Estradiol and tamoxifen mediate rescue of the dominant-negative effects of estrogen response element-binding protein in vivo and in vitro

Abstract

Biological responses to estrogens are dependent on the integrated actions of proteins, including the estrogen receptor (ER)-alpha, that regulate the transcription of estrogen response element (ERE)-containing target genes. We have identified a naturally occurring ERE antagonist, termed an ERE-binding protein (BP). To verify that ERE-BP can induce estradiol (E(2)) resistance in vivo, we generated transgenic mice that overexpress this protein in breast tissue. Female transgenic mice with high levels of ERE-BP were unable to lactate, and we hypothesized that this effect was dependent on the relative levels of ERE-BP and ERalpha ligand. To test this hypothesis, wild-type and ERE-BP-expressing female mice were implanted with capsules containing E(2), the selective estrogen receptor modulator tamoxifen, or placebo. Histological analysis of nonlactating mammary glands showed a 4.5-fold increase in gland branch number and 3.7-fold increase in ducts in ERE-BP mice treated with E(2) (7.5 mg, 21 d) compared with placebo-treated ERE-BP mice. Wild-type mice showed a 5.3-fold increase in branches and 1.4-fold increase in ducts under the same conditions. Similar results were obtained with tissue from lactating mice, in which tamoxifen also increased mammary gland branch number. Studies using ERE-BP-expressing MCF-7 breast cells showed that high doses of E(2) (1000 nM) restored normal ERalpha-chromatin interaction in these cells, whereas tamoxifen was able to achieve this effect at a dose of 10 nM. These data highlight the importance of ERE-BP as an attenuator of normal ERalpha signaling in vivo and further suggest that ERE-BP is a novel target for modulation by selective estrogen receptor modulators.

Figure 1

ERE-BP is expressed in ERα-positive cells in mammary gland ductal cells from ERE-BP+ mice. Immunofluorescence analysis of ERE-BP and ERα protein in sections of mammary gland tissue from a nonlactating ERE-BP+ female mouse. Coincident expression of ERα and ERE-BP is shown as merged immunofluorescence. Control staining is shown as shown as immunofluorescence for IgG as primary antibody.

Figure 2

Serum E₂ levels in female mice after treatment with SERMs. WT and ERE-BP transgenic mice (ERE-BP+) were implanted with 7.5 mg E₂ slow-release pellets and blood sampled at different time points to assess serum levels of E₂. Values shown are the mean of n = 5 mice for each treatment and time point. TAM, Tamoxifen.

Figure 3

Effect of E₂ on mouse mammary gland branching in virgin, nonlactating, female mice. Four-week-old virgin female WT and ERE-BP transgenic mice (ERE-BP+) were implanted with slow-release pellets for either E₂ (0.5 mg or 7.5 mg). After a further 21 d, mice were euthanized, and whole tissue mounted and stained with carmine aluminum solution (A); breast tissue sections assessed histologically after H&E staining (B; magnification, ×40). Quantification of carmine aluminum and H&E staining is shown in C and D, respectively (n = 3 separate slides with three separate fields of view for each slide), based on arbitrary units for branch area quantified by Adobe CS3. Statistical analysis of data in C and D is shown as P values calculated by ANOVA.

Figure 4

Effect of E₂ on mammary gland ducts in virgin, nonlactating, female mice. Four-week-old female WT (A) and ERE-BP transgenic mice (ERE-BP+) (B) were implanted with slow-release pellets for either E₂ (7.5 mg) or placebo. After a further 21 d, mice were euthanized and breast tissue sections assessed histologically after H&E staining. Inset on ×10 magnification images corresponds to field for ×40 magnification. Quantification of H&E staining is shown in C (n = 3 separate slides with three separate fields of view for each slide), based on arbitrary units for duct perimeter quantified by Adobe CS3. Statistical analysis of data in C is shown as P values calculated by ANOVA.

Figure 5

Effect of E₂ or tamoxifen on mammary gland development in lactating female mice. A, One-month-old female ERE-BP transgenic mice (ERE-BP+) were implanted with slow-release pellets for E₂ (0.72 or 7.5 mg), tamoxifen (TAM) (6.25 mg), or placebo. After expiration of the timed-release pellets (21 d for 7.5 mg E₂ and 6.25 mg TAM or 60 d for 0.72 mg E₂), mice underwent a 2-wk SERM-free hiatus. The resulting 9- or 14-wk-old females were then bred with ERE-BP+ male counterparts and breast tissue examined at d 3 of lactation after birth. Whole-mount breast tissue sections were stained with carmine aluminum solution (top panel) and breast tissue sections assessed histologically after H&E staining (middle and lower panels). In each case, insets show magnification values. B, Carmine aluminum and H&E staining in lactating mammary glands from WT mice. C, Quantification of H&E staining of mammary gland branches and ducts (n = 3 separate slides with three separate fields of view for each slide), based on arbitrary units for branch area and duct perimeter quantified by Adobe CSE. Statistical analysis of data in C is shown as P values calculated by ANOVA. *, Statistically different from placebo-treated ERE-BP+ mice P < 0.05; **, statistically different from placebo-treated ERE-BP+ mice P < 0.01; ***, statistically different from placebo-treated ERE-BP+ mice P < 0.001.

Figure 6

Effect of E₂ or tamoxifen (tam) on dysregulated ERα-chromatin interaction in WT and ERE-BP overexpressing MCF-7 cells. Nuclear chromatin from parental MCF-7 cells (A) and stable transfectant variants of MCF-7 overexpressing the ERE-BP (B; MCF-7 ERE-BP) were used in ChIP assays. Cells were treated with or without E₂ (10 or 1000 nm) or tamoxifen (10 or 1000 nm) for 15 min. Occupancy of the ERE in the pS2 gene promoter by ERα, ERE-BP, or IgG (negative control) was determined by PCR amplification of the estrogen-responsive DNA region of the pS2 promoter after precipitation with antibodies to ERα (ERAb), ERE-BP (ERE-BPAb), or IgG (see Materials and Methods).