Physical Interaction of p73 with c-Myc and MM1, a c-Myc-binding Protein, and Modulation of the p73 Function*

p73 shares high sequence homology with the tumor suppressor p53. Like p53, ectopic overexpression of p73 induces cell cycle arrest and/or apoptosis, and these biological activities are linked to its sequence-specific transactivation function. The COOH-terminal region of p73 is unique and has a function to modulate DNA-binding ability and transactivation activity. To identify and characterize cellular proteins that interact with the COOH-terminal region of p73α and regulate its activity, we employed a yeast-based two-hybrid screen with a human fetal brain cDNA library. We found MM1, a nuclear c-Myc-binding protein, was associated with p73α in both yeast two-hybrid and in vitro pull-down assays. In mammalian cells, MM1 co-immunoprecipitated with p73α, whereas p73β and tumor suppressor p53 did not interact with MM1. Overexpression of MM1 in p53-deficient osteosarcoma SAOS-2 cells enhanced the p73α-dependent transcription from the p53/p73-responsiveBax and PG13 promoters, whereas p73β- and p53-mediated transcriptional activation was unaffected in the presence of MM1. MM1 also stimulated the p73α-mediated growth suppression in SAOS-2 cells. More importantly, we found that c-Myc was physically associated with p73α and significantly impaired the transcriptional activity of p73α on Bax and p21 waf1 promoters. Expression of MM1 strongly reduced the c-Myc-mediated inhibitory activity on p73α. These results suggest that MM1 may act as a molecular partner for p73 to prevent the c-Myc-mediated inhibitory effect on its activity.

p73 shares high sequence homology with the tumor suppressor p53. Like p53, ectopic overexpression of p73 induces cell cycle arrest and/or apoptosis, and these biological activities are linked to its sequence-specific transactivation function. The COOH-terminal region of p73 is unique and has a function to modulate DNA-binding ability and transactivation activity. To identify and characterize cellular proteins that interact with the COOH-terminal region of p73␣ and regulate its activity, we employed a yeast-based two-hybrid screen with a human fetal brain cDNA library. We found MM1, a nuclear c-Myc-binding protein, was associated with p73␣ in both yeast two-hybrid and in vitro pull-down assays. In mammalian cells, MM1 co-immunoprecipitated with p73␣, whereas p73␤ and tumor suppressor p53 did not interact with MM1. Overexpression of MM1 in p53-deficient osteosarcoma SAOS-2 cells enhanced the p73␣-dependent transcription from the p53/p73-responsive Bax and PG13 promoters, whereas p73␤-and p53-mediated transcriptional activation was unaffected in the presence of MM1. MM1 also stimulated the p73␣-mediated growth suppression in SAOS-2 cells. More importantly, we found that c-Myc was physically associated with p73␣ and significantly impaired the transcriptional activity of p73␣ on Bax and p21 waf1 promoters. Expression of MM1 strongly reduced the c-Myc-mediated inhibitory activity on p73␣. These results suggest that MM1 may act as a molecular partner for p73 to prevent the c-Mycmediated inhibitory effect on its activity.
p73 is a new member of the p53 gene family (1). Like p53, p73 is a nuclear transcription factor, which carries an NH 2 -terminal transactivation domain, sequence-specific DNA-binding domain, and oligomerization domain. As expected from the significant amino acid sequence homology in the sequence-specific DNA-binding domain between p73 and p53, p73 recognizes and binds to the p53-responsive elements found within the promoter regions of the various p53-target genes. In transiently transfected mammalian cells, p73 transactivates the transcription from a variety of p53-responsive promoters to various degrees (1)(2)(3)(4)(5)(6)(7)(8). Artificially introduced mutation within the DNA-binding domain has significantly reduced the transactivation activity of p73, suggesting that the structural integrity of this domain is required for this activity (1,2). Similarly to p53, the cellular transcriptional coactivator p300/CBP interacts with the NH 2 -terminal transactivation domain of p73, resulting in stimulation of its activity (9). Furthermore, ectopic overproduction of p73 induces the cell cycle arrest and/or apoptosis in p53-deficient cultured cells (1)(2)(3)(4)(5)(6)10). Recently, it has been shown that p73 is stabilized and its apoptotic activity is enhanced in response to ionizing radiation or genotoxic agents such as cisplatin in a pathway depending on c-Abl (11)(12)(13). In addition, the endogenous level of p73 is increased during retinoic acid-induced differentiation in cultured neuroblastoma cells, and overexpression of p73 but not p53 caused neuronal differentiation (14). Fang et al. (7) have found that overexpression of p73 induces the growth arrest and the senescence-like phenotypes in human bladder carcinoma cells.
p73 is assigned to chromosome 1p36.3, which is a candidate tumor suppressor locus in a variety of human cancers (1). Although p73 mimics p53 in transcriptional activation as well as induction of apoptosis, p73 is infrequently mutated in many human tumors (15). In contrast to p53-knockout mice, p73 deficiency in mice did not lead to an increased susceptibility to spontaneous tumorigenesis (16). In addition, the elevated level of p73 expression was detected in some primary tumors, including breast and ovarian cancers (17,18). Subsequent work from several laboratories has shown that forced expression of cellular and viral oncogenes, such as E2F-1, c-Myc, and E1A, stimulated expression of the p73 gene (19 -22). Thus, there have been conflicting reports about the role of p73 in cellular function, although the reasons remain unclear.
