Review
While the hormonal treatment of oestrogen receptor (ER)-positive tumours has been a major advance in the treatment of breast cancer, a significant percentage of hormone receptor positive breast cancers are inherently antioestrogen resistant, or become antioestrogen resistant during treatment [1,2]. An understanding of how antioestrogens act and how resistance may develop is an important goal of breast cancer research. To date aberrations in a number of signalling pathways have provided potential mechanisms for antioestrogen resistance. These include cross-talk between the ER and cell-surface tyrosine kinase receptors, as well as alterations in the MAP kinase and PI3-kinase pathways [1,2]. An important focus of study has understandably been at the point of convergence of these mitogenic signalling pathways, that is control of cell proliferation, where the transcription factor c-Myc is an important player [3].
Studies in clinical cohorts have failed to demonstrate a relationship between c-Myc amplification or c-Myc expression and overall outcome, including response to therapy, in breast cancer [4,5]. This may relate to the wide variability of amplification and expression levels identified across different cohorts as a consequence of variation in experimental techniques, contamination by non-tumour cells, and heterogeneous study populations, as well as an absence of data pertaining to patient follow-up, tumour characteristics and treatment [4,5]. Nonetheless there are significant in vitro data implicating c-Myc in breast cancer cell proliferation, and evidence suggesting that it may also be involved in the development of antioestrogen resistance. Regulation of c-Myc is an early response to either oestrogen or antioestrogen treatment of breast cancer cells [6]. Overexpression of c-Myc reverses or attenuates antioestrogen inhibition of breast cancer cell proliferation [7,8], and antisense oligonucleotide-mediated suppression of c-Myc expression leads to an inhibition of proliferation that is accompanied by molecular changes that mimic the effects of antioestrogens [9].
In their recent article, Mukherjee and Conrad provide further mechanistic insight into the potential role of c-Myc in resistance to antioestrogens in breast cancer cells in vitro [10]. Consistent with a previous study using MCF-7 cells [7], inducible expression of c-Myc led to re-initiation of cell cycle progression in antioestrogen-arrested cells, although this was less effective than oestrogen treatment [10]. Thus, there appears to be a link between c-Myc overexpression and resistance to antioestrogens. Overall, the kinetics of cell cycle progression in response to oestrogen treatment or c-Myc induction were similar. However, some differences were apparent, particularly at later time points in the analysis, which may relate to differences in the magnitude and/or time course of c-Myc induction. However, it is also possible that they resulted from effects of oestrogen treatment that were independent of c-Myc, for example induction of cyclin D1 and activation of cyclin D1-CDK4 [7].
Consistent with previous data in which c-Myc decreased the transcription of the CDK inhibitor p21WAF1/CIP1 (p21) in other cell types [11–13], Mukherjee and Conrad demonstrated that decreased p21 expression was a pronounced response to c-Myc induction in MCF-7 cells [10]. Other studies have shown that oestrogen also decreased p21 transcription during re-initiation of cell cycle progression in antioestrogen-arrested cells and this was necessary for oestrogen-mediated CDK activation [14–17]. In addition, decreased p21 expression conferred resistance to the anti-proliferative effects of antioestrogens [18,19] and increased p21 expression inhibited the ability of oestrogen to overcome antioestrogen-mediated arrest [20]. Similarly, when Mukherjee and Conrad overexpressed p21 using a recombinant adenovirus, c-Myc could not promote proliferation in antioestrogen-arrested cells [10]. Furthermore, RNA interference experiments showed that when c-Myc expression was decreased, p21 expression remained high in the presence of oestrogen [10]. Collectively, these observations suggest that the ability of oestrogen to decrease p21 transcription might be mediated by c-Myc.
