Introduction
Progesterone receptor (PR) belongs to the superfamily of steroid receptors and mediates the action of progesterone in its target tissues [
1,
2]. In both humans and rodents, progesterone promotes the proliferation of epithelial cells that accompanies each menstrual/estrous cycle and pregnancy. In normal mammary glands of adult human and rodent females, PR expression is restricted to the luminal epithelial cells of the duct [
3]. Studies on PR-null mutant mice have revealed that PR is essential for progesterone-dependent proliferation of epithelial cells [
4].
PR exists in two isoforms, the A and B forms, and the expression of these, in an appropriate ratio, is critical for normal mammary development [
5]. As such, mammary development is abnormal in transgenic mice carrying either an additional A form of PR (PR-A transgenics) or the B form of PR (PR-B transgenics) [
6,
7]. In particular, mammary glands of PR-A transgenics are characterized by extensive lateral branching, ductal hyperplasia, a disorganized basement membrane (BM) and loss of cell-cell adhesion [
6]. Studies using the molecular markers for transformation, as defined by Medina [
8], revealed that these mammary glands contained at least two distinct populations of transformed epithelial cells. The ducts with normal histology contain cells resembling immortalized cells, while hyperplasias consist of cells in later stages of transformation associated with early preneoplasias [
9].
The development of cancer is also associated with disruption of tissue architecture. Branching morphogenesis in the mammary gland is the culmination of hormone-mediated proliferation and extracellular matrix (ECM) remodeling; these are each in turn dependent on the production of growth factors and the balance between ECM production and degradation [
10]. Once established, the mammary gland undergoes rounds of highly orchestrated proliferation and morphogenesis during pregnancy and involution, yet without losing the fundamental patterning of the gland. In contrast, hyperplasia is defined as loss of this patterning and is considered to be a precursor to neoplasia.
It is well established that PR-A can modulate the activities of both estrogen receptor (ER) alpha and PR [
11,
12]. Accordingly, either estrogen action or progesterone action or both, resulting from overexpression of PR-A, may mediate the abnormal mammary phenotype of PR-A transgenics. To this end, the objective of our present study was to identify the respective roles of estrogen and progesterone in the genesis of mammary hyperplasias/preneoplasias in PR-A transgenic mice.
Materials and methods
Mice, treatment with steroids and tissue preparations
All mice used in these studies were of FVB strain. The mice were housed and cared for in accordance with the National Institutes of Health guide to humane use of animals in research. All experiments were conducted with Lawrence Berkeley National Laboratory institutional review and approval. Animals were killed by carbon dioxide inhalation and cervical dislocation at the indicated times in accordance with Association for Assessment and Accreditation of Laboratory Animal Care guidelines.
PR-A transgenic mice [
6] and PR null mutant mice [
4] have been described previously. Nulliparous adult (9 to 12 weeks old) mice were used either intact or ovariectomized and/or treated with steroids for the indicated times. For zymography studies, prior to tissue collection, mice were perfused with 10 ml PBS and the tissues were frozen in liquid nitrogen and stored at -70°C until use.
For studies with steroids, nulliparous adult mice were used either as intact or after ovariectomy. Estradiol (1 μg/mouse) and/or progesterone (1 mg/mouse) were administered as described previously [
3]. Antiestrogen, ICI 182,780 (50 μg/mouse; Tocris Cookson Inc., Ellisville, MO, USA) or antiprogestin, mifepristone (RU486, 16 μg/g body weight; Sigma, St Louis, MO, USA) or ZK 98,299 (16 μg/g body weight; gift from Dr Ming-Wei Wang) were administrated daily for 4 days. For cell proliferation studies, mice were administered 160 μg/g body weight 5-bromo-2-deoxyuridine (BrdU) (Sigma) 2 hours prior to sacrifice.
