Báo cáo khoa học: TRAF6 and C-SRC induce synergistic AP-1 activation via PI3-kinase–AKT–JNK pathway pptx

TRAF6 and C-SRC induce synergistic AP-1 activation via PI3-kinase–AKT–JNK pathway Megumi Funakoshi-Tago 1 , Kenji Tago 2 , Yoshiko Sonoda 1 , Shin-ichi Tominaga 2 and Tadashi Kasahara 1 1 Department of Biochemistry, Kyoritsu College of Pharmacy, 1-5-30 Shibakoen, Minato-ku, Tokyo 105–8512, Japan; 2 Department of Biochemistry, Jichi Medical School, 3311–1 Minamikawachi-machi, Tochigi-ken, 329–0433, Japan Interleukin-1 (IL-1) induces multiple genes via activation of transcription factors that include NF-jB and activator protein-1 (AP-1). We found that IL-1-mediated c-Src acti- vation was required for AP-1 activation, but not for NF-jB activation and also revealed that c-Src-induced AP-1 acti- vation was enhanced synergistically by the coexpression of TNF receptor associated factor 6 (TRAF6). In addition, c-Src interacts with TRAF6 in response to IL-1 and this interaction is required for c-Src activity. However, neither dominant negative mutants of TRAF6 (TRAF6 DN) nor kinase-dead mutant of c-Src (c-Src KD) counteracted each- induced AP-1 activation, suggesting no hierarchy between these two molecules. During the TRAF6 and c-Src-induced AP-1 activation, phosphatidylinositol 3 (PI3)-kinase, its downstream signaling molecule, Akt and c-Jun N-terminal kinase (JNK) were significantly activated and inhibition of these kinase activities down-regulated AP-1 activation through the suppression of c-fos expression. Furthermore, TRAF6 and c-Src-induced JNK activation was significantly inhibited by PI3-kinase inhibitor or a dominant negative mutant of Akt (Akt DN). Taken together, our results demonstrate that c-Src and TRAF6 are key mediators of IL-1-induced AP-1 activation and provide evidence of cross talk between c-Src and TRAF6 molecules through PI3 kinase–Akt–JNK pathways. Keywords:NF-jB activation; activator protein-1 (AP-1); Src kinase; TRAF6; c-Jun N-terminal kinase (JNK). Interleukin-1 (IL-1) is a potent activator of immune and inflammatory responses that exert various biological acti- vities mostly through rapid and marked activation of transcription factors, nuclear factor-jB(NF-jB) and acti- vator protein-1 (AP-1). Both NF-jB and AP-1 are import- ant regulators of numerous cytokine genes including IL-6, IL-8, monocyte chemotactic protein-1 (MCP-1) and the induction of adhesion molecules [1–3]. Recently, intracel- lular events mediated by IL-1 receptor type I (IL-1RI) have been extensively explored. That is, binding of IL-1 to the IL)1RI receptor allows association with the IL-1R acces- sory protein (IL-1R AcP), that causes the recruitment of MyD88, an adapter protein, to the IL-1RI complex [4–6]. This in turn leads to the recruitment of a serine/threonine kinase, IL-1 receptor-associated kinases (IRAK) to the receptor complex. Then, IRAK interacts with tumor necrosis factor receptor-associated factor 6 (TRAF6) [5–9]. Kinase, TAK1 and two interacting proteins, TAB1 and TAB2, are also implicated in IL-1 signaling [10,11], in which TAK1 activates the kinase, NIK, directly, leading to the activation of NF-jB. IL-1-dependent signaling also lead to the activation of c-Jun N-terminal kinase (JNK) and other mitogen-activated protein kinase [6,9,12–14], that results in the phosphorylation and activation of AP-1. The above model is based mainly on results obtained through the isolation of specific protein complexes, analyses of protein–protein interactions (yeast two-hybrid screens) and data from gene knockout studies. However, the detailed mechanisms of IL-1-dependent signaling pathways are still not well defined. While some investigators have reported the involvement of tyrosine kinases in IL-1 signaling, the involvement of Src family kinases has received little attention [15,16]. c-Src tyrosine kinase is a member of a family of nine closely related tyrosine kinases defined by a common structure that includes SH2 and SH3 protein interaction domains, a catalytic region and a negative-regulatory tyrosine located near the carboxyl terminus [18–20]. Src kinases have been implicated in multiple signaling pathways that regulate cellular growth, migration and cell survival [19,20]. Functional roles of Src family kinases have been well documented in leukocyte signaling processes such as cell migration, adhesion, phagocytosis, cell survival [21–23] or immune response signalings through T cell and B cell receptors [24,25]. It has been demonstrated recently that TRANCE, a TNF family member, activates Akt/PKB in osteoclasts and dendritic cells through a signaling complex involving TRAF6 and c-Src [26]. Src family kinases are presumed to be a prerequisite for survival signaling using the TNF-family receptors, RANKL/TRANCE [27]. In spite of the widespread expression of c-Src, targeted disrup- tion of the c-src gene in mice leads to one predominant Correspondence to T. Kasahara, Department of Biochemistry, Kyoritsu College of Pharmacy, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan. Fax/Tel: +81 3 5400 2697, E-mail: kasahara-td@kyoritsu-ph.ac.jp Abbreviations:NF-jB, nuclear factor-kappaB; AP-1, activator protein-1; TRAF6, TNF receptor associated factor 6; PI3-kinase, phosphatidyl-inositol 3-kinase; JNK, c-Jun N-terminal kinase; DN, dominant negative mutant; IL-1, interleukin-1. (Received 29 September 2002, revised 15 December 2002, accepted 27 January 2003) Eur. J. Biochem. 