Comparison of tamoxifen ligands on estrogen receptor interaction with estrogen response elements

Molecular and Cellular Endocrinology 143 (1998) 79 – 90

Comparison of tamoxifen ligands on estrogen receptor interaction with estrogen response elements Carolyn M. Klinge a,*, April L. Studinski-Jones b, Peter C. Kulakosky a, Robert A. Bambara c, Russell Hilf c a

Department of Biochemistry and Molecular Biology, Uni6ersity of Louis6ille School of Medicine, Louis6ille, KY 40292, USA b Department of Genetics, Uni6ersity of Pittsburgh, Pittsburgh, PA 18, USA c Department of Biochemistry and Biophysics, Uni6ersity of Rochester School of Medicine, Rochester, NY 14642, USA Received 20 May 1998; accepted 24 June 1998

Abstract The estrogen receptor (ER) is a ligand-activated transcription factor that binds to specific DNA sequences, estrogen response elements (EREs). Estradiol-liganded ER (E2-ER) binds cooperatively to stereoaligned EREs that are surrounded by naturally-occurring AT-rich sequences with a stoichiometry of one E2-ER dimer per ERE. When ER is bound by 4-hydroxytamoxifen (4-OHT), the active metabolite of the widely used therapeutic antiestrogen tamoxifen (TAM), the receptor binds to EREs with high affinity. However, one molecule of 4-OHT ligand dissociates from the ER dimer apparently during the process of binding to DNA, yielding a stoichiometry of one [3H]4-OHT molecule per ERE. To determine whether DNA-binding induced ligand dissociation is a general property of type I antiestrogens that are not covalently attached to the ER, we examined the interaction of ER liganded by tamoxifen (TAM) with EREs. We demonstrate that TAM-ER binds EREs with lower affinity than E2-ER, 4-OHT-ER, or ER liganded by the covalent antiestrogen tamoxifen aziridine. Unlike E2-ER, both TAM and 4-OHT-ER bind EREs non-cooperatively. Like 4-OHT, TAM appears to dissociate from the liganded ER as the receptor binds EREs. Additionally, partial proteolysis of ERE-bound ER by trypsin revealed different cleavage patterns for E2 versus 4-OHT and TAM. These findings indicate that the behavior of the ER liganded by TAM is generally similar to that of the antiestrogen 4-OHT. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Estrogen receptor; Tamoxifen; 4-Hydroxytamoxifen; Tamoxifen aziridine; Estrogen response element (calf uterus)

1. Introduction Lifetime exposure to estrogens is widely accepted as a major risk factor for the development of breast cancer. Estrogens promote cell replication by binding to a nuclear protein, the estrogen receptor (ER), that is encoded by two genes, a and b (Katzenellenbogen and Korach, 1997). ER is a member of the steroid/ nuclear receptor superfamily of transcriptional enhancers (Mangelsdorf et al., 1995). Each acts as a * Corresponding author. Tel.: +1 502 8523668; fax: + 1 502 8526222; e-mail: [email protected]

ligand-dependent and sequence-specific transcription factor. Ligand, e.g. estradiol (E2), binding initiates a series of steps leading to formation of ‘activated’, dimeric E2-ER complex that binds with high affinity to specific DNA sequences: estrogen response elements (EREs), that are located adjacent to the coding regions of estrogen-responsive genes. Sequence analysis of the 5%-regulatory regions of numerous estrogen responsive genes led to derivation of a 13 bp palindromic inverted repeat (IR) consensus sequence: 5%GGTCAnnnTGACC-3% (Klein-Hitpass et al., 1986), referred to as the minimal ERE, where n equals any nucleotide in the center spacer region. EREc con-

0303-7207/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S0303-7207(98)00130-0

80

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

ferred estrogen responsiveness to reporter genes analyzed by transfection assay (Klein-Hitpass et al., 1986; Berry et al., 1989). Following ERE binding, the liganded-ER-ERE complex is thought to interact with other transcription components, e.g. co-activator or co-repressor proteins, other transcription factors, and components of the TATA binding factor complex, altering gene expression (Horwitz et al., 1996). Blocking estrogen action by tamoxifen (TAM) treatment is one adjuvant therapy for women with breast cancer. The primary mechanism by which antiestrogens inhibit estrogen action is by direct binding competition with E2 for the ligand binding domain (LBD) of the ER (Klinge et al., 1992a). Subsequent actions of the antiestrogen-liganded ER remain a matter of intense study and debate. TAM elicits mixed agonist-antagonist activity depending on cell type in vivo. Although 4-OHT binds ER with higher affinity than the parental TAM or the endogenous ligand E2 (Katzenellenbogen et al., 1984), no one has examined how TAM impacts the specific interaction of ER with EREs. To probe the mechanism of antiestrogen action, we measured the binding of antiestrogen-versus E2-liganded ER complexes to EREc38 in vitro (Klinge et al., 1992a,b; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996a,b, 1997a). We initially examined the ERE binding of [3H]4-hydroxytamoxifen-liganded ER ([3H]4OHT). The 4-OHT-ER complex readily binds the ERE with an affinity comparable to that of the E2-ER (Klinge et al., 1992a,c; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996a,b), indicating that antagonist activity does not derive from simple binding inefficiency. However, E2-ER, but not 4-OHT-ER, bound cooperatively to stereoaligned EREs surrounded by naturally occurring AT-rich flanking sequences (Klinge et al., 1992b,c; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996a). In contrast, EREs lacking these flanking sequences did not bind ER cooperatively, regardless of ERE spacing or stereoalignment (Klinge et al., 1992b; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996a,b). We demonstrated that when 4-OHT-ER binds an ERE in vitro, one 4-OHT ligand molecule dissociates. We suggested that this may be the basis for its antagonistweak agonist activity when compared to the pure agonist E2 (Klinge et al., 1996a,b). In the present study we examined the ERE binding affinity and stoichiometry of TAM-liganded ER interaction with selected EREs and their variants. We present data showing that TAM-liganded ER binds EREs with significantly lower affinity compared to E2-, 4-OHT-, or tamoxifen aziridine (TAz)-liganded ER. We also utilized partial proteolysis of the ER liganded with E2, 4-OHT, or TAM in the presence or absence of EREs to evaluate conformational changes elicited by ligand and ERE binding.

