosine phosphatase SHP2 (B). For detection of tyrosine- 5B), indicating that these cells are suitable for the following phosphorylated proteins, membranes were incubated using an mAb spe- experiments. cifically raised against phosphotyrosine motifs (4G10; 1/1000; A, upper For analyzing signaling through mutated receptor constructs we and third panels; B, upper panel). Blots were stripped. For A, membranes had to avoid stimulation of endogenous wild-type receptors. were cut in two parts at the level of ϳ80 kDa, and the upper part was incubated with polyclonal Ab raised against gp130 (1/1500; second panel), Therefore, chimeric receptor containing the extracellular domain whereas the lower part was incubated with polyclonal Ab for detection of of the Epo receptor and the transmembrane and wild-type (EG SHP2 (1/4000; last panel). For B, membranes were reprobed with poly- (YYYYYY)) or mutated cytoplasmic domains of gp130 (EG(Y clonal Ab specifically raised against SHP2 (1/4000; lower panel). C, For FYYYY)) or for control EG(FFFFFF)) were used. These con- determination of the STAT activation cells were stimulated with TNF-␣ structs were cotransfected together with a STAT3-responsive ␣ (10 ng/ml) or IL-6 (200 U/ml) for the time indicated. Determination of 2-macroglobulin promoter/luciferase reporter into NIH-3T3 STAT activation was performed after preparation of nuclear extracts. Five fibroblasts. 32 micrograms of nuclear protein was mixed with a STAT1/3-specific P-labeled Stimulation of cells expressing EG(YYYYYY) with Epo led to Ј oligonucleotide (mutated m67SIE probe of the c-Fos promoter 5 -GAT CCG a significant expression of the reporter. According to the findings GGA GGG ATT TAC GGG AA ATG CTG-3Ј), and EMSAs were performed. for IL-6-induced STAT3 activation, transcriptional activation of The positions of comigrating STAT1/3 heterodimer and STAT3 homodimer ␣ from IL-6 stimulated HepG2 cells are indicated by arrows. the reporter by Epo was inhibited by preincubation with TNF- (Fig. 5C). This inhibitory activity of TNF-␣ depended on the pres- ence of Y759 in gp130, since its mutation counteracted the inhib- itory function of TNF-␣ and led to increased IL-6-induced reporter The Journal of Immunology 261

observed by mutating tyrosine 759 of gp130 (compare panels 2 and 3 in Fig. 6, A and B). Furthermore, no additional effect of dominant negative p38MAPK was observed in cells expressing EG (YFYYYY) ( panels 3 and 4 in Fig. 6B). Thus, these data indicate that besides the inhibitory activity of p38MAPK further inhibitory signals are integrated to gp130-dependent signal transduction via tyrosine 759. To determine whether TNF-␣-induced recruitment of the pro- tein tyrosine phosphatase SHP2 to gp130 might depend on p38MAPK or tyrosine kinase activity, inhibitor studies were per- formed in RAW264.7 cells. As shown in the coimmunoprecipita- tion studies with SHP2-specific Abs in Fig. 7A, inhibition of p38MAPK activation by pretreatment with SB202190, an inhibitor supposed to be specific for p38MAPK, reduces TNF-␣-dependent SHP2 recruitment to gp130. These data suggest that activation of p38MAPK might be involved in the recruitment of SHP2 to gp130 upon stimulation with TNF-␣. On the other hand, pretreatment with the tyrosine kinase inhibitor genistein also inhibited SHP2 recruitment to gp130 upon stimulation with TNF-␣, indicating that tyrosine kinase activity is also required (Fig. 7B). Downloaded from

p38MAPK counteracts MAPK activity induced by TNF-␣, but does not play a specific role for IL-6-induced activation of MAPK activity, and the effects of TNF-␣ on this

