MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sreeramaneni, R.
Right arrow Articles by Inoue, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sreeramaneni, R.
Right arrow Articles by Inoue, K.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, January 2005, p. 220-232, Vol. 25, No. 1
0270-7306/05/$08.00+0     doi:10.1128/MCB.25.1.220-232.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Ras-Raf-Arf Signaling Critically Depends on the Dmp1 Transcription Factor

Ramesh Sreeramaneni,1,{dagger} Asif Chaudhry,1,{dagger} Martin McMahon,2 Charles J. Sherr,3 and Kazushi Inoue1*

Departments of Pathology and Cancer Biology, Wake Forest University Health Sciences, Winston-Salem, North Carolina,1 Cancer Research Institute and Department of Cellular and Molecular Pharmacology, UCSF/Mt. Zion Comprehensive Cancer Center, San Francisco, California,2 Howard Hughes Medical Institute, Department of Genetics and Tumor Cell Biology, St. Jude Children's Research Hospital, Memphis, Tennessee3

Received 6 July 2004/ Returned for modification 3 September 2004/ Accepted 8 October 2004


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Dmp1 prevents tumor formation by activating the Arf-p53 pathway. In cultured primary cells, the Dmp1 promoter was efficiently activated by oncogenic Ha-RasV12, but not by overexpressed c-Myc or E2F-1. Dmp1 promoter activation by RasV12 depended on Raf-MEK-ERK signaling. Induction of p19Arf and p21Cip1 by oncogenic Raf was compromised in Dmp1-null cells, which were resistant to Raf-mediated premature senescence. A RasV12-responsive element was mapped to the 5' leader sequence of the murine Dmp1 promoter, where endogenous Fos and Jun family proteins bind. Dmp1 promoter activation by RasV12 was strikingly impaired in c-Jun as well as in JunB knock-down cells, suggesting the critical role of Jun proteins in the activation of the Dmp1 promoter. A RasV12-responsive element was mapped to the unique Dmp1/Ets site on the Arf promoter, where endogenous Dmp1 proteins bind upon oncogenic Raf activation. Therefore, activation of Arf by Ras/Raf signaling is indirectly mediated by Dmp1, explaining why Dmp1-null primary cells are highly susceptible to Ras-induced transformation. Our data indicate the presence of the novel Jun-Dmp1 pathway that directly links oncogenic Ras-Raf signaling and p19Arf, independent of the classical cyclin D1/Cdk4-Rb-E2F pathway.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The INK4a-ARF locus on human chromosome 9p21 is disrupted in approximately 40% of human cancers (42). This unusual locus encodes two distinct tumor suppressor proteins, p16INK4a and p14ARF (p19Arf in the mouse), encoded in part via alternative reading frames. p16Ink4a binds to cyclin-dependent kinase 4 (Cdk4) to inhibit Rb phosphorylation, whereas p19Arf binds to the p53 negative regulator, Mdm2, thereby stabilizing and activating p53 (26, 47). Arf is induced by potentially harmful growth-promoting signals stemming from overexpression of a variety of oncoproteins, including c-Myc, E2F-1, mutated Ras, v-Abl, and ß-catenin (47). This forces incipient cancer cells to undergo p53-dependent proliferative arrest or apoptosis, providing a powerful mode of tumor suppression. In turn, Arf-null mice are highly prone to spontaneous tumor development and die of various forms of cancer by 15 months of age (21, 22). Recently, the creation of Arf-GFP knock-in mice has provided direct experimental evidence that the Arf promoter monitors latent oncogenic signals in vivo (56).

How Arf responds to oncogenic Ras signaling remains unclear. Ras family proteins play crucial roles in the control of cell growth and differentiation (29). Overexpression of activated Ras initiates DNA synthesis independent of growth factor stimulation. In immortal rodent cell lines, transformation by oncogenic Ras involves its ability to bind and activate a series of effector proteins, including Raf-1, phosphoinositide 3-OH kinase [PI(3)K], and Ral-GDS (23). Each of these molecules, in turn, activates distinct downstream targets, thereby producing different aspects of the transformed phenotype. The Ras-Raf interaction initiates the mitogen-activated protein kinase (MAPK) cascade, which involves the sequential activation of a series of protein kinases that transmit mitogenic signals to nuclear transcription factors. These kinases include Raf-1, the MEKs (MEK1 and MEK2), and the ERKs (ERK1 and ERK2). On the other hand, the ability of Ras to activate PI(3)K promotes membrane ruffling (20), and the Ral-GDS proteins act as exchange factors that can activate the Ral family of small GTPases (50). Although each of these effector pathways contributes to the transforming activity of Ras in established rodent fibroblast cell lines, activation of the Raf-MEK-ERK pathway is sufficient for transformation (10). Paradoxically, sustained overexpression of oncogenic Ras and its various effectors in nonimmortalized cells has the capacity to elicit irreversible cell cycle arrest by upregulating the levels of p16Ink4a, p19Arf, and p53 in mice and p16INK4a and p53 in humans (29, 33, 44, 48, 57). The ability of oncogenic Ras to induce premature senescence depends on the activity of the Raf-MEK-ERK pathway that mediates proliferation (25) but is nullified in primary mouse embryo fibroblasts (MEFs) lacking either Arf or p53 (21, 33, 48).

Among known Arf activators, the Dmp1 transcription factor (cyclin D-interacting Myb-like protein 1) is a bona fide tumor suppressor (18, 19). Dmp1 was originally isolated in a yeast two-hybrid screen of a murine T-lymphocyte library with cyclin D2 as bait (14). The protein binds to nonameric CCCG(G/T)ATG(T/C) DNA consensus sequences, a subset of which is also bound by proteins of the Ets family. Dmp1 can physically interact with any of the three D-type cyclins, each of which can interfere, in a Cdk4-independent manner, with Dmp1's ability to bind to DNA (15). Overexpression of Dmp1 in mouse fibroblasts arrests cell cycle progression, an effect that can be overridden by coexpression of D-type cyclins (15). Importantly, Dmp1 directly binds to the Arf promoter to activate its expression, thereby inducing p53-dependent cell cycle arrest (17).

Several lines of evidence have implicated Dmp1 in the process by which Ras induces Arf and p53. When primary Dmp1-null MEFs were explanted into culture and continuously passaged, p19Arf and p53 levels remained uncharacteristically low and the cells exhibited a prolonged proliferative capacity, readily yielding established cell lines that retained wild-type Arf and p53. Such cells were susceptible to transformation by oncogenic Ras alone without any requirement for an immortalizing oncogene, such as Myc or adenovirus E1A. Thus, the activity of the Arf-p53 pathway is strikingly impaired in Dmp1-null cells (18).

Dmp1-null mice are prone to spontaneous tumor development in their second year of life, and tumor formation was accelerated when the animals were neonatally treated with ionizing radiation or dimethylbenzanthracene (18, 19), a carcinogen that induces Ras mutations in vivo (37). When crossed onto a Dmp1+/– or Dmp1–/– background, lymphomas induced by an Eµ-Myc transgene were greatly accelerated with no differences between cohorts lacking one or two Dmp1 alleles. The retention and expression of the wild-type Dmp1 allele in tumors arising in heterozygotes indicated that Dmp1 is haplo-insufficient for tumor suppression (19, 38). Interestingly, the combined frequencies of p53 mutation and Arf deletion in the Dmp1–/– and Dmp1+/– lymphomas were significantly lower than those in Dmp1+/+ tumors (~14% versus ~50%). Thus, Dmp1 is a physiological regulator of the Arf-p53 pathway in vivo (19). The present studies were undertaken in an attempt to define the mechanism(s) by which Ras induces Arf. Here we show that Dmp1 is a key mediator of this process.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture and reporter assays. Wild-type and Dmp1-null MEFs were established from 13.5-day-old embryos as previously described (18). Cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 2 mM glutamine, 0.1 mM nonessential amino acids, 55 µM 2-mercaptoethanol, and 10 µg of gentamicin/ml. BALB/3T3, C33A, and IMR-90 cells were purchased from the American Type Culture Collection. For reporter assays, 1.5 x 105 cells were seeded into 60-mm diameter culture dishes 24 h before transfection. In order to study the responsiveness of the Dmp1 and Arf promoters to oncogenic Ha-RasV12, 4 µg of luciferase reporter DNA was cotransfected with 0.3 to 1 µg of Ras expression vectors and 4 µg of internal control actin promoter-secreted endocrine alkaline phosphatase vector (a gift from Michael Ostrowski, Ohio State University). Genejuice (Novagen) was used in all transfections. Specific MAPK inhibitors U0126 (for MEK/ERK), SP600125 (for JNK/SAPK), (L)-JNKI1 (for JNK/SAPK), and SB203580 (for p38) were purchased from Calbiochem. All of them were used at 10 µM.

