Transcriptional Regulation of Heme Oxygenase-1 Gene Expression by ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 278, No. 20, Issue of May 16, pp. 17927–17936, 2003 Printed in U.S.A.

Transcriptional Regulation of Heme Oxygenase-1 Gene Expression by MAP Kinases of the JNK and p38 Pathways in Primary Cultures of Rat Hepatocytes* Received for publication, April 23, 2002, and in revised form, March 11, 2003 Published, JBC Papers in Press, March 11, 2003, DOI 10.1074/jbc.M203929200

Thomas Kietzmann‡§, Anatoly Samoylenko‡, and Stephan Immenschuh¶ From the ‡Institut fu¨r Biochemie und Molekulare Zellbiologie, Georg-August-Universita¨t Go¨ttingen, D-37073 Go¨ttingen, Germany and the ¶Institut fu¨r Klinische Chemie und Pathobiochemie, Justus-Liebig-Universita¨t Giessen, D-35392 Giessen, Germany

Heme oxygenase-1 (HO-1) gene expression is induced by various oxidative stress stimuli including sodium arsenite. Since mitogen-activated protein kinases (MAPKs) are involved in stress signaling we investigated the role of arsenite and MAPKs for HO-1 gene regulation in primary rat hepatocytes. The Jun N-terminal kinase (JNK) inhibitor SP600125 decreased sodium arsenite-mediated induction of HO-1 mRNA expression. HO-1 protein and luciferase activity of reporter gene constructs with ⴚ754 bp of the HO-1 promoter were induced by overexpression of kinases of the JNK pathway and MKK3. By contrast, overexpression of Raf-1 and ERK2 did not affect expression whereas overexpression of p38␣, ␤, and ␦ decreased and p38␥ increased HO-1 expression. Electrophoretic mobility shift assays (EMSA) revealed that a CRE/AP-1 element (ⴚ668/ⴚ654) bound c-Jun, a target of the JNK pathway. Deletion or mutation of the CRE/AP-1 obliterated the JNK- and c-Jundependent up-regulation of luciferase activity. EMSA also showed that an E-box (ⴚ47/ⴚ42) was bound by a putative p38 target c-Max. Mutation of the E-box strongly reduced MKK3, p38 isoform-, and c-Max-dependent effects on luciferase activity. Thus, the HO-1 CRE/AP-1 element mediates HO-1 gene induction via activation of JNK/c-Jun whereas p38 isoforms act through a different mechanism via the E-box.

The microsomal enzyme heme oxygenase (HO,1 EC 1.14.99.3) catalyzes the first and rate-limiting step of heme degradation producing carbon monoxide (CO), Fe2⫹, and biliverdin, which is converted into bilirubin by biliverdin reductase (1). HO-1 is the inducible isoform of HO and is identical to heat shock protein 32 (HSP32) (2), whereas HO-2 and HO-3 are constitutive iso* This study was supported by the Deutsche Forschungsgemeinschaft SFB 402 Teilprojekt A1 and GRK335 (to T. K.) and SFB 402 Teilprojekt A8 and GRK335 (to S. I.). 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 with 18 U.S.C. Section 1734 solely to indicate this fact. § To whom correspondence should be addressed: Institut fu¨r Biochemie und Molekulare Zellbiologie, Humboldtallee 23, D-37073 Go¨ttingen, Germany. Tel.: 49-551-395952; Fax: 49-551-395960; E-mail: [email protected]. 1 The abbreviations used are: HO, heme oxygenase; bHLHzip, basic helix-loop-helix leucine zipper; dnRas, dominant-negative H-Ras; EMSA, electrophoretic mobility shift assay; ERK, extracellular signalregulated kinase; FL, firefly luciferase; JNK, c-Jun N-terminal kinase; Luc, luciferase; MAP, mitogen-activated protein; MAPK, MAP kinase; p38AF, activated form of p38; HA, hemagglutinin; MEK, MAPK/ERK kinase; MEKK, MEK kinase; MKK, MAP kinase kinase; CRE, cAMP response element; CREB, CRE-binding protein. This paper is available on line at http://www.jbc.org

forms (3). HO-1 expression is induced in response to its substrate heme by the second messengers cAMP and cGMP (5), by various stress stimuli such as UV light, heat-shock (6), sulfhydryl reactive reagents (7), heavy metals, anoxia, hypoxia, hyperoxia (8, 9), as well as by inflammatory cytokines such as tumor necrosis factor ␣ (TNF␣), interleukin-1␤ (IL-1␤), or interferon-␥ (IFN␥) (10, 11). The up-regulation of HO-1 expression by these stimuli and their observed toxic effects in various organs of HO-1-deficient mice (12) indicated that the enhancement of HO-1 activity was an adaptive and protective response to cellular stress (11, 13). The transcriptional activation of HO-1 and other genes is mediated by a network of signaling pathways, mostly by modulation of the activities of transcription factors and target gene expression. A central position in the signaling cascades regulating a number of cellular processes such as cell growth, differentiation, stress responses, and apoptosis occupy mitogenactivated protein kinases (MAPKs). Three major subfamilies of MAPKs have been described: extracellular signal-regulated kinases (ERK) (14), c-Jun N-terminal kinases (JNK) (15), and p38 (16, 17). In general, the ERK pathway mediates cellular responses to growth and differentiation factors including platelet-derived growth factor, epidermal growth factor, and IL-5, whereas JNK and p38 kinases are activated by stress-related stimuli, such as heat shock, inflammation, hyperosmolarity, ultraviolet, and ␥ irradiation (18). Arsenic in the form of sodium arsenite, like other heavy metals, is also a potent inducer of heme oxygenase (19); the mechanism of this induction seems to be cell- and species-dependent (20 –22). Arsenite is known to react with thiol groups of different proteins (e.g. receptors) and to be one of the inducers of ERK, JNK, and p38 cascades (23, 24). This appears to be mediated by the activation of Shc due to tyrosine phosphorylation (25) and recruitment of the Grb2-Sos complex which can activate the p21 small GTPases of the Rho family Ras and Rac. Ras, in turn, activates either the ERK pathway or via action on PAK the JNK or p38 pathway. Rac activation can lead again to PAK activation and subsequently to the activation of JNK and p38 (26). In liver cells the only detailed report so far regarding the role of MAPKs in HO-1 gene regulation had demonstrated that in the LMH chicken hepatoma cell line sodium arsenite-dependent induction of HO-1 gene expression was mediated by the ERK and p38 pathways (21). Despite the extensive characterization of the MAPK signal transduction pathways and the nearly universal induction of HO-1 expression by stress-related stimuli, conflicting data have been reported as to the involvement of MAPKs pathways in HO-1 gene activation in various tumor cell lines (20, 21, 27–29). Moreover, little is known about

17927

17928

Rat HO-1 Gene Expression Is Regulated by MAP Kinases HO-1 gene expression in response to the activation by arsenite and MAPKs in primary cultured cells. Thus, it was the goal of the present study to investigate the regulation of rat HO-1 gene expression by MAPKs using primary rat hepatocyte cultures as a model system. In our study basal levels of HO-1 mRNA expression as well as sodium arsenite-induced HO-1 mRNA levels were down-regulated in the presence of JNK inhibitor SP600125. It was demonstrated that the Ras pathway via JNK induced HO-1 gene expression while the Raf pathway via ERK was not involved in the regulation of HO-1 gene expression. The promoter sequence ⫺668/⫺654 of the rat HO-1 gene was shown to be responsible for HO-1 activation by JNK and Jun. The E-box element (⫺47/⫺42) was involved in the inhibition of HO-1 gene expression via an MKK3-independent mechanism by p38 and Max. EXPERIMENTAL PROCEDURES