270, 1257–1268 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03487.x phenotype, osteopetrosis – a failure to break down bone – resulting from an intrinsic defect in osteoclasts, a cell type that expresses high levels of Src [27,28]. IL-1 also regulates cytoskeletal organization in osteoclasts via a TRAF6/c-Scr complex [29]. These findings strongly support the fact that IL-1 signaling is closely related to that by TRAF6 and c-Src. However, downstream targets of c-Src kinase, particularly in IL-1-mediated signaling processes, that lead to NF-jBor AP-1 activation, remain to be defined. We describe here that IL-1 induces marked c-Src kinase activation and that c-Src activity is involved in AP-1 but NF-jB activation. In addition, c-Src-induced AP-1 activa- tion is synergistically augmented by the coexpression of TRAF6. This synergistic AP-1 activation is mediated mostly by the PI3-kinase/Akt/c-Jun N-terminal kinase (JNK) pathways. Experimental procedures Antibodies and reagents Mouse monoclonal antibodies against FLAG-peptides (M2), Myc-peptides (9E10) and HA-peptides were pur- chased from Sigma, Santa Cruz Biotechnology and Roche, respectively. Rabbit Ig against c-Src were purchased from Santa Cruz Biotechnology. Peroxidase-conjugated porcine anti-(rabbit IgG) and peroxidase-conjugated goat anti- (mouse IgG) and anti-(rabbit IgG) were obtained from Dako (Dako-Japan, Tokyo, Japan). Human recombinant IL-1a (IL-1) was kindly provided by Dainippon Pharma- ceutical Co. (Suitashi, Osaka, Japan). Src inhibitor, PP2, JNK inhibitor I, and PI3-kinase inhibitor, LY294002, were purchased from Calbiochem–Novabiochem Co. and Alexis Biochemicals, respectively. SAM68 and histone H1 were obtained from Santa Cruz Biotechnology and SigmaCell Culture. A human glioblastoma cell line, T98G (JCRB, Kamiyoga, Tokyo) and human embryonic kidney cells, HEK293T were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Nissui Seiyaku, Tokyo) supplemented with 10% heat-inactivated fetal bovine serum (Nippon Bio-Supply Center, Tokyo, Japan), 4 m M glutamine, 100 unitsÆmL )1 penicillin G and 100 lgÆmL )1 streptomycin. Construction of plasmid vectors A dominant negative mutant of TRAF6 (TRAF6 DN; mouse where TRAF6 lacks amino acids 1–288) and the full- length cDNA with the N-terminal FLAG sequence were generated from the total RNA of a murine thymoma cell line, EL4, by RT-PCR amplification as described by Cao et al. [7]. PCR products were purified and inserted into the EcoRI and BamHI site of pCMV5. The cDNA encoding mouse c-Src was also generated from the total RNA of EL4 by RT-PCR. A kinase dead mutant of c-Src cDNA (c-Src KD) contained one point mutation (Arg296fiLys). PCR products were inserted into the HindIII site of pCMV5. The cDNA encoding human Csk with the N-terminal FLAG sequence were generated by RT-PCR amplification and was inserted into the EcoRI and BamHI site of pCMV5. Transient transfection and luciferase assay plasmid DNAs were transfected into T98G or HEK293T cells by the calcium phosphate precipitation method. Final amounts of the transfected DNA were adjusted to 10 lgper60-mm dish by addition of empty vector, pCMV5. pCMV5-FLAG- TRAF6, and/or pCMV5-c-Src or pCMV5-c-Src KD (1 lg each) were cotransfected with 0.1 lg of pRL-TK (Promega) and 1 lgofpNF-jB-Luc, pAP-1-Luc, or pIL-8-Luc ()133 to +46 [31,32]). Cells were exposed for 1 day to DNA precipitation in DMEM containing 10% fetal bovine serum. After 48 h of transfection, cells were harvested and the luciferase activities were measured by Lumat LB9501 (Bertold Japan, Tokyo [33,34]). Efficiency of transfection was normalized with sea pansy luciferase activities. Immunoprecipitation and immunoblotting All cells were harvested and lysed in lysis buffer [10 m M Tris/HCl (pH 7.4), 158 m M NaCl, 1% Triton X-100, 1% sodium deoxycholate, 1 m M EGTA, 1 m M Na 3 VO 4 , 2 lgÆmL )1 aprotinin, 2 lgÆmL )1 leupeptin] on ice and cleaned by centrifugation at 13 000 g to obtain whole cell extracts. Aliquots (250 lg) of cell lysate were mixed with protein G-sepharose (Pharmacia-LKB Biotechnologies, Uppsala, Sweden) with each antibody for 2 h at 4 °C. Immune complexes were precipitated by centrifugation at 13 000 g and washed three times with lysis buffer and were boiled in Laemmli sample buffer. Boiled samples were separated by SDS/PAGE and the proteins were transferred to nitrocellulose membranes. Immunoblot and visualization using the enhanced chemiluminescence Western blotting detection system (Amersham) was described elsewhere [33,34]. In vitro kinase assay The immunoprecipitates were washed twice with lysis buffer and three times with kinase buffer (25 m M Hepes/NaOH (pH 7.5), 20 m M MgCl 2, 20 m M b-glycerophosphate, 0.1 m M Na 3 VO 4 ,2m M dithiothreitol, 20 m M p-nitrophe- nylphosphate). The kinase reaction (20 lL of kinase buffer, 10 l M ATP and 0.5 lg SAM68 for c-Src, Histon H1 for Akt, or GST-c-Jun for JNK as substrates) was carried out with [ 32 P]ATP for 15 min at 30 °C. Samples were resolved by 15 or 12% PAGE and phosphorylated SAM68, histone H1 and GST-c-Jun were visualized by autoradio- graphy. RNA isolation and PCR analysis Total RNA separation and RT–PCR analysis were per- formed according to the manufacturer’s protocols (Takara Shuzo, Shiga, Japan) using an oligo(dT) 20-primer and 1 lg total RNA for first strand cDNA synthesis. PCR was performed at an annealing temperature of 57 °Cand20 amplification cycles. The PCR products were resolved and electrophoresed on a 1% agarose gel in Tris/borate/EDTA. The primers used were as follows: human IL-8, 5¢-GAG CCAGGAAGAAACCACCGGA-3¢ (upstream) and 5¢-GCATCTGGCAACCCTACAACAGACC-3¢ (down- stream); human c-Fos, 5¢-CCGGGGATAGCCTCTCT TAC-3¢ (upstream) and 5¢-CTTCTCCTTCAGCAGGT TGG-3¢ (downstream); human c-Jun, 5¢-TTCTATGAC GATGCCCTCAA-3¢ (upstream) and 5¢-GTTGCTGAG GTTTGCGTAGA-3¢ (downstream); human GAPDH, 1258 M. Funakoshi-Tago et al. (Eur. J. Biochem. 