Involvement of Suppressor for Gal 1 in the Ubiquitin/Proteasome-mediated Degradation of Estrogen Receptors*

The proteasome-mediated pathway involves the degradation of several nuclear receptors. Previously we demonstrated that the interaction between the suppressor for Gal 1 (SUG1) and nuclear receptors, the vitamin D receptor, or the pregnane X receptor was involved in proteasome-mediated degradation. In our recent experiments, we examined the potential role of SUG1 in the proteasome-mediated degradation of estrogen receptors (ER)α and -β. Both ERs interacted with SUG1 in a ligand-dependent manner. Functionally, the overexpression of SUG1 inhibited both ERα- and ERβ-mediated transcription in the presence of ligands. Transient expression studies demonstrated that the overexpression of wild-type SUG1 generated proteolytic fragments of both ERs and that these products were blocked by a proteasome inhibitor. The overexpression of SUG1 also enhanced the formation of ubiquitinated proteins of both ERs in the presence of ligand. On the other hand, bisphenol A (BSA), which activated ER-mediated transcription, did not enhance the interaction between ERβ and SUG1. Furthermore, the degradation of ERβ was much slower in the presence of BSA than in the presence of estradiol or phthalate, which is another endocrine-disrupting chemical. Also, BSA had no effect on the formation of proteolytic fragments of ERβ, and neither did it have any effect on the ubiquitination of ERβ. These findings indicate that the ubiquitin/proteasome-mediated degradation of both ER proteins may involve the interaction of SUG1 with both ERs. Moreover, BSA strongly blocked the ubiquitination and degradation of ERβ compared with estradiol, suggesting that BSA may affect the ERβ-mediated transcription of target genes by inhibiting ERβ degradation.

Steroid hormones, including estrogen and progesterone, as well as non-steroid hormones, vitamin D, retinoids, thyroid hormone, and prostanoids regulate their specific genes by binding to their specific receptors, which comprises the nuclear receptor superfamily. These receptors form homodimers or heterodimers with retinoid X receptor and are associated directly with specific DNA sequences, known as hormone-responsive elements, at the upstream regions of specific genes (1,2). The DNA receptor complex interacts with basal transcriptional ma-chinery and nuclear receptor coactivator proteins resulting in the ligand-dependent induction of transcription (2,3,4). Ligand binding, in addition to altering the conformational change of the receptor to interact with coactivators, has been shown to influence the stability of the nuclear receptors. The half-life of estrogen receptor (ER) 1 ␣ particularly is approximately 5 h in the absence of estradiol, whereas estradiol binding reduces the half-life to 3-4 h (5, 6), suggesting that receptor degradation may be an important event in regulating the response duration of transactivation to the ligand binding.
We have demonstrated that several putative cofactor proteins, including steroid hormone receptor coactivator-1 (7), receptor interacting protein 140 (8), and suppressor for Gal 1 (SUG1) (9), interacted with vitamin D receptor (VDR) and pregnane X receptor (PXR) in a natural steroid-dependent manner (10,11). Expression of steroid hormone receptor coactivator-1 and receptor interacting protein 140 augments ligand-activated transcription by a variety of nuclear receptors, indicating that these proteins act as transcriptional coactivators (7,8). Although yeast SUG1 was originally identified as a transcription factor (12) and interacts with several nuclear receptors, such as VDR, retinoid X receptor, and thyroid hormone receptor in a ligand-dependent manner (4,9,11,13), more recent evidence indicates that this protein is actually a component of the 26 S proteasome complex (14). Moreover, overexpression of this protein enhanced the degradation of VDR (10) and PXR (15). Together, these finding suggested that SUG1 might be involved in the degradation of nuclear receptors by proteasome.
