The proto-oncogene c-Myc plays a central role in cell growth and the development of human tumors. c-Myc interacts with Max and Myc-Max complexes bind to E-box and related sequences to activate transcription. Max also interacts with Mnt but Mnt-Max complexes repress transcription when bound to these sequences. MNT maps to human chromosome 17p13.3, a region frequently deleted in various human tumors, including mammary gland tumors. Consistent with the possibility that Mnt functions as a Myc antagonist, Mnt-deficient fibroblasts exhibit many of the hallmark characteristics of cells that overexpress Myc, and conditional (Cre/Lox) inactivation of Mnt in mammary gland epithelium leads to adenocarcinomas. Here, we further characterize mammary gland tissue following conditional deletion of Mnt in the mammary gland. We show that loss of Mnt severely disrupts mammary gland involution and leads to hyperplastic ducts associated with reduced numbers of apoptotic cells. These findings suggest that loss of Mnt in mammary tissue has similarities to Myc overexpression. We tested this directly by using promoter array analysis and mRNA expression analysis by oligonucleotide arrays. We found that Mnt and c-Myc bound to similar promoters in tumors from MMTV-c-Myc transgenic mice, and mRNA expression patterns were similar between mammary tumors from MMTV-Cre/MntKO/CKO and MMTV-c-Myc transgenic mice. These results reveal an important role for Mnt in pregnancy-associated mammary gland development and suggest that mammary gland tumorigenesis in the absence of Mnt is analogous to that caused by Myc deregulation. (Cancer Res 2006; 66(11): 5565-73)

Myc family proteins are nuclear phosphoproteins that play a pivotal role in many cellular events, including the control of cell growth, proliferation, differentiation, and embryonic development (reviewed in ref. 1). Moreover, it is well known that the dysregulation of c-Myc is associated with tumorigenesis in a variety of contexts (reviewed in ref. 2). In many tumors and tumor-derived cell lines, myc genes are altered by gene amplification, chromosome translocations, and viral insertion, events that typically lead to deregulation of Myc with increased Myc expression (reviewed in ref. 2). The relevance of these alterations to tumorigenesis is suggested by the ability of forced overexpression of Myc in vivo to cause tumors in a wide variety of tissues, including mammary gland epithelium (35).

The Myc family of proteins (c-Myc, N-Myc, and L-Myc) require heterodimerization with Max through related bHLHZip motifs to bind DNA and function as transcription factors (1). Max also interacts with the four Mxd (formerly Mad) family members, Mad1, Mxi1, Mad3, and Mad4, as well as the Mad-related proteins Mnt and Mga, to form heterodimers via bHLHZip domains (reviewed in ref. 1). As with Myc, heterodimerization with Max allows these proteins to bind the E-box consensus sequence, CACGTG, and related sequences. The Mad family, Mnt and Mga, are transcriptional repressors that can compete with Myc for binding to Max, and it is hypothesized that all of these complexes can compete for binding to shared target sites in vivo (1). Recent experiments showing that Mnt-Max complexes decline when c-Myc is induced during cell cycle entry also suggest that Mnt and c-Myc compete for binding for a limiting supply of Max (6). Thus, activating Myc-Max complexes together with repressive Mnt-Max, Mad-Max, and Mga-Max complexes seem to function as an antagonistic network through the binding and transcriptional regulation of shared target genes.

The MNT gene maps to human 17p13.3, a region that is commonly deleted in a variety of tumors, including mammary gland tumors (710). Although no mutations in MNT have been described in human tumors to date (1113), previous studies have shown that frequent loss of heterozygosity (LOH) in 17p13.33, but also in 17p25.1, 8p22, 13q12, and 22q13, correlates with postoperative recurrence in breast cancers (14). LOH involves a region telomeric to p53 at 17p13.1 (14) and clearly indicates the presence of another tumor suppressor gene(s) located in 17p13.3. A candidate tumor suppressor gene in this region is HIC1 (hypermethylated in cancer), which is aberrantly hypermethylated and transcriptionally inactivated in several types of human cancers (1517). Heterozygous Hic1 mutant mice have no developmental phenotype but are predisposed to malignant tumors, including pulmonary carcinomas, lymphomas, and sarcomas (18). However, these mice do not display mammary gland cancer, suggesting that candidate tumor suppressor genes within 17p13.3 responsible for other malignant tumors, such as breast tumors, remain to be identified. A recent study suggests that MNT is a potential tumor suppressor gene in medulloblastoma (19). Chromosome 17p13.3 is also well defined as the Miller-Dieker syndrome critical region. Patients with Miller-Dieker syndrome display severe neuronal migration defects defined by lissencephaly (smooth brain) as well as dysmorphic features consistent with abnormalities in craniofacial development (reviewed in ref. 20).

We recently showed that loss of Mnt causes a hyperproliferation phenotype in mouse embryonic fibroblasts, including accelerated proliferation, increased sensitivity to apoptosis, sensitivity to transformation by H-RasV12 alone, and senescence bypass (ref. 21; also reviewed in ref. 22). In addition, conditional deletion of Mnt in mouse mammary gland epithelium led to the development of mammary adenocarcinomas after long latency. Mnt-deficient fibroblasts and mammary gland epithelium displayed decreased c-Myc, but increased cyclin E and cyclin-dependent kinase (Cdk)-4, consistent with a phenotype resulting from increased relative c-Myc activity. Here we extend these findings to show that loss of Mnt in mammary gland epithelium causes defects in mammary gland development and severely disrupts involution. We also used genome-wide location analysis (GWLA) to examine promoter binding of c-Myc and Mnt in tumors arising from overexpression of c-Myc, and oligonucleotide expression arrays to compare mRNA expression patterns between tumors resulting from loss of Mnt or overexpression of c-Myc. Our results indicate an important role for Mnt in mammary gland development and suggest that tumorigenesis caused by loss of Mnt and overexpression of c-Myc are functionally related.

Generation of Mnt mutant mice. The production of mice with a Mnt conditional knockout (MntCKO) allele by gene targeting is described elsewhere (23). Briefly, a loxP-flanked hygro resistance gene and third loxP were inserted into introns 3 and 6, respectively, of the Mnt gene by gene targeting. EIIa-Cre transgenic mice in an FVB/n background (24) were mated with inbred 129/S6 Mnt+/CKO mice to generate the MntKO allele in a mixed background (129/S6 × FVB/n) mice in vivo. To produce mice with loss of Mnt in the mammary gland (MMTV-Cre+/MntKO/CKO mice), MMTV-Cre mice were mated with Mnt+/KO mice to produce MMTV-Cre+/MntKO/+ mice. MMTV-Cre+/MntKO/CKO mice were produced by breeding MMTV-Cre+/MntKO/+ mice with MntCKO/CKO mice. We used MMTV-Cre/MntKO/CKO mice without MMTV-Cre as control. MMTV-Cre/MntKO/CKO mice displayed no mammary gland tumors.