2. Materials and methods

2.1. ER purification from calf uterus ER was partially purified from an ammonium sulfate fractionated, high speed cytosolic preparation from calf uterus by heparin agarose affinity chromatography (Klinge et al., 1992c). ER was liganded with either 17b-[2,4,6,7,16,17-3H]E2 (142 Ci/mmol from NEN), Z4[N-methyl-[3H] 4-OHT (71 Ci/mmol from Amersham), [3H]tamoxifen ([3H]TAM, 84.5 Ci/mmol from NEN), or [ring-3H]tamoxifen aziridine ([3H]TAz, 23 Ci/mmol from Amersham). The concentration of ER was determined by adsorption to hydroxylapatite (HAP) (Pavlik and Coulson, 1976). All receptor concentrations refer to dimeric ER, i.e. with two molecules of ligand bound. When using antiestrogens, the antiestrogen-ligandedER was protected from exposure to light.

2.2. Preparation of baculo6irus-expressed recombinant human ERa A recombinant AcMNPV containing the coding sequence for wild-type recombinant human ERa was generously provided by Dr Nicholas J. Koszewski of the University of Kentucky (Obourn et al., 1993). IPLB-SF-21AE insect cells were maintained in stationary flasks in TnMFH medium (Gibco, Grace’s supplemented) with 10% fetal bovine serum (Gibco) (Vaughn et al., 1977). The cells were syncronously infected at a density of 7.5× 104/cm2 (Richardson, 1995). After 3 days, i.e. before significant cell lysis, the cells were pelleted. The cell pellet was washed in phosphate buffered saline and resuspended in 2 packed cell volumes of hypotonic buffer (10 mM Tris–HCl pH 7.5, 10 mM KCl, 10% glycerol, 10 mM DTT, 1 mM EDTA, 1 mM PMSF, 2 mg/1 each in aprotinin, leupeptin, and pepstatin, and 10 mg/l E64 (Boehringer Mannheim). After a 20 min incubation on ice, the cells were homogenized with a Dounce homogenizer. Cell lysis and nuclear integrity were monitored by observation with a phase-contrast microscope and homogenization was stopped when 90% of the cells were lysed. Nuclei were pelleted and a nuclear extract (NE) was prepared by resuspending the nuclear pellet in 50 mM Tris–HCl pH 7.5, 0.6 M KCl, 10% glycerol, 10 mM DTT, 1 mM EDTA, 1 mM PMSF, 2 mg/l each in aprotinin, leopeptin, and pepstatin, and 10 mg/l E64. The nuclei were sonicated briefly with a Branson sonicator equipped with a microprobe and incubated for 3 h on ice. The NE was clarified in a micro-ultracentrifuge (Beckman TL100) using a TLS55 swinging bucket rotor at 259000× g for 20 min. The resulting supernatant is termed NE and was stored at − 70°C. The ER concentration in the NE was determined by HAP assay (Pavlik and Coulson, 1976).

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

2.3. Preparation of EREs A synthetic single-stranded oligonucleotide corresponding to each of the ERE sequences analyzed here is shown in Table 1. Double stranded oligomers were ligated into the Sma I restriction site of the vector pGEM-7Zf(+) (Promega) as described (Klinge et al., 1992c; Anolik et al., 1995). Although the inserts vary in lengths, linearized plasmid DNA was used for the microtiter plate assay experiments described below, obviating the small differences in insert length.

2.4. Microtiter plate assay of ER-ERE binding The microtiter (well) plate assay is an equilibrium binding assay that quantitates ER-ERE binding based on measuring [3H]liganded-ER binding to [35S]DNA (Ludwig et al., 1990). For each plate assay presented here, plasmid DNA was linearized with EcoRI. Aliquots of EcoRI-digested plasmid DNA were labeled by incorporation of [35S]dATP (\600 Ci/mmol, NEN) at the recessed 3% termini using the Klenow fragment of E. coli DNA polymerase I and mixed with unlabeled EcoRI-digested DNA for the desired final concentration. Briefly, for saturation binding analysis, various concentrations of heparin agarose affinity purified [3H]E2ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER were preincubated with one concentration (approximately 0.22 nM) of [35S]DNA (plasmid DNA with or without ERE) for 2.5 h at 4°C. A total of 50-ml aliquots of the receptor-DNA equilibrium mixture were then Table 1 Sequence of EREs used in experiments Name

Sequence

EREc38

5%-CCAGGTCAGAGTGACCTGAGCTAAAATAACACATTCAG-3% 5%-CCGGTCAGAGTGACC-3% 5%-CCAGGTCAGAGTGACCTG AGCTGACAGGACTCGACCAG-3% 5%-CCAGGTCACTGTGACCTGAGCTAAAATAACACATTCAG-CTAGCACTGACGCTAGCGAGCTAAAATAACACAT TCAG-CCAGGTCACTGTGACCTGAGCTAAAATAACACATTCAG3% 5%-CCTGGTCACTGTGACC GGGGTTGGGAAATCGATAAGCTT-GTTACAAGCTTGGATCGGAGAGCTCCCAACGCGTTGA-TGCAGGTCACTCTGACCTG GTGCA-3%

EREc EREGC ERE7.2AT+

ERE7.2AT−

EREc38 is a 38bp ERE consensus sequence (Peale et al., 1988). The underlined nucleotides correspond to the minimal core consensus ERE. GC, indicates that that AT-rich portion of EREc38 was substituted by a synthetic GC-enriched sequence; +, indicates the presence of the consensus AT-rich sequence and − denotes its absence (Anolik et al., 1995).

81

incubated in histone/gelatin-coated microtiter wells for 2.5 h at 4°C. Wells were rinsed, and the radioactivity remaining in the wells was counted using EcoScint A (National Diagnostics, Atlanta, GA). Calculation of specific [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER binding to EREs was previously described, with binding to pGEM-7Zf(+ ) parental plasmid alone subtracted from binding to plasmid containing an ERE construct (Ludwig et al., 1990; Klinge et al., 1992b,c; Kladde et al., 1996; Klinge et al., 1996a).

2.5. Electrophoretic mobility shift assay (EMSA) EMSA was performed as described (Klinge et al., 1996b), with the following [32P]EREc38 (77bp), obtained by EcoRI-BamHI digestion of insert-containing pGEM-7Zf(+ ). Binding reactions contained 10 fmol (25000 dpm) 32P-labeled DNA and ER, 5 mg poly d(I’C) (Midland Certified Reagent, Midland, TX) and other reaction components in the amounts indicated in the Figure legend. Binding reactions were incubated on ice for 2.5 h and 40 ml aliquots were loaded on 4% polyacrylamide non-denaturing gels and electrophoresed. Gels were dried under vacuum and autoradiographed on Kodak X-Omat film with an intensifying screen. A total of 1 ml of 1:10 diluted ER monoclonal antibody H222, a gift of Abbott Laboratories, was added to selected samples in each experiment to confirm the identity of ER protein in the shifted ER-ERE complexes. The amount of ER-ERE complex formed, and that of free ERE, was determined by excision of the corresponding regions from the dried gels into scintillation vials containing 3 ml of EcoScint A, and the radioactivity was counted. The fraction of total [32P]ERE in the ER-ERE complex was calculated as follows: F(t)= (cpm in the ER-ERE complex)/(total cpm in the lane), where the (total cpm in the lane) = (cpm in Free ERE)+ (cpm in the ER-ERE complex) (Hoopes et al., 1992).

