What hormone regulates mammary gland growth?

• PDF
• Split View
• • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data

MAMMARY gland development is under multihormonal control involving functional interplay between the ovary and pituitary (1–4). Classical endocrine ablation/hormone replacement studies demonstrated that ovarian estradiol is critical to the two major phases of mammary development; ductal elongation during puberty and lobuloalveolar development during pregnancy. However, the effects of estrogens on mammary growth appeared to require a functional pituitary gland (5, 6). Therefore, estrogens may contribute to mammary development by acting directly on the mammary gland and/or by indirect endocrine action on the hypothalamic/pituitary/gonadal axis.

Estradiol has been shown to act directly on the mammary gland to stimulate ductal morphogenesis during puberty (7, 8), whereas progesterone is the major stimulator of mammary epithelial DNA synthesis and alveolar development in 10-week-old mice (9). However, estrogen priming contributes to alveolar growth by inducing progesterone receptor (PR) and PRL receptor levels in the ductal epithelium (9–14). Therefore, the action of estradiol directly on the mammary gland involves ERα-mediated effects on the regulation of downstream mammogenic hormonal signaling systems.

The pituitary gland, which is responsive to estradiol, also contributes to mammary development. Estradiol is known to enhance PRL synthesis and secretion from the pituitary gland in two ways (2, 15). First, estradiol can induce PRL gene expression via a functional estrogen response element in the promoter of the PRL gene (16, 17). Second, estradiol can prevent the release of hypothalamic dopamine, an inhibitor of pituitary PRL synthesis and secretion (15). As a result, serum PRL levels are highest on the day of proestrus in rodents when estradiol levels are peaking (18–20). Pituitary-grafted mice with intact ovaries possessed highly elevated blood PRL levels and their mammary glands displayed terminal end bud (TEB) formation and alveolar development (21). These findings suggested that either elevated PRL alone was sufficient for mammary development or ovarian hormones (i.e. estradiol) in combination with PRL were required for the growth response.

The generation of gene knockout mice has helped to resolve the functions of various hormonal signaling systems involved in mammary development. Mice with disruption of genes encoding the PR (22), PRL receptor (23), PRL (24), and their downstream effectors, Stat5A (25), and cyclin D1 (26, 27), all possess a complete mammary epithelial ductal network but do not develop alveolar structures. The mammary development in these knockout mice indicated that the progesterone and PRL signaling systems are not required for ductal morphogenesis, but alveolar development appears to require the action of these mammogenic hormones.

Mammary glands from the ERα knockout (αERKO) mouse were undeveloped possessing only a rudimentary ductal structure that emanated from the nipple (28, 29). In contrast, ERβ knockout (βERKO) mice appear to undergo normal mammary development (30). Therefore, the morphology of the αERKO mammary gland in conjunction with the aforementioned knockouts indicated that estrogen/ERα signaling is required for ductal morphogenesis. ERα-mediated ductal morphogenesis and alveolar development may involve induction of estrogen-responsive genes within the mammary gland and in peripheral endocrine tissues that contribute to mammary gland development and function. Therefore, ERα gene disruption may result in reduced pituitary PRL secretion and absent mammary PR gene induction; both of which may contribute to the αERKO mammary phenotype (29). In support of deficient mammogenic signaling in the αERKO, we have previously reported that progesterone circulates at low basal concentrations in the αERKO female (31), and PR messenger RNA (mRNA) levels are markedly reduced and not inducible by estradiol in the αERKO mammary gland (29). In addition, PRL mRNA is reduced 15-fold in the αERKO pituitary (32).

In the present report, to determine if PRL signaling was also compromised in the αERKO, serum PRL was measured in intact wild-type (WT) and αERKO mice, and in ovariectomized (OVX) mice treated with estradiol. To determine whether the αERKO mammary gland was capable of undergoing ductal morphogenesis and/or develop alveolar structures in response to PRL, a single heterozygous ERα pituitary gland was implanted under the renal capsule of OVX or intact WT andα ERKO mice. Finally, the specific hormonal requirements for mammary gland growth in αERKO and WT mice were defined by implanting pellets of PRL, progesterone and estradiol individually or in various combinations and analyzing the mammary growth response.

Materials and Methods

Surgical manipulations

All mice for this study were generated by breeding male and female mice, heterozygous for the ERα gene (HET), to produce offspring with all three genotypes. The mice were housed and cared for in accordance with the NIH Guide to Humane Use of Animals in Research and all surgical procedures were approved by the NIEHS Animal Care and Use Committee. In experiments outlined below, a pituitary gland from age-matched HET mice was grafted under the renal capsule of host WT andα ERKO mice. Hormone pellets purchased from Innovative Research of America (IRA) were implanted sc above the shoulder region in host mice. Ovariectomy was performed at the time of pituitary grafting or pellet implantation.

(A) WT and αERKO mice (n = 5 per group, 12 weeks of age) were OVX and a placebo or estradiol pellet (0.05 mg) was implanted (sc). These mice were killed 21 days later and blood was collected for measurement of serum PRL and estradiol. According to IRA, the expected release of estradiol was 50–100 pg/ml. The level of serum estradiol in the ERKO mice at the time they were killed was 80[mean]± 8[sem] pg/ml.