Genotyping analysis. Adult mice were genotyped by PCR analysis using tail DNA. The following primers were used to detect wild-type, conditional, or deleted allele. MntWT-sense (5′-cagtccctgaagaggaagga-3′) and MntWT-Rev2 (5′-ccggagcacacgatctatct-3′) were employed for wild-type allele detection. MntKO3 (5′-caggtcctccaaaagagcag-3′) and MntKO4 (5′-ggagcaatgtggagagaagc-3′) were used for deleted allele detection. The conditions of amplification were 35 cycles at 94°C for 30 seconds, 68°C for 60 seconds, and 72°C for 120 seconds. This PCR reaction generates a 386-bp fragment from mutant allele and a 147-bp from wild-type allele. To distinguish wild-type and conditional mutant alleles, two primers (sense, 5′-cagattcagtgtcccctgct-3′; antisense, 5′-gtctcaagtcgtgggcattg-3′) were used to amplify a 178-bp product from the wild-type allele and a 1.9-kbp product from the conditional allele. The results from this PCR were confirmed by detection of the hygro resistance gene with primers for hygro (sense, 5′-gatgtaggagggcgtggata-3′; antisense, 5′-gatgttggcgacctcgtatt-3′). The MMTV-Cre allele was detected with the primers CREA1 (5′-ccgggctgccacgaccaa-3′) and CREA2 (5′-ggcgcggcaacaccattttt-3′).

Whole-mount analysis of mammary gland. Mammary gland tissues were dissected at various developmental stages from the two genotypes (MMTV-Cre+/MntKO/CKO and MMTV-Cre/MntKO/CKO), spread on glass slides, and fixed immediately with Carnoy's fixative (6 volumes ethanol; 4 volumes chloroform; 1 volume acetic acid) for 5 hours at room temperature. After 3 days of acetone treatment to remove lipid, tissues were hydrated with serially diluted ethanol. Hydrated tissues were stained with 0.6% carmine solution for 1 hour at room temperature, followed by treatment with differentiating solution (36% ethanol, 18% methanol). Dehydrated tissues with ethanol were cleared with benzyl alcohol/benzyl benzoate (1:2) solution.

Histologic analysis. Mammary gland tissues (n = 4-5 for each genotype) were dissected at various stages and fixed in 4% paraformaldehyde/PBS solution for 15 hours at 4°C. Fixed tissues were embedded in paraffin, sectioned at 5 μm, and stained with H&E.

Terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling analysis in the developing mammary gland. Mammary gland tissues fixed with 4% paraformaldehyde/PBS were embedded in paraffin, sectioned at 5 μm, and deparaffinized. Apoptotic cells were detected by using the ApopTag kit (Intergen Company, Purchase, NY) according to the instructions of the manufacturer. For quantification analysis, terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL)–positive cells in 100 × 150-μm area were counted in both MMTV-Cre+/MntKO/CKO mice and MMTV-Cre/MntKO/CKO mice.

Western blot analysis. Tumors and adjacent normal tissue from MMTV-Cre+/MntKO/CKO or from wild-type C57/Bl6 mice were dissected and lysates were prepared as previously described for Western blotting (25). Antibodies against Mnt and c-myc were used as previously described (26).

Kaplan-Meier analysis. To analyze mammary gland tumor development of MMTV-Cre+/MntKO/CKO mice, we longitudinally followed 10 MMTV-Cre+/MntKO/CKO mice and 9 MMTV-Cre/MntKO/CKO mice as controls. All mice were checked weekly for mammary gland tumor development by palpation and mice were sacrificed when they displayed a 1-cm mammary gland tumor. The time from birth was plotted on the X axis and mammary gland tumor-free survival (as percentage of total number of mice) was plotted on the Y axis. The censored subjects are shown as solid squares or triangles in the horizontal part of the staircase and include all deaths not associated with mammary gland tumors.

Statistical analysis. H&E-stained sections (five to seven sections per mouse) of mammary gland for each developmental stage were used to determine the ductal area by using Microanalyzer software (Nihon Poladigital, Tokyo, Japan). The percentage of enlarged ducts was calculated on the basis of the measurement of ductal area. “Enlarged duct” was defined as the ducts in which the area was larger than the area of wild-type (the average area of wild-type was 1,500 μm2). Statistical significance was determined by two-sided Student's t test.

Preparation of chromatin for microarray analysis. A DNA microarray containing PCR products spanning the proximal promoters of 1,920 mouse genes (M2K, Supplementary Table S1) was chosen from the NCBI Refseq database. These genes were selected for good gene annotation and a clear functional association. Tumors were harvested and cross-linked in formaldehyde [11% formaldehyde/0.1 mol/L NaCl/1 mmol/L Na-EDTA/0.5 mmol/L Na-EGTA/50 mmol/L HEPES (pH 8.0)] and incubated on ice for 10 minutes. The cross-linking reaction was stopped with a 1/20 volume of 2.5 mol/L glycine solution. Following cross-linking, chromatin was extracted in lysis buffer [0.2 mol/L NaCl/1 μmol/L EDTA/0.5 μmol/L EGTA/10 μmol/L Tris (pH 8)/protease inhibitor cocktail] for 10 minutes, followed by centrifugation at 2,000 × g for 10 minutes at 4°C.

Chromatin was fragmented using sonication on ice and immunoprecipitated using magnetic beads coupled to primary antibodies to c-Myc and Mnt as described (27). The beads were collected and the supernatant was removed. The chromatin was eluted from the beads using elution buffer [50 mmol/L Tris (pH 8)/10 mmol/L EDTA/1% SDS]. The cross-link was reversed and DNA was purified and prepared for labeling and microarray hybridization through random primer amplification. The samples were then hybridized to the M2K Microarray slides and analyzed using previously described protocols (28). The antibodies to c-Myc (28) and Mnt (6) have been validated previously for chromatin immunoprecipitation. To validate the array data, genomic sequence primers encompassing Ibd1, Sap18, Cdkn1b, Rab18, and Brca1 promoter regions that bound Mnt and c-myc were used to amplify immunoprecipitated DNA. The following primers, corresponding to promoter regions of each gene, were used for the chromatin immunoprecipitation analysis: Ibd1 sense 5′-CACCTGGGGTCAGAACATCT-3′, antisense 5′-CTGTTGGAAGATTGGGCTGT-3′; Sap18 sense 5′-CCTTGCCACTCGTTTCCTAT-3′, antisense 5′-CCGGCTCCTTCTTAATTTCC-3′; Cdkn1b sense 5′-CTAGCCACCGAAGCTCCTAA-3′, antisense 5′-GCCCCGGAGAAATTAAAAAT-3′; Rab18 sense 5′-TGATGTGGGGAGACACTTGA-3′, antisense 5′-GGCAGAGCTGCTGTAAATCC-3′; and Brca1 sense 5′-TCCGGTGTGAGAAGCTCTTT-3′, antisense 5′-AGCAGTTGCAAAACGAGGTC-3′. The following primers, corresponding to intragenic regions of each gene, were used as negative control: Ibd1 sense 5′-GCCTCCAAACTTTTGCTGAG-3′, antisense 5′-CTGCAGAGGCTAGGATACCG-3′; Sap18 sense 5′-TGAGAGGGGCTAAGAGGTGA-3′, antisense 5′-GTCCTCAAAATCAGGGCAAA-3′; Cdkn1b sense 5′-TGTCCTCCTAATGGCAAAGG-3′, antisense 5′-TGGCAGTGGGAATGTAGTCA-3′; Rab18 sense 5′-GCAATTGTCTGGCTGTGAGA-3′, antisense 5′-AAAGGCCTGAAGGGGTTAGA-3′; and Brca1 sense 5′-CAGGTCAGCTACCACAGCAA-3′, antisense 5′-TTGGACATTGTGGAGCTTGA-3′.

