J Biol Chem, Vol. 274, Issue 46, 32580-32587, November 12, 1999

Regulation of c-Myc through Phosphorylation at Ser-62 and Ser-71 by c-Jun N-Terminal Kinase^*

Kohji Noguchi, Chifumi Kitanaka, Hironobu Yamana, Akiko Kokubu, Toshihiro Mochizuki, and Yoshiyuki Kuchino^§¶

From the  Biophysics Division, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan and ^§ Core Research for Evolutional Science and Technology, Japan Science and Technology Corporation, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The expression of c-myc promotes cell proliferation and also sensitizes cells to various extracellular apoptotic stimuli. However, signal pathways regulating the function of Myc proteins during apoptosis are unknown. c-Jun N-terminal kinase (JNK) is activated by various apoptotic stimuli, but neither the target molecule(s) or the action of JNK has been identified in Myc-mediated apoptosis. Here, we found that JNK selectively interacted with, and phosphorylated, c-Myc at Ser-62 and Ser-71 as confirmed with phospho-c-Myc-specific antibodies. Interestingly, dominant negative mutant JNK(APF) impaired the c-Myc-dependent apoptosis, but not mutated c-Myc (S62A/S71A)-dependent apoptosis triggered by UV irradiation. Furthermore, c-Myc (S62A/S71A)-expressing NIH3T3 cells were not sensitized like wild type c-Myc-expressing NIH3T3 cells to JNK-activating apoptotic stimuli, such as UV and Taxol. These results indicate that the JNK pathway is selectively involved in the c-Myc-mediated apoptosis and that the apoptotic function of c-Myc is directly regulated by JNK pathway through phosphorylation at Ser-62 and Ser-71.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The Myc family proteins, such as c-, N-, L-, and s-Myc, are transcriptional factors of the basic-helix-loop-helix leucine zipper family and recognize the hexanucleotide sequence CAC(G/T)TG (E-box element) through heterodimer formation with Max protein (1). The c-Myc protein plays a important role in the cell transformation through the transcriptional regulation of target genes (1). However, the myc family genes are also implicated in the regulation of apoptosis, and both the mitogenic and proapoptotic properties of Myc are functionally inseparable (2). There is substantial evidence that c-myc and s-myc promote apoptosis induced by various stimuli including serum deprivation, Fas, UV irradiation, some antitumor agents, and T-cell receptor activation (3-7). The precise effect of myc expression in apoptosis is largely unclear, but Myc-mediated apoptosis is suppressed by bcl-2 family genes, ras, and insulin-like growth factor-1 and requires the caspase activation in the apoptotic processes (8-12). Importantly, loss of caspase-9/Apaf-1-mediated apoptotic machinery results in an increase of malignant cell transformation by c-myc (13). However, it is unclear what signal pathway triggers the Myc-dependent mechanism of apoptosis induction.

Extracellular stresses, including genotoxic stress stimuli, often activate the c-Jun N-terminal kinase (JNK)¹/p38 mitogen-activated protein kinase (MAPK) pathway (14, 15). JNK and p38 MAPK are activated by upstream MAPK kinases, such as MKK4/7 and MKK3/6, and then phosphorylate their targets molecules, such as c-Jun, ATF-2, Elk-1, or p53 transcription factors (16-20). Previous studies have shown that JNK and p38 MAPK are involved in apoptosis caused by nerve growth factor withdrawal in PC12 cells, antitumor agents in U937 cells, anisomycin in Jurkat cells, and glutamate-induced apoptosis in rat cerebellar granule cells (21-24). Moreover, JNK and/or p38 MAPK regulating upstream JNK kinase kinases, such as MEKK1 and ASK1, induce apoptotic cell death in some types of cells (24-26). The most direct observation is that jnk3-deficient mice are resistant to excitotoxicity-induced apoptosis in hippocampal neurons (27). However, the role of JNK-activating signals is confusing regarding cell survival or death (28, 29), and identifying the target molecules of JNK is important to elucidating the effect of JNK.

In the present study, we investigated whether there is a direct link between JNK pathway and the functional regulation of Myc through phosphorylation. We found that JNK directly phosphorylated c-Myc at Ser-62 and Ser-71, which affects the apoptosis-promoting activities of c-Myc. Therefore, our findings suggest a selective and direct role for JNK as a signal mediator of stress stimuli to c-Myc in apoptosis.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- pc3GF vector was constructed as follows. Briefly, the neomycin-resistant gene was replaced with the enhanced green fluorescence protein (EGFP) gene of pEGFP (CLONTECH Laboratories, Inc., Palo Alto, CA) in the pcDNA3 plasmid by BspMI digestion following blunt-end ligation, and this EGFP gene-expressing plasmid was termed pc3GF. Desired cDNAs were inserted at HindIII-XhoI sites to obtain the pc3GFhcmyc plasmid. Human c-myc cDNA was inserted into pCMV2-Flag plasmid (Eastman Kodak) to construct pflag-hcmyc. pSPhcmyc was described before (6, 30). pCR3JNK11 and pCR3JNK11(APF) were gifts from Dr. H. Seimiya (Cancer Chemotherapy Center, Japanese Foundation for Cancer Research, Tokyo, Japan).