FIG. 1. Involvement of MAP kinases in the induction of rat HO-1 gene expression by sodium arsenite. Primary hepatocytes were cultured for 42 h. 8 h before harvesting, the cells were treated with 5 ␮M sodium arsenite (NaAsO4) and with inhibitors of the MEK/ERK pathway PD98059 (5 ␮M) and U0126 (10 ␮M), of the JNK pathway SP600125 (25 ␮M) and of the p38 pathway SB203580 (20 ␮M). A, HO-1 mRNA levels were measured by Northern blotting. For Northern analysis 20 ␮g of total RNA prepared from the cultured hepatocytes were hybridized to digoxigenin-labeled HO-1 and ␤-actin antisense RNA probes (see “Experimental Procedures”). Autoradiographic signals were obtained by chemiluminescence and scanned by videodensitometry. The mRNA level in untreated hepatocytes was set equal to 100%. Values are means ⫾ S.E. of three independent culture experiments. Statistics, Student’s t test for paired values: *, significant difference sodium arsenite versus control; **, significant difference SP600125 and arsenite versus control and arsenite, p ⱕ 0.05. B, representative Northern blots. C, inhibition of arsenite-induced (A) phosphorylation of ERK, Jun, and

Materials—All biochemicals and enzymes were of analytical grade and were purchased from commercial suppliers. Cell Culture—Cells from male Wistar rats (200 –250 g) were used for culture experiments. Hepatocytes were isolated by collagenase perfusion (30). About 1 ⫻ 106 cells were cultured at 37 °C on 6-cm Falcon culture dishes under air/CO2 (19:5) in medium 199 with Earle’s salts containing bovine serum albumin (2 g/liter), NaHCO3 (20 mM), Hepes (10 mM), streptomycin sulfate (117 mg/liter), penicillin (60 mg/liter), insulin (1 nM), and dexamethasone (10 nM). Fetal calf serum (5%) was present in the initial 5 h after which cultures were incubated in serumfree medium for another 18 h. Then, medium was changed again, and the cells were further cultured in serum-free medium for 24 h. Hepatocytes were treated with 5 ␮M sodium arsenite 8 h before harvesting. The specific inhibitors of MEK PD98059 (New England BioLabs) (5 ␮M) and U0126 (Cell Signaling) (10 ␮M), of JNK SP600125 (Biomol) (25 ␮M) and of p38 SB203580 (Calbiochem) (20 ␮M) were added to the culture medium also 8 h before harvesting. Plasmid Constructs—The luciferase reporter gene constructs pHO754 and pHO-754⌬A were previously described (5). The luciferase reporter gene construct pHO-347 was generated by deletion of the ⫺754/ ⫺347 fragment of HO-754 by ApaI and KpnI followed by blunting of the remaining vector with Klenow enzyme and ligation. The luciferase construct pHO-754Em was generated with the QuickChangeTM XL site-directed mutagenesis kit (Stratagene) using the oligonucleotide 5⬘-GGCTCAGCTGGGCGGCCACctctagACTCGAGTAC-3⬘. The luciferase constructs p(CRE/AP-1)3-SV40Luc and p(CRE/AP-1)6-SV40Luc were generated by ligation of the oligonucleotide 5⬘-GATCCTGACTTCAGTCTGAATTCCTGACTTCAGTCTGACTTCAGTC-3⬘ containing 3 CRE/AP-1 elements into the BglII site of pGl3 prom (pSV40Luc) (Promega). Expression vectors for MAP kinase signaling pathway components and transcription factors have been described: ERK2 (31), MEKK1 (32), JNK1 (33), MKK3, dominant-negative MKK3, dominantnegative MKK4 (34), p38 isoforms, and dominant-negative p38 isoforms (16, 35–37), H-Ras (38), dominant-negative H-Ras (dnRas) (38), Raf-1 (39), c-Jun (40), c-Myc (41). RNA Isolation, Northern Blot Analysis, and Hybridization—Isolation of total RNA and Northern analysis were performed as described (42). Digoxigenin (DIG)-labeled antisense RNAs served as hybridization probes; they were generated as described (43) by in vitro transcription from pBS-HO-1 (800-bp cDNA fragment) using T3 RNA polymerase or from pBS-␤ actin (550-bp cDNA fragment) using T7 RNA polymerase and RNA labeling mixture containing 3.5 mM 11-DIG-UTP, 6.5 mM UTP, 10 mM GTP, 10 mM CTP, 10 mM ATP. Hybridizations and detections were carried out essentially as described before (42). Blots were quantified with a videodensitometer (Biotech Fischer, Reiskirchen). Western Blot Analysis—Western blot analysis was carried out as described (5). In brief, total protein from primary cultured hepatocytes was prepared and the protein content was determined using the Bradford method. 50 ␮g of protein were loaded onto a 10% SDS-polyacrylamide gel and after electrophoresis blotted onto nitrocellulose mem-

HSP27 by specific inhibitors PD98059, U0126, SP600125, and SB203580 detected by Western blotting with antibodies against phospho-ERK, phospho-Jun, and phospho-HSP27 (see “Experimental Procedures”). D, activation of the JNK and p38 pathway by overexpression of MEKK1 as detected by Western blotting with antibodies against phospho-Jun and phospho-p38 (see “Experimental Procedures”).

