The Cell Fate Determination Factor DACH1 Is Expressed in Estrogen Receptor-α–Positive Breast Cancer and Represses Estrogen Receptor-α Signaling

Estrogen receptor α (ERα) plays a major role in regulating the growth, survival, and differentiation of normal and malignant breast epithelial cells. ERα is overexpressed in 60% to 80% of human breast cancers, with ∼70% of patients responding to endocrine treatments (1, 2). The ERα encodes a modular nuclear receptor with an NH₂-terminal activation function 1 (AF-1), a COOH-terminal activation function 2 (AF-2), and a conserved ligand-binding domain within the COOH terminus. The ligand-bound receptor dimerizes and binds to ER response elements (ERE) in the promoter region of target genes. ERα can, via distinct cognate transcription factors (activator protein-1, nuclear factor-κB, SP-1, and FKHR), bind to alternate DNA sequences (3, 4). In addition to the predominantly nuclear location, ERα is located in the plasma membrane, cytoplasm, and mitochondria (5–8).

ERα participates in nongenomic (cytoplasmic and membrane-mediated) signaling via a multiprotein complex, which includes the ERα, Src kinase, phosphatidylinositol 3-kinase, SHC, and G proteins (3). Estrogen-mediated induction of cell survival and proliferation induces AKT and mitogen-activated protein kinase, which in turn can phosphorylate the ERα and its coregulators (9). Estrogen antagonists are widely used and represent an effective treatment for ERα-positive breast cancers. Antiestrogens, such as tamoxifen, inhibit estrogen binding to the ERα, whereas fulvestrant (ICI 182, 780, Faslodex) blocks dimerization (10) and reduces ERα protein abundance. Current therapies fail a proportion of patients who at initial diagnosis express ERα (de novo resistance) and the majority of patients who eventually become resistant (acquired resistance; refs. 11, 12). A continued role for ERα in most resistant tumors is underscored by the finding that most resistant tumors remain ERα positive. The identification of endogenous inhibitors of ERα in human breast cancer may provide alternative therapeutic targets for ERα-positive breast cancers that are resistant to current treatments.

Steroid receptor coactivators and corepressors interact with the ERα to regulate target gene expression and estradiol (E₂)-dependent biological effects (13). ERα corepressors encode, or are associated with, histone deacetylases (HDAC). A number of ERα coactivator and corepressor proteins are dysregulated in human breast cancer, which may contribute to altered cellular growth or therapeutic resistance (14, 15). Coactivators include the cointegrator proteins (p300/CREB binding protein); the steroid receptor coactivation proteins; and chromatin remodeling recruitment scaffolds such as proline, glutamic acid, and leucine rich protein 1 (PELP1). Initially cloned as an ERα coactivator (16), PELP1 encodes a proline-rich protein with interaction motifs that bind to FHA, SH2, SH3, PDZ, and WW domains and encodes nuclear receptor interacting boxes (LXXLL) that enable interaction with multiple nuclear receptors. Estrogen promotes PELP1 expression and PELP1 interaction with the AF-2 domain of ERα through the LXXLL motifs 4 and 5 of PELP1 (16–18). PELP1 is expressed in a variety of tissues including the breast, with the highest expression found in the mammary gland during pregnancy.

Recent studies of more than 2,200 human breast cancer samples showed that the reduction in abundance of a cell fate determination factor, DACH1, correlated with poor prognosis (19). Initially cloned in a screen for antagonists of the Drosophila hyperactive EGFR mutation ellipse (20), the dac gene is sufficient to induce ectopic eye formation (21). The human orthologue, DACH1 gene, encodes a cell fate determination protein with domains homologous to the Sno/Ski oncogenes. Dac functions in a complex regulatory network [Retinal Determination Gene Network (RDGN)] of genes including the twin of eyeless (toy), eyeless (ey), sin oculis (so), and eyes absent (eya) gene complex (22). The RDGN governs organismal fate in metazoans, including retinal, muscle, and gonodal development (23).