SHP2 have been demonstrated to serve as an adaptor protein link- http://www.jimmunol.org/ FIGURE 4. Influence of TNF-␣ pretreatment on IL-6-induced phos- ing gp130 signal transduction to Erk-type MAPK signaling (27). phorylation of STAT3- or Erk-type MAPK. After a preincubation period of We therefore asked whether p38MAPK might also be involved in ␣ 40 min with or without TNF- (10 ng/ml), RAW 264.7 cells were stimu- activation of Erk-1/2 by IL-6, TNF-␣, or TNF-␣ plus IL-6. To lated with IL-6 (200 U/ml), and incubation was continued for the times address this question, reporter gene assays were performed in NIH- indicated. Cells were then solubilized for immunoblotting. Tyrosine phos- phorylation of STAT3 (A) and tyrosine/threonine phosphorylation of Erk- 3T3 cells using a plasmid that expresses a fusion protein of the type MAPK (B) were determined by immunoblot analyses using specific activation domain of the transcription factor Elk1 (pFA-Elk1) Abs recognizing STAT3 phosphorylated at tyrosine 705 or the activated fused with the yeast GAL4 DNA binding domain. Phosphorylation (phosphorylated) form of Erk1/2 (upper panels). As a loading control, blots of the Elk1 activation domain induces transcriptional activation of where stripped and reprobed with Abs specific for STAT3 or Erk2 (lower a reporter plasmid encoding a sequence of five GAL4 binding by guest on October 1, 2021 panels). elements fused to the luciferase-encoding sequence (pFA-Luc). As shown in Fig. 8 transcriptional activation via the Elk1 trans- activating fusion protein is induced by stimulation of the cotrans- activity. Receptors lacking all cytoplasmic tyrosine motifs fected chimeric EG (YYYYYY) receptor with erythropoietin. Ac- (EG(FFFFFF)) did not mediate any significant gene induction. cording to the data shown for Erk-type MAPK activation in Fig. 4, TNF␣-mediated recruitment of SHP2 to gp130 and inhibition of costimulation with TNF-␣ or stimulation with TNF-␣ alone results IL-6-induced STAT activation involve tyrosine kinase activity in a much stronger expression of luciferase activity. However, co- and activation of p38MAPK transfection of an inactive mutant of p38MAPK leads to a significant enhancement of the transcriptional activation of Elk1 after co- Previous studies suggested that p38MAPK is involved in the inhi- stimulation with TNF-␣ plus Epo or stimulation with TNF-␣ bition of IL-6-induced STAT activation by proinflammatory me- alone, whereas basal and Epo-induced activation was only slightly diators such as TNF-␣ (31, 42, 43). Thus, it was intriguing to affected. Since activation of Elk1 is downstream from Erk/MAPK investigate the effect of p38MAPK on TNF-␣-induced SHP2 re- (44) one can conclude from these data that p38MAPK counteracts cruitment to gp130, inhibition of IL-6-induced STAT activation, MAPK activity induced by TNF-␣, but does not play a specific and gene expression. Until now, the role of p38MAPK for the in- role for IL-6-induced activation of MAPK activity and the effects hibitory effects of TNF-␣ on IL-6-induced STAT3 activation was of TNF-␣ on this. With respect to this, it is interesting to note that only demonstrated by studies using low m.w. inhibitors thought to p38MAPK is involved in the stress-induced expression of members be specific for p38MAPK. In this study we performed a more spe- of the MAPK phosphatase family, thereby counteracting the acti- cific experimental approach using dominant negative p38MAPK vation of MAPKs (45Ð47). Thus, one might speculate that TNF-␣ (37) to confirm the inhibitory function of p38MAPK. balances the activation of Erk-1/2 via p38MAPK-dependent activa- As shown in Fig. 6 the inhibition of gp130-mediated STAT3 tion of members of the MAPK phosphatase family. activation (Fig. 6A) and transcriptional activation of the STAT3- responsive reporter gene (Fig. 6B) by TNF-␣ pretreatment was completely restored by cotransfection of this inactive mutant of Involvement of the tyrosine phosphatase SHP2 in inhibition of ␣ p38MAPK (compare two left panels in Fig. 6, A and B). Thus, these IL-6-induced STAT activation by TNF findings provide further evidence that activation of p38MAPK is As shown in Fig. 9, the expression of SHP2 strongly suppressed indeed involved in inhibition of gp130-dependent STAT3 activa- STAT activation mediated by the wild-type EpoR/gp130 chimera, tion by TNF-␣. indicating the inhibitory potential of SHP2 with respect to gp130- However, the enhanced transcriptional activation achieved by dependent signal transduction. To further reveal the implication of cotransfection of the inactive p38MAPK was not as strong as that SHP2 in inhibition of IL-6 signaling by TNF-␣, the inhibition of 262 TNF-␣-INDUCED PREASSOCIATION OF SHP2 AND gp130 Downloaded from http://www.jimmunol.org/

FIGURE 5. Involvement of tyrosine 759 of the cytoplasmic tail of gp130 in TFN-␣-mediated inhibition of IL-6 signaling. A, NIH-3T3 cells were pretreated for 40 min with TNF-␣ (10 ng/ml) and then stimulated with IL-6 (200 U/ml) and soluble gp80 (10 ␮g/ml) for the times indicated. After preparation of nuclear extracts, STAT activation was analyzed by EMSA as described in Fig. 1 and Materials and Methods. B, After stimulation of NIH-3T3 cells with TNF-␣ (10 ng/ml) for the time indicated, cell lysates were prepared, and immunoprecipitation was performed using 1 ␮g of an Ab specific for SHP2. Subsequent to immunoblotting, membranes were cut into two parts at the level of ϳ80 kDa, and the upper part was incubated with polyclonal Ab raised against gp130 (1/2500; upper panel), whereas the lower part was incubated with polyclonal SHP2 Ab (1/4000; lower panel). C, NIH-3T3 cells were

␣ Ϫ ϩ by guest on October 1, 2021 cotransfected with a reporter gene construct containing the 2-macroglobulin promoter ( 209 to 8) fused to the firefly luciferase gene and expression vectors for chimeric receptors containing the extracellular domain of the Epo receptor and the transmembrane and wild-type (EG(YYYYYY)) or mutated cytoplasmic domains of gp130 where the tyrosine residue 759 of the SHP2/SOCS3 recruitment site (EG(YFYYYY)) or all six tyrosine residues (EG(FFFFFF)) were substituted by phenylalanine. An expression vector for Renilla luciferase was cotransfected for monitoring transfection efficiency. Two days after transfection, cells were preincubated with 10 ng/ml TNF-␣ for 40 min and stimulated with or without Epo (7 U/ml) for 10 h as shown. Luciferase activity in cellular extracts of these cells was determined and normalized to the Renilla luciferase activity as outlined in Materials and Methods.

