Since the early 1980s, numerous investigations have focused on c-Myc to explore its role in normal organ physiology, as well as in tumor biology 12. The focus of the present review, c-Myc (hereafter referred to as Myc), is the cellular homolog of the avian retroviral oncogene v-myc and, together with N-myc and L-myc, comprises the family of myc proto-oncogenes. The half-lives of Myc mRNA and protein are short, allowing for tight and rapid regulation of Myc levels, which occurs via numerous transcription factors (TFs) and signaling pathways. Proteins that directly bind the promoter or indirectly influence promoter activity have been reviewed recently 3. To provide some insight into the complexity of Myc regulation, we will mention a few of the factors and pathways that impact on its expression, many of which were shown to be essential during mammary gland development 4.

The myc promoter contains TF binding sites for Myc (auto-suppression), estrogen receptor (ER) alpha, T-cell factor (TCF) 4, Notch/C promoter-binding factor 1 (Cbf1), E2F, Fos/Jun, signal transducer and activator of transcription (Stat) 3, NF-κB, Smads and others. TFs that occupy or regulate the myc promoter without specific binding sites include p53, CCAAT/enhancer binding protein beta, and Stat5. Moreover, numerous signaling pathways that are frequently deregulated in human cancer influence myc expression; for example, rat sarcoma (Ras)/extracellular signal-related kinase (Erk) and phosphoinositide 3-kinase (PI3K)/serine/threonine kinase Akt (Akt). Post-translational modifications of Myc include phos-phorylation, ubiquitinylation and acetylation, and their effects on Myc activity have been reviewed 5.

The Myc protein is a basic helix-loop-helix TF that must heterodimerize with the abundantly expressed Max to regulate transcription. Myc-Max dimers bind to hexameric DNA sequences (E-box) and activate transcription by recruiting multiple coactivators 1. In contrast, when dimerized with basic helix-loop-helix proteins such as Mad or Mnt, Max binds to E-boxes but represses transcription. Myc can also act as a transcriptional repressor via different mechanisms, often involving interaction with Miz1 (for reviews of Myc transcriptional activity, see 16). It is now well accepted that Myc acts as a relatively weak activator of RNA polymerase II-driven transcription for a large set of target genes, thereby affecting cell cycle, cell growth and metabolism, cell death, adhesion, angiogenesis and other functions (to date, almost 1,700 targets on Myc Cancer Gene 78). Furthermore, Myc affects RNA polymerase I- and III-mediated transcription 1, thus regulating ribosome biogenesis and translation. Nontranscriptional roles for Myc in DNA replication and in translation have also been reported recently 9.

Considering the large number of receptors, hormones, paracrine factors and other signaling molecules that can impact on Myc levels, it is likely that Myc has a variety of functions throughout normal mammary gland development, downstream of one or more of these inputs. To date, however, there are only a few studies on the physiological role of Myc in the mammary gland. In this review we present what is known about Myc from transgenics and conditional knockout models, and also include indirect evidence implicating Myc based on results from other studies. We shall discuss the potential inputs activating Myc during development and the ensuing outputs of Myc activity. A summary of the discussion on the role of Myc in normal development is shown in Figure 1.

Development of the mammary gland begins at embryonic day 10 as an appendix of the ventral skin, with formation of the milk line followed by appearance of placodes 10. The ErbB4 ligand, Neuregulin3 (Nrg3), was identified as a specification signal for placode formation. Based on this, Keratin14 (K14)-Nrg3 transgenic mice expressing Nrg3 throughout the basal layer of the epidermis, including the stem and progenitor cells, were investigated 11. Ectopic Nrg3 expression resulted in hyperplastic epidermis and formation of supernumerary placodes. Interestingly, the skin of K14-Nrg3 mice displayed increased Myc expression and decreased levels of α₆-integrin and β₁-integrin, which are adhesion receptors highly expressed in adult mammary stem cells (SCs) 12. While the direct stimulus of Myc expression is unknown, it seems to be an important mediator of the phenotype observed in K14-Nrg3 mice, as strong similarities were found in a K14-Myc model where Myc is activated in the epidermis 11. Nrg3 therefore possibly has a role in promoting mammary lineage commitment and in the regulation of SC fate via Myc.

