Proc. Natl. Acad. Sci. USA Vol. 89, pp. 4037-4041, May 1992 Biochemistry Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover SOPHIE DAUVOIS, PAUL S. DANIELIAN, ROGER WHITE, AND MALCOLM G. PARKER* Molecular Endocrinology Laboratory, Imperial Cancer Research Fund, 44 Lincoln's Inn Fields, London WC2A 3PX, United Kingdom Communicated by Elwood V. Jensen, December 30, 1991 (received for review August 21, 1991) ABSTRACT The ability of estrogens to stimulate the tran- with a subsequent step required for receptor-mediated gene scriptional activity of the estrogen receptor can be inhibited by transcription (16, 17). a diverse range of estrogen antagonists. Here we show that the The variation in the ability of ICI 164,384 to inhibit the antiestrogen ICI 164,384, N-(n-butyl)-11-[3,1713-dihydroxy- DNA binding activity of the estrogen receptor might be estra-1,3,5(10)-trien-7a-yl]N-methylundecanamide, rapidly explained by differences in the stability of receptor dimers. reduces the levels of receptor protein transiently expressed in Since it may not always be possible to dissociate preformed cells without affecting receptor mRNA abundance. The reduc- dimers and inhibit DNA binding in cell-free extracts, we have tion in the levels of receptor protein is dose dependent, revers- investigated whether ICI 164,384 was able to prevent DNA ible by estradiol, and mediated by the hormone-binding do- binding in intact cells. In this paper we show that ICI 164,384 main ofthe receptor. Pulse-chase experiments indicate that the treatment causes a decrease in cellular content of estrogen half-life of the receptor is reduced from -5 hr in the presence receptor protein by markedly reducing its half-life and sug- of estradiol to <1 hr by ICI 164,384. A similar reduction in gest that this might be caused by impaired receptor dimer- estrogen receptor levels is demonstrated in human breast ization. cancer cells treated with ICI 164,384. We discuss the possibility that the increased turnover of the receptor might be a conse- quence of impaired receptor dimerization. MATERIALS AND METHODS Cell Culture and Transient Transfection Experiments. The Estrogens regulate cell growth and differentiation by binding expression of the MOR was analyzed by transient transfec- to specific receptors that function as transcription factors tion experiments. COS-1 cells were grown in Dulbecco's (1-3). Hormone binding is required for the dissociation of an modified Eagle's medium (DMEM) supplemented with 10% inactive oligomeric receptor complex to allow receptor dimer- (vol/vol) fetal calf serum (GIBCO). After suspension in ization and high-affinity DNA binding (4,5) and to produce the phosphate-buffered saline (PBS), cells (6 x 106 per ml) were full transcriptional activity of the receptor (6-9). Stimulation transfected by electroporation with 25 ,ug of pJ3MOR1-599 of transcription is mediated by two transcriptional activation (18), pJ3MOR1-339 (9), or pJ3f1 together with 6 gg of pJ3 functions (TAF)-namely, TAF-1 and TAF-2, the activity of luciferase (19) as an internal control for transfection effi- TAF-2 being dependent upon estrogen binding. Since estro- ciency as described (20). After 24 hr the medium was replaced gens have been found to act as mitogens in at least 30o of with the same medium containing 10 nM estradiol or 1 tM ICI breast cancers, a large number of estrogen antagonists have 164,384 or no added hormone as indicated. Whole-cell ex- been developed as therapeutic agents. One type, which in- tracts were prepared in a high-salt buffer containing 0.4 M cludes the nonsteroidal breast cancer drug tamoxifen (10), can KCI; 20 mM Hepes (pH 7.4); 1 mM dithiothreitol; 20% also act as partial agonists in a number of physiological (vol/vol) glycerol; and 0.5 mg ofbacitracin, 5,g ofaprotinin, responses; while another type, which includes ICI 164,384, 40,tg ofphenylmethylsulfonyl fluoride, 5,ug ofpepstatin, and has been reported to be devoid ofagonist activity (11, 12). ICI 5 ,g of leupeptin per ml (14). Their protein concentrations 164,384 is N-(n-butyl)-11-[3,17,3-dihydroxyestra-1,3,5(10)- were determined with a Bio-Rad protein assay reagent to trien-7a-yl]N-methylundecanamide. normalize the amount of extract examined in a gel-shift The mechanism of action of some steroid hormone antag- assay. onists