STAT3 and promoter activation was analyzed in murine fibro- tokine cross-talk on the level of signal transduction are scarcely blasts lacking exon 3 of SHP2. These cells express a mutant SHP2 known. Induction of SOCS3 expression, which was at least par- protein lacking 65 aa within the N-terminal SH2 domain (35). In tially dependent on activation of p38MAPK, has been proposed as a clear contrast to the corresponding wild-type fibroblasts (SHP2- possible mechanism for the inhibitory effects of TNF-␣ and LPS wt), IL-6-induced STAT3 tyrosine phosphorylation and DNA on IL-6 signaling (31). On the other hand, several lines of evidence binding in SHP2-mut cells was almost not affected by preincuba- exist that p38MAPK-dependent inhibition of IL-6-induced STAT3 tion with TNF-␣ (Fig. 10, A and B). Accordingly, preincubation of by IL-1␤ occurs in the absence of new protein synthesis. This wild-type cells with TNF-␣ inhibited gp130-dependent transcrip- indicates that in this context protein synthesis, and thus SOCS ␣ tional activation of the 2-macroglobulin promoter reporter gene expression, is not essential for inhibition of IL-6 signaling by construct, whereas no inhibitory effect of TNF-␣ could be ob- IL-1␤ (42, 43). Furthermore, analyzing the mechanism underlying served in cells expressing the SHP2-mutant (Fig. 10C). These data the IL-1␤-mediated attenuation of IL-6 signal transduction and clearly indicate that SHP2 is involved in inhibition of IL-6 signal acute phase protein synthesis, it was observed that IL-1␤ did not transduction by TNF-␣. induce SOCS3 gene expression in hepatic parenchymal cells and barely affected IL-6-mediated STAT3 activation, but did affect its Discussion binding to IL-6-responsive promoters (51). In these studies NF-␬B In the past it became increasingly evident that the IL-6 signal was identified as a mediator of IL-1␤-dependent negative regula- transduction is substantially modulated by proinflammatory medi- tion of IL-6-inducible genes (50, 51). Most likely, in this context ators such as TNF-␣, IL-1␤, and LPS. It was demonstrated that NF-␬B exerts its negative regulatory function by counteracting TNF-␣ and LPS inhibit IL-6-induced activation of STAT3 in sev- DNA binding of STAT3 at overlapping STAT3/NF-␬B binding eral types of macrophages, monocytic cell lines, and synovial fi- sites within the promoter region of the respective gene. This reg- broblasts (31, 42Ð43). Furthermore, IL-1␤ inhibits IL-6-induced ulatory mechanism represents a very fast and accurate mode by signal transduction and expression of acute phase proteins in liver which IL-1␤ modulates IL-6-induced gene expression of specific cells (48Ð51). The underlying molecular mechanisms for this cy- target genes (50, 51). The Journal of Immunology 263

FIGURE 6. Involvement of p38MAPK in inhibi- tion of IL-6 signaling by TNF-␣. A, NIH-3T3 cells were transiently cotransfected with either the empty KRSPA vector or with a flag-tagged kinase-deficient mutant of p38MAPK (AF; dominant negative (dn) p38MAPK) and chimeric receptor constructs containing the transmembrane and wild-type (EG(YYYYYY)) or mutated cytoplasmic domains (EG(YFYYYY)) of gp130. After 2 days cells were pretreated for 40 min with TNF-␣ as indicated and then stimulated with Epo (7 U/ml) for 30 min. Thereafter, cells were harvested, Downloaded from and nuclear extracts for the assessment of STAT acti- vation were prepared. STAT activation was analyzed by EMSA as outlined in Fig. 2. B, NIH-3T3 cells were transiently cotransfected with either the empty KRSPA (control) vector or with dn p38MAPK and the respective chimeric receptor construct EG(YYYYYY) or EG(Y- http://www.jimmunol.org/ FYYYY). For determination of the transcriptional ac- ␣ tivation the 2-macroglobulin promoter reporter gene construct was cotransfected. Transfection efficiency was monitored by cotransfected Renilla luciferase cDNA. Two days after transfection cells were prein- cubated with 10 ng/ml TNF-␣ for 40 min and stimu- lated with or without Epo (7 U/ml) for 10 h as shown. Luciferase activity in cellular extracts of these cells was determined and normalized to the Renilla lucif- erase activity as outlined in Materials and Methods. by guest on October 1, 2021