270) Ó FEBS 2003 5¢-GTCAGTGGTGGACCTGACCT-3¢ (upstream) and 5¢-TGAGGAGGGGAGATTCAGTG-3¢ (downstream). Electrophoretic mobility shift assay (EMSA) EMSAs were carried out as described previously [31,32]. The consensus double-strand oligodeoxynucleotide probes for AP-1 (Santa Cruz Biotechnology, Inc.) were radio- actively labeled using [c- 32 P]ATP and T4 polynucleotide kinase using standard procedures. Nuclear proteins (10 lg) prepared from cells were incubated with [c- 32 P]labeled double-strand oligonucleotide probe. The binding reaction was carried out at room temperature for 30 min in a total volume of 25 lL. Bound complexes were separated on 5% Tris-Glycine EDTA gel by electrophoresis in Tris-Glycine EDTA buffer, dried and visualized by autoradiography. Results IL-1-induced c-Src activation: c-Src kinase activity is required for IL-1-induced AP-1 but not NF-jB activation The IL-1/IL1R-mediated signaling pathway has been extensively explored and many signaling molecules have been identified so far. However, the cross talk between these signaling molecules is complicated and less well defined. To explore the possible involvement of tyrosine kinase and c-Src in IL-1 signaling, we first explored whether IL-1 induced activation of c-Src. Cell lysates of IL-1-treated T98G cells were assayed for c-Src kinase activity in vitro using SAM68, a known Src-family kinase substrate. As shown in Fig. 1A,B c-Src activity was detected at 15–30 min after IL-1 stimulation and the activity returned to basal levels by 60 min. In addition, most of the IL-1-induced Src activity was abrogated by pretreatment with PP2, a known Src family inhibitor, after 20 min (Fig. 1C). While it has been established that IL-1 rapidly activates the transcription factors, NF-jB and AP-1 [33,34], we explored how c-Src gene expression modulated IL-1- induced NF-jB and AP-1 activation. A luciferase reporter gene assay for NF-jB and AP-1 demonstrated that cells transfected with wild-type c-Src exhibited a substantial increase in NF-jB activation ( 4-fold) almost comparable to that by IL-1 stimulation alone ( sixfold; Fig. 2A). IL-1- induced NF-jB activation was markedly augmented by the transfection of wild-type c-Src ( 14-fold)aswellasbythe kinase-dead mutant of c-Src (c-Src KD) ( 15-fold). While either IL-1 or wild-type c-Src expression alone induced a marginal AP-1 activation, simultaneous stimulation with c-Src and IL-1 enhanced AP-1 activation substantially ( 2-fold) (Fig. 2B). Interestingly, in contrast to its effect on the NF-jB activation, c-Src KD virtually abrogated IL-1- induced AP-1 activation to the unstimulated level. Thus, these results suggested that c-Src kinase is more prominently involved more in AP-1 rather than NF-jB activation in IL-1 signaling. c-Src-induced AP-1 activation is augmented synergistically by the coexpression of TRAF6 As TRAF6 functions as a major signal transducer in IL-1signaling pathways [6–9], we attempted to explore how c-Src affects the TRAF6-induced signaling by coexpression experiments. Therefore, we first determined the effect of c-Src on TRAF6-induced NF-jB activation. T98G cells transfected with TRAF6 alone displayed a 10 to 15-fold increase in NF-jB activation that was augmented additively by the coexpression of c-Src (Fig. 2C). Furthermore, both TRAF6 and IL-1-induced NF-jB activation ( 18-fold) was augmented further by c-Src coexpression ( 32-fold). In contrast, TRAF6 or c-Src-transfection alone, as well as IL-1-stimulation alone, induced minimal levels of AP-1 activation (> twofold; Fig. 2D). IL-1-induced AP-1 acti- vation was augmented additively, though minimally, by TRAF6 or by c-Src. Of note is that cotransfection with both TRAF6 and c-Src that exhibited a dramatic activation of AP-1 ( 13-fold), to a level that was not achieved by the combination of either TRAF6 and IL-1 or c-Src and IL-1. AP-1 activation induced by TRAF6 and c-Src was not further augmented by the addition of IL-1. These data indicated that both TRAF6 and c-Src appear to be essential for the full activation of AP-1. Fig. 1. IL-1 induced c-Src activation. (A) T98G cells were stimulated with IL-1 for the indicated periods and cell lysates were immunopre- cipitated with anti-c-Src antibody. c-Src activity in the immunopre- cipitates was measured by an in vitro kinase assay using GST-SAM68. (B) c-Src activity in (A) was quantified by a BAS 2000 II (Fuji Bio- image analyzer) and fold increase was shown. Results are expressed as mean ± SD of three independent experiments. (C) c-Src activity was measured in the presence of PP2 (20 l M ) at 20 min after IL-1 stimu- lation (C). Ó FEBS 2003 Synergistic activation of AP-1 by TRAF6 and c-Src (Eur. J. Biochem. 270) 1259 TRAF6 interacts with c-Src in an IL-1-dependent manner To investigate the molecular mechanism of TRAF6 and c-Src-induced synergistic AP-1 activation in the IL-1 signaling pathway, we examined whether TRAF6 physically interacted with c-Src. As detected by a coimmunoprecipi- tation assay in Fig. 3A, endogenous c-Src was found to associate with ectopically expressed FLAG-tagged TRAF6 in an IL-1-dependent manner. Similarly, endogenous TRAF6 was coimmunoprecipitated with ectopic c-Src in response to IL-1 stimulation (Fig. 3B). We then examined whether the kinase activity of c-Src is required for the interactionwithTRAF6.FLAG-taggedTRAF6wascoex- pressed with wild-type or kinase-dead mutant of c-Src (c-Src KD) in T98G cells, and their association was tested by detecting c-Src in anti-FLAG immunoprecipitates. We could detect the strong association of TRAF6 with c-Src without IL-1 stimulation as shown in Fig. 3C. We could not detect interaction of TRAF6 with c-Src KD, in spite of the similar expression amounts of these two molecules. These data suggested that the kinase activity of c-Src is essential for the interaction with TRAF6. Synergistic AP-1 activation by TRAF6 and c-Src is induced through c-Fos expression To further delineate the molecular mechanism of synergistic AP-1 activation by Src and TRAF6, we used HEK293T (293T) cells in the subsequent studies, because of its high transfection efficiency. As shown in Fig. 4A, 293T cells transfected with TRAF6 displayed a moderate increase, i.e.  2-foldincrease,inAP-1activationcomparedtothe vector-transfected