In view of the finding that DACH1 expression is lost during progression of human breast cancer correlating with poor prognosis, and that ERα plays a key role in human breast cancer progression, we examined the interaction between DACH1 and estrogen signaling. DACH1 physically interacted with and inhibited ligand-dependent ERα activity. DACH1 inhibited ligand-dependent, ERα-mediated cellular proliferation and DNA synthesis. Through a proteomic analysis and direct sequencing, the ERα coregulator PELP1 was identified as a DACH1-binding PELP1 associated with DACH1 in an intranuclear/extranucleolar distribution and reverted DACH1-mediated ERα repression. With the DACH1 gene shown to encode a key component of the RDGN signaling network, these studies provide the first evidence for interaction between the RDGN network and hormone signaling in mammalian cells.

Plasmid constructions. The expression plasmids, which include an NH₂-terminal FLAG peptide for DACH1, DACH1 DS-domain alone (DS), or DACH1 DS-domain deleted (ΔDS), were previously described (24). The FLAG-tagged DACH1 cDNA was subcloned into the vector MSCV-IRES-GFP. The expression vector encoding wild-type (wt) and mutant ERα (25) and PELP1 (26, 27) mutants were previously described.

Cell culture, DNA transfection, and luciferase assays. Cell culture, DNA transfection, and luciferase assays were done as previously described (28–30). Cells were plated at a density of 1 × 10⁵ cells in a 24-well plate on the day before transfection with EasyTransgator according to the manufacturer's protocol (America Pharma Source). Luciferase activity was normalized for transfection efficiency using a β-galactosidase reporter as an internal control. Statistical analyses were done using the Mann-Whitney U test.

Western blot and immunoprecipitation assays. Western blotting for DACH1 (M2), FLAG tag (Sigma), and the loading control guanine dissociation inhibitor was conducted as previously described (24, 31). HEK293T cells were used for the detection of protein-protein interactions and immunoprecipitation was conducted as previously described using anti-FLAG–coated M2 beads. Immunoblotting was conducted with antibodies to ERα (H-184; Santa Cruz Biotechnology) and DACH1 (Abcam). HEK293T cells were cotransfected with 3xFLAG-ERα and T7-PELP1 in a 10-cm cell culture dish with 70% cell confluency. Five hours after transfection, cells were treated with either E₂ (10 nmol/L) or vehicle for 24 h. Cell lysates were prepared using TBS-T buffer [50 mmol/L Tris (pH 7.4), 150 mmol/L NaCl, 1% Triton X-100, 1 mmol/L EDTA, 1 mmol/L DTT, and proteinase inhibitors]. Immunoprecipitation was done on 1 mg of the whole-cell lysates using 30 μL of FLAG affinity gel (Sigma). 6xHIS-DACH1 proteins purified from bacteria were added to the immunoprecipitation reaction at increasing dose as indicated in the figure. The immunoprecipitate was resolved on 8% SDS-PAGE gel. Western blot was conducted using anti-FLAG antibody (Sigma) for ERα and T7 antibody (Bethyl Laboratories) for PELP1.

[³H]Thymidine uptake. [³H]Thymidine uptake was done as previously described (32). MCF-7 cells stably infected with either MSCV-IRES-GFP or MSCV-IRES-DACH1-GFP were plated in triplicate in 24-well plates in DMEM supplemented with 10% fetal bovine serum, 1× penicillin/streptomycin. Cells were allowed to attach overnight, after which medium was switched to phenol red–free DMEM supplemented with 5% charcoal stripped fetal bovine serum, 1× penicillin/streptomycin for 24 h. Cells at ∼70% confluency were treated with E₂ (10 nmol/L) or vehicle 7 h, after which cells were pulsed with [³H]thymidine for an additional 2 h (2.0 μCi/well). Cells were washed with cold PBS, and proteins precipitated with 10% trichloroacetic acid for 30 min at 37°C (100 μL/well). Precipitation was followed by additional washing with 10% trichloroacetic acid, after which cells were treated with 0.2 N NaOH and collected in scintillation vials. Radioactivity was read using a Wallac 1209 Rackbeta liquid scintillation counter.

