Background
In spite of recent advances in diagnosis and treatment, breast cancer remains the second leading cause of cancer-related death in women in the United States. The existence of multiple subtypes of breast cancer, each with unique clinical and/or molecular characteristics, is now well established [
1,
2]. Multiple genetic and environmental factors contribute to breast cancer development, and it is becoming increasingly clear that development of each breast cancer subtype is influenced by different sets of factors. Known risk factors include a family history of breast cancer, cumulative exposure to endogenous and exogenous estrogens and breast mammographic density [
3‐
9]. Although several genes have been identified that significantly impact breast cancer risk when mutated or aberrantly expressed, only a small fraction of the overall population risk can be attributed to these genes [
10‐
12]. Similarly, the genetic determinants of responsiveness to estrogens and mammographic density remain poorly defined.
We are using inbred ACI (August x Copenhagen, Irish), COP (Copenhagen) and BN (Brown Norway) rats to define the mechanisms through which estrogens contribute to mammary cancer development and identify genetic determinants of susceptibility to mammary cancer. When treated continuously with 17β-estradiol (E2), female ACI rats develop mammary carcinoma at an incidence approaching 100% [
13]. The mammary cancers that develop in E2 treated ACI rats express estrogen receptor-α (ERα) and progesterone receptor (Pgr), are dependent upon E2 for continued growth and survival, and frequently exhibit chromosome copy number changes and instability [
14‐
16]. Development of mammary cancer in E2 treated ACI rats is dramatically inhibited by concurrent treatment with tamoxifen, indicating a requirement for one or more estrogen receptor mediated mechanisms in tumor development [
17,
18]. Interestingly, tumor development in ACI rats also requires the action of progesterone [
13,
19]. By contrast, COP and BN rats are resistant to E2-induced mammary cancer [
20‐
22]. Multiple genetic determinants of susceptibility to E2-induced mammary cancer, designated
Emca1 (
E strogen-induced m ammary ca ncer 1) through
Emca9, have been mapped in crosses between susceptible ACI rats and resistant COP or BN rats [
21‐
24]. Each of the mapped quantitative trait loci (QTL) encompass segments of the rat genome that are orthologous to regions of the human genome linked to breast cancer risk in genome wide association studies (GWAS). Together, these data indicate that the ACI rat model of E2-induced mammary cancer is a physiologically relevant model for studying the molecular etiology of luminal type breast cancers.
The purpose of this study was to define, both qualitatively and quantitatively, the manner in which the mammary glands of susceptible ACI and resistant BN rats respond to E2. Dramatic differences in multiple cellular and molecular responses to E2 were observed when these two inbred rat strains were compared. These differences contributed to and/or were associated with differences in epithelial density, mammary gland differentiation and ECM, as well as differential expression of many genes of known significance to mammary gland development. We propose that the observed differences in responsiveness of the mammary gland to E2 represent phenotypes that underlie the documented strain differences in susceptibility to mammary cancer and may also contribute to and/or serve as biomarkers of breast cancer risk in humans.
Discussion
Data presented herein demonstrate that the mammary glands of ACI and BN rats exhibited marked quantitative and qualitative differences in their cellular and molecular responses to E2. The primary response exhibited by ACI rats, which are uniquely susceptible to mammary cancer when treated with estrogens, was a robust and sustained proliferation within the mammary epithelium. By contrast, the proliferative response of the mammary epithelium of BN rats, which are highly resistant to estrogen-induced mammary cancer, was restrained and transient. This difference in induction of cell proliferation, not a difference in apoptosis, appeared to be largely responsible for the quantitative differences in epithelial density observed when the mammary glands of ACI and BN rats were compared following 1, 3 and 12 weeks of E2 treatment. Moreover, the mammary glands of E2 treated BN rats, but not ACI rats, exhibited qualitative phenotypes consistent with differentiation to secretory epithelium, as well as luminal ectasia and associated changes in collagenous stroma. These differences in the responsiveness of the mammary glands of ACI and BN rats to E2 were apparent within one week of initiation of treatment, strongly suggesting that the molecular mechanisms responsible for the rat strain specific responses may be inherent within the mammary glands of these inbred rat strains.
Comparison of gene expression profiles for mammary glands of E2 treated ACI and BN rats revealed differential expression of multiple genes that may have contributed to the differences in luminal epithelial cell proliferation and lobuloalveolar hyperplasia observed upon comparison of these rat strains.