Rat HO-1 Gene Expression Is Regulated by MAP Kinases branes. The primary antibodies against rat HO-1 (Stressgen), rat HO-2 (Stressgen), phospho-c-Jun (Ser63) (Cell Signaling), phospho-HSP27 (Ser82) (Cell Signaling), phospho-p38 (Cell Signaling), H-Ras (Santa Cruz), and c-Fos (Santa Cruz) were rabbit and used at 1:1000 dilution. The primary monoclonal mouse antibody against ␤-actin (Sigma) was used at 1:5000 dilution, the primary mouse antibody against phosphop44/42 MAPK (Thr-202/Tyr-204) (Cell Signaling) was used at 1:2000 dilution, the primary mouse antibody against c-Jun (Santa Cruz Biotechnology), the primary monoclonal mouse antibody against HA tag (Cell Signaling) and primary monoclonal mouse antibody against FLAG M2 (Sigma) were used at 1:1000 dilution. The secondary antibodies were goat anti-rabbit IgG (Bio-Rad) and goat anti-mouse IgG (Bio-Rad) and used at 1:2000 dilution. The ECL Western blotting system (Amersham Biosciences) was used for detection. HO-1 was visible as a 32-kDa band, HO-2 as a 36-kDa band, ␤-actin as a 42-kDa band, H-Ras as a 21-kDa band, c-Fos as a 55-kDa band, c-Jun as a 39-kDa band, phosphoHSP27 as a 27-kDa band, phospho-ERK was seen as a double band of 42 and 44 kDa. HA-MEKK1 was seen as 186-kDa band, HA-ERK as a 42-kDa band, FLAG-Raf as a 68-kDa band, FLAG-MKK3 as a 40-kDa band, FLAG-JNK1 as a 56-kDa band, and FLAG-p38 as a 38-kDa band. Cell Transfection and Luciferase Assay—Freshly isolated rat hepatocytes (about 1 ⫻ 106 cells per dish) were transfected as described (5). In brief, rat hepatocyte cultures (about 1 ⫻ 106 cells per dish) were transiently transfected with 2.5 ␮g of plasmid DNA containing 500 ng of the Renilla luciferase construct pRL-SV40 (Promega) to control transfection efficiency and 2 ␮g of the appropriate HO-1 promoter Firefly luciferase (FL) construct. In every culture experiment two cultures were transfected per point. The DNA was transfected as a calcium phosphate precipitate (5) for 5 h. After removal of the medium cells were cultured under standard conditions without serum. 24 h later cells were treated with fresh medium and were cultured for another 12 h. 12 h before harvesting the cells were treated with PD98059 (5 ␮M), U0126 (10 ␮M), SP600125 (25 ␮M), SB203580 (20 ␮M), or LY294002 (Cell Signaling) (10 ␮M), as indicated. The Luc activity of 20 ␮l of cell lysate was recorded in a luminometer (Berthold) using the dual luciferase assay kit (Promega). Preparation of Nuclear Extracts and Electrophoretic Mobility Shift Assay (EMSA)—Nuclear extracts were prepared essentially as described (44). The sequences of the HO-1 oligonucleotides used for the EMSA are 5⬘-TGTGTCAGAGCCATGTGTCCTGACTTCAGTCT-3⬘ (spanning the CRE/AP-1 (⫺664/⫺657), Fig. 4) and 5⬘-GGCTCAGCTGGGCGGCCACCACGTGACTCGAGTAC-3⬘ (spanning the E-box (⫺47/ ⫺42), Fig. 6). Equal amounts of complementary oligonucleotides were annealed and labeled by 5⬘ end-labeling with [␥-32P]ATP (Amersham Biosciences) and T4 polynucleotide kinase (MBI). They were purified with the Nucleotide Removal kit (Qiagen). Binding reactions were carried out in a total volume of 20 ␮l containing 50 mM KCl, 1 mM MgCl2, 1 mM EDTA, 5% glycerol, 10 ␮g of nuclear extract, 250 ng of poly(d(I-C)), and 5 mM DTE. For competition analyses a 50-fold molar excess of unlabeled AP1 consensus oligonucleotide (Promega) was added. After preincubation for 5 min at room temperature, 1 ␮l of the labeled probe (104 cpm) was added, and the incubation was continued for an additional 10 min. For supershift analysis 1 ␮l of the ATF/ CREB-1 (24H4B), c-JunC (epitope corresponding to DNA binding domain), c-JunN (epitope mapping within the N-terminal domain), c-Fos (K-25), Myc (C33), Max (C17) or SP-1 (PEP2-G) antibody (all obtained from Santa Cruz Biotechnology) as well as a rabbit preimmune serum was added to the EMSA reaction, which was then incubated at 4 °C for 2 h. The electrophoresis was then performed with a 5% non-denaturing polyacrylamide gel in TBE buffer (89 mM Tris, 89 mM boric acid, 5 mM EDTA) at 200 V. After electrophoresis the gels were dried and exposed to a phosphorimager screen. RESULTS

Involvement of the JNK Pathway in the Sodium Arsenite-dependent HO-1 mRNA Expression—To investigate whether the sodium arsenite-dependent HO-1 induction occurs via the activation of various MAPK signaling pathways primary hepatocytes were treated with sodium arsenite in the presence of specific inhibitors of the ERK, JNK, or p38 pathways. 5 ␮M sodium arsenite induced HO-1 mRNA expression in primary rat hepatocytes after 8 h of treatment by about 8-fold (Fig. 1, A and B), in line with a previous study (45). The inhibitors of the ERK pathway PD98059 and U0126 as well as the p38 pathway inhibitor SB203580 did not change either basal or arseniteenhanced HO-1 mRNA levels. In contrast, the JNK inhibitor

17929

FIG. 2. Regulation of rat HO-1 protein expression by overexpression of H-Ras and components of MAPK pathways. Primary hepatocytes were transfected with 8 ␮g of Ras, Raf, MKK3, ERK2, MEKK1, JNK1, or p38␣ expression vectors or the control vector pCMV. After 5 h the medium was changed, and cells were cultured for 43 h. A, HO-1 protein levels were measured by Western blotting with an antibody against rat HO-1 (see “Experimental Procedures”). Autoradiographic signals were obtained by chemiluminescence and scanned by videodensitometry. The protein level in hepatocytes transfected with the control pCMV vector was set equal to 100%. Values are means ⫾ S.E. of three independent culture experiments. Statistics, Student’s t test for paired values: *, significant difference Ras versus control, MKK3 versus control, MEKK1 versus control, JNK1 versus control, p38␣ versus control, p ⱕ 0.05. B, HO-1, HO-2, ␤-actin, Ras, HA-ERK2, HA-MEKK1, FLAG-Raf, FLAG-MKK3, FLAG-JNK1, FLAG-p38 proteins were detected by Western blotting (see “Experimental Procedures”).

SP600125 attenuated the sodium arsenite-dependent HO-1 mRNA induction. Interestingly, combination of the JNK and p38 inhibitors led to a more pronounced inhibition of the

17930

Rat HO-1 Gene Expression Is Regulated by MAP Kinases

FIG. 3. Induction of HO-1 promoter controlled luciferase expression by overexpression of Ras and MEKK1. A, hepatocytes were transiently cotransfected with either Ras, dominant-negative Ras (dnRas), Raf, MEKK1, MKK3, or dominant-negative MKK3 (MKK3dn) and dominant-negative MKK4 (MKK4dn) expression plasmids and luciferase (Luc) gene constructs driven by wild-type 754 and 347 bp of the rat HO-1 promoter (pHO-754 and pHO-347) or the 754-bp promoter mutated at the previously described (5) CRE/AP-1 site (pHO-754⌬A). In control experiments Luc constructs were cotransfected with pCMV plasmid. In each experiment the fold induction of Luc activity was determined relative to the pHO-754, pHO-347, or pHO-754⌬A controls, which were set equal to 1. In pHO-754⌬A the wild-type HO-1 sequence is shown on the upper strand, deleted bases are indicated by ⫺ and mutated bases are shown in lowercase letters and indicated by *. Values represent means ⫾ S.E. of three independent experiments. Statistics, Student’s t test for paired values: *, significant difference pHO-754⫹Ras versus pHO-754 control, pHO-754⫹MEKK1 versus pHO-754 control, pHO-754⫹MKK3 versus pHO-754 control, pHO-754⌬A⫹Ras versus pHO-754⌬A control, pHO-754⌬A⫹MEKK1 versus pHO-754⌬A control, pHO-754⌬A⫹MKK3 versus pHO-754⌬A control, pHO-347⫹Ras versus pHO-347 control, pHO-347⫹MEKK1 versus pHO-347 control, pHO347⫹MKK3 versus pHO-347 control, p ⱕ 0.05. B, hepatocytes were transiently cotransfected with either MEKK1 expression plasmid or the empty control vector and luciferase (Luc) gene construct pHO-347. 12 h before harvesting the cells were treated with PD98059 (5 ␮M), U0126 (10 ␮M), SP600125 (25 ␮M), SB203580 (20 ␮M), or LY294002 (10 ␮M). In each experiment the fold induction of Luc activity was determined relative to the pHO-347 control, which was set equal to 1. Values represent means ⫾ S.E. of three independent experiments. Statistics, Student’s t test for paired values: *, significant difference pHO-347⫹SB203580 versus pHO-347 control, pHO-347⫹MEKK1⫹SB203580 versus pHO-347⫹MEKK1.

arsenite-dependent HO-1 mRNA induction pointing to the involvement of p38, which was overridden by the JNK pathway. The effectiveness of the kinase inhibitors at the concentrations

used was demonstrated by Western blot analysis with antibodies against phospho-ERK (for ERK pathway), phospho-Jun63 (for JNK pathway), and phospho-HSP27 (for p38 pathway)