Accumulating evidence indicates that homotypic and hetero-* This work was supported by a grant-in-aid from the Ministry of Health and Welfare for a New 10-Year Strategy for Cancer Control, a grant-in-aid for Scientific Research on Priority Areas, and a grant-inaid for Scientific Research (B) from the Ministry of Education, Science, Sports and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ Both authors contributed equally to this work.  (1,3). Di Como et al. (5) found that tumor-derived p53 mutants but not wild-type p53 were associated with p73␣ and thereby reduced its transactivation and pro-apoptotic function. Likewise, p53 mutant also interacted with the remaining p73 splicing isoforms (␤, ␥, and ␦) through its DNA-binding domain, and abrogated their transcriptional activities (26). The ability of mutant p53 to interact with p73 was regulated by the status of a common p53 polymorphism at codon 72 (27). Recently, Yang et al. (16) discovered the truncated p73 isoform (⌬Np73), 1 which lacks the NH 2terminal transactivation domain. ⌬Np73 was generated from an alternative promoter located within the intron 3 of the p73 gene and lost the transactivation ability toward the p53-responsive promoter. Of note, ⌬Np73 was predominantly expressed in the developing brain and sympathetic neurons, and inhibited the pro-apoptotic function of p53 by hetero-oligomerization (29). These homotypic and heterotypic interactions among p53 family members give a complexity to the understanding of the p73 signaling in vivo.
Various lines of evidence suggest that the extreme COOHterminal region of p53 has a function of negative regulator. The COOH terminus of p53 directly binds and masks its central DNA-binding domain (30,31). Indeed, its inhibitory effect was removed by structural modifications such as phosphorylation, glycosylation, acetylation, or deletions (32)(33)(34)(35)(36). Additionally, physical interaction of Ref-1 (previously identified as the AP-1-stimulating protein (37)) or 14-3-3 with the COOH-terminal region of p53 stimulated its DNA-binding activity (38,39). Like p53, the COOH-terminal deletion of p73␣ caused the significant increase in its DNA-binding and transactivation activities (8,25,40). These observations suggest that the p73-DNA crystal structure might be similar to that of p53, although the COOH-terminal region of p73 does not share amino acid sequence similarity with that of p53.
The purpose of this study was initially to isolate and characterize cellular protein(s) that could associate with the COOH-terminal region of p73␣. By using a yeast-based twohybrid screening, we identified MM1, which had been reported to be a c-Myc-binding protein (41), as a p73␣ COOH-terminal region-binding protein. We found that overexpression of MM1 stimulated the p73␣-mediated transcription from some p53/ p73-responsive promoters as well as growth suppression. Moreover, c-Myc bound to p73␣ and inhibited the p73␣-dependent transactivation. Of interest, overexpression of MM1 antagonized the inhibitory effect of c-Myc on p73␣.

EXPERIMENTAL PROCEDURES
Cell Culture-Human osteosarcoma SAOS-2, COS7, and human embryonic kidney 293 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen, Grand Island, NY) and penicillin (100 IU/ml)/streptomycin (100 g/ml). HL60 cells were grown in RPMI 1640 supplemented with 15% heat-inactivated fetal bovine serum and antibiotic mixture. Cultures were maintained at 37°C in a water-saturated atmosphere of 5% CO 2 in air.
Transfection-Transient transfection of SAOS-2 cells was performed by LipofectAMINE Plus reagent according to the manufacturer's recommendations (Invitrogen). COS7 and 293 cells were transfected with the indicated expression plasmids using FuGENE6 transfection reagent (Roche Molecular Biochemicals, Indianapolis, IN) in accordance with the manufacturer's specifications.
Yeast Two-hybrid Screening-The MATCHMAKER Two-Hybrid System 2 was purchased from CLONTECH Laboratories, Inc. (Palo Alto, CA). The cDNA encoding the extreme COOH-terminal region of p73␣ (amino acid residues 551-636) was amplified by PCR-based strategy using the full-length p73␣ cDNA as a template. The PCR product, which was produced by an additional EcoRI site in 5Ј-upstream and BamHI site in 3Ј-downstream, was digested completely with EcoRI and BamHI and subcloned in-frame into the identical restriction sites of pAS2-1 to generate a "bait" plasmid, pAS2-1-p73␣COOH. The resulting bait plasmid was used to isolate the cDNA for p73␣-binding protein from a cDNA library derived from human fetal brain or 293 cells in the pACT yeast expression vector (CLONTECH Laboratories, Inc.). Both plasmids were introduced into the Saccharomyces cerevisiae strain CG1945 (MATa, ura3-52, his3-200, ade2-101, lys2-801, trp1-901, leu2-3,112, gal4 -542, gal80 -538, cyh r 2, LYS2::GAL1 UAS -GAL1 TATA -HIS3, URA3:: GAL4 17-mers(X3) -CYC1 TATA -lacZ) using the lithium acetate/heat-shock protocol (42). The transformed yeast cells were plated on SD medium lacking tryptophan, leucine, and histidine in the presence of 20 mM 3-amino-1,2,4-triazole and incubated for 1 week at 30°C. Positive colonies were picked up and assayed for lacZ activity using a filter ␤-galactosidase assay, as described previously (43). The interacting cDNA clones were rescued from the selected yeast transformants. ␤-Galactosidase activity in the yeast two-hybrid system was determined by a liquid assay using an o-nitrophenylgalactopyranoside according to the manufacturer's instructions. The nucleotide sequences of the positive cDNA clones were determined by the dideoxy terminator cycle sequencing method using an automated Prism 377 DNA sequencer (PerkinElmer Life Sciences and Applied Biosystems, Foster City, CA) and nucleotide sequence data bases were searched for homologous sequences using the BLAST program.