cells. Similarly, c-Src-transfected 293T cells exhibited a substantial increase in AP-1 activation ( seven- to eightfold). Interestingly, simultaneous expres- sion of TRAF6 and c-Src induced a dramatic increase of AP-1 activity (> 20-fold). This enhanced AP-1 activation was suppressed partially by the cotransfection of Csk, C-terminal Src kinase (data not shown), which inhibits c-Src kinase activity by inducing conformational changes [19,20], suggesting that synergistic AP-1 activation by c-Src and TRAF6 requires c-Src kinase activity. To assess the effect of TRAF6 and c-Src on AP-1 activation further, we measured AP-1 DNA binding by an EMSA assay, using oligo- nucleotide probes for AP-1. Consistent with the results of luciferase assay, we confirmed that synergistic AP-1 activa- tion was induced by coexpression of TRA6 and c-Src (Fig. 4B). This DNA binding activity was not detected using mutant oligonucleotide with a substitution in the AP-1 binding motif. Furthermore, the band of AP-1/DNA complex disappeared by adding a specific antibody against c-Fos to the nuclear extracts (Fig. 4C). Thus, a supershift assay suggested that TRAF6- and c-Src-induced AP-1 DNA binding proteins contain c-Fos but not c-Jun. The above observation was further confirmed at the mRNA level. Namely, either TRAF6 or c-Src transfection alone induced moderate c-Fos mRNA expression. Of note is that the dramatic expression of c-Fos mRNA was induced Fig. 2. Synergistic activation of AP-1 by TRAF6 and c-Src: requirement of c-Src kinase activity for the IL-1-induced AP-1 activation. T98G cells were transiently cotransfected with NF-jB-responsive reporter (pNF-jB-Luc) (A) or AP-1-responsive reporter (p AP-1-Luc) (B), with either an empty vector (none) or plasmids bearing wild-type (WT) or c-Src KD. T98G cells were transiently cotransfected with pNF-jB-Luc (C) or pAP-1-Luc (D), with expression plasmids bearing FLAG-TRAF6 and/or c-Src (1 lg each). Cell lysates were harvested at 48 h after transfection, and luciferase activity was assayed as described in Experimental procedures. Results are expressed as mean ± SD of three independent experiments. 1260 M. Funakoshi-Tago et al. (Eur. J. Biochem. 270) Ó FEBS 2003 by the coexpression of TRAF6 and c-Src (Fig. 4D). In contrast, levels of c-Jun mRNA was changed little by the overexpression of TRAF6 and/or c-Src. Thus, TRAF6 and c-Src appeared to induce synergistic AP-1 activation through enhancement of c-Fos expression. TRAF6 and c-Src induce AP-1 activation independently Then, in order to investigate the mutual dependency of TRAF6 and c-Src in AP-1 activation, we cotransfected c-Src KD with TRAF6 and determined its effect on the TRAF6-induced AP-1 activation. As shown in Fig. 5A c-Src KD did not down-regulate the TRAF6-induced AP-1 activation at all, suggesting that c-Src activity is not required for TRFA6 to induce AP-1 activation. Similarly, TRAF6 DN was cotransfected with c-Src, no reduction in c-Src-induced AP-1 activation was observed as well. This observation was further substantiated by the DNA binding and mRNA levels. Namely, as shown in Fig. 5B, neither c-Src KD nor TRAF6 DN counteracted each- induced AP-1 DNA binding activity. Furthermore, both TRAF6 and c-Src-induced c-Fos mRNA expression were not inhibited by c-Src KD or TRAF6 DN, respectively (Fig. 5C). Thus, the above data suggested that there is no hierarchy between TRAF6 and c-Src in AP-1 activation, and that each works in an independent manner. JNK and PI3-kinase inhibitors down-regulate TRAF6 and c-Src-induced AP-1 activation To further explore the signals involved in the synergistic AP-1 activation by TRAF6 and c-Src, we focused on the MAP kinase pathways involving ERK, JNK and p38 MAPK activation. In addition, involvement of PI3-kinase was also studied as it is known to be involved in TNFa or IL-1 signaling that leads to NF-jB and AP-1 activation [26,33,35]. To date, no direct interplay between c-Src or TRAF6 and PI3-kinase or MAP kinases have been defined. To address this question, we tested several inhibitors for MEK/ERK, JNK, p38MAPK, and PI3- kinase on the TRAF6 and c-Src transfection system. Using a luciferase assay, neither PD98059 (PD) nor SB203580 (SB), that are specific inhibitors for MEK/ERK or p38MAPK, respectively, affected TRAF6- or c-Src- induced AP-1 activation (data not shown). In contrast, a JNK inhibitor I (JI) or LY294002 (LY), a specific PI3- kinase inhibitor, significantly reduced TRAF6- or c-Src- induced AP-1 activation by approximately 50% (data not shown). These two inhibitors attenuated synergistic AP-1 activation by the coexpression of TRAF6 and c-Src to approximately 40% (Fig. 6A). Consistent with Fig. 5A, both JI and LY significantly reduced AP-1 DNA binding activity induced by TRAF6 and c-Src (Fig. 6B). In addition, both JI and LY significantly reduced c-Fos mRNA expression induced by TRAF6 and c-Src, while PD or SB did not (Fig. 6C). These results suggested that c-Src and/or TRAF6- induced AP-1 activation is dependent predominantly on the JNK and PI3-kinase signaling pathways, but not on the ERK or p38MAPK pathways. Synergistic activation of AP-1 by TRAF6 and c-Src via PI3-kinase-Akt pathway As PI3-kinase was presumed to be involved in the TRAF6 and c-Src–induced AP-1 activation as described above, we assumed that Akt/PKB, which is activated via PI3-kinase [26], was also involved in this pathway. To further confirm the direct involvement of Akt, 293T cells with AP-1-Luc were cotransfected with TRAF6 and/or c-Src, as well as with the dominant negative form of Akt Fig. 3. TRAF6 interacts with c-Src in response to IL-1. (A) T98G cells were transiently transfected with expression plasmids bearing FLAG- tagged TRAF6. FLAG-tagged TRAF6 was immunoprecipitated from the whole cell lysates (WCL) using anti-FLAG antibodies. (B) T98G cells were transiently transfected with expression plasmids bearing c-Src. c-Src was immunoprecipitated from the WCL using antic-Src antibodies. (C) T98G cells were transiently cotransfected with expression plasmids bearing FLAG-tagged TRAF6 with either plas- mids bearing wild-type (WT) or c-Src KD. FLAG-tagged TRAF6 was immunoprecipitated from the whole cell lysates using anti-FLAG antibodies. Immunoprecipitates (IP) and WCL were analysed by immunoblot analysis to detect c-Src and TRAF6 as indicated. Ó FEBS 2003 Synergistic activation of AP-1 by TRAF6 and c-Src (Eur. J. Biochem. 