Proc. Natd. Acad. Sci. USA

Vol. 91, pp. 11502-11506, November 1994 Biochemistry

A low-affinity estrogen-binding site in pregnant rat uteri: Analysis and partial purification WESLEY G. N. GRAY*t*, ESTHER E. BISWAS*, NASIR BASHIRELAHIt, AND SUBHASIs B. BISWAS*§ *Department of Pediatrics, School of Medicine, and tDepartment of Biochemistry, Dental School, University of Maryland, Baltimore, MD 21201

Communicated by Jack Gorski, July 25, 1994 (received for review January 19, 1994)

We have identified a low-affinity (type U) estrogen-binding site (EBS) that is expressed at high levels

ABSTRACT

also been shown that the type II EBS appears present in certain breast tumor tissues (6, 7) and in a wide variety of lymphoblastoid and endometrial cancers (22-26). Studies with type II EBS in extracts of estradiol-stimulated chicken oviduct and rat uteri demonstrated that estradiol binding to the low-affinity EBS is apparently cooperative in nature, as it displays a sigmoidal binding curve (1, 2, 8). In this report we have demonstrated that the levels of a low-affinity EBS increase enormously during pregnancy in rat uteri, making pregnant rat uteri an abundant source of type II EBS. We also describe methodology for the purification of type II EBS from pregnant rat uteri to near homogeneity.

durin pregnancy in rat uteri. Although this activity was dtectble I nonpegnt rat uteri, it was present in mounts (0.094 pmol/g of uteri) that were severalfold lower than the -affnity type I estrogen receptor (0.57 pmol/g of uteri). During prgnancy at 19-20 days of gestation, the low-affinity type II EBS became the mior (288%) esogen-binding ste in rat uteri. The increase in the level of low-affinity EBS (7.9 pmol/g) in uteri was 45-fold with an w20-fold increase in the specific activity (0.39 pmol/mg) of this form, whereas the hi h-anity form emaed relatively un d. We report here a method of p tin of type U EBS from pregnant rat uteri and present an analysis of Its DNA and steroid-binding properties. Estradlol-binding studies and Scatchard analysis showed that the type U EBS had an apparent estradlol-binding affinity Of >24 nM. Gel trion and SDS/PAGE analysis indicated that the type U EBS was a monomeric 73-kDa l bnding remained apparently uninhibprotein. The ited in the praesence of a large excess of tamoxifen, nafoidine, or dihydrnEstradl, diethystilberol, and quercitin (a type HI EBS-specfi inhibitor) competed efficiently. The purified low-affinity EBS did not have sequence-specfic DNAbndin activit with the etron-responsive element, which indicated that it differs in funion from the type I estrogen

MATERIALS AND METHODS Materials. Pregnant and nonpregnant rats were obtained from Charles River Breeding Laboratories. Pregnant rats were sacrificed on the 19-20th day of gestation. The uteri were removed immediately after sacrifice, cleaned, frozen in liquid nitrogen, and stored at -800C until use. Rat uteri were gifts from I. Gewolb and J. Torday of the Department of Pediatrics. Steroid Binding and Scatchard Analysis. Single-point steroid-binding measurement was used to monitor the purification of type H EBS with the hydroxyapatite method (27). A standard 100-A4 binding mixture contained 25 nM [3H]estradiol, 10 mM Tris'HCl (pH 7.4), 1.5 mM EDTA, 2% glycerol, and samples as indicated. The binding reactions were allowed to incubate for 15 min at room temperature followed by chilling in an ice/water slurry for 5 min. The remainder of the assay followed the standard hydroxyapatite method, as described (27). Saturation binding and competition analyses were also done by using a modification of this assay. Preparation of Type I and Type U EBS Extrc. Preparation of cytosol from rat uteri and (NH4)2SO4 fractionation were done essentially as described (8). Mobility-Shift DNA-Binding Analysis. Mobility-shift DNAbinding analyses were done with a synthetic 18-mer ERE element corresponding to the sequence 5'-TCACjGICACAG.WA(CCTG-3'. The underlined consensus sequence, containing two inverted repeats, appears essential for estrogen inducibility (17). The assay was done as described (14, 28, 29).

receptor.

