This Article Abstract Full Text (PDF) All Versions of this Article: 20/11/2695    most recent Author Manuscript (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager NURSA Molecule Pages Link Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jaber, B. M. Articles by Smith, C. L. Search for Related Content PubMed PubMed Citation Articles by Jaber, B. M. Articles by Smith, C. L.

The Pure Estrogen Receptor Antagonist ICI 182,780 Promotes a Novel Interaction of Estrogen Receptor- with the 3',5'-Cyclic Adenosine Monophosphate Response Element-Binding Protein-Binding Protein/p300 Coactivators

Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas 77030

Address all correspondence and requests for reprints to: Carolyn L. Smith, Ph.D., Molecular and Cellular Biology, One Baylor Plaza, Houston, Texas 77030. E-mail: carolyns{at}bcm.tmc.edu.

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Estrogen receptor- (ER) is a member of the nuclear receptor superfamily of ligand-activated transcription factors. Abundant evidence demonstrates that ER agonists promote, whereas antagonists inhibit, receptor binding to coactivators. In this report we demonstrate that binding of the ICI 182,780 (ICI) pure antiestrogen to ER promotes its interaction with the cAMP response element-binding protein-binding protein (CBP)/p300 but not the p160 family of coactivators, demonstrating the specificity of this interaction. Amino acid mutations within the coactivator binding surface of the ER ligand-binding domain revealed that CBP binds to this region of the ICI-liganded receptor. The carboxy-terminal cysteine-histidine rich domain 3 of CBP, rather than its amino-terminal nuclear interacting domain, shown previously to mediate agonist-dependent interactions of CBP with nuclear receptors, is required for binding to ICI-liganded ER. Chromatin immunoprecipitation assays revealed that ICI but not the partial agonist/antagonist 4-hydroxytamoxifen is able to recruit CBP to the pS2 promoter, and this distinguishes ICI from this class of antiestrogens. Chromatin immunoprecipitation assays for pS2 and cytochrome P450 1B1 promoter regions revealed that ICI-dependent recruitment of CBP, but not receptor, to ER targets is gene specific. ICI treatment did not recruit the steroid receptor coactivator 1 to the pS2 promoter, and it failed to induce the expression of this gene. Taken together, these data indicate that recruitment of the CBP coactivator/cointegrator without steroid receptor coactivator 1 to ER is insufficient to promote transcription of ER target genes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
ESTROGEN RECEPTOR- (ER) is a member of the nuclear receptor superfamily of ligand-dependent transcription factors that plays an important role in many biological processes, including growth, differentiation, and development (1). The transcriptional activity of ER is dependent on the receptor’s two activation functions (AFs); AF-1 in the N terminus (A/B domain), and AF-2 in the C terminus (E domain). The AF-1 functions in a ligand-independent manner, whereas AF2 function is ligand dependent (2, 3). Although these two domains can function independently, maximal ER transcriptional activity is achieved when the two AFs synergize (4). These domains are sites of interaction for a large number of coactivators: proteins that facilitate the interaction between ER and the general transcriptional machinery (5, 6). Several coactivators also possess histone acetyltransferase, arginine methyltransferase, and ubiquitin ligase activities (6), and receptor recruitment of coactivators possessing these enzymatic activities contributes to chromatin remodeling and the dynamic occupancy of target gene promoters by receptors, coactivators, and RNA polymerase II (7, 8, 9).

The ligand binding domain (LBD) of ER consists of 12 -helices arranged as a three-layered antiparallel -helical sandwich that forms a hydrophobic pocket to which ligands bind (10). Upon binding to an agonist, helix 12 is oriented over the ligand-binding pocket, allowing helices 3, 5, and 12 to generate a functional AF-2 domain consisting of a hydrophobic groove on the LBD surface (11, 12). This, in turn, enables the agonist-bound ER to interact with coactivators, such as p160 [steroid receptor coactivator (SRC)] family coactivators, and ultimately stimulate target gene expression. Coactivator interaction with agonist-bound LBD is achieved through a highly conserved signature motif found within many coactivators termed the nuclear receptor box, which consists of the consensus sequence LXXLL (where L is leucine and X is any amino acid) (13). Extensive work has gone into characterizing the interaction of LXXLL-containing peptides with nuclear receptors, and although the affinities of the various p160 LXXLL motifs for ER vary, they are greater than the ER-binding affinity of the CBP LXXLL motifs, which are considered to be sufficiently low that CBP cannot bind to ER without the assistance of other coregulators (13, 14, 15). Indeed, CBP/p300 binds to ER as a component of a ternary complex consisting of CBP/p300, p160 coactivators, and ER (16), and SRC-1 and CBP synergistically enhance ER transcriptional activity (17, 18). Moreover, mutation of the CBP-binding region of SRC-1 blocks the ability of CBP to contribute to ligand-stimulated gene expression (14, 19, 20)

Because of the importance of ERs in the reproductive, skeletal, cardiovascular, and other systems, significant efforts have been extended to develop ligands that can positively or negatively regulate the transcriptional activity of the receptors. There are two types of antiestrogens: type I antiestrogens such as 4-hydroxytamoxifen (4HT) and raloxifene can function as partial agonists and antagonists of ER, whereas type II antiestrogens, typified by ICI 164,384 and ICI 182,780 (ICI), are devoid of estrogenic activity. The partial agonist/antagonist activity of type I antiestrogens is derived from their ability to inhibit AF-2 but not AF-1 activity (6). x-Ray crystallographic studies of tamoxifen or raloxifene bound to the ER LBD reveal that helix 12 adopts a distinct orientation in which it binds to the coactivator recognition groove of the LBD, thereby preventing coactivator binding to the receptor (12). In contrast, type I antiestrogens do not appear to block the interaction of coactivators, such as SRC-1 or CBP/p300 with the amino terminus of ER, thereby enabling the receptor to exert AF-1-dependent transcriptional activity (21, 22, 23, 24). Consistent with the cell- and tissue-specific nature of AF-1 activity (4, 22, 25), type I antiestrogens exert agonist activity in a promoter-, cell-, and tissue-specific manner, and they have therefore been classified as selective ER modulators (SERMs).