17931

FIG. 4. Mutation of the CRE/AP1 site abolished the JNK-dependent induction of HO-1 promoter-controlled luciferase expression. Hepatocytes were transiently cotransfected with either ERK2 or JNK1 expression plasmids and Luc gene constructs driven by wild-type 754 and 347 bp of the rat HO-1 promoter (pHO-754 and pHO-347) or the 754-bp promoter mutated at the CRE/AP-1 site (pHO-754⌬A). In control experiments Luc constructs were cotransfected with pCMV plasmid. Tratment with the JNK inhibitor SP600125 (25 ␮M) was performed as in Fig. 3. In each experiment the fold induction of Luc activity was determined relative to the pHO-754, pHO-347, or pHO-754⌬A controls, which were set equal to 1. In pHO-754⌬A the wild-type HO-1 sequence is shown on the upper strand, deleted bases are indicated by ⫺, and mutated bases are shown in lowercase letters and are indicated by *. The values represent means ⫾ S.E. of three independent experiments. Statistics, Student’s t test for paired values: *, significant difference pHO-754⫹JNK1 versus pHO-754 control, p ⱕ 0.05.

(Fig. 1C). These results indicated that kinases of the JNK and p38 pathways were involved in the regulation of HO-1 expression by sodium arsenite. Since the JNK and p38 pathways can be activated by MEKK1, the question of whether overexpression of MEKK1 activates these pathways in hepatocytes was examined. Indeed, MEKK1 overexpression led to the activation of the JNK and p38 pathway as demonstrated by the phosphorylation of cJun and p38 (Fig. 1D). Regulation of Transfected HO-1 Gene Promoter Constructs by Overexpression of MAP Kinases in Primary Rat Hepatocytes—To further investigate the regulatory role of MAP kinases for HO-1 gene expression primary cultured rat hepatocytes were transfected with expression vectors for human H-Ras, Raf-1, MEKK1, MKK3, ERK2, JNK1, and p38. Overexpression of these proteins was confirmed by Western blot analysis. Hepatocytes transfected with H-Ras, MEKK1, JNK1, or MKK3 expression vectors showed enhanced HO-1 protein levels. By contrast, transfection with Raf-1 as well as ERK2 had no effect on the HO-1 expression while transfection with p38 led to a significant decrease of HO-1 protein levels (Fig. 2). The observed pattern appeared to be specific for HO-1 since neither construct changed expression of HO-2 or actin. To examine the molecular mechanism of HO-1 gene regulation by MAPKs luciferase reporter gene constructs from the rat HO-1 promoter (46) (pHO-754, pHO-750⌬A (5), and pHO-347) were cotransfected with expression vectors encoding various components of the MAPK signaling pathways (Fig. 3). Luciferase expression of pHO-754 was up-regulated by constitutive active H-Ras, MEKK1, and MKK3 but not by Raf-1 or dominant-negative Ras (dnRas). After cotransfection of both H-Ras and dnRas expression vectors the stimulatory effect of H-Ras was attenuated by dnRas (Fig. 3). Cotransfection of MEKK1 and dominant-negative mutants of MKK3 and MKK4 downregulated HO-1 promoter activity. The up-regulation of Luc activity of pHO-754⌬A and pHO-347, both of which lack the previously described HO-1 CRE/AP-1 element (5), by overexpressed MEKK1 was reduced compared with pHO-754 (Fig. 3A). Furthermore, the dominant-negative MKK3 still abolished Luc activity whereas the dominant-negative MKK4 did not. The same pattern was seen with pHO-347. This indicated that the JNK pathway acted primarily through the CRE/AP-1 ele-

ment, and the p38 pathway via an element inside the first ⫺347 bases of the HO-1 promoter. However, the remaining induction of pHO-347 by MEKK1 was not attenuated by the inhibitors of either ERK, JNK, p38, or PI3K/PKB because neither PD98059, U0126, SP600125, SB203580, nor the phosphatidylinositol 3-kinase inhibitor LY294002 down-regulated MEKK1-dependent enhancement of Luc activity. By contrast, the inhibitor of the p38 pathway SB203580 led to an induction of Luc activity. Because MEKK1 activates the JNK pathway, which is known to have the transcription factor AP-1 as a nuclear target (47, 48) the regulation of HO-1 gene promoter constructs was investigated in the presence of overexpressed JNK1, and for comparison also ERK2. JNK1 enhanced Luc activity of pHO754, which was abolished in the presence of the JNK inhibitor SP600125. A JNK-dependent induction was not observed with pHO-754⌬A or pHO-347. By contrast, ERK2 did not affect Luc expression of the transfected HO-1 reporter gene constructs (Fig. 4). Taken together, these results indicate that activation of the JNK pathway may induce HO-1 gene expression through the HO-1 CRE/AP-1 site. Role of the HO-1 CRE/AP-1 Element in the Activation of HO-1 Promoter Luc Constructs by c-Jun—Since the transcription factor Jun is a substrate of JNK it was investigated whether c-Jun can bind directly to the CRE/AP-1 element of the HO-1 promoter. EMSA studies with nuclear extracts of cultured rat hepatocytes and an oligonucleotide containing the HO-1 CRE/ AP-1 site formed a strong protein-DNA complex. The intensity of this complex was reduced by an unlabeled AP-1 consensus oligonucleotide added in a 50-fold molar excess to the binding reaction (Fig. 5). To investigate the presence of AP-1 and members of the ATF/CREB family in this complex antibodies against c-Jun, c-Fos, and ATF/CREB were included in the binding reactions. In the presence of the ATF/CREB antibody a slightly supershifted complex was observed (Fig. 5). In addition, the binding of the protein complex to the CRE/AP-1 nucleotide was enhanced. Addition of antibodies against c-Fos, as well as against the GC-box binding factor SP-1 or a rabbit preimmune serum, did not result in a supershift or inhibition of complex forma-

17932


FIG. 5. Binding of c-Jun to the CRE/AP-1 site of the rat HO-1 promoter. A, The CRE (79) and AP-1 (69, 70) consensus sequences and the rat HO-1 promoter sequence ⫺666/⫺655 containing the CRE/AP1 site are shown. Bases matching the consensus sequences are in bold. B, EMSA: the 32P-labeled HO-1 CRE/AP1 oligonucleotide (see “Experimental Procedures”) was incubated with 10 ␮g of protein of nuclear extracts from primary hepatocytes. In EMSA with antibodies the nuclear extracts were preincubated with 1 ␮l of the JunC, JunN, Fos, SP-1, ATF/CREB antibodies or a rabbit preimmune serum for 2 h at 4 °C before adding the labeled probe. For competition analyses a 50-fold molar excess of unlabeled AP1 consensus oligonucleotide (Promega) was added. The DNA-protein binding was analyzed by electrophoresis on 5% native polyacrylamide gels. C, constitutive ATF/Jun-containing complex; S, supershift.