Expression arrays analysis. Expression microarrays were done on the Affymetrix 430 2.0 mouse microarray chips. This chipset represents some 45,000 genes and single-nucleotide polymorphisms from the mouse genome. Total RNA was isolated from two MntKO/CKO; MMTV-Cre mammary tumors and two c-Myc transgenic mammary tumors. The RNA was labeled and hybridized to the array using standard protocols. Microarray data were preprocessed via the Robust Multichip Average method (29) as implemented in Bioconductor,10

a suite of programs for the R statistical programming language.11
11

R Project for Statistical Computing: http://www.r-project.org/.

Robust Multichip Average preprocessing consists of three steps: background adjustment, normalization, and summarization (29). Analysis of expression was done using bioweight analysis (30). Bioweight is designed to produce a higher true positive rate compared with traditionally used t test– and P value–based methods of analysis, especially for smaller sample sets. Bioweight takes into consideration the across-replicate effect of traditional fold change analysis as well as the negative decimal log of P values for the gene specific t tests. This method allows the user to retain overexpressed/underexpressed genes that may be of biological significance as well as retain statistical stringency corresponding to a small P value. Therefore, the bioweight value can be described as the product of absolute average fold change and negative log of P value, accounting for small P values and large fold changes and providing a smooth transition between the two. For a detailed explanation of bioweight analysis, see Rosenfeld et al. (30).

Morphologic analysis of mammary gland development. To determine if loss of Mnt in the mammary gland was associated with developmental abnormalities, we analyzed mammary gland development during pregnancy, lactation, and involution in MMTV-Cre+/MntKO/CKO mice and MMTV-Cre/MntKO/CKO control mice by whole-mount (Fig. 1A-L) and histologic sections (Fig. 1M-X). Virgin mammary glands (V, Fig. 1A,, G and M, S) and mammary gland development during pregnancy (P17, Fig. 1B,, H and N, T) and during and lactation (L8, Fig. 1C,, I and O, U) were indistinguishable between genotypes. However, mammary glands from MMTV-Cre+/MntKO/CKO mice displayed defects characterized by enlarged ducts during involution after weaning (arrows in Fig. 1J,, K, L, W, X). In control mice, involution was advanced between 2 days postweaning (I2, 2 days after removal of pups; Fig. 1D and P) and 5 days postweaning (I5, 5 days after removal of pups; Fig. 1E and Q) whereas the mammary gland epithelial tree was greatly reduced by day 16 (I16, 16 days after removal of pups; Fig. 1F and R). By contrast, in MMTV-Cre+/MntKO/CKO mice, involution was delayed at I2 (Fig. 1J and V), I5 (Fig. 1K and W), and I16 (Fig. 1L and X) postweaning compared with controls. At the end of involution, the ducts of MMTV-Cre+/MntKO/CKO mice were still enlarged in comparison with those of MMTV-Cre/MntKO/CKO mice (Fig. 1L, and X, arrows). These changes were quantitated by calculating the area of ducts in histologic sections (Fig. 2), showing significant differences between genotypes at I5 and I15.

Figure 1.

Morphologic analysis by whole-mount analysis (A-L) and histologic sections with H&E staining (M-X) of developing mammary gland in MMTV-Cre/MntKO/CKO mice (A-F, M-R) and MMTV-Cre+/MntKO/CKO mice (G-L, S-X). Note the enlarged ducts in MMTV-Cre+/MntKO/CKO mice (J-L, W, and X, arrows). V, virgin; P17, 17 days of pregnancy; L8, 8 days of lactation; I2, I5, and I16, 2, 5, and 16 days after removal of pups after 12 days of lactation.

Figure 1.

Morphologic analysis by whole-mount analysis (A-L) and histologic sections with H&E staining (M-X) of developing mammary gland in MMTV-Cre/MntKO/CKO mice (A-F, M-R) and MMTV-Cre+/MntKO/CKO mice (G-L, S-X). Note the enlarged ducts in MMTV-Cre+/MntKO/CKO mice (J-L, W, and X, arrows). V, virgin; P17, 17 days of pregnancy; L8, 8 days of lactation; I2, I5, and I16, 2, 5, and 16 days after removal of pups after 12 days of lactation.

Close modal
Figure 2.

Quantification of the ductal area in developing mammary glands from MMTV-Cre/MntKO/CKO mice (white columns; n = 3 for each stage) and MMTV-Cre+/MntKO/CKO mice (gray columns; n = 3-4 for each stage). Columns, mean; bars, SD. **, P < 0.01; ***, P < 0.001.

Figure 2.

Quantification of the ductal area in developing mammary glands from MMTV-Cre/MntKO/CKO mice (white columns; n = 3 for each stage) and MMTV-Cre+/MntKO/CKO mice (gray columns; n = 3-4 for each stage). Columns, mean; bars, SD. **, P < 0.01; ***, P < 0.001.

Close modal

Reduced apoptosis during involution of the postlactation mammary gland. As involution occurs predominantly by apoptosis, we did TUNEL analysis on sections from involution days 2, 5, and 16 (I2, I5, and I16). In MMTV-Cre/MntKO/CKO mammary alveoli, apoptosis peaked at involution day 2 (Fig. 3A,, C, and E). In contrast, apoptosis peaked at involution day 5 in MMTV-Cre+/MntKO/CKO mice (Fig. 3B,, D, and F). The percentage of TUNEL+ cells was calculated in alveoli (Fig. 3G) and ducts (Fig. 3H), respectively. In alveoli, the peak level of apoptosis in MMTV-Cre+/MntKO/CKO mice was delayed compared with MMTV-Cre/MntKO/CKO mice (Fig. 3G). In addition, a decreased number of apoptotic cells were observed in the ducts of MMTV-Cre+/MntKO/CKO mice (Fig. 3H). These results are consistent with an important function for Mnt in the regulation of apoptosis in both alveoli and ducts during involution.

Figure 3.

Apoptotic defects during involution in MMTV-Cre+/MntKO/CKO mice. A-F, TUNEL staining was done on sections of mammary gland of MMTV-Cre/MntKO/CKO mice (A, C, and E; n = 3 for each stage) and MMTV-Cre+/MntKO/CKO mice (B, D, and F; n = 3 for each stage) on involution day 2 (A and B), day 5 (C and D), and day 16 (E and F). Magnification, × 400. G and H, quantification of TUNEL+ cell in alveoli (G) and ducts (H) of MMTV-Cre/MntKO/CKO mice (white columns) and MMTV-Cre+/MntKO/CKO mice (solid columns). Note delayed apoptosis in alveoli of MMTV-Cre+/MntKO/CKO mice and decreased number of apoptosis in those ducts. Columns, mean; bars, SD.

Figure 3.

Apoptotic defects during involution in MMTV-Cre+/MntKO/CKO mice. A-F, TUNEL staining was done on sections of mammary gland of MMTV-Cre/MntKO/CKO mice (A, C, and E; n = 3 for each stage) and MMTV-Cre+/MntKO/CKO mice (B, D, and F; n = 3 for each stage) on involution day 2 (A and B), day 5 (C and D), and day 16 (E and F). Magnification, × 400. G and H, quantification of TUNEL+ cell in alveoli (G) and ducts (H) of MMTV-Cre/MntKO/CKO mice (white columns) and MMTV-Cre+/MntKO/CKO mice (solid columns). Note delayed apoptosis in alveoli of MMTV-Cre+/MntKO/CKO mice and decreased number of apoptosis in those ducts. Columns, mean; bars, SD.