Although a structure for ER bound to pure ICI 164,384 and ICI antiestrogens is not yet available, the structure of the closely related ERß in the presence of ICI 164,384 reveals a distinct conformation in which the position of helix 12 cannot be determined (26). This is likely due to the long alkylamide side chain of ICI 164,384 that extends out of the ligand-binding pocket and toward the coactivator-binding groove (26). Neither AF-1 nor AF-2 activities are manifested for ER in the presence of ICI compounds, and a variety of additional mechanisms have been proposed to account for the pure antiestrogen effects of these ligands, including inhibition of receptor dimerization and DNA binding, acceleration of ER down-regulation, and nuclear export of ER (27, 28, 29, 30, 31, 32). It also has been shown that ICI induces ER interaction with the C terminus of the nuclear receptor corepressor and can promote histone deacetylase-mediated repression of an estrogen response element (ERE)-responsive reporter (33). In addition, ICI promotes an AF-2 independent interaction between ER and CIA, a coregulator that possesses corepressor and coactivator functions (34), but inhibits interactions between ER and SRC family coactivators (35). The extent to which each of these mechanisms contributes to the antiestrogen activity of ICI has not been precisely defined.

In this study the ability of antiestrogens to regulate interactions between ER and various coactivators was investigated, and our results reveal that pure antiestrogens promote ER binding to CBP and p300, but not to p160 coactivators. An LXXLL-independent binding site within CBP responsible for interaction with ICI-bound ER was defined, and ICI was shown to selectively recruit CBP/p300 to target gene promoters by chromatin immunoprecipitation (ChIP) assay. Taken together, our data provide evidence that CBP/p300 can bind to ER independently of p160s in the presence of ICI and suggests that recruitment of CBP to target gene promoters is insufficient to promote transcription of ER target genes.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
ICI Enhances the Interactions between CBP/p300 and the ER LBD
Binding of the pure antiestrogen ICI to ER is associated with loss of transcriptional activity by the receptor (36), and presumably is dependent on interactions of receptor with corepressors. Moreover, current hypotheses suggest that antiestrogens do not promote interactions of ER with coactivators; this was investigated in the current study in which the effect of ICI on interactions between ER and CBP were examined. To accomplish this, we employed a mammalian two-hybrid assay in which HeLa cells were cotransfected with a GAL4 responsive reporter and vectors for the GAL4 DNA-binding domain (DBD) (GAL) or chimeras of GAL fused to the amino termini of either full-length CBP or p300, in the presence of the VP16 activation domain alone (VP16) or VP16 fused to the amino terminus of the ER LBD (VP16-LBD). Cells were treated with either vehicle, 10 nM 17ß-estradiol (E2) or 100 nM ICI for 20–24 h. As expected, E2 increased the ER LBD-CBP interaction as well as the ER LBD-p300 interaction (Fig. 1, A and B). Surprisingly, treatment with ICI increased the ER LBD-CBP interaction to an extent even greater than for E2. Similar results were obtained for the ICI-induced interaction between ER LBD and p300.

View larger version (19K): [in this window] [in a new window]   Fig. 1. The Pure Antagonist ICI Enhances Interactions between CBP/p300 and the ER LBD HeLa cells were transfected with expression vectors for the GAL4 DBD alone (GAL) or GAL-CBP (panel A) or GAL-p300 (panel B) in the presence of expression vectors for the VP16 activation domain (VP16) or a chimera consisting of the VP16 activation domain fused to the amino terminus of the ER LBD (VP16-ERLBD) along with the pG5-Luc reporter plasmid. Cells were treated with 0.1% ethanol (vehicle), 10 nM E2, or 100 nM ICI. Values are normalized to those obtained for GAL-CBP (panel A) or GAL-p300 (panel B) and VP16 in the presence of vehicle, which was defined as 1. Bars represent the average ± SEM of six experiments.

Interaction of p160 Coactivators with the ER LBD

View larger version (14K): [in this window] [in a new window]   Fig. 2. Mammalian Two-Hybrid Assay of SRC Family Coactivator Interactions with the ER LBD HeLa cells were transfected with expression vectors for the GAL4 DBD alone (GAL) or GAL-SRC-1 (SRC-1), GAL-TIF2 (TIF2), or GAL-RAC3 (RAC3) chimeras in the presence of an expression vector for the VP16 activation domain (VP16) or a VP16-LBD chimera along with the pG5-Luc reporter plasmid. Cells were treated with 0.1% ethanol (vehicle), 10 nM E2, or 100 nM ICI and subsequently assayed for luciferase activity. Values are normalized to that obtained for GAL-SRC-1 and VP16 in the presence of vehicle, which was defined as 1. Bars represent the average ± SEM of three experiments.

ICI Promotes CBP Interaction with Full-Length ER

View larger version (42K): [in this window] [in a new window]   Fig. 3. CBP and SRC-1 Interaction with Full-Length ER A, Full-length recombinant ER (100 ng) previously incubated with either ethanol [vehicle (V)], 10 nM E2, or 100 nM ICI was mixed with Hela cell lysates and subjected to coimmunoprecipitation with antibodies against CBP (top) or SRC-1 (bottom). The ER content of the immunocomplex was assessed by Western blot with anti-ER antibody (Cell Signaling Technology, Beverly, MA). The Western blot is a representative of two experiments. B, MCF-7 cells were treated with 0.1% ethanol (V), 10 nM E2, or 1 µM ICI. Western blot of CBP (top), ER (middle), and actin (bottom) expression in total lysates of treated cells (lanes 1–3) or in complexes immunoprecipitated from cell lysates with an antibody to CBP (lanes 4–6) or IgG (lane 7). The blot is representative of five experiments. IB, Immunoblot; IP, immunoprecipitation.