In this paper we identified the recruitment of SHP2 to the cy- Thus, both SHP2 and SOCS3 might contribute to TNF-␣-depen- toplasmic part of the gp130 signal-transducing subunit of the dent signal attenuation. Indeed, the expression of SOCS3 occurs IL-6R complex as another fast-acting mechanism crucial for the upon stimulation of macrophages with TNF-␣ (31), and on the TNF-␣-dependent negative regulation of IL-6-mediated STAT3 other hand, specific disturbance of SHP2 recruitment to gp130 also tyrosine phosphorylation and activation in macrophages and fibro- diminishes the inhibitory activity of TNF-␣ (Fig. 10). The indi- blast cell lines. Inhibitor studies revealed that TNF-␣-induced vidual contributions of SHP2 and SOCS3 in the context of TNF-␣ binding of SHP2 to gp130 might depend on protein tyrosine kinase remains to be established. In this respect, it was recently demon- activity and activation of p38MAPK (Fig. 7). In line with these strated that SHP2 and SOCS3, although recruited to the same site findings, the expression of a dominant negative p38MAPK com- of gp130, are able to exert their inhibitory function independently pletely restores inhibition of IL-6-induced STAT activation by of each other (33). Since SOCS3 protein remains detectable longer TNF-␣, further indicating the importance of p38MAPK for mediat- than SHP2 tyrosine phosphorylation, and recruitment of SHP2 to ing the effect of TNF-␣ on IL-6 signal transduction (Fig. 6). Fur- gp130 Ϫ SHP2 may be important for early signal attenuation, thermore, evidence is given that tyrosine phosphorylation of the whereas SOCS3 acts later. tyrosine 759 of the cytoplasmic tail of gp130 is a central event for As mentioned above, p38MAPK was demonstrated to be involved the inhibitory action of TNF-␣ on IL-6 signaling (Figs. 5 and 6). in the inhibitory action of proinflammatory mediators (31, 42, 43). Phosphorylation of this tyrosine has been shown to be crucial for Particularly, activation of p38MAPK was shown to be essential for the recruitment of both the protein tyrosine phosphatase SHP2 and the TNF-␣- and LPS-induced expression of SOCS3 (31), but is the SOCS3 to the gp130 signal-transducing subunit (15, 39, 41). also involved in the autoregulatory SOCS3 expression induced by 264 TNF-␣-INDUCED PREASSOCIATION OF SHP2 AND gp130

FIGURE 8. p38MAPK counteracts MAPK activity induced by TNF-␣, but does not play a specific role for IL-6-induced activation of MAPK activity and the effects of TNF-␣ on this. NIH-3T3 cells were cotransfected

with the chimeric receptor (EG(YYYYYY), the pFR-Luc plasmid and the Downloaded from pFA2 Elk1 plasmid of the PathDetect trans-reporting system from Strat- agene and either the empty KRSPA vector or with a flag tagged kinase- deficient mutant of p38MAPK (AF; dominant negative (dn) p38MAPK). Two days after transfection cells were preincubated with 10 ng/ml TNF-␣ for 40 min and stimulated with or without Epo (7 U/ml) for 10 h as shown. Luciferase activity in cellular extracts of these cells was determined and

normalized to the Renilla luciferase activity as outlined in Materials and http://www.jimmunol.org/ Methods.

FIGURE 7. Influence of the activation of p38MAPK or protein tyrosine gp130 cytoplasmic tail, containing one consensus phosphorylation kinase on TNF-␣-mediated recruitment of SHP2 to gp130. Following a site for MAPKs and a serine-rich region (15). Thus, this mem- preincubation period for 40 min with SB 202190 (10 ␮M; A)or40min brane-proximal region of gp130 might represent another target for ␮ with Genistein (100 M; B), RAW 264.7 cells were treated with 10 ng/ml the inhibitory effects of p38MAPK on gp130-dependent signaling. TNF-␣ as indicated (40 min in A). The extract of total protein was sub- Despite its inhibitory activity on IL-6-induced activation of mitted to immunoprecipitation as described in Materials and Methods us- by guest on October 1, 2021 ing 1 ␮g of Abs specific for the SHP2. After immunoblotting membranes STAT factors, SHP2 has been further shown to serve as an adaptor were cut in two parts at the level of ϳ80 kDa and the upper part was protein linking gp130 signal transduction to the Ras/Raf/Erk cas- MAPK incubated with polyclonal Ab specifically raised against gp130 (1/2500; cade (27, 54). On the other hand, p38 interferes with TNF- upper panel), whereas the lower part was incubated with polyclonal Ab ␣-induced recruitment of SHP2 to gp130. Thus, one might expect specifically raised against SHP2 (1/4000; lower panel). that inhibition of p38MAPK somehow affects TNF-␣/IL-6-induced

IL-6 (52). The induction of SOCS3 has been suggested to be part of the molecular mechanism underlying the inhibitory effects of TNF-␣ and LPS. The expression of a dominant negative mutant of p38MAPK completely rescues inhibition of gp130-mediated activa- ␣ tion of STAT3 and the STAT3-dependent 2-macroglobulin pro- moter (Fig. 6). It is further demonstrated that inhibition of p38MAPK using an inhibitor specific for p38MAPK attenuates TNF- ␣-induced recruitment of SHP2 to the gp130 subunit (Fig. 7A). These data indicate that p38MAPK activation by TNF-␣ might be involved in the TNF-␣-dependent recruitment of SHP2 to the gp130 receptor subunit. p38MAPK is a serine/threonine protein ki- nase (53); thus, a direct influence on the tyrosine phosphorylation of gp130 is impossible. We could not detect reliable changes in TNF-␣-induced tyrosine phosphorylation of SHP2 and gp130 upon pretreatment with a p38MAPK-specific inhibitor (data not shown), indicating that an involvement of p38MAPK in regulation FIGURE 9. Effects of SHP2 on IL-6-induced STAT-activation. NIH- of tyrosine phosphorylation of gp130 is unlikely. However, the 3T3 cells were cotransfected with either the empty pCBC1 vector or with the expression plasmid pCBC1-SHP2WT (SHP2 wt) encoding for the regulatory effect of p38MAPK on SHP2 recruitment to gp130 might wild-type protein tyrosine phosphatase SHP2 and the chimeric receptors also depend on phosphorylation of additional sites representing a (EG(YYYYYY) or EG(YFYYYY). After 2 days cells were pretreated for MAPK consensus motif. With respect to this, it is interesting to 40 min with TNF-␣ as indicated and then stimulated with Epo (7 U/ml) for note that using truncated gp130 receptor constructs, Ahmed et al. 30 min. Thereafter cells were harvested, and nuclear extracts for the as- MAPK (42) restricted the target of the p38 -mediated inhibitory ef- sessment of STAT activation were prepared. STAT activation was ana- fects on IL-6 signaling to the membrane-proximal 113 aa of the lyzed by EMSA as outlined in Fig. 2. The Journal of Immunology 265