tion. The Jun antibody, which was raised against the highly conserved DNA binding domain of c-Jun (c-JunC) abolished DNA-protein complex formation. Addition of the c-JunN antibody generated against the N-terminal domain formed a strong supershifted complex, indicating that the CRE/AP-1 site was bound mostly by c-Jun. To verify the functional relevance of c-Jun binding to the HO-1 CRE/AP-1 element the pHO-754 and pHO-754⌬A luciferase constructs were cotransfected with expression vectors for c-Jun and c-Fos. Overexpression of both c-Jun and c-Fos was confirmed by Western blot analysis. Overexpression of c-Jun, but not c-Fos, strongly induced luciferase activity of pHO754 (Fig. 6A). No induction was observed for the pHO-754⌬A construct in the presence of overexpressed c-Jun and c-Fos. Similarly, c-Jun but not c-Fos induced luciferase activity of SV40 promoter luciferase gene constructs containing 6 (p(CRE/AP1)6-SV40Luc) or 3 (p(CRE/AP-1)3-SV40Luc) copies of the CRE/ AP-1 element (Fig. 6B). Cotransfection of c-Jun did not induce luciferase activity of the control pSV40Luc vector. Thus, it

appears that c-Jun is the major protein that interacts with the CRE/AP-1 element. Inhibition of HO-1-driven Luciferase Activity by p38 and c-Max Is Abolished by Mutation of the HO-1 E-box Element— Induction of HO-1 protein expression by overexpression of MKK3 and its repression by p38 suggested the presence of response elements for the p38 pathway within the HO-1 promoter. Candidate transcription factors mediating the activation by p38 are Myc and Max (49). These basic helix-loop-helix leucine zipper (bHLHzip) proteins (50) are known to bind to E-box elements one of which is also found in the rat HO-1 promoter (⫺47/⫺42) (51). The potential binding of nuclear proteins to oligonucleotide probes spanning the E-box sequence of the rat HO-1 promoter was examined by EMSA studies. The oligonucleotide ⫺66/⫺32 containing the E-box formed two protein complexes, which were no longer detectable with a mutated oligonucleotide (Fig. 7). To investigate the presence of bHLHzip factors in these complexes antibodies against Myc and Max were included in the binding reaction. Addition of Max antibody led to formation of a supershifted complex. Surprisingly, incubation of the E-box oligonucleotide with an antibody against Myc did not result in a supershift or inhibition of the complex indicating Max homodimer formation. To ensure specificity of the supershift mediated by the Max antibody the EMSA was also performed in the presence of an antibody against the GC-box binding factor SP-1. The SP-1 antibody did not affect the complex formation with the E- box oligonucleotide. To investigate the functional role of the E-box in the regulation of HO-1 gene expression by MAPK the wild-type pHO-347 and the construct pHO-347Em, in which 3 of 6 bp of the E-box (⫺47/⫺42) were mutated, were transfected along with expression vectors for MKK3, different isoforms of wild-type p38 (p38␣, ␤, ␥, ␦) or their respective dominant-negative forms (p38␣, ␤, ␥, ␦, AF) and the p38 target Max. Overexpression of MKK3 enhanced pHO-347-dependent Luc activity, and this enhancement was reduced in the pHO-347Em-transfected cells. Transfection of pHO-347 with vectors for p38␣, p38␤, and p38␦ as well as Max inhibited Luc activity whereas p38␥ enhanced Luc activity. The effects of the different p38 isoforms were reversed by use of the respective dominant-negative mutants p38␣AF, p38␤AF, p38 ␥AF, p38␦AF. Furthermore, when pHO-347Em was used neither a repression nor an induction of Luc activity was observed after cotransfection of the vectors for the different p38 isoforms and their mutants (Fig. 8). These data indicate that the HO-1 E-box is responsible for the p38␣, p38␤, and p38␦-dependent down-regulation and the p38␥-mediated induction of HO-1 expression. DISCUSSION

The present investigation has shown that kinases of the JNK pathway are involved in the induction of rat HO-1 gene expression by sodium arsenite in primary cultured rat hepatocytes. Rat HO-1 gene expression is induced by Ras via MAPKs through the JNK (MEKK1, JNK) pathway but not via MAPKs of the ERK (Raf, ERK2) pathway. JNK activation was mediated via a CRE/AP-1 element (⫺668/⫺654) which was bound mainly by c-Jun. Furthermore, different roles of the p38 kinases could be demonstrated because MKK3 acted as an inducer of HO-1, and the specific inhibitor of p38 pathway alone did not influence HO-1 induction by sodium arsenite. The MKK3dependent induction was partially mediated by p38␥ via an E-box element (⫺47/⫺42), which also appeared to be the target for direct inhibition of HO-1 expression by the p38 kinasese ␣, ␤, and ␦. The target site for p38 was bound by Max. Overexpression of Max, like the p38 ␣, ␤, and ␦ isoforms, inhibited HO-1 expression.


17933

FIG. 6. Overexpression of c-Jun induced luciferase expression controlled by either 754 bp of the HO-1 promoter or by 6 and 3 copies of the HO-1 CRE/AP1 element. A, hepatocytes were transiently cotransfected with either c-Jun, c-Fos, or both expression plasmids and Luc gene constructs driven by the rat HO-1 promoter (pHO-754 and pHO-754⌬A) or by the SV40 promoter regulated by 6 or 3 copies of the HO-1 CRE/AP1 elements (p(CRE/AP-1)6-SV40Luc and p(CRE/AP-1)3-SV40Luc). In control experiments Luc constructs were cotransfected with pCMV plasmid and pSV40Luc construct was cotransfected with c-Jun, c-Fos, or both expression vectors. In each experiment the fold induction of Luc activity was determined relative to the pHO-754, pHO-754⌬A, p(CRE/AP-1)6-SV40Luc, p(CRE/AP-1)3-SV40Luc, or pSV40Luc controls, which were set equal to 1. In pHO-754⌬A the wild-type HO-1 sequence is shown on the upper strand, deleted bases are indicated by ⫺, and mutated bases are shown in lowercase letters and are indicated by *. The values represent means ⫾ S.E. of three independent experiments. Statistics, Student’s t test for paired values: *, significant difference pHO-754⫹c-Fos versus pHO-754 control, pHO-754⫹c-Jun⫹c-Fos versus pHO-754 control, p(CRE/AP1)6-SV40Luc⫹c-Jun versus p(CRE/AP-1)6-SV40Luc control, p(CRE/AP-1)6-SV40Luc⫹c-Jun⫹c-Fos versus p(CRE/AP-1)6-SV40Luc control, p(CRE/ AP-1)3-SV40Luc⫹c-Jun versus p(CRE/AP-1)3-SV40Luc control, p ⱕ 0.05. B, overexpression of Jun and Fos proteins was detected by Western blotting (see “Experimental Procedures”).

Cell-specific HO-1 Gene Activation by MAPKs—HO-1 gene expression is induced by various stress stimuli that activate MAPKs. Conversely, overexpression of HO-1 and HO-1-derived CO were shown to exert anti-inflammatory and cytoprotective effects via MAPKs as mediators (52, 53). The role of MAPKs have previously been demonstrated in various cell culture systems (20, 21, 27–29, 54), and contradictory results on the regulatory role of different MAPK pathways for HO-1 gene expression were observed. Results obtained in HeLa cells indicated that the induction of human HO-1 by NO was mediated via ERK and p38 pathways but not via the SAPK/JNK pathway (29). In another human cell line, HepG2, overexpression of MEKK1, a kinase upstream of JNK, led to the induction of HO-1 protein levels (28). On the other hand, it was reported that the induction of HO-1 mRNA expression by cadmium, arsenite, and hemin was not mediated via MAPKs in Hela cells (20). In the same cells SB203580, an inhibitor of p38 MAPK, had no effect on the cadmium-, arsenite-, and hemin-dependent HO-1 mRNA induction. In contrast, in the chicken hepatoma cell line LMH HO-1

gene induction by sodium arsenite was mediated via ERK and p38 kinase pathways (21). Transfection of LMH cells with expression vectors for wild-type and activated forms of the components of the ERK pathway and the p38 cascade induced chicken HO-1 promoter-driven luciferase activity whereas dominant-negative forms of MAPKs blocked luciferase activity. Cotransfected expression vectors encoding wild-type and dominant-negative components of the JNK pathway (JNK, MEKK1, MLK3) with chicken HO-1 promoter Luc reporter genes did not affect Luc expression and its induction by arsenite. However, data from our study show that induction of rat HO-1 expression by sodium arsenite was mediated via the JNK pathway (Fig. 1) and that HO-1 can be induced via overexpression of JNK but not ERK2 (Figs. 2– 4). Accordingly, inefficiency of the ERK pathway was demonstrated in the cadmium-dependent activation of the mouse HO-1 gene, thereby pointing for a role of the p38 cascade (27). A major reason for these regulatory discrepancies could be species- or cell-specific variations that may affect the regulation of HO-1 gene expression via MAPKs and different mechanisms of