ICI Promotes ER LBD-CBP Interaction in a Dose-Dependent Manner and Is Antagonized by 4HT

View larger version (23K): [in this window] [in a new window]   Fig. 4. Dose-Dependent Antagonism of ICI-Induced CBP-ER LBD Interaction by 4HT Cells were transfected with expression vectors for GAL-CBP and VP16-LBD chimeras along with the pG5-Luc reporter plasmid and treated with either 0.1% ethanol vehicle (lane 1), 10 nM E2 (lane 2), increasing concentration of 4HT (1–1000 nM; lanes 3–6), or ICI (1–1000 nM; lanes 7–10), or a combination of 100 nM ICI plus 1–1000 nM 4HT (lanes 11–14) for 20–24 h. The cells were subsequently harvested and assayed for luciferase activity. Values are normalized to those obtained for GAL-CBP and VP16 in the presence of vehicle, which was defined as 1 (not shown in the graph). Bars represent the average ± SEM of three experiments.

SERMs Do not Enhance ER-CBP Interaction

View larger version (31K): [in this window] [in a new window]   Fig. 5. Effect of Pure Antagonists and SERMs on the Interactions between CBP and the ER LBD HeLa cells were transfected with expression vectors for the GAL4 DBD alone (GAL) or a GAL-CBP (panels A and B) or GAL-SRC-1 (panel B) chimera in the presence of an expression vector for VP16 alone or VP16-LBD along with the pG5-Luc reporter plasmid. Cells were treated with 0.1% ethanol (vehicle), 10 nM E2, or 100 nM of 4HT, raloxifene, idoxifene, ICI 164,380, ICI, ZK-703, or ZK-253 for 20–24 h, and subsequently harvested and assayed for luciferase activity. Values are normalized to those obtained for GAL-CBP and VP16 in the presence of vehicle which was defined as 1. Bars represent the average ± SEM of three experiments.

Mutation of the ER LBD Blocks Interactions between CBP and ER

View larger version (23K): [in this window] [in a new window]   Fig. 6. Mutation of the ER LBD Inhibits CBP-Receptor Interactions HeLa cells were transfected with expression vectors for GAL-TIF2 (panel A) or GAL-CBP (panel B) along with plasmids for VP16 alone or VP16 fused to the amino termini of wild-type (WT) or mutant forms (I358D, K362D V376D or L539A) of the ER LBD along with the pG5-Luc reporter plasmid. Cells were treated with either ethanolic vehicle, 10 nM E2, or 100 nM ICI for 20–24 h, and subsequently harvested and assayed for luciferase activity. Values are normalized to those obtained for GAL-CBP and VP16 in the presence of vehicle, which was defined as 1. Bars represent the average ± SEM of four independent experiments.

To ensure that differences in ER mutant binding to ICI did not account for the lack of ICI-induced CBP-ER LBD interaction, relative binding affinity (RBA) assays were conducted in which the ability of ICI to bind to wild-type ER and each of the coactivator-binding groove mutants in comparison with E2 was determined. Scatchard plots had previously demonstrated that these mutations had little effect on the receptor’s binding affinity for E2 (39). As shown in Table 1, the RBA of the mutant ERs varied little in comparison with the RBA of wild-type ER for ICI. This indicates that the inability of CBP to bind to the ER mutants is not due to lack of ICI binding to ER(I358D), ER(K362D), ER(V376D), or ER(L539A).

View this table: [in this window] [in a new window]   Table 1. RBA of Wild-Type (WT) and Mutant ER for ICI 182,780

View larger version (29K): [in this window] [in a new window]   Fig. 7. Mammalian Two-Hybrid Assay of CBP Domain Interactions with the ER LBD A, Schematic representation showing the major CBP interaction domains. B, HeLa cells were transfected with GAL chimeric expression vectors for full-length CBP (CBP) or the NID, KIX, CH1, CH2, or CH3 domains, in the presence of an expression vector for the VP16 activation domain or the VP16-LBD along with the pG5-Luc reporter plasmid. Cells were treated with 0.1% ethanol [vehicle (V)], 10 nM E2, or 100 nM ICI for 20–24 h. Values are normalized to those obtained for GAL-CBP and VP16 in the presence of vehicle, which were defined as 1. Bars represent the average ± SEM of three independent experiments. C, Interaction of in vitro translated [35S]ER-179C with GST or GST-CH3 in the presence of ethanolic vehicle, 10 nM E2, or 100 nM ICI.

View larger version (32K): [in this window] [in a new window]   Fig. 8. Mutation of CBP’s CH3 Domain Blocks Its Interaction with ER A, Amino acid sequence of CH3 domain showing the zinc finger-containing domains. Residues shown in bold are responsible for coordinate binding of zinc. Asterisks (*) indicate the amino acids that were mutated to alanine for each of the mutants, m1, m2, and m3. B, HeLa cells were transfected with expression vectors for the GAL-CH3 (CH3) or GAL-CH3 zinc finger mutants (m1, m2, and m3) along with the expression vector for the VP16 activation domain or VP16-ER LBD along with the pG5-Luc reporter plasmid. Cells were treated with 0.1% ethanol (vehicle), 10 nM E2, or 100 nM ICI for 20–24 h. Values are normalized to those obtained for GAL-CH3 and VP16 in the presence of vehicle, which were defined as 1. Bars represent the average ± SEM of at least three independent experiments. C, HeLa cells were transfected with 2 µg of either GAL alone or GAL-CH3 or its corresponding mutants. Protein (30 µg) was resolved on 10% SDS-PAGE, and GAL4 chimeric proteins were detected by Western blot using HRP-conjugated antibody against the GAL4 DNA binding domain.