(B) WT and αERKO females (n = 6 per group, 25 days and 12 weeks of age) were OVX or left intact and a single pituitary from an age-matched HET female was grafted under the renal capsule of the host mice. The host mice were housed for 45 days and then killed. Blood was collected for PRL and progesterone measurement, and mammary glands excised for analysis.

(C) WT and αERKO females were OVX and received a single age-matched HET pituitary graft as described above. Pituitary-grafted mice of both genotypes (n = 4 per group, 12 weeks of age) then received a single pellet of estradiol (0.05 mg) or placebo implanted (sc) as described above. After 21 days, a second placebo or estradiol pellet was implanted (sc) and then the host mice were killed 21 days later (day 42). The expected release of estradiol over 42 days was 50–100 pg/ml according to IRA. Blood was collected for PRL measurement and mammary glands excised for analysis.

A group of pituitary-grafted αERKO mice (n = 3, 12 weeks of age) were implanted (sc) with SILASTIC brand silicon tubing (id 0.062 inches and od 0.125 inches, Dow Corning Corp.) filled with 5 mg crystalline DHT (Steraloids) (33), and killed 42 days later. The expected sustained release of DHT into the circulation was approximately 2 ng/ml. Blood was collected for PRL measurement and mammary glands excised for analysis.

Pituitary-grafted WT and αERKO mice (n = 4 per group, 12 weeks of age) received a 35 mg progesterone pellet (sc) 21 days after pituitary grafting, and the mice were killed 21 days later (on day 42). The expected release of progesterone over the 21 days was approximately 50 ng/ml. Mammary glands were excised for analysis.

(D) WT and αERKO mice (n = 4 per group, 12 week of age) were OVX and implanted (sc) with pellets of estradiol (2 × 0.1 mg), progesterone (1 × 35 mg), or rat PRL (2 × 5 mg) individually or in various combinations (see Results and Table 1). According to IRA, the expected sustained release of the various hormones over 21 days is as follows: approximately 250 pg/ml estradiol (2 × 0.1 mg pellets); approximately 50 ng/ml progesterone (1 × 35 mg pellet); approximately 50 ng/ml PRL (2 × 5 mg pellets). The actual serum hormone levels measured after 21 days are expressed as the mean ± sem (n = 4 for each individual hormone treatment). Serum hormone levels in WT mice are as follows: estradiol (662 ± 141 pg/ml); progesterone (40 ± 6 ng/ml); PRL (7± 2 ng/ml). Serum hormone levels in αERKO mice are as follows: estradiol (938 ± 25 pg/ml); progesterone (46 ± 7 ng/ml); PRL (10 ± 1 ng/ml).

Table 1

Summary of wild-type and αERKO mammary gland development induced by the various hormone pellet treatments in ovariectomized mice

Blood and tissue collection

To minimize stress, the mice were housed individually and transferred to the biopsy room the night before they were killed. The next morning each mouse was quickly killed by decapitation and blood was collected from the trunk. After clotting on ice, the blood was centrifuged and the serum was collected and stored at −80 C.

The inguinal (#4 and #5) mammary glands were excised from the skin and subjected to whole mount staining analyses (see below). The right kidney was also removed to confirm the presence of the pituitary graft. The ovaries from mice with pituitary grafts were excised and fixed in 10% formalin overnight. The ovaries were transferred to 70% ethanol at 4 C and then embedded in paraffin, sectioned at 5 μm thickness, and stained with hematoxylin and eosin for analyses. The ovaries were photographed using an Olympus Corp. BX-50 microscope and a DP-10 digital camera.

Mammary gland whole mount analysis

Mammary glands were fixed and stained in carmine alum solution as described (34). Excised mammary glands were placed on glass slides and immersed in tissue fixative (25% glacial acetic acid, 75% ethanol) for a minimum of 1 h at room temp. The glands were placed in 70% ethanol for 15 min and then rinsed in distilled water for 5 min. The mammary glands were stained overnight at room temperature in carmine alum solution (1 g carmine natural red (Sigma, St. Louis, MO), 2.5 g aluminum potassium sulfate (Sigma) in 500 ml water). The glands were then dehydrated progressively in 70–95-100% ethanol for 15 min at each step. The mammary fat pads were cleared in xylene for 15 min before mounting on slides with Permount (Fisher Scientific, Suwanee, GA). The mammary whole mounts were photographed using a Leica Corp. (Deerfield, IL) MZ6 dissecting microscope and a Canon EOS 35-mm camera.

Serum hormone measurements

Serum PRL concentration was measured using mouse and rat PRL RIAs consisting of reagents provided by Dr. A. F. Parlow, Scientific Director of the National Hormone and Pituitary Program. Progesterone and estradiol were measured according to the protocol of commercially available RIA kits (Diagnostics Systems Laboratories, Inc.). All hormone measurements are expressed as mean ± sem. Values for circulating hormones were tested for homoscedascity using Levene’s test. In cases where data exhibited significant heteroscedascity, the data were log-transformed before conducting ANOVA. Posthoc analyses were conducted using the Fisher’s protected LSD.