Anti-Flag M5 antibody was purchased from Kodak. Anti-JNK1(C-17), anti-phospho-JNK (G-7), anti-ERK2 (C-14), anti-phospho-ERK (E-4), anti-p38 (C-20-G), and control rabbit antibodies were from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Anti-phospho-c-Myc(Thr58/Ser62) antibodies were from New England Biolabs Inc., (Beverly, MA). Taxol (paclitaxel) was from Sigma, and PD98059 was from Calbiochem-Novabiochem (La Jolla, CA).

Preparation of Recombinant Proteins-- For the preparation of various GST-c-Myc proteins, cDNA fragments were generated from each c-myc cDNA containing template plasmids by polymerase chain reaction, and inserted into pGEX2T plasmid at BamHI and EcoRI sites. All polymerase chain reaction-generated fragments were sequenced for verification. BamHI fragment of mouse s-myc gene was inserted into pGEX2T at BamHI site, and pGEX2T/Max was constructed as described (6). pGEX5X/c-Jun (1-92) and pGEX5X/JNK11 were gifts from Dr. H. Seimiya (Cancer Chemotherapy Center, Japanese Foundation for Cancer Research). His-ATF2 was a gift from Prof. E. Nishida (Kyoto University). GST proteins were induced in Escherichia coli JM109 by addition of isopropyl--thiogalactopyranoside, purified with glutathione-Sepharose 4B beads, and eluted by 20 mM glutathione in 50 mM Tris-HCl (pH 8.8). GST-JNK3, GST-JNK3 (K55R), and GST-ERK2 were from Upstate Biotechnology (Lake Placid, NY).

Mutagenesis of c-myc-- Mutagenesis of human c-myc cDNA was performed with the polymerase chain reaction-based QuikChange^TM site-directed mutagenesis kits (Stratagene, La Jolla, CA) according to the instruction manual. In brief, synthetic DNA primers (5'-GAAATTCGAGCTGCTGCCCGCCCCGCCCC-TGTCCCCTAGCCGC-3' and 5'-GCGGCTAGGGGACAGGGGCGGGGCGGGCAGCAGCTCGAATTTC-3' for c-MycT58A, 5'-CTGCTGCCCACCCC-GCCCCTGGCCCCTAGCCGCCGCTCCGGGCTC-3' and 5'-GAGCCCGGAGCGGCGGCTAGGGGCCAGGGGCGGGGTGGGCAGCAG-3' for c-MycS62A, 5'-GCCGCCGCTCCGGGCTCTGCGCGCCCTCCTACGTTGC-GGTCAC-3' and 5'-GTGACCGCAACGTAGGAGGGCGCGCAGAGCCCGGAGCGGCGC-3' for c-MycS71A, and 5'-CGCCCTCCTACGTTGCGGTCGCACCCTTCTCCCTTCGGGGAGAC-3' and 5'-GTCTCCCCGAAGGGAGAAGGGTGCGAC-CGCAACGTAGGAGGGCG-3' for c-MycT78A) were used in the polymerase chain reaction with template pcDNA3hcmyc plasmid. All mutated c-myc cDNA clones were fully sequenced and subcloned into pc3GF plasmid.

In-gel Kinase Assay-- Cells were lysed in WCE buffer (25 mM Hepes-KOH, pH 7.7, 0.3 M NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.1% Triton X-100, 20 mM -glycerophosphate, 0.1 mM Na₃VO₄, 2 mM DTT, 0.5 mM PMSF). JNK1 and ERK2 were immunoprecipitated from the cell extracts (1.6 mg of protein) by using 10 µg of anti-JNK1 and anti-ERK2 antibodies bound with protein-G4FF-Sepharose beads. Immunocomplexes were resolved on 11% SDS-polyacrylamide gels polymerized with GST-c-Jun (100 µg/ml) or GST-c-Myc (1-262) (100 µg/ml). After electrophoresis, each gel was washed twice for 30 min with 20% 2-propanol, 50 mM Hepes-KOH (pH 7.6) to remove SDS, and again twice for 30 min with Buffer A (50 mM Hepes-KOH, pH 7.6, 5 mM 2-mercaptoethanol). The gel was denatured in 6 M urea in Buffer A for 1 h and then subjected to serial incubations in Buffer A containing 0.05% Tween-20 and either 3, 1.5, or 0.75 M urea. After several washes at room temperature and one overnight wash at 4 °C in Buffer A containing 0.05% Tween-20, the gel was once washed in kinase buffer (50 mM Hepes-KOH, pH 7.6, 0.5 mM EGTA, 20 mM MgCl₂, 2 mM DTT) for 20 min at room temperature. The kinase reaction proceeded in kinase buffer containing 20 µM ATP and 20 µCi/ml [-³²P]ATP at 30 °C for 1 h. Finally, the gel was washed with 5% trichloroacetic acid and 1% sodium pyrophosphate at room temperature several times and then overnight at 4 °C before being dried. Kinase activities were analyzed with a BAS2000 bio-image analyzer (Fujix, Tokyo).

Kinase Assay of JNK-- For the in vitro kinase assay, 5 µg of each recombinant substrate was incubated with 100 ng of GST-JNK11 in Buffer K (25 mM Hepes-KOH, pH 7.6, 20 mM MgCl₂, 20 mM -glycerophosphate, 20 mM p-nitrophenyl phosphate, 1 mM Na₃VO₄, 2 mM DTT, 20 µM ATP, 100 µCi/ml [-³²P]ATP) at 30 °C for 30 min, and the reaction was stopped by adding 2× Laemlli buffer. Phosphorylated proteins were resolved by SDS-PAGE, stained with Coomassie Blue, and analyzed.