Endocrine-disrupting chemicals (EDCs) have been defined as exogenous agents that interfere with the synthesis, secretion, transport, binding, action, or elimination of the body's natural hormones, which are responsible for the maintenance of homeostasis, reproduction, development, and/or behavior (16). These chemicals can alter endocrine functions through a variety of mechanisms, including steroid hormone receptor-mediated changes in protein synthesis, interference with membrane receptor binding, steroidogenesis, or the synthesis of other hormones (17). Although major chemicals, such as phthalates, alkylphenols, bisphenol A (BSA), and 1,1,1-trichloro-2,2-bis(pchlorophenyl)ethane, have been shown to disrupt estrogenic actions mainly through their binding to estrogen or androgen receptors (17), it is not clear whether EDCs directly affect endocrine functions in vivo.
We have examined the potential role of SUG1 in the proteasome-mediated degradation of ER␣ and ER␤. In this aspect, we checked whether or not SUG1 interacts with ERs in the presence of several ligands, including estradiol and EDCs, which activate ER-mediated transcription. Next, the effects of ligands for ERs on the steady states of both ERs were examined because we had observed a different effect of BSA, an EDC, on the interaction between SUG1 and ER␤ compared with the interaction between SUG1 and ER␣. We also examined the effect of SUG1 on ER-mediated transcription in the presence of these ligands. Furthermore, we used a transient transfection assay to determine whether or not the interaction between SUG1 and either of the ERs is involved in this degradation system. Finally, we checked whether or not SUG1 plays a role in the ubiquitination of ER proteins. The results of these inquiries FIG. 1. Differential effect of EDCs on the interactions between ERs and SUG1. A, yeasts expressing the pAS1-ER␣, -ER␤, -VDR, -PXR, and pAD-GAL4-SUG1 two-hybrid plasmids were grown for 24 h at 30°C in a selection medium (SC-Leu-Trp) in the absence and presence of 10 Ϫ6 M endogenous ligand for each receptor. The interactions between SUG1 and nuclear receptors were assessed in a ␤-Gal assay. Results are presented as mean Ϯ S.D. of triplicate independent cultures. B, yeasts expressing the pAS1-ER␣, -ER␤, and pAD-GAL4-SUG1 two-hybrid plasmids were grown for 24 h at 30°C in a selection medium (SC-Leu-Trp) in the absence or presence of 10 Ϫ6 M estradiol or 10 Ϫ6 M EDC. ER-SUG1 interaction was assessed in a ␤-Gal assay. Results are presented as mean Ϯ S.D. of triplicate independent cultures. C, yeasts expressing the pAS1-ER␣ and pAD-GAL4-SUG1 two-hybrid plasmids were grown for 24 h at 30°C in a selection medium (SC-Leu-Trp) in the absence or presence of increasing concentrations of estradiol or EDCs. ER␣-SUG1 interaction was assessed in a ␤-Gal assay. Results are presented as mean Ϯ S.D. of triplicate independent cultures (*, p Ͻ 0.01). D, yeasts expressing the pAS1-ER␤ and pAD-GAL4-SUG1 two-hybrid plasmids were grown for 24 h at 30°C in a selection medium (SC-Leu-Trp) in the absence or presence of increasing concentrations of estradiol or EDCs. ER␤-SUG1 interaction was assessed in a ␤-Gal assay. Results are presented as mean Ϯ S.D. of triplicate independent cultures (*, p Ͻ 0.01). E, a pAD-GAL4-SUG1 (wild-type (WT) or mutant (K196H)) plasmid was cotransformed with pAS1-ER␣ or -ER␤ into the yeast strain Hf7c. Wild-type or mutant SUG1 interactions with both ERs were quantitated in a ␤-Gal assay after 24 h growth at 30°C in a selection medium (SC-Leu-Trp) in the presence of 10 Ϫ6 M estradiol or 10 Ϫ6 M EDCs. Results are presented as mean Ϯ S.D. of triplicate independent cultures. indicate that the interactions between ERs and SUG1 may be involved in the ubiquitin/proteasome-mediated degradation of ER proteins. Moreover, BSA may affect ER␤-mediated transcription of target genes by inhibiting ER␤ degradation.