270) 1261 (Akt DN). Virtually most of the TRAF6 or c-Src- induced AP-1 activation was abrogated by the increasing doses of Akt DN (data not shown). Similarly, synergistic AP-1 activation by TRAF6 and c-Src was attenuated thoroughly by the Akt DN (Fig. 7A). Inability of TRAF6 and/or c-Src to stimulate AP-1 activation in the presence of the Akt DN was not due to the suppression of TRAF6 and/or c-Src expression vectors, since cell extracts displayed similar protein expression levels of transfected cells (shown in the blots). In addition, synergistic AP-1 DNA binding activity and c-Fos expression induced by TRAF6 and c-Src cotrans- fection was similarly reduced by the presence of Akt DN (Fig. 7B,C). To further delineate how TRAF6 and/or c-Src modulate Akt activation, Akt activity was measured in an in vitro kinase assay using histone H1 as a substrate. Akt expression alone induced slight phosphorylation of histone H1 as with control vectors. When Akt was transfected with TRAF6 or c-Src, significant Akt activity, i.e. Akt-induced histone H1 phosphorylation, was observed (Fig. 8A). Of note is that coexpression of the three constructs, i.e. Akt, TRAF6 and c-Src, resulted in marked Akt activity that was significantly blocked in the presence of LY but not JI (Fig. 8B). These data indicated that the PI3-kinase/Akt-dependent pathway plays a critical role in the synergistic AP-1 activation by the TRAF6 and c-Src. Synergistic activation of JNK by TRAF6 and c-Src: involvement of Akt pathway To confirm the involvement of JNK as shown in Fig. 6, we examined whether expression of TRAF6 and/or c-Src directly modulates JNK activity, by measuring in vitro kinase activity with GST-c-Jun as a substrate. When JNK was cotransfected, either with TRAF6 or c-Src, only moderate phosphorylation of c-Jun was detected. However, coexpression of the three constructs, i.e. TRAF6, c-Src and JNK, resulted in a marked JNK activity (Fig. 9A), which was inhibited by both JI and LY (Fig. 9B). Furthermore, the expression of Akt DN inhibited TRAF6- and c-Src- induced JNK activation, while wild-type Akt enhanced JNK activation. These data in addition to Fig. 6 support the notion that synergistic AP-1 activation induced by coexpression of TRAF6 and c-Src is mediated through a JNK-dependent signaling pathway. Thus, these data suggest that the PI3- kinase/Akt pathway also plays a critical role in TRAF6 and c-Src-induced JNK activation. TRAF6 and c-Src induced marked IL-8 production via JNK and PI3 kinase pathway We finally examined how TRAF6 and c-Src would partici- pate in the IL-1-induced IL-8 expression using an IL-8 Fig. 4. Synergistic AP-1 activation induced by TRAF6 and c-Src through significant expression of c-Fos. (A) 293T cells were transiently cotransfected with pAP-1-Luc and expression plasmids bearing FLAG-TRAF6 and/or c-Src (1 lg each). Cell lysates were harvested at 48 h after transfection and luciferase activity was assayed as described in Experimental procedures. Results are expressed as mean ± SD of three independent experiments. Total protein was isolated from representative transfection experiments, and immunoblot analysis was performed for transfected gene expression. Protein samples were probed with a FLAG-specific antibody for FLAG-tagged TRAF6 and with a c-Src specific antibody for c-Src. (B) 293T cells were transiently cotransfected with expression plasmids bearing FLAG-TRAF6 and/or c-Src (1 lg each). Nuclear extracts were prepared for detection of AP-1 activity using EMSA with radiolabeled probe containing the consensus AP-1 binding site (AP-1 oligonucleotide 5¢-CGCTTGATGACTCAGCCGGAA-3¢, mutant oligonucleotide; AP-1 mut: 5¢-CGCTTGATGACTTGGCCGGAA-3¢ obtained from Santa Cruz). The arrows denote the specific AP-1-DNA complex. (C) For supershift assays, nuclear extracts were incubated in the presence of 2 lgof specific antibodies against each c-Fos and c-Jun. (D) Total RNA were prepared and analysed by RT-PCR. The PCR products were electrophoresed on 2% agarose gels and visualized by ethidium bromide staining. 1262 M. Funakoshi-Tago et al. (Eur. J. Biochem. 270) Ó FEBS 2003 promoter assay. Transfection of either TRAF6 or c-Src alone induced a minimal level of IL-8 promoter activity when compared to IL-1 stimulation (Fig. 10A). Simultaneous expression of both TRAF6 and c-Src induced much higher IL-8 promoter activity (mean 22-fold) than IL-1 stimulation (8-fold). Most prominent activation was obtained by the coexpression of TRAF6 and c-Src with IL-1 stimulation (50-fold). NF-jB and also AP-1 activation by TRAF6 and c-Src are required for optimal and maximal IL-8 promoter activation. IL-8 promoter activity was partially, but signifi- cantly down-regulated by JI, LY, or by Akt DN (Fig. 10B), indicating again that JNK and PI3-kinase-mediated path- ways play substantial roles in the TRAF6 and c-Src-induced IL-8 gene activation. IL-8 mRNA expression by RT-PCR (shown at the bottom panel) confirmed the results obtained by the promoter assay. Fig. 5. Src activates AP-1 independently with TRAF6. (A) 293T cells were transiently cotransfected with pAP-1-Luc and expression plasmids bearing FLAG-TRAF6, TRAF6 DN, and/or c-Src, c-Src KD (1 lg each). Cell lysates were harvested at 48 h after transfection, and luciferase activity was assayed as described in Experimental procedures. Results are expressed as mean ± SD of