For the immunocomplex kinase assay, JNK was immunoprecipitated as above, and the immunocomplex was washed with WCE buffer (1 ml, three times) followed by Buffer K without [-³²P]ATP (1 ml, one time). The kinase reaction proceeded as described above with GST-c-Jun (1-92) as a substrate.

Generation of Anti-phospho-c-Myc (Ser-71) Polyclonal Antibody and Detection of Phosphorylated c-Myc-- To generate anti-phospho-c-Myc (Ser-71) polyclonal antibody (-P-Ser-71), a synthetic phospho-peptide (RSGLC(phospho-S)PSYVA) corresponding to amino acids 66-76 in the human c-Myc protein was used as an antigen. After five immunizations in Japanese White rabbits, serum was prepared, and -P-Ser71 was purified on peptide-affinity columns. For the detection of phosphorylated c-Myc of Ser-71 in vivo, 293T cells were transfected with pc3GFhcmyc. Two days after transfection, cells were suspended in cytosol buffer (0.2% Nonidet P-40, 20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 10 mM NaF, 20 mM -glycerophosphate, 1 mM Na₃VO₄, 2 mM MgCl₂, 1 mM PMSF, 1 mM DTT) and the particle nuclear pellet was lysed with particle buffer (1% Nonidet P-40, 0.5% SDS, 0.5% sodium deoxycholate, 20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 10 mM NaF, 20 mM -glycerophosphate, 1 mM Na₃VO₄, 1 mM PMSF, 1 mM DTT) by sonication. Equal amounts of nuclear protein (32 µg) were resolved by SDS-PAGE and transferred to filters. Filters were blocked with 5% lowfat milk, Tris-buffered saline, 0.1% Tween-20 at room temperature, and phosphorylation of Thr-58/Ser-62 or Ser-71 of c-Myc was detected using anti-phospho-c-Myc (Thr-58/Ser-62) antibody (-P-Thr-58/Ser-62) and -P-Ser-71. First antibodies diluted with 3% bovine serum albumin, Tris-buffered saline were used for an overnight incubation at 4 °C, and after hybridization with secondary antibodies conjugated with horseradish peroxidase, signals were detected with ECL detection reagent (Amersham Pharmacia Biotech).

In Vitro and in Vivo Binding of JNK1 and c-Myc-- For the pull-down assay, in vitro transcribed and translated ³⁵S-Met-labeled JNK11 and GST protein (30 µg)/glutathione-Sepharose beads were mixed in a dilution buffer (20 mM Hepes-KOH, pH 7.6, 20 mM MgCl₂, 0.5 mM EGTA, 2 mM DTT, 0.1 mM PMSF) and rotated at 4 °C for 2 h. Beads were washed with binding buffer (20 mM Hepes-KOH, pH 7.6, 50 mM NaCl, 2.5 mM MgCl₂, 0.05% Triton X-100, 1 mM DTT, 1 mM PMSF) (1 ml, five times), and bound proteins were resolved by 12% SDS-PAGE. The binding proteins were analyzed with a BAS2000 bio-image analyzer.

For the detection of an in vivo association between JNK1 and c-Myc, 293T cells (3 × 10⁶ cells) were transfected with pflag-hcmyc (10 µg) using SuperFect reagents (Qiagen GmbH, Hilden, Germany). Two days after transfection, cells were either irradiated with UVC (300 J/m²) or left unirradiated. After a further 30 min of incubation in the presence of chemical cross-linker dithiobis(succinimidylpropionate) (Pierce) (1.5 mg/ml), cells were lysed in Buffer S (20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM PMSF) following brief sonication. Immunoprecipitated complex from equal amounts of each cell lysate (1.6 mg) was extensively washed with Buffer S (1 ml, five times) and subjected to 10% SDS-PAGE, and flag-tagged c-Myc was detected by Western blot analysis using anti-Flag M5 antibody.

Western Blot Analysis-- To detect the c-Myc and mutated Myc proteins, cells were lysed in the Buffer S and cleared lysates were obtained by centrifugation. Equal amounts of protein were resolved by 10% SDS-PAGE following transfer to nitrocellulose membrane. Filters were blocked in 5% lowfat milk, PBS, 0.1% Tween 20 and hybridized with first antibodies, such as anti-c-Myc monoclonal antibody 9E10 (Calbiochem-Novabiochem). For the detection of JNK and ERK, cells after treatment were once washed with PBS and frozen with liquid nitrogen. Cells were lysed in 1× Laemlli sample buffer or Buffer S containing phosphatase inhibitors (20 mM NaF, 1 mM Na3VO4, 20 mM -glycerophosphate), heat-denatured, and sonicated. Equal amounts of protein were resolved by 12% SDS-PAGE, transferred, and hybridized with a first antibody, such as anti-phospho-JNK (G-7) or anti-phospho-ERK (E-4), diluted with 3% bovine serum albumin/Tris-buffered saline, or anti-JNK1(C-17) and anti-ERK2 (C-14) diluted with 5% lowfat milk, PBS, 0.1% Tween 20. After hybridization with secondary antibodies conjugated with horseradish peroxidase, immunocomplex was detected with the ECL detection reagent.