For mammary whole mounts, one of the Number 4 inguinal mammary glands was fixed in Carnoy's solution and stained in Alum Carmine [0.2% carmine/0.5% aluminum potassium sulfate (both from Sigma, St Louis, MO, USA)]. For immunohistochemical analyses on paraffin sections, mammary tissues were collected, fixed in 4.7% buffered formalin (Fisher Scientific, Pittsburgh, PA, USA), dehydrated, embedded in paraffin and cut into 5 μm thick sections. For immunofluorescence analyses on frozen sections, the entire Number 4 inguinal mammary glands were mounted and quick-frozen in Optimal Cutting Temperature compound (Ted Pella, Inc., Redding, CA, USA). Cryostat sections (5 to 10 μm thick) were cut and mounted onto glass slides and were fixed for 2 minutes in methanol/acetone (1:1).
Antibodies
The antibodies used were: anti-BrdU, rat monoclonal antibody (Harlan Sera-Lab Ltd, Loughborough, UK); goat anti-latency-associated peptide (anti-LAP) and chicken anti-transforming growth factor beta 1 (anti-TGFβ1) (R&D Systems, Minneapolis, MN, USA); MMP-2 (Millipore, Bilerica, MA, USA); collagen I, collagen III and collagen IV (Southern Biotechnology Assoc. Inc., Birmingham, AL, USA); laminin-1 (Telios Pharmaceuticals, San Diego, CA, USA); and laminin-5 (gift from Dr V Quaranta).
Immunohistochemistry and immunofluorescence
BrdU-positive cells were analyzed in paraffin-embedded sections as described previously [
9]. Briefly, paraffin sections were deparaffinized and rehydrated prior to antigen retrieval and treatment with specified antibodies. The antigen-antibody complexes were identified using the Universal DAKO LSAB2-labeled streptavidin-biotin peroxidase kit (DAKO, Carpinteria, CA, USA). The sections were counterstained with Mayer's hematoxylin solution (DAKO). After counterstaining, nuclei negative for the antigen appeared purple-blue and positive nuclei appeared brown.
Immunofluorescence assays were performed as described previously [
3]. Briefly, nonspecific binding sites were blocked by incubation in a blocking buffer (PBS containing either 1% casein or 15% FCS and 0.2% Tween 20), for 30 minutes at ambient temperature. Sections were then treated with primary antibodies, washed with PBS and incubated with fluorescein isothiocyanate-conjugated secondary antibodies (dissolved in the blocking buffer) for 30 minutes at ambient temperature. Slides were then washed five times with PBS and the nuclei were stained with 4',6-diamino-2-phenylindole and mounted with Vectashield (Vector Laboratories, Burlingame, CA, USA).
Image acquisition and processing
Immunofluorescence images were obtained using a 40 × 0.75 numerical aperture Zeiss Neofluor objective on a Zeiss Axiovert equipped with epifluorescence (Carl Zeiss MicroImaging, Inc., Thornwood, NY, USA). A multiband-pass dichroic mirror, barrier filter and differential wavelength filter wheel combination was used to selectively excite fluorochromes in sequence. Images were captured using a scientific-grade 12-bit charged coupled device (KAF-1400, 1,317 × 1,035, 6.8 μm square pixels) on a Xillix digital camera (Vancouver, Canada). Relative intensity of images was maintained when constructing figures using Scilimage (TNO Institute of Applied Physics, Delft, the Netherlands) to scale the 12-bit data to a common 8-bit scale using the dataset minimum and maximum. False color was assigned accordingly. Internal standardization was achieved by comparing only images stained with the same antibodies in the same experiment, captured with identical parameters and scaled and displayed identically.
Zymography
Mammary glands were homogenized on ice in RIPA insoluble buffer (50 mM Tris, pH 8.0, containing 150 mM NaCl, 0.1% sodium dodecyl sulfate, 0.5% deoxycholate and 1% NP40). The homogenates were centrifuged for 15 minutes at 4°C and the supernatants were subjected to electrophoresis on gelatin substrate gels (8.8% sodium dodecyl sulfate-polyacrylamide slab gels containing 1 mg/ml gelatin), as described previously [
13]. Subsequently, the gels were treated with 2.5% Triton X-100 for 30 minutes, followed by incubation for 24 hours at 37°C in a buffer containing 100 mM Tris- HCl, pH 7.4, and 15 mM CaCl
2. The gels were stained with Coomassie Blue R-250 and destained with water until clear zones indicative of proteolytic activity emerged against a blue background. The zymograms were scanned and subjected to densitometric analyses using the PC version of NIH image (Scion Corporation, Frederick, MD, USA).