17934


FIG. 7. Binding of Max to the E-box element of the rat HO-1 promoter. A, E-box consensus sequence and the rat HO-1 promoter sequence ⫺50/⫺39 containing the wild-type and the mutated E-box are shown. Bases matching the consensus sequences are in bold, mutated bases are shown in lowercase letters. B, EMSA: the 32P-labeled HO-1 wild-type (WT) and mutated (M) E-box oligonucleotides were incubated with 10 ␮g of protein of nuclear extracts. In EMSA with antibodies the nuclear extracts were preincubated with 1 ␮l of the Myc, Max, or Sp1 antibodies for 2 h at 4 °C before adding the labeled probe. The DNAprotein binding was analyzed by electrophoresis on 5% native polyacrylamide gels. S, supershifted Max complex.

arsenite-dependent HO-1 induction. JNK-mediated HO-1 Gene Activation in Primary Rat Hepatocytes—Different regulation of the rat and chicken gene by MAPKs could be caused by the lack of an AP-1 site at ⫺668/ ⫺654 in the chicken HO-1 promoter, which mediates HO-1 gene activation via the JNK and Jun pathway (Fig. 4). Although the rat and mouse HO-1 gene 5⬘-flanking regions exhibit a much higher sequence similarity than the rat and the chicken HO-1 gene 5⬘-flanking regions (46, 55, 56) different regulatory patterns were observed. It was shown that the two distal enhancer regions E1 (⫺4 kb) and E2 (⫺10 kb) containing several antioxidant response elements that are regulatory sites for the transcription factor Nrf2 are mainly responsible for

stress-induced activation of mouse HO-1 gene expression (22, 55, 57). In contrast to the mouse gene, in the rat gene a number of apparently inducer-specific regulatory elements are located in the proximal promoter. These include the CRE/AP-1 site, which was shown in the present study to be responsible for the induction of rat HO-1 via the JNK pathway, a prostaglandin J2 element (⫺673/⫺668) and a heat shock element (⫺278/⫺263) (5, 57–59). However, the ⫺754 bp rat HO-1 promoter does not contain a binding site for Nrf2. Accordingly, overexpression of Nrf2 did not enhance rat HO-1 promoter activity in primary rat hepatocytes although it was able to increase Luc activity from the Nrf2 responsive mouse HO-1 pE1-Luc gene (kindly provided by J. Alam) (data not shown). However, this does not preclude a role for Nrf2 in the regulation of the rat HO-1 gene, in particular, because HO-1 induction is lost in Nrf2-deficient mice (60). Thus, there could be an Nrf2 binding site in the yet unknown rat HO-1 far upstream regulatory sequences, which in addition to the sequences identified in this study might have a potential role in rat HO-1 gene activation. In the study with the mouse HO-1 gene, however, dominantnegative mutants of JNK1 and JNK2 did not reduce basal and cadmium-induced pE1-Luc activity suggesting that the cadmium-dependent activation of the HO-1 gene was not mediated via the JNK pathway (27). This result is in contrast to the findings of this study and to the observation with rat liver in which induction of HO-1 mRNA expression by the glutathione depletor phorone was mediated by JNK1 and c-Jun (54). However, in that study with rat liver the effects of MAPKs other than JNK1 and JNK2 on rat HO-1 gene expression and the localization of the JNK target site in the rat HO-1 promoter were not examined. That glutathione depletion can regulate Ras/MAP kinases was previously demonstrated also in rat pulmonary artery smooth muscle cells (61) and rat mesangial cells (62). It is conceivable that the cascade identified in this study via Ras/MEKK1/JNK/Jun is responsible for the rat HO-1 gene activation by phorone. The finding that the CRE/AP1 element is the target site for the transcription factor c-Jun may also agree with previous studies showing that cGMP/protein kinase G (PKG)-dependent HO-1 gene activation is mediated via this site (5). Since the PKG signal can couple to the MEKK-1 pathway (63) it is conceivable that agents that stimulate the cGMP signaling like atrial natriuretic peptide (ANP) (64) activate HO-1 gene expression via the PKG-MEKK1-JNK-Jun cascade. In addition, AP-1 has previously been demonstrated as a target of the cGMP-dependent gene activation (5, 65, 66). The most prominent AP-1 dimer is the c-Jun/c-Fos heterodimer, which appears to be involved in the activation of a number of target genes (67). However, AP-1 may consist also of heterodimers of c-Jun and a member of the ATF/CREB family especially ATF-2 (68). The c-Jun/c-Fos heterodimer binds to the TGANTCA consensus sequence whereas the c-Jun/ATF heterodimer binds preferentially to the TGANNTCA consensus site (69, 70). The rat HO-1 promoter CRE/AP-1 site is homologous to the latter consensus sequence, which may be bound by c-Jun/ATF proteins. This was supported by the data of EMSA studies in which antibodies against c-Fos failed to produce a supershifted complex (Fig. 5). Furthermore, overexpression of c-Jun induced HO-1 promoter activation whereas overexpression of c-Fos showed a negative effect. This negative effect could be explained by the assumption that overexpressed Fos dimerized with Jun competing with the ATF/CREB. Thus, it appears that the HO-1 CRE/AP-1 site is a binding site for either Jun homodimers or Jun/ATF heterodimers. p38-mediated HO-1 Gene Regulation—The p38 pathway has


17935

FIG. 8. Mutation of the E-box abolished the p38 and c-Max-dependent inhibition of HO-1 promoter-controlled luciferase expression. Hepatocytes were transiently cotransfected with either p38␣, p38␤, p38␥, p38␦ and the mutants, p38␣AF, p38␤AF, p38␥AF, p38␦AF, Max, or MKK3 expression plasmids and Luc gene constructs driven by the wild-type 347 bp of the rat HO-1 promoter (pHO-754) or the 347-bp promoter mutated at the E-box site (pHO-347Em). In control experiments Luc constructs were cotransfected with the empty control vectors. In pHO-347Em the wild-type HO-1 sequence is shown on the upper strands, and mutated bases are shown in lowercase letters and are indicated by *. In each experiment the percent of Luc activity was determined relative to the pHO-347 or pHO-347Em controls, which were set equal to 100%. The values represent means ⫾ S.E. of three independent experiments. Statistics, Student’s t test for paired values: *, significant difference pHO-347⫹MKK3 versus pHO-347 control, pHO-347⫹p38␣ versus pHO-347 control, pHO-347⫹p38␤ versus pHO-347 control, pHO-347⫹p38␥ versus pHO-347 control, pHO-347⫹p38␥ versus pHO-347 control, pHO-347⫹Max versus pHO-347 control, p ⱕ 0.05.

