MM-1, a Novel c-Myc-associating Protein That Represses Transcriptional Activity of c-Myc*

We have isolated the cDNA encoding a novel c-Myc-binding protein, MM-1, by the yeast two-hybrid screening of a human HeLa cell cDNA library. The protein deduced from the cDNA comprises 167 amino acids and was localized in the nucleus of introduced COS-I cells. The MM-1 mRNA was highly expressed in human pancreas and skeletal muscle and moderately in other tissues. As for the c-Myc binding, glutathione S-transferase MM-1 expressed in Escherichia coli bound in vitro to c-Myc translated in reticulocyte lysate, and almost whole, the MM-1 molecule was necessary for the binding in the yeast two-hybrid system. The mammalian two-hybrid assays in hamster CHO cells revealed that MM-1 interacts in vivo with the N-terminal domain covering themyc box 2, a transcription-activating domain, of c-Myc. Furthermore, MM-1 repressed the activation of E-box-dependent transcription by c-Myc.

c-Myc, a proto-oncogene product, is a key factor regulating cell growth, transformation, and apoptosis induction (for reviews, see Refs. [1][2][3][4]. c-Myc recognizes the E-box sequence together with Max as a partner (5)(6)(7) and activates the transcription due to the sequence. Several genes containing the E-box sequence were identified as candidates to be regulated by c-Myc; the genes for ␣-prothymosin (8,9), ECA39 (10, 11), eIF4E (12,13), eIF-2␣ (12), ornithine decarboxylase (14), carbamoyl-phosphate synthase (15), and cdc25A (16) are included. c-Myc comprises several domains necessary for functions: the N-terminal domain covering the amino acid sequences called myc boxes, conserved among various myc gene families, and the C-terminal domain containing basic-helix-loop-helix-leucine zipper structure are required for transcriptional activity. Various factors have been reported to bind to each domain of c-Myc and, thus, to modulate the c-Myc functions either positively or negatively. Max (5)(6)(7), Nmi (17), and YY-1 (18) bind to the C-terminal domain of c-Myc, whereas p107 (19 -21), Bin1 (22), TBP 1 (23,24), and ␣-tubulin (25) bind to the N-terminal do-main of c-Myc. Max is a partner protein with c-Myc to bind to the E-box sequence for transcription activation, whereas p107 or Bin1 inhibits transcription or transformation activity of c-Myc (5-7, 19 -21, 22). Functions of other proteins associated with c-Myc have not yet been well characterized. Although the associating factors and the candidate genes for c-Myc targets have been identified as above, precise mechanisms of the regulation of versatile c-Myc functions as well as signal transduction pathways to c-myc remain unclear. Further clarification of the mechanisms necessarily involves cloning and characterization of yet unknown proteins around c-Myc. We describe here a novel protein termed MM-1, which associates with c-Myc and represses the transcription activity of c-Myc.

Cells
Human HeLa, monkey CV-1, monkey COS-I and hamster CHO cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% calf serum.

Cloning of MM-1 by Two-hybrid System
Saccharomyes cerevisiae HF7C cells containing the lacZ gene driven by the GAL1 promoter were transformed first with pGBT-c-Myc. The transformant yeast cells were subsequently transformed with Human HeLa MATCHMAKER cDNA (CLONTECH), a cDNA library expressing the GAL4 activation domain fused to cDNAs from human HeLa cells. About 2 ϫ 10 5 colonies were screened for ␤-galactosidase expression, and the plasmid DNAs in the LacZ-positive cells were extracted by the procedure described in the protocols from CLONTECH. A plasmid thus obtained was named pGAD-#9 and analyzed. Because the EcoRI-XhoI fragment insert of cDNA in pGAD-#9 did not contain the ATG initiation codon, the gt11 human placenta cDNA library (CLON-TECH) was further screened with the labeled insert of pGAD-#9 as a probe. The longest cDNA obtained was inserted to the EcoRI site of pBluescript SK(Ϫ), and the construct was named pBS-MM-1.

In Vitro Binding Assay
GST-MM-1 and GST were purified from a 1,000-ml culture of Escherichia coli BL21(DE3) transformed with pGEX-MM-1 and pGEX-6P-1, respectively, as described previously (26). Two g of the purified GST-MM-1 or GST were first applied to GST-Sepharose 4B (Amersham Pharmacia Biotech) in a buffer containing 20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.5% Nonidet P-40, and 1 mM EDTA. The 35 S-labeled c-Myc or its deletion mutants synthesized in vitro using pGEM-cmyc(ATG) or pcDNA3-MYC deletions, respectively, as a template in the coupled transcription-translation system (Promega) were then applied to the column. After extensive washing of the column with the same buffer as above, the proteins bound to the resin were recovered, separated in a 7.5% polyacrylamide gel containing SDS, and visualized by fluorography.