Estrogen receptors have been studied from a wide variety of tissue sources and cancer cell lines. Investigations in several laboratories during the last few years have clearly demonstrated the existence of two subtypes of estrogen receptors (1-8). The two classes of estrogen receptors are differentiated with respect to their affinities toward estradiol binding: the high-affinity type I with a Kd of 0.1-1 nM and the low-affinity type II with a Kd .:20 nM (3-8) and generally termed as type II estrogen-binding site(s) (EBS). The structure and function of the high-affinity type I estrogen receptor has been the subject of a wide variety of extensive studies (9-11). In addition to its binding to estradiol, it binds to estrogen-responsive elements (EREs) and initiates transcription of estrogen-responsive genes in target tissues; functioning as a transcription initiation factor (12-17). The type I estrogen receptor gene has been cloned, sequenced, and expressed in prokaryotic and eukaryotic systems (18-21). However, the structure and function of type II EBS remain unknown. The existence of an EBS with a low affinity for estradiol was first demonstrated during exogenous estradiol stimulation studies in rats. The uteri of the estradiol-treated rats appeared to contain the type II form of EBS (1, 2, 8). Analogous results were obtained with chicken oviduct under similar conditions. A majority of the studies to date have been done by using these sources of type II EBS. However, it has

RESULTS Low- and High-Affnity EBS in Pregnant Rat Uteri. Estradiol-binding assays of cytosol and (NH42SQ4 fractions, done as described by Densmore et al. (8), indicated that the total cytosolic estradiol-binding activity increased significantly in Abbreviations: EBS, estrogen-binding site; ERE, estrogen-responsive element; DES, diethylstilbestrol. tPresent address: Department of Biochemistry, University of Wisconsin School of Medicine, Madison, WI 53706. §To whom reprint requests should be addressed.

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

11502

Biochemistry: al.al. Gray ~~~Proc. Nati. Acad. Sci. etet Gray Biochemistry: Table 1. Distribution of type I and type II EBS Total Total Specific Uterus protein, activity,* activity,* Fraction units ± SEM unit/mg sample mg 190 7.5 ± 0.50 0.04 Nonpregnant I Cytosol II 40% 55 10.7 ± 0.06 0.20 (18.7 g) 1.76 ± 0.05 III 70%/ 88 0.02 I Cytosol 0.12 825 99.55 ± 0.05 Pregnant II 40%/ 214 20.07 ± 0.09 0.09 (19 g) III 70% 384 150 ± 0.06 0.39 = Averages of three (n 3) separate experiments are given. *One unit ofestradiol binding is defined as 1 pmol of estradiol binding under our assay conditions.

USA 91 (1994)

11503

200 B

15-6 E

B5

0-

t-5;

C.) 15

0

m

H1

0

01

.70ci'