gests that p38MAPK activity reduces transcriptional activation of Elk-1 by TNF-␣, but is not specifically involved in IL-6-induced Erk activation and the effect of TNF-␣ on it. However, further investigation is required to clarify the interplay between p38MAPK and Erk activation upon costimulation with IL-6 and TNF-␣ as well as the role of SHP2 with respect to the activation of Erk-type MAPKs. As demonstrated in this study, TNF-␣-induced recruitment of SHP2 to the cytoplasmic tail of gp130 is involved in the inhibitory effects of TNF-␣ on IL-6 signal transduction via STAT3. The ob- servation that IL-6-induced Erk activation is enhanced rather than inhibited by costimulation with TNF-␣ suggests that SHP2 recruit- ment toward gp130 might function as a kind of molecular switch, inhibiting signal transduction via STAT factors, whereas signaling via Erk-type MAPKs is still admitted or even strengthened. In some aspects this situation resembles to the COOH-terminal gp130⌬STAT knockin mutation that deleted all STAT binding sites. Mice with this gp130⌬STAT mutation displayed gastrointestinal ul- ceration and severe joint disease with features of chronic synovitis, cartilaginous metaplasia, and degradation of the articular cartilage Downloaded from (55). In these animals mitogenic hyper-responsiveness of synovial cells to the leukemia inhibitory factor/IL-6 family of cytokines was related to sustained gp130-mediated SHP2/Ras/Erk activation and the lack of the activation of STAT factors. Thus, the pathologic changes observed in gp130 mice⌬STAT are likely to arise from the disturbance of the otherwise balanced activation of the SHP2/Ras/ http://www.jimmunol.org/ Erk and STAT signaling cascades emanating from gp130. The similarity of the modifications of gp130-dependent signal trans- duction in gp130⌬STAT mice and those presented here are inter- esting, since TNF-␣ also blocks IL-6-induced STAT activation, but not the activation of Erk-type MAPKs (Fig. 4). Thus, one might speculate that TNF-␣ provokes a pathophysiological situa- tion in IL-6 signaling similar to that achieved with the COOH- terminal gp130⌬STAT knockin mutation that deleted all STAT by guest on October 1, 2021 binding sites.

Acknowledgments Murine fibroblasts lacking exon 3 of SHP2 were kindly provided by B. G. Neel (Boston, MA). We thank M. Ruhl and J. Matthes for technical FIGURE 10. Role of SHP2 for the inhibitory effects of TNF-␣ on IL- assistance. 6-signaling. A, Murine fibroblasts lacking exon 3 of SHP2 (SHP2 mut) derived from the respective knockout mice and the corresponding wild- References type fibroblasts (SHP2 wt) were pretreated with 10 ng/ml TNF-␣ for 40 1. Bode, J. G., and P. C. Heinrich. 2001. Interleukin-6 signaling during the acute min and stimulated with 150 U/ml of IL-6 and soluble IL-6 receptor phase response of the liver. In The Liver: Biology and Pathobiology, 4th Ed. (sgp80; 10 ␮g/ml) for the times indicated. Thereafter, cells were harvested, I. M. Arias, J. L. Boyer, N. Fausto, F. V. Chisari, H. Schachter, and D. A. Shafritz, eds. Lippincott Williams & Wilkins, Philadelphia, p. 565. and nuclear extracts for the assessment of STAT activation and tyrosine 2. Heinrich, P. C., I. Behrmann, G. Mu¬ller-Newen, F. Schaper, and L. Graeve. 1998. phosphorylation were prepared. STAT activation was analyzed by EMSA Interleukin-6-type cytokine signalling through the gp130/Jak/STAT pathway. as outlined in Fig. 2. B, Tyrosine phosphorylation of STAT3 was assessed Biochem. J. 334:297. by immunoblot analyses using Abs recognizing STAT3 tyrosine phosphor- 3. Gabay, C., and I. Kushner. 1999. Acute-phase proteins and other systemic re- ylated at tyrosine 705. As a loading control blots were stripped and re- sponses to inflammation. N. Engl. J. Med. 340:448. 4. Ramadori, G., and B. Christ. 1999. Cytokines and the hepatic acute-phase re- probed with Abs specific for STAT3 (second panel). C, SHP2 wt and SHP2 sponse. Semin. Liv. Dis. 19:141. ␣ mut fibroblasts were cotransfected with the 2-macroglobulin promoter 5. Barton, B. E. 1996. The biological effects of interleukin-6. Med. Res. Rev. 16:87. reporter gene construct and the chimeric receptors EG(YYYYYY). Two 6. Romano, M., M. Sironi, C. Toniatti, N. Polentarutti, P. Fruscella, P. Ghezzi, days after transfection, cells were preincubated with 10 ng/ml TNF-␣ for R. Faggioni, W. Luini, V. van Hinsbergh, S. Sozzani, et al. 1997. Role of IL-6 and its soluble receptor in induction of chemokines and leukocyte recruitment. 40 min and stimulated with or without Epo (7 U/ml) for 10 h as shown. Immunity 6:315. Luciferase activity in cellular extracts of these cells was determined and 7. Riedy, M. C., and C. C. Stewart. 1992. Inhibitory role of interleukin-6 in mac- normalized to the Renilla luciferase activity as outlined in Materials and rophage proliferation. J. Leukocyte Biol. 52:125. Methods. 8. Ulich, T. R., K. Z. Guo, D. Remick, J. del Castillo, and S. M. Yin. 1991. Endo- toxin-induced cytokine gene expression in vivo. III. IL-6 mRNA and serum pro- tein expression and the in vivo hematologic effects of IL-6. J. Immunol. 146: 2316. 9. Oh, J. W., N. J. Van Wagoner, S. Rose-John, and E. N. Benveniste. 1998. Role Erk activation. As shown in Fig. 8, cotransfection of a dominant of IL-6 and the soluble IL-6 receptor receptor in inhibition of VCAM-1 gene MAPK negative mutant of p38 leads to enhanced transcriptional ac- expression. J. Immunol. 161:4992. tivation of Elk-1 in the presence of TNF-␣, but not upon activation 10. Gatsios, P., H. D. Haubeck, E. Van de Leur, W. Frisch, S. S. Apte, H. Greiling, P. C. Heinrich, and L. Graeve. 1996. Oncostatin M differentially regulates tissue of the Epo/gp130 receptor chimera with Epo. Considering that inhibitors of metalloproteinases TIMP-1 and TIMP-3 gene expression in human Elk-1 is downstream from Erk/MAPK activation (44), this sug- synovial lining cells. Eur. J. Biochem. 241:56. 266 TNF-␣-INDUCED PREASSOCIATION OF SHP2 AND gp130