2.6. Association time course gel mobility shift assays For determining the effect of ligand on the association rate and time to achieve ER-ERE binding saturation, liganded ER was incubated with [32P]EREc38, as above, for 0–360 min at 4°C prior to gel shift analysis of ER-ERE complex formation. The binding reactions were timed so that immediately after addition of [32P]EREc38 to the final (to) tube, 40 ml of this sample was loaded into a lane of the gel and current was applied. The to value was counted as 0.5 min. The running of the gel and quantitation of the resulting [32P]DNA are described above.

82

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

2.7. Partial ER proteolysis Recombinant hERa was prepared as a NE from SF21 cells as described above. The concentration of ER was determined by HAP assay (Pavlik and Coulson, 1976). The NE was incubated with 1000-fold molar excess of the indicated ligand and either TE or the 25 nM 4(EREc38) (four tandem, head-to-tail copies of EREc38, sequence given in Table 1) in a total reaction volume of 18 ml in TDPEK100 buffer (40 mM Tris– HCl, pH 7.5; l mM DTT, 0.5 mM PMSF, 1 mM EDTA, 100 mM KCl) containing 10% glycerol. After a 30 min incubation at 30°C, the tubes were equilibrated for 5 min at RT. To each tube 2 ml of trypsin stock was added to give the final concentration shown in the Figure legend. The trypsin digestion was terminated by addition of 3 ml of Laemmeli SDS-PAGE loading buffer and microwaving the samples (Horscroft and Roy, 1997). Samples were applied to a 10% polyacrylamide mini-SDS PAGE gel. Conditions for running the gel and Western blotting have been described (Klinge et al., 1997b). AER320 monoclonal antibody to ER was used in immunoblotting (Neomarkers, Lab Vision, Fremont, CA). The interacting proteins were detected by chemiluminescence (ECL, Amersham) on BIOMAX ML (Kodak) film (Klinge et al., 1997b).

3. Results

3.1. Saturation analysis of specific liganded-ER binding to EREc38 One of the more intriguing observations from our study differentiating the agonist versus antagonist actions of ER ligands at the level of ERE binding was the finding that one of the two ER-bound 4-OHT molecules dissociated from the ER when the ER complex bound to ERE constructs in vitro (Klinge et al., 1992b, 1996a). In order to determine whether ligand dissociation is a general property of a triphenethylene antiestrogens, we studied the interaction of TAM-liganded ER to selected EREs in vitro. The binding of E2-ER, 4-OHT-ER, TAM-ER, or TAz-ER to EREc38 was quantitated by an equilibrium binding plate assay (Fig. 1). The saturation binding profiles for E2-ER and TAz-ER were comparable, indicating a stoichiometry of one E2-ER or one TAz-ER dimer bound per EREc38 (Table 2). In contrast, at saturation, 4-OHTER-EREc38 binding was about 50% of that for E2-ER or TAz-ER-ERE, giving an apparent stoichiometry of 0.5 4-OHT-ER dimer, or one 4-OHT-ER monomer, per EREc38. TAM-ER-EREc38 binding at saturation was less than that of 4-OHT-ER, giving an apparent stoichiometry of 0.3 TAM-ER dimer, or less than one TAM-ER monomer, per EREc38 (Table 2). It is impor-

Fig. 1. Saturation analysis of [3H]liganded-ER binding to one copy of EREc38. EcoRI-linearized [35S]dATP-end labeled plasmid DNA, either the parental plasmid alone or plasmid containing a single copy of EREc38 (sequence in Table 1) was incubated with increasing concentrations of heparin-agarose-purified [3H]4-OHT-ER (open circles), [3H]E2-ER (closed circles), [3H]TAz-ER (closed triangles), or [3H]TAM-ER (open inverted triangles). The data points shown are the average of quadruplicate determinations 9S.E.M. from three to seven representative plate assay experiments, and are calculated for binding to 11 fmol of DNA/well as described in Section 2.

tant to note that all preparations of ER bound identical molar amounts of [3H]E2, [3H]TAz, [3H]4-OHT, or [3H]TAM, thus obviating miscalculation of the quantity of ER based on radioactivity. Additionally, all liganded ER preparations displayed similar binding profiles, indicating that differences in EREc38 binding for E2-ER or TAz-ER versus 4-OHT-ER and TAM-ER were independent of receptor preparation. Scatchard analysis revealed that E2-ER, 4-OHT-ER, TAM-ER and TAzER bound to one EREc38 with similar high affinity (Table 3). Hill coefficients (nH) were calculated (Table 3). The nH values of approximately 1 indicate non-cooperative binding of liganded ER to EREc38 in vitro. To determine whether ER-ERE binding kinetics are altered with TAM versus 4-OHT, TAz, or E2 as ER ligand, the affinity of variously liganded ER-ERE binding was measured to two, three, or four tandem copies of EREc38 under equilibrium binding conditions. These data are summarized in Fig. 3. As reported, for 2, 3 or 4 tandem copies EREc38, the binding stoichiometry of either E2- or TAz- ER at saturation was one ER dimer per ERE and that of 4-OHT-ER binding was consistently lower (Klinge et al., 1996a). Likewise, the stoichiometry of TAM-ER-EREc38 binding was also lower than the expected one ER dimer per ERE. Approximately 0.5 4-OHT-ER dimer, or one 4-OHT-ER monomer, was bound per ERE, whereas the stoichiometry of TAM-ER-EREc38 binding was approximately 0.3, which is less than one ER monomer/ERE (Table 2). Although the molecular mechanism for this result is unclear, one possibility given the 2-fold lower

0.97 2.09 3.15 4.10

One Two Three Four

0.97 1.09 1.05 1.10

Ratio E2-ERERE 0.41 1.03 1.78 2.14

Ratio 4-OHT-ER-plasmid 0.41 0.52 0.59 0.54

Ratio 4-OHT-ERERE 0.31 0.62 0.52 0.82

0.31 0.31 0.15 0.22

Ratio TAM-ER-plas- Ratio TAM-ERmid ERE

1.01 1.95 3.00 4.01

Ratio Taz-ER-plasmid

1.01 0.97 1.00 1.00

Ratio Taz-ERERE

Saturation analyses were performed using a fixed concentration of plasmid DNA and increasing concentrations of heparin-agarose purified [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER as described in Section 2. [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER binding to pGEM-7Zf(+) plasmid alone or containing the number of tandem copies of EREc38 was measured as detailed in Section 2. The binding ratios presented were calculated from values taken at saturation (77 – 140 fmol/well [input 11 fmol DNA] [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER dimer added) from which background and non-specific binding to plasmid without inserts has been subtracted (Klinge et al., 1996a, 1992c).