Cy.,tosol 40%:-

E' CL

pregnant rat uteri compared with that in an equal weight of nonpregnant rat uteri (Table 1 and Fig. 1). The estradiolbinding activities in both 40%o and 70% (NH4)2S04 fractions were stimulated. Our results indicated that most of the estradiol-binding activity (88%) precipitated at 70% (NH4)2S04 saturation, whereas only 12% of the activity precipitated at 40%o (NH4)2S04 saturation. Comparison of these results with that of nonpregnant rat uteri (Table 1) indicated that the estrogen-binding activity that precipitated at 70% (NH4)2S04 saturation significantly increased with pregnancy. Total activity of the 70%6 (NH4)2504 fraction of cytosol increased 85-fold with a 20-fold increase in the specific activity. Total activity of the 40% (NH4)2S04 fraction, on the other hand, increased only 2-fold in pregnant rats, whereas the specific activity decreased by 50%6. We have done saturation binding and Scatchard analysis (30) of both of these EBS fractions (Fig. 1 C-F). Scatchard analysis presented in Fig. 1 indicated that the Kd of unpurified type II EBS was =40 nM, whereas, the Kd of type I estrogen receptor was -t0.3 nM. Although, a high-affinity component was present in the 40% (NH4)2S04 fraction of pregnant rat uteri, most of the activity in the 40%1, and almost all of the activity in the 70% (NH4)2S04 fraction, contained lowaffinity EBS (Fig. 1 D and F). Consequently, it appears that the level of low-affinity EBS was significantly stimulated in pregnancy and that the 40% (NH4)2S04 fr-action also contained a significant amount of low-affinity EBS. Separation and Analysis of Hg- and Low-Affinity EBS from Nonpregnant Uteri. When cytosol of nonpregnant rat uteri was fractionated, ==86%6 of the activity precipitated at 40%6 (NH4)2S04 (fraction IIA) saturation, and --14%o of the activity precipitated at 70%6 AS (fr-action IIB) saturation (Table 1). The 40% (NH4)2S04 fraction of estrogen receptor showed the binding characteristics of type I estrogen receptor (data not shown). Nevertheless, the 70%6 fraction appeared to contain type II EBS, although at very low levels (Table 1). Purification of Type H EBS. All purification procedures were done at 40C unless otherwise indicated and in the presence of the following protease inhibitors: pepstatin A and leupeptin each at 1.0 pg/ml chymostatin and antipain each at 0.1 p~g/mi, 1 miM each of (tosyl)phenylalanine chloromethyl ketone (TPCK) and (tosyl)lysine chloromethyl ketone (TLCK), and soybean trypsin inhibitor at 5 ug/ml. In a standard procedure, 100 g of rat uteri from pregnant rats was used to prepare fraction IIB containing low-affinity type HI EBS as described in Materials and Methods. Fraction IIB was dissolved in buffer A (50 mM Tris-HCl, pH 7.5/10% glycerol/1.5 mM EDTA) at 1/10th the original volume of cytosol and then dialyzed against buffer A/20%6 sucrose for %-2.5 hr. The dialysate was diluted to the conductivity of buffer A/25 mM NaCl using 10 mM Tris-HCl, pH 7.5/1.5 mM EDTA/101% glycerol. The dialysate was then loaded on a tandem column system composed of S-Sepharose (5.0 cm x 10.0 cm) -.-) CM-Affl-Gel Blue (3.0 cm x 15.0 cm) --) Q-Sepharose (2.5 cm x 25 cm) in a series that was equili-

01

.0

aI HE 0

a)

C,)

v0

0 2030 4050

.02 .04 .06

.08

Bound, pmol/mg

Estradiol, nM

F E

0.03

1Z0.02

.0 0

0

.a

0.0)i

0.

;0

CO

0 0

20

40

60

80

100

-----------------------

.0

Estradiol, nM

FIG.

1.

Relative

-,

.2

.6

.4

.8

Bound, pmol/mg

occurrence

of type

Iland type

EBS in pregnant

Estradiol-binding analyses of cytosol and its 40% (HA) and 70%/ (IIB) (NH4h2SO4 fr-actions, prepared from nonpregnant rat uteri (A) and pregnant rat uteri (B). Each value represents the mean ±t SEM calculated from three identical experand nonpregnant rat uteri.

iments. (C-F) Saturation binding and Scatchard analyses of estrogen binding from pregnant rat uteri. (C and E) Saturation binding analysis of EBS HA (C) and UB (E); (D and F) Scatchard plots for EBS HA (D) and IIB (F). Each point was the mean ± SEM for three identical analyses. B/F, bound/free.

brated with buffer

columns

were

A/25

mM NaCl. After

Q-Sepharose column

The

loading, the tandem A/25 m.M NaCl.

washed with 200 ml of buffer was

50 column volumes of buffer

disconnected and washed with

A/70

m.M NaCl until

detectable in the eluant. The column

was

was

no

protein

then eluted

a 600-mi gradient of buffer A/70 m.M NaCl --+ buffer A/300 mM NaCi. The fractions were assayed for protein (3 1),1 estradiol, and DNA-binding activities. The DNA-binding activity was seen to elute at a higher ionic strength (0.24 M NaCl) than the estradiol-binding activity (0. 12 M NaCl) (Fig. 2). The peak estradiol-binding fr-actions were pooled (fraction III) and adjusted to 10 mM KPO4, pH 7.4. The tandem chromatography resulted in --300-fold purification of the estradiol-binding activity compared with fr-action I (Table 2). Saturation analysis and Scatchard plot showed that fr-action

with

III contained

a

low-affitnity estrogen-binding protein (data not

shown). Fraction

III1 was loaded onto a hydroxyapatite column (1.5

equilibrated with buffer B (10 mM KPO4, pH 7.4/10%o glycerol/1.5 mM EDTA) and eluted with a gradient of buffer B --+ buffer B/100 m.M KPO4. The active estradiolbinding fractions were pooled (fraction IV) and concentrated to 10 mg/mi by ultrafiltration using an Amicon YM30 memcm

x

brane.