Close modal

Accumulated apoptotic defects resulted in the enlargement of mammary ducts. We calculated the percentage of enlarged ducts (Fig. 4A) and average area of ducts in aged virgin female mice (Fig. 4B) and aged nonvirgin female mice that had undergone multiple (68) pregnancies (Fig. 4C). Nonvirgin MMTV-Cre+/MntKO/CKO mice displayed ∼90% enlarged ducts in comparison with MMTV-Cre/MntKO/CKO mice, which had only 40% enlarged ducts, and virgin MMTV-Cre/MntKO/CKO mice and MMTV-Cre+/MntKO/CKO mice, which showed 30% and 50% enlarged ducts, respectively. The enlargement of ductal area is obvious in mammary glands after multiple pregnancies (Fig. 4C) in comparison with virgin mice (Fig. 4B). These results strongly suggest that the high percentage of enlarged ducts in nonvirgin MMTV-Cre+/MntKO/CKO mice results from accumulation of ductal material due to decreased apoptosis during successive involutions postpregnancy.

Figure 4.

Multiple pregnancies accelerate the ductal enlargement. A, percentage of enlarged ducts in MMTV-Cre/MntKO/CKO mice and MMTV-Cre+/MntKO/CKO virgin mice or mice that have had multiple pregnancies. B and C, ductal area in MMTV-Cre/MntKO/CKO mice (white boxes) and MMTV-Cre+/MntKO/CKO mice (shaded boxes) that are virgin (B; n = 4 for each genotype) or have had multiple pregnancies (C; n = 5 for each genotype). Columns, mean; bars, SD. *, P < 0.05; **, P < 0.01.

Figure 4.

Multiple pregnancies accelerate the ductal enlargement. A, percentage of enlarged ducts in MMTV-Cre/MntKO/CKO mice and MMTV-Cre+/MntKO/CKO virgin mice or mice that have had multiple pregnancies. B and C, ductal area in MMTV-Cre/MntKO/CKO mice (white boxes) and MMTV-Cre+/MntKO/CKO mice (shaded boxes) that are virgin (B; n = 4 for each genotype) or have had multiple pregnancies (C; n = 5 for each genotype). Columns, mean; bars, SD. *, P < 0.05; **, P < 0.01.

Close modal

c-Myc and Mnt expression during mouse mammary gland development. To support the notion that Mnt is essential for mammary gland development during pregnancy, lactation, and involution, we analyzed Mnt and c-Myc expression by Western blotting (Fig. 5). Mnt was detectable at different stages during mammary gland development in MMTV-Cre+/MntKO/CKO mice but its expression did show small fluctuations. In MMTV-Cre+/MntKO/CKO mice, Mnt expression was strongly reduced as expected. Compared with Mnt, c-Myc expression was highly variable during mammary gland development. c-Myc increased in MMTV-Cre+/MntKO/CKO mice during pregnancy, whereas during involution, there was a reduction then an increase in c-Myc. Similar changes in c-Myc were observed in MMTV-Cre+/MntKO/CKO mice. These results suggest that reduced expression of Mnt causes an imbalance of c-Myc expression, and this may be one cause of mammary tumor development in MMTV-Cre+/MntKO/CKO mice.

Figure 5.

Western blot analysis of Mnt and c-Myc expression during pregnancy, lactation, and involution. Protein lysates from MMTV-Cre/MntKO/CKO mice (lanes 1-4) and MMTV-Cre+/MntKO/CKO mice (lanes 5-8) were analyzed by Western blot for the expression levels of Mnt and c-Myc. β-Actin was used as a loading control.

Figure 5.

Western blot analysis of Mnt and c-Myc expression during pregnancy, lactation, and involution. Protein lysates from MMTV-Cre/MntKO/CKO mice (lanes 1-4) and MMTV-Cre+/MntKO/CKO mice (lanes 5-8) were analyzed by Western blot for the expression levels of Mnt and c-Myc. β-Actin was used as a loading control.

Close modal

Mammary gland–specific loss of Mnt results in mammary adenocarcinoma. As previously described (21), we used MMTV-Cre mice (31, 32) to produce 11 MMTV-Cre+/MntKO/CKO mice that were continuously mated to promote loss of Mnt specifically in mammary gland epithelial cells. Nine continuously mated female MMTV-Cre/MntKO/CKO mice were used as controls. Here we present further details of the mammary gland tumor phenotype of these mice, including Kaplan-Meier survival analysis and a more complete histologic analysis of tumors.

Seven of eleven MMTV-Cre+/MntKO/CKO mice developed mammary adenocarcinomas (Fig. 6A). Tumor latency ranged from 6 to 24 months (Fig. 6B). No tumors were observed in MMTV-Cre/MntKO/CKO mice (Fig. 6A) and mammary gland morphology was normal in these mice (Fig. 6C). Some of the tumors showed lymphocyte invasion (Fig. 6D). The tumors were invasive ductal carcinoma with varying histology, including solid tubular carcinoma (Fig. 6F), tubular carcinoma (Fig. 6G), and carcinoma with sarcomatous change (Fig. 6H). In areas adjacent to the tumors, the nontumor tissue seemed to display enlarged ductal morphology although we cannot rule out that such areas were early preneoplastic lesions (Fig. 6E).

Figure 6.

Mammary gland tumor development in MMTV-Cre+/MntKO/CKO mice. A, mammary gland tumor was developed in 22-month-old MMTV-Cre+/MntKO/CKO mice (top) compared with age-matched control MMTV-Cre/MntKO/CKO mice (bottom). B, incidence of mammary gland tumors in MMTV-Cre+/MntKO/CKO mice (dotted line) and MMTV-Cre/MntKO/CKO mice (solid line). MMTV-Cre+/MntKO/CKO mice and MMTV-Cre/MntKO/CKO mice were monitored over a period of 24 months. D, carcinoma with lymphocyte infiltration (arrows). Magnification, ×200. E, enlarged ducts. F, solid tubular carcinoma. G, tubular carcinoma. H, carcinoma with sarcomatous change. Note that MMTV-Cre+/MntKO/CKO mice develop some different types of mammary adenocarcinomas (F-H) and have enlarged ducts (E) compared with age-matched MMTV-Cre/MntKO/CKO mice (C). Magnification, ×200 (D), ×400 (C and E-H).

Figure 6.

Mammary gland tumor development in MMTV-Cre+/MntKO/CKO mice. A, mammary gland tumor was developed in 22-month-old MMTV-Cre+/MntKO/CKO mice (top) compared with age-matched control MMTV-Cre/MntKO/CKO mice (bottom). B, incidence of mammary gland tumors in MMTV-Cre+/MntKO/CKO mice (dotted line) and MMTV-Cre/MntKO/CKO mice (solid line). MMTV-Cre+/MntKO/CKO mice and MMTV-Cre/MntKO/CKO mice were monitored over a period of 24 months. D, carcinoma with lymphocyte infiltration (arrows). Magnification, ×200. E, enlarged ducts. F, solid tubular carcinoma. G, tubular carcinoma. H, carcinoma with sarcomatous change. Note that MMTV-Cre+/MntKO/CKO mice develop some different types of mammary adenocarcinomas (F-H) and have enlarged ducts (E) compared with age-matched MMTV-Cre/MntKO/CKO mice (C). Magnification, ×200 (D), ×400 (C and E-H).