Plasmid DNA. pBabepuro-Ha-RasV12, RasV12S35, RasV12G37, and RasV12C40 viral vectors were obtained from Scott Lowe, Cold Spring Harbor Laboratory, and from Christopher Counter, Duke University (13). Expression vectors for Ha-RasV12, RasV12S35, RasV12G37, and RasV12C40 were created by recloning the cDNA from the pBabepuro vector into the pcDNA3 vector (Invitrogen). Retroviral expression vectors for {Delta}Raf:ER[DD] and empty estrogen receptor (ER) vector were described previously (28). Rc/CMV-c-Myc was obtained from John Cleveland, St. Jude Children's Research Hospital. pcDNA-E2F-1 was received from Joseph Nevins, Duke University. Expression vectors for c-Fos and c-Jun family proteins driven by the cytomegalovirus (CMV) promoter were obtained from Tom Curran, St. Jude Children's Research Hospital. Expression vectors for D-type cyclins, cyclin A, and cyclin H were described previously (15).

Molecular cloning of the murine Dmp1 promoter. The murine Dmp1 promoter was cloned from the Bacterial Artificial Chromosome library derived from 129/Svj mice (Mouse ES Release II; Genome Systems Inc.) with 60-bp synthetic oligonucleotides covering the 5' end of the murine Dmp1 cDNA (14). A 1.8-kb PstI fragment hybridizing with the probe was cloned into the artificially created PstI site of the pGL2-basic vector to generate the –1787 PstI promoter construct. In order to create deletion mutants, the plasmid DNA was digested with SmaI plus SpeI, BstXI, NsiI, or ApaI, and the gel-purified DNA fragments were filled with Klenow enzyme or T4 DNA polymerase and then ligated. The transcription initiation sites on the murine Dmp1 promoter were determined by using the SMART rapid amplification of cDNA ends (RACE) kit (Clontech) with total RNA isolated from NIH 3T3 cells.

In vitro mutagenesis. The murine Dmp1 promoter deletion point mutants were generated by use of an in vitro mutagenesis kit (Stratagene). In order to introduce point mutations at the Ets site (–9 to –17) on the murine Dmp1 promoter, PCR was performed by using oligonucleotides 5'-GCCTCGCGGCTCCGTCGTAGGTGGCTGGTTGCGC-3' and its reverse complementary sequence. The mutated Ets site is underlined. In order to delete the Ets site, oligonucleotide 5'-GCCTCGCGGCTCCGTGGTGGCTGGTTGCGC-3' and its reverse complementary sequence were used. In order to delete the 5' leader sequence on the Dmp1 promoter, oligonucleotide 5'-TGGCTGGTTGCGCTGCAGGCTAGCTCGA-3' and its reverse complementary sequence were used. In order to mutate the AP-1 and AML1 consensus-like sequences on the 5'-untranslated region of the murine Dmp1 promoter, oligonucleotide 5'-GGCTGGTTGCGCTCGAAAACCCCAGCTGCAGGC-3' and its reverse complementary strand and 5'-GGTTGCGCTCGCTCAATCTAGCTGCAGGCTAGCTCG-3' and its reverse complementary strand were used, respectively. The mutated consensus sequences for AP-1 and AML1 are underlined. All the Dmp1 promoter mutants were subjected to sequencing analysis to confirm the presence of mutation or deletion.

Retroviruses and RNAi. Ecotropic retroviruses encoding Ha-RasV12, {Delta}Raf:ER[DD], or empty (ER) vector were prepared by transfecting 293T cells with a helper ecotropic retrovirus plasmid defective in psi-2 packaging sequences together with pBabepuro vectors containing Ha-RasV12, or {Delta}Raf:ER[DD] or with empty (ER) vector. Viruses were harvested every 6 h for 24 to 72 h after transfection, pooled, filtered, and stored at –80°C until use. Ecotropic retroviruses to knock-down mouse c-Jun, JunB, JunD, c-Fos, Fra-1, or Fra-2 were prepared by using the pSuper RNA interference (RNAi) system (Oligoengine). The 19-bp target sequence was 5'-GCGCATGAGGAACCGCATT-3' for c-Jun, 5'-GACCAGGAGCGCATCAAAG-3' for JunB, 5'-AAGCCAGAACACCGAGCTG-3' for JunD, 5'-GCGGAGACAGATCAACTTG-3' for c-Fos, 5'-ATTGGAGGATGAGAAATCG-3' for Fra-1, and 5'-TCAACGCCATCACCACCAG-3' for Fra-2. The effectiveness of down regulation of each gene product was studied by Western blotting with specific antibodies. In order to create growth curves of MEFs with activated c-Raf, Dmp1+/+ and Dmp1–/– MEFs were infected with retroviruses expressing {Delta}Raf:ER or empty ER. Forty-eight hours after infection, cells were selected with 2 µg of puromycin/ml for 48 h. A total of 105 puromycin-resistant cells were seeded in 60-mm-diameter culture dishes and then treated with 1 µM 4-hydroxytamoxifen (4-HT; Sigma). The medium was changed every 48 h with fresh 4-HT, and cells were counted.

Northern and Western blotting. Total RNA was extracted by using TRIzol (Invitrogen) from MEFs infected with retroviruses encoding Ha-RasV12, {Delta}Raf:ER[DD], or empty vector. Northern blotting was performed with 10 µg of total RNA by using Turboblotter (Schleicher & Schuell). The filter was hybridized with a murine Dmp1-specific probe (KpnI-NcoI fragment; 426 bp) and then with a mouse ß-actin-specific probe. For Western blotting, proteins were extracted with EBC buffer (14) with proteinase inhibitor cocktail III and leupeptin (Calbiochem). For detection of Dmp1, affinity-purified RAF antibodies (14) were used. The following antibodies (all from Santa Cruz Biotechnology) were used for the detection of AP-1 proteins in Western blotting, an electrophoretic mobility shift assay (EMSA), and a chromatin immunoprecipitation (ChIP) assay: c-Fos (sc-52x), FosB (sc-48x), Fra-1 (sc-605x), Fra-2 (sc-171x), c-Jun (sc-1694x), phospho-c-Jun (sc-7981x), JunB (sc-8051x), JunD (sc-74x), Fos family (sc-253x), Jun family (sc-44x), ATF-1 (sc-243x), ATF-2 (sc-187x), ATF-3 (sc-188x), CREB1 (sc-58x), and CREB 2 (sc-200x). For Western blotting of other proteins, the following antibodies were used: p-ERK (sc-7383), cyclin D1 (sc-450), cyclin D2 (sc-34B1-3), cyclin D3 (sc-18B-10), cyclin A (sc-751), cyclin H (D-10), p16Ink4a (sc-1207), p21Cip1 (sc-6246), and actin (sc-1615). For detection of p19Arf, affinity-purified rabbit antibodies were used (17, 55). For detection of p53, Ab7 (Oncogene Science) was used.

EMSA. In order to detect proteins that bind to the 5'-untranslated region of the murine Dmp1 promoter, lysates were prepared from MEFs expressing {Delta}Raf:ER[DD] with or without treatment with 2 µM 4-HT for 16 h. The lysate was incubated with 32P-labeled oligonucleotide probe covering the 5' leader sequence of the murine DMP1 cDNA obtained by annealing sense oligonucleotide 5'-TGGTTGCGCTCGCTCACCCCAGCTGCAGCCA-3' and its reverse complementary sequence (the AP-1-like sequence is underlined, and the AML1 consensus sequence is italicized). For competition assays, a 100-fold excess of unlabeled oligonucleotides was added to reaction mixtures before probe incubation. To verify the identity of the proteins in shifted complexes, reaction mixtures were incubated with control nonimmune rabbit serum or with specific antibodies to Fos, Jun, ATF, and CREB family proteins (all from Santa Cruz Biotechnology).

ChIP. ChIP were performed as described previously (8). Briefly, MEFs expressing {Delta}Raf:ER[DD] were either left untreated or treated with 2 µM 4-HT for 0, 8, 16, and 24 h. The lysates were precipitated with specific antibodies to Fos, Jun family proteins, or with anti-Dmp1 antibody (RAF) and incubated at 4°C overnight. The immunoprecipitated DNA was detected by PCR, including 1 µCi of [{alpha}-32P]dATP (Amersham Pharmacia) separated on a 10% nondenaturing polyacrylamide gel. For detection of the endogenous Dmp1 on the Arf promoter, sense primer 5'-AAAGGGCGCAGCTACTGCTA-3' and anti-sense primer 5'-TCTTTGCTCCACGCCCATCT-3' were used. For detection of the endogenous AP-1 family transcription factors on the murine Dmp1 promoter, sense primer 5'-CTCGCGGCTCCGTTTCCG-3' and antisense primer 5'-CCTGAAGGTTCCATCGCACT-3' were used. For the control amplification of 2-kb upstream sequence on the Dmp1 promoter, sense primer 5'-TCTCCATAGCAATGCCCTTTAC-3' and antisense primer 5'-CGAGCCATTTGGGTATGTGTA-3' were used.