Cell Culture and Transient Transfection Studies-COS-7, HEC-1, and Ishikawa cells were cultured in Dulbecco's modified Eagle's medium without phenol red. The medium was supplemented with 10% charcoal-striped fetal bovine serum. COS-7 cells were cotransfected with 1 g of a reporter gene construct ((ERE) 2 -G-CAT) and 0.5 g of a receptor expression vector (pSG5-ER␣ or ER␤) or of an empty vector (pSG5). Ishikawa cells were transfected with 1 g of a reporter gene construct ((ERE) 2 -G-CAT) or of an empty vector (G-CAT) without the ER expression vector. For SUG1 expression, pcDNA3-SUG1 (wild-type or K196H) or pcDNA3 alone was transfected into the cells. All transfections were liposome-mediated by the use of lipofectAMINE (Invitro-gen) according to the manufacturer's protocol. Transfected cells were treated for 36 h, either with the vehicle alone or with the indicated concentrations of estradiol or EDCs. Cell extracts were prepared and assayed for CAT activity. The amount of CAT was determined using a CAT enzyme-linked immunosorbent assay kit (Roche Diagnostics) according to the manufacturer's protocol.
Preparation of Two-hybrid Expression Vectors and ␤-Gal Assays-All two-hybrid plasmid constructs used the pAS1 (18) and pAD-GAL4 yeast expression vectors (Stratagene, La Jolla, CA). The AS1-VDR, -PXR, -ER␣, and -ER␤ constructs have been previously described (11). The pAD-GAL4-SUG1 (wild-type or K196H) was cotransformed with pAS1-ER␣ or ER␤ into yeast strain Hf7c. Transformants were plated on synthetic complete medium lacking leucine and tryptophan (SC-Leu-Trp) and were grown for 4 days at 30°C to yield yeast that had acquired both plasmids. Triplicate independent colonies from each plate were grown overnight in 2 ml of SC-Leu-Trp with or without the indicated concentrations of estradiol, progesterone, 1,25(OH) 2 D 3 , or EDCs. Cells were harvested and assayed for ␤-Gal activity as described previously (4).
Enrichment of Ubiquitinated Protein-Polyubiquitin affinity beads and control beads were provided from Calbiochem-Novabiochem. Whole cell extracts were obtained from Ishikawa cells using M-Per reagent (Pierce) according to the manufacturer's protocol. An equivalent amount of whole cell extract (1 mg) was incubated with 40 l of beads for 4 h at 4°C with constant mixing in lysis buffer. This buffer consisted of 50 mM HEPES, pH 7.5, 5 mM EDTA, 150 mM NaCl, and 1% Triton X-100. The beads were washed three times, solubilized in SDS buffer, and analyzed by Western blotting using anti-ubiquitin (Santa Cruz Biotechnology, Inc.), -ER␣ or -ER␤ antibody.
Statistical Analysis-Statistical analysis was evaluated by one factor ANOVA followed by Dunnett's test. Data are mean Ϯ S.D. p Ͻ 0.05 denoted statistical significance.

FIG. 2. Effect of proteasome inhibitors on ER protein levels in endometrial cancer cell lines.