previously been shown to activate the mouse and chicken HO-1 genes (21). In the mouse gene a stress response element in the upstream E1 enhancer as target for Nrf2 was shown to be involved in p38-dependent induction (27). Our data indicate that in rat hepatocytes different p38 isoforms exert different patterns in the regulation of HO-1 gene expression (Fig. 8). We observed a p38␥-dependent induction and inhibition of the HO-1 promoter activity by the other p38 isoforms. This observation could explain why the p38 inhibitor SB203580 alone was not effective on HO-1 expression. Furthermore, such different actions of p38 isoforms have been observed in other cell types (71). Interestingly, an inhibitory role for p38 has also been previously described by others (28, 72). In particular, it was shown that the p38 pathway may suppress Ras proliferative signaling in NIH3T3 cells (72). Furthermore, an inhibitory effect of p38 on gene expression was also demonstrated (28, 73). Thereby, overexpression of wild-type p38 not only inhibited basal expression of an antioxidant response element (ARE)-dependent luciferase reporter gene construct but also suppressed ARE-Luc activation by MEKK1 in HepG2 cells (28). In the same study MEKK1 induced HO-1 expression (28). In our study the E-box element (⫺47/⫺42) was identified as target for p38-dependent induction and inhibition of HO-1 gene expression. The HO-1 Luc reporter gene construct containing a mutated E-box was not repressed by cotransfection of vectors encoding wild-type or mutated p38 (Fig. 8). The E-box sequence (CACGTG) is known to be bound by members of the bHLH leucine zipper (bHLHzip) protein family such as upstream stimulatory factors (USF) or the Myc/Max dimer (74). Among bHLHzip proteins the transcription factor Max (49) is known to be a target for p38 kinase. It was further demonstrated that Myc in contrast to Max did not bind the E-box element (⫺47/⫺42) of the rat HO-1 promoter (Fig. 7). Although this finding was surprising, Max not only forms heterodimers with Myc, but also forms homodimers (75) and het-

erodimers with other members of the bHLHzip family such as Mad1, Mad3, Mad4, and Mnt (76, 77). It was shown that in contrast to the stimulatory effect of Myc/Max heterodimers on transcription, Max homodimers, or heterodimers between Max and members of the MAD family of bHLH-zip proteins, repress transcription (78). In our study an inhibitory effect of Max on HO-1-driven Luc activity was also observed. Mutation of the HO-1 E-box abolished this effect. Thus, the E-box of the rat HO-1 gene promoter is a target site for p38- and Maxmediated inhibition. Acknowledgments—We thank Dr. R. J. Davis for the gift of MKK3 and MKK4 dominant-negative mutants and Dr. Jiahuai Han for the generous supply with the p38 vectors. REFERENCES 1. Tenhunen, R., Marver, H. S., and Schmid, R. (1968) Proc. Natl. Acad. Sci. U. S. A. 61, 748 –755 2. Maines, M. D. (1988) FASEB J. 2, 2557–2568 3. McCoubrey-WK, J., Huang, T. J., and Maines, M. D. (1997) Eur. J. Biochem. 247, 725–732 4. Immenschuh, S., Kietzmann, T., Hinke, V., Wiederhold, M., Katz, N., and Muller, E. U. (1998) Mol. Pharmacol. 53, 483– 491 5. Immenschuh, S., Hinke, V., Ohlmann, A., Gifhorn, K. S., Katz, N., Jungermann, K., and Kietzmann, T. (1998) Biochem. J. 334, 141–146 6. Keyse, S. M., and Tyrrell, R. M. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 99 –103 7. Applegate, L. A., Luscher, P., and Tyrrell, R. M. (1991) Cancer Res. 51, 974 –978 8. Lee, P. J., Jiang, B. H., Chin, B. Y., Iyer, N. V., Alam, J., Semenza, G. L., and Choi, A. M. (1997) J. Biol. Chem. 272, 5375–5381 9. Lee, P. J., Camhi, S. L., Chin, B. Y., Alam, J., and Choi, A. M. (2000) Am. J. Physiol. Lung Cell Mol. Physiol. 279, L175–L182 10. Rizzardini, M., Terao, M., Falciani, F., and Cantoni, L. (1993) Biochem. J. 290, 343–347 11. Choi, A. M., and Alam, J. (1996) Am. J. Respir. Cell Mol. Biol. 15, 9 –19 12. Poss, K. D., and Tonegawa, S. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 10919 –10924 13. Maines, M. D. (1997) Annu. Rev. Pharmacol. Toxicol. 37, 517–554 14. Gomez, N., and Cohen, P. (1991) Nature 353, 170 –173 15. Davis, R. J. (1999) Biochem. Soc. Symp. 64, 1–12 16. Han, J., Lee, J. D., Bibbs, L., and Ulevitch, R. J. (1994) Science 265, 808 – 811 17. Nebreda, A. R., and Porras, A. (2000) Trends. Biochem. Sci. 25, 257–260 18. Kyriakis, J., and Avruch, J. (2001) Physiol. Rev. 81, 807– 869

17936


19. Sardana, M. K., Drummond, G. S., Sassa, S., and Kappas, A. (1981) Pharmacology 23, 247–253 20. Masuya, Y., Hioki, K., Tokunaga, R., and Taketani, S. (1998) J. Biochem. (Tokyo) 124, 628 – 633 21. Elbirt, K. K., Whitmarsh, A. J., Davis, R. J., and Bonkovsky, H. L. (1998) J. Biol. Chem. 273, 8922– 8931 22. Alam, J., Stewart, D., Touchard, C., Boinapally, S., Choi, A. M., and Cook, J. L. (1999) J. Biol. Chem. 274, 26071–26078 23. Ludwig, S., Hoffmeyer, A., Goebeler, M., Kilian, K., Hafner, H., Neufeld, B., Han, J., and Rapp, U. R. (1998) J. Biol. Chem. 273, 1917–1922 24. Liu, Y., Guyton, K. Z., Gorospe, M., Xu, Q., Lee, J. C., and Holbrook, N. J. (1996) Free Radic. Biol. Med. 21, 771–781 25. Chen, W., Martindale, J. L., Holbrook, N. J., and Liu, Y. (1998) Mol. Cell. Biol. 18, 5178 –5188 26. Takai, Y., Sasaki, T., and Matozaki, T. (2001) Physiol. Rev. 81, 153–208 27. Alam, J., Wicks, C., Stewart, D., Gong, P., Touchard, C., Otterbein, S., Choi, A. M., Burow, M. E., and Tou, J. (2000) J. Biol. Chem. 275, 27694 –27702 28. Yu, R., Chen, C., Mo, Y. Y., Hebbar, V., Owuor, E. D., Tan, T. H., and Kong, A.-N. T. (2000) J. Biol. Chem. 275, 39907–39913 29. Chen, K., and Maines, M. D. (2000) Cell Mol. Biol. Noisy. le. grand. 46, 609 – 617 30. Immenschuh, S., Iwahara, S., Satoh, H., Nell, C., Katz, N., and Muller, E. U. (1995) Biochemistry 34, 13407–13411 31. Whitmarsh, A. J., Yang, S. H., Su, M. S., Sharrocks, A. D., and Davis, R. J. (1997) Mol. Cell. Biol. 17, 2360 –2371 32. Xu, S., Robbins, D. J., Christerson, L. B., English, J. M., Vanderbilt, C. A., and Cobb, M. H. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 5291–5295 33. Derijard, B., Hibi, M., Wu, I. H., Barrett, T., Su, B., Deng, T., Karin, M., and Davis, R. J. (1994) Cell 76, 1025–1037 34. Raingeaud, J., Whitmarsh, A. J., Barrett, T., Derijard, B., and Davis, R. J. (1996) Mol. Cell. Biol. 16, 1247–1255 35. Han, J., Lee, J. D., Jiang, Y., Li, Z., Feng, L., and Ulevitch, R. J. (1996) J. Biol. Chem. 271, 2886 –2891 36. Jiang, Y., Chen, C., Li, Z., Guo, W., Gegner, J. A., Lin, S., and Han, J. (1996) J. Biol. Chem. 271, 17920 –17926 37. Jiang, Y., Gram, H., Zhao, M., New, L., Gu, J., Feng, L., Di Padova, F., Ulevitch, R. J., and Han, J. (1997) J. Biol. Chem. 272, 30122–30128 38. Feig, L. A., and Cooper, G. M. (1988) Mol. Cell. Biol. 8, 3235–3243 39. Dent, P., Reardon, D. B., Morrison, D. K., and Sturgill, T. W. (1995) Mol. Cell. Biol. 15, 4125– 4135 40. Schonthal, A., Buscher, M., Angel, P., Rahmsdorf, H. J., Ponta, H., Hattori, K., Chiu, R., Karin, M., and Herrlich, P. (1989) Oncogene 4, 629 – 636 41. Kretzner, L., Blackwood, E. M., and Eisenman, R. N. (1992) Nature 359, 426 – 429 42. Kietzmann, T., Roth, U., Freimann, S., and Jungermann, K. (1997) Biochem. J. 321, 17–20 43. Kietzmann, T., Hirsch, E. K., Kahl, G. F., and Jungermann, K. (1999) Mol. Pharmacol. 56, 46 –53 44. Kietzmann, T., Roth, U., and Jungermann, K. (1999) Blood 94, 4177– 4185 45. Jacobs, J. M., Nichols, C. E., Andrew, A. S., Marek, D. E., Wood, S. G., Sinclair, P. R., Wrighton, S. A., Kostrubsky, V. E., and Sinclair, J. F. (1999) Toxicol. Appl. Pharmacol. 157, 51–59 46. Muller, R. M., Taguchi, H., and Shibahara, S. (1987) J. Biol. Chem. 262, 6795– 6802 47. Tibbles, L. A., and Woodgett, J. R. (1999) Cell Mol. Life Sci. 55, 1230 –1254 48. Vuong, H., Patterson, T., Shapiro, P., Kalvakolanu, D. V., Wu, R., Ma, W. Y., Dong, Z., Kleeberger, S. R., and Reddy, S. P. (2001) J. Biol. Chem. 275, 32250 –32259