Close modal

Mnt and Myc bind similar promoters. The notion that Mnt may act as a counterregulator of c-Myc predicts that c-Myc and Mnt bind to similar sequences in the promoter regions of regulated genes by competitively binding to those locations. To assess this hypothesis, Genome Wide Location Analysis (GWLA) was used to determine if c-Myc and Mnt have similar promoter occupancy in MMTV-c-Myc transgenic mammary tumors (33). One Mnt and two independent c-Myc location-analysis experiments were done on a M2K microarray chip containing ∼2,000 mouse gene promoters (Supplementary Table S1) as previously described (28). The P value was set at <0.001 to maintain a strict stringency for comparisons of promoter binding, and Mnt and Myc bound to a very similar set of genes by these strict criteria.

Promoter DNA sequences that coimmunoprecipitated with Mnt were rank ordered by their P value for binding affinity versus control, unenriched chromatin DNA. After this rank order was established, promoters that coprecipitated with c-Myc were aligned with the established order to determine if there was a similarity in P values for a subset of promoters. It was determined that at P < 0.001 (which was considered significant for this study), there were an extremely high number of genes that bound both Mnt and c-Myc with similarly significant P values.

c-Myc bound 138 promoter sequences in one trial and 114 unique gene promoters in another trial whereas Mnt bound 147 of a possible 1,920 genes (Fig. 7A-C). When the 138 unique promoters of the first c-Myc location-analysis were compared with the 147 promoters from Mnt, 111 were found to be common (Fig. 7D). Of 114 unique promoters from the second c-Myc experiment, 84 were fond to be in common with the same set of Mnt bound genes (Fig. 7E). The c-Myc experiments yielded 80 common significantly bound promoters when compared with each other (P < 0.001). Of these genes, 75 were also found to bind Mnt at a significant level (Supplementary Table S2), indicating only 5 of the 80 genes were not bound by Mnt at a significant level (Fig. 7F). We validated these results with chromatin immunoprecipitation analysis of five genes that bound both Mnt and c-Myc, Ibd1, Sap18, cdkn1b, Rab18, and Brca1, showing that Mnt and c-Myc bound specifically to these promoters but not sequences 3′ to the transcriptional start site of these genes (Supplementary Fig. S1). These results show a significant overlap in promoter binding of c-Myc and Mnt in tumor tissue.

Figure 7.

GWLA profiles of Myc and Mnt binding to M2K promoter arrays using tumors from MMTV-c-Myc transgenic mice and the overlap of significant binding. A, there were 138 promoters bound in one tumor sample by Myc at P < 0.001. B, Mnt bound 147 genes at a significant level; C, the second Myc bound 114 genes at a significant level. D, when Myc 1 binding is compared with Mnt, 111 genes were found to be common. Myc2 and Mnt shared 84 promoters (E) whereas Myc1 compared with Myc2 had 80 genes in common (F); all three combined shared 75 genes, indicating these 75 genes are probably biologically significant.

Figure 7.

GWLA profiles of Myc and Mnt binding to M2K promoter arrays using tumors from MMTV-c-Myc transgenic mice and the overlap of significant binding. A, there were 138 promoters bound in one tumor sample by Myc at P < 0.001. B, Mnt bound 147 genes at a significant level; C, the second Myc bound 114 genes at a significant level. D, when Myc 1 binding is compared with Mnt, 111 genes were found to be common. Myc2 and Mnt shared 84 promoters (E) whereas Myc1 compared with Myc2 had 80 genes in common (F); all three combined shared 75 genes, indicating these 75 genes are probably biologically significant.

Close modal

Expression analysis of genes preferentially bound by Mnt and c-Myc. If Mnt acts to regulate c-Myc, and the lack of Mnt results in tumorigenesis by removing a mechanism to counter c-Myc expression, we reasoned that there may be a similarity in gene expression between the two types of tumors. To test this hypo-thesis, we compared the expression of 45,000 genes in MMTV-c-Myc transgenic and MntCKO/CKO; MMTV-Cre mammary tumors. The expression levels were analyzed using bioweight analysis, which is a combined measure of fold change and P values from t tests (30). This method has been shown to have more statistical power for small sample sets than traditional means of microarray analysis based on the t test. A large bioweight value suggests that the gene is likely to have a small P value, a large fold change, and is more likely to have statistically significant differences in expression. This allows one to rank order the genes from most to least similar in expression based on two methods of measurement. We were interested to see if the different tumors from MMTV-c-Myc and MntCKO/CKO; MMTV-Cre mice show any similarity in gene expression pattern, indicated by a low bioweight value.

A volcano plot of the 45,000 genes indicates that there are very few genes with significant bioweights (Fig. 8), suggesting an overall expression pattern that is very similar between c-Myc transgenic and MMTV-CRE+/MntCKO/CKO mammary tumors. In this plot, the X axis represents the across-replicate binary fold change between c-Myc transgenic and MMTV-CRE+/MntCKO/CKO. The Y axis represents the negative decimal log of P values for the gene specific t tests using the null hypothesis that the average log-ratios are zero (no difference between groups). The most statistically significant genes are located on the top left and top right boxed areas of the volcano plot. Left-to-right variability corresponds to a large absolute fold change. This method of measurement is able to represent the relationship between “biological” and “statistical” significance in one graph. Because only two genes are located in either of these boxes, the vast majority of genes are similarly expressed in tumors from c-Myc transgenic and MMTV-CRE+/MntCKO/CKO.

Figure 8.

Bioweight analysis reveals a tight clustering of expression in both the ranges of P values and fold change in expression between MMTV-c-Myc transgenic and MntCKO/CKO; MMTV-Cre tumors. Gray-shaded points, ∼45,000 genes from the Affymetrix 430 2.0 mouse microarray chip. Black-shaded points, 75 genes identified by GWLA. Horizontal dotted line, P value significance threshold corresponding to a type I error rate of 0.05. Vertical dotted lines, fold change thresholds corresponding to a binary log fold difference of 1 and −1 (which is equivalent to an absolute regular fold change of 2). Genes significantly different between tumors from MMTV-c-Myc transgenic and MntCKO/CKO; MMTV-Cre mice are in the top left and top right areas delineated by the dotted lines, and only two genes were in those quadrants. Of the 75 genes clustered tightly by GWLA, only one was expressed at a level above 1 binary log fold difference and another gene exceeded the P value threshold.

Figure 8.

Bioweight analysis reveals a tight clustering of expression in both the ranges of P values and fold change in expression between MMTV-c-Myc transgenic and MntCKO/CKO; MMTV-Cre tumors. Gray-shaded points, ∼45,000 genes from the Affymetrix 430 2.0 mouse microarray chip. Black-shaded points, 75 genes identified by GWLA. Horizontal dotted line, P value significance threshold corresponding to a type I error rate of 0.05. Vertical dotted lines, fold change thresholds corresponding to a binary log fold difference of 1 and −1 (which is equivalent to an absolute regular fold change of 2). Genes significantly different between tumors from MMTV-c-Myc transgenic and MntCKO/CKO; MMTV-Cre mice are in the top left and top right areas delineated by the dotted lines, and only two genes were in those quadrants. Of the 75 genes clustered tightly by GWLA, only one was expressed at a level above 1 binary log fold difference and another gene exceeded the P value threshold.