While these results suggest a role for Myc in embryonic development, its role in adult mammary SCs has not yet been analyzed. A function for Myc in mammary SCs seems likely, however, based on its role in other well-characterized models 113. In the hematopoietic system, the balance between SC self-renewal and differentiation is controlled by Myc levels, which in turn regulate expression of adhesion molecules such as N-cadherin and β₁-integrin 14. In addition, the Wnt and Notch signaling pathways have been proposed to play important roles in mammary SCs 13, and their effector proteins - β-catenin/TCF and Notch intracellular domain (N^IC)/Cbf1, respectively - each have binding sites on the Myc promoter 3. Furthermore, a transgenic model expressing stabilized β-catenin in basal cells, which are believed to contain the SC population 12, displayed upregulation of Myc 15. Finally, the HC11 mammary epithelial cell line, which has SC-like properties, might be an interesting model to explore Myc function, since Myc levels are downregulated when these cells are induced to differentiate 16.

The steroid hormones estrogen and progesterone, as well as the prolactin receptor binding peptides prolactin and placental lactogen, dominate the extensive developmental changes occurring during puberty and pregnancy 417. Both estrogen and progesterone are able to directly stimulate Myc expression via an estrogen response element 18 and a progesterone receptor regulatory element 19. While Myc is expressed at low levels in prepubertal and virgin glands, it reaches its highest expression levels between days 6.5 and 12.5 of pregnancy, after which the RNA level slowly returns to baseline until parturition 20.

Interestingly, Myc was shown to be directly downstream of estrogen and progesterone in breast cancer cells, stimulating their proliferation 2122, but one of the major differences between normal breast and malignant breast is that the estrogen- and progesterone receptor-positive cells do not proliferate during normal development. Instead, by producing paracrine mediators, estrogen and progesterone stimulate proliferation of neighboring cells via amphiregulin 23 and Wnt4 24, respectively. Other growth factors, such as epidermal growth factor or receptor activator of NF-κB ligand, have also been shown to act as paracrine mediators during mammary gland development 25; and importantly, Wnt, epidermal growth factor and receptor activator of NF-κB ligand might all affect Myc levels directly or indirectly 3.

Although no study has yet addressed the contribution of Myc to pubertal development and early pregnancy, it is very likely that Myc is induced via paracrine signals, and, at least early in pregnancy when Myc levels are highest, Myc might promote proliferation. Importantly, different Myc targets such as cyclin-dependent kinase (Cdk) 4, nucleophosmin and nucleolin are also highly expressed at this time 20 - prompting us to speculate that Myc might not only have a direct role in proliferation, but also in the synthesis of ribosomal components required for rapid growth during pregnancy.

A transgenic model revealed why it is important for Myc levels to decrease, starting at 12.5 days of pregnancy, and to remain low until parturition. Using a doxycycline inducible model (MMTV-rtTA/TetO-MYC (MTB/TOM)) 26, it was shown that transient overexpression of Myc between days 12.5 and 15.5 of pregnancy induced a lactation failure 27. Abnormal Myc expression was shown not only to induce proliferation, but also to promote precocious Stat5 activation and differentiation, followed by premature involution, triggered by milk stasis. Decreased levels of Caveolin1 (Cav1), a direct target of Myc repression 7, were shown to be responsible for the phenotype. Cav1 is a negative regulator of Janus kinase2-Stat5 signaling, and Cav1^-/-mammary epithelial cells show hyperactivation of Stat5 and spontaneous milk production 28, similar to what is observed when Myc is elevated late in pregnancy.

These results demonstrate an important characteristic of Myc - namely, the effects of Myc are dependent upon the developmental stage of the mammary gland. Myc overexpression between 12.5 and 15.5 days of pregnancy was necessary and sufficient to induce the observed phenotype, while overexpression during other short intervals (for example, 9.5 to 12.5 days) did not result in lactation failure 27. Deregulated Myc therefore leads to a premature decrease in Cav1, thereby removing its restraining influence on prolactin receptor-Janus kinase2-Stat5 signaling.