Ratio E2-ERplasmid

cTandem EREc38

Table 2 Stoichiometric relationship of E2-ER, 4-OHT-ER, TAM-ER or TAz-ER-EREc38 interaction

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90 83

0.249 0.01 0.23 9 0.03

One Two Three Four

1.169 0.03 1.249 0.04 2.17 90.46 1.869 0.17

nH 24 22 23 22

n 0.169 0.01 0.69 90.02 0.409 0.01 3.55 9 0.18

Kd (nM)

4-OHT-ER n 13 12 8 11

nH B1a 0.74 90.02 0.71 90.01 1.01 90.04

nH

1.14 9 0.1b 1.1 90.02 1.46 90.08b 1.0 90.2 1.36 90.11b 1.1 90.12 1.55 90.04b 1.18 9 0.11

Kd (nM)

TAM-ER

15 25 21 23

n 0.27 90.03

Kd (nM)

TAz-ER

1.1 9 0.3 2.41 9 0.27 2.31 9 0.57 2.66 9 0.01

nH

Saturation analyses were performed using the plate assay, at a fixed concentration of plasmid DNA (11 fmol/well) and increasing concentrations of [3H]liganded-ER as described in Section 2. Data were plotted according to the methods of Scatchard and Hill to derive Kd values and Hill coeffficients (nH), respectively. The values shown were derived from two to four separate experiments for a total of n different concentrations of receptor (Klinge et al., 1996a). n, Number of different concentrations of [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER assayed. a Since 4-OHT-ER binding to 1 EREc38 is less than one 4-OHT-ER dimer per ERE, Log (Y/1−Y) is less than zero. The slope of the best fit line for the data was B1. b The value given is significantly different from that of E2-, 4-OHT-, or TAz- ER- binding to that number of tandem copies of EREc38 (PB0.001).

17 12 14 32

Kd (nM)

cERE c38 n

E2-ER

Table 3 Comparison of E2-ER, 4-OHT-ER, TAM-ER, or TAz-ER binding to EREc38

84 C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

85

Table 4 Comparison of E2-ER and TAM-ER binding to various ERE constructs ERE

EREc38 EREc ERE7.2(+) ERE7.2(−)

E2-ER

4-OHT-ER

TAM-ER

n

Kd (nM)

n

Kd (nM)

n

Kd (nM)

17 15 28 24

0.24 90.01 0.81 9 0.04a Cooperative 0.5 90.02

24 ND 14 14

0.16 90.01 ND 0.49 9 0.04 1.02 9 0.20

13 14 12 10

1.14 9 0.1b 1.79 9 0.11b 3.00 90.15b 2.7 90.1b

Saturation analyses were performed by the plate assay with 11 fmol plasmid DNA/well and increasing concentrations of 3H-liganded ER (Ludwig et al., 1990). Data were plotted according to the method of Scatchard and linear regression was used to calculate the Kd values. The values shown were derived from the saturation binding data of four separate experiments for a total of number (n) different concentrations of ER, each assayed in quadruplicate. Cooperative indicates that an accurate Kd value could not be obtained because of the positive cooperative appearance of the Scatchard Plot. In this case, Hill analysis revealed an nH value of 1.89 90.20 (Anolik et al., 1996). n, Number of different concentrations of [3H]E2 ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER assayed. Cooperatives, when curvilinear Scatchard plots, characteristic of positive cooperativity, were obtained, accurate Kd values can not be estimated. ND =not determined. a The value given is significantly different from that of E2-ER interaction with EREc38 (PB0.005). b The value given is significantly different from that of E2- or 4-OHT-ER-binding to that ERE sequence (PB0.001).

affinity of TAM versus 4-OHT for the ER (Katzenellenbogen et al., 1984) is that one or both molecules of TAM dissociate from the ER homodimer, leaving the ER bound to the ERE. We and others have demonstrated that ER does not bind EREs as a monomer (Klinge et al., 1992c; Anolik et al., 1996; Klinge et al., 1996a,b) and we do not detect a reduction in the mobility of the TAM-ER-ERE complex in gel mobility shift assays that would reflect dissociation of the ER dimer to ER monomers (Fig. 2 and data not shown). Thus, we provide evidence that ER remains bound to EREs as a homodimer despite the loss of [3H]TAM or [3H]4-OHT from the ER. These data clearly illustrate the distinction between ERE binding of E2-ER and TAz-ER versus 4-OHT- or TAM-ER. Despite differences in the amounts of [3H]ligand bound to ERE-bound ER, based on [3H]ligand counts in the plate assay, gel shift assays revealed similar mobilities of E2-ER, 4-OHT-ER, TAMER, and TAz-ER complexed with 2, 3, or 4 copies of EREc38 (data not shown). Consistent with earlier reports (Klinge et al., 1992a), the mobility of the 4-OHTER-, TAM-ER-, and TAz-ER -EREc38 complexes was slightly more retarded than the E2-ER-EREc38 complex. As reported, liganded ER binds one or more tandem copies of EREc38 with high affnity (Klinge et al., 1992b,c; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996a,b) (Table 3). While TAM-ER bound one or more tandem copies of EREc38 with Kd values in the nM range, the values were statistically different from those of either E2-, 4-OHT, or TAz- ER-EREc38 interaction. In each case, TAM-ER binding was of significantly lower affinity with the exception of four copies of EREc38. The reason for the low affinity for 4-OHT-ER binding to four copies of EREc38 is uncertain, but could involve steric constraints. As reported, binding of both E2-ER and TAz-ER to three tandem EREc38 was

cooperative, evidenced by nH values \ 1 (Klinge et al., 1996a). In contrast, neither 4-OHT- nor TAM- ER bound cooperatively to multiple tandem copies of EREc38. Since 4-OHT-ER and TAM-ER binding to a single copy of EREc38 was B50% of that for E2- or TAz- ER-ERE binding, steric constraints, or other interference between ERE sites, is not a logical explanation for the observed reduction in the stoichiometry of 4-OHT- or TAM- ER-EREc38 binding.