Plasmid Constructs-The mammalian expression plasmid encoding hemagglutinin (HA) epitope-tagged p73␣ or p73␤ was a generous gift from Dr. Mourad Kaghad (Sanofi Recherche, Paris, France). The p53 expression plasmid and the p53/p73-responsive reporter constructs were kindly provided by Dr. Bert Vogelstein (Johns Hopkins University, Baltimore, MD). The truncated form of p73␣ was generated as described previously (40). The full-length cDNA encoding human MM1 was amplified by PCR using the C42 cDNA present in pACT2 as a template. The primers used for PCR were: sense, 5Ј-CCAGGAATTCG-GGGTTGATGTCATGACTGTAGTCGGCCTTCCCAACATGGCAGTC-TATTAACATCACGGAGCTGAAT-3Ј (the EcoRI recognition site is underlined); antisense, 5Ј-GCAACTCGAGTCAGGCCTTAGCAG-3Ј (the XhoI restriction site is underlined). The PCR primers contained the restriction enzyme sites to facilitate the subsequent cloning step. The PCR product was subcloned into pGEM-T Easy (Promega), and its nucleotide sequence was verified by automated dideoxy terminator cycle sequencing. The PCR product, which was produced by additional EcoRI site in the 5Ј-upstream and XhoI site in the 3Ј-downstream, was digested with EcoRI and XhoI and subcloned into the identical restriction sites of the pcDNA3-FLAG (kindly provided by Dr. Toshiharu Suzuki, University of Tokyo, Tokyo, Japan) to give pcDNA3-FLAG-MM1.
Generation of a Bacterial Expression Construct-The cDNA clone (C42) encoding the 150-amino acid sequence of human MM1 (amino acid residues 18 -167) was digested with EcoRI and XhoI and subcloned into the EcoRI and SalI restriction sites of pMAL-cRI, a maltosebinding protein (MBP) fusion vector (New England BioLabs Inc., Beverly, MA), to create pMAL-MM1- (18 -167). Cultures of Escherichia coli DH5␣ harboring the pMAL-cR1 or pMAL-MM1-(18 -167) were induced with 1 mM isopropyl-␤-D-thiogalactopyranoside for 6 h at 37°C. Cells were harvested by centrifugation at 4000 rpm for 5 min at 4°C, and the bacterial pellet was resuspended in NETN buffer (50 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.1% Nonidet P-40, and 1 mM EDTA). After the brief sonication, the lysate was centrifuged at 12,000 rpm for 30 min at 4°C to remove insoluble materials. The cleared supernatant was incubated with 1/10th volume of pre-equilibrated amylose resin (New England BioLabs Inc.) for 30 min at 4°C. MBP fusion proteins bound to the beads were recovered with an elution buffer containing 10 mM maltose and dialyzed against ice-cold phosphate-buffered saline (PBS). The purity of the recovered proteins was analyzed by SDSpolyacrylamide gel electrophoresis and Coomassie Brilliant Blue staining. Protein concentration was determined using Bradford protein assay (Bio-Rad, Hercules, CA).
In Vitro Interaction Assay-p73␣ or p53 was generated in vitro in the presence of [ 35 S]methionine using the quick-coupled in vitro transcrip-tion and translation system (TnT) according to the procedure suggested by the manufacturer (Promega Corp., Madison, WI). The quality of the synthesized proteins was verified by electrophoresis through 10% SDSpolyacrylamide gel and autoradiography. For the in vitro pull-down assays, radiolabeled p73␣ or p53 (10 l of a 50-l reaction) was mixed with the MBP or MBP fusion protein (1 g) in NETN buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and 1 mg/ml of bovine serum albumin. Following incubation in the presence of amylose resin for 2 h at 4°C with gentle shaking, the beads were washed three times with 1 ml of the binding buffer. The bound 35 S-labeled proteins were then eluted by boiling in the SDS sample buffer (62.5 mM Tris-Cl, pH 6.8, 2% SDS, 10% glycerol, 0.1 M dithiothreitol, and 0.01% bromphenol blue) for 5 min and analyzed on a 10% SDS-polyacrylamide gel. Following electrophoresis, gel was destained, dried, and exposed to an x-ray film with an intensifying screen at Ϫ80°C.
Immunocytochemistry-COS7 cells were doubly transfected with 1 g of pcDNA3-FLAG-MM1 and 1 g of the expression vector for HA-p73␣ using FuGENE6 reagent. Forty-eight hours post-transfection, cells grown on glass coverslips were fixed with 3.7% formaldehyde in PBS for 30 min at room temperature, permeabilized with 0.2% Triton X-100 for 5 min at room temperature and blocked for 1 h in PBS containing 3% bovine serum albumin. After washing extensively with PBS, cells were incubated with the polyclonal anti-HA antibody (diluted 1:500, Medical and Biological Laboratories, Nagoya, Japan), and the monoclonal anti-FLAG M2 antibody (diluted 1:50, Sigma Chemical Co., St. Louis, MO) for 1 h at room temperature. After incubation with primary antibodies, cells were washed and incubated with FITC-or rhodamine-conjugated secondary antibodies (Invitrogen) diluted 1:200 for 1 h at room temperature. Cells were finally washed in PBS, the coverslips were removed from the dishes, mounted onto slides, and observed under Fluoview laser scanning confocal microscope (Olympus, Tokyo, Japan). Production of Polyclonal Anti-MM1 Antibody-The polyclonal anti-MM1 antibody was raised against the glutathione S-transferase (GST)-MM1-(1-167) fusion protein. The specificity of the antibody was checked on its ability to immunoprecipitate MM1 expressed in 293 cells and its ability to detect MM1 by Western blot analysis.