11. Tilg, H., E. Trehu, M. B. Atkins, C. A. Dinarello, and J. W. Mier. 1994. Inter- 33. Lehmann, U., J. Schmitz, M. Weissenbach, R. M. Sobota, M. Hortner, leuki-6 (IL-6) as an anti-inflammatory cytokine: induction of circulating IL-1 K. Friederichs, I. Behrmann, W. Tsiaris, A. Sasaki, J. Schneider-Mergener, et al. receptor antagonist and soluble tumor necrosis factor receptor p55. Blood 83:113. 2003. SHP2 and SOCS3 contribute to Y759-dependent attenuation of IL-6-sig- 12. Tilg, H., E. Vannier, G. Vachino, C. A. Dinarello, and J. W. Mier. 1993. Anti- naling through gp130. J. Biol. Chem. 278:661. inflammatory properties of hepatic acute phase proteins: preferential induction of 34. Weiergra¬ber, O., U. Hemmann, A. Ku¬ster, G. Mu¬ller-Newen, J. Schneider, interleukin 1 IL-1 receptor antagonist over IL-1␤ synthesis by human peripheral S. Rose-John, P. Kurschat, J. P. J. Brakenhoff, M. H. L. Hart, S. Stabel, et al. blood mononuclear cells. J. Exp. Med. 178:1629. 1995. Soluble human interleukin-6 receptor: expression in insect cells, purifica- 13. Lu¬tticken, C., U. M. Wegenka, J. Yuan, J. Buschmann, C. Schindler, tion and characterization. Eur. J. Biochem. 234:661. A. Ziemiecki, A. G. Harpur, A. F. Wilks, K. Yasukawa, T. Taga, et al. 1994. 35. Oh, E. S., H. Gu, T. M. Saxton, J. F. Timms, S. Hausdorff, E. U. Frevert, Association of transcription factor APRF and protein kinase Jak 1 with the In- B. B. Kahn, T. Pawson, B. G. Neel, and S. M. Thomas. 1999. Regulation of early terleukin-6 signal transducer gp 130. Science 263:89. events in integrin signaling by protein tyrosine phosphatase SHP-2. Mol. Cell. 14. Stahl, N., T. G. Boulton, T. Farruggella, N. Y. Ip, S. Davis, B. A. Witthuhn, Biol. 19:3205. F. W. Quelle, O. Silvennoinen, G. Barbieri, S. Pellegrini, et al. 1994. Association 36. Andrews, N. C., and D. V. Faller. 1991. A rapid micropreparation technique for and activation of Jak-Tyk kinases by CNTF-LIF-OSM-IL-6 ␤ receptor compo- extraction of DNA-binding proteins from limiting numbers of mammalian cells. nents. Science 263:92. Nucleic Acids Res. 19:2499. 15. Stahl, N., T. J. Faruggella, T. G. Boulton, Z. Zhong, J. E. Darnell, Jr., and 37. Ludwig, S., A. Hoffmeyer, M. Goebler, K. Kilian, H. Ha¬fner, B. Neufeld, G. D. Yancopoulos. 1995. Choice of STATs and other substrates specified by H. Jiahuai, and U. R. Rapp. 1998. The stress inducer arsenite activates mitogen- modular tyrosine-based motifs in cytokine receptors. Science 267:1349. activated protein kinases extracellular signal-regulated kinases 1 and 2 via a 16. Heim, M. H., I. M. Kerr, G. R. Stark, and J. E. Darnell, Jr. 1995. Contribution of MAPK kinase 6/p38-dependent pathway. J. Biol. Chem. 273:1917. STAT SH2 groups to specific interferon signaling by the Jak-STAT pathway. 38. Flory, E., A. Hoffmeyer, U. Smola, U. R. Rapp, and J. T. Bruder. 1996. Raf-1 Science 267:1347. kinase targets GA-binding protein in transcriptional regulation of the human im- 17. Hemmann, U., C. Gerhartz, B. Hessel, J. Sasse, G. Kurapkat, J. Gro¬tzinger, munodeficiency virus type 1 promoter. J. Virol. 70:2260. Wollmer A, Z. Zhong, J. E. Darnell, Jr., L. Graeve, et al. 1996. Differential 39. Schmitz, J., M. Weissenbach, S. Haan, P. C. Heinrich, and F. Schaper. 2000. activation of acute phase response factor/Stat3 and Stat1 via the cytoplasmic SOCS3 exerts its inhibitory function on interleukin-6 signal transduction through domain of the interleukin 6 signal transducer gp130. II. Src homology SH2 do- the SHP2 recruitment site of gp130. J. Biol. Chem. 275:12848. 40. Wagner, B. J., T. E. Hayes, C. J. Hoban, and B. H. Cochran. 1990. The SIF