three independent experiments. (Blots) Total protein was isolated from representative transfection experiments, and immunoblot analysis was performed for the transfected gene expression, by probing a FLAG specific antibody for FLAG-tagged TRAF6 and TRAF6 DN, and by a c-Src specific antibody for c-Src and c-Src KD. (B) 293T cells were transiently cotransfected with expression plasmids bearing FLAG-TRAF6, TRAF6 DN, and/or c-Src, c-Src KD (1 lg each). Nuclear extracts were prepared for detection of AP-1 activity using EMSA. The arrows denote the specific AP-1-DNA complex. (C) Total RNA were prepared and analysed by RT-PCR. The PCR products were electrophoresed on 2% agarose gels and visualized by ethidium bromide staining. Fig. 6. Down-regulation of TRAF6 and c-Src-induced AP-1 activation by JNK and PI3 kinase inhibitors. (A) 293T cells were transiently cotransfected with pAP-1-Luc and expression plasmids bearing FLAG-TRAF6 and/or c-Src (1 lg each). Cells were treated with PD98059 (10 l M ), SB203580 (10 l M ), JNK inhibitor I (10 l M ) and LY294002 (10 l M ) for 6 h. Results are expressed as means ± SD of three independent experiments. (Blots) Total protein was isolated from representative transfection experiments, and immunoblot analysis was performed for the transfected gene expression. Protein samples were probed with a FLAG specific antibody for FLAG-tagged TRAF6, with a c-Src specific antibody for c-Src. (B) 293T cells were transiently cotransfected with expression plasmids bearing FLAG-TRAF6 and c-Src (1 lg each). After the treatment of various inhibitors for 6 h, nuclear extracts were prepared for detection of AP-1 activity using EMSA. The arrows denote the specific AP-1-DNA complex. (C) Total RNA were prepared and analysed by RT-PCR. The PCR products were electrophoresed on 2% agarose gels and visualized by ethidium bromide staining. Ó FEBS 2003 Synergistic activation of AP-1 by TRAF6 and c-Src (Eur. J. Biochem. 270) 1263 Fig. 7. Down-regulation of TRAF6 and c-Src induced activation of Akt by dominant negative mutant of Akt (Akt DN). (A) 293T cells were transiently cotransfected with pAP-1-Luc and expression plasmids bearing FLAG-TRAF6 and c-Src (1 lg each) and varying amounts of Akt DN. Cell lysates were harvested at 48 h after transfection, and luciferase activity was assayed. Results are expressed as means ± SD of three independent experiments. (Blots) Blots were probed with a FLAG specific antibody for FLAG-tagged TRAF6, with a c-Src specific antibody for c-Src, and Myc specific antibody for Myc-Akt DN. (B) 293T cells were transiently cotransfected with expression plasmids bearing FLAG-TRAF6, c-Src and Myc- Akt DN (1 lg each). Nuclear extracts were prepared for detection of AP-1 activity using EMSA. The arrows denote the specific AP-1–DNA complex. (C) Total RNA were prepared and analysed by RT-PCR. The PCR products were electrophoresed on 2% agarose gels and visualized by ethidium bromide staining. Fig. 8. TRAF6 and c-Src induced marked Akt activation. (A) 293T cells were transiently cotransfected with expression plasmids bearing FLAG- TRAF6, and/or c-Src and Myc-Akt (1 lg each). Cells were treated by LY294002 (10 l M ) and JNK inhibitor I (10 l M ) for 6 h (B). Cell lysates were harvested at 48 h after transfection, and the relative activity of Akt in the immunoprecipitates was measured by in vitro kinase assay using with histone H1 as substrate. Blots were probed with a FLAG specific antibody for FLAG-tagged TRAF6, with a c-Src specific antibody for c-Src and with Myc specific antibody for Myc-tagged Akt. 1264 M. Funakoshi-Tago et al. (Eur. J. Biochem. 270) Ó FEBS 2003 These data suggested that TRAF6 and c-Src are signi- ficant signal molecules involved in the AP-1 activation as well as IL-1-induced IL-8 expression, acting particularly through the PI3-kinase/Akt/JNK. Discussion IL-1 plays an important role in immunological and inflammatory reactions by rapidly activating transcription factors, NF-jB and AP-1, that induce various inflamma- tory genes [1–3]. Recently, a number of molecules involved in the IL-1 signaling pathway have been identified [4–11]. While details of the IL-1/IL-1R-mediated signal cascade have been uncovered, the roles of many regulatory and interacting molecules remain to be eluci- dated, particularly those from the tyrosine kinase families. Our initial observation revealed that IL-1-induced IL-8 production was attenuated in the presence of Src family kinase inhibitor, suggesting the involvement of Src family kinases in IL-1 signaling (data not shown). As NF-jBis a master transcription factor involved in the inflammatory responses elicited by IL-1 and is necessary for IL-8 Fig. 9. TRAF6 and c-Src induced marked JNK activation through PI3-kinase-Akt pathway. 293T cells were transiently cotransfected with expression plasmids bearing FLAG-TRAF6, and/or c-Src (A, B), and/or Myc-Akt WT and Akt DN and HA-JNK (1 lg each) (C) Cells were treated with JNK inhibitor I (10 l M ) and LY294002 (10 l M ) for 6 h (B). Cell lysates were harvested at 48 h after transfection, and the relative activity of JNK in the immunoprecipitates was measured by in vitro kinase assay using with GST-c-Jun as substrate. (Blots) Protein extracts were probed with a FLAG specific antibody for FLAG-tagged TRAF6, with c-Src specific antibody for c-Src, with Myc specific antibody for Myc-tagged Akt WT and Akt DN, and with HA specific antibody for HA-tagged JNK. Fig. 10. TRAF6 and c-Src activate IL-8 promoter activity through JNK and PI3-kinase pathways. 293T cells were transiently cotransfected with pIL- 8-Luc ()133 to +46) and FLAG-TRAF6, c-Src (A), and/or Akt DN (1 lg each), in the presence or absence of JNK inhibitor I (10 l M )or LY294002 (10 l M ) (B). Cell lysates were harvested at 48 h after transfection, and luciferase activity was assayed. Results are expressed as means ± SD of three independent experiments. (Bottom) Total RNA were prepared and analysed by RT-PCR. PCR products were electro- phoresed on 1% agarose gels and visualized by ethidium bromide staining. Ó FEBS 2003 Synergistic activation of AP-1 by TRAF6 and c-Src (Eur. J. Biochem. 