A, subconfluent HEC-1 and Ishikawa cells were treated with Me 2 SO or protease inhibitors for 6 h, and nuclear extracts were prepared as described under "Experimental Procedures." An equivalent amount of each extract was resolved by 10% SDS-PAGE. ER protein levels were determined by Western blotting using anti-ER␣ and -ER␤ antibody. B, subconfluent Ishikawa cells were treated with cycloheximide (10 mg/ml media) for 10 min prior to the addition of ligands. Cells were then treated in the presence of 10 Ϫ6 M estradiol, EDCs, or vehicle. Nuclear extracts were prepared, and ER protein levels were examined as described above. This level of cycloheximide inhibited Ͼ95% of 35 S-labeled methionine incorporation into trichloroacetic acid-precipitated protein (data not shown). Fig. 1A, SUG1 interacted with VDR, PXR, ER␣, and ER␤ in the presence of endogenous ligands for each receptor in the two-hybrid system. ER␣ interacted with SUG1 in the presence of EDCs, phthalate and BSA, both of which have been demonstrated to activate ER␣-mediated transcription (19). On the other hand, 1,1,1-trichloro-2,2bis(p-chlorophenyl)ethane, which had no transcriptional activity on ER␣, had no effect on the interaction. In contrast, interaction between ER␤ and SUG1 was not observed in the presence of BSA (Fig. 1B). The effect on the interaction of SUG1 with ER␣ depended on the concentration of ligands, and this effect increased significantly at 10 pM estradiol, 1 nM BSA, or 10 nM phthalate (Fig. 1C, p Ͻ 0.01 compared with ethanol). The effect on the interactions between SUG1 and ER␤ also depended on the concentration of ligands and significantly increased at 10 pM estradiol or 10 nM of phthalate, but BSA had no effect on the interaction at any concentration ( Fig. 1D; *, p Ͻ 0.01 compared with ethanol). Also, we checked the interaction between ERs and SUG1 (K196H), which in previous studies did not interact with VDR or ER␣ (10,13). The presence of any of the ligands had no effect on the interaction between ERs and the SUG1 mutant (Fig. 1E).

Differential Effects of EDCs on the Interaction between ER and SUG1-As illustrated in
Effect of Proteasome Inhibitors on ER Protein Levels in Endometrial Cancer Cell Lines-Because the turnover of ER␣ has been demonstrated to be involved in the proteasome-mediated degradation system (20, 21), we examined whether or not ER␤ as well as ER␣ is also a target for proteasome-mediated degradation in endometrial cancer cell lines, Ishikawa cells, and HEC-1 cells. The protein levels of ER␣ and ER␤ were examined in Ishikawa cells and HEC-1 cells that had been exposed to proteasome inhibitors. The protein levels of both ER␣ and ER␤ were markedly increased in the presence of 5 mM MG132 or 0.05 mM ␤-lactone, both of which strongly inhibit proteasome activities (22,23) with or without estradiol, in HEC-1 cells and Ishikawa cells (Fig. 2A). Furthermore, Ishikawa cells were treated with cycloheximide (10 mg/ml), and the effect of estradiol or either of two EDCs, phthalate and BSA, on ER turnover was determined. In the presence of estradiol, the protein levels of both ERs were rapidly degraded compared with ethanoltreated cells (Fig. 2B, lanes 2 and 6). The effects of BSA and phthalate on ER␣ turnover were similar to those in the presence of estradiol (Fig. 2B, lanes 3 and 4), but the degradation of ER␤ was blocked in the presence of BSA (Fig. 2B, lane 7) compared with that in the presence of estradiol or phthalate.
Suppression of ER-mediated Transcription by Overexpressed Wild-type SUG1-We used transient reporter expression assays in COS-7 cells and Ishikawa cells to examine whether or not SUG1 affected ER␣-and ER␤-mediated transcription. Expression of SUG1 suppressed estradiol-mediated transactivation of both overexpressed ER␣ and overexpressed ER␤, but SUG1 overexpression had no effect on basal transcription using COS-7 cells (Fig. 3A). The suppressive effect on estradiol-dependent transcription was the result of an interaction between ERs and SUG1 because the expression of mutant SUG1 (K196H), which interacted with neither of the ERs (Fig. 1E), had no significant effect on ER-mediated transactivation in this system (Fig. 3A). And, as described in Fig. 3B, SUG1 suppressed the ER␣-mediated transcription in the presence of BSA or phthalate, both of which enhanced the interaction between ER␣ and SUG1 (Fig. 1B). In contrast, SUG1 had no suppressive effect on ER␤-mediated transcription in the presence of BSA, which had no effect on the interaction between SUG1 and ER␤ (Fig. 1B). Mutant SUG1 had no effect on ER-mediated transcription in any ligands. Moreover, SUG1 had a negative effect on endogenous ER-mediated transcription in the presence of phthalate or estradiol, but there was no change in the presence of BSA in Ishikawa cells (Fig. 3C). We also examined the effect of a proteasome inhibitor, MG132, on ER-mediated transcription using Ishikawa cells. MG132 weakly suppressed basal transcription and ER-mediated transcription in the presence of estradiol or BSA and completely blocked the suppressive effect of SUG1 on ER-mediated transcription in the presence of estradiol or phthalate (Fig. 3D), suggesting that SUG1 affected the transcription through the proteasome-mediated degradation.
SUG1 Overexpression Enhances ER Proteolysis in Ishikawa Cells-To test whether or not SUG1-ER interactions are involved in ER degradation in proteasome-mediated systems, wild-type and mutant SUG1 were transiently overexpressed in Ishikawa cells, and their respective effects on ER protein levels were examined by Western immunoblotting (Fig. 4A). In the absence of estradiol, overexpression of wild-type or mutant SUG1 did not significantly affect the protein levels of ER␣ (Fig.  4A, lanes 1-3) or ER␤ (Fig. 4A, lanes 7-9). However, in the presence of estradiol, novel proteolytic fragments of both ERs were observed when Ishikawa cells were transfected with the wild-type SUG1 expression vector (Fig. 4A, lane 5 and 11). The ER proteolytic fragments were not observed in the cells transfected with mutant SUG1 (Fig. 4A, lanes 6 and 12). The overexpression of SUG1 was confirmed using anti-SUG1 antibody. Moreover, proteolytic fragments of ER␣ were also observed in the presence of BSA or phthalate (Fig. 4B, lanes 2 and 3). However, we observed proteolytic fragments of ER␤ only in the presence of phthalate and not in the presence of BSA (Fig. 4B,  lanes 8 and 9), suggesting that this phenomenon requires interaction between ERs and SUG1, whereas MG132 completely blocked the formation of SUG1-dependent fragments of both ERs (Fig. 4B, lanes 4 -6 and 10 -12). Then, we checked the dominant-negative effect of SUG1 (K196H) on wild-type SUG1induced proteolysis of both ERs. As described in Fig. 4C, mutant SUG1 inhibited the proteolytic fragments of both ERs by overexpressed wild-type SUG1 in a dose-dependent manner.
Wild-type SUG1 Enhanced the Ubiquitination of ER-Because the ubiquitination of the target protein has been demonstrated to be important in the proteasome-mediated degradation system (24, 25), we checked whether or not SUG1 is involved in the process of ER ubiquitination. We used polyubiquitin affinity beads to pull polyubiqutinated proteins down FIG. 4. Enhancement of wild-type SUG1 on ER proteolysis in Ishikawa cells. A, Ishikawa cells on 150-mm plates were transfected with 10 g of pcDNA3-SUG1 (wild-type (WT) or K196H) or pcDNA3 expression vector. The cells were treated with an ethanol vehicle or 10 Ϫ6 M estradiol for 24 h, and nuclear extracts were prepared. The protein levels of ERs were examined as described in the legend to Fig. 2. The expression of overexpressed SUG1 was confirmed using anti-SUG1 antibody. B, Ishikawa cells were transfected with 10 g of pcDNA3-SUG1 (wildtype (WT)). The cells were treated with 10 Ϫ6 M estradiol or EDCs and exposed to 5 mM MG132 or vehicle for 24 h, and nuclear extracts were prepared. The protein levels of the ERs were examined as described under " Fig. 2." C, Ishikawa cells were transfected with 10 g of pcDNA3-SUG1 (wild-type (WT)) and increasing amount of pcDNA3-SUG1 (K196H). The cells were treated with 10 Ϫ6 M estradiol for 24 h, and nuclear extracts were prepared. The protein levels of the ERs were examined as described in the legend to Fig. 2. from whole cell extracts of Ishikawa cells (Fig. 5A). We detected ubiquitinated proteins of both ER␣ and ER␤ with overexpressed wild-type SUG1 alone; none were detected with the pcDNA3 expression vector alone or with SUG1 (K196H) (Fig.  5B). No nonspecific interaction with control beads was observed. Moreover, ubiquitinated ER␣ protein was observed in the presence of BSA or phthalate as well as estradiol, but the ubiquitination of ER␤ was detected in the presence of phthalate or estradiol (Fig. 5C), suggesting that ubiquitination of ER proteins requires interaction between SUG1 and the ER. Then, the dominant-negative effect of mutant SUG1 on the ubiquitination of both ERs was examined. This mutant SUG1 inhibited the ubiquitination of both ERs in a dose-dependent manner (Fig. 5D). DISCUSSION Proteasome is a major cytosolic and nuclear protease complex that is responsible for an ATP-dependent, extra-lysosomal proteolytic pathway. This complex is responsible for the degradation of most cellular proteins, and proteasome activity is necessary for cell viability (24,25). Proteasome is highly conserved throughout eukaryotic evolution, and it exists as two major complexes; 20 S proteasome, which contains multiple peptidase activities, and 26 S proteasome, which contains the 20 S subunit as well as a 19 S regulatory complex composed of multiple ATPases and components necessary for binding protein substrates (24,25). To date, a wide variety of substrates for proteasome have been identified, including rate-limiting enzymes such as ornithine decarboxylase, transcriptional regula-tors such as c-Jun, p53, and NF-B, and critical regulatory proteins such as cyclins and tyrosine kinase receptors (24,25). In addition, we and others (10,15,20,21) have demonstrated that the proteasome system might be involved in the degradation of nuclear receptors, including vitamin D receptor, pregnane X receptor, progesterone receptor, and estrogen receptor. A common feature of proteasome-mediated degradation is the covalent attachment of ubiquitin to lysine residues of target proteins followed by polyubiquitin chains attached covalently to the proteins. A polyubiquitin chain of a target protein is important for specific targeting to the proteasome (24,25). Some nuclear receptors have been reported to be polyubiquitinated and degraded by proteasome (20,26).
In the present study, some evidence was obtained that SUG1 might play a role in the ubiquitin/proteasome-mediated degradation of both ERs. First, two relatively selective inhibitors of the proteasome pathway, MG-132 and ␤-lactone, dramatically increased the steady state levels of native ER␣ and ER␤ proteins in nuclear extracts obtained from Ishikawa cells. Second, overexpression of wild-type SUG1 in Ishikawa cells resulted in the appearance of proteolytic derivatives of both ERs in the presence of estradiol. ER-SUG1 interaction was also required for the formation of these proteolytic fragments, because overexpression of mutant SUG1 did not produce a similar effect. Moreover, the ubiquitinated proteins of both ERs were observed in the presence of overexpressed wild-type SUG1 in estradiol-treated cells. Furthermore, MG132 abolished the formation of proteolytic fragments of both ERs. Taken together, FIG. 5. Enhancement of wild-type SUG1 on the ubiquitination of ERs. A, equivalent amounts of whole cell extract from Ishikawa cells (1 mg) were incubated with 40 l of polyubiquitin affinity beads for 4 h at 4°C with constant mixing. After washing three times, the ubiquitinated proteins were analyzed by Western blotting using anti-ubiquitin antibody. B, Ishikawa cells on 150-mm plates were transfected with 10 g of pcDNA3 expression vector or pcDNA3-SUG1 (wild-type (WT) or K196H). The cells were treated with 10 Ϫ6 M estradiol for 24 h, and whole cell extracts were incubated with 40 l of beads as described above. The ubiquitinated proteins were analyzed by Western blotting using anti-ER␣ or -ER␤ antibody. C, Ishikawa cells were transfected with 10 g of pcDNA3 expression vector or pcDNA3-SUG1 (wild-type (WT) or K196H). The cells were treated with ethanol, 10 Ϫ6 M estradiol, or 10 Ϫ6 M EDCs for 24 h, and whole cell extracts were incubated with 40 l beads as described above. The ubiquitinated proteins were analyzed by Western blotting using anti-ER␣ or -ER␤ antibody. D, Ishikawa cells were transfected with 10 g of pcDNA3-SUG1 (wild-type (WT)) and increasing amounts of pcDNA3-SUG1 (K196H). The cells were treated with 10 Ϫ6 M estradiol for 24 h, and whole cell extracts were incubated with 40 l of beads as described above. The ubiquitinated proteins were analyzed by Western blotting using anti-ER␣ or -ER␤ antibody. these results suggested that SUG1 interacted with both ERs and targeted these ERs for degradation by proteasome machinery through ubiquitination.
Deletion analysis of SUG1 demonstrated that mutant SUG1 (which abolished ATPase activity) as well as the N-or Cterminal domain of SUG1 acted as a dominant-negative in the proteasome-dependent degradation of transcription factor Sp1 (27). It has also been determined that the coiled coils in the N-terminal region of proteasomal ATPases including SUG1 direct the placement of these proteins within the proteasome (28). In our experiments, mutant SUG1, which eliminated ATPase activity, abolished the interaction between SUG1 and ERs and exhibited a dominant-negative effect on the ubiquitination and proteasome-mediated degradation of ERs. This suggested that mutant SUG1 might inhibit the recruitment of wild-type SUG1 to proteasome by occupying the available docking sites. Further analysis will be required to resolve in detail the question of how SUG1 participates in ubiquitination and targeting to proteasome machinery.
A variety of putative pathways by which EDCs affect the endocrine system have been reported (16,17). We have already presented one such potential pathway, the PXR-mediated changes of steroidogenesis (11). We have also demonstrated the existence of a general mechanism for receptor down-regulation that may involve proteasome-mediated proteolysis with the interaction between PXR and SUG1 (15). In the present experiments, we demonstrated that SUG1 interacted differently with ER␣ than with ER␤ in the presence of EDC, especially BSA. BSA had a positive effect on the interaction between SUG1 and ER␣, but did not affect the interaction with ER␤. And, in the presence of BSA, the degradation of ER␤ was much slower than that in the presence of estradiol or phthalate, another EDC. In addition, the formation of proteolytic fragments and the ubiquitination of ER␤ by overexpressed SUG1 were not observed in the presence of BSA. This suggested that BSA blocked the turnover of ER␤ by inhibiting the ubiquitin/ proteasome-mediated degradation of ER␤, resulting in changes in ER␤ protein levels that in turn may have affected ER␤mediated gene regulation. If so, this chain of events may be a potential mechanism by which EDCs affect endocrine function.
In summary, we examined whether or not the interactions between SUG1 and ERs were involved in the ubiquitin/proteasome-mediated degradation. Both estrogen receptors ␣ and ␤ interacted with SUG1 in an estradiol-dependent manner, and the protein levels of both ERs were markedly increased in the presence of proteasome inhibitors. Functionally, the expression of SUG1 inhibited both ER␣-and ER␤-mediated transcription in the presence of ligands. The transient expression studies demonstrated that overexpression of wild-type SUG1 generated proteolytic fragments of both ERs, and these products were blocked by a proteasome inhibitor. We also found that overexpression of SUG1 enhanced the formation of ubiquitinated proteins of both ERs in the presence of ligands. On the other hand, BSA, which activated ER-mediated transcription, did not enhance the interaction between ER␤ and SUG1. In the presence of BSA, ER␤ was degraded much more slowly than it is in the presence of estradiol and phthalate, another EDC. Also, BSA had no effect on the formation of proteolytic fragments of ER␤ or on ubiquitination of ER␤ in the presence of overexpressed wild-type SUG1. These findings indicate that the interactions between the ERs and SUG1 may be involved in the ubiquitin/proteasome-mediated degradation of both ER proteins. Moreover, compared with estradiol, BSA strongly blocked the ubiquitination and degradation of ER␤, suggesting that BSA may affect the ER␤-mediated transcription of target genes by inhibiting ER␤ degradation.