Nucleotide sequence accession number. The gene accession number for the murine Dmp1 promoter used in this study is AY702209.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Molecular cloning of the murine Dmp1 promoter and its responsiveness to oncogenes. In order to study regulation of Dmp1 transcription, we cloned the murine Dmp1 promoter from a commercially available murine genomic library using synthetic oligonucleotide probes covering the 60 bp of the 5' end of the murine Dmp1 cDNA. A 1.8-kb PstI fragment hybridizing with the probe was cloned into the pGL2-basic vector (Promega), and a series of deletion mutants were created (Fig. 1A). The Dmp1 promoter has multiple transcription initiation sites; however, the most commonly used site was mapped to the guanine residue 2 bp upstream from the 5' end of the Dmp1 cDNA first isolated from CTLL-2 cells (14). We tested basal Dmp1 promoter-reporter activity in mouse NIH 3T3 cells, BALB/3T3 cells, and also in human carcinoma C33A cells. Deletion from the –1787 PstI site to the –374 NsiI site did not influence basal promoter activity, while deletion from the –374 NsiI site to the –88 ApaI site (Fig. 1, Del 5) dramatically decreased basal Dmp1 promoter function in the three cell lines.



View larger version (44K):
[in this window]
[in a new window]
  		FIG. 1. Cloning of the murine Dmp1 promoter and the sequence of the proximal region. The murine Dmp1 promoter was cloned from the Bacterial Artificial Chromosome library derived from 129/Svj mice (Mouse ES Release II; Genome Systems Inc.) with 60-bp synthetic oligonucleotides covering the 5' end of the murine Dmp1 cDNA. The 1.8-kb PstI fragment hybridizing with the probe was cloned into the polylinker site of the pGL2-basic vector (Promega). (A) Deletion analysis of the Dmp1 promoter. NIH 3T3, BALB/3T3, and C33A cells were transfected with 4 µg of murine Dmp1 promoter-luciferase constructs and 4 µg of secreted alkaline phosphatase expression vector driven by the actin promoter. Relative luciferase levels corrected by the internal control alkaline phosphatase levels are shown. Deletion of the promoter up to the NsiI site did not influence the endogenous promoter activity, while deletion of the promoter to the ApaI site resulted in a 4- to 10-fold decrease in the relative luciferase levels. (B) Nucleotide sequences of the proximal region (–414 to +65). The major transcription initiation sites determined by 5'-RACE are shown in large capital letters. Consensus sequences for possible transcription factor binding are also shown.

 
The minimal Dmp1 promoter is TATA-less, GC rich, and contains multiple transcription factor binding sequences (Fig. 1B). In addition, AP-1-like and AML1-like consensus binding sequences are present in the 5' leader sequence. We tested whether the Dmp1 promoter can be activated by overexpression of c-Myc, E2F-1, or by oncogenic Ha-RasV12, all of which can upregulate p19Arf mRNA and protein levels (3, 33, 55). Neither c-Myc nor E2F-1 activated the murine Dmp1 promoter, whereas oncogenic RasV12 weakly stimulated the promoter in Ink4a-Arf-null NIH 3T3 cells (Fig. 2A). The same data were obtained in 293T cells, where both Rb and p53 were functionally inactivated (data not shown). Oncogenic Ras overexpression induces transformation in immortalized fibroblasts but causes p19Arf (and/or p16Ink4a) and p53-dependent cellular senescence in primary fibroblasts (33, 44). The Dmp1 promoter was more efficiently activated by H-RasV12 in primary wild-type MEFs (Fig. 2B), as well as in IMR-90 cells, primary diploid fibroblasts derived from human fetal lung (data not shown). Overexpression of c-Myc and E2F-1 had no inductive effects on the murine Dmp1 promoter even in primary MEFs (data not shown). As for NIH 3T3 cells, RasV12 did not efficiently activate the Dmp1 promoter in immortalized Arf-null MEFs; however, a response to RasV12 was restored by cotransfecting a p19Arf expression vector (Fig. 2B). Thus, the ability of RasV12 to activate Dmp1 is maximized in nonimmortalized cells that retain functional Arf and p53 activities.



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 2. Responsiveness of the Dmp1 promoter to various oncogenic stimuli. (A) The Dmp1 promoter is weakly activated by oncogenic RasV12 but not by c-Myc or E2F-1 in 3T3 cells. NIH 3T3 cells were transfected with the luciferase reporter –374 NsiI with expression vectors for c-Myc, E2F-1, and Ha-RasV12 driven by the CMV promoter. The numbers show the fold activation of the luciferase reporter corrected by the internal control secreted endocrine alkaline phosphatase levels. (B) p19Arf dependence of the response of the Dmp1 promoter to RasV12. Reporter assays were performed in passage-5 wild-type MEFs (left panel), Arf-null MEFs (middle panel), and Arf-null MEFs with p19Arf expression vector (right panel). The Dmp1 promoter was efficiently activated by RasV12 in wild-type cells but not in Arf-null cells. The loss of responsiveness of the Dmp1 promoter to RasV12 in Arf-null cells was restored by cotransfecting the RasV12 and p19Arf expression vectors. (C) The Dmp1 promoter is activated by RasV12S35 but not by RasV12G37 or by RasV12C40. A Dmp1 reporter assay was performed in wild-type MEFs with RasV12 double mutants (RasV12S35, RasV12G37, and RasV12C40). The Dmp1 promoter was activated most efficiently by RasV12S35, suggesting that the MAPK pathway plays the most significant role in Dmp1 promoter activation in response to oncogenic Ras signaling. (D) Activation of the Dmp1 promoter by MAPK pathways is inhibited by the MEK/ERK inhibitor U0126. In order to study which MAPK pathway is key to RasV12-mediated Dmp1 promoter activation, reporter assays were performed with pCMV-RasV12 in the presence of 10 µM U0126, 10 µM SP600125, 10 µM SB203580, or a combination of these compounds. Activation of the Dmp1 promoter by RasV12 was efficiently blocked by U0126.

 
Activation of the Dmp1 promoter by RasV12 is dependent on the Ras-MEK-ERK pathway. The transforming activity of activated Ras depends on at least three downstream effectors, including Raf-1/MAPK, Ral-GDS, and PI(3)K, which mediate different aspects of oncogenic transformation. Oncogenic H-RasV12 variants with mutations T35S, E37G, or Y40C bind to and primarily activate Raf/MAPK, PI(3)K, or Ral-GDS, respectively (20). The Dmp1 promoter was most efficiently activated by RasV12S35, followed by RasV12G37, while RasV12C40 had a minimal effect (Fig. 2C). Mammalian MAPKs branch into three major pathways that involve MEK/ERK, JNK/SAPK, and p38 signaling. These three pathways are mainly activated by mitogens, by inflammatory cytokines, UV, and {gamma}-irradiation, and by osmotic stress signaling, respectively. In order to study which of these pathways is important in RasV12-mediated Dmp1 promoter activation, reporter assays were performed with an Ha-RasV12 expression vector in the presence of MAPK inhibitors (Fig. 2D). Activation of the Dmp1 promoter by RasV12 was efficiently blocked by U0126 alone (MEK/ERK pathway), whereas SP600125 (targeting JNK/SAPK) and SB203580 (targeting p38) had partial inhibitory effects. When combined, U0126 plus SP600125 synergistically blocked Dmp1 promoter activation by Ras, whereas the combination of SP600125 and SB203580 exhibited no additive effects (Fig. 2D). The effects of the SP compound were confirmed by an independent reporter assay with another JNK inhibitor, (L)-JNKI1 (Calbiochem). Together, these data suggest that the Raf-MEK-ERK pathway plays the major role in Dmp1 promoter activation by RasV12.

Activated RasV12 and c-Raf induce Dmp1 mRNA. Northern blotting analysis demonstrated accumulation of Dmp1 transcripts in cells infected with retroviruses encoding Ha-RasV12 (Fig. 3A, left panel). To determine if activation of the Raf-MEK-ERK pathway can lead to alterations in Dmp1 mRNA and protein expression, wild-type MEFs were infected with a retrovirus expressing {Delta}Raf:ER[DD], a mutated Raf kinase whose activity is regulated by tamoxifen (28, 54). Infected cells were either left untreated or were treated with 2 µM 4-HT for 0 to 36 h. Cells were harvested at various times thereafter and were analyzed by Northern and Western blotting with specific probes and antibodies to Dmp1. As an additional control, wild-type MEFs were infected with an empty ER vector and treated with 2 µM 4-HT. Treatment of control MEFs with 4-HT did not significantly change the levels of Dmp1 mRNA or protein (Fig. 3A and B). Activation of {Delta}Raf:ER by 4-HT increased the Dmp1 mRNA threefold by 8 h and eightfold by 16 to 24 h (Fig. 3A). Proportionate increases in Dmp1 protein expression were observed (Fig. 3B). Dmp1 is a phosphoprotein with different isoforms of various molecular masses (120 to 130 kDa) (14). We noticed that some additional immunoreactive species of both higher (~135 kDa) and lower (~110 kDa) molecular masses were observed at 16 to 36 h after {Delta}Raf:ER activation, suggesting that kinases regulated by the Raf-MEK-ERK pathway could further modify the Dmp1 protein (Fig. 3B). The rate of Dmp1 protein accumulation was several hours slower than that of phospho-ERK but was very similar to that of cyclin D1 induction, which is also positively regulated by the MEK-ERK pathway (Fig. 3B).