Cell Culture and Apoptosis Assay-- Human cervical carcinoma HeLa cells, mouse fibroblast NIH3T3 cells, and human embryonic kidney transformed 293T cells were grown in Dulbecco's modified Eagle's medium (Nissui, Tokyo, Japan) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Cytosystems, Castle Hill, Australia). To establish human c-myc-transfected NIH3T3 clones, cells were cotransfected using pcDNA3 with each myc-expressing pSP271 plasmid (pSPhcmyc and pSPhcmyc(S62A/S71A)). Transfected cells were cultured in the presence of G418 (200 µg/ml) for 2 weeks and cloned. Deregulated expression of exogenous human c-myc gene products was confirmed by both Northern and Western blot analysis. To isolate JNK(APF)-expressing HeLa cells, pCR3JNK11(APF) was transfected into HeLa cells, and transfected cells were cultured with G418 (400 µg/ml) for 2 weeks. After screening jnk1(APF) expression, a HeLa/APF cell clone was established.

For transient apoptosis assay of HeLa cells, HeLa (4 × 10⁵ cells/60-mm dish) cells were transfected with plasmid using SuperFect reagents. At 40 h posttransfection, cells were either irradiated by UVC (100 J/m²) or not and then incubated for 4 h in the original medium containing 10% FBS. Next, the cells were fixed with 1% formaldehyde-PBS for 1 min, and the EGFP-positive cells were examined for normal or apoptotic morphology under a fluorescence microscope as described (6). Apoptosis was also determined by trypan blue dye exclusion assay.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Phosphorylation of c-Myc by JNK-- We assumed that some molecules might directly mediate stress signals to Myc, and modulate the functions of Myc. Our preliminary experiments suggested that 42-55-kDa protein kinase(s) could be activated by UVC and phosphorylate c-Myc detected by in-gel kinase assay.² UV irradiation is known to activate the JNK and p38 MAPK pathway, and the molecular masses of these MAPK family proteins are around 42-55 kDa. As reported before, UV irradiation activates JNK and a mitogenic MAPK protein, ERK2 pathway in HeLa cells (17), and ERK2 phosphorylates c-Myc at Ser-62 (31). Furthermore, the JNK pathway contributes to the apoptosis induction (21, 22). Then, we examined whether JNK could phosphorylate c-Myc, as does ERK2. JNK and ERK2 were immunoprecipitated from control and UV-irradiated HeLa cell lysates, and immunocomplex was subjected to in-gel kinase assay with GST-c-Myc (1-262) as a substrate (Fig. 1A). We found that not only ERK2 but also activated JNK phosphorylated c-Myc, and surprisingly, phosphorylation of c-Myc by activated JNK was likely to be stronger than that by activated ERK2 in HeLa cell lysate (Fig. 1A, right). ERK2 was activated about 2-fold by UV irradiation when GST-c-Myc was used as a substrate, but JNK was activated about 50-fold when GST-c-Jun (1-92) was used as a positive control substrate (Fig. 1A, left). These observations suggest that the total kinase activities of both JNK and ERK2 will affect the phosphorylation status and the function of c-Myc protein, regardless of differences in the total amount and in the substrate specificity of JNK and ERK2.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   JNK phosphorylates c-Myc. A, phosphorylation of c-Myc was demonstrated by in-gel kinase assay. Immunoprecipitated JNK1 and ERK2 were resolved by 11% SDS-polyacrylamide gels polymerized with GST-c-Jun (1-92) or GST-c-Myc (1-262) (100 µg/ml). A representative result of in-gel kinase assay is shown. B, substrate reactivity of JNK against Mycs was studied by in vitro kinase assay. The top panel shows the autoradiography, and the bottom panel shows the Coomassie Blue staining pattern of this gel. The arrow indicates the autophosphorylation of GST-JNK1, and stars indicate the recombinant substrates.

Next, we examined the phosphorylation of several c-Myc substrates by JNK (Fig. 1B). In vitro kinase assay by GST-JNK1 demonstrated that JNK1 phosphorylated GST-c-Myc (1-139), (1-181), and (1-262), but less GST-c-Myc (1-66) and little GST-c-Myc (1-48, 101-262) (Fig. 1B). This study indicated that c-Myc was phosphorylated among amino acids 49-100. Furthermore, interestingly, JNK1 did not phosphorylate GST-s-Myc (1-337) and GST-Max (Fig. 1B) Thus, c-Myc might be a substrate for JNK, but s-Myc and Max are not. In addition, GST-JNK2 and GST-JNK3 could also phosphorylate GST-c-Myc as JNK1 does,² although the phosphorylation ratio of c-Myc was lower than that of ATF-2 and of c-Jun.