In many cell types Myc is downregulated when cells undergo terminal differentiation. Indeed, Myc RNA levels do drop during lactation to below those levels found in the virgin gland 20; however, the molecular reason for this dramatic decrease is not known. The mammary gland as a milk factory produces immense amounts of lipids, lactose and proteins, and most of its energy is devoted to milk component synthesis. Considering the importance of Myc in energy and glucose metabolism as well as ribosome biogenesis and translation 1, it is possible that, even despite its low levels, Myc has an essential function during lactation.

Indeed, data from our laboratory revealed a novel role for Myc in the mammary gland using a conditional knockout approach. In c-myc^fl/flWAP(whey acidic protein)iCre mice, loss of Myc occurs exclusively in luminal alveolar cells starting from mid-pregnancy. We show that milk production was reduced in Myc mutant mothers, while milk composition was unchanged between wild-type and mutant mothers 29. Electron microscopy revealed that there were less secretory vesicles budding from the endoplasmic reticulum in lactating mutant cells, suggesting a decreased protein synthesis. In polysomal fractionation experiments we found that translation efficiency was generally decreased in lactating, Myc-deficient mammary glands. Furthermore, we observed reduced expression levels of ribosomal proteins and RNA, as well as proteins involved in translation and ribosome biogenesis. Although compensation by N-Myc or L-Myc cannot be excluded, neither was found upregulated in c-Myc mutant glands. These results highlight Myc's importance for mammary gland function even when endogenous levels are low.

The impact of Myc on apoptosis has been widely studied in many systems 2. In the mammary gland, the high levels of apoptosis during the first phase of involution are promoted by the leukemia inhibitory factor-Stat3 axis 30. Compared with its low expression in lactating glands, higher Myc levels are detected during involution 20. Importantly, a role for Myc during the first apoptotic phase was uncovered in mice with a conditional deletion of Socs3 (Socs3^-/flWAPiCre), a negative regulator of leukemia inhibitory factor-Stat3 signaling 31. Socs3-deficient glands displayed accelerated apoptosis accompanied by elevated levels of p-Stat3 and Myc, which is a direct Stat3 target gene.

To further analyze Myc's function in apoptosis, the doxycycline-inducible MTB/TOM model described above 26 was used, giving more evidence for a direct role of Myc in involution. Overexpression of Myc prior to forced weaning caused a dramatic acceleration of involution, accompanied by increased apoptosis and high levels of the pro-apoptotic proteins Bax, E2F-1 and p53, which have all been described as direct or indirect Myc targets.

The conclusions based on those two models suggest that Myc acts as a central mediator of apoptotic signaling in the mammary gland, being a direct target of Stat3 and inducing expression of pro-apoptotic genes.

Alterations in Myc have been found in many types of tumors. At the genomic level these include gene amplification, chromosomal translocations and point mutations. Furthermore, Myc is regulated by multiple signals that control promoter activity, transcriptional elongation and translation, as well as by post-translational modifications that control Myc's transcriptional targets as well as protein stability. Since most tumors have numerous alterations in signaling cascades, Myc is likely to be deregulated by some mechanisms in most cancers.

Considering breast cancer, amplification is the most commonly described alteration. The MYC amplicon on chromosome 8q23-24 was one of the first consistent genetic alterations found 32. Results from a meta-analysis of breast tumors yielded a frequency of 15.7% for the MYC amplicon, with a range of 4 to 52% depending on the study 2133. MYC amplification is found in a high proportion of tumors with Brca1 alterations, as well as in ERα-negative, basal-like tumors 3435. Despite intense screening efforts, point mutations of Myc have not been described in breast or other carcinomas 36. Other mechanisms promoting increased Myc levels, however, have been found. The ubiquitin ligase F-box and WB repeat domain containing 7, which catalyzes polyubiquitination of Myc and ensuing degradation, is often mutated or downregulated in breast cancer 37. Moreover, the de-ubiquitinating enzyme ubiquitin-specific protease 28-which antagonizes F-box and WB repeat domain containing 7, thereby stabilizing Myc - has been found overexpressed in a small panel of breast tumors 38. In addition to these breast tumor-specific alterations, it is possible that Myc is deregulated in most breast tumors since the normal tight control that is exerted on Myc at multiple levels is impaired in essentially all cancer cells.