View larger version (54K):
[in this window]
[in a new window]
  		FIG. 3. Dmp1 plays a key role in c-Raf-induced senescence. (A) Ha-RasV12 and activated {Delta}Raf:ER induce Dmp1 mRNA. Passage-5 wild-type MEFs were infected with retroviruses expressing Ha-RasV12 (left panel) or {Delta}Raf:ER (right panel) and selected with puromycin. Northern blotting was performed with a Dmp1-specific probe with ß-actin as an internal control. Expression of Ha-RasV12 increased Dmp1 mRNA by threefold at day 3 to 4 (left panel). Activation of {Delta}Raf:ER by 2 µM 4-HT increased Dmp1 mRNA by threefold at 8 h and by eightfold at 16 to 24 h (right panel). (B) Activated {Delta}Raf:ER induces the Dmp1 protein. The increase of Dmp1 mRNA resulted in Dmp1 protein induction by threefold at 8 h and eightfold by 24 h. The kinetics was slower than for phosphor-ERK accumulation but very similar to that of cyclin D1 induction. (C) Dmp1-null cells are resistant to Raf-induced cell cycle arrest. In order to study the biological effects of c-Raf activation in Dmp1–/– cells, both wild-type and Dmp1-null MEFs were infected with retroviruses encoding {Delta}Raf:ER or the control ER vector. After selection with puromycin, 105 cells were plated in 60-mm-diameter dishes and 1 µM 4-HT was added to activate {Delta}Raf:ER. (Open symbols, with 4-HT; closed symbols, without 4-HT). Note that Dmp1-null cells expressing {Delta}Raf:ER (triangles) grew at the same rate as those expressing ER vector alone (rectangles), while wild-type cells expressing activated {Delta}Raf:ER underwent irreversible cell cycle arrest. (D) p19Arf does not increase in response to {Delta}Raf:ER in Dmp1-null cells. None of p19Arf, p53, and p21Cip1 increased in response to {Delta}Raf:ER in Dmp1-null cells, while p16Ink4a levels remained constant in both wild-type and Dmp1-null cells.

 
Dmp1 is a key target in Raf-induced cellular senescence. We next infected wild-type and Dmp1-null MEFs with retroviruses encoding {Delta}Raf:ER or the control ER vector. After selection with puromycin, 105 cells were plated in 60-mm-diameter dishes and 1 µM 4-HT was added to activate {Delta}Raf:ER. The total cell number did not increase in 4HT-treated wild-type cells expressing {Delta}Raf:ER, since oncogenic c-Raf activation induces replicative senescence in primary MEFs (Fig. 3C, left panel). There were no significant differences in the proliferation of cells infected with the control ER vector with or without 4-HT and untreated cells infected with the {Delta}Raf:ER virus (Fig. 3C, left panel). In contrast, Dmp1-null cells expressing activated {Delta}Raf:ER grew exponentially after treatment with 4-HT and reached the same saturation density as control populations (Fig. 3C, right panel). When cell lysates prepared from wild-type and Dmp1-null cells treated with 4-HT were fractionated on denaturing gels and blotted with various antibodies, we observed accumulation of p19Arf (fourfold) and p53 (twofold) from 16 to 36 h in wild-type cells expressing activated {Delta}Raf:ER. However, the basal levels of p19Arf were very low in Dmp1-null cells (18), and neither p19Arf nor p53 was induced by activation of {Delta}Raf:ER (Fig. 3D). We also observed very rapid accumulation of p21Cip1 in wild-type MEFs (twofold increase at 3 h and eightfold increase by 8 to 36 h), whereas the levels of p16Ink4a did not appreciably change upon Raf activation (Fig. 3D). Thus, Dmp1 contributes to p19Arf and p21Cip1 upregulation and proliferative arrest in response to oncogenic Raf signaling.

Mapping of RasV12-responsive elements in the Dmp1 promoter. Deletion of the Dmp1 promoter 5' to the –374 NsiI site did not significantly change its responsiveness to RasV12 (Fig. 4A). However, when we deleted the 286-bp fragment from –374 NsiI to –88 ApaI, we observed a twofold increase in Dmp1 promoter activity in response to oncogenic Ras (fourfold versus eightfold). Ets1/Ets2 transcription factors have been reported to play crucial roles in the regulation of the human p16INK4A promoter in response to Ras-Raf-MEK-ERK signaling (32). EGR-1 has also been reported to play important roles in Ras signaling (2). However, mutation or deletion of these Ets/Elk, EGR-1 sites did not alter the responsiveness of the Dmp1 promoter to oncogenic RasV12, although mutation or deletion of the former significantly lowered the basal levels (Fig. 4A; data not shown for EGR-1). In contrast, deletion of the 14-bp 5'-untranslated region (from +4 cytosine to +17 guanine) resulted in elimination of the Dmp1 promoter response to RasV12 (Fig. 4A). The presence of the Ras-responsive elements in the 5'-untranslated region was confirmed by eliminating the 14-bp fragment from the –374 NsiI promoter reporter construct (Fig. 4A).



View larger version (76K):
[in this window]
[in a new window]
  		FIG. 4. Mapping of the RasV12-responsive element on the murine Dmp1 promoter. (A) Mapping of the RasV12-responsive element on the Dmp1 promoter based on a luciferase assay. Reporter assays were performed in IMR-90 cells with the deletion mutants described in Fig. 1A and their derivatives with (white columns) or without (black columns) Ha-RasV12 expression vector. (B) Kinetics of Fos and Jun family protein accumulation in cells with activated c-Raf. Wild-type MEFs were infected with vector ER or {Delta}Raf:ER retroviruses and were stimulated with 4-HT. The lysates were analyzed by Western blotting with specific antibodies. (C) Identification of the transcription factors that bind to the AP-1-like sequence on the Dmp1 promoter by EMSA. Lysates were prepared from wild-type MEFs expressing {Delta}Raf:ER or ER alone and stimulated with 2 µM 4-HT for 16 h. Complex A formation was antagonized by a 100-fold excess of cold oligonucleotides derived from the AP-1-like sequence on the Dmp1 promoter, as well as classical AP-1 or CREB consensus sequences reported earlier. The complex was supershifted only with antibodies to c-Fos, Fra-1, c-Jun, phospho-c-Jun, JunB, and JunD (arrows, S). (D) ChIP assay of the Dmp1 promoter. Significant amounts of c-Fos, Fra-1, c-Jun, JunB, and JunD were detected on the Dmp1 promoter in response to activated Raf signaling. The specificity of the Fos/Jun signals was confirmed by control PCR amplification of the Dmp1 promoter sequence located 2 kb upstream from the transcription initiation site (24 h cont).

 
The Ras-responsive element in the Dmp1 promoter contains an AP-1-like sequence (CTCGCTCA) as well as an AML1-like sequence (ACCCCA) (Fig. 1B). Both AP-1 and AML1 can be activated in oncogenic Ras signaling (6, 51). The Ras response of the Dmp1 promoter was significantly compromised by mutating the AP-1-like sequence (–88 ApaI AP-1 mut and –1787 PstI AP-1 mut), but not by mutating the AML1 consensus sequence (–88 ApaI AML1 mut), suggesting that proteins of the AP-1 family could play key roles in Dmp1 regulation (Fig. 4A).

We therefore studied the kinetics of accumulation of Fos and Jun family members in response to {Delta}Raf:ER activation. We observed a rapid increase of c-Fos (3 h) and Fra-1 and also increases in the JunB and c-Jun proteins and, later, JunD (Fig. 4B). FosB was not detected in MEFs with or without Raf activation (data not shown), and we did not observe any significant change of Fra-2 levels upon Raf activation (Fig. 4B). We also searched for transcription factors that could bind to the 14-bp Ras-responsive element in an EMSA performed with nuclear extracts isolated from MEFs with activated {Delta}Raf:ER (16 h). A major complex (labeled A in Fig. 4C) was formed using nuclear extracts from induced cells (left panel, lane 2), and both classical AP-1 (TGACTCA) and CREB (TGACGTCA) consensus oligonucleotides blocked its formation (lanes 5 and 6). Antibodies to Fos family and Jun family proteins supershifted the AP-1-like complex (Fig. 4C, right panel, lanes 17 and 18). Fos-family proteins, except for FosB, were detected on the 31-bp oligonucleotides covering the Ras-responsive element (Fig. 4C, right panel, lanes 4 to 7). Among the Jun family proteins, c-Jun and especially phosphorylated c-Jun, as well as JunB, appeared to contribute to the major complexes, although a small amount of JunD was also detected (Fig. 4C, right panel, lanes 8 to 11). Although formation of complex A was completely inhibited by CREB-consensus oligonucleotides, none of the ATF (ATF-1, -2, and -3) or CREB family proteins (CREB1 and CREB2) was detected in the AP-1-like complex on the Dmp1 promoter (Fig. 4C, right panel, lanes 12 to 16). Complex A was not supershifted with antibodies to CREM1, AML1, or c-Maf (data not shown).