View larger version (40K): [in this window] [in a new window]   Fig. 9. Down-Regulation of ER Expression by ICI MCF-7 cells were plated in phenol red-free DMEM containing 10% sFBS and treated with 100 nM ICI for 0–6 h. Cell lysates were analyzed by Western blot analysis for expression of ER (top) or actin (bottom). Shown is a representative blot of four experiments.

View larger version (37K): [in this window] [in a new window]   Fig. 10. ER Selectively Recruits CBP to Gene Promoters MCF-7 cells were treated with -amanitin for 1.5 h followed by either ethanol, 10 nM E2, 100 nM 4HT, or 100 nM ICI for 30 min (panel A), 45 min (panel B), or 60 min (panels C and E). Cell lysates were subjected to immunoprecipitation with anti-ER or anti-CBP (A-22) or anti-SRC-1 antibodies. Immunoprecipitations were collected using protein A-sepharose beads, and purified DNA was quantified by real-time PCR analysis using specific primers for pS2 promoter (panels A–C) or CYP1B1 promoter (panel E). Values were normalized to input, and values for vehicle treatment were defined as 1. The values for ChIP experiments represent an average of two to four experiments (except n = 1 for 4HT in panel B). Levels of mRNAs for pS2 (panel D) or CYP1B1 (panel F) were measured for cells treated with the indicated hormones for 8 or 24 h and normalized to levels of 18S RNA. Values represent the average ± SEM of four experiments.

This prompted us to examine whether ICI recruitment of CBP may be associated with a gene the expression of which is induced by ICI treatment. Recent microarray analyses indicated that ICI increases mRNA expression of the cytochrome P450 CYP1B1 gene (46), and, in our MCF-7 cells, 8 h of ICI treatment stimulates a significant induction of CYP1B1 mRNA (P < 0.05; Fig. 10F). Using primers that span the functional ERE within the CYP1B1 promoter (47, 48), ChIP assays demonstrated that both E2 and ICI recruited ER to the CYP1B1 promoter (Fig. 10E). Consistent with our previous demonstration of SRC-3 recruitment to the CYP1B1 promoter (48), ligand-dependent interaction of CBP and SRC-1 was observed only after E2 treatment; no recruitment was observed for ICI-treated cells. Similar results were obtained for cells treated with ligands for 30 min (data not shown). Taken together, our results indicate that ICI-dependent recruitment of CBP to ER-regulated genes is promoter specific. Moreover, the inability of ICI to stimulate pS2 gene expression indicates that recruitment of CBP was insufficient to initiate activation of this gene’s expression.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
It has been well established that interactions between coactivators and ER are required for maximal transcriptional activation of target genes (5, 49). Although many studies have demonstrated the important contribution of the p160 and CBP/p300 coactivators to the ability of ER to activate target genes (18, 50, 51), the detailed mechanism by which these cofactors coordinate their function has not been elucidated. In the present study, we employed assays to observe interactions between ER and CBP/p300 in the presence of agonist or antagonists. Surprisingly, the pure antagonist ICI, which blocks ER transcriptional activity, increased ER-CBP and ER-p300 interactions to levels higher than those observed for E2. These interactions are unique because ICI was unable to promote ER binding to any of the p160 coactivators. Although ICI has been demonstrated to enhance ER binding to CIA [a cofactor containing LXXLL coactivator and XX corepressor motifs (34)], and promote the recruitment of silencing mediator of retinoid and thyroid hormone receptor and nuclear receptor corepressor to ER (33), to the best of our knowledge, this is the first example of a pure antagonist strongly promoting the recruitment of a coactivator to ER.

Numerous studies have demonstrated that LXXLL motifs are major determinants of coactivator interactions with agonist-bound nuclear receptors. There are two LXXLL motifs located within CBP; one in the NID (amino acids 68–75) and another in the CH1 (amino acids 356–366) domain (13, 14). As expected, E2-induced binding was observed for ER with the NID, but not the CH1 domain. This is likely attributable to differences in the relative affinities of specific LXXLL motifs for binding to receptors (52, 53). In contrast to the CH1 domain, the CH3 domain, which does not contain any LXXLL motifs, showed strong interaction with the ER-LBD in the presence of either E2 or ICI, as measured by mammalian two-hybrid, coimmunoprecipitation, and GST-pull-down assays. Our findings differ from a previous report on ER interactions with the CH3 domains of CBP and p300 that were insensitive to either agonist or antagonist ligands (54). The reason for the apparent difference is unknown, but may relate to the previous report’s reliance on in vitro assays and/or experimental conditions. Mutation of residues important for binding each of the three zinc atoms in the CH3 domain disrupted ER interaction with this domain, suggesting that its overall structure is important for supporting hormone-dependent interaction with the receptor.