270) 1265 production [31–34,36], we investigated how c-Src influen- ces the IL-1-induced NF-jB activation. A marked c-Src kinase activity was induced by IL-1, but both wild-type and a mutant of c-Src, the latter lacking its kinase activity, could activate NF-jB in the presence or absence of IL-1. Thus, IL-1-induced c-Src kinase activity per se was not necessarily required for the IL-1-induced NF-jB activation, but c-Src appeared to activate NF-jB directly or to act as a scaffold protein or an adapter protein. A similar observation has been suggested elsewhere; namely, while c-Src deficient mice exhibit severe osteopetrosis (caused by an intrinsic defect in osteoclast function), it is rescued by kinase-deficient c-Src mutants (Y416F, K295M) as well as by a wild-type c-Src [20,26,27]. However, when expressed beyond normal levels, the inactive c-Src kinase mutant fails to rescue c-Src deficient mice by inducing osteoclasts to undergo apoptosis [37]. Therefore, c-Src kinase activity is likely to be important for regulating at least some aspects of osteoclasts (such as cell morphology and survival). In this study, kinase activity of c-Src appeared to be indispensable for IL-1-induced AP-1 activation and bind- ing to TRAF6 (Figs 2 and 3). Furthermore, our surprising finding is that TRAF6 and c-Src augmented AP-1 activation dramatically (Figs 2 and 4). It is also the case that the TNF-related activation-induced cytokine (TRANCE), which plays an essential role in bone resorption and osteoclast differentiation, utilizes TRAF6 and c-Src as important downstream molecules mediating this effect [38,39]. TRANCE activates a serine/threonine kinase, Akt (through TRAF6), that appears capable of binding to and activating c-Src [26]. Activated c-Src could then recruit PI3-kinase, which generates D3-phospho- inositides for Akt activation. In addition, TRANCE and CD40L-mediated Akt activation in dendritic cells was regulated by cytoplasmic adapter molecules, Cbl family proteins [40]. The above observation, however, might be different from the finding in our IL-1 study. That is, although TRAF6 and c-Src could form the complex in an IL-1-dependent manner and lead to significant AP-1 activation, the hierarchy between these two molecules did not exist in IL-1-induced AP-1 activation. Although the kinase activity of c-Src is required for IL-1-induced AP-1 activation, a kinase-dead mutant of c-Src had no effect on TRAF6-induced AP-1 activation (Fig. 5). More- over, a recent study by Zhang et al. [41] indicated that TNF-a markedly accelerates osteoclastogenesis induced by TRANCE, and deletion of TNF receptor 1 (TNFR1) abrogates this response. Enhanced osteoclastogenesis was thus associated with high expression of TNF and TRANCE-signaling mediators, including c-Src, TRAF2, TRAF6 and MEKK-1, and the levels of these signal molecules were notably reduced in TNFR1 knockouts. In addition, MEKK1 is a potential target for TRAF2 and TRAF6, two related signal transducers to TNFa and IL-1, respectively, and is critically required for JNK activation in the study using MEKK1-deficient embryonic stem cells [42–44]. Although the detailed mechanisms of osteoclastogenesis by these molecules is not clarified, these studies also mentioned that c-Src, TRAF2, TRAF6 and MEKK-1 work cooperatively as potent osteoclastogenic factors. That is, TRAF6-c-Src complex leads to NF-kB activation, whereas TRAF2-MEKK1 primarily activates ERKandJNKintheTRANCEandTNFa signaling pathway. Our next interest was the signaling pathways downstream of TRAF6 and c-Src. As demonstrated in Fig. 6, we found that PI3-kinase–Akt pathway and JNK pathway but not ERK and p38 MAP kinase pathways are important for AP-1 activation induced by TRAF6 and c-Src. Further- more, we presented that PI3-kinase inhibitor or Akt DN inhibited TRAF6 and c-Src-induced JNK activation, sug- gesting that Akt activates JNK at downstream of TRAF6 and c-Src. However, a recent study demonstrated that JNK activity could be antagonized by Akt in various cell systems. Kim et al. [45] have reported that Akt1 inhibits JNK activation, where Akt1 interaction with JNK interacting protein 1 (JIP1) inhibited JIP1 binding to specific JNK pathway kinases, such as mixed-lineage kinase (MLK) and JNK. One possible interpretation of the difference in the function of Akt on JNK activation we presume is that JNK activation induced by TRAF6 and c-Src in our study may not be mediated by JIP. Actually, Kim et al. indicated that Akt1 specifically inhibited JNK1 activation induced by JIP1 but not JNK activation induced by expression of MKK7 or by MLK3, thus, indicating that Akt1 specifically inhibits the JIP-mediated JNK activity. On the other hand, Yuan et al. [46] have reported on the negative regulation mechanism of JNK that is required for kinase activity of Akt2. In their system, stress-induced Akt2 interacts with and phosphory- lates IKKa, leading to NF-jB activation. Moreover, Akt2- induced IKKa phosphorylation and NF-jB activation is required for inhibition of JNK activation. Thus, they have proposed that Akt2–NF-jB-upregulated XIAP inhibits the activation of MKK7/JNKK2, upstream of JNK [46]. In our study, a kinase-dead mutant of c-Src was able to induce NF-jB activation (Fig. 2A) but was unable to induce AP-1(Figs 2B and 5A) or Akt activation (data not shown). Therefore, it is presumed that activation of NF-jBbyc-Src is not induced through the activation of Akt. Thus, inhibition of JNK by Akt through the NF-jB pathway does not function in TRAF6 and c-Src-induced AP-1 activation. AP-1 is usually composed of a heterodimer of Jun and Fos proteins. AP-1 