Pgr,
Wnt4,
Tnfsf11 (
RankL),
Prlr,
Stat5a,
Areg and
Gata3 were expressed at higher levels in mammary glands of E2 treated ACI rats, relative to identically treated BN rats. The protein products encoded by these genes play well defined important roles in mammary gland development. Expression of
Pgr in mammary epithelium is induced by E2 and progesterone, acting through Pgr, plays a requisite role in stimulating lobuloalveolar development during pregnancy [
30‐
32]. Moreover, studies summarized above have demonstrated a requisite role for progesterone in the induction of mammary cancer development by E2 in ACI rats [
13,
19]. Both Wnt4 and RankL have been demonstrated to function downstream of Pgr in stimulating lobuloalveolar development and have more recently been shown to be requisite paracrine mediators of the actions of progesterone in the regulation of mammary stem cell number [
33‐
37]. Prlr and Stat5a are both required for induction of lobuloalveolar development by prolactin, a second major hormonal regulator of lobulogenesis during pregnancy [
38,
39]. Areg functions as an important paracrine mediator of the actions of estrogens and ERα on induction of mitogenesis in the mammary epithelium [
40,
41]. Finally, Gata3 is required for elongation of mammary ducts at puberty and maintenance of differentiated luminal epithelium, and also acts as a positive regulator of expression of
Esr1, the gene encoding ERα [
42,
43]. Additional studies are needed to establish whether differential expression of these genes is the cause or the consequence of the observed differences in epithelial cell proliferation and lobuloalveolar hyperplasia exhibited by E2 treated ACI and BN rats.
Other differentially expressed genes encode protein products that are functionally associated with mammary gland differentiation, lactation and/or post-lactational involution.
Spp1 and
Lcn2 are among those genes that were most highly expressed at the mRNA level in mammary glands of E2 treated BN rats, relative to identically treated ACI rats.
Spp1 encodes a secreted phosphoprotein that is highly expressed in the mammary gland during lactation and involution [
44‐
46].
Spp1 has also been demonstrated to be more highly expressed in mammary glands of parous mice and rats, compared to nulliparous controls [
47,
48]. Inhibition of
Spp1 expression in the luminal epithelium of the mouse mammary gland inhibits lobuloalveolar development, expression of genes encoding milk proteins and milk production [
49]. Moreover,
Spp1 underlies a quantitative trait locus (QTL) in dairy cattle that controls milk yield and protein content [
45]. Together, these data suggest that Spp1 regulates multiple processes in the mammary epithelium during pregnancy, lactation and/or mammary gland involution.
Lcn2 encodes a secreted glycoprotein that is highly expressed within the luminal epithelium of the mammary gland during pregnancy and lactation as well as during mammary gland involution [
50,
51]. Lcn2 is known to bind a diverse group of ligands, including retinoids, fatty acids, bacterial siderophores and specific MMPs. Suggested functions of Lcn2 in the mammary gland include modulation of inflammation and immunity, ECM remodeling and regulation of iron homeostasis. The functional significance of differential expression of
Spp1 and
Lcn2 in the mammary glands of ACI and BN rats remains under investigation.
As noted above, luminal ectasia and associated collagenous stroma were qualitative phenotypes unique to the mammary glands of E2 treated BN rats. Several genes that encode proteins that are known to modify the extracellular microenvironment were observed to be differentially expressed between E2 treated ACI and BN rats. Two examples are
Mmp7 and
Mmp9, both of which were expressed and activated to a greater degree in the mammary glands of BN rats, relative to ACI rats. Known functions of these MMPs include ECM remodeling and production of active forms of multiple growth factors, cytokines and chemokines [
52].
Mmp7 is unique among the MMPs in that its expression in the mammary gland is largely restricted to the glandular epithelium [
53]. A role for Mmp7 in mammary gland development is suggested by a study that demonstrated that expression of MMP7 in the mammary epithelium of nulliparous mice under control of the mouse mammary tumor virus (MMTV) promoter induced production of milk proteins, suggesting MMP7 may play a role in mammary gland differentiation [
54]. Other studies support a role of Mmp7 in mammary cancer development and/or progression. For example, expression of an
MMTV-MMP7 transgene in the mammary epithelium resulted in development of hyperplastic alveolar nodules in a large fraction of aged multiparous mice and shortened the time to onset of mammary tumors in mice that also expressed an
MMTV-Neu transgene [
53]. Moreover, single nucleotide polymorphisms in
MMP7 have been associated with disease free and/or overall survival in two breast cancer case control studies [
55,
56]. Mmp9 is expressed by the mammary epithelium, stromal fibroblasts and infiltrating immune/inflammatory cells. The highest levels of Mmp9 expression in the mammary gland occur during pregnancy and involution [
57,
58]. However, the roles of Mmp9 at these developmental stages are not well defined. Mmp9 contributes to mammary cancer metastasis in mouse models and nucleotide variants within
MMP9 have been associated with breast cancer metastasis in humans [
59,
60]. Interestingly, Mmp9 has been demonstrated to form a binary complex with Lcn2, leading to activation and stabilization of this matrix metalloproteinase [
61‐
64]. These data suggest a potential mechanism for the enhanced activation of Mmp9 observed in the mammary glands of E2 treated BN rats.