Previous studies have demonstrated that p160 coactivators, as part of a ternary complex, are required for efficient, agonist-induced interaction of CBP with nuclear receptors (17, 55). The ability of the I358D, V376D, and L539A mutations to inhibit ER-CBP interactions suggests that the ER coactivator-binding groove is important for receptor binding to CBP/p300. However, the differing sensitivity of CBP vs. p160 coactivator binding to the K362D mutant form of ER, and the ability of CBP/p300 to bind to the ICI-occupied ER LBD independent of p160s suggest that CBP employs a distinct means of binding to receptor in the presence of ICI. The chemical structures of ICI and ICI 164,384 consist of a steroidal ring with a bulky alkylamide side chain protruding from the 7 position (30). Based on a structure determined for ICI 164,384 binding to ERß, the displacement of the H12 by the bulky side chain results in the enlargement of the ligand-binding pocket, thus creating an 11ß-channel, which can accommodate the 7-side chain extending out of the ER ligand-binding pocket toward the coactivator-binding surface (26). The inability of the SERMs, 4HT, raloxifene, and idoxifene, to promote CBP interaction with the ER LBD indicates that these ligands do not induce an ER conformation able to bind directly to CBP. Neither ligands possess a side chain similar in size to the 7-side chains of the ICI compounds (11, 56), and the structures of ERs bound to 4HT or raloxifene indicate that in both cases, H12 binds to a portion of the receptor’s coactivator binding groove, which blocks coactivators from binding to this region of the receptor (11, 12). Based on these results, we propose that a ligand side chain of sufficient bulk that can disrupt H12 binding to the ER LBD is important for promoting direct interactions between ER and CBP/p300. In support of this concept, the ZK-703 compound, which belongs to a group of compounds referred to as specific ER destabilizers and also has a bulky 7-side chain structure like ICI, can promote ERLBD-CBP interaction to a similar extent as ICI. The inability of the ZK-253 compound to promote ER-CBP binding is likely due to the extensive incorporation of the fluoride atoms into the 7-side chain (five in ZK-703 and ICI vs. 9 in ZK-253) and possible steric interference. A detailed structural determination of ICI-occupied ER bound to the CH3 domain of CBP will be required to determine whether the ICI side chain contributes directly to this interaction.

Previous in vivo footprinting studies demonstrated that ICI treatment of MCF-7 cells promoted the recruitment of multiple proteins to the pS2 promoter in a pattern distinct from that obtained for E2 or 4HT (57). Based on our results, as well as the ability of ICI to promote corepressor binding to ER (33), it is likely that the ICI-specific footprinting pattern reflects recruitment of CBP/p300 as well as corepressor proteins to the pS2 promoter. We and others (43, 46, 48) have shown that ICI treatment suppresses pS2 gene expression, indicating that recruitment of CBP to this gene’s promoter does not support its transcription, and we recently demonstrated that RNA polymerase II is not recruited to the pS2 gene in ICI-treated MCF-7 cells (48). This is reminiscent of the experiment that demonstrated that addition of CBP before SRC-1 blocked progesterone receptor activity in in vitro transcription assays (58). In ChIP time course experiments, SRCs are recruited to the pS2 promoter before CBP (7, 9) and it is possible that ordered recruitment of SRCs before CBP is important for gene expression. A more active role for CBP/p300 relative to ICI inhibition of pS2 gene expression also may be possible based on the coactivator’s ubiquitin ligase activity (59) and the well-documented ability of ICI to promote rapid degradation of ER through a proteasome-dependent pathway (42, 60). It has been demonstrated that p300, in conjunction with MDM2, serves as both coactivator and ubiquitin ligase for p53, and is required for polyubiquitination and subsequent proteasome-mediated degradation of this transcription factor (59). It is possible that a component of the ICI-dependent inhibition of pS2 gene expression is due to recruitment of the ubiquitin ligase activity of CBP/p300 and subsequent proteasome-dependent degradation of ubiquitinated ER and/or other coregulators; this possibility is currently being investigated. Thus, in the context of pure antiestrogens, CBP/p300 does not serve as a coactivator of transcription and may potentially play a more active role in inhibiting gene expression.

Although the expression of pS2 and many other ER target genes are inhibited by ICI treatment, other genes have been identified the expression of which is stimulated by this pure antiestrogen. One such gene is the monoxygenase, cytochrome P450 CYP1B1, which can modify E2 by 4-hydroxylation as well as metabolize a wide range of toxicants and carcinogenic chemicals (46, 61). Microarray analyses demonstrated that CYP1B1 expression also is increased by E2, 4HT, and raloxifene (46). Estrogen regulation of CYP1B1 mRNA is mediated via ER binding to an imperfect ERE located in the gene’s promoter (47), and we have shown previously by ChIP assay that both ER and SRC-3 are recruited to this region in response to E2 treatment (48). Although data in this report indicate that ER is recruited to the promoter regardless of whether cells were treated with E2 or ICI, we detected binding of SRC-1 and CBP only in E2-treated cells. The difference in ICI-induced CBP binding to pS2 vs. CYP1B1 gene promoters indicates that this recruitment is promoter specific, perhaps due to differences within the ERE sequences or their context within the respective promoters.

At the present time, it is unclear how ICI increases CYP1B1 mRNA expression, and whether this is a transcriptional event. However, examination of other genes positively regulated by ICI suggests several possibilities. Expression of heat shock protein 27 is induced by ICI 164,384, and this is thought to be mediated through binding of ER via Sp1 to a GC-rich region of the promoter (62). Conversely, induction of p21 mRNA expression after ICI treatment of MCF-7 cells is the result of release of ER and histone deacetylase 1 from Sp1 sites located in the gene’s promoter (63). The ICI antiestrogen also has been associated with induction of gene expression via ER binding to AP-1 sites, and evidence suggests that this may be achieved via a squelching mechanism that removes repressors from binding to AP-1 (64, 65). Whether ICI recruitment of CBP plays a role in regulation of gene expression under any of these contexts remains to be determined.