Results from the EMSA were confirmed by ChIP with specific antibodies (Fig. 4D). The levels of Fos and Jun family proteins were very low on the Dmp1 promoter before addition of 4-HT, except for Fra-2. Significantly increased levels of c-Fos (8 to 24 h), Fra-1 (24 h), c-Jun (8 to 24 h), JunB (16 to 24 h), and JunD (8 to 24 h) proteins were detected on the Dmp1 promoter after addition of 4-HT in cells expressing {Delta}Raf:ER (Fig. 4D). The specificity of binding of Fos and Jun proteins to the region of interest was confirmed by PCR amplification of the sequences located 2 kb upstream of the transcription initiation site with samples harvested at 24 h (Fig. 4D, 24 h control). Thus, Fos and Jun family proteins, especially c-Fos, c-Jun, JunB, and JunD, bind to the endogenous Dmp1 promoter 5'-untranslated region in response to oncogenic Ras-Raf signaling.

Oncogenic Ras does not activate the Dmp1 promoter in c-Jun and JunB knock-down cells. Detection of Fos and Jun proteins on the Dmp1 promoter does not necessarily mean that they are transcriptional activators of the Dmp1 promoter. In order to test which of these proteins is physiologically relevant, we performed reporter assays with Fos and Jun family proteins alone or in combination. When tested alone, only c-Jun activated the promoter to significant levels. (Fig. 5A, left panel). All of the Jun proteins synergistically activated the Dmp1 promoter by collaborating with c-Fos or Fra-1, whereas no synergism was observed among Jun family proteins (Fig. 5A, right panel).



View larger version (47K):
[in this window]
[in a new window]
  		FIG. 5. Synergism of Dmp1 promoter activation by AP-1 proteins and RNAi assay. (A) Synergistic activation of the Dmp1 promoter by AP-1 family proteins. Synergism among Fos and Jun family proteins was tested in IMR-90 cells. For reporter assays, 0.5 to 1 µg of Fos/Jun expression vectors per dish were used for transfections in the left panel, while 0.25 µg of expression vectors were used per dish in the right panel to study synergism. Significant synergism was found between c-Jun and c-Fos or Fra-1, JunB and c-Fos or Fra-1. (B) Downregulation of Fos and Jun family proteins by siRNA. Wild-type MEFs were infected with retroviruses that produce siRNA, and puromycin-resistant cells were expanded. The lysates were studied for target protein expression with specific antibodies. (C) RasV12 does not activate the Dmp1 promoter in cells with downregulated c-Jun or JunB. Reporter assays were performed with the –88 ApaI Dmp1 promoter in MEFs with downregulated AP-1 proteins. Dmp1 promoter activation by RasV12 was completely attenuated in cells where c-Jun or JunB proteins were knocked down, whereas c-Fos, Fra-1, Fra-2, or JunD proteins were dispensable for Dmp1 promoter activation by Ras. (D) Restoration of the Ras responsiveness of the Dmp1 promoter in Jun knock-down cells. Reporter assays were performed in c-Jun or JunB knock-down (K.D.) MEFs with c-Jun or JunB and RasV12 expression vectors. The responsiveness of the Dmp1 promoter to RasV12 was completely restored by transfection of c-Jun or JunB knock-down cells with a c-Jun or JunB expression vector, respectively.

 
We next used short interfering RNAs (siRNAs) (7) to knock down the level of c-Fos, Fra-1, Fra-2, c-Jun, JunB, or JunD in MEFs. Figure 5B shows that the endogenous protein levels were significantly downregulated by 80 to 90% in cells that were infected with retroviruses prepared by use of the pSUPER.retro.puro RNAi system (Oligoengine). Oncogenic RasV12 completely failed to activate the Dmp1 promoter in MEFs in which c-Jun or JunB levels were reduced, whereas it activated the Dmp1 promoter three- to fourfold in cells with downregulated c-Fos, Fra-2, or JunD (Fig. 5C). The responsiveness of the Dmp1 promoter to Ras in c-Jun or JunB knock-down cells was completely restored by cotransfecting a c-Jun or JunB expression vector with RasV12 (Fig. 5D). We confirmed by Western blotting that overexpression of c-Jun/JunB under the control of the CMV promoter was strong enough to overcome the effects of siRNAs (data not shown). These data underscore the critical roles of c-Jun and JunB transcription factors in the activation of Dmp1 in response to oncogenic Ras signaling.

Mapping of Ras-responsive elements on the murine Arf promoter. Although oncogenic Ras signaling increases both p19Arf mRNA and protein (33), the mechanism of Arf induction by RasV12 has not yet been clarified. In wild-type MEFs, RasV12 induced a sevenfold activation of the Arf promoter, which was compromised in an Arf promoter with disrupted Dmp1/Ets binding sequence (Fig. 6A, left panel) (17). In order to demonstrate the relative importance of Dmp1 versus Ets family proteins in the Arf promoter activation by Ras, we repeated these assays using both wild-type and Dmp1-null MEFs. This revealed a dramatic reduction in the response of the Arf promoter to RasV12 in cells lacking Dmp1 (Fig. 6A, middle panel). In order to rule out the possibility that events secondary to Dmp1 loss were responsible for the failure of the Arf promoter to respond to oncogenic Ras, we performed a reporter assay on the Arf promoter by transfecting Dmp1–/– MEFs with Dmp1 and RasV12 expression vectors. Our data indicated that ectopic expression of Dmp1 restores the responsiveness of the Arf promoter to oncogenic Ras (Fig. 6A, right panel). This result is consistent with those of previous studies in which Dmp1-null cells exhibited only minor accumulation of p19Arf and p53 in response to oncogenic Ras, enabling these cells to be transformed by oncogenic Ras alone (18).



View larger version (52K):
[in this window]
[in a new window]
  		FIG. 6. Dmp1 plays a key role in Arf promoter activation by RasV12. (A) Mapping of the RasV12-responsive element on the Arf promoter. (Left panel) Reporter assay performed with either wild-type or mutant murine Arf promoter to study the importance of the Dmp1/Ets site in response to oncogenic Ras activation. Arf promoter activation was almost completely abolished by mutating the Dmp1/Ets binding site (17). (Middle panel) Reporter assay performed with wild-type Arf promoter in both wild-type and Dmp1-null MEFs. Oncogenic Ras does not activate the Arf promoter in Dmp1-null cells. (Right panel) Restoration of the Ras responsiveness of the Arf promoter in Dmp1-null cells. The responsiveness of the Arf promoter to RasV12 was completely recovered by transfection of Dmp1-null cells with the Dmp1-expression vector. (B) ChIP assay on the Arf promoter. Endogenous Dmp1 protein was found on the Arf promoter in response to oncogenic Raf signaling (lanes 1 and 2). Lanes 3 and 4, no antibodies were used for the immunoprecipitation; lanes 5 and 6, signals from total chromatin samples. (C) Cyclin D1 does not inhibit the Dmp1 activity on the Arf promoter. A reporter assay was performed in NIH 3T3 cells with expression vectors for Dmp1 and cyclins. Dmp1 and D-type cyclins additively activated the Arf promoter, while other cyclins (cyclin A and cyclin H) had little effect on Arf promoter activation by Dmp1. The K114E cyclin D1 mutant (D1 KE) that does not interact with Cdk4 did not influence Dmp1's activity on the Arf promoter, suggesting that the collaborative effect was dependent on Cdk4 activation. (D) Detection of cyclins in transfected cells. The expression of cyclins was confirmed by Western blotting of lysates from the luciferase assay with specific antibodies. Lane 1, pFLEX1 vector only; lane 2, Flag-tagged cyclin D1; lane 3, cyclin D1 KE mutant; lane 4, cyclin D2; lane 5, cyclin D3; lane 6, cyclin A; lane 7, cyclin H.

 
We then used a ChIP assay to determine whether endogenous Dmp1 binds to the Arf promoter in vivo. Dmp1 bound to the endogenous Arf promoter after activation of {Delta}Raf:ER; we saw an increased signal even though the input chromatin was 50% lower (Fig. 6B). We conclude that the RasV12-responsive element falls within the Dmp1/Ets consensus binding sequence on the Arf promoter and that Dmp1 activity is required to efficiently enable Arf induction by oncogenic Ras.