Immunoblotting-COS7 cells were transfected with 2 g of the indicated expression plasmid and harvested at 48 h after transfection. Cells were washed with ice-cold PBS, lysed in 400 l of EBC buffer (50 mM Tris-Cl, pH 7.5, 120 mM NaCl, 0.5% Nonidet P-40, and 1 mM PMSF) containing protease inhibitor mixture (Sigma), and the extracts were sonicated for 10 s and centrifuged at 15,000 rpm for 10 min to remove insoluble materials (2). The protein concentrations were determined by Whole cell lysates were immunoprecipitated with the monoclonal antibodies to p53 (DO-1 and PAb1801) and subjected to the immunoblot analysis using the polyclonal anti-MM1 antibody (top). The expression of endogenous p53 and FLAG-MM1 was monitored by Western blot analysis (middle and bottom panels, respectively). E, nuclear co-localization of MM1 and p73␣. COS7 cells were transiently cotransfected with the expression plasmids for HA-p73␣ and FLAG-MM1. Forty-eight h post-transfection, cells were fixed, and incubated with the polyclonal anti-HA and the monoclonal anti-FLAG antibodies. Expression of HA-p73␣ and FLAG-MM1 was visualized with FITC-conjugated anti-rabbit IgG (green) and with rhodamine-conjugated anti-mouse IgG (red), respectively. The merged image suggests the nuclear co-localization of MM1 and p73␣. with 1 g of the expression plasmid for HA-p73␣, and 1 g of pcDNA3-FLAG-MM1 using FuGENE 6 reagent. Forty-eight h posttransfection, cells were harvested and cell lysates were prepared in 400 l of the EBC buffer, which were spun at 15,000 rpm for 20 min at 4°C to remove insoluble materials. The precleared soluble supernatants were mixed with the polyclonal anti-MM1 antibody and incubated for 2 h at 4°C. Protein A-Sepharose beads were then added to the reaction mixtures and incubated for 1 h at 4°C. The immune complexes were washed with the lysis buffer three times at 4°C, and the bound proteins were eluted by boiling in the SDS-sample buffer. Proteins were then analyzed by 10% SDS-polyacrylamide gel electrophoresis, semi-dryblotted onto a nitrocellulose membrane, and probed with the monoclonal anti-p73 antibody (Ab-4, NeoMarkers, Inc.). Immuno-complexes were detected by ECL (Amersham Biosciences, Inc.).
Luciferase Assays-SAOS-2 cells were plated for transfection at a density of 5 ϫ 10 4 cells/well in a 12-well tissue culture dish for 24 h. Cells were co-transfected with 100 ng of the indicated p53/p73-responsive reporter plasmid (p21 waf1 , MDM2, Bax, or PG13), 10 ng of pRL-TK Renilla luciferase cDNA, and 25 ng of the indicated expression plasmid (p53, p73␣, p73␣-(1-427), or p73␤) in the presence or absence of the increasing amounts of the expression plasmid for FLAG-MM1. The total amount of DNA was kept constant (510 ng) with pcDNA3 (Invitrogen, Carlsbad, CA) per transfection. Forty-eight hours post-transfection, cells were lysed and luciferase activity was measured by using the dual-luciferase reporter assay system (Promega, Corp.) according to the manufacturer's instructions. The transfection efficiency was standardized against Renilla luciferase.
Cell Growth Assay-SAOS-2 cells were plated on a 12-well tissue culture dish (5 ϫ 10 4 cells/well) and transiently transfected with 125 ng of the reporter plasmid pCH110 (Amersham Biosciences, Inc.), which encodes E. coli ␤-galactosidase, plus 12.5 ng of the expression plasmid for p73␣ or p53 in the presence or absence of 187 ng of the expression plasmid encoding FLAG-MM1 by the LipofectAMINE procedure. The total amount of transfected plasmid DNA was kept constant (500 ng/ transfection) by adding pcDNA3. At 48 h post-transfection, cells were fixed with 0.25% glutaraldehyde in PBS, the transfected cells were detected by staining with 5-bromo-4-chloro-3-indolyl-␤-D-galactopyranoside (X-gal), and the number of ␤-galactosidase expressing cells was scored. The experiments were repeated three times.
Etoposide Treatment-HL60 cells were seeded at a density of 5 ϫ 10 5 cells/tissue culture plates in RPMI medium. Etoposide was then added to the media to a final concentration of 68 M and incubated for 0 -8 h as indicated. At the indicated time periods after the treatment of etoposide, total RNA was extracted by using the RNeasy Mini kit (Qiagen, Inc., Valencia, CA) according to the manufacturer's protocol. The cDNA was synthesized from 5 g of total RNA using reagents from Super-Script II (Invitrogen) and then used for PCR-based amplification. Oligonucleotides used to amplify c-myc, MM1, and p73 mRNAs were as follows: sense primer for c-myc, 5Ј-CTCGACTACGACTCGGTGCA-3Ј and antisense primer for c-myc, 5Ј-TGGTGGGCGGTGTCTCCTCA-3Ј; sense primer for MM1, 5Ј-GGGGGTTGATGTCATGACTG-3Ј and antisense primer for MM1, 5Ј-TCAGGCCTTAGCAGTAGCCT-3Ј; sense primer for p73, 5Ј-CTTTGAGGGCCGCATCTG-3Ј and antisense primer for p73, 5Ј-GCCCCAGGTTGGGTGTAG-3Ј. The PCR products were electrophoretically separated on 1.5% agarose gels and visualized by ethidium bromide staining. Flow cytometric analysis was used to determine the apoptotic cells. At the indicated time periods after the addition of etoposide, cells were trypsinized, fixed with ice-cold ethanol, and stained with propidium iodide (44). The DNA content of the cells was analyzed by flow cytometry (FACScan, Becton Dickinson, Oxford, UK).