Plasmids
Nucleotide sequences of oligonucleotides used were as follows: cmyc(GKPG)ATG, 5Ј-GGAATTCGGTAAACCGGGAATGCCCTCAACG-TTAGCTTC-3Ј; c-myc(end), 5Ј-GGAATTCAGATGGTAAGCAT-3Ј; MM-1Bam(ATG), 5Ј-GGGGATCCTGATGGTCCATCCAA-3Ј; RVX, 5Ј-GAG-CGGATAACAATTTCACACACAGG-3Ј; FLAG-A, 5Ј-CATGGACTACA-* This work was supported by a Grant-in-aid from the Ministry of Education, Science, Culture, and Sports in 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D89667.
¶ To whom correspondence should be addressed: Faculty of Pharmaceutical Sciences, Hokkaido University, Kita 12 Nishi 6, Kita-ku, Sapporo 060, Japan. Tel. The c-Myc cDNA was prepared by PCR using pGEM-c-myc(ATG) as a template and c-myc(GKPG)ATG and c-myc(end) as primers. The PCR contained 10 ng of pGEM-c-myc(ATG), 10 pmol each of the primers, and 2.5 units of Taq polymerase (Boehlinger Mannheim). The reactions were carried out for first 5 min at 94°C, then for 30 cycles of 1 min at 94°C followed by 2 min at 55°C, and finally for 3 min at 72°C. The amplified fragment was digested with EcoRI and cloned to the EcoRI site of pGBT9. All the PCR in this study used the same conditions as above.
pGEX-MM-1-The BamHI-EcoRI fragment of the polymerase chain reaction product with MM-1(ATG) and RVX primers on pBS-MM-1 template was inserted to the respective site of pGEX-5X-1 (Amersham Pharmacia Biotech).
pSV-FLAG-The 5Ј-ends of the oligonucleotides FALG-A and FLAG-B were phosphorylated by T4 polynucleotide kinase. The two oligonucleotides were annealed and inserted to pSV2␤. The multiclon- ing sites from SpeI to XhoI in pGBT9 were then inserted to the respective sites of the construct above.
pGEX-6P-MM-1-The EcoRI fragment of the polymerase chain reaction product using MM-1(EcoTTC) and RVX primers on the pBS-MM-1 template was inserted to the respective site of pGEX-6P-1 (Amersham Pharmacia Biotech).

Indirect Immunofluorescence
COS-I cells were transfected with pFLAG-MM-1 or pEF-c-myc (26) by the calcium phosphate precipitation technique (27). Forty-eight h after transfection, the cells were fixed with the solution containing acetone-methanol (3:7) and reacted with a mouse anti-FLAG monoclonal antibody (M2, Eastman Kodak Co.) or an anti-c-Myc monoclonal antibody (C-33, SantaCruz). The cells were then reacted with an fluorescein isothiocyanate-conjugated anti-rabbit IgG and observed with a fluorescence microscope.

Luciferase Assay
CV-1 cells subconfluent in a 6-cm dish were transfected with various amounts of effector plasmids expressing the wild type or mutants of MM-1 and c-Myc in addition to 1 g of pCMV-␤-gal and 1 g of p4xE-SVP-Luc by the calcium phosphate precipitation method. Forty-eight h after transfection, cell extracts were prepared by adding a solution containing 25 mM Tris-phosphate (pH 7.8), 2 mM dithiothreitol, 10% glycerol, 1% Triton X-100, and 2 mM 1,2-diaminocyclohexane-N,N,NЈ,NЈ-tetraacetic acid (Promega). After standardization of the transfection efficiencies by ␤-galactosidase assays, the luciferase activity was examined in the luciferase assay mixture supplied by Promega.

Mammalian Two-hybrid Assay
CHO cells were transfected with 2 g of effector plasmids expressing the wild type or mutants of MM-1 and c-Myc fused to GALBD or VP16 in addition to 1 g of pCMV-␤-gal and 1 g of pH17-MX-Luc-Luc and followed by luciferase assay as above.