Our present study demonstrates that {Delta}Raf:ER induces Dmp1 and cyclin D1 proteins almost with the same kinetics (Fig. 3B). Because gross overexpression of D-type cyclins can inhibit Dmp1-mediated transactivation in a Cdk-independent fashion (15, 16), we investigated whether D-type cyclins could influence Arf promoter activation by Dmp1 (Fig. 6C). The Arf promoter was weakly activated by overexpression of cyclin D1, D2, or D3 (~2-fold) (data not shown), consistent with the concept that the Arf promoter responds to hyperproliferative oncogenic signaling (47). Moreover, the D-type cyclins collaborated with Dmp1 in activating the Arf promoter (Fig. 6C). Importantly, this additive effect was dependent on the activation of the Rb-E2F pathway, since a cyclin D1 mutant that does not interact with Cdk4 (D1K114E) (15) did not activate the Arf promoter. Thus, while overexpressed D-type cyclins can inhibit Dmp1 activity in a Cdk4-independent manner, their ability to activate Cdk4 conveys an opposing signal. Interestingly, the stimulatory effects on the Arf promoter were limited to D-type cyclins, since none of the other cyclins (cyclin A or cyclin H) affected Arf promoter activation by Dmp1 (Fig. 6C).

Our model of the signaling cascades that link the oncogenic Ras-Raf-MEK-ERK pathway and the Arf-Mdm2-p53 tumor surveillance pathway is depicted in Fig. 7.



View larger version (15K):
[in this window]
[in a new window]
  		FIG. 7. The novel Jun-Dmp1 pathway. Dmp1 is a key molecule linking Ras-Raf-MEK-ERK oncogenic signaling and the Arf-Mdm2-p53 tumor suppressor pathway. The pathway mediated by Dmp1 directly links oncogenic Ras-Raf signaling and p19Arf through activation of c-Jun/JunB proteins. Oncogenic Ras activates both the Jun-Dmp1 pathway and the classical cyclin D1/Cdk4-Rb-E2F pathway to activate Arf gene expression to achieve premature senescence. Since the Ras-Raf pathway activates the Mdm2 promoter as well, the levels of p53 are delicately regulated by the activity of the Dmp1-Arf and AP-1/Ets-Mdm2 pathways.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Overexpression of oncogenic RasV12 acts through the Raf-MEK-ERK pathway and AP-1 signaling to activate Dmp1. In turn, Dmp1 activates the Arf promoter, ultimately leading to a p53 response that limits oncogene-induced cell proliferation and tumorigenesis. Although our investigators previously recognized that the Dmp1 protein was induced by explantation of early-passage MEFs and that its accumulation preceded that of p19Arf as primary cells were passaged in culture (18), we now appreciate that the Dmp1 promoter can respond to a specific subset of Ras-dependent signals. Among the three different oncogenes that we studied (c-Myc, E2F-1, and Ha-RasV12), each of which can induce Arf (3, 33, 55), only RasV12 activated the Dmp1 promoter. Our findings indicate that RasV12 signaling to Arf is mediated primarily by Dmp1, whereas E2F-1 and Myc address Arf via other mechanisms.

Ras transformation correlates with increased c-Jun transcriptional activity and an increase in AP-1-mediated gene expression (reviewed in references 30 and 46). In experiments employing the {Delta}Raf:ER system, c-Jun protein accumulation peaked at 16 to 24 h after 4-HT treatment, the timing of which was coincident with expression of Dmp1 and p19Arf. Significantly increased levels of the c-Jun protein were detected on the Dmp1 promoter after Raf activation, suggesting that endogenous c-Jun can activate the Dmp1-p19Arf axis in response to oncogenic Ras-Raf signaling. Overexpression of c-Jun alone activated the Dmp1 promoter, and Dmp1 promoter activation by RasV12 was completely abolished in c-Jun knockdown cells. Thus, we conclude that c-Jun is the most prominent AP-1 protein that activates the Dmp1 promoter in response to oncogenic Ras-Raf signaling. Control of cell cycle progression by c-Jun was shown to be p53 dependent (43). Indeed, c-Jun is upregulated in many carcinomas (11), and it plays a key role in chemically induced hepatocellular carcinoma in mice (11). Thus, the concept that c-Jun is an oncogene (27) fits with the observation that c-Jun activates the Dmp1 promoter in response to an oncogenic stimulus.

The finding that activation of the Dmp1 promoter by Ras was significantly attenuated in JunB knock-down cells was quite unexpected, since c-Jun and JunB are known to have antagonistic functions (30). JunB can suppress cell proliferation by transcriptional activation of p16Ink4a in MEFs (35). Since JunB alone does not activate the Dmp1 promoter, JunB heterodimers, possibly in association with Fos family proteins, should play important roles in the Dmp1 promoter activation response to oncogenic Ras. Transcriptionally less active JunB can substitute for c-Jun in mouse development and cell proliferation, and JunB can restore the expression of genes regulated by Jun/Fos, but not those regulated by Jun/ATF (36).

Although oncogenic Ras has been reported to induce premature senescence by upregulating p19Arf and p53 levels, the mechanism of p19Arf activation has remained poorly understood. Activated Ras induces cyclin D1 and accelerates its assembly with Cdk4 (1, 9), and it has generally been assumed that E2F family proteins play important roles in Arf regulation in response to Ras signaling (Fig. 7). However, results from two different groups suggest that oncogenic Ras can increase Arf mRNA levels in an E2F-independent manner (34, 41). The RasV12-responsive element was mapped to the Dmp1/Ets binding sequence in the murine Arf promoter. We also observed increased binding of the endogenous Dmp1 protein to the Dmp1/Ets site within the Arf promoter in response to activation of {Delta}Raf:ER. In contrast, activated {Delta}Raf:ER failed to stop the growth of Dmp1-null cells. These data strongly indicate that Dmp1 plays a major role in conveying oncogenic Ras-Raf signaling to the Arf-p53 pathway in rodent fibroblasts. We propose the presence of the Jun-Dmp1 pathway that directly links Ras-Raf signaling and p19Arf. This novel pathway collaborates with the classical cyclin D1/Cdk4-Rb-E2F pathway to activate the Arf gene expression in response to oncogenic Ras signaling (Fig. 7).

The induction of p21Cip1 by activated Raf was also impaired in Dmp1-null cells. A simple interpretation is that loss of Dmp1 attenuates Arf-p53 function, preventing the accumulation of canonical p53-responsive genes, p21Cip1 among them. However, this is unlikely to be the only explanation, because the rate of p21Cip1 accumulation in response to Raf expression was faster than that of p19Arf in wild-type MEFs, and because high-intensity B-Raf and c-Raf signaling can cause p21Cip1-dependent cell cycle arrest in NIH 3T3 cells that lack the Ink4a-Arf locus (45, 54). Since both the mouse and human p21Cip1 promoter lacks typical Dmp1-consensus sequences, it is unlikely that Dmp1 binds and activates the p21Cip1 promoter directly. Further studies will be required to clarify the roles of Dmp1 in p21Cip1 regulation in response to Ras-Raf signaling.

Mdm2 acts as a major regulator of the tumor suppressor p53 by targeting its destruction, inhibiting its transcriptional activation, and by accelerating nuclear-to-cytoplasmic shuttling of p53. p19Arf can antagonize each of these processes (49). The Mdm2 intronic promoter is regulated by oncogenic Ras signaling as well as by p53 signaling (39). Activation of the Mdm2 promoter by Ras is also dependent on AP-1 and Ets-like elements and is p53 independent (39). Interestingly, the structures of the murine Dmp1 promoter and 5'-untranslated region are very similar to that of the Mdm2 promoter, although the Ets site was dispensable for Dmp1 promoter activation by RasV12. Since oncogenic Ras activates both the Mdm2 and Dmp1-p19Arf pathways (Fig. 7), the levels of p53 in response to oncogenic Ras should be delicately controlled by opposing effects between Mdm2 and Dmp1-p19Arf-regulated signaling pathways.

Overexpressed D-type cyclins antagonize Dmp1 transcriptional activity in a Cdk-independent fashion when tested using artificial promoter-reporter plasmids containing concatamerized Dmp1 consensus binding sequences or with some natural promoters, such as that derived from the CD13/aminopeptidase N gene (15, 16). However, the results were reversed on the Arf promoter, where D-type cyclins cooperated to enhance the activity of Dmp1 in a Cdk4-dependent manner. The Arf promoter contains both Dmp1- and E2F-binding sites, enabling RasV12-induced cyclin D1 to assemble with Cdk4, promote the release of E2Fs from Rb, and thereby collaborate with Dmp1 in activating Arf gene expression (17). On the other hand, the CD13/aminopeptidase N promoter, which lacks E2F consensus sequences, can be experimentally suppressed by D-type cyclins which, when overexpressed, can interfere with Dmp1 binding to DNA. The Dmp1/Ets consensus sequences found within these two promoters are completely identical (CCCGGATGC) (16, 17), consistent with the idea that sequences flanking the Dmp1 binding site determine the responsiveness of the promoter to D-type cyclins. It is important to emphasize that interference with Dmp1 activity by D-type cyclins has never been demonstrated in a setting in which D-type cyclins accumulate to physiological levels. Indeed, the level of cyclin D1 achieved after Ha-RasV12 expression was 10-fold lower than that generated by the cyclin D1 expression vector itself (R. Sreerameneni and K. Inoue, unpublished data).