Results

Serum PRL levels in αERKO females are low and not stimulated by estradiol

To determine if circulating PRL levels were altered in the αERKO mouse, sera from adult WT, HET and αERKO females were analyzed by a mouse PRL RIA. Serum PRL in the αERKO (3.58[mean] ± 1.28[sem] ng/ml) was reduced compared with the WT (18.76 ± 10.66 ng/ml) and HET (12.07 ± 6.70 ng/ml) mice (Fig. 1A). PRL levels were not statistically different between WT and HET females and varied widely within each group. The variability is attributable to differences in the stage of the estrus cycle.

Open in new tabDownload slide

Serum PRL levels in untreated mice and in ovariectomized mice treated with estradiol. A, Serum PRL was measured in untreated WT (n = 8), heterozygous ERα (HET, n = 7) andα ERKO (n = 6) mice and expressed as the mean ± sem. B, Adult WT (n = 5) and αERKO (n = 5) mice were ovariectomized and implanted (sc) with a placebo or estradiol (0.05 mg) pellet. After 21 days, blood was collected and serum PRL was measured and expressed as mean ± sem (A) One-way ANOVA revealed no significant differences between genotypes. B, Two-way ANOVA revealed significant genotype, treatment and interactive effects (P < 0.01). By Fisher’s PLSD: **, P < 0.001 vs. all other groups.

To control for cycling in the WT group and to determine the effect of estradiol treatment on PRL secretion, 12-week-old WT and αERKO mice were OVX and implanted with an estradiol (0.05 mg) or placebo pellet. Mice were killed 21 days later and blood was collected. Serum PRL levels were 4-fold greater in estradiol-treated WT mice (28.09 ± 3.57 ng/ml) compared with the placebo control group (7.68 ± 0.16 ng/ml) (Fig. 1B). PRL levels in αERKO mice treated with estradiol (5.32 ± 0.32 ng/ml) or placebo (4.09 ± 0.86 ng/ml) remained consistently low. Mammary glands from these αERKO mice did not exhibit any growth at this dose of estradiol (data not shown).

A pituitary graft and ovarian hormones induce αERKO mammary gland development

A single pituitary gland from age-matched HET female donors was grafted under the renal capsules of 25 day-old αERKO mice to determine whether the αERKO mammary epithelium was capable of undergoing ductal morphogenesis and/or develop alveolar structures in response to elevated PRL. HET pituitaries were used as grafts since they secrete PRL at levels similar to WT pituitaries (Fig. 1A). In addition, HET mice are the most abundant genotype in the colony therefore it was more practical to use them as pituitary donors to their WT and αERKO siblings. At the time of grafting, the host mice were either OVX or left intact to determine whether ovarian hormones act in concert with the pituitary graft to induce a mammary growth response. Forty-five days after pituitary grafting and ovariectomy, the mammary glands were excised from the host mice and analyzed.

In OVX WT host mice grafted on day 25, neither ductal morphogenesis nor alveolar development occurred in the mammary gland (Fig. 2A), which is similar in appearance to the undeveloped mammary gland from OVX αERKO hosts (Fig. 2B). In contrast, ovarian-intact WT host mice developed a ductal network, and alveolar structures developed along the length of the mammary ducts (Fig. 2C). Ovarian intact αERKO host mice displayed a dramatic mammary growth response as illustrated by the generation of a ductal network and marked alveolar development (Fig. 2D).

Open in new tabDownload slide

Whole mount analyses of mammary glands from intact or ovariectomized mice with a pituitary graft. Twenty-five-day-old (A–D) or 12-week-old (E–H) host WT and αERKO mice were OVX or left intact (INT), and a single pituitary gland from an age-matched heterozygous ERα mouse was grafted under the host renal capsule. After 45 days, the mammary glands were excised and stained with carmine red. The arrows in panels A, B, and F indicate the unstimulated mammary ductal rudiment in immature WT or αERKO mice. Scale bar, 2.5 mm for all panels.

To resolve whether adult αERKO mammary glands can respond to PRL and ovarian hormones, 12-week-old host mice were grafted with an age-matched pituitary. OVX WT host mice possessed a complete ductal network, which is typical of a mammary gland from an adult WT female, but did not display any lobuloalveolar development (Fig. 2E). Ovarian intact WT host mice displayed a complete ductal network and extensive lobuloalveolar development when analyzed 45 days later (Fig. 2G). Pituitary grafting did not stimulate any mammary growth in the OVXα ERKO host mouse (Fig. 2F). In contrast, intact αERKO host mice displayed an expanded mammary ductal network and lobuloalveolar development (Fig. 2H), similar to their counterparts treated at 25 days (Fig. 2D).