In conclusion, our study demonstrates that CBP/p300 binding to ER in the presence of the pure antagonist ICI is distinct from p160 interaction with the ER ligand-binding domain and that this interaction depends on the nature of a long side chain protruding from the 7-position of the steroidal backbone of these antiestrogens. Recruitment of CBP to the promoters of ER target genes is gene specific, suggesting that binding of CBP to ICI-bound ER contributes to the array of mechanisms that determine the gene- and cell-specific action of the ligands for this receptor. Finally, this work provides novel insight into the mechanistic differences between type I and II (partial and pure, respectively) antiestrogens that should help to clarify the distinct effects of these pharmacological agents critical for prevention and treatment of hormone-dependent cancers.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Plasmids
The pBIND vector encoding the GAL4-DNA binding domain (DBD; amino acids 1–147) and pACT plasmid encoding the herpes simplex virus VP16 activation domain (amino acids 411–456), as well as the pG5-Luc reporter gene containing five DNA binding sites for the GAL4 DBD, a TATA box and luciferase reporter gene were purchased from Promega Corp. (Madison, WI). The pBIND expression vectors for GAL-SRC-1e, GAL-TIF2, GAL-RAC3, and GAL-CBP have been described previously (43). An expression vector for GAL-p300 was obtained from Dr. Tso-Pang Yao (Duke University, Durham, NC). The VP16-ER LBD chimera expression vector has been described previously (45). The mammalian expression vectors for the VP16-LBD point mutants (I358D, K362D, V376D, and L539A) were generated by PCR amplification. A portion of the LBD cDNA encoding amino acids 302–595 was amplified using the full-length hER version of these mutants (60) as template with the 5'- and 3'-primers, 5'-CGGGATCCCTAAGAAGAACAGCCTGGCCT-3' and 5'-GCTCTAGATCAGACTGTGGCAGGGAAACC-3', respectively, and the resulting PCR products were subcloned into pCR3.1 vector using a TA Cloning kit (Invitrogen, Carlsbad, CA). From these plasmids, BamHI-NotI restriction enzyme fragments were purified and subsequently subcloned into the corresponding sites of pACT such that the LBDs are in frame with and downstream of the VP16 activation domain. To generate cDNAs of CBP domains for the mammalian two-hybrid assays, appropriate regions of the CBP cDNA were PCR amplified using primers listed in Table 2 and subsequently subcloned into pBIND using PCR and the TA Cloning kit as described above. The CH3 domain mutants were generated by site-directed mutagenesis using the appropriate primers (Table 3). The GST-CH3 construct was made by a similar method as for pBIND-CH3 except with the use of the pGEX-4T-1 expression vector (Amersham Biosciences, Piscataway, NJ). The pCR3.1-hER-179C construct consists of amino acids 179–595 of ER and has been described previously (66). All constructs made through use of PCR were sequenced to verify that mutations did not occur during DNA amplification.

Mammalian Two-Hybrid Assay
In most mammalian two-hybrid assays, HeLa cells were transfected with 100 ng of the expression vectors for GAL-p160 coactivators (pBIND-SRC-1e, pBIND-TIF2, pBIND-RAC3, or pBIND-CBP) or GAL-CBP domains. Along with pBIND expression vectors, 1000 ng of expression vector for VP16-ER-LBD (pACT-hER-LBD) and 1000 ng of pG5-Luc reporter plasmid were used. In p300-ER interaction assays, 50 ng of GAL-p300 was transfected along with 1000 ng of VP16-ER-LBD and 1000 ng of pG5-Luc reporter. Control experiments employed equivalent amounts of the pACT and pBIND empty vectors. Cells were treated with either 0.1% ethanolic vehicle, 10^–8 M E2 (Sigma-Aldrich, St. Louis, MO), 10^–7 M 4HT (kindly supplied by D. Salin-Drouin, Laboratoires Besin Iscovesco, Paris, France), raloxifene, or idoxifene (kind gift of GlaxoSmithKline, King of Prussia, PA), ICI, or ICI 164,384 (kindly supplied by A. Wakeling, Zeneca Pharmaceuticals, London, UK), or ZK-703 or ZK-253 (kind gift from J. Hoffmann, Schering AG, Berlin, Germany) for 20–24 h before cell harvest and assay as described above.

RBA Assays
The RBA of wild-type and mutant ERs for ICI was determined in vivo as previously described (67). Briefly, 2 µg of the expression plasmids for wild-type ER (pCR3.1-hER) or for the corresponding I358D, K362D, V376D, or L539A mutants (60) were transfected into HeLa cells using Lipofectamine Plus according to the manufacturer’s instructions. Media were aspirated from wells 16–18 h later and replaced with phenol-red free DMEM containing 5% sFBS, approximately 1.5 pmol [³H]estradiol (250 mCi) (PerkinElmer Life Sciences, Wellesley, MA), and increasing concentrations (from 10^–10 to 10^–3 M) of either unlabeled E2 or ICI. After incubation for 2 h at 37 C, media were aspirated, and cells were washed three times with ice-cold PBS and incubated in 100% ethanol for 10 min at room temperature to extract bound steroid. The amount of ER-bound [³H]E2 in the ethanol extract was quantified with a Beckman LS 6500 scintillation counter (Beckman Instruments, Fullerton, CA) and Biodegradable Counting Scintillant (Amersham Biosciences). The RBA was calculated as the ratio of the concentration of ICI in comparison with E2 required to reduce [³H]E2 binding by 50% times 100. The RBA value of nonradiolabeled E2 was defined as 100.