Isolation of MM1 as a p73␣-binding Protein in a Two-hybrid
Screen-To isolate one or more cellular proteins that could interact with the unique COOH-terminal region of p73␣ and regulate its activity, we used a yeast-based two-hybrid system to screen a human fetal brain cDNA library with a "bait" plasmid (pAS2-1-p73␣COOH) encoding the extreme COOHterminal portion (amino acid residues 551-636) of p73␣ (Fig.  1A). A yeast strain CG1945 was co-transformed with the bait plasmid and the human fetal brain cDNA library, and the yeast colonies showing positive signals for nutrient (His) selection and ␤-galactosidase activity were selected. Of a total of 1 ϫ 10 6 primary transformants, 34 colonies grew on the selection medium lacking tryptophan, leucine, and histidine, and 10 out of the 34 His-positive transformants formed blue colonies. Plasmids carrying these 10 positive candidates were rescued into E. coli, and their nucleotide sequences were determined. One clone, termed C42, contained a partial human cDNA for MM1, which had previously been reported to be a c-Myc-binding protein (41). The cDNA insert of this clone encoded a peptide ranging from amino acids 18 to 167 of MM1 (Fig. 1B). To assess the binding activity of C42 with the COOH-terminal region of p73␣, yeast cells were co-transformed with various combinations of the constructs, and the ␤-galactosidase activity (Miller units) of each transformant was measured and compared. Transformants lacking the p73␣ COOH-terminal region or C42 were negative for the ␤-galactosidase activity, whereas those expressing both constructs induced the enzymatic activity (Fig.  1C). These data suggested that MM1 interacts with p73␣ in yeast.
Specificity of the MM1⅐p73␣ Interaction in Vitro-To confirm the results obtained from the yeast two-hybrid assay, we first examined the MM1⅐p73␣ interaction with in vitro pull-down assay using MBP-MM1-(18 -167) fusion protein ( Fig. 2A). Amylose resin bearing MBP or MBP-MM1-(18 -167) was incubated with [ 35 S]methionine-labeled full-length p73␣ generated in vitro in the coupled transcription/translation system. After washing the beads extensively with the binding buffer, the MBP-or MBP-MM1-(18 -167)-bound proteins were eluted by boiling. Following the SDS-polyacrylamide gel electrophoresis, protein complexes were detected by autoradiography. As shown in Fig. 2B, we observed that the MBP fusion protein containing MM1-(18 -167) was associated with radiolabeled p73␣, whereas the MBP control was not. We also tested MM1 for its ability to interact with p53. As seen in Fig. 2B, we were unable to detect the interaction between MBP-MM1-(18 -167) and radiolabeled p53. These results suggested that p73␣ but not p53 directly binds to MM1.
In Vivo Interaction between MM1 and p73␣-To demonstrate the interaction of MM1 with p73␣ in mammalian cells, human full-length MM1 tagged with the FLAG peptide on its NH 2 terminus (FLAG-MM1) was prepared and introduced into COS7 cells. Whole cell lysates from COS7 cells transiently overexpressing HA-p73␣, FLAG-MM1, or HA-p73␣, and FLAG-MM1 were immunoprecipitated with the polyclonal anti-MM1 antibody. The immunoprecipitates were analyzed by SDS-polyacrylamide gel electrophoresis and subsequent immunoblotting with the monoclonal anti-p73 antibody. As shown in Fig.  3A, HA-p73␣ was co-immunoprecipitated with FLAG-MM1. The expression of HA-p73␣ and FLAG-MM1 in COS7 cells was confirmed by immunoblotting with anti-p73 and anti-MM1 antibody, respectively (Fig. 3A, lower panels). To determine the MM1-binding region of p73␣, p73␣ deletion mutants were used for the immunoprecipitation experiments with MM1. As expected, these COOH-terminal deletion mutants failed to coprecipitate with MM1 (Fig. 3B). We also tested MM1 for its ability to interact with p73␤ or p53 by co-immunoprecipitation analysis. As shown in Fig. 3 (C and D), FLAG-MM1 was not co-immunoprecipitated with HA-p73␤ or p53. These results suggested that p73␣ but not p53 and p73␤ interacted with MM1 in mammalian cells, and the extreme COOH-terminal region of p73␣ was required for this interaction.
To confirm the above co-immunoprecipitation analysis, we examined the subcellular distribution of MM1 and p73␣. COS7 cells were transiently co-transfected with the expression plasmids for FLAG-MM1 and HA-p73␣, and were double-stained with anti-FLAG and anti-HA antibodies. The transfected cells were then observed under a confocal laser scanning microscope. As reported previously (2, 41), MM1 or p73␣ was almost exclusively located in the nucleus (Fig. 3E). Upon close inspection of FIG. 4. Differential effects of MM1 on the p73␣-mediated transcriptional activation. p53-deficient SAOS-2 cells (5 ϫ 10 4 cells/well) were transiently co-transfected with 25 ng of the expression plasmid for p73␣ (A), p73␤ (B), p73␣-(1-427) (C), or p53 (D) along with the luciferase the individual and merged images, MM1 signal appeared to overlap with that of p73␣ (Fig. 3E), indicating that these two proteins co-localize in the nucleus.