Serum PRL levels are elevated in pituitary-grafted mice

Mammary gland growth only occurred in pituitary-grafted WT andα ERKO host mice that possessed ovaries. One possible explanation is that ovarian estradiol is required to drive synthesis and secretion of PRL from the HET pituitary graft to stimulate mammary growth. To address this issue, serum PRL was measured in host mice that received a pituitary graft at 12 weeks with or without ovarian ablation. PRL levels were elevated in both WT and αERKO host regardless of ovarian status (Fig. 3A). The serum PRL levels in OVX (145.80 ± 15.26 ng/ml) and intact (119.36 ± 26.21 ng/ml) WT hosts were similar. PRL levels in αERKO hosts that were OVX (180.07 ± 29.06 ng/ml) or intact (217.68 ± 45.85 ng/ml) were also comparable. Therefore, in mice receiving a pituitary graft at 12 weeks, high circulating PRL levels persisted despite the absence of ovaries.

Open in new tabDownload slide

Serum PRL and progesterone levels from intact or ovariectomized mice with a pituitary graft. Twenty-five day or 12-week-old host WT and αERKO mice were OVX or left intact (INT), and a single pituitary gland from an age-matched heterozygous ERα mouse was grafted onto the host renal capsule. After 45 days, blood was collected from the WT (black bars) and αERKO (open bars) mice and measured for serum PRL (A) and progesterone (B) was using RIAs. Normal mice in (B) refer to serum progesterone levels in untreated control mice. The data are expressed as the mean ± sem (n = 6 for each treatment). A, Two-way ANOVA revealed only significant genotype effects (P < 0.05) for the 12-week grafted animals. Two-way ANOVA revealed significant genotype (P < 0.05) and treatment effects (P < 0.001) for the 25-day grafted animals. (B) Two-way ANOVA was performed on intact animals grafted at 12 weeks and 25 days. Significant genotype, treatment and interactive effects were found (P < 0.01). By Fisher’s PLSD: **, P < 0.001 vs. all other groups.

In contrast, the absence of ovaries did appear to affect PRL levels in mice grafted at 25 days (Fig. 3A). PRL levels were 4-fold greater in intact WT (85.19 ± 18.27 ng/ml) vs. OVX WT (21.53 ± 3.23 ng/ml) mice. Serum PRL was 10-fold greater in intact αERKO mice (259.32 ± 35.80 ng/ml) than in OVXα ERKO mice (26.97 ± 3.20 ng/ml). OVX αERKO mice had a similar PRL concentration (26.97 ± 3.20 ng/ml) as their OVX WT counterparts (21.53 ± 3.23 ng/ml); however PRL levels were 3-fold greater in intact αERKO mice (259.32 ± 35.80 ng/ml) compared with intact WT mice (85.19 ± 18.27 ng/ml). Therefore in mice receiving a graft at 25 days, the presence of ovaries and genotype of the host mouse appeared to influence PRL secretion.

The age of the pituitary gland at the time of grafting also affected PRL levels (Fig. 3A). When used as a graft in OVX WT mice, the pituitary gland from a 12-week-old HET produced 7-fold more PRL (145.80 ng/ml) than a pituitary graft from a 25-day-old HET (21.53 ng/ml). The same observation held true when comparing PRL levels in OVXα ERKO mice grafted with a pituitary from a 12-week-old (180.07 ng/ml) vs. 25 day old (26.97 ng/ml) HET female.

Mammary growth is not stimulated in pituitary-grafted αERKO mice treated with estradiol or DHT

Mammary gland development does not occur in pituitary-grafted WT or αERKO mice with ablated ovaries despite high circulating concentrations of PRL. Therefore, ovarian hormone(s) likely participate in the mammary growth response initiated by PRL signaling. To address this issue, 12-week-old WT and αERKO mice were OVX and received a single pituitary graft. These mice were then treated with a single pellet containing placebo, estradiol (0.05 mg), or DHT (5 mg,α ERKO only) for 42 days before they were killed (see Materials and Methods). These hormones were chosen because estradiol and testosterone are elevated in αERKO females (31, 35) and could potentially contribute to the mammary growth response. PRL levels were elevated in both placebo (185.48 ± 29.50 ng/ml) and estradiol-treated (323.20 ± 59.78 ng/ml) WT mice and the estradiol-treated WT mice exhibited limited alveolar bud growth (data not shown). PRL levels were also elevated in αERKO mice treated with either placebo (128.5 ng/ml), estradiol (122.40 ± 19.51 ng/ml) or DHT (223.50 ± 56.47 ng/ml). Even though the pituitary graft produced markedly elevated PRL levels, the concomitant steroid treatments did not generate a mammary growth response in the αERKO (data not shown).

Serum progesterone levels are elevated in pituitary- grafted mice

PRL is known to induce synthesis and secretion of ovarian progesterone as a mechanism to stimulate mammary alveolar growth (21, 36, 37). Therefore, serum progesterone was measured in WT and αERKO female mice grafted with a pituitary at 12 weeks (Fig. 3B). Progesterone levels were elevated 7-fold in intact WT hosts (12.11± 3.37 ng/ml) compared with ungrafted WT females (1.76 ± 0.74 ng/ml). The concentration of serum progesterone in intact αERKO hosts (95.5 ± 6.79 ng/ml) was 60-fold greater than ungrafted αERKO mice (1.60 ± 0.19 ng/ml). As expected, progesterone levels in OVX WT (0.39 ± 0.14 ng/ml) and OVX αERKO (0.58 ± 0.16 ng/ml) hosts approached the lower detection limit of the RIA.