mains define the specificity of Stat factor activation. J. Biol. Chem. 271:12999. Downloaded from 18. Zhong, Z., Z. Wen, and J. E. Darnell, Jr. 1994. Stat3: a STAT family member binding element confers sis/PDGF inducibility onto the c-fos promoter. EMBO J. activated by tyrosine phosphorylation in response to epidermal growth factor and 9:4477. interleukin-6. Science 264:95. 41. Nicholson, S. E., D. De Souza, L. J. Fabri, J. Corbin, T. A. Willson, J. G. Zhang, 19. Wegenka, U. M., J. Buschmann, C. Lu¬tticken, P. C. Heinrich, and F. Horn. 1993. A. Silva, M. Asimakis, A. Farley, A. D. Nash, et al. 2000. Suppressor of cytokine Acute-phase response factor, a nuclear factor binding to acute-phase response signaling-3 preferentially binds to the SHP-2-binding site on the shared cytokine elements, is rapidly activated by interleukin-6 at the posttranslational level. Mol. receptor subunit gp130. Proc. Natl. Acad. Sci. USA 97:6493. Cell. Biol. 13:276. 42. Ahmed, S. T., and L. B. Ivashkiv. 2000. Inhibition of IL-6 and IL-10 signaling 20. Haspel, R. L., M. Salditt-Georgieff, and J. E. Darnell, Jr. 1996. The rapid inac- and STAT activation by inflammatory and stress pathways. J. Immunol. 165: tivation of nuclear tyrosine phosphorylated Stat1 depends upon a protein tyrosine 5227. http://www.jimmunol.org/ phosphatase. EMBO J. 15:6262. 43. Ahmed, S. T., A. Mayer, J. D. Ji, and L. B. Ivashkiv. 2002. Inhibition of IL-6 signaling by a p38-dependent pathway occurs in the absence of new protein 21. Kim, T. K., and T. Maniatis. 1996. Regulation of interferon-␥-activated STAT1 synthesis. J. Leukocyte Biol. 72:154. by the ubiquitin-proteasome pathway. Science 273:1717. 44. Pearson, G., F. Robinson, T. B. Gibson, B. E. Xu, M. Karandikar, K. Berman, and 22. Chung, C. D., J. Y. Liao, B. Liu, X. P. Rao, P. Jay, P. Berta, and K. Shuai. 1997. M. H. Cobb. 2001. Mitogen-activated protein (MAP) kinase pathways: regulation Specific inhibition of Stat3 signal transduction by PIAS3. Science 278:1803. and physiological functions. Endocr. Rev. 22:153. 23. Liu, B., J. Y. Liao, X. P. Rao, S. A. Kushner, C. D. Chung, D. D. Chang, and 45. Li, J., M. Gorospe, D. Hutter, J. Barnes, S. M. Keyse, and Y. Liu. 2001. Tran- K. Shuai. 1998. Inhibition of Stat1-mediated gene activation by PIAS1. Proc. scriptional induction of MKP-1 in response to stress is associated with histone H3 Natl. Acad. Sci. USA 95:10626. phosphorylation-acetylation. Mol. Cell. Biol. 21:8213. 24. Starr, R., T. A. Willson, E. M. Viney, L. J. Murray, J. R. Rayner, B. J. Jenkins, 46. Bokemeyer, D., A. Sorokin, M. Yan, N. G. Ahn, D. J. Templeton, and T. J. Gonda, W. S. Alexander, D. Metcalf, N. A. Nicola, et al. 1997. A family of M. J. Dunn. 1996. Induction of mitogen-activated protein kinase phosphatase 1