RESULTS AND DISCUSSION
Cloning of a cDNA Encoding the c-Myc Binding Protein-To screen cDNAs encoding c-Myc-associating proteins, the entire coding sequences of the c-myc cDNA were fused to the GALDB and introduced to S. cerevisiae HF7C cells. A human cDNA library prepared from HeLa cells and cloned in pGADGH was then introduced to the transformant yeast cells, and the colonies resistant to His marker were selected. Among a total of 2 ϫ 10 5 transformant cells, 17 colonies were His-positive, and 15 of the 17 His-positive colonies yielded a high expression of ␤-galactosidase activity. Restriction enzyme analyses revealed that the cDNA carried in the 15 colonies were of 2 classes. The cDNA of one class, named No. 9, did not contain the initiation codon ATG, and another cDNA library prepared from human placenta and cloned in gt11 was hence screened by hybridization with a labeled No. 9 cDNA. Two cDNA clones thus obtained contained the same coding sequence, and the protein coded by the longer cDNA was termed MM-1 (Myc Modulator -1) and further characterized.
The MM-1 cDNA comprised 1,029 nucleotides with a stop codon in-frame at the nucleotide 285 upstream from the first ATG at 418 (Fig. 1A). It was therefore determined that MM-1 cDNA encodes 167 amino acids starting from the nucleotide 418. There exists a putative leucine zipper structure from amino acid 28 to 49. It is not clear at present whether MM-1 makes complexes with other proteins via the leucine zipper, but no homodimer of MM-1 was observed with the purified recombinant MM-1 (data not shown). The EST data base search revealed several cDNA clones homologous to the MM-1 cDNA including accession numbers W25318, AA287397, and H03166. These clones were obtained via Genome Systems Inc., and the nucleotide sequences of the inserts were determined (Fig. 1B). The clone W25318 contained 10 original nucleotides followed by the same 589 nucleotides as in the MM-1 cDNA from 441 to 1029, and the sequence at the junction between the W25318 original 10 nucleotides and the following nucleotides common to W25318 and MM-1 was TA 3 AG, which is a consensus sequence of donor/acceptor sites for splicing. Since a stop codon, TAA, thus appeared at the junction, the ATG at the nucleotide 457 referred to the MM-1 cDNA is a putative initiation site for the clone W25318. Both the clones AA287397 and H03166 are probably derived from different splicing of the same gene as MM-1 (Fig. 1B). These two clones did not contain stop codons in-frame upstream from the ATG at the nucleotide 457, and the clones hence may not contain the complete coding sequences.
The expression of MM-1 was examined in various human tissues by Northern bolt analyses (Fig. 2). At least four distinct bands were detected corresponding to the sizes of 0.7, 1.15, 2.9, and 4.4 kb. The 0.7-kb mRNA was vigorously expressed in all the tissues examined. The 1.15-kb mRNA corresponding to the size of the MM-1 cDNA cloned here was strongly expressed in pancreas, weakly in kidney, skeletal muscle, and placenta and faintly in liver and lung (Fig. 2). To verify the cellular localization of MM-1, an expression vector for MM-1 fused to FLAG was transfected to monkey COS-I cells. Two days after transfection, the cells were stained with an anti-FLAG antibody. An expression vector for c-Myc was similarly transfected to COS-I cells as a control. FLAG-MM-1 was mainly observed in nuclei except for nucleoli and a little in cytoplasm (Fig. 3). The c-Myc introduced was confirmed to be located in nuclei as reported (data not shown).
Association of MM-1 with c-Myc-To examine the interaction between MM-1 and c-Myc, GST-MM-1 or GST was expressed in E. coli and applied to a glutachione-Sepharose column. The 35 S-labeled c-Myc synthesized in vitro in the coupled transcription-translation system (Promega) was then applied to the column, and the labeled protein bound to the GST-MM-1 or GST on the column was separated in a gel and visualized (Fig.  4A, lanes 1-3). The result showed that the labeled c-Myc bound to GST-MM-1 but not to GST, and the direct binding between MM-1 and c-Myc was thus suggested. To assess the binding domain of MM-1 with c-Myc, the wild type and deletion mutants of MM-1 were fused to the GALAD, the wild type c-Myc were fused to the GALBD, and the interaction of the fusion proteins was assayed in yeast. All the deletion mutants of MM-1 except for ⌬13, deleting 13 amino acids