Sera from 70-day-old WT and αERKO mice grafted at 25 days were also measured for progesterone (Fig. 3B). Again, the progesterone levels were barely detectable in OVX WT hosts (0.91 ± 0.50 ng/ml) and below the level of detection in OVX αERKO hosts (<0.3 ng/ml). Intact WT mice grafted at 25 days possessed similar levels of progesterone (7.63 ± 5.68 ng/ml) as their WT counterparts grafted at 12 weeks (12.11 ± 3.37 ng/ml). Intact αERKO mice grafted at 25 days possessed the highest levels of progesterone (17.08 ± 6.22 ng/ml) in this age group. Although this level was much lower than serum progesterone in αERKO mice grafted at 12 weeks (95.50 ± 6.79 ng/ml), it was still sufficient to drive mammary growth in the αERKO (Fig. 2D).

Pituitary-grafted αERKO mice possess hypertrophied corpora lutea

Progesterone levels were elevated in intact αERKO mice containing a pituitary graft compared with ungrafted αERKO mice. Therefore, ovaries from grafted and ungrafted αERKO mice were sectioned and stained with H&E for analyses and compared with their WT counterparts. Ovaries from ungrafted WT mice revealed follicles at different stages of development including a corpus luteum (Fig. 4A). αERKO ovaries possessed many immature follicles at the primary to tertiary stage, in addition to atretic, hemorrhagic follicular cysts (Fig. 4B). The αERKO ovaries do not ovulate as demonstrated by the absence of corpora lutea (CL). The ovaries from a grafted WT mouse appeared normal with the presence of some CL (Fig. 4, C and E). In contrast, the ovaries from graftedα ERKO mice were very enlarged relative to ovaries from grafted WT mice (compare Fig. 4, C and D). Furthermore, the αERKO ovary contained many CL comprised of hypertrophied luteal cells (Fig. 4D). The cytoplasm of the luteal cells was highly vacuolated with lipid droplets, indicative of steroidogenesis (Fig. 4F). Corpora lutea formation also occurred in the ovaries of αERKO mice grafted at 25 days (data not shown). These studies have described the first observations of CL formation in the αERKO ovary under any experimental conditions to date.

Open in new tabDownload slide

Histology of ovaries from untreated or pituitary-grafted adult mice. Ovaries from adult WT (panels C and E) and αERKO (panels D and F) mice grafted with a heterozygous ERα pituitary, as described in Fig. 2, were sectioned and stained with H&E and compared with untreated adult WT (panel A) and αERKO (panel B) controls. Corpora lutea are within the boxed regions of panels C and D, and are enlarged in panels E and F, respectively. Scale bar, 500 μm (panels A–D) and 200 μm (panels E and F).

Hormonal requirements for mammary gland growth in αERKO mice

Intact ovaries in concert with a pituitary graft are required to induce mammary growth in the αERKO. To determine the specific hormonal requirements for mammary growth in αERKO and WT mice, a group of 12-week-old WT and αERKO mice were OVX. Pellets of estradiol (2 × 0.1 mg), progesterone (1 × 35 mg), or rat PRL (2× 5 mg) were implanted (sc) individually or in various combinations. The expected hormone release rates were chosen to mimic the circulating levels of estradiol in αERKO mice, and PRL and progesterone found in pregnant mice (see Materials and Methods). The mice were killed 21 days later and the mammary glands were analyzed by whole mount. The mammary growth responses to the different treatments are summarized in Table 1. Serum hormone levels after the individual hormone treatments are reported in the footnotes to Table 1.

WT mice displayed a fully developed network of thin ducts occupying the entire fat pad (Fig. 5A. In contrast, αERKO mice possessed only a rudimentary ductal structure emanating from the nipple (Fig. 5B). Estradiol induced limited dilation of the ducts in WT mice (Fig. 5C). Pharmacological doses of estradiol stimulated TEB formation in 2 of 4 αERKO mice, and ductal elongation beyond the lymph node was clearly evident in one of these mice (Fig. 5D). Progesterone treatment resulted in side-branching and limited alveolar budding in WT mammary glands (Fig. 5E) but had no effect on the αERKO ductal rudiment (Fig. 5F). PRL exposure had little effect on the WT gland, perhaps generating limited ductal dilation (Fig. 5G). The αERKO mammary rudiment was unaffected by PRL treatment (Fig. 5H). Estradiol + PRL treatment induced some alveolar budding in the WT mammary gland (Fig. 5I). After estradiol + PRL treatment, 3 of 4α ERKO mice developed TEBs, and ductal elongation occurred in 2 of these mice (Fig. 5J). The effect of progesterone + PRL on mammary growth was no different than either hormone alone in both WT andα ERKO mice (Fig. 5, K and L). Estradiol + progesterone stimulated ductal dilation and lobuloalveolar development in WT mice (Fig. 5M). Interestingly, estradiol + progesterone stimulated TEB formation, ductal side-branching and alveolar development in αERKO mammary glands (Fig. 5N). However, the addition of PRL to the estradiol + progesterone regimen had no further effect on mammary development in WT and αERKO mice (Fig. 5, O and P).