Close modal

The 75 genes discovered to have similarly bound promoters using GWLA at P ≤ 0.001 were then assessed. Many of these genes are of obvious biological relevance in tumorigenesis, cell cycle regulation, proliferation, and transcriptional regulation (data not shown; Supplementary Table S2). The bioweight values for these genes varied from 0.0001 to 0.73, with the majority decreasing below 0.2 (63 of 75 were below 0.2). These low values indicate that the overall expression pattern is similar between the tumor types and these directly regulated genes are not different in expression in mammary gland tumors resulting from loss of Mnt or overexpression of c-Myc.

Elevated c-Myc expression is a common feature of human breast cancers (reviewed in ref. 34) and overexpression of c-Myc in mammary gland tissue of transgenic mice leads to mammary gland cancer (reviewed in ref. 35). Thus, mammary gland epithelial tissue offers an excellent model system to study the function of the putative Myc antagonist Mnt. Consistent with the proposed Myc-antagonist function of Mnt, we previously showed that deletion of Mnt specifically in mammary gland tissue using a Mnt conditional knockout led to the formation of adenocarcinomas (21). Here we show that Mnt also plays an important role in mammary gland development preceding tumor formation, as loss of Mnt impairs mammary gland involution. In addition, we further characterize tumor formation caused by loss of Mnt and show that the gene expression profiles of tumors that lack Mnt are similar to those caused by Myc overexpression.

The finding that loss of Mnt can lead to mammary gland tumorigenesis is consistent with our results showing that loss of Mnt causes a hyperproliferation phenotype in mouse embryonic fibroblasts, including accelerated proliferation, increased sensitivity to apoptosis, sensitivity to transformation by H-RasV12 alone, and senescence bypass (21). In both mouse embryonic fibroblasts and mammary gland tumors, loss of Mnt was found to cause up-regulation of cyclin E and Cdk4 (21), both of which have been implicated in mammary gland tumorigenesis (reviewed in ref. 36). Mnt has also been shown to preferentially bind similar promoter regions to c-Myc in mammary tumor tissues from c-Myc transgenic mice and the corresponding proteins were shown to be expressed at similar levels in both Mnt knockout and c-Myc transgenic tumors. Furthermore, because it is well established that c-Myc promotes mammary gland tumorigenesis in transgenic mice (33), these results are consistent with the proposed role of Mnt as a Myc antagonist and tumor suppressor.

Depending on the in vivo cellular context, dysregulation of c-Myc can either promote proliferation or induce apoptosis (3739). Induction of apoptosis, and possibly induction of cell cycle arrest (40), likely provide protection against the proliferative effects of over-activity of Myc family proteins. Indeed, Myc-driven tumorigenesis is dramatically accelerated in mouse models when apoptosis is suppressed by expressing antiapoptotic proteins (39, 4143) or by disruption of p53 pathway function (25, 4347). Because, in general, cells are sensitized to apoptosis by Myc overexpression and we previously found apoptosis levels increased in fibroblasts lacking Mnt, it was predicted that loss of Mnt and its Myc-antagonistic activity would result in increased apoptosis in mammary gland tissue (21). However, just the opposite was observed during mammary gland involution (Fig. 3), a process characterized by high levels of apoptosis (48). It is possible that cell type differences in the response to loss of Mnt are responsible for this apparent discrepancy or that the decreased apoptosis is the result of impaired initiation of the involution process that precedes active apoptosis. Another possible explanation is that the observed decreased apoptosis is a secondary response to hyper-Myc activity caused by loss of Mnt. In the later scenario, decreased apoptosis during involution may be a manifestation of prior elimination—by apoptosis—of cells important for the involution process. Under such conditions, cells with innate or acquired resistance to apoptosis would have a selective advantage and perhaps are the precursors to tumors caused by loss of Mnt. In situations of Myc overexpression, variant cells less prone to apoptosis are selected that have disrupted ADP ribosylation factor (ARF)/murine double minute-2/p53 pathway function (25, 46, 49). Similarly, primary Mnt−/− fibroblasts seem to sustain p53 mutation or loss of ARF more rapidly than wild-type cells during serial passaging (21) and tumors arising from MMTV-Cre+/MntKO/CKO mice express elevated p53, which is often indicative of disabling mutations in p53 (data not shown). Furthermore, there is a modest increase in p53 and p19ARF in the phenotypically normal mammary gland tissue of some MMTV-Cre+/MntKO/CKO mice (data not shown), raising the possibility that the decrease in apoptosis during involution is associated with the evolution of cells in mammary gland tissue that have sustained mutations in the p53 gene. Interestingly, p53−/− mice on a BALB/c genetic background exhibit a delayed involution phenotype and develop mammary tumors (50). The role of the ARF/murine double minute-2/p53 pathway and suppression of apoptosis in tumor formation by MMTV-Cre+/MntKO/CKO mice can be tested by determining whether mammary gland tumor formation in the absence of Mnt is accelerated by, for example, loss of ARF or p53 or by overexpression of Bcl-2 or other antiapoptotic proteins.

We found that mice with multiple pregnancies had a higher percentage of enlarged ducts and larger average size of ductal lumen than MMTV-Cre+/MntKO/CKO virgin mice (Fig. 4). The accumulation of apoptotic defects in the involution of each pregnancy may be part of the reason for causing the enlargement of mammary ducts. However, virgin MMTV-Cre+/MntKO/CKO mice showed higher percentage of enlarged ducts than virgin wild-type mice. This result suggests that there are other causes than the accumulation of apoptotic defects. Hyperproliferation of the ductal epithelial cells may be a possible explanation but we could not detect any difference in proliferating cell nuclear antigen expression between MMTV-Cre+/MntKO/CKO mice and wild-type mice (data not shown). Although no changes were noted in levels of Mnt expression during mammary gland development, c-Myc levels varied greatly during mammary gland development. Thus, in the absence of Mnt, these changes in c-Myc will be unopposed and may lead to the developmental defects observed.

Similar promoter binding of c-Myc and Mnt in c-Myc mammary tumors (Fig. 7) supports the hypothesis that Mnt may be an important direct regulator of c-Myc in vivo by competing with c-Myc for binding to Max and shared target genes. Although the Mad protein family may be important in this regulation under certain conditions, evidence presented here, as well as data showing concurrent expression of c-Myc and Mnt in dividing cells (21), suggests that Mnt antagonism may be particularly relevant to the control of cell proliferation. In addition, our results suggest that global gene expression is virtually indistinguishable in tumors of c-Myc transgenic mice and tumors of MntKO/CKO; MMTV-Cre conditional knockout mice (Fig. 8). Taken together with previous cell culture studies (21), these findings support the hypothesis that Mnt acts in direct functional opposition to c-Myc in regulating transcription and that loss of Mnt is, to a large extent, functionally equivalent to elevated and/or deregulated c-Myc expression.