Coimmunoprecipitation
For studies employing recombinant ER, HeLa cells (5 x 10⁶) were harvested and incubated for 30–60 min on ice in lyses buffer [50 mM HEPES (pH 7.5), 100 mM KCl, 0.2 mM EDTA, and 0.1% Nonidet P-40 (NP40)] supplemented with Complete Mini-Tablets protease inhibitor tablets (Roche Applied Sciences, Indianapolis, IN). Thereafter, the cell lysate was centrifuged for 5 min at 21,000 x g, and its protein content was quantified using Bio-Rad Protein Assay reagent. Before immunocomplex preparation, 100 ng of recombinant ER (PanVera, Madison, WI) was incubated with either ethanol or 270 ng of estradiol or ICI in 30 µl of buffer [50 mM Tris-HCl (pH 7.5) 10 mM MgCl₂, 1 mM EGTA, and 2 mM dithiothreitol] for 30 min. The immunocomplex was prepared in a total volume of 1 ml containing 60 µl (50% slurry) of prewashed protein G+ agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), 0.5 mg cell lysate, 100 ng of ligand-bound ER, and either 1 µg of anti-CBP (C-1) antibody (Santa Cruz), or 1 µg of anti-SRC-1 antibody (GeneTex, San Antonio, TX). The immunoprecipitation reaction was incubated at 4 C for 1.5 h with constant rotation and then centrifuged and washed three times with lysis buffer. Subsequently, the immunocomplex was boiled for 5 min in 50 µl of 2x Laemmli solution and resolved by 7.5% SDS-PAGE. The Western blot was probed with H222 hER antibody (gift of Abbot Laboratories, Abbot Park, IL) in 1% dry milk powder-Tris-buffered saline-Tween, followed by mouse antirat antibody conjugated to horseradish peroxidase (HRP; ICN Biochemicals, Inc., Aurora, OH). Protein bands were detected using ECL plus Western blotting detection system (Amersham Biosciences) and X-Omat film (Perkin Elmer).

For studies examining the interaction between endogenous ER and CBP, MCF-7 cells (5 x 10⁶ per 100-mm dish) were grown in phenol-red free DMEM containing 10% sFBS until they reached approximately 80% confluency, and were then treated with either vehicle (0.1% ethanol), 10^–6 M ICI, or 10^–8 M E2 for 15 min. Cells were harvested and incubated in lysis buffer [20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 5% glycerol, 1 mM Na₃VO₄, and 1 mM NaF) supplemented with Complete Mini-Tablets protease inhibitor tablets at 4 C with rotation for 30–60 min. The cell lysates were centrifuged for 5 min at 20,000 x g, and protein contents were quantified using Bio-Rad protein assay reagent. After preclearing with prewashed protein G agarose beads (Amersham Biosciences) for 2 h at 4 C with rotation, the cell lysates (0.5 mg) were incubated with either a human CBP antibody (2 µg; BD Biosciences, San Jose CA) or normal mouse IgG (2 µg; Santa Cruz) at 4 C with rotation for 2 h. Thereafter, 25 µl of a 50% slurry of prewashed protein G agarose beads was added and the incubation continued for another 2 h at 4 C, followed by four washes with lysis buffer. Subsequently, the immunocomplex was boiled at 75 C for 10 min in 25 µl of 1x Laemmli buffer and resolved by 4–12% SDS-PAGE (Invitrogen), and transferred to nitrocellulose. The blot was probed with antibodies directed against human ER (HC-20), mouse CBP-NT (Upstate Biotechnology, Inc., Charlottesville, VA) or ß-actin (Chemicon International, Temecula, CA) in PBS containing 5% skim milk powder (wt/vol) and 0.05% Tween 20, followed by the appropriate HRP-conjugated antimouse or antirabbit antibody (ICN). Protein bands were detected using ECL plus Western blotting detection system and X-Omat film.

ChIP Assay
MCF-7 cells (6 x 10⁶) per 150-mm plate were grown for 4 d in DMEM supplemented with 10% sFBS. After that, cells were treated with 2.5 µM -amanitin (Sigma-Aldrich, St. Louis, MO) for 1.5 h and then washed twice with serum-free DMEM media. Thereafter cells were treated with either 0.1% ethanol, 10 nM E₂, or 100 nM ICI for 15–60 min. Cross-linking was done as follows: the media were replaced with PBS and then cells were treated with 1% formaldehyde for 10 min at 37 C. Cells were then washed twice with ice-cold PBS and harvested by scraping into 1 ml of ice-cold PBS supplemented with Complete Mini-Protease Inhibitor cocktail. Cells were centrifuged and then incubated in 500 µl of lysis buffer [1% sodium dodecyl sulfate (SDS), 10 mM EDTA, 50 mM Tris-HCl (pH 8.0)] for 30 min on ice or 1 ml of nuclei preparation buffer [5 mM piperazine-N,N'-bis (2-ethanesulfonic acid) (pH 8.0), 85 mM KCl, 0.5% NP-40 and Complete Mini-Protease Inhibitors] at 4 C for 30 min, followed by sonication pulses of 8 sec that were repeated four times. After centrifugation, 20 µl of the supernatants was retained as inputs, and the remainder was diluted 10-fold in dilution buffer [16.7 mM Tris-HCl (pH 8.1), 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 100 mM NaCl and a Complete Mini-Protease Inhibitor tablet]. The diluted lysate was precleared with 40 µl of a 50% slurry protein-A sepharose beads (Sigma-Aldrich) mixed with 4 µg of sheared salmon sperm DNA at 4 C for 30 min. This diluted fraction was subjected to immunoprecipitation overnight using the appropriate antibody, after which the immunocomplexes were recovered by 1 h incubation at 4 C with 60 µl of protein A-Sepharose and 4 µg of sheared salmon sperm DNA. Precipitates were washed serially with 1 ml Washing Buffer I [20 mM Tris-HCl (pH 8.0), 2 mM EDTA, 0.1% SDS, 1% Triton X-100, and 150 mM NaCl], Washing Buffer II [20 mM Tris-HCl (pH 8.0) and 2 mM EDTA], and Washing Buffer III [10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 1% NP-40, 1% deoxycholate, and 0.25 M LiCl] and then twice with 1x TE [10 mM Tris-HCl (pH 8.0), 1 mM EDTA]. Precipitated chromatin complexes were eluted from the beads by two 15-min incubations with 100 µl of elution buffer (1% SDS and 0.1 M NaHCO₃). Cross-linking was reversed by overnight incubation at 65 C, and DNA was purified with QIAquick columns (QIAGEN, Valencia, CA) and then subjected to quantification applying 40 cycles of amplification using pS2 primers and probe (Table 4) or CYP1B1 primers (48) and the Real Time PCR system protocol (Applied Biosystems, Foster City, CA).