Open in new tabDownload slide

Open in new tabDownload slide

Whole mount analyses of mammary glands from ovariectomized mice treated with various hormones. Twelve-week-old WT and αERKO mice were ovariectomized and implanted (sc) with estradiol (E2, 2 × 0.1 mg), progesterone (Prog, 1 × 35 mg) or PRL (Prl, 2 × 5 mg) hormone pellets in various combinations as summarized in Table 1. After 21 days, the mammary glands were excised and stained with carmine red. The arrow in panels B, F, H, and L indicate the unstimulated mammary ductal rudiment. The arrow in panels D, J, N, and P indicate ductal expansion and possible terminal end buds. Scale bar, 2.5 mm for all panels.

Serum PRL levels measured in PRL pellet-treated WT and αERKO mice were lower than expected at the time the mice were killed, although PRL levels may have been higher earlier in the treatment. A false negative result (no mammary growth) due to inadequate PRL levels is possible. However, we believe this is not the case since we have already demonstrated that elevated PRL levels (145–180 ng/ml) generated from a pituitary graft at 12 weeks (Fig. 3A) did not induce mammary growth in OVX αERKO and WT mice (Fig. 2, E and F). We also determined that OVX, adult αERKO mice implanted with a pituitary graft (as a source of PRL) and a progesterone pellet did not exhibit any mammary growth (data not shown).

Discussion

The previously described αERKO mouse clearly demonstrated that estrogen/ERα signaling is required for postnatal mammary gland development (28, 29). Estrogen is thought to regulate the hormonal signaling systems of other mammogenic hormones such as progesterone (38) and PRL (39). Therefore, disruption of the ERα gene might be expected to diminish PRL and progesterone signaling necessary for normal mammary gland development that contributes to the αERKO mammary phenotype. In this report, we have established that serum progesterone and PRL levels are low in the female αERKO mouse. By using pituitary grafting and hormone replacement techniques in theα ERKO mouse, we have demonstrated that progesterone, in the presence of estradiol, directly stimulated αERKO mammary gland development. In contrast, PRL contributed to mammary development by inducing corpora lutea formation in the αERKO ovary thereby leading to elevated serum progesterone levels. In addition, pharmacological doses of estradiol could stimulate ductal morphogenesis in the αERKO mammary gland. Therefore, the manifestation of the αERKO mammary phenotype is due to the lack of estrogen/ERα action both directly at the mammary gland and indirectly on the induction of mammogenic hormones associated with the hypothalamic/pituitary/gonadal axis.

Estradiol stimulates PRL secretion by inducing pituitary PRL gene expression and by reducing dopamine release from the hypothalamus (15–17). Accordingly, we report that circulating PRL levels are reduced at least 5-fold in the αERKO females and are not inducible by estradiol. The low serum PRL is consistent with a 15-fold reduction in PRL mRNA levels in the αERKO pituitary (32). The absence of ERα in the αERKO pituitary and hypothalamus abolished estradiol-induced PRL secretion normally associated with proestrus (18, 40). Indeed, the 5-fold difference between WT and αERKO serum PRL levels would likely be much greater if WT mice in proestrus were analyzed. Thus, estrogen-induced PRL secretion is mediated by ERα, and not ERβ, in the hypothalamus/pituitary axis.

PRL levels were dramatically elevated in WT and αERKO host mice receiving the age-matched graft at 12 weeks regardless of ovariectomy. These results are consistent with previous observations that PRL secretion occurs at a high spontaneous rate when the pituitary is transplanted to a site distant from the inhibitory actions of the hypothalamus (15). Interestingly, serum PRL levels in OVX mice receiving an age-matched pituitary graft at 25 days were much lower than their ovarian-intact counterparts. This observation suggests that the 25-day-old pituitary graft is at a less mature stage of development and requires a period of estrogen exposure to develop into a more functional pituitary gland. The αERKO pituitary gland has been shown to possess a modest decrease in lactotroph cell number in the absence of estrogen/ERα stimulation (32). Therefore, the estrogen-induced maturation effect on the HET pituitary graft does not involve development of the lactotroph cell lineage but may involve lactotroph proliferation (41).

The pituitary gland also secretes GH that contributes to mammary development (42, 43). However, unlike PRL secretion that is under tonic hypothalamic inhibition, GH secretion requires positive stimulation from hypothalamic GHRH (44). In accordance with the requirement for GHRH stimulation, mice with long-term pituitary grafts possessed elevated PRL with no stimulation of GH levels (45, 46). Therefore, it is unlikely that the pituitary graft contributed to αERKO mammary gland growth by ectopic GH secretion.