is activated by various stimuli such as growth factors, cytokines, neurotransmitters, cell–matrix interactions and bacterial infections as well as by physical and chemical stresses and its function appear complex. AP-1 is thus recognized as an important regulator of cell function and this has been revealed recently by studying mice lacking AP-1 components [36,43,44]. AP-1 components can be activated through direct phosphorylation, i.e. c-Jun is phosphorylated at its N-terminal region by JNK, which in turn is activated in response to growth factors and cytokines [47,48]. An alternative mechanism for c-Jun activation is enhanced transcription of c-Jun mRNA by ERK and p38 MAP kinases [49]. In addition, it is also reported that c-Fos protein expression, that is another typical component of AP-1, is induced by ERK or JNK in growth factor or CD28 signaling pathway, and consequently, AP-1 DNA binding activity as a whole is enhanced [50]. As AP-1 activation induced by the overexpression of TRAF6 and/or c-Src was inhibited more strongly by JNK inhibitor than by PD98059 or SB203580 [Fig. 6], JNK rather than ERK1/2 1266 M. Funakoshi-Tago et al. (Eur. J. Biochem. 270) Ó FEBS 2003 [...]... the novel existence of the synergistic activation of the AP-1 pathway induced by TRAF6 and c-Src, particularly emphasizing the IL-1-induced IL-8 production, it is of great interest to us that the same signaling pathway also exists in the bone resorption and maturation system In this sense, we should add that in our preliminary study, TRAF6 and c-Src induced synergistic AP-1 activation with or without...Ó FEBS 2003 Synergistic activation of AP-1 by TRAF6 and c-Src (Eur J Biochem 270) 1267 or p38MAPK appeared to be more intimately involved in the IL-1-mediated signaling in the 293T and glioblastoma cell line Thus, we monitored whether the expression level of c-Jun and c-Fos mRNA are enhanced by expression of TRAF6 and/ or c-Src In our study, coexpression of TRAF6 and c-Src induced marked c-Fos... which can develop osteoclast-like cells (data not shown) In conclusion, we have described here the finding of a synergistic activation of AP-1 by TRAF6 and c-Src, and have presented data that PI3 kinase-Akt and JNK pathways are responsible for the AP-1 activation AP-1 activation is involved in IL-1-induced IL-8 induction Acknowledgements We would like to express thanks to N Mukaida, Cancer Institute, Kanazawa... presumed that the activation of AP-1by TRAF6 and c-Src was mediated through expression of c-Fos Lakshminarayanan et al [51] have reported on the mechanism of H2O2 induction of IL-8 mRNA which accompanies AP-1 activation with c-Fos and JunD Therefore, it may be reasonable that c-Jun mRNA expression was not changed even in the induction by TRAF6 and c-Src Knockout mice of either TRAF6, c-Src or c-Fos exhibit... RANKL-induced osteoclastogenesis via coupling of TNF type 1 receptor and RANK signaling pathways J Biol Chem 276, 563– 568 Xia, Y., Makris, C., Su, B., Li, E., Yang, J., Nemerow, G.R & Karin, M (2000) MEK kinase 1 is critically required for c-Jun N-terminal kinase activation by proinflammatory stimuli and growth factor-induced cell migration Proc Natl Acad Sci USA 97, 5243–5248 .Karin, M., Liu, Z & Zandi,... (2001) Differential involvement of p38 mitogen-activated protein kinase and phosphatidyl inositol 3-kinase in the IL-1mediated NF-jB and AP-1 activation Int Immunopharmacol 1, 595–604 32 Funakoshi, M., Tago, K., Sonoda, Y., Tominaga, S.-I & Kasahara, T (2001) A MEK inhibitor, PD98059 enhances IL-1induced NF-jB activation by the enhanced and sustained degradation of IjBa Bichem Biochem Res Commun 283, 248–253... substrates and functions of src Biochem Biophys Acta 1287, 121–149 20 Thomas, S.M & Brugge, J.S (1997) Cellular functions regulated by Src family kinases Annu Rev Cell Dev Biol 13, 513–609 21 Volonte, D., Galbiati, F., Pestell, R.G & Lisanti, M.P (2001) Cellular stress induces the tyrosine phosphorylation of caveolin-1 (Tyr (14) via activation of p38 mitogen-activated protein kinase and c-Src kinase... cellular stress- and tumor necrosis factor alpha-induced AKT2 through activation of the NF kappa B pathway in human epithelial Cells J Biol Chem 277, 29973– 29982 Su, B & Karin, M (1996) Mitogen-activated protein kinase cascades and regulation of gene expression Curr Opin Immunol 3, 402–411 Hibi, M., Lin, A., Smeal, T., Minden, A & Karin, M (1993) Identification of an oncoprotein- and UV-responsive... Drab-Weiss, E.A & Roebuck, K.A (1998) H2O2 and tumor necrosis factor-alpha induce differential binding of the redox-responsive transcription factors AP-1 and NF-kappaB to the interleukin-8 promoter in endothelial and epithelial cells J Biol Chem 273, 32670–32678 Wang, Z.Q., Ovitt, C., Grigoriadis, A.E., Mohle-Steinlein, U., Ruther, U & Wagner, E.F (1992) Bone and haematopoietic defects in mice lacking... osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL Proc Natl Acad Sci USA 95, 3597–3602 Arron, J.R., Vologodskaia, M., Wong, B.R., Naramura, M., Kim, N., Gu, H & Choi, Y (2001) A positive regulatory role for Cbl family proteins in tumor necrosis factor-related activation- induced cytokine (trance) and CD40L-mediated Akt activation J Biol Chem 276, 30011–30017 Zhang, Y.H., . expression. TRAF6 and c-Src induce AP-1 activation independently Then, in order to investigate the mutual dependency of TRAF6 and c-Src in AP-1 activation, we cotransfected c-Src KD with TRAF6 and determined. coexpression of TRAF6 and c-Src (Fig. 4D). In contrast, levels of c-Jun mRNA was changed little by the overexpression of TRAF6 and/ or c-Src. Thus, TRAF6 and c-Src appeared to induce synergistic AP-1 activation through. independent manner. JNK and PI3-kinase inhibitors down-regulate TRAF6 and c-Src- induced AP-1 activation To further explore the signals involved in the synergistic AP-1 activation by TRAF6 and c-Src, we focused