Immunohistochemistry. To analyze the coexpression of DACH1 and ERα in human breast tissue, microarrays were constructed from paraffin-embedded tissues using the Cutting Edge Matrix Assembly technique (33) and consisted of 140 invasive ductal carcinomas and 40 normal breast tissues. Immunohistochemistry was conducted as described (2) using antibodies specific for DACH1 (Abcam) and ERα (DAKO). Immunofluorescent images were taken using a PM2000 microscope (HistoRX; magnification, ×60). These images were then analyzed for intensity of DACH1 and ERα expression in human breast epithelial cells using the Image J program (NIH) to quantify expression in each single cell with an image of breast tissue. These data were then graphically represented using the GraphPad Prism version 5.0 software (GraphPad Software, Inc.). This study did not need approval from the Institutional Review Board, Thomas Jefferson University, as the human breast tissue was more than 10 y old.

Confocal microscopy. Cells grown in four-well chamber slides were fixed with 4% paraformaldehyde for 20 min at the room temperature. The slides were rinsed with PBS and permeated with 0.05% NP40. The primary antibodies used were mouse monoclonal anti-T7 for PELP1 (Bethel Laboratory) and rabbit polyclonal anti-FLAG. The secondary antibodies used were Alexa Fluor 633–conjugated F(ab′)2 fragment of goat anti-mouse IgG (Molecular Probes, Inc.; 1/250) and rhodamine red X–conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Lab; 1/50). Fluorescence confocal imaging was acquired with a 63× objective of a Zeiss 510 Meta laser confocal microscope. The images were processed with MetaMorph (Molecular Devices).

Chromatin immunoprecipitation. Chromatin immunoprecipitation analysis was done following a protocol provided by Upstate Biotechnology under modified conditions (34). MCF-7 cells (1 × 10⁶) were grown in DMEM with 10% charcoal dextran-stripped serum for 3 d. Cells were treated with E₂ for the time points indicated in the figure legends (45 min, 1 h, 12 h). On E₂ (100 nmol/L) stimulation, the cells were cross-linked by directly adding 1.1% formaldehyde buffer containing 100 mmol/L sodium chloride, 1 mmol/L EDTA-Na (pH 8.0), 0.5 mmol/L EGTA-Na, Tris-HCl (pH 8.0) to culture medium for 10 min at 37°C. The medium was aspirated; cells were washed thrice using ice-cold PBS containing 10 mmol/L DTT and protease inhibitors, lysed in warm 1% SDS lysis buffer, and incubated for 10 min on ice. The cell lysates were sonicated to shear DNA to lengths between 200 and 1,000 bp and the samples were diluted 10-fold in chromatin immunoprecipitation dilution buffer. To reduce nonspecific background, the cell pellet suspension was precleared with 60 μL of salmon sperm DNA/protein A agarose-50% slurry (Upstate Biotechnology) for 2 h at 4°C with agitation. The following primers were used for PCR of the pS2 promoter: sense, 5′-GGCCATCTCTCACTATGAATCACTTCTGC-3′; antisense, 3′-GCAGAAGTGATTCATAGTGAGAGATGGCC-5′.

In-gel trypsin digestion. Gels were stained with Coomassie G-250 to visualize proteins. To identify DACH1-associated proteins, direct sequencing of coprecipitated proteins was conducted. All surfaces and utensils were cleaned with methanol/water. Inside a chemical hood, one-dimensional SDS-PAGE gel bands were cut from the gel and then diced into small pieces (∼1 mm³) and put into microcentrifuge tubes. Ammonium bicarbonate (25 mmol/L)/50% acetonitrile was added to the tubes to cover the pieces and incubated for 10 min with periodic vortexing. The supernatant was removed and discarded. These two steps were repeated until the Coomassie stain was removed. Acetonitrile (100%) was added to each tube to cover the pieces and incubated for 5 min. The supernatant was removed and discarded, and the gel pieces were allowed to dry in the chemical hood. Twenty-five microliters of 10 mmol/L DTT in 25 mmol/L ammonium bicarbonate were added to each tube, vortexed, and then spun. The gel pieces were then incubated at 56°C for 1 h. The supernatant was removed and 25 μL of iodoacetamide were added to the gel pieces, vortexed, and then spun. The gel pieces were incubated for 45 min in thedark at room temperature. The supernatant was removed and discarded. Acetonitrile (100%) was added to cover the gel pieces and thenincubated for 5 min. The supernatant was removed and discarded and the gel pieces were dried in the chemical hood. Just before use, sequencing-grade modified trypsin (Promega) was diluted to 20 mg/mL and kept on ice. Trypsin was added to cover the gel pieces and incubated for 30 min on ice. After incubation and gel swelling, excess trypsin was removed and then 20 mmol/L ammonium bicarbonate was added to cover the gel pieces. The gel pieces were incubated 16 to 24 h at 37°C. The digest solution was transferred to clean microcentrifuged tubes, after which 50% acetonitrile/5% formic acid was added to cover the gel pieces. The gel pieces were incubated for 30 min and the supernatant was added to the previously collected digest solution. The collected digest solution was dried using a speed vacuum, resuspended in 0.1% trifluoroacetic acid, and analyzed by mass spectrometry (MS).