It is known that the MMTV promoter contains a hormonal regulatory element, which can be induced by progesterone, glucocorticoids, androgens, and prolactin (5153). We have continuously mated MMTV-Cre+/MntKO/CKO mice to promote the expression of Cre transgene. There are many reports that hormones are important factors to alter mammary tumor development (5456). Unlike female MMTV-Cre+/MntKO/CKO mice, male mice did not develop mammary tumors (data not shown). Male MMTV-Cre+/MntKO/CKO mice might develop mammary tumors after the administration of hormones and may be a good tool to analyze the effect of hormones in the development of mammary tumors in MMTV-Cre+/MntKO/CKO mice. However, it remains to be determined whether hormones have a role in Myc/Mnt signaling or only affect the direct Cre transgene induction in MMTV-Cre+/MntKO/CKO mice.

We also produced mice with a null mutation of Mnt using the conditional allele and germ line Cre deleter strains (21, 23). Almost all homozygous mutants (Mnt−/−) in a mixed or inbred genetic background died perinatally. Of the <2% of Mnt−/− mice in a mixed genetic background and ∼3.5% Mnt−/− mice in an inbred background that survived to adulthood, all died within 1 year and most of these died within several months of birth from currently unknown causes. Therefore, it has not been possible to address whether mice completely deficient for Mnt develop mammary gland cancers and/or other tumors and we have not detected any tumors in Mnt heterozygotes. Thus, Mnt conditional knockout mice will be required to address the possibility that Mnt functions as a tumor suppressor in other cell types and to determine whether MNT is the tumor suppressor or one of the general tumor suppressors suspected to reside at human chromosome 17p13.3 (57, 58).

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

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.

We thank Hideki Wanibuchi for helpful discussions.