Effect of MM1 on the p73-mediated Transcriptional Activation and Growth Suppression-To determine whether or not MM1 affects the p73-or p53-dependent transcriptional activation in cells, p53/p73-responsive Bax, PG13, which carries 13 copies of the consensus p53-responsive element, p21 waf1 , or MDM2-luciferase reporter construct was co-transfected with the expression plasmid for HA-p73␣, HA-p73␤, HA-p73␣-(1-427), or p53 in the presence or absence of the increasing amounts of FLAG-MM1 expression plasmid into p53-deficient human osteosarcoma SAOS-2 cells. Under our experimental conditions, endogenous p73 and MM1 could not be detected, and the levels of the ectopically expressed proteins were not affected in the presence of MM1 (data not shown). Expression of p73␣, p73␤, p73␣-(1-427), or p53 successfully activated transcription of each of those p53/p73-responsive reporters compared with the empty plasmid controls, and MM1 alone failed to induce these reporters (Fig. 4). The ability of p73␣ to drive transcription from Bax and PG13 reporter was enhanced by MM1 in a dose-dependent manner (Fig. 4A), whereas MM1 did not elevate the luciferase activity of those reporters induced by p73␤ or p53 (Fig. 4, B and D). These observations were consistent with the results showing that MM1 interacted with p73␣ but not with p73␤ or p53. In addition, MM1 failed to affect the transactivation function of p73␣-(1-427), which lacked the COOH-terminal region of p73␣ (Fig. 4C). On the other hand, MM1 did not enhance the MDM2-or p21 waf1 -driven transcription mediated by p73␣ (Fig. 4A). Similarly, p53-dependent luciferase activity driven by the p53/p73-responsive p21 waf1 or MDM2 promoter was not affected in the presence of exogenously expressed MM1 (data not shown). These results indicated that MM1 might determine the differential response of target genes to p73, although the precise mechanism defining those specificities remained to be determined.
In view of the ability of MM1 to modulate cellular function of p73␣, we next examined whether MM1 could affect the p73␣mediated growth inhibition. To this end, SAOS-2 cells were co-transfected with the expression plasmid for p53 or p73␣ in the presence or absence of MM1 expression plasmid together with ␤-galactosidase expression construct (pCH110) to identify transfected cells. The number of ␤-galactosidase-positive cells was measured 48 h post-transfection. As shown in Fig. 5 (top  panel), ectopic overexpression of p73␣ or p53 led to the significant decrease in number of ␤-galactosidase-positive cells compared with that transfected with the empty plasmid alone (compare lane 1 with lanes 3 or 6), whereas the number of ␤-galactosidase-positive cells expressing MM1 was similar to that observed in the empty plasmid-transfected cells (compare lane 1 with lane 2). At expression plasmid concentrations ranging from 12.5 to 375 ng, p73␣ and p53 decreased the number of ␤-galactosidase-positive cells in a dose-dependent manner (data not shown). To facilitate the detection of the possible effect of MM1 on p73␣ or p53, SAOS-2 cells were transfected with 12.5 ng of p73␣ or p53 expression plasmid, together with the expression plasmid for MM1. Under these experimental conditions, endogenous p73␣ and MM1 were undetectable, and the ectopically expressed proteins were readily detected by Western analysis (Fig. 5, bottom panel) As seen in Fig. 5 (top  panel), co-expression of p73␣ with MM1 significantly reduced the number of ␤-galactosidase-positive cells compared with that transfected with p73␣ expression plasmid alone, whereas expression of MM1 had no significant effect on the growth suppression induced by p53, suggesting that MM1 might cooperate with p73␣ in inhibiting the cell growth.
Physical and Functional Interaction among p73, MM1, and c-Myc-As reported previously, MM1 bound to the NH 2 -terminal transactivation domain of c-Myc and inhibited its E-boxdependent transcriptional activity (41). Recently, Ceballos et al. have found that c-Myc counteracts the transactivation of p21 waf1 promoter mediated by p53 (45). These observations prompted us to examine the role of c-Myc on p73 function. To determine the functional relevance of the p73⅐c-Myc interaction, we examined the effect of c-Myc on the p73␣-mediated transcriptional activation. SAOS-2 was co-transfected with the expression plasmid encoding p73␣ or p53 along with the p53/ p73-responsive Bax or p21 waf1 promoter luciferase reporter construct in the presence of c-Myc expression plasmid. Consistent with the previous results (45), p53-dependent transcriptional activation was repressed by co-expression with c-Myc (Fig. 6). Similarly, c-Myc abrogated the p73␣-mediated transcriptional activation. Intriguingly, unlike c-Myc, overexpression of N-Myc exhibited no detectable effect on the transactivation function of p73␣ and p53. The protein levels of ectopically expressed c-Myc and N-Myc were easily detected by reporter constructs (100 ng) as indicated, and 10 ng of the Renilla luciferase plasmid (pRL-TK) in the presence or absence of the increasing amounts of pcDNA3-FLAG-MM1 (125, 250, or 375 ng). The total amount of plasmid DNA per transfection was kept constant (510 ng) with pcDNA3. All transfections were performed in triplicate. Forty-eight h after transfection, cells were lysed, and the firefly luciferase activities were determined. The transfection efficiency was standardized against Renilla luciferase. Results are shown as -fold induction of the firefly luciferase activity compared with control cells transfected with pcDNA3 alone. Western analysis, whereas endogenously expressed c-Myc and N-Myc could not be detected (data not shown). We then examined whether MM1 could antagonize the inhibitory effect of c-Myc on the p73 function. To this end, we performed the luciferase reporter assays in transiently transfected SAOS-2 cells using the Bax promoter luciferase reporter construct. As expected, co-expression of MM1 with c-Myc and p73␣ apparently rescued the reporter activity of p73␣ suppressed by c-Myc (Fig. 7A). Unlike p73␣, c-Myc-mediated repression of the reporter activity of p53 was not affected in the presence of MM1 expression (Fig. 7B). These studies raised the possibility that MM1 is associated with p73␣ and/or c-Myc and suppresses the functional interaction between p73␣ and c-Myc.