Fig. 2. Time course of liganded TAM-ER-EREc38 binding measured by EMSA. Reactions were initiated by addition of 10 fmol of [32P]EREc38 oligomer into the reaction mixture containing 122 fmol TAM-ER and poly d(I’C) (Section 2). Reactions were terminated by pipetting 40 ml of the reaction onto 4% non-denaturing polyacrylamide gel. Time indicates total incubation time in minutes. (ERERE) is the retarded liganded ER-EREc38 complex. EREc38 is the free ERE oligomer. SS indicates the ‘supershifted’ ER-EREc38 complex in the presence of anti-ER antibody H222 (H in lane 1).

86

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

3.3. Specific liganded-ER binding to sequence 6ariant EREs

Fig. 3. Comparison of the binding of [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER to a series of tandem EREs. Binding of [3H]E2-ER, [3H]4-OHT-ER, [3H]TAM-ER, or [3H]TAz-ER was measured at saturation (60–125 fmol of ER dimer added) to the indicated number of tandem (head-to-tail) copies of EREc38 (sequence in Table 1). The data shown are the mean of quadruplicate determinations9 S.E.M. and are calculated for binding to 11 fmol DNA/well (Klinge et al., 1992b, 1996a; Anolik et al., 1995; Klinge et al., 1992c).

3.2. Time course of liganded ER-EREc38 association in 6itro To test how the ER ligand affected ER-EREc38 association kinetics, we calculated the rate of EREREc38 binding using EMSA. We examined the rate of binding of ER liganded with E2, 4-OHT, or TAM to EREc38 in vitro. The ER concentrations selected were expected to achieve saturation binding, based on earlier analyses (Klinge et al., 1992b,c, 1996b). Fig. 2 shows a representative time course EMSA using TAM-ER. The ER-ERE complex contained ER protein, since addition of ER-specific monoclonal antibody H222 resulted in the appearance of a ‘supershifted’ complex (Fig. 2, lane 1). The complex remaining in lane 1 contains COUP-TF (Klinge et al., 1997b). Specificity of the ER-ERE complexes was demonstrated by the dose-dependent inhibition of ERERE complex formation upon addition of excess unlabeled EREc38, but not by an ERE half-site (data not shown). The ER-ERE interaction was quantitated and the initial rates of ER-ERE complex formation were determined (Table 5). As expected there was no effect of ER ligand on the rate of ER-EREc38 binding. Thus, differences in the association kinetics of liganded ER-ERE binding do not account for the differences detected in Kd values.

We also examined the specific binding of TAM-ER to sequence variant EREs selected to display the most varied response to E2- versus 4-OHT- ER binding (Klinge et al., 1992b; Anolik et al., 1993, 1995; Klinge et al., 1996a). We compared these binding data with those reported for E2-, 4-OHT-, and TAz- ER-ERE interaction (Anolik et al., 1996; Klinge et al., 1996a). As was the case for EREc38, TAM-ER bound each variant ERE with lower affinity compared to E2 or 4-OHT-ER (Table 4). To examine the effect of sequences flanking the ERE palindrome on TAM-ER-ERE interaction, we measured the binding of partially purified TAM-ER to EREc, a construct containing a single 13 bp ERE palindrome lacking the AT-rich flanking region (Table 1). We found that unlike the 4-fold reduction in binding affinity of E2-ER for EREc versus EREc38, TAM-ER-EREc binding was reduced only 2-fold compared to EREc38 (Table 4). This indicates that TAM-ER-ERE binding is less affected by the presence of an adjacent AT-rich region compared to E2ER binding. We also addressed the ability of TAM-ER to bind constructs containing two EREs separated (center-tocenter) by 7.2 helical turns. The rationale for this constuct is based on the distance between the first and third EREs in a construct containing three head-totail tandem copies of EREc38 to which both E2- and TAz- ER bound cooperatively (Klinge et al., 1996a; Anolik et al., 1996). TAM-ER bound the two ERE7.2 constructs with comparable affinity, regardless of the presence of the consensus AT-rich region that conferred higher binding affinity for 4-OHT-ER and that allowed E2-ER to bind cooperatively (Anolik et al., 1995). Table 5 Association rates of binding of liganded ER to EREc38 in vitro as determined by gel shift assay E2-ER

4-OHT-ER

Association 1.26 90.35 rates DF(t)/min

TAz-ER

TAM-ER

1.3 9 0.32

1.4 90.04

1.29 9 0.37

Heparin agarose-purified ER liganded with E2, 4-OHT, TAz, or TAM (0.2 – 0.8 nM) was incubated with [32P]EREc38 (20 000 dpm = 10 fmol) for 0 – 4 h at 4°C. The reaction was terminated by loading 40 ml of the binding reaction onto a 4% polyacrylamide gel and applying current as described in Section 2. The association rate values were calculated by linear regression of the binding data taken from t0−t30 min (n = 6 points, r= 0.73 – 0.97). Values are the mean9 S.E.M. for two to four separate experiments (Klinge et al., 1996a). Statistical evaluation was performed by Student’s one-tailed t-test.

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

87

Fig. 4. Trypsin digestion patterns of EREc38-bound ERa vary with ligand. A nuclear extract of recombinant human ERa baculovirus-infected SF21 cells was incubated with the indicated ER ligand and either TE buffer or a 191 bp DNA oligomer bearing 4 tandem (head-totail) copies of EREc38 Klinge et al., 1992c. The concentration of ER in the reaction was 1.1 nM and the DNA concentration was 25 nM. After a 30 min incubation, trypsin was added to give the indicated concentrations in the binding reaction (20 ml). The reaction was terminated and samples were applied to a 10% polyacrylamide mini-SDS PAGE gel. Western blotting utilized AER320 monoclonal ER antibody. Additional details of the experiment are provided in Section 2. Sizes of protein markers are indicated in kDa. This experiment was repeated twice with similar results.

3.4. TAM ligand dissociation from ER upon DNA binding We reported the DNA-dependent dissociation of [3H]4-OHT from dimeric [3H]4-OHT-ER (Klinge et al., 1996a). To examine whether the lower stoichiometry of [3H]TAM-ER-ERE binding was caused by [3H]TAM dissociation, [3H]TAM-ER, with no excess free [3H]TAM, was incubated with 10 ml of Sepharose to which 8 head-to-tail tandem copies of EREc38 were covalently liganded (Peale et al., 1989; Klinge et al., 1996a). Similar to [3H]4-OHT, 62% of the [3H]TAM counts dissociated within 30 min of incubation. In contrast, no [3H]TAz and less than 10% of the bound [3H]E2 dissociated from ER (Klinge et al., 1996a).