Stress-activated protein kinases such as JNK1, JNK2, and JNK3, belong to the MAPK superfamily, and they phosphorylate Ser/Thr residues following Pro residue in the target substrate (32). Our data indicated that phosphorylation sites for JNK were located between amino acids 49 and 100. In vivo phosphorylation of c-Myc protein has been investigated, and phosphorylation at Thr-58, Ser-62, and Ser-71, each of which was followed by a Pro residue, was identified within amino acids 49-100 (33). From these observations, we tested the phosphorylation at these residues by JNK using mutant GST-c-Myc substrates such as GST-c-Myc (T58A), GST-c-Myc (S62A), GST-c-Myc (S71A), GST-c-Myc (T78A), and GST-c-Myc (S62A/S71A) (Fig. 2A). The in vitro kinase assay of both JNK1 and JNK3 revealed that mutation at Ser-62 or Ser-71 to Ala reduced the phosphorylation, and double mutation completely abolished it. On the other hand, the in vitro kinase assay of ERK2 against the same substrates confirmed that ERK2 phosphorylated only Ser-62, and the mutation at Ser-71 did not result in a reduction of phosphorylation in c-Myc.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Ser-62 and Ser-71 of c-Myc are phosphorylated by JNK. A, substitutions of Ser-62 and Ser-71 by Ala abrogate the phosphorylation of c-Myc by JNK. The phosphorylation of mutated GST-fused c-Myc proteins by GST-JNK1, GST-JNK3, and GST-ERK2 was examined by in vitro kinase assay. The substrates (5 µg) were subjected to in vitro kinase assay and resolved by SDS-PAGE. Arrowheads indicate the 32P-labeled substrates, and the bottom panel shows the Coomassie Blue staining pattern of this gel (CBB). B, JNK phosphorylates c-Myc at Ser-62 and Ser-71 as detected by phospho-c-Myc-specific antibodies. Anti-phospho-c-Myc (Ser-71) polyclonal antibody (-P-S71) was raised against a synthetic phosphopeptide (NH2-RSGLC(phospho-S)PSYVA-COOH) corresponding to amino acids 66-76 in the human c-Myc protein. After the in vitro kinase assay, phosphorylated c-Myc was detected by subsequent Western blot analysis using anti-phospho-c-Myc(Thr-58/Ser-62) polyclonal antibody (-P-T58/S62) and -P-S71. Arrowheads indicate the phosphorylated substrates, and the bottom panel shows the Coomassie Blue staining pattern of these substrates (CBB). C, in vivo phosphorylation of c-Myc at Ser-71 was detected by Western blotting. c-Myc was transiently expressed in 293T cells and either irradiated by UVC (300 J/m2) or not irradiated. After a further 30 min of incubation, the nuclear fraction was prepared, and phosphorylation of c-Myc at Ser-71 in the treated cells was detected by Western blot analysis using -P-S71 (top panel). The expression level of c-Myc was determined with anti-human c-Myc antibody (bottom panel).

To identify the phosphorylation of Ser-62 and Ser-71 in c-Myc, we generated a phosphorylated Ser-71-reactive polyclonal antibody (-P-Ser-71). Then, we utilized the -P-Thr-58/Ser-62 and -P-Ser71 polyclonal antibodies and examined the phosphorylation of c-Myc by Western blot analysis (Fig. 2B). This experiment directly clarified that JNK phosphorylated c-Myc at both Ser-62 and Ser-71 in vitro, and the substitution of Ala for Ser-62 and Ser-71 resulted in the disappearance of phosphorylated signals by JNK. By contrast, Western blot analysis showed that ERK2 did not phosphorylate c-Myc at Ser-71 and that the loss of Ser-71 did not affect the phosphorylation of c-Myc at Ser-62 by ERK2². However, we noticed that loss of Ser-62 did not abolish the phosphorylation of c-Myc by ERK2 completely, whereas a double mutation of Ser-62 and Ser-71 did (see Fig. 2A). Thus, ERK2 might phosphorylate c-Myc at some residue(s) other than Ser-62 that might be affected by Ser-71. Next, the in vivo phosphorylation of c-Myc at Ser-71 was examined by Western blot analysis using -P-Ser-71. Consistent with a previous report (34), phosphorylation of c-Myc at Ser-71 was observed in untreated c-Myc-expressing cells (Fig. 2C, +c-myc, UV). Moreover, phosphorylation of Ser-71 was relatively increased after UV irradiation (Fig. 2C, +c-myc, UV+). Collectively, these results indicate that Ser-62 and Ser-71 in c-Myc are target phosphoacceptor sites for JNK.

Interaction between c-Myc and JNK-- To test the physical interaction of JNK and c-Myc, we conducted an in vitro pull-down assay using GST-fusion proteins and in vitro translated JNK1 (Fig. 3A). In vitro translated JNK1 was co-precipitated with GST-c-Myc (1-262) but not with GST-s-Myc (1-337) or GST-Max. Furthermore, the binding ability for JNK1 with GST-c-Myc (1-262) was similar to that for GST-c-Jun (1-92). The in vivo association of JNK and c-Myc was confirmed by coimmunoprecipitation assay in the transient transfection system using 293T cells (Fig. 3B). Flag-tagged c-Myc was detected in the immunocomplex precipitated by anti-JNK1 antibody, and the formation was likely formed independent of the stress stimuli, such as UV. The amount of c-Myc coimmunoprecipitated by anti-JNK antibody was not high, but the efficiency was similar to that by anti-ERK2 antibody.² No corresponding signals were detected in the immunocomplex by control and anti-p38 antibody. Collectively, these results show that JNK interacts with c-Myc protein but not with s-Myc and Max.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Interaction between c-Myc and JNK. A, binding between JNK and c-Myc was examined by pull-down assay. In vitro translated JNK1 was mixed with each GST protein, and the precipitated complex was resolved by SDS-PAGE (top panel). An arrowhead indicates the in vitro translated JNK1. The bottom panel shows the Coomassie Blue staining pattern of this gel (CBB), and asterisks indicate GST proteins. B, complex formation of JNK and c-Myc in vivo was examined by coimmunoprecipitation assay. Flag-tagged c-Myc was transiently expressed in 293T cells and either irradiated by UVC (300 J/m2) or not irradiated. After a further 30 min of incubation in the presence of the chemical cross-linker dithiobis(succinimidylpropionate) (1.5 mg/ml), cells were lysed, and JNK1 and p38 were immunoprecipitated by each antibody (C-17 for JNK1, C-20-G for p38 and species-specific control antibody). Co-precipitated flag-tagged c-Myc was detected by anti-Flag M5 monoclonal antibody. As a positive control, 1% of the UV-subjected cell lysate used for immunoprecipitation was also assayed. Arrowheads indicate the immunoprecipitated Flag-c-Myc and cross-reacted immunoglobulin heavy chain (IgH).