cytokine-inducible inhibitors of signalling. Nature 387:917. by the stress-activated protein kinase signaling pathway but not by extracellular by guest on October 1, 2021 25. Endo, T. A., M. Masuhara, M. Yokouchi, R. Suzuki, H. Sakamoto, K. Mitsui, signal-regulated kinase in fibroblasts. J. Biol. Chem. 271:639. A. Matsumoto, S. Tanimura, M. Ohtsubo, H. Misawa, et al. 1997. A new protein 47. Schliess, F., S. Heinrich, and D. Haussinger. 1998. Hyperosmotic induction of the containing an SH2 domain that inhibits JAK kinases. Nature 387:921. mitogen-activated protein kinase phosphatase MKP-1 in H4IIE rat hepatoma cells 26. Naka, T., M. Narazaki, M. Hirata, T. Matsumoto, S. Minamoto, A. Aono, Arch. Biochem. Biophys. 351:35. N. Nishimoto, T. Kajita, T. Taga, K. Yoshizaki, et al. 1997. Structure and func- 48. Deon, D., S. Ahmed, K. Tai, N. Scaletta, C. Herrero, I. H. Lee, A. Krause, and tion of a new STAT-induced STAT inhibitor. Nature 387:924. L. B. Ivashkiv. 2001. Cross-talk between IL-1 and IL-6 signaling pathways in 27. Fukada, T., M. Hibi, Y. Yamanaka, M. Takahashitezuka, Y. Fujitani, rheumatoid arthritis synovial fibroblasts. J. Immunol. 167:5395. T. Yamaguchi, K. Nakajima, and T. Hirano. 1996. Two signals are necessary for 49. Andus, T., T. Geiger, T. Hirano, T. Kishimoto, and P. C. Heinrich. 1988. Action cell proliferation induced by a cytokine receptor gp130: involvement of STAT3 of recombinant human interleukin 6, interleukin 1␤ and tumor necrosis factor ␣ in anti-apoptosis. Immunity 5:449. on the mRNA induction of acute-phase proteins. Eur. J. Immunol. 18:739. 28. Kim, H. K., T. S. Hawley, R. G. Hawley, and H. Baumann. 1998. Protein tyrosine 50. Zhang, Z., and G. M. Fuller. 2001. Interleukin 1␤ inhibits interleukin 6-mediated phosphatase 2 (SHP2) moderates signaling by gp130 but is not required for the rat ␥ fibrinogen gene expression. Blood 96:3466. induction of acute-phase plasma protein genes in hepatic cell. Mol. Cell. Biol. 51. Bode, J. G., R. Fischer, D. Ha¬ussinger, L. Graeve, P. C. Heinrich, and F. Schaper. 18:1525. ␤ ␣ 2001. The inhibitory effect of IL-1 on IL-6 induced 2-macroglobulin expres- 29. Symes, A., N. Stahl, S. A. Reeves, T. Farruggella, T. Servidei, T. Gearan, sion is due to activation of NF-␬B. J. Immunol. 167:1469. G. Yancopoulos, and J. S. Fink. 1997. The protein tyrosine phosphatase SHP-2 52. Bode, J. G., S. Ludwig, C. A. Correia Freitas, F. Schaper, M. Ruhl, S. Melmed, negatively regulates ciliary neurotrophic factor induction of gene expression. P. C. Heinrich, and D. Ha¬ussinger. 2001. The MKK6/p38 mitogen activated Curr. Biol. 7:697. protein kinase pathway is capable to induce SOCS3 gene expression and inhibits 30. Schaper, F., C. Gendo, M. Eck, J. Schmitz, C. Grimm, D. Anhuf, I. M. Kerr, and IL-6-induced gene induction. Biol. Chem. 382:1447. P. C. Heinrich. 1998. Activation of SHP2 via the IL-6 signal transducing receptor 53. Han, J., J. D. Lee, L. Bibbs, and R. J. Ulevitch. 1994. A MAP kinase targeted by protein gp130 requires JAK1 and limits acute-phase protein expression. Biochem. endotoxin and hyperosmolarity in mammalian cells. Science 265:808. J. 335:557. 54. Takahashi-Tezuka, M., Y. Yoshida, T. Fukada, T. Ohtani, Y. Yamanaka, 31. Bode, J. G., A. Nimmesgern, J. Schmitz, F. Schaper, M. Schmitt, W. Frisch, D. K. Nishida, K. Nakajima, M. Hibi, and T. Hirano. 1998. Gab1 acts as an adapter Ha¬ussinger, P. C. Heinrich, and L. Graeve. 1999. LPS and TNF␣ induce SOCS3 molecule linking the cytokine receptor gp130 to ERK mitogen-activated protein mRNA and inhibit IL-6-induced activation of STAT3 in macrophages. FEBS kinase. Mol. Cell. Biol. 18:4109. Lett. 463:365. 55. Ernst, M., M. Inglese, P. Waring, I. K. Campbell, S. Bao, F. J. Clay, 32. Stoiber, D., P. Kovarik, S. Cohney, J. A. Johnston, P. Steinlein, and T. Decker. W. S. Alexander, I. P. Wicks, D. M. Tarlinton, U. Novak, et al. 2001. Defective 1999. Lipopolysaccharide induces in macrophages the synthesis of the suppressor gp130-mediated signal transducer and activator of transcription (STAT) signaling of cytokine signaling 3 and suppresses signal transduction in response to the results in degenerative joint disease, gastrointestinal ulceration, and failure of activating factor IFN-␥. J. Immunol. 163:2640. uterine implantation. J. Exp. Med. 194:189.