To determine whether p73␣ could physically bind to c-Myc, and/or whether p73␣ could be a part of the c-Myc⅐MM1 complex, co-immunoprecipitation analysis was carried out. COS7 cells were transfected with the indicated combinations of the expression plasmids, and the whole cell lysates were immunoprecipitated with antibody to MM1 or p73 followed by immunoblotting with antibody to c-Myc. Consistent with the previ-ous report (41), the anti-MM1 immunoprecipitates contained c-Myc (Fig. 8A). Analysis of the anti-p73 or the anti-c-Myc immunoprecipitates also indicated that c-Myc can form a physical complex with p73␣ in vivo (Fig. 8B). In contrast, we could not detect the physical interaction between p73␣ and N-Myc (data not shown). To evaluate their interaction by confocal immunofluorescence microscopy, HA-p73␣ was co-expressed with FLAG-c-Myc in COS7 cells. Fig. 8C shows the confocal images from the same cell expressing both proteins. A significant co-localization of p73␣ and c-Myc to the nucleus was observed when both images were overlaid. To determine whether c-Myc could be associated with p73␣ in the presence of MM1, HA-p73␣ and FLAG-c-Myc were co-expressed with the increasing amounts of MM1 in COS7 cells. As shown in Fig. 8D, the amounts of p73␣ co-precipitated with c-Myc was decreased in the presence of the increasing amounts of MM1 (compare lanes 1, 2, and 3). Taken together, these results suggested that MM1 might prevent the c-Myc⅐p73␣ interaction and/or directly bind to c-Myc to inhibit its activity and thereby stimulate the p73␣ activity.
Etoposide Induces Expression of p73 and Down-regulation of c-myc-Recently, it has been shown that expression of p73 is significantly induced in response to various DNA-damaging agents such as etoposide, doxorubicin, and camptothecin (46). To determine the effect of etoposide on p73, MM1, and c-myc, HL60 cells (which do not contain p53 protein) were treated with etoposide, and the expression levels of p73, MM1, and c-myc were examined by RT-PCR analysis. As described previously (47), incubation of HL60 cells with etoposide resulted in the remarkable increase in the number of cells in sub-G 1 phase indicating the apoptosis (Fig. 9A). Under our experimental conditions, sub-G 1 fractions were not increased by the treatment with the dimethyl sulfoxide solvent alone (data not shown). We then prepared the total RNA from HL60 cells at the indicated time points after the treatment with etoposide, and RT-PCR analysis was carried out. In agreement with the previous reports, the expression level of p73 was significantly increased in response to etoposide (Fig. 9B). On the other hand, the level of MM1 was unaffected in the presence of etoposide, whereas etoposide treatment caused a significant decrease in c-myc level. Similar results were also obtained by Western blot analysis (Fig. 9C).

DISCUSSION
Our present results have revealed for the first time that p73 is directly associated with c-Myc and its repressor MM1. However, the functional interactions among them appear to be finely regulated. MM1 is able to bind only to p73␣ but not to p73␤ and stimulates transactivation ability of the former. Only c-Myc but not N-Myc binds to p73␣ and inhibits its transcriptional activity. In addition, the regulatory effect of MM1 on p73␣ function seems to be dependent on each target gene promoter. Nevertheless, our finding of a direct link through forming a complex among c-Myc oncogene product, p73 tumor suppressor, and MM1 should give a new insight into better understanding of molecular and cellular mechanism of cell cycle control and apoptosis.
MM1 has bound to p73␣ at the extreme COOH-terminal region, which regulates its transactivation activity and DNAbinding ability. Like p53, p73-mediated transcription and apoptosis are reported to be markedly enhanced by binding to p300/CBP (9), or the COOH-terminal deletions (8,25,40). The importance of the COOH-terminal region of p53 has so far been insisted. The extreme COOH-terminal region of p53 is closely involved in its pro-apoptotic function by protein⅐protein interactions (48). The physical interaction of the NH 2 -terminal region of p53 with the components of RNA polymerase II transcriptional machinery or with the transcriptional coactivators such as p300/CBP enhanced its transcriptional activity (49 -51). Similarly, the structural modifications within the COOHterminal region of p53 significantly facilitated its sequencespecific DNA binding (32)(33)(34)(35)(36)(37)(38)(39). In relation to these observations on p53, the importance of the COOH-terminal extension of p73␣ has also been implicated. There exist various functional motifs, including a PPPPY motif (amino acid residues 482-488) and a SAM domain (amino acid residues 484 -549) within that region. The SAM domain seems to be a protein⅐protein interaction module found in a variety of proteins involved in developmental regulation (52). Recently, it has been shown that Yes-associated protein is associated with p73␣ via its PPPPY motif and enhances the transcriptional activity of p73␣ (53). In addition, Minty et al. (54) have found that p73␣ but not p73␤ bound to SUMO-1, and the extreme COOH-terminal Lys residue of p73␣ (at position 627) is the major site for SUMO-1 modification. SUMO-1 modification alters the subcellular distribution of p73, although it did not affect the transcriptional activity of p73. Furthermore, the extreme COOH-terminal region of p73␣ is suggested to act as a negative regulator of its own function (8,40). In conjunction with those observations reported, our data suggest that the interaction of MM1 with the extreme COOH-terminal region allows p73␣ to take the conformation competent to express its activity, although the precise mechanism remains to be clarified.