MS analysis. Mass spectra were acquired using an Applied Biosystems model 4800 MALDI TOF/TOF and 4000 Series Explorer software (version 3.5.1). Mass spectra at m/z 800 to 4,000 were acquired for each digested one-dimensional SDS-PAGE band using 1,000 laser shots. MS/MS spectra were derived from up to 2,000 laser shots depending on the fragmentation of the precursor ion. Fragmentation of the precursor ions was induced by the use of atmosphere as a collision gas with a pressure of 1 × 10⁻⁶ Torr and in MS/MS 2 kV mode. For peak detection, cluster area S/N optimization using a S/N threshold value of 30 was selected.

Peptide identifications. Peptide identifications were done using the MASCOT search engine (version 1.9, Matrix Science) via GPS Explorer (version 3.6). Each MS/MS spectrum was searched against the National Center for Biotechnology Information number database of human protein sequences. Protein identifications with confidence interval scores of >95% were retained. The MASCOT search parameters were as follows: methionine oxidation and carbamidomethylation were selected as variable modifications; 1 missed cleavage was allowed; precursor error tolerance of 100 ppm and MS/MS fragment tolerance of 0.3 Da were selected; and charge was set to +1.

Analyses of DACH1 and ERα abundance were conducted on normal and tumorous human breast samples. Cellular quantitative immunohistochemical staining was conducted using the AQUA-PM 2000 platform on human breast cancer samples (Fig. 1A). ERα expression was lost or reduced in DACH1-expressing breast cancer epithelial cells. In normal breast tissue, DACH1 localized in the nucleolus, whereas in breast cancer samples, DACH1 localized diffusely within either the nucleus or cytoplasm (Fig. 1B). Analysis of ductal carcinoma in situ lesions and early invasive breast cancer showed that nucleolar DACH1 distribution was lost in early invasive breast cancer (Fig. 1C and D). ERα expression levels were inversely correlated with DACH1 expression levels in breast cancer samples (Fig. 1C). Although the relative abundance of DACH1 was reduced in many of the breast cancer samples, further immunohistochemical staining of ∼2,000 breast cancer samples identified either nuclear or cytoplasmic DACH1 in 7% and 50%, respectively, of the samples. DACH1 expression (cytoplasmic) correlated inversely with ERα expression (data not shown).

In view of the overall inverse correlation between ERα and DACH1 expression in patients with breast cancer, we investigated the functional interactions between the two proteins. Immunoprecipitation-Western blotting was conducted to examine the physical interaction between ERα and DACH1. Cells were cotransfected with expression vectors encoding ERα and an NH₂-terminal FLAG epitope–tagged DACH1. Immunoprecipitation was conducted with an antibody to the FLAG epitope of the DACH1 protein and Western blotting was conducted with an antibody directed to the ERα. ERα wt bound to DACH1 wt in immunoprecipitation-Western blot analysis (Fig. 2A , lane 3). Addition of E₂ induced a modest increase in DACH1-associated ERα.

To analyze the domains of DACH1 required for binding to the ERα, a series of DACH1 mutants were transiently transfected with the wt ERα into HEK293T cells. The DS domain is known to bind HDAC3 and contribute to repression of c-Jun (32). Immunoprecipitation assays directed toward the DACH1 FLAG epitope coprecipitated with ERα (Fig. 2A). Deletion of the DACH1 DS domain did not reduce ERα binding; however, deletion of the DACH1 COOH terminus abolished DACH1-ERα binding (Fig. 2A). To determine the domains of the ERα required for binding to DACH1, four fragments that cover the full length of ERα were fused with glutathione S-transferase (GST). Equal amounts of GST proteins were incubated with in vitro translated ³⁵S-labeled DACH1 (Fig. 2B). The ERα 282–337 domain is required for binding DACH1. This domain of ERα is known to undergo posttranslational modification by acetylation, which can alter the binding affinity for coactivator and corepressor proteins (14, 25).

E₂ enhanced the ERα luciferase reporter activity ∼20-fold (Fig. 3A,, left). The expression of DACH1 inhibited ligand-dependent ERα activity by ∼60% (Fig. 3A,, left) and deletion of the conserved DS domain abrogated repression of ERα activity (Fig. 3A,, right). E₂ induces DNA synthesis in MCF-7 cells. MCF-7 cells were stably transfected with either MSCV-IRES-GFP or MSCV-IRES-DACH1-GFP. Control MCF-7 cells treated with E₂ (10 nmol/L) for 9 hours showed a 2-fold induction of DNA synthesis, measured by [³H]thymidine uptake, consistent with prior studies (35). DACH1 expression in MCF-7 cells decreased both basal and ligand-induced DNA synthesis (Fig. 3B). To further validate the effect of DACH1 on ERα activity, the pS2-Luc promoter reporter gene was used in MCF-7 cells transfected with DACH1. The pS2 gene was induced by ERα activity. Expression of DACH1 in MCF-7 cells inhibited the promoter activity of a 1-kb pS2 estrogen-responsive promoter fragment (Fig. 3C,, left). These findings are consistent with prior published studies in which the E₂-responsive cyclin D1 gene was induced by shRNA to DACH1 in MCF-7 cells (19, 36). DACH1 expression inhibited E₂-dependent induction of the endogenous pS2 gene (Fig. 3C , right). DACH1 expression was verified by quantitative reverse transcription-PCR (data not shown).

DACH1 inhibited ERE reporter gene activity and inhibited both E₂-induced and basal DNA synthesis. DACH1 has been shown to inhibit basal level expression of c-jun (37) and cyclin D1 (19). Because cyclin D1 and c-jun have both been shown to enhance basal DNA synthesis in MCF-7 human breast cancer cell lines, DACH1 repression of these proteins likely also contributes to the inhibition of basal DNA synthesis. The pS2 minimal ERE reporter was examined in the presence of E₂ or the E₂ antagonist ICI 182, 780. The E₂ antagonist reduced E₂-induced pS2 reporter activity by 80%, and additional ICI 182, 780 reduced activity a further 10% (Fig. 3D). These findings suggest that basal ERE reporter activity may reflect low levels of agonist-bound ERα activity.

ERα activity is modulated by coactivators and corepressors, which modulate local chromatin structure (13). To investigate the possibility that DACH1 physically associated with ERα coregulators, FLAG-tagged DACH1 was used as a molecular probe to identify co-associated proteins. Cell extracts were prepared and a high-flow-through anti-FLAG column was used to bind DACH1. 3xFLAG peptide was used to elute DACH1 and DACH1-associated proteins. The proteins of interest were excised and analyzed by MALDI MS/MS with subsequent analysis by peptide sequencing. The analysis showed the presence of a prominent band with a molecular weight of ∼160 kDa (Fig. 4A). Western blotting using an antibody to FLAG showed the presence of DACH1 in the 97-kDa peak. Peptide sequence analysis identified the ∼160-kDa (DACH1-associated) protein as PELP1. The physical association between PELP1 and DACH1 was further assessed by immunoprecipitation-Western blotting. Expression vectors encoding PELP1 and DACH1 or a COOH-terminal deletion mutant of DACH1 were transduced into HEK293T cells. PELP1 was identified by Western blotting in the DACH1 immunoprecipitation, but was not present in the cells transduced with PELP1 and the DACH1 construct deleted of the COOH terminus (Fig. 4B). To identify the domains of PELP1 required for the association with DACH1, series of PELP1 mutants were transfected into HEK293T cells and immunoprecipitation-Western blotting was conducted. The mean data from immunoprecipitation-Western blots were tabulated and a representative example is shown (Fig. 4C). Thus, DACH1 association requires a region of PELP1 at amino acids 400 to 600. Immunohistochemical analysis identified DACH1 and PELP1 ascolocalized in an intranuclear, extranucleolar distribution (Fig. 4D).