2.5 cm)

11504

Biochemistry: Gmy et al.

Proc. Nati. Acad. Sci. USA 91 (1994)

A

.0'

0

0 ~0

1-

0

.i

0

c

4:

z a

Cl

Fraction number B

FRACTION NUMBER

28 30 32 34 36 31 40 42 44 4 64

ER-ERE COMPLEX

50 52

So 58 60 6264

66 61 70

72 74

76

78

0

..e

FIG. 2. Fractionation of the estradiol and DNA-binding activities by Q-Sepharose chromatography. Details of the fractionation are presented in Results. (A) Estradiol binding analysis using 10-I4 aliquots of the indicated fractions. (B) Autoradiogram of the mobility-shift ERE-binding assay of the fractions. It should be noted that only the areas corresponding to detectable protein-DNA complex formation are shown.

The EBS was further purified by size-exclusion HPLC on a TSK-250 column (21.5 mm x 60 cm) with buffer A/150 mM NaCl. The active estrogen-binding fractions were pooled (fraction V) and concentrated; size-exclusion HPLC resulted in an increase in the specific activity from 1822 pmol/mg to 5625 pmol/mg (Table 2). Fraction V was refractionated by size-exclusion HPLC using a Pharmacia Superdex 200 HR column (10 mm x 30 cm) using the same buffer. The steroid-binding activity in the final size-exclusion HPLC eluted with the major protein peak (Fig. 3A). The native molecular mass, as determined from the eluted volume of size markers, was 73 kDa. Analysis across the fractions on a 5 -* 15% SDS/PAGE gel showed a doublet band between 68 and 73 kDa (Fig. 3B). Densitometric analysis of the two bands indicated that the upper 73-kDa protein band more closely matched with the estradiol-binding activity than did the lower 68-kDa band. The peak estrogen-binding fractions were pooled (fraction VI). The type II EBS was very sensitive to repeated freeze-thawing, and excessive dilution durTable 2. Purification table Fraction I Cytosol

II (NH42SO4 III Tandem columns IV Hydroxyapatite V SE-HPLC

Specific PurificaProtein, Activity,* activity,* tion, Yield, units % mg units/mg -fold 0.8 7000 9000 1.0 100 2.24 3200 7200 2.8 80 23

5301

230

288

59

1.7 34 3062 1822 2278 0.32 1800 5625 7032 20 SE-HPLC, size-exclusion HPLC. *One unit of estradiol binding is defined as 1 pmol of estradiol binding under our assay conditions.

ing purification was avoided. In the absence of protease inhibitors the type II EBS polypeptide underwent protease degradation during purification (W.G.N.G. and S.B., unpublished observation). Table 2 summarizes the results from a typical purification. Densmore et al. (8) have described the isolation of a type II EBS from estrogen-implanted chicken oviduct. Their estimation of the size ofthe type II EBS was 40 kDa. However, the induction and purification of type H EBS were very different from the procedure described there--e.g.-lack of use of protease inhibitors, different chromatographic steps, etc. Consequently, at the present time it is premature to compare these two reports. Etradiol-Binding C cteri s of Type I EBS. Saturation analysis was done by using fraction VI type II EBS. The binding was saturated at 60 nM estradiol and did not display cooperative binding (Fig. 4A). Scatchard plot analysis indicated one binding site with a Kd of 24 nM and a B. of 340 pmol/mg of protein. (Fig. 4B). Analysis of Steroid Sp yof Type U EBS. The steroid specificity of type II EBS was determined by using 25 nM estradiol (Fig. 4C). Competition analysis showed that the purified type II EBS could be challenged effectively with estradiol, quercetin, and diethylstilbestrol (DES) with K, values of 12, 30, and 23.3 nM, respectively. Nafoxidine, tamoxifen, and dihydrotestosterone did not appear to compete in our assays (data not shown).