Our present data have shown that the ability of p73␣ to drive the transcription from the Bax and the PG13 promoter is enhanced by overexpression of MM1. However, the effect of MM1 on the p73␣-mediated transcription from the MDM2 and p21 waf1 promoter is barely detectable. The similar pattern of target gene induction is also observed in some other cases. Mts1, which is associated with the extreme COOH-terminal region of p53, differentially regulates the transactivation function of p53 (55). Mts1 significantly inhibited the p53-dependent transcription from the p21 waf1 promoter as measured in transient reporter assays, whereas its effect on the Bax promoter was negligible. Similarly, WT1 exhibited an inhibitory effect on the p53-regulated activation of the MDM2 promoter, whereas the repression of the Bax promoter by WT1 was not significant (56). Recently, Thornborrow and Manfredi (57) have reported that there exists an Sp1-binding site immediately adjacent to the p53-responsive element of the Bax promoter, and p53 may require the cooperation of Sp1 to activate the Bax promoter. In contrast, the p53⅐Sp1 complex may function in an antagonistic manner for other promoters (58). These findings suggest that MM1-mediated differential regulation of the p53/p73-responsive promoters is due to a requirement for additional cofactors specific to each promoter, although this requires further elucidation.
Recently, Ceballos et al. (45) have found that c-Myc significantly inhibits the transactivation and pro-apoptotic function of p53. Consistent with their report, our data have shown that co-expression of p53 with c-Myc results in a remarkable reduction of the p53-mediated transcriptional activation of the Bax and p21 waf1 promoter. Using the immunoprecipitation and the luciferase reporter assays, we have demonstrated that c-Myc physically interacts with p73␣ and thereby inhibits its transactivation function. In contrast, N-Myc, the other member of myc family, is unable to affect the p73␣-or p53-dependent transcriptional activation. As reported previously, targeted homozygous disruption of the c-myc or N-myc gene resulted in embryonic lethality (59 -61). When c-myc was replaced by Nmyc, mice did not show any evident defects (62), suggesting that c-Myc and N-Myc are functionally redundant. However, c-myc and N-myc have shown a distinct expression pattern in the tissues or organs during embryogenesis. Therefore, our present results may suggest that there is tissue-specific signaling or regulation in the p73 function, to which c-Myc and N-Myc are differently concerned.
MM1 has been originally identified as a c-Myc-binding protein and blocked the E-box-dependent transcriptional activation of c-Myc (41). Recently, Satou et al. (28) found that MM1 recruits an HDAC complex and thereby inhibits the c-Mycmediated transactivation. According to our present results, FIG. 9. Etoposide-induced apoptosis and overexpression of p73 in HL60 cells. A, HL60 cells (5 ϫ 10 5 cells/tissue culture plate) were treated with etoposide (68 M) and incubated for up to 8 h. Cells were collected at each of the indicated time periods. After fixing and staining with propidium iodide, the DNA content of the cells was analyzed by flow cytometry. B, at the indicated time periods after the treatment of etoposide, total RNA was extracted by using the RNeasy Mini kit, and the expression levels of p73, c-myc, and MM1 were examined by the RT-PCR analysis. GAPDH mRNA expression levels were measured as an internal control. C, whole cell lysates were prepared at the indicated time periods after the treatment with etoposide, and the expression levels of p73␣, c-Myc, and MM1 were analyzed by Western blotting using the monoclonal anti-p73, the monoclonal antic-Myc, and the polyclonal anti-MM1 antibodies, respectively. The MM1 blot was reprobed for actin to ensure equal protein loading (bottom). MM1 attenuated the c-Myc-dependent down-regulation of p73␣, and the amount of p73␣ co-precipitated with c-Myc was decreased in the presence of MM1. It is unclear at this time whether HDAC-mediated repression of the transcriptional activity of c-Myc is required for MM1 to abolish the ability of c-Myc to inhibit p73␣.
Intriguingly, when HL60 cells (which do not contain p53 protein) were exposed to etoposide, the expression level of p73␣ protein clearly increased 4 h after the treatment and continued to increase at least until 8 h, being strongly associated with increase in the number of cells in sub-G 1 . On the contrary, the levels of c-Myc expression decreased 4 h after the treatment with etoposide. Given that c-Myc inhibits the transcriptional activity of p73␣, it is possible that MM1 interferes with the ability of c-Myc to inhibit the pro-apoptotic function of p73. It is necessary to be clarified whether the c-Myc-binding site in p73␣ is localized close to the MM1-binding site within the extreme COOH-terminal region. However, MM1 may not only bind to p73␣ but also inhibit the interaction between c-Myc and p73␣ and thereby allow p73␣ to change to a conformation that is competent to express its activity.