1
Grandori C, Cowley SM, James LP, Eisenman RN. The Myc/Max/Mad network and the transcriptional control of cell behavior.
Annu Rev Cell Dev Biol
2000
;
16
:
653
–99.
2
Nesbit CE, Tersak JM, Prochownik EV. MYC oncogenes and human neoplastic disease.
Oncogene
1999
;
18
:
3004
–16.
3
Stewart TA, Pattengale PK, Leder P. Spontaneous mammary adenocarcinomas in transgenic mice that carry and express MTV/myc fusion genes.
Cell
1984
;
38
:
627
–37.
4
Schoenenberger CA, Andres AC, Groner B, van der Valk M, LeMeur M, Gerlinger P. Targeted c-myc gene expression in mammary glands of transgenic mice induces mammary tumours with constitutive milk protein gene transcription.
EMBO J
1988
;
7
:
169
–75.
5
Leder A, Pattengale PK, Kuo A, Stewart TA, Leder P. Consequences of widespread deregulation of the c-myc gene in transgenic mice: multiple neoplasms and normal development.
Cell
1986
;
45
:
485
–95.
6
Walker W, Zhou ZQ, Ota S, Wynshaw Boris A, Hurlin PJ. Mnt-Max to Myc-Max complex switching regulates cell cycle entry.
J Cell Biol
2005
;
169
:
405
–13.
7
Devilee P, Cornelisse CJ. Somatic genetic changes in human breast cancer.
Biochim Biophys Acta
1994
;
1198
:
113
–30.
8
McDonald JD, Daneshvar L, Willert JR, Matsumura K, Waldman F, Cogen PH. Physical mapping of chromosome 17p13.3 in the region of a putative tumor suppressor gene important in medulloblastoma.
Genomics
1994
;
23
:
229
–32.
9
Williamson MP, Elder PA, Knowles MA. The spectrum of TP53 mutations in bladder carcinoma.
Genes Chromosomes Cancer
1994
;
9
:
108
–18.
10
Biegel JA, Burk CD, Barr FG, Emanuel BS. Evidence for a 17p tumor related locus distinct from p53 in pediatric primitive neuroectodermal tumors.
Cancer Res
1992
;
52
:
3391
–5.
11
Takahashi T, Konishi H, Kozaki K, et al. Molecular analysis of a Myc antagonist, ROX/Mnt, at 17p13.3 in human lung cancers.
Jpn J Cancer Res
1998
;
89
:
347
–51.
12
Sommer A, Waha A, Tonn J, et al. Analysis of the Max-binding protein MNT in human medulloblastomas.
Int J Cancer
1999
;
82
:
810
–6.
13
Nigro CL, Venesio T, Reymond A, et al. The human ROX gene: genomic structure and mutation analysis in human breast tumors.
Genomics
1998
;
49
:
275
–82.
14
Hirano A, Emi M, Tsuneizumi M, et al. Allelic losses of loci at 3p25.1, 8p22, 13q12, 17p13.3, and 22q13 correlate with postoperative recurrence in breast cancer.
Clin Cancer Res
2001
;
7
:
876
–82.
15
Fujii H, Biel MA, Zhou W, Weitzman SA, Baylin SB, Gabrielson E. Methylation of the HIC-1 candidate tumor suppressor gene in human breast cancer.
Oncogene
1998
;
16
:
2159
–64.
16
Morton RA, Jr., Watkins JJ, Bova GS, Wales MM, Baylin SB, Isaacs WB. Hypermethylation of chromosome 17P locus D17S5 in human prostate tissue.
J Urol
1996
;
156
:
512
–6.
17
Wales MM, Biel MA, el Deiry W, et al. p53 activates expression of HIC-1, a new candidate tumour suppressor gene on 17p13.3.
Nat Med
1995
;
1
:
570
–7.
18
Chen WY, Zeng X, Carter MG, et al. Heterozygous disruption of Hic1 predisposes mice to a gender-dependent spectrum of malignant tumors.
Nat Genet
2003
;
33
:
197
–202.
19
Cvekl A, Jr., Zavadil J, Birshtein BK, Grotzer MA, Cvekl A. Analysis of transcripts from 17p13.3 in medulloblastoma suggests ROX/MNT as a potential tumour suppressor gene.
Eur J Cancer
2004
;
40
:
2525
–32.
20
Dobyns WB, Reiner O, Carrozzo R, Ledbetter DH. Lissencephaly. A human brain malformation associated with deletion of the LIS1 gene located at chromosome 17p13.
JAMA
1993
;
270
:
2838
–42.
21
Hurlin PJ, Zhou Z-Q, Toyo-oka K, et al. Deletion of Mnt leads to disrupted cell cycle control and tumorigenesis.
EMBO J
2003
;
22
:
1
–13.
22
Nilsson JA, Cleveland JL. Mnt: master regulator of the Max network.
Cell Cycle
2004
;
3
:
588
–90.
23
Toyo-oka K, Hirotsune S, Gambello MJ, et al. Loss of the Max-interacting protein Mnt in mice results in decreased viability, defective embryonic growth and craniofacial defects: relevance to Miller-Dieker syndrome.
Hum Mol Genet
2004
;
13
:
1057
–67.
24
Lakso M, Pichel JG, Gorman JR, et al. Efficient in vivo manipulation of mouse genomic sequences at the zygote stage.
Proc Natl Acad Sci U S A
1996
;
93
:
5860
–5.
25
Eischen CM, Weber JD, Roussel MF, Sherr CJ, Cleveland JL. Disruption of the ARF-Mdm2-53 tumor suppressor pathway in Myc-induced lymphomagenesis.
Genes Dev
1999
;
13
:
2658
–69.
26
Hurlin PJ, Steingrimsson E, Copeland NG, Jenkins NA, Eisenman RN. Mga, a dual-specificity transcription factor that interacts with Max and contains a T-domain DNA-binding motif.
EMBO J
1999
;
18
:
7019
–28.
27
Metivier R, Penot G, Hubner MR, et al. Estrogen receptor-α directs ordered, cyclical, and combinatorial recruitment of cofactors on a natural target promoter.
Cell
2003
;
115
:
751
–63.
28
Li Z, Van Calcar S, Qu C, Cavenee WK, Zhang MQ, Ren B. A global transcriptional regulatory role for c-Myc in Burkitt's lymphoma cells.
Proc Natl Acad Sci U S A
2003
;
100
:
8164
–9.
29
Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data.
Nucleic Acids Res
2003
;
31
:
e15
.
30
Rosenfeld S, Wang T, Kim Y, Milner J. Numerical deconvolution of cDNA microarray signal: simulation study.
Ann N Y Acad Sci
2004
;
1020
:
110
–23.
31
Xu X, Wagner KU, Larson D, et al. Conditional mutation of Brca1 in mammary epithelial cells results in blunted ductal morphogenesis and tumour formation.
Nat Genet
1999
;
22
:
37
–43.
32
Wagner KU, Wall RJ, St Onge L, et al. Cre-mediated gene deletion in the mammary gland.
Nucleic Acids Res
1997
;
25
:
4323
–30.
33
Hutchinson JN, Muller WJ. Transgenic mouse models of human breast cancer.
Oncogene
2000
;
19
:
6130
–7.
34
Liao DJ, Dickson RB. c-Myc in breast cancer.
Endocr Relat Cancer
2000
;
7
:
143
–64.
35
Amundadottir LT, Merlino G, Dickson RB. Transgenic mouse models of breast cancer.
Breast Cancer Res Treat
1996
;
39
:
119
–35.
36
Landberg G. Multiparameter analyses of cell cycle regulatory proteins in human breast cancer: a key to definition of separate pathways in tumorigenesis.
Adv Cancer Res
2002
;
84
:
35
–56.
37
Pelengaris S, Littlewood T, Khan M, Elia G, Evan G. Reversible activation of c-Myc in skin: induction of a complex neoplastic phenotype by a single oncogenic lesion.
Mol Cell
1999
;
3
:
565
–77.
38
Pelengaris S, Khan M, Evan G. c-MYC: more than just a matter of life and death.
Nat Rev Cancer
2002
;
2
:
764
–76.
39
Pelengaris S, Khan M, Evan GI. Suppression of Myc-induced apoptosis in β cells exposes multiple oncogenic properties of Myc and triggers carcinogenic progression.
Cell
2002
;
109
:
321
–34.
40
Felsher DW, Zetterberg A, Zhu J, Tlsty T, Bishop JM. Overexpression of MYC causes p53-dependent G2 arrest of normal fibroblasts.
Proc Natl Acad Sci U S A
2000
;
97
:
10544
–8.
41
Strasser A, Harris AW, Bath ML, Cory S. Novel primitive lymphoid tumours induced in transgenic mice by cooperation between myc and bcl-2.
Nature
1990
;
348
:
331
–3.
42
Jager R, Herzer U, Schenkel J, Weiher H. Overexpression of Bcl-2 inhibits alveolar cell apoptosis during involution and accelerates c-myc-induced tumorigenesis of the mammary gland in transgenic mice.
Oncogene
1997
;
15
:
1787
–95.
43
Jacobs JJ, Scheijen B, Voncken JW, Kieboom K, Berns A, van Lohuizen M. Bmi-1 collaborates with c-Myc in tumorigenesis by inhibiting c-Myc-induced apoptosis via INK4a/ARF.
Genes Dev
1999
;
13
:
2678
–90.
44
Blyth K, Terry A, O'Hara M, et al. Synergy between a human c-myc transgene and p53 null genotype in murine thymic lymphomas: contrasting effects of homozygous and heterozygous p53 loss.
Oncogene
1995
;
10
:
1717
–23.
45
Elson A, Deng C, Campos Torres J, Donehower LA, Leder P. The MMTV/c-myc transgene and p53 null alleles collaborate to induce T-cell lymphomas, but not mammary carcinomas in transgenic mice.
Oncogene
1995
;
11
:
181
–90.
46
Schmitt CA, McCurrach ME, de Stanchina E, Wallace Brodeur RR, Lowe SW. INK4a/ARF mutations accelerate lymphomagenesis and promote chemoresistance by disabling p53.
Genes Dev
1999
;
13
:
2670
–7.
47
Eischen CM, Rehg JE, Korsmeyer SJ, Cleveland JL. Loss of Bax alters tumor spectrum and tumor numbers in ARF-deficient mice.
Cancer Res
2002
;
62
:
2184
–91.
48
Walker NI, Bennett RE, Kerr JF. Cell death by apoptosis during involution of the lactating breast in mice and rats.
Am J Anat
1989
;
185
:
19
–32.
49
Zindy F, Eischen CM, Randle DH, et al. Myc signaling via the ARF tumor suppressor regulates p53-dependent apoptosis and immortalization.
Genes Dev
1998
;
12
:
2424
–33.
50
Blackburn AC, Jerry DJ. Knockout and transgenic mice of Trp53: what have we learned about p53 in breast cancer?
Breast Cancer Res
2002
;
4
:
101
–11.
51
Darbre P, Dickson C, Peters G, Page M, Curtis S, King RJ. Androgen regulation of cell proliferation and expression of viral sequences in mouse mammary tumour cells.
Nature
1983
;
303
:
431
–3.
52
Chalepakis G, Arnemann J, Slater E, Bruller HJ, Gross B, Beato M. Differential gene activation by glucocorticoids and progestins through the hormone regulatory element of mouse mammary tumor virus.
Cell
1988
;
53
:
371
–82.
53
Haraguchi S, Good RA, Engelman RW, Day NK. Human prolactin regulates transfected MMTV LTR-directed gene expression in a human breast-carcinoma cell line through synergistic interaction with steroid hormones.
Int J Cancer
1992
;
52
:
928
–33.
54
Cavalieri EL, Stack DE, Devanesan PD, et al. Molecular origin of cancer: catechol estrogen-3,4-quinones as endogenous tumor initiators.
Proc Natl Acad Sci U S A
1997
;
94
:
10937
–42.
55
Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study.
J Natl Cancer Inst
1998
;
90
:
1371
–88.
56
Yang X, Edgerton SM, Kosanke SD, et al. Hormonal and dietary modulation of mammary carcinogenesis in mouse mammary tumor virus-c-erbB-2 transgenic mice.
Cancer Res
2003
;
63
:
2425
–33.
57
Hurlin PJ, Queva C, Eisenman RN. Mnt, a novel Max-interacting protein is coexpressed with Myc in proliferating cells and mediates repression at Myc binding sites.
Genes Dev
1997
;
11
:
44
–58.
58
Meroni G, Reymond A, Alcalay M, et al. Rox, a novel bHLHZip protein expressed in quiescent cells that heterodimerizes with Max, binds a non-canonical E box and acts as a transcriptional repressor.
EMBO J
1997
;
16
:
2892
–906.

Supplementary data