What is the output of deregulated Myc in breast cancer? Myc levels respond to proliferative and anti-proliferative stimuli, and many reported Myc target genes, such as cyclin D₂, Cdk4 and the Cdk inhibitor p21^Cip1, are important regulators of proliferation 2. A major mechanism underlying Myc's role in breast cancer is activation of cyclin E-Cdk2 via repression of p21^Cip1 21. Myc deregulation not only impacts on proliferation, but on various other processes like survival and apoptosis.

Here we would like to mention Myc and translational control, since we have recently shown an important role for Myc in translation also during mammary gland development 29. Generation of the Eμ-Myc B-cell leukemia model in mice with heterozygosity in the gene encoding the L24 ribosomal protein restored normal levels of protein synthesis in the leukemic cells, thereby suppressing Myc's oncogenic potential 39. These results show that, in addition to proliferative effects, abnormal Myc activation also deregulates protein synthesis, which in this model is necessary for oncogenesis.

Considering the broad-ranging effects of Myc, the outcome of its activation in breast cancer is likely to be dependent upon the cellular context. Indeed, using a siRNA approach to knockdown Myc in a panel of breast cancer cell lines, and combining this with a genomic and phenotypic analysis, revealed that selectively regulated target genes in each cell line were responsible for the differential effects resulting from Myc loss. A comprehensive list of potential Myc targets in the BT-474, MCF-7 and MDA-MB-231 breast cancer cell lines can be found in Cappellen and colleagues' study 40. A summary of the discussion on the role of Myc in cancer is shown in Figure 2.

Myc was the first oncogene tested for mammary tumor-forming potential using the mouse mammary tumor virus (MMTV)-long terminal repeat to drive its expression. Tumor incidence in MMTV-Myc transgenic females was high; however, the kinetics of tumor appearance suggested that Myc expression alone was not sufficient to induce cancer 41. Indeed, double transgenics expressing Myc and mutant Hras showed more rapid mammary tumor development 2. More recently, it was shown that Myc induction using the inducible MTB/TOM model described above 26 results in mammary tumors, with approximately one-half also harboring activating Kras2 mutations. Interestingly, these tumors did not regress following Myc deinduction, demonstrating that Hras mutations not only change tumor kinetics but also cause progression to Myc independency 26.

Each membrane spanning Notch receptor is proteolytically processed in response to ligand binding, releasing N^IC, which converts the nuclear Cbf1 repressor to a transcriptional activator. A link between aberrant Notch signaling and mammary cancer was first discovered in MMTV-induced tumors with proviral DNA integrated within the Notch4 gene, leading to constitutive N^IC expression. Different mechanisms activate Notch signaling in human breast cancer 4243. For example, coexpression of Jagged1 ligand and Notch receptors has been found in breast cancers, in particular the triple-negative (negative for ERα, for progesterone receptor and for ErbB2) subtype 44, suggesting an autocrine mechanism of Notch pathway activation. Furthermore, levels of Numb, a negative regulator of Notch, were shown to be low in ~50% of primary breast tumors 45, which could contribute to maintenance of pathway activity.

Based on the observation that mammary tumors in the MMTV-N^ICtransgenics presented elevated Myc, its role in Notch transformation was examined in mice with floxed Myc alleles 43. Conditional ablation of Myc using the WAPCre transgene revealed that Myc was indispensable for development of N^IC-driven mammary tumors. This contribution of Myc to Notch-induced tumorigenesis is interesting especially when comparing it with Wnt pathway-driven models (see below). Moreover, Myc was shown to be a direct target of the Notch pathway, since a complex of N^IC and Cbf1 was detected on the Myc promoter. The Cbf1 binding site on the human Myc promoter is conserved, and immunohisto-chemistry revealed that there was a significant correlation between high Myc levels and N^IC in human breast tumors 43. It is intriguing that coexpression of Jagged ligand and Notch receptors is found in triple-negative breast tumors 44, a subgroup that also has high Myc activity 46.