Mukherjee and Conrad used luciferase reporter assays to show that a fragment of the p21 promoter beginning at −194 bp, encompassing a region previously shown to bind c-Myc [13], was repressed by c-Myc in MCF-7 cells [10], with the implication that oestrogen/antioestrogen-mediated regulation of c-Myc would lead to regulation of p21 via the same promoter element. Others have localised the sequence required for increased p21 promoter activity in response to antioestrogen treatment to a similar region, between −143 and +8 [21]. Mutation of individual Sp1 sites within this region substantially attenuated the induction, and further assays including chromatin immunoprecipitation suggested a model in which an ER-Sp1 complex binds to this region in cycling cells, recruiting histone deacetylases that repress p21 transcription [21]. Treatment with the antioestrogen ICI 182,780 led to disruption of this association and enhanced recruitment of Sp1 to the p21 promoter [21]. Gartel et al [12] demonstrated that c-Myc bound Sp1 but did not directly interact with the p21 promoter, suggesting that c-Myc may repress p21 by titrating Sp1 away from the p21 promoter. In hormone-responsive cells, ER and c-Myc may collectively modulate Sp1 regulation of p21 transcription. Thus although the experiments of Mukherjee and Conrad suggest that oestrogen/antioestrogen effects on p21 transcription are primarily mediated by c-Myc, the precise mechanism by which this occurs requires further study.
An increase in p21 levels when c-Myc is decreased has not been observed in another study using MCF-7 cells [9], perhaps due to differences in experimental design, whereby p21 expression was evaluated after 8–16 h of antisense c-Myc-oligonucleotide treatment as compared to 48 h of c-Myc RNA interference. However, despite the apparent differences between these studies, the overall conclusion is that regulation of p21 expression is central to the ability of c-Myc to modulate cell cycle progression in these cells, since overexpression of p21 using a recombinant adenovirus abrogated the ability of c-Myc to induce proliferation in antioestrogen-arrested cells [10], and decreased p21 expression allowed continued proliferation despite decreased c-Myc expression [9].
In a further set of experiments Mukherjee and Conrad investigated the potential role of c-Myc and p21 in resistance to antioestrogens in LCC9 breast cancer cells. These cells are an ER-positive, but oestrogen-independent and antioestrogen-resistant derivative of MCF-7 cells. They displayed no alteration in c-Myc or p21 expression after oestrogen or antioestrogen treatment [10]. When c-Myc was suppressed in LCC9 cells using RNA interference, p21 expression increased after antioestrogen treatment. It has been shown elsewhere that proliferation of antioestrogen-sensitive cells continues in the face of antioestrogen treatment if the increase in p21 is prevented using antisense oligonucleotides, or if c-Myc expression is maintained [8,18,19]. Collectively, these data suggest that the antioestrogen resistance observed in LCC9 cells may be due to altered p21 regulation by c-Myc.
The potential clinical importance of this article lies in its description of a mechanism through which c-Myc may mediate antioestrogen resistance. The questions that now arise include whether p21 modulation accounts for the majority of the antioestrogen resistance associated with c-Myc overexpression, and whether other upstream mediators are also involved. Although a previous study suggested that c-Myc induction resulted in cell cycle progression equivalent to the oestrogen response [7], Mukherjee and Conrad concluded that c-Myc induction alone was less effective than oestrogen treatment. Other investigators have demonstrated that c-Myc confers only partial antioestrogen resistance [8]. Thus the quantitative relationship between the level of c-Myc expression and oestrogen/antioestrogen sensitivity, and whether there is a level of c-Myc that can confer complete antioestrogen resistance, is unclear. An additional question relates to the degree to which c-Myc expression may influence antioestrogen resistance in the clinic. Further studies are required to dissect out other downstream targets of c-Myc with potential roles in steroid sensitivity, and upstream influences on p21 activity in order to direct research towards future therapeutic strategies for women with antioestrogen-resistant breast cancers.
Thus the important mechanistic study of Mukherjee and Conrad provides a framework for further testing these hypotheses in the laboratory and in tumour material from women with known therapeutic response to endocrine therapies.
Acknowledgments
Dr. McNeil is supported by the National Health and Medical Research Council, and the Cancer Institute of New South Wales. Associate Professor Musgrove is supported by the National Health and Medical Research Council, The Cancer Council NSW, the Association for International Cancer Research and is a Cancer Institute of New South Wales Fellow.