A further complication stems from observations that efficient activation of the Dmp1 promoter by RasV12 was limited to nonimmortalized cell strains, such as primary MEFs and human IMR-90 cells, whereas Dmp1 induction was compromised in Arf-null cells and other established cell lines. Oncogenic Ras has differential effects on primary cell strains versus established cell lines, inducing replicative senescence in the former but stimulating proliferation and transformation in the latter (44). When we reexpressed p19Arf in Arf-null cells, the response of the Dmp1 promoter to RasV12 was at least partially restored. We do not consider the effect of Ras in regulation of the Dmp1 expression is a consequence of growth inhibition because (i) Dmp1 is induced by {Delta}Raf:ER at the same kinetics as that of cyclin D1 before the cell growth is inhibited, (ii) the magnitude of Dmp1 promoter activation by Ras is higher in early-passage MEFs than in late-passage MEFs that are becoming senescent (Sreeramaneni and Inoue, unpublished), and because (iii) Dmp1–/– cells are morphologically transformed by RasV12 (18), suggesting that the Dmp1 promoter activation is an integral part of Ras-Arf-p53 signaling. Although we cannot yet explain why Dmp1 activation by Ras depends upon Arf in NIH 3T3 cells and MEFs and upon p53/Rb in 293T cells, these results suggest that the presence of a functional Arf-p53 pathway and an intact p53 G1 checkpoint may be required. Arf may modify the activities of nuclear proteins that regulate Dmp1 transcription, possibly in a cell cycle-specific manner. The Arf status of a cell can also determine the transcriptional activity of NF-{kappa}B in response to an oncogene, although these effects are independent of p53 and Mdm2 (40). Indeed, many genes are induced or suppressed by p19Arf (24), some of which might well modify the Dmp1 response. The durability of the Dmp1-Arf response to Ras might well depend upon feedback control, in which the integrity of Arf signaling, whether p53 dependent or not, appears to reinforce the ability of Ras to trigger Dmp1 gene expression.

Recently, studies have shown that lower physiological levels of mutant Ras from a single-copy oncogenic Ras allele activate the p19Arf-p53 pathway in MEFs, but not to the same extent as overexpressing oncogenic Ras driven by the retroviral promoter (12, 52). Lower levels of mutant Ras promoted cell proliferation in MEFs in culture and also in epithelial cells in vivo despite a lack of obvious cooperating events, due to the incomplete activation of the p19Arf-p53 pathway (12, 52). Since human tumors that contain oncogenic Ras tend to have single mutant alleles, such studies are a good representation of early-stage human cancer. However, duplication or amplification of the mutant Ras gene should have more of a growth advantage than single Ras mutant cells, and it has been reported that both human and mouse cancers often amplify or overexpress the mutant Ras gene with progression of the disease (4, 5, 31, 53). Since the level of induction of Dmp1 and Arf by oncogenic Ras-Raf pathway is dose dependent (Fig. 2 and 6) (Sreeramaneni and Inoue, unpublished), our study presents the mechanism of prevention of tumor progression induced by overexpressed oncogenic Ras.


   ACKNOWLEDGMENTS
 
We thank Martine Roussel and Yoshiaki Tsuji for helpful discussions. We also thank Mark Willingham and Ali Mallakin for critical reading of the manuscript. We are grateful to Michael Ostrowski, Scott Lowe, Christopher Counter, John Cleveland, Joseph Nevins, Tom Curran, and Andrew Thorburn for generous gifts of plasmid DNAs. We also thank Madeline Coombes, Paloma Giangrande, and Kenji Tago for ChIP protocols and Oktay Kaplan for technical assistance.

This work was supported by American Cancer Society grant 93-035-09 and National Institutes of Health grant CA106314-01 (to K.I.). C.J.S. is an Investigator of the Howard Hughes Medical Institute.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Pathology, Wake Forest University Health Sciences, 2102 Gray Building, Medical Center Blvd., Winston-Salem, NC 27157. Phone: (336) 716-5863. Fax: (336) 716-6757. E-mail: kinoue{at}wfubmc.edu. Back

{dagger} R.S. and A.C. contributed equally to this work. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Albanese, C., J. Johnson, G. Watanabe, N. Eklund, D. Vu, A. Arnold., and R. G. Pestell. 1995. Transforming p21ras mutants and c-Ets-2 activate the cyclin D1 promoter through distinguishable regions. J. Biol. Chem. 270:23589-23597.[Abstract/Free Full Text]

2. Alexandropoulos, K., S. A. Qureshi, J. T. Bruder, U. Rapp, and D. A. Foster. 1992. The induction of Egr-1 expression by v-Fps is via a protein kinase C-independent intracellular signal that is sequentially dependent upon HaRas and Raf-1. Cell Growth Differ. 3:731-737.[Abstract]

3. Bates, S., A. C. Phillips, P. Clarke, F. Stott, G. Peters, R. L. Ludwig, and K. H. Vousden. 1998. E2F-1 regulation of p14ARF links pRB and p53. Nature 395:124-125.[CrossRef][Medline]

4. Bos, L. J. 1988. The ras gene family and human carcinogenesis. Mutat. Res. 195:255-271.[CrossRef][Medline]

5. Bremner, R., and A. Balmain. 1990. Genetic changes in skin tumor progression: correlation between presence of a mutant ras gene and loss of heterozygosity on mouse chromosome 7. Cell 61:407-417.[CrossRef][Medline]

6. Bruder, J. T., G. Heidecker, and U. R. Rapp. 1992. Serum-, TPA-, and Ras-induced expression from Ap-1/Ets-driven promoters requires Raf-1 kinase. Genes Dev. 6:545-556.[Abstract/Free Full Text]

7. Brummelkamp, T. R., R. Bernards, and R. Agami. 2002. A system for stable expression of short interfering RNAs in mammalian cells. Science 296:550-553.[Abstract/Free Full Text]

8. Chadee, D. N., M. J. Hendzel, C. P. Tylipski, C. D. Allis, D. P. Bazett-Jones, J. A. Wright, and J. R. Davie. 1999. Increased Ser-10 phosphorylation of histone H3 in mitogen-stimulated and oncogene-transformed mouse fibroblasts. J. Biol. Chem. 274:24914-24920.[Abstract/Free Full Text]

9. Cheng, M., V. Sexl, C. J. Sherr, and M. F. Roussel. 1998. Assembly of cyclin D-dependent kinase and titration of p27kip1 regulated by mitogen-activated protein kinase kinase (MEK1). Proc. Natl. Acad. Sci. USA 95:1091-1096.[Abstract/Free Full Text]

10. Cowley, S., H. Paterson, P. Kemp, and C. J. Marshall. 1994. Activation of MAP kinase kinase is necessary and sufficient for PC12 differentiation and for transformation of NIH 3T3 cells. Cell 77:841-852.[CrossRef][Medline]

11. Eferl, R., R. Ricci, L. Kenner, R. Zenz, J. P. David, M. Rath, and E. F. Wagner. 2003. Liver tumor development. c-Jun antagonizes the proapoptotic activity of p53. Cell 112:181-192.[CrossRef][Medline]

12. Guerra, C., N. Mijimolle, A. Dhawahir, P. Dubus, M. Barradas, M. Serrano, V. Campuzano, and M. Barbacid. 2003. Tumor induction by an endogenous K-ras oncogene is highly dependent on cellular context. Cancer Cell 4:111-120.[CrossRef][Medline]

13. Hamad, N. M., J. H. Elconin, A. E. Karnoub, W. Bai, J. N. Rich, R. T. Abraham, C. J. Der, and C. M. Counter. 2002. Distinct requirements for Ras oncogenesis in human versus mouse cells. Genes Dev. 16:2045-2057.[Abstract/Free Full Text]

14. Hirai, H., and C. J. Sherr. 1996. Interaction of D-type cyclins with a novel myb-like transcription factor, DMP1. Mol. Cell. Biol. 16:6457-6467.[Abstract]

15. Inoue, K., and C. J. Sherr. 1998. Gene expression and cell cycle arrest mediated by transcription factor DMP1 is antagonized by D-type cyclins through a cyclin-dependent kinase-independent mechanism. Mol. Cell. Biol. 18:1590-1600.[Abstract/Free Full Text]

16. Inoue, K., C. J. Sherr, and L. H. Shapiro. 1998. Regulation of the CD13/aminopeptidase N gene by DMP1, a transcription factor antagonized by D-type cyclins. J. Biol. Chem. 273:29188-29194.[Abstract/Free Full Text]