3.5. DNA binding and ER-ligand-media ted changes in ERa conformation are reflected in altered trypsin digestion pattern Differences in the conformation of ER in the presence of E2 versus antiestrogen have been studied using a variety of techniques. To our knowledge, only one report examined how DNA binding affects ER conformation using limited trypsin digestion (Fritsch et al., 1992). That report demonstrated that rat uterine cytosolic ER conformation was unaffected by DNA binding

(Fritsch et al., 1992). Here we utilized recombinant human ERa, expressed using a baculovirus expression system in insect cells (Obourn et al., 1993), to examine how ER ligands, E2, TAM, or 4-OHT, and binding of ERa to 4(EREc38) affected the susceptibility of the receptor to proteolysis by limited concentrations of trypsin (Fig. 4). Without trypsin treatment, immunoblotting of E2-ER with a C-terminal specific monoclonal antibody to ER, AER 320 (Abbondanza et al., 1993), showed that a 66 kDa band was the most prominent band (Fig. 4A). This corresponds in size to intact human ERa. A higher MW band of approximately 140 kDa was also visible, likely corresponding to the ERa homodimer. No difference in ERa bands was detected whether ER was incubated at 4°C or at RT prior to trypsin digestion (data not shown), indicating that RT incubation did not result in ER proteolysis. Pre-incubation of E2-ERa with 4(EREc38) did not alter the appearance of the receptor prior to trypsin treatment. Trypsin digestion of E2-ERa preincubated with 4(EREc38) resulted in a gradual disappearance of both the 140 and 66 kDa bands, followed by increased then decreased intensity of bands of approximately 60 and 55 kDa, respectively, with a concomitant appearance of 45 and 42 kDa bands.

88

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

The trypsin proteolysis pattern of TAM-ERa showed less of the ERa homodimer, even with longer exposure (Fig. 4B and data not shown). In contrast to the E2-ERa, bands of 60 or 55 kDa were not detected. However, bands of 45 and 42 kDa appeared with 0.5 mg/ml of trypsin. An additional difference is the more rapid loss of the 66 kDa intact ERa band with 1.3 mg/ml of trypsin. These differences in size and pattern of bands detected after trypsin proteolysis imply that some conformational differences occur between EREbound E2-ERa versus ERE-bound TAM-ERa. In an identical experiment, the trypsin proteolysis pattern of 4-OHT-ERa was examined in the presence of 4(EREc38) (Fig. 4C). As seen with E2- and TAM-ERa, the 66 kDa band was most prominent in the nontrypsin-treated 4-OHT-ERa. In a result similar to that seen for TAM-ERa, and different from that for E2ERa, there was little or no 60 kDa band detected. However, in contrast to the trypsin digestion of TAMERa, trypsin digestion of 4-OHT-ERa resulted in the appearance of a 55 kDa band, as seen for E2-ER.

4. Discussion One intriguing observation from our study defining the agonist versus antagonist actions of ER ligands at the level of ERE binding was the finding that one of the two ER-bound 4-OHT molecules dissociated from the ER when the ER complex bound to ERE constructs in vitro (Klinge et al., 1992b; Anolik et al., 1993, 1995, 1996; Klinge et al., 1996b). That finding led us to hypothesize that ligand dissociation may result in the observed weak agonist activity of 4-OHT despite the comparable ERE binding affinity of 4-OHT-ER and E2-ER. It is important to ascertain whether this dissociation of 4-OHT upon ER binding to EREs is a general attribute of mixed agonist/antagonist antiestrogens or whether this phenomenon is unique to 4-OHT. Here we observed that TAM-ER binds EREs with significantly lower affinity than E2-ER. We showed ER-ERE interaction impacts ligand binding with DNA binding apparently causing a loss of [3H]TAM from the [3H]TAM-ER dimer in a manner resembling that observed for [3H]4-OHT-ER-DNA interaction (Klinge et al., 1996a). Similar to the apparent lack of cooperativity of 4-OHT-ER binding to tandem repeats of EREc38, TAM-ER-EREc38 binding was non-cooperative. In contrast, E2-ER and TAz-ER bind certain tandem repeats of EREc38 cooperatively (Klinge et al., 1996a). Most natural estrogen target genes lack tandem copies of perfectly palindromic EREs like the ones used here (Anolik et al., 1995). Instead estrogen-responsive genes usually contain single or multiple imperfect EREs and multiple ERE half-sites (Klinge et al., 1996a). The exact relationships between ER binding, ERE structure,

the position of the ERE within the promoter, and transcriptional activation remain to be defined. These issues are under investigation in our laboratories. The overall similarity in structure between TAM and 4-OHT is expected to elicit a similar conformation within the LBD of the ER homodimer, allowing dissociation of at least one molecule of ligand when the receptor interacts with DNA. Because the molar amount of [3H]TAM bound per ER-ERE complex was less than that calculated for [3H]4-OHT, we believe that DNA binding promotes [3H]TAM dissociation from both ER monomers in the ERE-bound ER dimer complex. We suggest that ERE-bound TAM- or 4-OHT-liganded ER exhibit a distribution in the number of molecules bound from the expected two, i.e. one per ER monomer, to none, i.e. when both molecules have dissociated. The stoichiometry of one ER homodimer bound per ERE when ER is liganded by E2 or TAz supports the model that each ER monomer interacts with half of the palindromic ERE and agrees with the predicted stoichiometry of one ER dimer per ERE based on NMR and crystallographic studies of the ER DNA binding domain (Schwabe et al., 1990, 1993). We interpret the results of the trypsin proteolysis experiments to suggest that, when ERa is bound to 4(EREc38) and liganded by either E2, TAM, or 4-OHT, that the overall ER conformation is similar, but not identical. This result reflects the differences in mobility of these complexes in EMSA. It is possible that TAM and 4-OHT ligand dissociation from the ERE-bound ER is at least partially responsible for the altered trypsin sensitivity. Recent studies of the X-ray crystal structure of the LBD of ERa liganded by E2 and by the ‘selective ER antagonist’ (SERM) raloxifene (RAL) demonstrated that E2 and RAL form specific hydrogen bonds within the ligand binding cavity resulting in specific conformational differences in the surface shape of the LBD (Brzozowski et al., 1997). Because TAM lacks the 4%-OH on the A ring that is present in 4-OHT, we hypothesize that TAM and 4-OHT exhibit different interactions with a helices H3 and H6 which are the sites of the interaction of the 4%OH phenol in the A ring of both E2 and RAL (Brzozowski et al., 1997). Amino acid 351, aspartate, was recently reported to be critical for the antiestrogenic activity of RAL (Levenson and Jordan, 1998). The current model of how ligand binding impacts ER activity predicts that the nature of the transcriptional response of an estrogen-responsive gene depends on coordination of diverse signalling pathways. According to this model, ligand-specific conformational differences between E2-liganded ER versus type I antiestrogen-liganded ER, e.g. 4-OHT, and RAL, act as ‘switches’ for the recruitment of either co-activator or co-repressor complexes to the ER. When ERE-bound ER is liganded by E2, co-activators, e.g. SRC-1, TIF-1, and