from the N terminus, entirely lost or hardly retained the binding activity to c-Myc (Fig. 4B). The binding activity of MM-1 to c-Myc was hence suggested to require nearly the whole molecule of MM-1, probably because the appropriate stereostructures of the rather small molecule is important for the interaction. To determine the binding domain on the c-Myc side, similar assays applying the yeast two-hybrid system were first attempted. The background GAL4 activity, however, varied among the wild type and deletion mutants of c-Myc when the proteins were fused to GALBD, and the assays did not work when the c-Myc variants were fused to GALAD and used as a bait. We hence examined the c-Myc domains for interaction with MM-1 by in vitro binding assays. The deletion mutants of c-Myc as well as the wild type protein were synthesized and used for in vitro binding assay. The mutant protein lacking Myc-box II (⌬MII) or the amino acids 1 to 147 lost the binding activity, whereas the mutant deleted of Myc-box I (⌬MI) or leucine zipper (⌬ZIP) retained the binding activity (Fig. 4A, lanes  4 -15). The binding activity of c-Myc to MM-1 was further tested by the mammalian two-hybrid assays, which were successfully applied to see the interaction of c-Myc with p107, RB, or TBP (21,28). c-Myc and its deletion mutants were fused to the VP16 activation domain (VP16AD), whereas MM-1 was fused to GALBD. Hamster CHO cells were transfected with the expression vectors for MM-1-GALBD and c-Myc or its mutants and fused to VP16AD in addition to pHE17-MX-Luc, which contains 6xGAL4, and the HTLV-1 promoter was linked to the luciferase gene. First, for testing the effect of the transactiva- tion domain of c-Myc in this system, the cells were transfected with wild type, or the segment N-containing transactivation domain of c-Myc was fused to VP16AD together with pHE17-MX-Luc. These Myc constructs gave no luciferase activities in the absence of MM-1-GALBD because of the lack of the DNA binding domain to GAL4 (data not shown). In the cells introduced with GALBD-MM-1 and c-Myc-VP16AD (the wild type c-Myc fused to VP16AD), the luciferase activity was seven times as high as that in the cells with GALBD-MM-1 and VP16AD, indicating that MM-1 interacted with c-Myc in vivo (Fig. 4C). Among the deletion mutants of c-Myc, the activity was lost for ⌬S, ⌬MII, ⌬6, ⌬147, and ⌬177, whereas ⌬MI, ⌬ZIP, N, ⌬43, and ⌬103 still retained the activity (Fig. 4C). Deletion of Zip to ⌬MI, ⌬S, and ⌬MII had little effect (Fig. 4C, ⌬MI⌬Zip,  ⌬S⌬Zip, and ⌬MII⌬Zip, respectively). The results indicate that MM-1 binds in vivo to the domain around myc box II from the amino acid 104 to 166 but not to the basic-helix-loop-helixleucine zipper of c-Myc in CHO cells. Similar results were obtained when the same plasmids were transfected to human HeLa cells (data not shown).
Transcriptional monkey CV-1 cells were transfected with the expression vectors for c-Myc and MM-1 together with the luciferase gene linked to the tetramerized E-box sequence (4xE) followed by the SV40 promoter. As previously reported, c-Myc stimulated the E-box-dependent luciferase activity in a dose-dependent manner (Fig. 5A, wt). Neither a Zip-deleting mutant (⌬Zip) nor another deleting the transactivation domain of c-Myc (⌬MII) nor vector (SR␣) showed the luciferase activity (Fig. 5A). When CV-1 cells were co-transfected with various amounts of pCMV-F-MM-1, an expression vector of MM-1, or of pCMV-F vector in addition to 0.5 g of pSR␣-c-myc, an expression vector of c-myc, or of pSR␣ vector, the luciferase activity enhanced by c-Myc (in the presence of pSR␣-c-myc) was repressed by co-introduced pCMV-F-MM-1 in a dose-dependent manner, whereas the basal activity in the presence of pSR␣-c-myc was not affected by pCMV-F-MM-1 (Fig. 5B). The vector pCMV-F alone showed no effect on either the activity enhanced by pSR␣-c-myc or the basal activity in the presence of pSR␣. The expression vector of MM-1⌬N, a mutant protein lacking the c-Myc binding activity, also lost the repression activity on c-Myc (Fig. 5B). The results suggest that MM-1 modulates the transcriptional activity of c-Myc via binding to c-Myc.