Ser-62 and Ser-71 for the Proapoptotic Activity of c-Myc-- To clarify the effect(s) of phosphorylation by JNK on c-Myc, we compared the functions of mutant and wild type c-Myc in stress-triggered apoptosis. First, we examined the proapoptotic activities of c-Mycs in UV-induced apoptosis by transient transfection assay in HeLa cells and the expression of exogenous c-Mycs was detected by Western blot analysis (Fig. 4A). In this case, after UV irradiation, 44% of c-Myc (S62A)-expressing cells rapidly underwent apoptosis the same as wild type c-Myc. However, the ratio of apoptotic cells in mutated c-Myc (S71A) was reduced to 35%, and in double mutated c-Myc (S62A/S71A)-expressing cells, it was about 30%. To test the contribution of the JNK pathway in c-Myc-mediated apoptosis, we used a dominant-negative mutant of JNK1 called JNK(APF). JNK(APF), in which the phosphorylation site Thr-Pro-Tyr is changed to Ala-Pro-Phe, is thought to behave as a competitive inhibitor of JNK signaling (17, 35, 36). We established a JNK(APF)-expressing stable HeLa cell line, HeLa/APF, in which activation of JNK but not ERK2 was selectively suppressed after UV irradiation (Fig. 4B). Compared with in parental HeLa cells, wild type c-Myc-mediated apoptosis was suppressed in HeLa/APF, but c-Myc (S62A/S71A)-mediated apoptosis caused by UV was not (Fig. 4C). Thus, the suppressive effect of JNK(APF) was restricted to wild type c-Myc-dependent apoptosis. These results indicate that phosphorylation at both Ser-62 and Ser-71 in c-Myc is functionally important for the apoptosis induced by the JNK pathway, and phosphorylation of Ser-71 might be more significant than that of Ser-62 for the proapoptotic effects of c-Myc.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   JNK signaling associated with c-Myc-mediated apoptosis. A, the apoptosis-inducing ability of mutant c-Myc was examined by transient assay in HeLa cells. Each c-myc-expressing pc3GF plasmid was transfected in HeLa cells, and after 2 days, cells were irradiated by UVC (100 J/m2). The expression of each c-Myc was detected by Western blotting as shown in the top panel. At 5 h after irradiation, cells were fixed, and cell morphologies of EGFP-positive cells were examined under a fluorescence microscope. More than 300 cells were analyzed. B, expression of jnk1(APF) was detected by Northern blotting in stable HeLa transfectant HeLa/APF cells (top panel). The arrow indicates exogenous jnk1(APF) mRNA, and the arrowhead indicates endogenous jnk1 mRNA. Activations of JNK1 and ERK2 by UV irradiation (100 J/m2) in HeLa and HeLa/APF cells were evaluated by immunocomplex kinase assay (for JNK; bottom left) and by in-gel kinase assay (for ERK2; bottom right), respectively. Kinase activities were determined from the results of two independent experiments, and values relative to those of nonirradiated cells were shown. C, apoptosis by c-Myc and c-Myc(S62A/S71A) in HeLa/APF cells. c-Myc and c-Myc(S62A/S71A) were transiently expressed in HeLa and HeLa/APF cells, and apoptosis induction was determined as in Fig. 4A.