DISCUSSION The low-affinity (type II) EBS has been shown to be induced

in uteri by prolonged exogenous estradiol administration in chickens and rats (1-3). We have examined uterine tissues for the possible induction of type II EBS by elevated levels of endogenous estradiol during pregnancy, a physiological con-

Biochemistry: Gray et al.

Proc. NatL. Acad. Sci. USA 91 (1994)

A

tors in the cytosol and (NH4)2SO4 fractions indicated that type II EBS [in 40-70% (NH4)2S04 fraction] are stimulated -85-fold in pregnant rat uteri, whereas, the type I estrogen receptor [in the 0-40% (NH4)2S04 fraction] is stimulated only

0

T-

x

20

EQ. 0

0

4=

VI

10

e0 0)

0.

CiO

- - - - --

0'

B

--JLv---- -

E

40 50 Fraction number

60

FRACTION NUMBER

48

49 50

51

52 53

54

55

58

57

.0-

TYPE 11 ER

11505

_

FIG. 3. Fractionation of type II EBS by size exclusion HPLC. Type II EBS (fraction V) was chromatographed on a Pharmacia-LKB Superdex 200HR gel filtration column as described. (A) Aliquots of the fractions were assayed for estradiol binding and protein. Molecular weight markers were IgG, bovine serum albumin, and soybean trypsin inhibitor. (B) Silver-stained

5

--

15%

SDS/PAGE analysis

of

the fractions in A. ER, estrogen receptor.

dition which in that respect is somewhat analogous to in vivo exogenous estradiol stimulation. The use of pregnant rat uteri as a source of type II EBS may provide an alternative to exogenous estrogen administration.

We have compared uteri from nonpregnant rats and rats at 19-20 days of gestation. The distribution of estrogen recep-

2-fold (Table 1). Moreover, Scatchard analysis indicated that the estrogen receptor in 40%o (NH4)2S04 fraction of pregnant rat uteri cytosol contained a substantial fraction of type H EBS because of its large proportion in the cytosol (Fig. 1). Scatchard analysis presented in Fig. 1 indicated that the Kd of unpurified type II EBS was -40 nM, whereas, the Kd of type I estrogen receptor was =0.3 nM. This large induction of type II EBS could not be explained simply by the fact that during pregnancy there is an accompanying increase in uterine size and protein synthesis. The specific activity of type II EBS increased -20-fold, whereas the specific activity of the type I estrogen receptor actually decreased. Thus, the low-affinity type II EBS was the predominant EBS induced during pregnancy. The levels of type II EBS were quite high in pregnant rat uteri, and the pregnant uterus may be one of the richest sources of this form; consequently we used it for purification of type II EBS. The purification procedure consisted of a series ofdifferent chromatographic steps to selectively separate type II EBS from type I estrogen receptor, as well as other uterine proteins. The preparation was devoid of any tyrosinase activity [as reported by Garai et al. (32)], which was removed by the S-Sepharose column in the tandem column chromatography of fraction II (S. Pomerantz, personal communication). Type II EBS was unique in that it did not bind to cation-exchange columns such as S-Sepharose, CMSepharose, Affi-Gel Blue, as well as phosphocellulose and DNA-cellulose columns. However, type II EBS binds to anion-exchange (Q-Sepharose and DEAE-cellulose) column matrices. The differential binding of type II EBS suggested that this protein had a strong negatively charged surface. Marsigliante et aL (6) have shown that the type II EBS in breast tumors have a pI of 7.0, significantly higher than the type I estrogen receptor. The

purified type

EBS

100

IB a

.

0

10F

x E

a

low

affinity (Kd

24

cIN

75 0) C

:5

U. co

0

displayed

nM) for estradiol. The estradiol binding was inhibited by sulfhydryl group reagents, which was consistent with earlier reports (2, 8) characterizing crude type II EBS (data not shown). However, the purified type II EBS did not show the sigmoidal-like saturation binding behavior seen in crude preparations. The binding was saturated at 60 nM estradiol and did not display cooperative binding (Fig. 4A). It is interesting to note that Williams and Gorski (33) explored the equilibrium estradiol binding to isolated uterine cell suspension and whole uteri in vitro and detected no cooperative binding. Perhaps the cooperativity seen with crude uterine

50

C

a)

0

5

Oh

U1)

a.