Comparison of gene expression profiles for mammary glands from E2 treated ACI and BN rats also revealed differential expression of many genes that encode proteins that reside on the cell surface and function in cell-cell or cell-ECM interactions. One such gene,
Cd44, was observed to be expressed at an approximate 10-fold higher level in BN rats than in ACI rats. Cd44 is expressed by the myoepithelium in developing mammary gland and by luminal epithelium in adult mouse mammary gland and human breast [
65].
Cd44 null mice exhibit a lactation defect which appears to result from reduced activation of heparin binding epidermal growth factor (HB-EGF) and downstream signaling through ErbB4 [
66].
Cd44 null mice also exhibit delayed ductal outgrowth and small TEBs and these phenotypes were attributed to aberrant interactions between myoepithelium and luminal epithelium [
65]. Multiple studies have demonstrated a physical interaction between CD44 and Spp1 in a wide variety of cell types, including breast cancer cells, which alters an array of cellular phenotypes including motility and invasiveness [
67‐
70]. CD44 has also been demonstrated to interact physically and functionally with Mmp7 and Mmp9 in multiple cell types, and by doing so enhances the activities of Mmp7 and Mmp9 on specific substrates within the extracellular environment [
66,
71‐
73]. Interestingly, the interaction between CD44 and Mmp9 in PC3 prostate cancer cells has been demonstrated to be induced by Spp1 [
74]. CD24 and CD52 were observed to be expressed at higher levels in mammary glands from E2 treated ACI rats, relative to BN rats.
CD24 encodes a cell surface glycoprotein that has emerged as a marker for mammary stem cells [
75,
76]. In the mouse mammary gland, Cd24 is expressed in the luminal epithelium and to a lesser extent in the basal epithelium [
77,
78]. Mice that are homozygous for a
Cd24 null allele exhibit accelerated ductal elongation and increased branching morphogenesis in the mammary gland [
78].
CD52, which is paralagous to
CD24, is expressed by lymphocytes and other types of immune cells. Virtually nothing is known regarding the role of CD52 in mammary gland development or function. Ongoing studies are focused on identifying and quantifying the cell types in the mammary glands of ACI and BN rats that express these different proteins.
We hypothesize that variation in a subset of the cellular and molecular phenotypes described herein is heritable and underlies the differing susceptibilities of the ACI and BN rats to E2-induced mammary cancer. We are currently testing this hypothesis by evaluating these phenotypes in a panel of unique congenic rat strains that were developed to characterize the QTL that were identified as genetic determinants of susceptibility to E2-induced mammary cancer in intercrosses between susceptible ACI and resistant BN rats [
22,
23]. Our working model is that genetic variants within the
Emca QTL impact expression of genes that function downstream of E2 and progesterone to control proliferation, survival and/or differentiation within the mammary epithelium and/or the cellular composition of the stroma and thereby influence susceptibility to E2-induced mammary cancer. Supporting this model is a recently published study in which it was demonstrated that congenic rats that harbor, on the ACI genetic background, BN alleles across the
Emca8 locus on rat chromosome 5 exhibited significantly reduced susceptibility to E2-induced mammary cancer that was accompanied by reduced expression in the mammary gland of
Pgr,
Wnt4 and
Cd52 and increased expression of
Spp1, relative to E2 treated ACI rats [
24]. We further hypothesize that variation in the different cellular and molecular phenotypes observed in E2 treated ACI and BN rats is representative of variation that would exist within the genetically heterogeneous human population. For example, the difference in mammary epithelial density exhibited by E2 treated ACI and BN rat may be analogous to variation in breast mammographic density in humans, which is known to be modified by estrogens as well as other hormonal, genetic and environmental factors and has been strongly associated with breast cancer risk. Additional studies are required to establish cause and effect relationships between the cellular, molecular and mammary cancer susceptibility phenotypes in the rat and to translate the knowledge gained to humans.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Conception and design, LD, JDS; development of methodology, LD, JDS; acquisition of data, LD, YZ, CLW, RS, KWE, JDS; analysis and interpretation of data, LD, YZ, CLW, RS, KWE, JDS; writing of the manuscript: LD, RS, JDS; study supervision: JDS. All authors read and approved the final manuscript.