Wnt1 was the first identified oncogene activated by MMTV insertional mutagenesis. Wnt-mediated activation of the canonical pathway leads to β-catenin stabilization, TCF binding and transcriptional activation of Myc. Mammary tumors arising in Wnt1 transgenics 47 and models driven by a β-catenin-stabilized mutant 15 show elevated levels of Myc. Human breast tumors, unlike colon cancer, do not possess Wnt pathway-activating mutations. Deregulation of Wnt signaling appears to occur by autocrine mechanisms, however, since multiple Wnt ligands and Frizzled receptors are coexpressed 48, and the negative Wnt pathway regulator - secreted Frizzled-related protein (sFRP1) - is often absent 49. A positive feedback loop has also been described for Myc and the Wnt pathway. In telomerase-immortalized, Myc-transformed human mammary epithelial cells, Myc was shown to repress sFRP1 and Dickkopf 1, another negative pathway regulator, thereby contributing to activation of canonical Wnt signaling 50. Along the same lines, Myc knockdown in MDA-MB-231 tumor cells increased Dickkopf 3 expression 40.

What is Myc's role in tumors induced by Wnt pathway activation? The dependency of Wnt-driven mouse mammary tumors on Myc expression has not been tested. In other tumor models driven by adenomatosis polyposis coli (APC) loss, the importance of Myc has been examined using organ-specific Cre recombinase-mediated deletion of floxed Myc alleles. In the intestines, Myc deletion reversed the tumor phenotype induced by APC loss, and it was shown that the majority of Wnt targets in the intestines were Myc dependent 51. In striking contrast, Myc deletion had no affect on the phenotype of APC loss in the liver, where most Wnt target genes were β-catenin dependent but Myc independent 52.

These two studies 5152 reveal that the importance of Myc in a particular tumor model is very specific and can differ as already discussed for the normal organ-specific Myc functions. As discussed above, deregulated Myc might enforce autocrine Wnt pathway activity in human tumors by repressing negative regulators such as sFRP1. Other potential roles for Myc have not been examined; however, blockade of the Wnt pathway usually results in lowering Myc levels. Stable expression of sFRP1 in MDA-MB-157 and MDA-MB-231 breast tumor cell lines blocks proliferation of both cell lines, and Myc RNA was decreased in the former 53 while Myc protein was lower in the latter 49. In a panel of breast cancer cell lines, siRNA-mediated knockdown of Dishevelled, an essential mediator of Wnt signaling, led to lower Myc and decreased proliferation in most cell lines 54.

In summary, the current data suggest that, in mouse and human mammary tumors with constitutive Wnt signaling, Myc levels are elevated and might have a role in transformation.

Amplification of ERBB2 leading to receptor overexpression is found in 20 to 25% of primary breast tumors. In these tumors, constitutive ErbB2 activation stimulates numerous intra-cellular signaling pathways including Ras/Erk and PI3K/Akt, both of which impact on Myc transcription and protein stability. The role of Myc has been examined in the ErbB2-overexpressing SKBr3 and BT-474 breast tumor cell lines. Treatment of both with the ErbB2-specific antibody trastuzu-mab caused a cell cycle block that was accompanied by a decrease in PI3K/Akt pathway activity, and by downregulation of Myc and D-type cyclins 55. Interestingly, ectopic expression of Myc in SKBr3 cells partially rescued the cells from functional ErbB2 inactivation 56, pointing to the importance of Myc as an ErbB2 effector.

In the normal breast, both rodent and human, a major role of the ERα-positive cells is to act as sensors to relay a proliferative signal to neighboring cells. In contrast, many breast tumor cells are ERα-positive, and they have not only acquired the potential to proliferate in response to steroid hormones but are also dependent upon them for survival 2157. Since Myc is an ERα target, it is important to understand whether Myc has a role in acquisition of this phenotype. Moreover, since patients whose tumors are ERα-positive are treated with anti-estrogen therapy, Myc's role in response or resistance is also of high interest. Unfortunately, consistent clinical data relating MYC amplification or expression levels to endocrine therapy response are not available 21. Overexpression of ErbB2 has been correlated with de novo and acquired endocrine resistance 57, however, and Myc is an ErbB2 effector.

The role of Myc in ERα signaling has been well characterized in the MCF-7 breast tumor cell line 21. Myc RNA increases rapidly in response to estrogen treatment of these cells, and Myc knockdown impairs the ability of estrogens to stimulate proliferation. Furthermore, Myc overexpression in cells arrested by ERα antagonists overcomes the proliferative block. Interestingly, adaptation of MCF-7 cells to growth in estrogen-deprived medium is associated with upregulation of ERα-regulated target genes, including Myc 58, suggesting a mechanism whereby Myc maintains an important role in proliferation and survival even in the absence of ERα activity. It should also be mentioned that ERα cross-regulates ErbB2, which in turn impacts on Myc via activation of downstream signaling pathways 57.