Previous studies showed that UV- and Taxol-induced cell death required JNK activation (35, 36). To demonstrate the involvement of c-Myc during cell death triggered by these stimuli, we next established wild type c-Myc (CM-8 and CM-9) and mutated c-Myc (S62A/S71A)-expressing NIH3T3 clones (S6271A-4, and S6271A-13) in which comparable amounts of c-Myc was expressed at the protein level (Fig. 5A, left). Consistent with the transient assay with HeLa cells, the wild type c-Myc-expressing clone CM-9 was highly sensitive to UV irradiation, but the mutated c-Myc-expressing clone S6271A-13 showed a sensitivity similar to that of the parent NIH3T3 cells (Fig. 5A, right). We also confirmed that UVC and Taxol, but not serum deprivation, transiently activated the JNK pathway in parental NIH3T3 cells, as judged by Western blot analysis of phosphorylated JNK (Fig. 5B). Cell sensitivities against these stimuli were examined, and the results clearly showed that mutated c-Myc (S62A/S71A)-expressing cells were not sensitized, unlike wild type c-Myc-expressing cells against JNK-activating UV and Taxol (Fig. 5C). Interestingly, we also found that c-Myc (S62A/S71A)-expressing cells were sensitive to serum deprivation-induced apoptosis the same as wild type c-Myc-expressing cells (Fig. 5C, right graph). These observations indicated that whereas c-Myc mediated the apoptotic JNK signaling through Ser-62 and Ser-71 phosphorylation, such phosphorylation of c-Myc would not be required for the apoptosis induced by serum deprivation.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Ser-62 and Ser-71 residues of c-Myc are required for the cell sensitization to UV and Taxol. A, expression of human c-Myc protein was detected by Western blotting in stable NIH3T3 transfectants using anti-human c-Myc antibody (9E10) (left). CM-8 and -9 clones express wild type human c-Myc, and S6271A-4 and -13 clones express Ser-62 and Ser-71 mutated human c-Myc. Parental NIH3T3, wild type c-Myc transfectant CM-9, and (Ser-62/Ser-71) double mutated c-Myc transfectant S6271A-13 cells (7 × 105 cells/60-mm dish) were irradiated by UVC (100 J/m2). After 20 h of culture in the same medium containing 10% FBS, cell morphology was photographed (× 100) (right). B, transient activation of JNK pathway. Parental NIH3T3 cells (7 × 105 cells/60-mm dish) were treated with each stimulus, and then cell lysates were prepared at the indicated times. Activated JNK (P-JNK) and total expression of JNK1 were examined with anti-phospho-JNK and anti-JNK1 antibodies, respectively. C, sensitivities of parental NIH3T3 and c-Myc-expressing transfectants. Parental NIH3T3 and stable c-Myc-expressing transfectant cells (7 × 105 cells/60-mm dish) were treated with UV (100 J/m2), Taxol (10 µM), and serum deprivation (0.1% FBS). Cell viability (at 20 h for UV and Taxol and at 44 h for 0.1% FBS) was determined by trypan blue dye exclusion assay, and the percentage of cell death was determined. Data show the means and standard deviations from the results of two or three independent experiments. D, the ERK2 pathway is not associated with c-Myc-mediated apoptosis. Parental NIH3T3 cells were treated with each stimulus, as in B, or cells were also co-treated with a MEK inhibitor, PD98059, which was added 30 min before UV and Taxol treatment. Then, cell lysates were prepared at the indicated times, and the expression level of the activated ERK (P-p42/44 ERK) and total ERK2 were examined with anti-phospho-ERK and with anti-ERK2 antibodies, respectively (left panels). In the right panel, wild type c-Myc-expressing CM-9 cells were treated with UV or Taxol in the presence of PD98059 as in the left panels, and after 20 h, cell viability was determined by trypan blue dye exclusion assay. Data show the means and standard deviations from the results of two independent experiments.

Because ERK2 is activated by UV and phosphorylates c-Myc at Ser-62, we examined the activation of the MEK-ERK pathway by Taxol treatment using anti-phospho-ERK-specific antibody in the presence or absence of a MEK inhibitor PD98059. This Western blot analysis showed that UV irradiation (100 J/m²) transiently activated the phosphorylation of p42/44 ERK at 30 min (Fig. 5D, top left) as shown earlier in Fig. 1A, but Taxol (10 µM) did not activate the ERK pathway but rather inhibited the phosphorylation of p42/44 ERK (Fig. 5D, bottom left). In addition, we found that PD98059 (10 µM) did not inhibit the apoptosis induction by UV and Taxol in wild type c-Myc-expressing CM-9 cells, despite the strong inhibition of ERK phosphorylation by PD98059 (Fig. 5D). Thus, the activation of the ERK pathway did not correlate with c-Myc-dependent apoptosis by Taxol, suggesting that the contribution of MEK-ERK pathway was negligible in c-Myc-mediated apoptosis. Collectively, these results suggest that phosphorylation at both Ser-62 and Ser-71 is required for the cell sensitization by c-Myc to JNK-activating stimuli, but not for the apoptosis caused by other stimuli, such as serum deprivation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The expression of myc family genes is frequently deregulated in cancer, affecting cell transformation by deregulating the cell cycle progression (1). The deregulated expression of c-myc or s-myc also sensitizes cells to apoptotic stimuli (2). In this study, we found that the proapoptotic function of c-Myc was regulated by JNK pathway, and the effect of JNK was dependent on the Ser-62 and Ser-71 residues in c-Myc.

Phosphorylational regulation of c-Myc has been investigated regarding cell growth and transformation (34, 37), and previous findings have suggested a functional association between phosphorylation at Thr-58/Ser-62 by glycogen synthase kinase 3, cyclin-dependent kinase, and ERK2 in cell proliferation and cell cycle regulation, although controversy surrounds the role of Thr-58/Ser-62 phosphorylation in c-Myc transactivation activity (33, 34, 37-40). Our unpublished study² also indicated that the functional role of Ser-62/Ser-71 residues in c-Myc transactivation activity appeared to depend on the cell type. This contrariety may depend on the assay system, because various Myc-binding and regulating molecules, such as BIN-I, TRRAP, p107, YY-I, MizI, and AP-2, were identified, and c-Myc both activates and represses various target genes (41). Furthermore, the transactivation domain of c-Myc is reported to be involved in the protein stabilization/degradation mechanism as part of the ubiquitin-proteasome system (42), and phosphorylation at Thr-58 and Ser-62 is likely to participate in the c-Myc protein degradation/stabilization (43, 44). Therefore, the effect of phosphorylation at Thr-58/Ser-62 may require another regulator(s) in the cells, and further investigation will clarify the role of phosphorylation at the N-terminal transactivation domain in the biological effect of c-Myc in each assay system.