0

25

0.

Ci)

0

;

50 Estradiol, nM

20 60 100 140 180 220 260 300

Bound, pmol/mg

i In

"Mid -

..

. ~-A..I.^

10-9 10-C 10-7 10-6 10-5 10-4 Competitor, M

FIG. 4. Analysis of the estradiol binding of type II EBS (fraction VI). (A) Saturation binding analysis was done as described. (B) Scatchard plot of data in A. B/F, bound/free. (C) Competition analysis was done as described with the following competitors: estradiol (e), quercitin (n), and DES (i). Each point is the mean SEM for three separate identical experiments.

11506

Biochemistry: Gray et al.

extract (Fig. 1) was an artifact of the protein-extraction procedure or due to the presence of another protein factor that is lost during purification. Purified type II EBS did not bind the synthetic ERE sequence, and the ERE binding seen in the crude type II EBS fractions was clearly separated in the Q-Sepharose column of the tandem column chromatography (Fig. 2). Lack of ERE binding is an additional distinguishing feature between type I and type II EBSs. The type II estrogen binding remained apparently unaffected by tamoxifen, nafoxidine, and dihydrotestosterone, all of which are potent inhibitors of type I estrogen receptor. Estradiol, quercitin, and DES inhibited the type II EBS comparably. The methodology presented here allowed us to purify the type II EBS to near homogeneity. Two major protein bands were present in the final preparation (Fig. 3B), one of which, the 73-kDa polypeptide, clearly comigrated with the estradiol-binding activity of type II EBS, and the other 67-kDa protein was likely the rat serum albumin or a similar protein. Rat -fetoprotein is of similar size and known to bind estrogen. Therefore, we have carefully evaluated the possibility of its identity with type II EBS. Type II EBS could be distinguished from a-fetoprotein by the fact that the later has a higher Ki for DES (- 200 nM) compared with the K, of 23.3 nM reported here for type II EBS (34). Cloned and highly purified rat a-fetoprotein binds estradiol with a Kd of '10.8 nM (35), which is significantly lower than that seen with type II EBS (w24 nM). In addition, a-fetoprotein is produced in fetal liver, and in some cases it is transmitted to maternal blood in the instance of certain birth defects. a-Fetoprotein is not known to be produced in the maternal uteri or in other organs. As we extensively wash the uteri before freezing and extraction, it is very unlikely that there is any appreciable quantity of a-fetoprotein in the uteri cytosol. a-Fetoprotein and albumin are members of the same gene family (36). At this time, however, we cannot rule out the possibility that the purified type II type II EBS is a product of this gene family. We thank Dr. Seymour Pomerantz of this University for helpful advice and critical review of the manuscript. This work was supported in part by the Breast Cancer Research Program of the U.S. Army (Grant DAMD17-94-J4326). 1. Eriksson, H., Upchurch, S., Hardin, J. W., Peck, E. J. & Clark, J. H. (1978) Biochem. Biophys. Res. Commun. 81, 1-7. 2. Markaverich, B. M., Roberts, R. R., Finney, R. W. & Clark, J. H. (1983) J. Biol. Chem. 258, 11663-11671. 3. Chae, K., Johnson, H. S. & Korach, S. K. (1991) J. Steroid Biochem. 36, 35-42. 4. Swanek, G. E., Alvarez, J. M. & Sufrin, G. (1982) Biochem. Biophys. Res. Commun. 106, 1441-1447. 5. Yu, M., Cates, J., Leav, I. & Ho, S.-M. (1989) J. Steroid Biochem. 33, 449-457. 6. Marsigliante, S., Biscozzo, L., Puddefoot, J. R., Vinson, G. P. & Steorelli, C. (1992) J. Steroid Biochem. Mol. Biol. 42, 777-781.