αERKO females are anovulatory and accumulate developmentally arrested follicles (47). Consequently, the αERKO ovaries do not generate CL and serum progesterone levels remain low. In contrast, pituitary-grafted αERKO mice developed large spontaneous CL that secreted progesterone in the absence of true ovulation. A similar ovarian phenotype was reported in mice possessing an LHβ-transgene and elevated serum LH levels (48). PRL is thought to produce luteotropic effects on the ovary by elevating the level of LH receptors on granulosa and thecal cells (49–51). Because serum LH levels and ovarian LH receptors were already elevated in the αERKO (31, 47, 52), it is possible that the granulosa cells were sensitized toward the luteal phenotype. Consequently, the granulosa cells were induced to terminally differentiate into luteal cells by elevated PRL from the pituitary graft. Hypergonadotropic stimulation by chronically elevated LH was shown to be responsible for the αERKO ovarian polycystic phenotype (52). Thus, the higher progesterone levels in αERKO mice grafted at 12 weeks may be due to the greater number of arrested follicles capable of responding to PRL compared with the mice grafted at 25 days.

To date, ductal elongation had never been detected in mammary glands from αERKO female mice. However, estradiol at pharmacological circulating levels was capable of inducing TEB development and limited ductal morphogenesis in some αERKO mice. ERβ is a candidate to mediate the effects of pharmacological estradiol levels in the αERKO mammary gland. Two recent reports have indicated that ERβ protein is detectable in mammary epithelial cells of mice (53) and rats (54). Our laboratory has been unable to detect ERβ mRNA in the mouse mammary gland by RPA (55); however, we and others have detected mammary ERβ mRNA by RT-PCR (31, 56). These results indicate that the RPA is not sensitive enough to detect very low levels of ERβ mRNA that are further diluted out in the total RNA isolated from a mammary gland. The fact that ERβ protein is concentrated in the nuclei of predominantly epithelial cells that are only weakly immunostained suggests a low abundance of ERβ in mammary cells (53). If ERβ protein is present at low levels in the αERKO mammary gland, the requirement for pharmacological doses of estradiol to stimulate mammary ductal growth may reside in the fact that ERβ has weak estradiol-induced AF-2 function and an estradiol-repressed AF-1 function (57, 58).

Estradiol + progesterone were capable of inducing ductal growth and alveolar development in the αERKO mammary gland. In contrast, progesterone alone or in combination with PRL did not induce αERKO mammary growth. These results suggest that progesterone stimulation of mammary development is dependent on estrogen action to induce mammary PR gene expression (9–11). Indeed, mammary PR mRNA was modestly elevated 20–30% in OVX αERKO mice treated with estradiol + progesterone compared with OVX controls (data not shown). Although this modest elevation in PR mRNA is likely due to increased mammary epithelial cell content, the requirement for estradiol to facilitate the action of progesterone in the αERKO mammary gland is clear. Therefore, ERβ may also indirectly mediate the effect of progesterone action in the αERKO mammary gland.

Another possible explanation for the estrogen-induced effects observed in the αERKO mammary gland is that they may be mediated by the previously described E1 ERα variant (59). The E1 mRNA variant results from the splicing out of the neo gene from the disrupted ERα gene in the αERKO. If translated into protein, the E1 mRNA variant would produce an ERα that lacks amino acids 92–155. αERKO uterine samples bound estradiol at 3–9% of WT levels, however E1 protein was not detectable by Western analysis (59). When expressed in COS cells, the E1 variant displayed estrogen-induced transcriptional activity that was 35% of WT in a reporter gene assay. If present in the αERKO mammary gland, the E1 variant might mediate the effects of long-term pharmacological estradiol treatment used in these studies.

Mammary growth in the αERKO could result from the action of catecholestrogens arising from metabolism of pharmacological estradiol levels. Catecholestrogens have been shown to mediate estrogen-like effects in WT and αERKO mouse uteri including increases in lactoferrin and PR mRNA, and water imbibition that are not abrogated by the antiestrogen, ICI182780 (60, 61). The double knockout mouse lacking ERα and ERβ (62), which has a mammary phenotype similar to theα ERKO, may be a useful tool to address whether αERKO mammary growth induced by pharmacological doses of estradiol is mediated by E1, ERβ, or another receptor-based mechanism stimulated by catecholestrogens.

Elevated PRL levels alone did not stimulate mammary gland growth in mice lacking ovaries, indicating that PRL has no direct effect on the mammary gland. However, our studies do not address the direct role of PRL in lobuloalveolar expansion and differentiation into a secretory lactational phenotype during pregnancy that appears to be mediated by the PRL receptor in the mammary gland (39). In addition, placental lactogens are thought to contribute to the differentiation of mammary lobuloalveoli into secretory structures (39). Because the αERKO females are infertile, the effect of PRL on mammary development during pregnancy may not be realized out of its proper physiological context. Our findings are consistent with a recent report indicating that PRL acts indirectly via ovarian progesterone to induce ductal side-branching in cycling virgin mice, and directly on mammary epithelium to induce lobuloalveolar differentiation during pregnancy (63).