We have here described MM-1, a new partner protein of c-Myc. MM-1 is a nuclear protein like c-Myc and binds to the N-terminal domain of c-Myc covering myc box II and possessing transcriptional activation activity. Binding between MM-1 and c-Myc was observed both in vitro and in vivo in yeast and transfected mammalian cells. The search through the EST data base implied the presence of several splicing forms of the transcript from the MM-1 gene. In the Northern blot analyses, at least 4 distinct bands were detected, and the 0.7-kb mRNA was ubiquitously expressed in various human tissues. The 1.15-kb mRNA corresponding to the cDNA we obtained here was also expressed in several tissues but, except in pancreas, to much weaker levels than the 0.7-kb mRNA. The 0.7-kb mRNA may correspond to the clone W25318 (see Fig. 1B), which lacks only 13 amino acids from the N terminus of the MM-1 cDNA. We hence constructed a ⌬13 by deleting the N-terminal 13 amino acids from the wild type MM-1 and tested the mutant for biological functions. No difference was observed between ⌬13 and the wild type MM-1 in terms of the activities of c-Myc binding and transcriptional repression (data not shown).
MM-1 was thus suggested to repress the transactivation activity of c-Myc through association with c-Myc. To verify this further, c-Myc mutants lacking the MM-1 binding activity but still sustaining transactivation activity were desirable. Such mutants, however, could not be constructed, because the regions of c-Myc for both MM-1 binding and transactivation activities are overlapped. The c-Myc mutants ⌬S and ⌬MII, which lost the MM-1 binding activity (Fig. 4), hardly activated the E-box-dependent transcription (Fig. 5A). The result with an MM-1 mutant lacking the c-Myc binding activity (Fig. 5B) supported the suggestion above. The lack of c-Myc binding activity resulted in the loss of the repression activity on the transcriptional activity of c-Myc.
The region around myc box II of c-Myc associating with MM-1 is known to be essential for c-Myc to sustain the activities of transcription, cell transformation, and apoptosis induction (1)(2)(3)(4). The region has been also reported to be bound by p107, Bin1, TBP, or ␣-tubulin (19 -25) to modulate the functions of c-Myc. Among the binding proteins, p107 and TBP bind to the epitope around myc box II, the amino acids from 100 to 148 (21) overlapping the MM-1 binding domain, the amino acids from 104 to 166. p107 (19 -21) as well as Bin1 (22) is classified as a tumor suppresser protein and represses the transcription or transforming activity of c-Myc as MM-1 does. Two mechanisms have been considered to explain the transcriptional inhibitory activity of p107 to c-Myc. One is that p107 recruits cyclinA/Cdk2 to phosphorylate Ser-62 in c-Myc, and Thr-58 is subsequently phosphorylated to lead the downmodulation of the c-Myc transcriptional activity (21). This was based on the fact that the c-Myc transcriptional activity was not inhibited by p107 in several Burkitt lymphoma cells in which mutations were frequently found at the Ser-62 and Thr-58 of c-Myc (19,21). The conflicting observation, on the other hand, was reported that p107 similarly inhibited the transcriptional activity of both the wild type and mutated c-Myc in Burkitt lymphoma cells (29), proposing another mechanism that p107 competitively interferes with the binding of positive regulatory factors including TBP. Since MM-1 shares both the binding domain and the transrepression activity on c-Myc with p107, similar mechanisms may exist for the negative regulation of c-Myc transcriptional activity by MM-1. Further examination of the interaction among c-Myc, MM-1, TBP, p107, and possibly other associating proteins yet unidentified is required to assess such possibilities.  Fig. 4C) of c-Myc, or vector (SR␣) together with p4xE-SVP-Luc as a reporter plasmid. Two days after transfection, the luciferase assay was carried out. Relative luciferase activities to that of p4xE-SVP-Luc alone are shown. B, CV-1 cells were transfected with an expression vector for MM-1 (pCMV-F-MM1), MM-1⌬N (pCMV-F-⌬N), or vector alone (pCMV-F) together with p4xE-SVP-Luc as a reporter plasmid in addition to 0.5 g of pSR␣-c-myc or pSR␣. Two days after transfection, the luciferase assay was carried out. Relative luciferase activities to that of the cells cotransfected with the reporter plasmid and 0.5 g of pSR␣ are shown.
After the submission of this manuscript, two other groups reported that MM-1 may be a human homologue of GIM5 or PFD5, a subunit of the complex promoting the formation of functional actin or tubulin (30,31). Relationships among c-Myc, MM-1, and the complex have not been clarified.