Proc. NatL. Acad. Sci. USA 91 (1994) 7. Marsigliante, S., Puddefoot, Barker, S., Gledhill, J. & Vinson, J. R. (1992) J. Steroid Biochem. Mol. Biol. 37, 643-648. 8. Densmore, C. L., Markaverich, B. M., O'Malley, B. W. & Clark, J. H. (1989) Biochemistry 28, 7788-77%. 9. Gorski, J., Hansen, J. C. & Welshons, W. V. (1987) Adv. Exp. Med. Biol. 230, 13-29. 10. Greco, T. L. & Gorski, J. (1989) Endocrinology 124,2837-2843. 11. Stack, G., Korach, K. & Gorski, J. (1989) Steroids 54, 227-243. 12. Kumar, V., Green, S., Staub, A. & Chambon, P. (1986) EMBO J. 9, 2231-2236. 13. Kline-Hitpass, L., Schorpp, M., Wagner, U. & Ryffel, G. U. (1986) Cell 46, 1053-1061. 14. Klein-Hitpass, L., Tsai, S. Y., Green, G. L., ClarkJ. H., Tsai, M. J. & O'Malley, B. W. (1989) Mol. Cell. Biol. 9, 43-49. 15. Buch, J. B. E., Evans, M. I., Friedman, T. M. & O'Malley, B. J. (1988) Mol. Cell. Biol. 8, 1123-1131. 16. Maurer, R. A. & Notides, A. C. (1987) Mol. Cell. Biol. 7, 4247-4254. 17. Jost, J.-P., Seldran, M. & Geiser, M. (1984) Proc. Natl. Acad. Sci. USA 81, 429-433. 18. Walter, P., Green, S., Green, G., Krust, A., Bornert, J.-M., Staub, A., Jensen, E., Scrace, G., Waterfield, M. & Chambon, P. (1985) Proc. Natl. Acad. Sci. USA 82, 7889-7894. 19. Koike, S., Sakai, M. & Muramatsu, M. (1987) Nucleic Acids Res. 15, 2499-2512. 20. Ponglikitmongkol, M., Green, S. & Chambon, P. (1988) EMBO J. 11, 3385-3388. 21. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J. R. & Chambon, P. (1987) Cell 51, 941-951. 22. Scambia, G., Ranelletti, F. O., Panici, P. B., Piantelli, M., Bonanno, G., De Vincenzo, R., Ferranadina, G., Pierelli, L., Capelli, A. & Mancuso, S. (1991) Cancer Chemother. Pharmacol. 28, 248-255. 23. Scambia, G., Ranelletti, F. O., Panici, P. B. & Mancuso, S. (1990) Int. J. Cancer 46, 1112-1116. 24. Lopes, M. T. P., Liberato, M. H., Widman, A. & Bretani, M. M. (1987) J. Steroid Biochem. 26, 219-226. 25. Kornyei, J. L., Csermely T., Keller, G. & Vertes, M. (1993) Gynecol. Oncol. 48, 94-103. 26. Syne, J. S., Markaverich, B. M., Clark, J. H. & Panko, W. B. (1982) Cancer Res. 42, 4443-4448. 27. Panko, W. B., Watson, C. S. & Clark, J. H. (1981) J. Steroid Biochem. 14, 1311-1316. 28. Biswas, S. B. & Biswas, E. E. (1990) Mol. Cell. Biol. 10, 810-815. 29. Biswas, E. E., Stefanec, M. J. & Biswas, S. B. (1990) Proc. Natl. Acad. Sci. USA 87, 6689-6692. 30. Scatchard, G. (1949) Ann. N. Y. Acad. Sci. 51, 660-672. 31. Bradford, M. M. (1976) Anal. Biochem. 72, 248-252. 32. Garai, J., Tiller, A. A. & Clark, J. H. (1992) Steroids 57, 183-188. 33. Williams, D. & Gorski, J. (1974) Biochemistry 13, 5537-5542. 34. Savu, L., Benassayag, C., Vallette, G., Christeff, N. & Nunez, E. (1981) J. Biol. Chem. 256, 9414-9418. 35. Nishi, S., Matsue, H., Yoshida, H., Yamaoto, R. & Sakai, M. (1991) Proc. Natl. Acad. Sci. USA 88, 3102-3105. 36. McLeod, J. F. & Cooke, N. E. (1989) J. Biol. Chem. 36, 21760-21769.