Statistical analysis
The data are presented as the mean ± standard error of the mean. The differences between the various experimental groups were analyzed by Student's t test and were considered significant when P < 0.05 was obtained. A Kolmogorov-Smirnov comparison of two datasets was used to compare the TGFβ1 and LAP data, with P < 0.05 considered significant.
Discussion
We have previously shown that mammary gland development in PR-A transgenic mice is abnormal, characterized by extensive ductal growth, lateral branching and loss of basement membrane integrity and cell-cell adhesion [
6]. In the present report we show that the proliferative phenotype of PR-A mice is dependent on ovarian steroids, where hyperplasia is mediated by distinct mechanisms. PR-A transgenic expression augments ER-α mediated proliferation while PR-A the degradation of the basement membrane, both essential requirements for the development of hyperplasias. Moreover, even though progesterone is required for the manifestation of the hyperplasias, antiprogestins have no effect on the established hyperplastic phenotype.
Signaling through estradiol/ER is required for proliferation, but is not enough to give rise to overt hyperplasias. Once established, these hyperplasias are dependent on ER signaling, since treatment with the antiestrogen ICI 182,780 abolishes them; we have not determined, however, the relative role of ERα and ERβ. On the other hand, the overexpression of PR-A leads to antiprogestin resistance, since neither ZK 98,299 nor RU486 had any effect on the expression of the hyperplasia in the intact mice. Both antiprogestins have similar mechanisms of action inducing PR binding to the DNA, but causing conformational changes that block co-activator binding to the receptor, thus making it transcriptionally inactive [
24]. The antagonistic activity in both cases, however, has been shown to depend on the cellular context that determines the degree of the antagonist/agonist activity of these compounds, which relies on the balance between expression and availability of coactivators and corepressors and the context of specific target promoters available in any given cell type [
24]. Furthermore it is clear from the ovariectomized mice that progesterone is necessary to induce the full phenotype. Overexpression of PR-A in the mouse mammary gland therefore seems to generate a cellular context that promotes hyperplasia, but whose maintenance is not dependent on progesterone signaling
per se.
The hyperproliferation observed in the mammary glands of PR-A transgenic mice may be due to the diminished expression found for both latent and active TGFβ1. Activation of TGFβ1 is regulated by estradiol and progesterone, which in turn restricts the proliferative response to these steroids [
14]. Our studies suggest that physiological estradiol drives inhibition of TGFβ1 that in turn increases proliferation in response to estradiol, which is consistent with its action in estrus shown using TGFβ1 heterozygote mice [
17]. Notably, the mammary glands of
Tgfb1 heterozygote mice are two to four times more proliferative yet there is no evidence of hyperplasia [
17]. While similar to
Tgfb1 with regards to decreased TGFβ1 levels and increased proliferative response to steroid hormones, an additional action of PR-A must therefore be necessary for the expression of hyperplasia.
Since our studies show that signaling through progesterone has no impact on proliferation in hyperplastic PR-A mice but is required for the manifestation of the hyperplasias, we postulated that another effect must be required. We show that signaling through PR modulates the activation of MMP-2, and that areas of BM disruption coincide with loss of latent MMP-2. The integrity of BMs is dictated by the composition of MMPs, their inhibitors and the status of their activation [
25]. MMPs have been shown to be key regulators of branching morphogenesis in the mouse mammary gland [
20,
26,
27]. On the other hand it has long been known that both estrogen and progesterone affect mammary gland development regulating ductal elongation, side branching and differentiation. The development of PR and ER transgenic and knockout mice in the past decade has provided additional evidence to confirm these results [
4,
6,
28]. To date, however, there is insufficient evidence to address the intriguing question of how hormonal action, MMP production and activation are linked in the mouse mammary gland.