From transcriptome and network analyses, the close link between Myc and ERα signaling has become even more apparent. MCF-7 cells show a high level of overlap between ERα-regulated and Myc-regulated genes. Indeed, more than 50% of the estrogen-responsive genes are also Myc targets 59. Moreover, a meta-analysis of transcriptional and pathway data, carried out on primary breast tumors, revealed that Myc activity is elevated in ERα-negative, basal-like breast cancers, as measured by target gene levels 46.

In summary, these data led to the proposal that elevated Myc activity present in ERα-negative breast tumor cells mimics the activity of estrogen on ERα-positive breast tumor cells 4659.

Considering that Myc deregulation is so common in breast cancer, Myc has been examined as a prognostic factor and as a predictive factor. Indeed, MYC amplification is associated with aggressive clinical features including high grade and lymph node positivity, and correlates with poor patient outcome 34. A major interest in breast cancer is the use of genetic alterations to categorize patients into treatment groups; a good example being detection of the ERBB2 amplicon for trastuzumab treatment. Gene expression signatures are also being generated in order to provide predictive data on patient response to standard chemotherapeutics and to targeted therapy. In one study of chemotherapy-treated breast cancer patients, an attempt to correlate Myc pathway activity with response yielded inconsistent results. The group with Myc and Ras pathway activation had a high percentage of responders, while patients whose tumors had Myc and E2F pathway activity responded poorly 60. These results underscore the difficulty of using Myc levels alone as a predictive or prognostic factor, and emphasize the fact that the cellular context of Myc expression and activity determine the outcome.

Approaches to target Myc activity in tumors are also under consideration 3461. Although targeting Myc is appealing, there are many difficulties associated with altering transcription factor activity. Considering the variety of kinase inhibitors currently available, it is worth considering their use in breast tumors with Myc deregulation. Based on the molecular concepts analysis 62 of the Myc pathway activation signature in cancer, and on the identification of the genes downregulated in MCF-7 cells treated with wortmannin and LY-294002, there is reason to believe that PI3K inhibitors might be especially potent in breast cancers with high Myc activity. As discussed 46, ERα-negative, basal-like breast tumors with high Myc activity might be particularly susceptible to PI3K inhibitors.

Clearly, there are numerous signaling events in the development of the mammary gland that can be mediated, at least in part, via Myc. Using transgenic models, functions for Myc - some potential, others data based - have been discussed for embryonic development, pregnancy, lactation and involution. Furthermore, results from other model systems suggest that Myc could play a role in stem cell fate, as well as during early pregnancy, where Myc expression levels are highest in the normal gland. Even more efforts have been made to investigate Myc's role in transformation, as deregulation of Myc via amplification, overexpression or stabilization of the protein is a frequent event in breast cancer. Signaling pathways implicated in breast cancer, such as ERα, ErbB2, Notch and Wnt, all contribute to aberrant Myc levels or activity. The challenge for future studies will be to reveal the suitability of targeting Myc for breast cancer treatment, either by direct inhibition or by indirectly targeting another pathway.

Akt: serine/threonine kinase Akt; APC: adenomatosis polyposis coli; Cav1: Caveolin1; Cbf1: C promoter-binding factor 1; Cdk: cyclin-dependent kinase; ER: estrogen receptor; Erk: extracellular signal-related kinase; K14: Keratin14; MMTV: mouse mammary tumor virus; MTB/TOM: MMTV-rtTA/TetO-MYC; NF: nuclear factor; N^IC: Notch intracellular domain; Nrg3: Neuregulin3; PI3K: phosphoinositide 3-kinase; Ras: rat sarcoma; SC: stem cell; sFRP1: secreted Frizzled-related protein 1; Stat: signal transducer and activator of transcription; TCF: T-cell factor; TF: transcription factor.

The authors declare that they have no competing interests.

This article is part of a review series on Key signalling nodes in mammary gland development and cancer, edited by Adrian Lee and Charles Streuli.

Other articles in the series can be found online at: http://breast-cancer-research.com/series/bcr_signalling_nodes