Unlike for the Thr-58 and Ser-62 residues of c-Myc, the significance of phosphorylation at Ser-71 in vivo under stress stimuli is unknown, and the corresponding kinase(s) is yet to be identified. Here, we investigated the role of phosphorylation at the N-terminal transactivation domain of c-Myc under apoptotic stress conditions, such as following UV treatment. As a result, we demonstrated that JNK could phosphorylate c-Myc at Ser-62 and Ser-71. Moreover, we showed that UV irradiation, Taxol, and cisplatin,² which activate JNK pathway, required the Ser-62 and Ser-71 residues in c-Myc for efficient cell sensitization to apoptosis. Interestingly, our data shown in Figs. 4 and 5D suggest that ERK2, which phosphorylates c-Myc at Ser-62, is not associated with apoptosis and that a single mutation of c-Myc at Ser-62 shows no effect on the proapoptotic activity of c-Myc. Thus, the proapoptotic activity of c-Myc would be controlled by some extracellular stress-triggered signaling through phosphorylation at both Ser-62 and Ser-71. The region at amino acids 49-100 in c-Myc is not conserved, nor especially, is the Pro-72 residue after Ser-71, among other Myc proteins, such as s-Myc and N-Myc. The Pro residue after the Thr/Ser residue is required for the phosphorylation by MAPK family protein kinases including JNK (32). Therefore, we speculate that JNK would not phosphorylate s-Myc or other Myc family proteins and that the JNK pathway regulates only the c-Myc-mediated effect. Consistent with our hypothesis, a recent study showed that only c-Myc enhanced the chemosensitivity, whereas N- and L-Myc produced a significantly resistant phenotype, although the three Mycs were equally proficient at accelerating the apoptosis of 32D cells in response to interleukin-3 withdrawal (45). In addition, another study also demonstrated that c-MycS, a transactivation-defective form of c-Myc in which Ser-62 and Ser-71 are deleted, was able to induce apoptosis on serum deprivation (46). From these observations, it appears that the removal of cell survival signals might trigger different mechanisms of stress stimuli for the activation of apoptotic machinery. As c-MycS fails to transactivate through E-box elements but retains transrepression activity, the apoptosis caused by serum deprivation in c-Myc-expressing cells may be mediated in part by the transrepression activity of c-Myc.

JNK has been implicated in the regulation of apoptosis, but the role of the JNK pathway in apoptosis has been controversial (24, 47-49). One of the major targets of JNK is AP-1, and the positive regulation of AP-1-mediated gene expression, including Fas ligand up-regulation, is involved in the JNK-mediated apoptosis (50, 51). However, the functional effect of AP-1 in physiological apoptosis is unclear (28, 29). JNKs have at least 10 isoforms, which are classified into three groups, JNK1, JNK2, JNK3, based on their primary structures, and the substrate specificity of each isoform is relatively different (52). Because the cell response, cell death, or survival/proliferation will depend on the various effector activities, it is important to clarify the target molecule(s) for the effect of JNK pathway in the different systems. Intriguingly, among JNK family proteins, we observed that JNK3 phosphorylated c-Myc more effectively than did JNK1.² These observations suggest the involvement of c-Myc in JNK3-mediated apoptosis under physiological conditions.

In conclusion, we demonstrated a selective association between JNK and c-Myc and showed that JNK directly connects the stress signals with c-Myc to stimulate proapoptotic activity of c-Myc through phosphorylation at Ser-62 and Ser-71. However, we also demonstrated that c-Myc (S62A/S71A) can induce JNK-independent apoptosis. Our observations suggest that the proapoptotic activities of Myc family proteins would be regulated by distinct signal pathways that depend on the apoptotic stimulus and that the contribution of the target gene expression/repression by Myc may differ. Our study also suggests that the JNK-activating signals will become apoptotic signals when c-Myc is deregulated in cells such as tumor cells. Further explorations focused on both the regulation of Myc by upstream signals and identifying the effector(s) in each situation will provide new insights in the understanding of the Myc-mediated cell sensitization for apoptosis induction and the role of JNK signaling in apoptosis.

	ACKNOWLEDGEMENTS

We thank Drs. T. Tsuruo, E. Nishida, M. Natio, H. Seimiya, and N. Fujita for providing reagents and valuable suggestions.

	FOOTNOTES

^* This work was supported by grants from the Ministry of Education, Science, Sports and Culture of Japan for Cancer Research (to K. N. and Y. K.), by the Haraguchi Memorial Fund (to K. N.), and in part by a grant-in-aid from the Ministry of Health and Welfare of Japan for the Second-term Comprehensive 10-year Strategy for Cancer Control (to Y. K).The costs of publication of this article were defrayed in part by the payment of page charges. The 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. Tel.: 81-3-3542-2511, ext. 4600; Fax: 81-3-3546-1369; E-mail: ykuchino@ncc.go.jp.

² K. Noguchi, unpublished observation.

	ABBREVIATIONS

The abbreviations used are: JNK, c-Jun N-terminal kinase; MAPK, mitogen-activated kinase; ERK, extracellular signal-regulated kinase; MEK, MAPK/ERK kinase; EGFP, enhanced green fluorescence protein; GST, glutathione S-transferase; PAGE, polyacrylamide gel electrophoresis; DTT, dithiothreitol; PMSF, phenylmethylsulfonyl fluoride; FBS, fetal bovine serum.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES