Turning Off Estrogen Receptor β-Mediated Transcription Requires Estrogen-Dependent Receptor Proteolysis

Expression vectors.The ERα/β and CHIP expression plasmids and their deletion mutants have been described previously (48). ERE-TATA-Luc and ERE-TATA-LucCP were constructed by inserting three consensus estrogen response elements into a luciferase reporter plasmid encoding a TATA box, pGL3-Basic or phRG(R2.2)-Basic. pGL3-Basic, phRG(R2.2)-Basic, and pGL3-Control vectors were purchased from Promega (Madison, WI).

Antibodies.Anti-FLAG-M2 mouse monoclonal antibody (Sigma, St. Louis, MO), anti-hemagglutinin (HA) mouse monoclonal antibody (Roche), anti-human ERα antibody (Chemicon, Temecula, CA), anti-human ERβ antibody (PPMX), and anti-glyceraldehyde-3-phosphate-dehydrogenase mouse monoclonal antibody (American Research Products, Inc.) were used at appropriate dilutions, according to the manufacturers' instructions. A rabbit polyclonal antibody directed against CHIP was provided by T. Chiba (Tokyo Metropolitan Institute of Medical Science).

Cell culture and transfection.MDA-MB231 breast cancer cells and human embryonic kidney 293 cells were routinely maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Twenty-four hours before transfection, the medium was changed to phenol red-free DMEM containing 4% charcoal-stripped FBS. Transfection was performed using PerFectin transfection reagent (Gene Therapy Systems) or TransFast transfection reagent (Promega) according to the manufacturers' protocols. The cells were treated with or without MG132 (10⁻⁶ M) and estrogen (10⁻⁸ M). Twenty-four hours after the addition of estrogen, the cells were harvested and analyzed by Western blotting using the appropriate antibodies.

Coimmunoprecipitation and Western blotting.The 293 cells were transfected with the appropriate plasmids and lysed in TNE buffer (10 mM Tris-HCl [pH 7.8], 0.5% Nonidet P-40 [NP-40], 0.15 M NaCl, 1 mM EDTA, 1 μM phenylmethylsulfonyl fluoride [PMSF], 1 μg/ml aprotinin). Extracted proteins were immunoprecipitated with antibody-coated protein A/G-Sepharose (Amersham) or anti-FLAG M2 agarose (Sigma). The bound proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto polyvinylidene difluoride membranes (Millipore), and detected with the appropriate antibodies and secondary antibodies conjugated with horseradish peroxidase. Specific proteins were detected using an enhanced chemiluminescence Western blot detection system (Amersham).

Luciferase assay.For luciferase assays, ERE-TATA-Luc plasmids were cotransfected with expression vector encoding ERβ or its mutants. As a reference plasmid with which to normalize transfection efficiency, the pRSVβGAL vector or the pRL-CMV vector was cotransfected in all experiments. Twenty-four hours after transfection, the culture medium was replaced with fresh medium containing 0.2% FBS, either estrogen (10⁻⁸ M) or ethanol vehicle was added, and the cells were incubated for an additional 24 h. Cell extracts were prepared, and luciferase assays were performed following the manufacturer's protocol (Promega). β-Galactosidase activity was measured to control for the efficiency of each transfection. Individual transfections, each consisting of triplicate wells, were repeated at least three times.

Protein purification.Glutathione S-transferase (GST) protein or an immobilized 37-amino-acid region that we designated the “domain required for degradation” (DRD) fused to GST (GST-DRD) was preincubated for 1 h at 4°C in GST-binding buffer (20 mM Tris-HCl [pH 7.9], 180 mM KCl, 0.2 mM EDTA, 0.5 mM PMSF, and 1 mM dithiothreitol) containing bovine serum albumin (BSA; 1 mg/ml). Bead-immobilized proteins were then incubated at 4°C for 6 to 10 h with 293 cell extracts. After the beads had been washed three times with GST buffer (GST-binding buffer with 0.1% NP-40), they were washed again with GST buffer containing 0.2% N-lauroyl sarcosine. Proteins bound to the DRD were eluted with 15 mM reduced glutathione in elution buffer (50 mM Tris-HCl [pH 8.3], 150 mM KCl, 0.5 mM EDTA, 0.5 mM PMSF, 5 mM NaF, 0.08% NP-40, 0.5 mg/ml BSA, and 10% glycerol). Recombinant CHIP was expressed in Escherichia coli and purified using Ni-agarose.

Immunofluorescence.The 293 cells were grown on poly-l-lysine-coated 12-well culture dishes and transfected with plasmids. Twenty-four hours after transfection, the cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 10 min and permeabilized with Triton buffer (50 mM Tris-HCl [pH 7.5], 0.5% Triton X-100, 150 mM NaCl, 2 mM EDTA) for 15 min. The cells in each well were blocked with PBS containing 1% BSA and 0.5% goat serum for 3 h at 37°C. The cells were incubated with the appropriate antibody in PBS containing 1% BSA for 2 h at 37°C. After the cells had been washed with PBS, they were incubated with Alexa-Fluor-488-conjugated goat anti-rat immunoglobulin G or Alexa-Fluor-594-conjugated goat anti-mouse immunoglobulin G (Molecular Probes) for 1 h at 37°C and washed with PBS. The samples were mounted with Vectashield mounting medium (Vector Laboratories) and analyzed using a Keyence Biozero BZ-8000.

Pulse-chasing.The 293 cells were transfected with ERβ, and 48 h after transfection the cells were labeled for 30 min at 37°C with [³⁵S]methionine (50 μCi/ml) in methionine-free DMEM. The cells were then washed twice and incubated in DMEM containing 0.2% FBS for the indicated periods (chase). At each time point, cell lysates were immunoprecipitated with anti-ERβ antibody. The immunoprecipitates were resolved by SDS-PAGE and visualized by autoradiography. A phosphorimager was used to quantify the metabolically labeled ERβ present at each time point.

RNAi.The 293gp cells were transfected with pSINsi-hU6 vector (Takara) containing the target sequence of CHIP or LacZ (control) and with vesicular stomatitis virus-encoded envelope protein G to generate the small interfering RNA (siRNA) retroviral supernatant. MDA-MB-231 cells were transfected with retroviral supernatant in the presence of Polybrene (8 μg/ml). Twenty-four hours after transfection, the viral supernatant was replaced with fresh DMEM containing 10% FBS. The transfected cells were selected with G418 (1 mg/ml). The target sequences were 5′-GCACGACAAGTACATGGCGGA-3′ for CHIP and 5′-GCTACACAAATCAGCGATT-3′ for LacZ.

An N-terminal 37-amino-acid region is essential for ERβ degradation.The level of ERβ protein was reduced by the addition of estrogen to the MDA-MB-231 breast cancer cell line. This degradation was inhibited by the proteasome inhibitors MG132 (Fig. 1A) and lactacystin (data not shown). In a ubiquitination assay, ERβ was ubiquitinated in both the presence and the absence of estrogen (Fig. 1B). Thus, ERβ is degraded via ubiquitin-proteasome pathways. Preliminary biochemical study showed that a single K48-linked polyubiquitin chain is sufficient to target a substrate to the 26S proteasome. Therefore, we investigated the ubiquitin linage of the chains using a ubiquitin mutant lacking specific lysines, Ub(K48R). HA-tagged ubiquitin (HA-Ub) or Ub(K48R) was transfected into 293 cells, and the level of ERβ protein and the ubiquitination status were determined by Western blotting using an anti-FLAG-M2 or an anti-HA antibody. Expression of Ub(K48R) reduced the degradation (Fig. 1C) and the ubiquitination status (Fig. 1D) of ERβ, suggesting that the polyubiquitin chains of ERβ are linked via K48. These results indicate that whereas ERβ is polyubiquitinated via K48 in ubiquitin in both the presence and the absence of estrogen, the proteolysis is estrogen dependent. Our observations imply that ubiquitination in itself is not the only way to regulate ERβ.

• Open in new tab
• Download powerpoint

FIG. 1.

ERβ is degraded through ubiquitin-proteasome pathways. (A) ERβ is degraded in an estrogen-dependent manner. ERβ-transfected MDA-MB-231 cells were cultured in the presence or the absence of estrogen (10⁻⁸ M) and the proteasome inhibitor MG132 (10⁻⁶ M). ERβ level was analyzed by use of Western blots probed with the indicated antibody. (B) ERβ is degraded via ubiquitin-proteasome pathways. ERβ-transfected MDA-MB-231 cells were cultured in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). ERβ was immunoprecipitated (IP) using anti-ERβ antibody. Upper panel: the ubiquitination status of ERβ was analyzed by Western blotting (IB) and probed with an anti-ubiquitin antibody. Lower panel: immunoprecipitated ERβ was detected by Western blotting and probed with an anti-ERβ antibody. IgG, immunoglobulin G; M, molecular mass. (C) The estrogen-dependent degradation of ERβ is inhibited by mutated ubiquitin Ub(K48R). FLAG-tagged ERβ and DNA encoding HA-Ub (1,000 ng) or HA-Ub(K48R) (1,000 ng) were transfected into 293 cells in the presence or the absence of estrogen (10⁻⁸ M). ERβ was detected by use of Western blots probed with anti-FLAG-M2 antibody. (D) The polyubiquitin chain of ERβ is linked via K48. FLAG-tagged ERβ and DNA encoding HA-Ub (1,000 ng) or HA-Ub(K48R) (1,000 ng) were transfected into 293 cells in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). FLAG-tagged ERβ was immunoprecipitated with anti-FLAG-M2 antibody. Upper panel: the ubiquitination status of ERβ was analyzed by use of Western blots probed with anti-HA antibody. Lower panel: immunoprecipitated ERβ was detected by use of Western blots probed with anti-FLAG antibody. (E) The A/B domain is necessary for the estrogen-dependent degradation of ERβ. DNA (500 ng) encoding the indicated FLAG-tagged ER deletion mutants was transfected into 293 cells. These cells were cultured in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). To evaluate the protein levels of the ER mutants, Western blot analysis was performed using anti-FLAG-M2 antibody.

To investigate the molecular mechanism by which ERβ is degraded, DNA sequences encoding truncated forms of ERα or ERβ were transfected into 293 cells, and the protein levels were examined by Western blot analysis. ERα(Δ181), which does not contain the A/B domain, still underwent ligand-dependent degradation (Fig. 1E). In contrast, truncation of the A/B domain of ERβ, forming ERβ(Δ140), produced no ligand-dependent degradation, suggesting that the A/B domain of ERβ is necessary for its ligand-dependent degradation (Fig. 1E).

We then investigated the region responsible for the degradation of ERβ. The levels of truncated ERβ protein in the presence and the absence of estrogen were examined by Western blotting. As shown in Fig. 2A, ERβ(Δ20), in which the 20 N-terminal amino acids are deleted, was elevated in the presence of estrogen compared with wild-type ERβ. ERβ(Δ37) exhibited almost no degradation, indicating that the N-terminal 37 amino acids are essential for the ligand-dependent degradation of ERβ (Fig. 2A). To assess whether this region is sufficient for the ligand-dependent degradation of ERβ, it was fused to ERβ(Δ140), which does not contain the A/B domain and shows little ligand-dependent degradation. In contrast to that of ERβ(Δ140), the level of ERβ(Δ38-140) protein, in which the 37-amino-acid region is fused to ERβ(Δ140), was reduced in an estrogen-dependent manner (Fig. 2A). This reduction was abrogated by the addition of MG132, indicating that ERβ(Δ38-140) is degraded by proteasomal pathways. Interestingly, the degradation of ERβ(mC), which has three amino acid substitutions in the DNA-binding domain (C domain) and has almost no ability to bind DNA (30), was similar to that of wild-type ERβ (Fig. 2A), indicating that binding to the DNA element is not necessary for ligand-dependent receptor degradation. Similar results were obtained when other cell lines, such as MCF-7 or MDA-MB-231, were used (data not shown). From these observations, we concluded that the N-terminal 37-amino-acid region of ERβ is involved in its ligand-dependent degradation and designated this region the DRD. We next determined the ubiquitination status of the truncated forms of ERβ. The DRD deletion accumulated a monoubiquitinated form and showed a reduced level of the polyubiquitinated form (Fig. 2B), suggesting that the DRD is required for the polyubiquitination of ERβ.

• Open in new tab
• Download powerpoint

FIG. 2.

DRD-mediated degradation is involved in the downregulation of ERβ-mediated transcription. (A) The N-terminal 37-amino-acid region is essential for estrogen-mediated ERβ degradation. DNA sequences (500 ng) encoding the indicated FLAG-tagged ERβ deletion mutants were transfected into 293 cells. These cells were cultured in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). To evaluate the protein levels of the ERβ mutants, Western blot analysis was performed using an anti-FLAG-M2 antibody. (B) DRD-regulated ubiquitination of ERβ. DNA (500 ng) encoding the indicated FLAG-tagged ERβ mutants and DNA (1,000 ng) encoding HA-tagged ubiquitin were transfected into 293 cells in the presence or the absence of estrogen (10⁻⁸ M). Upper panel: FLAG-tagged ERβ mutants were immunoprecipitated (IP) with anti-FLAG-M2 antibody. The ubiquitination status of ERβ was analyzed by use of Western blots probed with anti-HA antibody. Lower panel: immunoprecipitated ERβ mutants were detected by use of Western blots probed with an anti-FLAG-M2 antibody. I.B., Western blotting.

Purification and identification of an E3 ubiquitin ligase that specifically binds to the DRD of ERβ.To investigate why the DRD induces ubiquitination, we examined two possibilities. First, the DRD may contain a lysine residue that is ubiquitinated by a ubiquitin ligase specific for ERβ. Second, there may be a ubiquitin ligase that specifically recognizes and binds to the DRD to ubiquitinate ERβ.

To test the first possibility, we introduced an amino acid substitution at the lysine residue in the DRD. ERβ(K4A), in which the lysine residue in the DRD is replaced by alanine, showed ligand-dependent degradation similar to that of the wild type (Fig. 2A), implying that the lysine residue in the DRD is not a target for ubiquitination. Therefore, we investigated the other possibility.

We generated a GST-fused 37-amino-acid region (GST-DRD) with which to purify proteins that specifically bind to the DRD. Whole-cell extracts prepared from 293 cells were incubated with either GST-DRD or GST. The bound proteins were precipitated with glutathione beads and separated using SDS-PAGE (Fig. 3A). Mass fingerprinting methods revealed that the proteins that specifically bound to the DRD were the carboxyl terminus of CHIP, HSP70, and HSP90 (Fig. 3A).

• Open in new tab
• Download powerpoint

FIG. 3.

The DRD associates with a protein complex containing the carboxyl terminus of HSC70-interacting protein (CHIP) and HSC/HSP70. (A) Purification and identification of DRD-interacting proteins. Whole-cell extracts (WCE) prepared from 293 cells were incubated with GST-DRD or with GST as the control. Bound proteins were precipitated with GST-conjugated beads, eluted with glutathione, and subjected to SDS-PAGE followed by silver staining. The fractions eluted from GST and GST-DRD are shown. Proteins eluted from the DRD peptide were examined by mass spectrometry. (B) CHIP binds the DRD. Recombinant CHIP was incubated with GST-DRD or GST. The bound proteins were precipitated with glutathione beads. CHIP bound to GST-DRD was separated by SDS-PAGE. (C) Interaction between ERβ and CHIP in vivo. ERβ-transfected MDA-MB-231 cells were lysed and immunoprecipitated (I.P.) with anti-ERβ antibody in the presence or the absence of estrogen (10⁻⁸ M). The precipitates were Western blotted (I.B.) and probed with antibodies directed against CHIP and HSP70. IgG, immunoglobulin G. (D) DRD is necessary for the interaction between ERβ and CHIP. ERβ or ERβΔDRD was transfected into 293 cells. These cells were lysed and immunoprecipitated with anti-FLAG-M2 antibody. The precipitates were Western blotted and probed with antibodies directed against CHIP and HSP70. (E) CHIP induces the ubiquitination of ERβ. DNA (500 ng) encoding FLAG-tagged ERβ and DNA (1,000 ng) encoding HA-tagged ubiquitin were transfected into 293 cells with or without DNA (50 ng) encoding CHIP in the presence or the absence of estrogen (10⁻⁸ M). FLAG-tagged ERβ or ERβ(ΔDRD) was immunoprecipitated using anti-FLAG-M2 antibody. The ubiquitination status of ERβ was analyzed by use of Western blots probed with anti-HA antibody.

CHIP has E3 ubiquitin ligase activity mediated by its carboxyl-terminal U-box domain and has the ability to bind to chaperones HSP/HSC70 via its tetratricopeptide repeat (TPR) domain (7, 48). A pull-down assay showed that recombinant CHIP bound to the GST-DRD (Fig. 3B). A coimmunoprecipitation assay demonstrated specific binding of CHIP and HSP70 to ERβ in both the absence and the presence of estrogen (Fig. 3C). ERβΔDRD showed no ability to bind CHIP (Fig. 3D). Therefore, we tested whether CHIP enhances ERβ ubiquitination. When ERβ was coexpressed with CHIP, smeary bands of ubiquitin-conjugated ERβ products were observed (Fig. 3E). In contrast, CHIP expression had no effect on the ubiquitination status of ERβ(ΔDRD). Thus, CHIP appears to mediate the ubiquitination of ERβ.

CHIP participates in the degradation of ERβ.ERβ is known to localize predominantly in the nucleus. Although a large proportion of the CHIP protein was localized in the cytoplasm, a small amount was observed in the nucleus by immunostaining (Fig. 4A) and Western blot analysis (Fig. 4B, left panel). To test whether CHIP-mediated ubiquitination induces the degradation of ERβ, DNA encoding ERβ was transfected into 293 cells with or without DNA encoding CHIP. Western blot analysis revealed that the steady-state level of ERβ decreased in both the nucleus and the cytoplasm when CHIP was produced (Fig. 4B, left panel). ERβ in MDA-MB-231 cells also decreased with the production of CHIP (Fig. 4B, right panel). We further determined CHIP's function by producing MDA-MB-231 cells in which endogenous CHIP expression was suppressed by the introduction of a siRNA complementary to sequences present in the CHIP mRNA. The introduction of the siRNA vector into MDA-MB-231 cells suppressed the expression of CHIP mRNA (data not shown) and protein (Fig. 4C, left panel) and the accumulation of ERβ protein (Fig. 4C, right panel). In contrast, a control vector failed to alter the CHIP or ERβ protein levels (Fig. 4C, right panel). Thus, CHIP appears to be involved in estrogen-dependent ERβ degradation. To confirm this, pulse-chase experiments were performed. In the presence of estrogen, the half-life of ERβ exceeded 8 h, but it had a half-life of about 4 h when CHIP was expressed (Fig. 4D).

• Open in new tab
• Download powerpoint

FIG. 4.

CHIP enhances the degradation of ERβ. (A) CHIP protein was observed in both the cytoplasm and the nucleus. The 293 cells were transiently transfected with plasmids encoding HA-tagged CHIP and FLAG-tagged ERβ. Mounted cells were examined by immunofluorescence microscopy as described in Materials and Methods. (a) The distribution of CHIP in the cell is shown. (b) The distribution of ERβ is shown. (c) Merged images of panels a and b. (B) CHIP expression reduced the steady-state level of ERβ. DNA (50 ng) encoding HA-tagged CHIP was cotransfected into 293 or MDA-MB-231 cells with DNA (500 ng) encoding FLAG-tagged ERβ in the absence or the presence of estrogen (10⁻⁸ M). The level of ERβ protein was examined by use of Western blots probed with anti-FLAG (293) or anti-ERβ (MDA-MB-231) antibody. (C) ERβ accumulated after the siRNA-mediated suppression of endogenous CHIP. Plasmids encoding a siRNA specific for CHIP or the control vector were introduced into MDA-MB-231 cells. These cells were selected with G418. Levels of CHIP and ERβ proteins were assessed by immunoblotting the whole-cell lysates with specific antibodies, as indicated. (D) CHIP facilitates the degradation of ERβ. 293 or MDA-MB-231 cells were transfected with DNA (50 ng) encoding CHIP or with DNA (500 ng) encoding FLAG-tagged ERβ in the presence of estrogen (10⁻⁸ M). Cells were pulse-labeled with [³⁵S]methionine and then chased for the indicated times in medium containing unlabeled methionine. ³⁵S-labeled ERβ in the immunoprecipitate was quantified by phosphorimaging, and the levels in control cells (closed circles) and CHIP-expressing cells (open circles) were plotted relative to the amount present at time zero.

To confirm that the DRD is necessary for CHIP-dependent ERβ degradation, DNA sequences encoding truncated forms of ERβ were transfected into 293 cells with or without DNA encoding CHIP. CHIP expression enhanced the degradation of ERβ, ERβ(mC), ERβ(DRD;Δ140), and ERβ(AF-1), all of which contain the DRD, but did not enhance the degradation of ERβ(ΔDRD) or ERβ(Δ140), neither of which contains the DRD (Fig. 5A). To test the specificity of this effect, we created constructs in which the TPR and U-box domains of CHIP were deleted (ΔTPR and ΔU-box, respectively). CHIP binds to HSP/HSC70 via its TPR motif and displays E3 ubiquitin ligase activity mediated by its U-box domain. Although the expression of these truncated proteins was similar to that of full-length CHIP (data not shown), deletion of either of these domains abolished the effects of CHIP on ERβ (Fig. 5B). Such a requirement for a TPR motif indicates that CHIP may have to interact with HSP/HSC70 to promote ERβ degradation. The functional requirement for the U-box confirms that CHIP regulates ERβ ubiquitination.

• Open in new tab
• Download powerpoint

FIG. 5.

CHIP regulates DRD-mediated ubiquitination and degradation. (A) The DRD is necessary for CHIP-dependent ERβ degradation. DNA (500 ng) encoding the indicated FLAG-tagged ERβ mutants was transfected into 293 cells, with or without DNA encoding CHIP. These cells were cultured in the presence or the absence of estrogen (10⁻⁸ M). To evaluate the levels of ERβ mutant proteins, Western blot analysis was performed using anti-FLAG-M2 antibody. (B) Both the TPR and U-box domain of CHIP are necessary for ERβ degradation. DNA (50 ng) encoding CHIP, CHIPΔTPR, or CHIPΔUbox was transfected into 293 cells with or without DNA (500 ng) encoding FLAG-tagged ERβ. Levels of ERβ protein were examined by use of Western blots probed with anti-FLAG-M2 antibody. (C) The CHIP-dependent ERβ degradation is not affected by ERα. DNA (500 ng) encoding FLAG-tagged ERβ was transfected into 293 cells with or without DNA (500 ng) encoding FLAG-tagged ERα and DNA (50 ng) encoding CHIP. Levels of ERβ and ERα protein were examined by use of Western blots probed with anti-FLAG-M2 antibody.

In a previous report, we showed that CHIP binds ERα in the absence of estrogen (48). Since ERα and ERβ heterodimerize readily in the presence of each other, we examined the effect of cointroducing ERα on CHIP-dependent degradation of ERβ. As shown in Fig. 5C, the degradation of ERβ was not affected by coexpression of ERα.

The F domain regulates ERβ degradation negatively.Our results showed that whereas CHIP binds and ubiquitinates ERβ in both the presence and the absence of estrogen, the degradation of ERβ is estrogen dependent. To investigate this discrepancy, DNA sequences encoding truncated forms of ERs were transfected into 293 cells and the protein levels were examined. The AD core domain is necessary for estrogen-dependent degradation of ERα (29, 48). Consistent with this observation, ERα(ΔAD) was stabilized by ligand binding and thus accumulated in response to estrogen (Fig. 6A). Interestingly, the deletion of the AD core of ERβ resulted in a dramatic reduction in ERβ in the absence of estrogen (Fig. 6A). This observation raises the possibility that the C-terminal region of ERβ is necessary for the stabilization of ERβ. To test this possibility, DNA sequences encoding deletion mutants of the C terminus of ERβ were transfected into 293 cells and levels of expressed protein were determined by Western blot analysis. When the F domain was deleted [ERβ(ΔF)], a dramatic reduction in ERβ protein was observed in the absence of estrogen (Fig. 6B). This reduction in ERβ was inhibited by MG132, indicating that the F domain suppresses the proteasome-dependent degradation of ERβ. The degradation of ERβ(ΔF) required the DRD, because ERβ(ΔDRD;ΔF) was not degraded in the absence of estrogen. In contrast, the degradation of ERβ(Δ140;ΔF) was enhanced by fusion with the DRD [ERβ(DRD;Δ140;ΔF)] in both the presence and the absence of estrogen (Fig. 6B). These results suggest that the F domain suppresses the DRD-dependent degradation of ERβ in the absence of estrogen.

• Open in new tab
• Download powerpoint

FIG. 6.

The F domain suppresses DRD-mediated proteolysis of ERβ. (A) AD core deletion in ERβ increased the degradation of ligand-unbound receptor. DNA sequences (500 ng) encoding the indicated FLAG-tagged ER deletion mutants were transfected into 293 cells. These cells were cultured in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). To evaluate the protein levels of the ER mutants, Western blot analysis was performed using anti-FLAG-M2 antibody. (B) F domain protects ERβ from DRD-mediated degradation. DNA sequences (500 ng) encoding the indicated FLAG-tagged ERβ deletion mutants were transfected into 293 cells cultured in the presence or the absence of estrogen (10⁻⁸ M) and MG132 (10⁻⁶ M). To evaluate the protein levels of the ERβ mutants, Western blot analysis was performed using anti-FLAG-M2 antibody.

In contrast to what was seen for ERβ, the deletion of the F domain in ERα did not affect its degradation. To study this, we generated a chimeric ERβ, ERβ (αF), in which the F domain was substituted with an ERα F domain. However, there is almost no homology between the ERα F and ERβ F domains, and the F domain of ERα was able to protect ERβ from proteolysis in the absence of estrogen (Fig. 6B).

Next, we determined the ubiquitination statuses of ERβ and ERβ(ΔF). That of ERβ(ΔF) was similar to that of full-length ERβ (Fig. 7A), indicating that the F domain does not regulate the ubiquitination step of ERβ. Thus, the F domain may inhibit the accession of 26S proteasome to ERβ structurally. Therefore, we evaluated the binding between ERβ and proteasomes. FLAG-tagged ERβ or ERβ(ΔF) was transfected into 293 cells with SUG-1, which is one of the components of regulatory subunit of 26S proteasome. A coimmunoprecipitation experiment revealed that deletion of the F domain enhanced association between ERβ and SUG-1 (Fig. 7B). In addition, SUG-1 efficiently reduced ERβ(ΔF) rather than ERβ (Fig. 7C). Thus, in the absence of estrogen, the F domain appears to abrogate the binding of the 26S proteasome to ERβ to protect the receptor from proteolysis.

• Open in new tab
• Download powerpoint

FIG. 7.

F domain abrogates the association of proteasomes with ERβ. (A) Deletion of the F domain does not alter the ubiquitination status of ERβ. DNA sequences (500 ng) encoding the indicated FLAG-tagged ERβ mutants and DNA (1,000 ng) encoding HA-tagged ubiquitin were transfected into 293 cells in the presence or the absence of estrogen (10⁻⁸ M). Upper panel: FLAG-tagged ERβ mutants were immunoprecipitated (I.P.) with anti-FLAG-M2 antibody. The ubiquitination status of ERβ was analyzed by use of Western blots probed with anti-HA antibody. I.B., Western blotting. Lower panel: immunoprecipitated ERβ mutants were detected by use of Western blots probed with anti-FLAG antibody. (B) The F domain abrogates the association of SUG-1 with ERβ. DNA sequences (500 ng) encoding either ERβ or ERβ(ΔF) were transfected into 293 cells. FLAG-tagged ERβ or ERβ mutants were immunoprecipitated with anti-FLAG-M2 antibody. The precipitates were Western blotted and probed with an antibody against SUG-1. T.F., transfection. (C) The F domain protects ERβ from SUG-1-dependent degradation. DNA sequences (100 ng) encoding SUG-1 were cotransfected into 293 cells with or without DNA sequences (500 ng) encoding FLAG-tagged ERβ or ERβ(ΔF). The levels of ERβ or ERβ(ΔF) protein were examined by Western blotting and probed with anti-FLAG antibody.

Receptor degradation is involved in the downregulation of transcription of the gene for ERβ after ligand withdrawal.Recent evidence indicates that receptor degradation is coupled to transactivation and that the transcriptional activity of the gene for ERα is abrogated by the addition of MG132. The transcriptional activity of ERβ was also reduced by MG132 (Fig. 8A). Thus, it is likely that receptor degradation is necessary for the ligand-dependent transactivation of the ERβ gene. To investigate the association between CHIP-mediated receptor degradation and transactivation, we evaluated the transcriptional activities of ERβ deletion mutants by use of a transient reporter assay. As shown in Fig. 8B, ERβ(ΔDRD) exhibited a transcriptional activity higher than that of full-length ERβ. However, ERβ(Δ140) still showed transcriptional activity. When the DRD was fused to ERβ(Δ140) to form ERβ(DRD;Δ140), its transcriptional activity was reduced (Fig. 8B). Similarly, ERβ(Δ140; ΔF) exhibited higher transcriptional activity than ERβ(ΔF), but fusion of DRD decreased its transcriptional activity (Fig. 8B). In agreement with these results, CHIP expression downregulated the transcriptional activity of the ERβ gene (Fig. 8C). In contrast, CHIP siRNA enhanced transcriptional activity (Fig. 8C). The transcriptional activity of ERβ(ΔDRD) was not affected by expression of CHIP (Fig. 8C). Overall, our results indicate that CHIP-mediated degradation is not necessary for receptor-mediated transactivation. In addition, CHIP is able to ubiquitinate and to degrade ERβ(mC) (Fig. 5A), indicating that binding to the DNA element is not necessary for CHIP-mediated receptor degradation. These results show that large amounts of liganded ERβ protein are selectively and rapidly degraded independently of receptor-mediated transcription by CHIP.

• Open in new tab
• Download powerpoint

FIG. 8.

CHIP decreases transcriptional activity of ERβ by degradation. (A) Proteasome-dependent degradation is necessary for the transactivation of ERs. DNA sequences (0.05 ng) encoding the ERβ or ERα were cotransfected into 293 cells with ERE-TATA-Luc (50 ng) and pRSVβGAL (50 ng) in the presence or absence of MG132 (10⁻⁶ M), and the cell extracts were used in a luciferase assay. (B) DRD-mediated degradation is not coupled to ligand-dependent receptor transactivation. DNA sequences (0.05 ng) encoding the indicated ERβ mutants were cotransfected into 293 cells with ERE-TATA-Luc (50 ng) and pRSVβGAL (50 ng), and the cell extracts were used in a luciferase assay. (C) CHIP production reduced the transcriptional activity of ERβ. ERE-TATA-Luc (50 ng), pRSVβGAL (50 ng), and DNA sequences (0.05 ng) encoding either ERβ or ERβ(ΔDRD) were cotransfected into 293 cells with or without DNA encoding CHIP and CHIP RNAi, and the cell extracts were used in a luciferase assay.

To investigate the biological significance of this “wastefulness,” we hypothesized that CHIP-mediated degradation selectively and rapidly degrades the active form of ERβ to shut off ERβ-mediated transcription when the estrogen dose is reduced. To test this hypothesis, we measured the transcriptional activities of constructs encoding ERβ and ERβ(ΔDRD) after estrogen withdrawal. DNA sequences encoding either ERβ or ERβ(ΔDRD) were transfected into 293 cells together with a plasmid in which the PEST sequence was fused to the luciferase gene (ERE-TATA-LucCP), and the luciferase activity was measured at various time points after estrogen withdrawal. As shown in Fig. 9A, the transcriptional activity of the ERβ sequence was reduced much faster than that of ERβ(ΔDRD). Using MDA-MB-231 cells that stably express either control siRNA or CHIP siRNA, we tested whether CHIP is involved in the downregulation of ERβ-mediated transcription after reduction of the ligand dose. Plasmids containing the PEST-fused luciferase gene were transfected into MDA-MB-231 cells expressing either control siRNA or CHIP siRNA. ERβ gene-mediated transcription was downregulated faster in the MDA-MB-231 cells expressing control siRNA than in the cells expressing CHIP siRNA (Fig. 9A).

• Open in new tab
• Download powerpoint

FIG. 9.

CHIP switches off ERβ-mediated transcription after the estrogen dose is reduced. (A) Downregulation of CHIP inhibits the turning off of ERβ-mediated transcription after estrogen withdrawal. Upper panel: DRD-mediated degradation is involved in the downregulation of ERβ-mediated transcription after estrogen withdrawal. The 293 cells were transfected with expression plasmids encoding FLAG-tagged ERβ or ERβ(ΔDRD), together with ERE-TATA-LucCP (50 ng) and pRSVβGAL (50 ng). Twenty-four hours after transfection, the cells were treated with estrogen (10⁻⁸ M) for an additional 24 h, after which the estrogen was withdrawn (time 0). These cells were harvested for luciferase assays and Western blotting at the indicated time points after estrogen withdrawal. Lower panel: CHIP is involved in the downregulation of ERβ-mediated transcription when the ligand dose is reduced. Plasmids encoding a siRNA specific for CHIP or control vector were introduced into MDA-MB-231 cells. These cells were selected with G418. FLAG-tagged ERβ (100 ng), ERE-TATA-LucCP (400 ng), and pGL3-Control (10 ng) were transfected into these cell lines. Twenty-four hours after transfection, the cells were treated with estrogen (10⁻⁸ M) for an additional 24 h, after which the estrogen was withdrawn (time 0). These cells were harvested for luciferase assays and Western blotting at the times indicated after estrogen withdrawal. (B) Dose-response curve of the estrogen-induced CHIP-dependent degradation of the ERβ. DNA (500 ng) encoding the FLAG-tagged ERβ was transfected into 293 cells, with or without DNA encoding CHIP. These cells were cultured in the presence of the indicated concentration (conc.) of estrogen (from 10⁻¹¹ M to 10⁻⁷ M). Levels of ERβ protein were examined by use of Western blots probed with anti-FLAG-M2 antibody.

The serum estrogen concentration varies from 1.5 × 10⁻¹⁰ M to 1.8 × 10⁻⁹ M during the menstrual cycle. Therefore, we investigated the dose response of the CHIP-dependent degradation of ERβ. As shown in Fig. 9B, CHIP-dependent ERβ degradation was changed in the levels of estrogen in the blood during the menstrual cycle.

Overall, these results suggest that CHIP-mediated degradation selectively and rapidly eliminates the active form of the NR to downregulate receptor-mediated transcription after ligand withdrawal.