49. Zervos, A. S., Faccio, L., Gatto, J. P., Kyriakis, J. M., and Brent, R. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 10531–10534 50. Murre, C., Bain, G., van, D. M., Engel, I., Furnari, B. A., Massari, M. E., Matthews, J. R., Quong, M. W., Rivera, R. R., and Stuiver, M. H. (1994) Biochim. Biophys. Acta 1218, 129 –135 51. Maeshima, H., Sato, M., Ishikawa, K., Katagata, Y., and Yoshida, T. (1996) Nucleic Acids Res. 24, 2959 –2965 52. Takeda, A., Perry, G., Abraham, N. G., Dwyer, B. E., Kutty, R. K., Laitinen, J. T., Petersen, R. B., and Smith, M. A. (2000) J. Biol. Chem. 275, 5395–5399 53. Otterbein, L. E., Bach, F. H., Alam, J., Soares, M., Tao, L. H., Wysk, M., Davis, R. J., Flavell, R. A., and Choi, A. M. (2000) Nat. Med. 6, 422– 428 54. Oguro, T., Hayashi, M., Nakajo, S., Numazawa, S., and Yoshida, T. (1998) J. Pharmacol. Exp. Ther. 287, 773–778 55. Alam, J., Cai, J., and Smith, A. (1994) J. Biol. Chem. 269, 1001–1009 56. Lu, T. H., Lambrecht, R. W., Pepe, J., Shan, Y., Kim, T., and Bonkovsky, H. L. (1998) Gene (Amst.) 207, 177–186 57. Alam, J., Camhi, S., and Choi, A. M. (1995) J. Biol. Chem. 270, 11977–11984 58. Okinaga, S., and Shibahara, S. (1993) Eur. J. Biochem. 212, 167–175 59. Koizumi, T., Odani, N., Okuyama, T., Ichikawa, A., and Negishi, M. (1995) J. Biol. Chem. 270, 21779 –21784 60. Ishii, T., Itoh, K., Takahashi, S., Sato, H., Yanagawa, T., Katoh, Y., Bannai, S., and Yamamoto, M. (2000) J. Biol. Chem. 275, 16023–16029 61. Lander, H. M., Tauras, J. M., Ogiste, J. S., Hori, O., Moss, R. A., and Schmidt, A. M. (1997) J. Biol. Chem. 272, 17810 –17814 62. Sandau, K. B., Callsen, D., and Brune, B. (1999) Mol. Pharmacol. 56, 744 –751 63. Soh, J. W., Mao, Y., Liu, L., Thompson, W. J., Pamukcu, R., and Weinstein, I. B. (2001) J. Biol. Chem. 276, 16406 –16410 64. Silberbach, M., and Roberts, C. T. J. (2001) Cell. Signal. 13, 221–231 65. Gudi, T., Huvar, I., Meinecke, M., Lohmann, S. M., Boss, G. R., and Pilz, R. B. (1996) J. Biol. Chem. 271, 4597– 4600 66. Pilz, R. B., Suhasini, M., Idriss, S., Meinkoth, J. L., and Boss, G. R. (1995) FASEB J. 9, 552–558 67. Karin, M., Liu, Z., and Zandi, E. (1997) Curr. Opin. Cell Biol. 9, 240 –246 68. Shaulian, E., and Karin, M. (2001) Oncogene 20, 2390 –2400 69. Ivashkiv, L. B., Liou, H. C., Kara, C. J., Lamph, W. W., Verma, I. M., and Glimcher, L. H. (1990) Mol. Cell. Biol. 10, 1609 –1621 70. Hai, T., and Curran, T. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 3720 –3724 71. Pramanik, R., Qi, X., Borowicz, S., Choubey, D., Schultz, R. M., Han, J., and Chen, G. (2003) J. Biol. Chem. 278, 4831– 4839 72. Chen, G., Hitomi, M., Han, J., and Stacey, D. W. (2000) J. Biol. Chem. 275, 38973–38980 73. Lavoie, J. N., L’Allemain, G., Brunet, A., Muller, R., and Pouyssegur, J. (1996) J. Biol. Chem. 271, 20608 –20616 74. Sawadogo, M., and Roeder, R. G. (1985) Cell 43, 165–175 75. Tchan, M., Choy, K., Mackay, J., Lyons, A., Bains, N., and Weiss, A. (2001) J. Biol. Chem. 275, 37454 –374561 76. Hurlin, P. J., Queva, C., Koskinen, P. J., Steingrimsson, E., Ayer, D. E., Copeland, N. G., Jenkins, N. A., and Eisenman, R. N. (1995) EMBO J. 14, 5646 –5659 77. Meroni, G., Reymond, A., Alcalay, M., Borsani, G., Tanigami, A., Tonlorenzi, R., Nigro, C. L., Messali, S., Zollo, M., Ledbetter, D. H., Brent, R., Ballabio, A., and Carrozzo, R. (1997) EMBO J. 16, 2892–2906 78. Zhang, H., Fan, S., and Prochownik, E. V. (1997) J. Biol. Chem. 272, 17416 –17424 79. Hoeffler, J. P., Meyer, T., Yun, Y., Jameson, J. L., and Habener, J. F. (1988) Science 242, 1430 –1433