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

CBP/p300, are recruited to form a ternary complex with P/CAF and other proteins (Goldman et al., 1997; Shibata et al., 1997; Scolnick et al., 1997; Torchia et al., 1997; The´not et al., 1997; Kurokawa et al., 1998). This co-activator complex has histone acetylase activity which creates a more ‘open’ nucleosomal structure for activation of transcription (Sternglanz, 1996; Pennisi, 1997). Interestingly, two recent reports indicate that 4-OHT-ER interacts specifically with NCoR (Lavinsky et al., 1998; Zhang et al., 1998). NCoR is a component of a co-repressor complex that includes m Sin 3A which associates with histone deacetylase 1 (HDAc1) (Zhang et al., 1998). Histone deacetylation represses gene transcription (Wolffe, 1996; Heinzel et al., 1997; Nagy et al., 1997). The precise mechanisms by which ligand-specific conformational differences, possibly facitiliated in part by ligand dissociation, in ER impact upon ER interaction with co-activator and co-repressor complexes remain to be explored.

Acknowledgements We thank Dr Nicholas J. Koszewski of the University of Kentucky for the gift of the ERa baculovirus plasmid and Abbott Labs for the gift of H222. This work was supported by a grant from the Cancer Research Foundation of America, N.I.E.H.S. lP20 ES06832-12, and a University of Louisville Research Initiation Grant to C.M.K. and in part by NIH grant HD24459 to R.H.

References Abbondanza, C., deFalco, A., Nigro, V., Medici, N., Armetta, I., Molinari, A.M., Moncharmont, B., Puca, G.A., 1993. Characterization and epitope mapping of a new panel of monoclonal antibodies to estradiol receptor. Steroids 58, 4–12. Anolik, J.H., Klinge, C.M., Bambara, R.A., Hilf, R., 1993. Differential impact of flanking sequences on estradiol- versus 4-hydroxytamoxifen-liganded estrogen receptor binding to estrogen responsive element DNA. J. Steroid Biochem. Molec. Biol. 46, 713–730. Anolik, J.H., Klinge, C.M., Hilf, R., Bambara, R.A., 1995. Cooperative binding of estrogen receptor to DNA depends on spacing of binding sites, flanking sequence, and ligand. Biochemistry 34, 2511 – 2520. Anolik, J.H., Klinge, C.M., Brolly, C.L., Bambara, R.A., Hilf, R., 1996. Stability of the ligand of estrogen response element-bound estrogen receptor depends on flanking sequences and cellular factors. J. Steroid Biochem. Molec. Biol. 59, 413–429. Berry, M., Nunez, A.M., Chambon, P., 1989. Estrogen responsive element of the human pS2 gene is an imperfectly palindromic sequence. Proc. Natl. Acad. Sci. USA 86, 1218–1222. Brzozowski, A.M., Pike, A.C., Dauter, Z., Hubbard, R.E., Bonn, T., Engstrom, O., Ohma, L., Greene, G.L., Gustafsson, J.A., Carlquist, M., 1997. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389, 753–758. Fritsch, M., Welch, R.D., Murdoch, F.E., Anderson, I., Gorski, J., 1992. DNA allosterically modulates the steroid binding domain of

89

the estrogen receptor. J. Biol. Chem. 267, 1823 – 1828. Goldman, P.S., Tran, V.K., Goodman, R.H., 1997. The multifunctional role of the co-activator CBP in transcriptional regulation. Rec. Prog. Hormone Res. 52, 103 – 119. Heinzel, T., Lavinsky, R.M., Mullen, T.-M., Soderstrom, M., Laherty, C.D., Torchia, J., Yang, W.-M.G.B., Ngo, S.D., Davie, J.R., Seto, E., Eisenman, R.N., Rose, D.W., Glass, C.K., Rosenfeld, M.G., 1997. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 387, 43 – 48. Hoopes, B.C., LeBlanc, J.F., Hawley, D.K., 1992. Kinetic analysis of yeast TFIID TATA box complex formation suggests a multi-step pathway. J. Biol. Chem. 267, 11539 – 11547. Horscroft, N.J., Roy, P., 1997. Thermal denaturation of proteins for SDS-PAGE analysis by microwave irradiation. Biotechniques 22, 224 – 226. Horwitz, K., Jackson, T., Bain, D., Richer, J., Takimoto, G., Tung, L., 1996. Nuclear receptor coactivators and corepressors. Mol. Endocrinol. 10, 1167 – 1177. Katzenellenbogen, B.S., Korach, K.S., 1997. Editorial: A new actor in the estrogen receptor drama-enter ERb. Endocrinology 138, 861 – 862. Katzenellenbogen, B.S., Norman, M.J., Eckert, R.L., Peltz, S.W., Mange, W.F., 1984. Bioactivities, estrogen receptor interactions, and plasminogen activator-inducing activities of tamoxifen and hydroxytamoxifen isomers in MCF-7 human breast cancer cells. Cancer Res. 44, 112 – 119. Kladde, M.P., Xu, M., Simpson, R.T., 1996. Direct study of DNA– protein interactions in repressed and active chromatin in living cells. EMBO J. 15, 6290 – 6300. Klein-Hitpass, L., Schorpp, M., Wagner, U., Ryffel, G.U., 1986. An estrogen responsive element derived from the 5’ flanking region of the Xenopus vitellogenin A2 gene functions in transfected human cells. Cell 46, 1053 – 1061. Klinge, C.M., Bambara, R.A., Hilf, R., 1992a. What differentiates antiestrogen liganded versus estradiol-liganded estrogen receptor action? Oncology Res. 4, 1073 – 1081. Klinge, C.M., Bambara, R.A., Hilf, R., 1992b. Antiestrogen-liganded estrogen receptor interaction with estrogen responsive element DNA in vitro. J. Steroid Biochem. Molec. Biol. 43, 249–262. Klinge, C.M., Peale, F.V. Jr., Hilf, R., Bambara, R.A., Zain, S., 1992c. Cooperative estrogen receptor interaction with consensus or variant estrogen responsive elements in vitro. Cancer Res. 52, 1073 – 1081. Klinge, C.M., Traish, A.M., Bambara, R.A., Hilf, R., 1996a. Dissociation of 4 hydroxytamoxifen, but not estradiol or tamoxifen aziridine, from the estrogen receptor when the receptor binds estrogen response element DNA. J. Steroid Biochem. Molec. Biol. 57, 51 – 66. Klinge, C.M., Traish, A.M., Driscoll, M.D., Hilf, R., Bambara, R.A., 1996b. Site directed estrogen receptor antibodies stabilize 4-hydroxytamoxifen ligand, but not estradiol, and indicate ligand-specific differences in the recognition of estrogen response element DNA in vitro. Steroids 61, 278 – 289. Klinge, C.M., Brolly, C.L., Bambara, R.A., Hilf, R., 1997a. Hsp70 is not required for high affinity binding of purified calf uterine estrogen receptor to estrogen response element DNA in vitro. J. Steroid Biochem. Molec. Biol. 63, 283 – 301. Klinge, C.M., Silver, B.F., Driscoll, M.D., Sathya, G., Bambara, R.A., Hilf, R., 1997b. COUP-TF interacts with estrogen receptor, binds to estrogen response elements and half sites, and modulates estrogen-induced gene expression. J. Biol. Chem. 272, 31465– 31474. Kurokawa, R., Kalatus, D., Ogliastro, M.H., Kioussi, C., Xu, L., Torchia, J., Rosenfeld, M.G., Glass, C.K., 1998. Differential use of CREB binding protein-coactivator complexes. Science 279, 700 – 703.