In Vitro Protein-Protein Interaction by GST Pull-Down Assay
The GST-CH3 fusion protein was expressed in isopropyl-ß-D-thiogalactopyranoside-induced Escherichia coli BL21 cultures as recommended by the manufacturer (Amersham Biosciences). Bacteria were collected by centrifugation and resuspended in 5 ml sonication buffer [20 mM HEPES (pH 7.9), 80 mM KCl, 6 mM MgCl₂, 1 mM dithiothreitol, 1 mM ATP, 1% Triton X-100, 1% NP-40 and 1% glycerol (33)], and sonicated four times with 10-sec pulses. The sonicated material was centrifuged and the supernatant was incubated with prewashed Glutathione Sepharose 4 Fast Flow beads (Amersham Biosciences) for 1 h with rotation at 4 C. After that, beads were centrifuged at 500 x g and washed three times with 1x TE. The GST-CH3-bound beads were resuspended in 100 µl binding buffer [20 mM HEPES (pH 7.9), 125 mM KCl, 10 mM MgCl₂, 1 mM dithiothreitol, 1 mM EDTA (pH 8.0), 2 µg BSA, 0.08% TritonX-100, 0.08% NP40 and 1% glycerol]; CH3 protein was eluted from 10 µl of beads and examined on Coomassie-stained SDS-PAGE to evaluate relative protein expression. Three aliquots of 25 µl of CH3-bound beads were used for the pull-down assay. Each aliquot was added to a tube containing 400 µl of binding buffer, 5 µl of ER-179C that was previously expressed using TNT in vitro expression kit (Promega), and labeled with [³⁵S]methionine (Amersham Biosciences). Four microliters of either ethanol, estradiol (100 µM), or ICI (1 mM) was added to the mix. The pull-down mix was incubated at 4 C for 2 h with rotation. Thereafter the beads were washed three times with NETN buffer [20 mM Tris-HCl (pH 8.0), 1 mM EDTA, 0.5% NP40, and 0.5 M NaCl]. The pull-down complex was denatured by 1x reducing agent (Invitrogen) at 70 C for 10 min and resolved by 8% SDS-PAGE followed by electrotransferring to nitrocellulose membrane. The resulting membrane was dried at room temperature and then exposed to X-Omat Blue Film (PerkinElmer) for 16 h at –80 C.

	ACKNOWLEDGMENTS

 
We thank Dr. Martin Dutertre and Dr. Kevin Colman for making available several of the expression vectors used in this study. The assistance of Laura C. Savery for the RBA assays, and the technical support of Judy Roscoe, Cheryl Parker, and Estrella Foster are gratefully acknowledged.

	FOOTNOTES

 
This work was supported by grants from the National Institutes of Health (DK53002 and DK64038) and the Department of Defense (W81XWH-04–1-0424) (to C.L.S.).

Present address for B.M.J.: Department of Biological Sciences, The University of Jordan, Amman 11942, Jordan.

First Published Online July 13, 2006

Abbreviations: AF, Activation function; CBP, cAMP response element-binding protein (CREB)-binding protein; ChIP; chromatin immunoprecipitation assay; CH1, -2, and 3, cysteine-histidine rich domains 1–3; CYP1B1, cytochrome P450 1B1; DBD, DNA-binding domain; ER, estrogen receptor; ERE, estrogen response element; GST; glutathione-S-transferase; HRP, horseradish peroxidase; 4HT, 4-hydroxytamoxifen; KIX, CREB binding domain; LBD, ligand-binding domain; NID, nuclear receptor interacting domain; NP40, Nonidet P-40; RAC3, receptor-associated coactivator 3; RBA, relative binding affinity; SDS, sodium dodecyl sulfate; SERM, selective ER modulator; sFBS, charcoal-stripped fetal bovine serum; SRC-1, steroid receptor coactivator-1; TIF2, transcription intermediary factor 2.

Received for publication June 1, 2005. Accepted for publication July 6, 2006.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 

1. Nilsson S, Mäkelä S, Treuter E, Tujague M, Thomsen J, Andersson G, Enmark E, Pettersson K, Warner M, Gustafsson J-Å 2001 Mechanisms of estrogen action. Physiol Rev 81:1535–1565[Abstract/Free Full Text]
2. Tora L, White JH, Brou C, Tasset DM, Webster NJG, Scheer E, Chambon P 1989 The human estrogen receptor has two independent nonacidic transcriptional activation functions. Cell 59:477–487[CrossRef][Medline]
3. Lees JA, Fawell SE, Parker MG 1989 Identification of two transactivation domains in the mouse oestrogen receptor. Nucleic Acids Res 17:5477–5487[Abstract/Free Full Text]
4. Tzukerman MT, Esty A, Santiso-Mere D, Danielian P, Parker MG, Stein RB, Pike JW, McDonnell DP 1994 Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol 8:21–30[Abstract]
5. McKenna NJ, O’Malley BW 2002 Combinatorial control of gene expression by nuclear receptors and coregulators. Cell 108:465–474[CrossRef]