In conclusion, the αERKO mammary ductal rudiment is not refractory to stimulation by estradiol and progesterone. This is consistent with our previous report that the αERKO mammary epithelial rudiment remains susceptible to mitogenic stimulation by an oncogene (64). We have demonstrated that if the αERKO mouse was capable of initiating and maintaining a normal estrus cycle, the mammary gland could be stimulated to grow by the elevated progesterone levels induced by PRL. It is interesting that the αERKO mammary growth response induced by progesterone remains dependent on the presence of estradiol. Our experiments may have unmasked a potential role for ERβ or another receptor-associated pathway to mediate the action of estradiol in theα ERKO mammary gland.

Acknowledgments

The authors thank Selena Mistich at the U.S. EPA for the serum PRL RIA. At NIEHS, we gratefully acknowledge Dr. Barbara Davis for histological examination of the ovaries, and Dr. Stefan Mueller for help in excising tissues for this study. We thank John Couse for helpful scientific discussions and editing of the manuscript. We also appreciate the editorial review of the manuscript by Dr. Diane Klotz and Wendy Jefferson.

References

8

, , ,

Essential role of endogenous estrogen in directly stimulating mammary growth demonstrated by implants containing pure antiestrogens.

9

, ,

Mammary gland growth and development from the postnatal period to postmenopause: ovarian steroid receptor ontogeny and regulation in the mouse.

J Mammary Gland Biol Neoplasia

18

, ,

Prolactin and growth hormone levels in different inbred strains of mice: patterns in association with estrous cycle, time of day, and perphenazine stimulation.

20

, , ,

Lactoferrin expression in the mouse reproductive tract during the natural estrous cycle: correlation with circulating estradiol and progesterone.

21

, , , , , ,

Interrelationship between pituitary and ovarian hormones in normal and neoplastic growth of mammary glands of mice.

22

, , , , , , , ,

Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities.

23

, , , , , , , , , ,

Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse.

24

, , , , , , , ,

Defective mammopoiesis, but normal hematopoeisis, in mice with a targeted disruption of the prolactin gene.

26

, , , , , , , , ,

Cyclin D1 provides a link between development and oncogenesis in the retina and breast.

28

, , , , , , , , , , ,

Estrogen receptor gene disruption: molecular characterization and experimental and clinical phenotypes.

30

, , , , , , , , ,

Generation and reproductive phenotypes of mice lacking estrogen receptor β.

33

, , , , ,

Effects of castration and chronic steroid treatments on hypothalamic gonadotropin-releasing hormone content and pituitary gonadotropins in male wild-type and estrogen receptor-α knockout mice.

34

, , , ,

Synergistic interaction of transforming growth factor α and c-myc in mouse mammary and salivary gland tumorigenesis.

37

, , , ,

Prolactin (PRL) and its receptor: actions, signal transduction pathways and phenotypes observed in PRL receptor knockout mice.

41

, , ,

Differential regulation by estrogens of growth and prolactin synthesis in pituitary cells suggests that only a small pool of estrogen receptors is required for growth.

45

, ,

Ectopic pituitary grafts in mice: hormone levels, effects on fertility, and the development of adenomyosis uteri, prolactinomas, and mammary carcinomas.

46

, , , , , ,

Kinetics of mammary epithelial cell proliferation in pituitary isografted BALB/c mice.

47

, , , , , ,

Targeted disruption of the estrogen receptor-α gene in female mice: characterization of ovarian responses and phenotype in the adult.

48

, , , , ,

Targeted overexpression of luteinizing hormone in transgenic mice leads to infertility, polycystic ovaries, and ovarian tumors.

52

, , , ,

Prevention of the polycystic ovarian phenotype and characterization of ovulatory capacity in the estrogen receptor-α knockout mouse.

53

, , ,

Murine progesterone receptor expression in proliferating mammary epithelial cells during normal pubertal development and adult estrous cycle: association with ERα and ERβ status.

55

, , , ,

Tissue distribution and quantitative analysis of estrogen receptor-α (ERα) and estrogen receptor-β (ERβ) messenger ribonucleic acid in the wild-type and ERα-knockout mouse.

56

57

, , , , , , , , , ,

The estrogen receptor enhances AP-1 activity by two distinct mechanisms with different requirements for receptor transactivation functions.

59

, , , , , , ,

Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene.

60

, , , , ,

Estrogenic responses in estrogen receptor-α deficient mice reveal a distinct estrogen signaling pathway.

61

, , ,

Differential spatiotemporal regulation of lactoferrin and progesterone receptor genes in the mouse uterus by primary estrogen, catechol estrogen, and xenoestrogen.

63

, , , , , , ,

Prolactin controls mammary gland development via direct and indirect mechanisms.

64

, , , ,

A mouse mammary tumor virus-Wnt-1 transgene induces mammary gland hyperplasia and tumorigenesis in mice lacking estrogen receptor-α.

Reproduction-Development