We have previously shown that there was loss of laminin-1 in areas of aberrant epithelial morphology in PR-A transgenic mice [
6]. We now report that, in addition to laminin-1, immunoreactivity of laminin-5, collagen IV and collagen III is also lost. In the mammary glands of PR-A transgenics, therefore, the concurrent loss of laminin and collagen indicate that the ECM is broadly affected. More importantly the fact that the phenotype was reversible by ovariectomy points to a role for hormonal action in the maintenance of ECM integrity, which is supported by our observation that the administration of estradiol and progesterone to ovariectomized PR-A transgenic mice led to BM degradation, strongly supporting this notion.
Alterations in the mammary gland ECM have been previously described in transgenic mice expressing an active form of MMP-3 in the epithelium [
21]. The fact that total MMP-2 levels in extracts derived from PR-A transgenic mice were not significantly different could be due to the heterogeneous phenotype, as previously shown [
6], and is consistent with the statistically significant decrease we observed in MMP-2 levels only in PR null mutant mice. Moreover, immunolocalization studies support the notion of a restricted phenotype as we found decreased latent MMP-2 only in hyperplastic regions where laminin-5 was absent.
The finding that treatment of ovariectomized wild-type mice with estradiol and progesterone leads to the activation of MMP-2 is the first direct demonstration that hormonal action modulates MMP activity in the mouse mammary gland. This result suggests an important role for MMP-2 during proliferation and differentiation that takes place during pregnancy. It is significant to note that while in wild-type ovariectomized mice progesterone alone had no effect on MMP-2 activation (because there is little or no PR-A), it caused an increase in the activation of MMP-2 in mammary glands of ovariectomized PR-A transgenics because they still expressed a high level of PR-A. In wild-type mice, ER would therefore be necessary to support the levels of PR, but signaling through PR-A would be important for the activation of MMP-2.
Using MMP-2 knockout mice, Wiseman and collaborators have shown that during puberty MMP-2 positively regulates invasion of terminal end buds but that it inhibits side branching from mature ducts. They also postulated TGFβ1 as the candidate for the inhibitory action of MMP-2 on lateral branching at puberty [
26]. This speculation is supported by recent modeling studies that demonstrate the effect of TGFβ1 as a branching inhibitor [
29]. In mammary glands of PR-A transgenics there is a positive relationship between the steroid hormone-dependent increase in MMP-2 activity and lateral branching, accompanied by a reduction in both the latent and active forms of TGFβ1. It is possible that other proteases are also involved, however, given that, although statistically significant, the degree of activation of MMP-2 was modest both in the wild-type and in the PR-A transgenics.
Given that MMP-2 is expressed by the periductal stroma and weakly by the adipose tissue [
26], and that PR is expressed exclusively in the epithelial compartment [
30], cross-talk between both compartments seems to be a necessary condition for MMP-2 activation. A possible mediator of this cross-talk may be MT1-MMP, which is expressed both by the stroma and the epithelium [
26]. This hypothesis is supported by the fact that we have detected an increase in the levels of MT1-MMP in the mammary glands of pregnant mice, and that MT1-MMP levels were high in certain areas in mammary glands of PR-A transgenic mice (Simian M and Shyamala G, unpublished observations). Additionally, degradation of collagen IV appears to require the action of MT1-MMP [
31]. Whether PR-positive cells are also MT1-MMP-positive remains to be determined.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MS performed the zymography studies, prepared the results and contributed to the drafting of the manuscript. MJB participated in the conception of the study, the direction of the research and the correction of the manuscript. MHB-H participated in the conception of the study, carried out the TGFβ1 studies and the immunofluorescence of the BM components, and corrected the manuscript. GS conceived the study, directed the research, analyzed the data, kept the transgenic mice, carried out the whole mounts and BrdU studies, and drafted and corrected the manuscript.