90

C.M. Klinge et al. / Molecular and Cellular Endocrinology 143 (1998) 79–90

Lavinsky, R., Jepsen, K., Heinzel, T., Torchia, J., Mullen, T.M., Schiff, R., Del-Rio, A.L., Ricote, M.N.S., Gemsch, J., Hilsenbeck, S.G., Osborne, C.K., Glass, C.K., Rosenfeld, M.G., Rose, D.W., 1998. Diverse signaling pathways modulate nuclear receptor recruitment of N-CoR and SMRT complexes. Proc. Natl. Acad. Sci. USA 95, 2920–2925. Levenson, A.S., Jordan, V.C., 1998. The key to the antiestrogenic mechanism of raloxifene is amino acid 351 (aspartate) in the estrogen receptor. Cancer Res. 58, 1872–1875. Ludwig, L.B., Peale, F.V., Klinge, C.M., Bambara, R.A., Zain, S., Hilf, R., 1990. A microtiter well assay for quantitative measurement of estrogen receptor binding to estrogen-responsive elements. Mol. Endocrinol. 4, 1027–1033. Mangelsdorf, D.J., Thummel, C., Beato, M., Herrlich, P., Schutz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., Chambon, P., Evans, R.M., 1995. The nuclear receptor superfamily: the second decade. Cell 83, 835–839. Nagy, L., Kao, H.-Y., Chakravarti, D., Lin, R.J., Hassig, C.A., Ayer, D.E., Schreiber, S.L., Evans, R.M., 1997. Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and histone deacetylase. Cell 89, 373–380. Obourn, J.D., Koszewski, N.J., Notides, A.C., 1993. Hormone- and DNA-binding mechanisms of the recombinant human estrogen receptor. Biochemistry 32, 6229–6236. Pavlik, E.J., Coulson, P.B., 1976. Hydroxylapatite ‘batch’ assay for estrogen receptor: Increased sensitivity over present receptor assays. J. Steroid Biochem. 7, 357–368. Peale, F.V., Ludwig, L.B., Zain, S., Hilf, R., Bambara, R.A., 1988. Properties of a high-affinity DNA binding site for estrogen receptor. Proc. Natl. Acad. Sci. USA 85, 1038–1042. Peale, F.V. Jr., Ishibe, Y., Klinge, C.M., Zain, S., Hilf, R., Bambara, R.A., 1989. Rapid purification of the estrogen receptor by sequence-specific DNA affinity chromatography. Biochemistry 28, 8671 – 8675. Pennisi, E., 1997. Opening the way to gene activity. Science 275, 155 – 157.

.

Richardson, C.D., 1995. In: Methods in Molecular Biology, Vol. 29, Humana Press, Towata. Schwabe, J.W.R., Neuhaus, D., Rhodes, D., 1990. Solution structure of the DNA binding domain of the oestrogen receptor. Nature 348, 458 – 461. Schwabe, J.W.R., Chapman, L., Finch, J.T., Rhodes, D., 1993. The crystal structure of the estrogen receptor DNA-binding domain bound to DNA: How receptors discriminate between their response elements. Cell 75, 567 – 578. Scolnick, D.M.N.H.C., Stavridi, E.S., Lien, M.S., Caruso, L., Moran, E., Berger, S.L., Halazonetis, T.D., 1997. CREB-binding protein and p300/CBP-assocaited factor are transcriptional coactivators of the p53 tumor suppressor protein. Cancer Res. 57, 3693–3696. Shibata, H., Spencer, T.E., Onate, S.A., Jenster, G., Tsai, S.Y., Tsai, M.J., O’Malley, B.W., 1997. Role of co-activators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog. Horm. Res. 52, 141 – 164. Sternglanz, R., 1996. Histone acetylation: a gateway to transcriptional activation. Trends Biochem. Sci. 21, 357 – 358. The´not, S., Henriquet, C., Rochefort, H., Cavailles, V., 1997. Differential interaction of nuclear receptors with the putative human transcriptional coactivator hTIF1. J. Biol. Chem. 272, 12062– 12068. Torchia, J., Rose, D.W., Isnostroza, J., Kamei, Y., Westin, S., Glass, C.K., Rosenfeld, M.G., 1997. The transcriptional co-activator p/CIP binds CBP and mediates nuclear receptor function. Nature 387, 677 – 684. Vaughn, J.L., Goodwin, R.H., Tompkins, G.J., McCawley, P.T., 1977. The establishment of two cell lines from the insect Spodoptera frugiperda ( Lepidoptera: Noctuidae). In Vitro 13, 213 – 217. Wolffe, A.P., 1996. Histone deacetylase: a regulator of transcription. Science 272, 371 – 372. Zhang, X., Jeyakumar, M., Petukhov, S., Bagchi, M.K., 1998. A nuclear receptor corepressor modulates transcriptional activity of antagonist-occupied steroid hormone receptor. Mol. Endocrinol. 12, 513 – 524.