To further examine the direct binding of DACH1 to ERα and PELP1, in vitro translated ³⁵S-labeled Myc-DACH1 and ERα were incubated in the presence of E₂ or vehicle and immunoprecipitated with anti-Myc antibody (Fig. 5A). DACH1 directly bound ERα. In the same assay, in vitro translated ³⁵S-labeled T7-PELP1 was used to confirm direct binding between DACH1 and PELP1 (Fig. 5A). To examine further the interaction between ERα, PELP1, and DACH1, HEK293T cells were transfected with T7-tagged PELP1 and 3xFLAG-tagged ERα. Immunoprecipitation was done on 1 mg of whole-cell lysates using 30 μL of FLAG affinity gel. 6xHIS-DACH1 protein purified from bacteria was added to the immunoprecipitation reaction at increasing dose (0–5 μg). Increasing amounts of DACH1 in the presence of 10 nmol/L E₂ caused a dissociation of PELP1 from ERα (Fig. 5B).

To determine the role of DACH1 in regulating ERα occupancy in the context of local chromatin, chromatin immunoprecipitation assays were conducted using the pS2 gene promoter. PELP1 is known to associate in the context of local chromatin at the pS2 gene promoter. DACH1 occupied the ERE of the pS2 promoter in chromatin immunoprecipitation assay (Fig. 6A). DACH1 is known to recruit HDAC1 and NCoR corepressor complexes in the context of local chromatin. Chromatin immunoprecipitation assays at the pS2 promoter identified with DACH1 the presence of PELP1, ERα, HDAC1, and histone H3. The presence of E₂ increased ERα abundance and reduced the relative binding of HDAC1 in chromatin immunoprecipitation at the ERE (Fig. 6A). Quantitative analysis of multiplicate experiments showed that treatment correlated with an increased abundance of DACH1 at the ERE of the pS2 promoter by ∼50% and reduced abundance of HDAC1 (∼50%). To determine the importance of PELP1 in HDAC recruitment to the ERE, MCF-7 cells were transduced with PELP1 shRNA (Fig. 6B). HDAC1 chromatin immunoprecipitation analysis showed that PELP1 shRNA did not affect the basal abundance of HDAC1 at the ERE. The E₂-mediated reduction in HDAC1 occupancy was, however, reduced by PELP1 shRNA. Thus, PELP1 shRNA increased HDAC1 occupancy at an ERE in the presence of E₂ (Fig. 6B , line 7 versus line 8).

To further examine the functional interaction between PELP1 and DACH1 at an ERE, a luciferase reporter gene was deployed. DACH1 repressed the transcription of the estrogen-responsive pS2 gene by ∼3 fold. E₂ enhanced ERE luciferase activity and cotransfection of PELP1 enhanced ERE activity in the presence of ligand (Fig. 6C). DACH1 repression of ERE activity was abrogated by coexpression of PELP1.

Collectively, these studies are consistent with a model in which DACH1 functions as an endogenous inhibitor of ERα function and transcriptional activity. DACH1 physically associates with PELP1 through the COOH terminus of DACH1 (Supplementary Fig. S1). In the presence of E₂, the DACH1-associated repressor complex dissociates from the endogenous ERα, resulting in an ERα/PELP1 coactivator complex and the induction of estrogen-dependent gene expression (Fig. 6D).

A comprehensive body of evidence has shown the importance of estrogen activity in the onset and progression of human breast cancer (1). Tamoxifen treatment, which inhibits ERα function, reduces the rate of onset of breast cancer in patients with BRCA2 mutations (38). Inhibition of ERα by aromatase treatment improves prognosis in patients with ERα-positive breast cancer. The constitutive endogenous inhibitors of ERα signaling, which are lost in human breast cancer, represent an ideal new therapeutic target. The ERα corepressor complex encodes a NCoR/HDAC1/BRCA1 complex (39); however, there is no evidence that the abundance of NCoR or HDAC is reduced during human breast cancer progression. In the current studies, DACH1, the expression of which is lost during breast cancer progression (19), was identified as a repressor of endogenous ERα activity.

Using a proteomic approach, DACH1 was found to bind the ERα coactivator PELP1. ERα is predominantly nuclear in the absence of ligand (8, 40). PELP1 binds the ERα in a ligand-dependent manner. Herein DACH1 inhibited the PELP1/ERα binding interaction and repressed PELP1-dependent activation of ERα activity. PELP1 bound to the COOH terminus of DACH1. DACH1 encodes a conserved Dac domain 1 (DS), which binds HDAC1/HDAC3/c-Jun and is conserved from Drosophila to humans (32, 41). The CREB binding protein coactivator also binds DACH1 (19, 32, 42, 43). The internal deletion of the DACH DS domain abrogated repression of ligand-dependent ERα activity. Thus, the recruitment of corepressor complexes by DACH1 is required for inhibition of ERα activity.

Immunohistochemical staining of normal mammary epithelial cells showed that DACH1 localizes to the nucleus, primarily within the nucleolus. In breast cancer cell lines, DACH1 distribution is primarily nuclear, and DACH1 and PELP1 colocalize in ∼80% of cells (Fig. 4D). These studies suggest that the relative balance of DACH1 and PELP1 in breast cancer cells modulates ERα signaling. PELP1 expression is increased in human breast cancer compared with normal human breast epithelium (16) and a substantial number of breast tumors coexpress PELP1 and ERα (44). In support of this model, estrogen-induced DNA synthesis and basal DNA synthesis were inhibited by DACH1 expression. DACH1 inhibition of estrogen-dependent DNA synthesis may involve the cyclin D1 gene because cyclin D1 is induced by E₂ and the relative abundance of cyclin D1 governs basal and E₂-induced MCF-7 cell DNA synthesis (45). DACH1 is known to inhibit fibroblast DNA synthesis induced by serum addition (19). Furthermore, genetic deletion of the cyclin D1 gene and cyclin D1 siRNA studies showed a key role for cyclin D1 in DACH1-mediated inhibition of DNA synthesis in fibroblasts (19). PELP1 plays a role in cell cycle progression through regulation of cyclin D1. PELP1 alters local chromatin structure required for optimal transcriptional responses to ligand-bound nuclear receptors (46) and is recruited to the promoter of ERα target genes including pS2, IGF, and cyclin D1 (27).

The ERα binding coactivator protein PELP1 functions as a molecular scaffold to recruit a variety of coactivators and histone modifying proteins, and interacts with ERα through the LXXLL motif 1 to 5 (16). In addition to the ERα, PELP1 interacts with the androgen receptor, glucocorticoid receptor, and progesterone receptor in a ligand-dependent manner and with retinoid X receptor α (16, 47). ERα complexes, which include PELP1, participate in membrane-mediated signaling by stimulating Src kinase, cytokines, mitogen-activated protein kinase, phosphatidylinositol 3-kinase, and signal transducer and activator of transcription 3 (48). Estrogen promotes the interaction between the AF-2 domain of ERα and PELP1. Recent studies have suggested that PELP1 plays a role in promoting cellular proliferation and inhibiting apoptosis via the phosphatidylinositol 3-kinase/AKT pathway (49). Conversely, PELP1 functions as a corepressor of several transcription factors, including the DNA binding transcription factors, serum response factor, nuclear factor-κB, and activator protein-1 proteins (50). The finding that DACH1 binds to PELP1 implicates DACH1 in a broad variety of signal transduction pathways, which remain to be explored.

No potential conflicts of interest were disclosed.

Grant support: R01CA70896, R01CA75503, R01CA86072 (R.G. Pestell) and Susan Komen Breast Cancer Foundation grant BCTR0504227 (C. Wang). The Kimmel Cancer Center is supported by NIH Cancer Center Core grant P30CA56036 (R.G. Pestell). This project is funded by a generous grant from the Dr. Ralph and Marian C. Falk Medical Research Trust Foundation and the Margaret Q. Landenburger Foundation (K. Wu) and supported in part by a grant from the Pennsylvania Department of Health (R.G. Pestell and C. Wang). The Department disclaims responsibility for any analysis, interpretations, or conclusions.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Tim R. Zacharewski (Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI) for the pS2-LUC reporter and Atenssa L. Cheek for preparation of the manuscript.