17. Inoue, K., M. F. Roussel, and C. J. Sherr. 1999. Induction of ARF tumor suppressor gene expression and cell cycle arrest by transcription factor DMP1. Proc. Natl. Acad. Sci. USA 96:3993-3998.[Abstract/Free Full Text]

18. Inoue, K., R. Wen, J. E. Rehg, M. Adachi, J. L. Cleveland, M. F. Roussel, and C. J. Sherr. 2000. Disruption of the ARF transcriptional activator DMP1 facilitates cell immortalization, Ras transformation, and tumorigenesis. Genes Dev. 14:1797-1809.[Abstract/Free Full Text]

19. Inoue, K., F. Zindy, D. H. Randle, J. E. Rehg, and C. J. Sherr. 2001. Dmp1 is haplo-insufficient for tumor suppression and modifies the frequencies of Arf and p53 mutations in Myc-induced lymphomas. Genes Dev. 15:2934-2939.[Abstract/Free Full Text]

20. Joneson, T., M. A. White, M. H. Wigler, and D. Bar-Sagi. 1996. Stimulation of membrane ruffling and MAP kinase activation by distinct effectors of RAS. Science 271:810-812.[Abstract]

21. Kamijo, T., F. Zindy, M. F. Roussel, D. E. Quelle, J. R. Downing, R. A. Ashmun, G. Grosveld, G., and C. J. Sherr. 1997. Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF. Cell 91:649-659.[CrossRef][Medline]

22. Kamijo, T., S. Bodner, E. van de Kamp, D. H. Randle, and C. J. Sherr. 1999. Tumor spectrum in ARF-deficient mice. Cancer Res. 59:2217-2222.[Abstract/Free Full Text]

23. Katz, M. E., and F. McCormick. 1997. Signal transduction from multiple Ras effectors. Curr. Opin. Genet. Dev. 7:75-79.[CrossRef][Medline]

24. Kuo, M. L., E. J. Duncavage, R. Mathew, W. den Besten, D. Pei, D., Naeve, T. Yamamoto, C. Cheng, C. J. Sherr, and M. F. Roussel. 2003. Arf induces p53-dependent and -independent antiproliferative genes. Cancer Res. 63:1046-1053.[Abstract/Free Full Text]

25. Lin, A. W., M. Barradas, J. C. Stone, L. van Aelst, M. Serrano, and S. W. Lowe. 1998. Premature senescence involving p53 and p16 is activated in response to constitutive MEK/MAPK mitogenic signaling. Genes Dev. 12:3008-3019.[Abstract/Free Full Text]

26. Lowe, S., and C. J. Sherr. 2003. Tumor suppression by Ink4a-Arf: progress and puzzles. Curr. Opin. Genet. Dev. 13:77-83.[CrossRef][Medline]

27. Maeda, S., and M. Karin. 2003. Oncogene at last: c-Jun promotes liver cancer in mice. Cancer Cell 3:102-104.[CrossRef][Medline]

28. McMahon, M. 2001. Steroid receptor fusion proteins for conditional activation of Raf-MEK-ERK signaling pathway. Methods Enzymol. 332:401-417.[Medline]

29. McMahon, M., and D. Woods. 2001. Regulation of the p53 pathway by Ras, the plot thickens. Biochem. Biophys. Acta 1471:M63-M71.[Medline]

30. Mechta-Grigoriou, F., D. Gerald, and M. Yaniv. 2001. The mammalian Jun proteins: redundancy and specificity. Oncogene 20:2378-2389.[CrossRef][Medline]

31. Mueller, M. M., W. Peter, M. Mappes, A. Huelsen, H. Steinbauer, P. Boukamp, M. Vaccariello, J. Garlick, and N. E. Fusenig. 2001. Tumor progression of skin carcinoma cells in vivo promoted by clonal selection, mutagenesis, and autocrine growth regulation by granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor. Am. J. Pathol. 159:1567-1579.[Abstract/Free Full Text]

32. Ohtani, N., Z. Zebedee, T. J. Huot, J. A. Stinson, M. Sugimoto, Y. Ohashi, A. D. Sharrocks, G. Peters, and E. Hara. 2001. Opposing effects of Ets and Id proteins on p16INK4a expression during cellular senescence. Nature 409:1067-1070.[CrossRef][Medline]

33. Palmero, I., C. Pantoja, and M. Serrano. 1998. p19ARF links the tumour suppressor p53 to ras. Nature 395:125-126.[CrossRef][Medline]

34. Palmero, I., M. Murga, A. Zubiaga, and M. Serrano. 2002. Activation of ARF by oncogenic stress in mouse fibroblasts is independent of E2F1 and E2F2. Oncogene 21:2939-2947.[CrossRef][Medline]

35. Passegue, E., and E. F. Wagner. 2000. JunB suppresses cell proliferation by transcriptional activation of p16INK4a expression. EMBO J. 19:2969-2979.[CrossRef][Medline]

36. Passegue, E., W. Jochum, A. Behrens, R. Ricci, and E. F. Wagner. 2002. JunB can substitute for Jun in mouse development and cell proliferation. Nat. Genet. 30:158-166.[CrossRef][Medline]

37. Quintanilla, M., K. Brown, M. Ramsden, and A. Balmain. 1986. Carcinogen-specific mutation and amplification of Ha-ras during mouse skin carcinogenesis. Nature 322:78-80.[Medline]

38. Quon, K. C., and A. Berns. 2001. Haplo-insufficiency? Let me count the ways. Genes Dev. 15:2917-2921.[Free Full Text]

39. Ries, S., C. Biederer, D. Woods, O. Shifman, S. Shirasawa, T. Sasazuki, M. McMahon, M. Oren, and F. McCormick. 2000. Opposing effects of Ras on p53: transcriptional activation of mdm2 and induction of p19ARF. Cell 103:321-330.[CrossRef][Medline]

40. Rocha, S., K. J. Campbell, and N. D. Perkins. 2003. p53- and Mdm2-independent repression of NF-kappa B transactivation by the ARF tumor suppressor. Mol. Cell 12:15-25.[CrossRef][Medline]

41. Rowland, B. D., S. G. Denissov, S. Douma, H. G. Stunnenberg, R. Bernards, and D. S. Peeper. 2002. E2F transcriptional repressor complexes are critical downstream targets of p19ARF/p53-induced proliferative arrest. Cancer Cell 2:55-65.[CrossRef][Medline]

42. Ruas, M., and G. Peters. 1998. The p16INK4a/CDKN2A tumor suppressor and its relatives. Biochem. Biophys. Acta Rev. Cancer 1378:F115-F177.[CrossRef]

43. Schreiber, M., A. Kolbus, F. Piu, A. Szabowski, U. Mohle-Steinlein, J. Tian, M. Karin, P. Angel, and E. F. Wagner. 1999. Control of cell cycle progression by c-Jun is p53 dependent. Genes Dev. 13:607-619.[Abstract/Free Full Text]

44. Serrano, M., A. W. Lin, M. E. McCurrach, D. Beach, and S. W. Lowe. 1997. Oncogenic ras provokes premature senescence associated with accumulation of p53 and p16INK4a. Cell 88:593-602.[CrossRef][Medline]

45. Sewing, A., B. Wiseman, A. C. Lloyd, and H. Land. 1997. High-intensity Raf signal causes cell cycle arrest mediated by p21Cip1. Mol. Cell. Biol. 17:5588-5597.[Abstract]

46. Shaulian, E., and M. Karin. 2002. AP-1 as a regulator of cell life and death. Nat. Cell Biol. 4:E131-E136.[CrossRef][Medline]

47. Sherr, C. J. 2001. The INK4a/ARF network in tumor suppression. Nat. Rev. Mol. Cell Biol. 2:731-737.[CrossRef][Medline]

48. Sherr, C. J., and R. A. DePinho. 2000. Cellular senescence: mitotic clock or culture shock? Cell 102:407-410.[CrossRef][Medline]

49. Sherr, C. J., and J. D. Weber. 2000. The ARF/p53 pathway. Curr. Opin. Genet. Dev. 10:94-99.[CrossRef][Medline]

50. Spaargaren, M., and J. R. Bischoff. 1994. Identification of the guanine nucleotide dissociation stimulator for Ral as a putative effector molecule of R-ras, H-ras, K-ras, and Rap. Proc. Natl. Acad. Sci. USA 91:12609-12613.[Abstract/Free Full Text]

51. Tanaka, T., M. Kurokawa, K. Ueki, K. Tanaka, Y. Imai, K. Mitani, K. Okazaki, N. Sagata, Y. Yazaki, Y. Shibata, T. Kadowaki, and H. Hirai. 1996. The extracellular signal-regulated kinase pathway phosphorylates AML1, an acute myeloid leukemia gene product, and potentially regulates its transactivation ability. Mol. Cell. Biol. 16:3967-3979.