is beginning to emerge. It has been shown that tamox- Human breast cancer cells, ZR-75-1, were grown in phenol ifen promotes high-affinity DNA binding of the receptor but red-free DMEM supplemented with 10% charcoal-stripped fails to induce the formation ofTAF-2 in the hormone-binding fetal calf serum containing no hormone or 10 nM estradiol or domain, which accounts for its antagonist activity (7, 9). The 1 uM ICI 164,384. Whole-cell extracts were prepared as agonist effect of tamoxifen has been proposed to be derived described above. from TAF-1 in the N-terminal domain of the receptor, which Gel-Shift Assay. DNA binding activity of estrogen recep- is constitutively active (13). The effect of the so-called pure tors was assayed by gel-shift analysis. Whole-cell extracts antiestrogen ICI 164,384 is controversial. In our studies with from transfected COS-1 cells (2 jig of protein) or from human the mouse estrogen receptor (MOR), we demonstrated that breast cancer ZR-75-1 cells (8 jig of protein) were incubated ICI 164,384 inhibited DNA binding of the receptor and with a [32P]DNA probe containing an estrogen response proposed that this may be a consequence of impaired recep- element (9) and the indicated steroids as described (15). An tor dimerization (14). This impairment might result from anti-MOR antibody, MP16 (15), or preimmune serum was steric interference produced by the large 7a-alkylamide ex- also added where indicated. tension present on ICI 164,384 (12), since we have shown that Enzyme Immunoassay of Estrogen Receptors. For immu- the steroid binding pocket is at or near the dimer interface noassays, COS-1 cell extracts were diluted 1:20, and the (15). However, other workers have found that ICI 164,384 did immunoreactive receptors were measured by using mono- not inhibit DNA binding and proposed that it also interferes Abbreviations: MOR, mouse estrogen receptor; ICI 164,384, N-(n- The publication costs of this article were defrayed in part by page charge butyl)-11-[3,17p-dihydroxyestra-1,3,5(10)-trien-7a-yl]N-methylun- payment. This article must therefore be hereby marked "advertisement" decanamide. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 4037 Downloaded by guest on September 29, 2021 4038 Biochemistry: Dauvois et al. Proc. Natl. Acad. Sci. USA 89 (1992) clonal antibodies against the human estrogen receptor in a A commercial kit (Abbott) as instructed by the manufacturer. RNA (Northern) Blot Analysis. Total RNA was isolated from transfected COS-1 cells using phenol-chloroform (1:1, vol/vol) extraction as described (21) and separated by elec- trophoresis in agarose gels containing 2.2 M formaldehyde - 0e d^ (22). After transfer to a nylon membrane (Pall Biodyne), the blots were hybridized with 1 x 106 dpm of riboprobes for the wo estrogen receptor (18) and y-actin (23) per ml as described (18, 24). Radiolabeling of MORs. COS-1 cells were cotransfected with pJ3MOR1-599 (containing the wild-type receptor) and pJ3MOR1-339 (containing a deletion mutant lacking the hor- mone binding domain) by electroporation as described above. Cells were radiolabeled with 200,iCi (1 ACi = 37 kBq) of [35S]methionine in phenol red-free, methionine-free DMEM supplemented with 10%o dialyzed charcoal-stripped fetal calf serum for 1 hr and then incubated for increasing periods of time in medium containing nonradioactive methi- onine and the indicated ligands. Cells were lysed with 0.3 ml FIG. 1. Effect of pretreatment of COS-1 cells with ligand on the of high-salt buffer containing 50 mM Hepes (pH 7.0); 500 mM ability of MORs to bind DNA. Cells were pretreated with no hormone (lanes NH), 10 nM estradiol (lanes E2), or 1 ,uM ICI 164,384 NaCl; 1% Nonidet P-40; and 0.5 mg bacitracin, 5 ,ug of (lanes ICI) for 24 hr. Receptor extracts were incubated in the absence aprotinin, 40 jig of phenylmethylsulfonyl fluoride, 5 Ag of (lanes -) or presence (lanes +) of MP16 before the addition of pepstatin, and 5 ,ug of leupeptin per ml for 10 min at 0°C and radiolabeled estrogen response element and then were tested for centrifuged at 10,000 x g for 10 min; the supernatant was DNA binding activity in a gel-shift assay. (A) Cells were transfected stored at -70°C. with pJ3MOR1-599 or pJ3fl (first lane on left). (B) Cells were Immunoprecipitation. [35S]Methionine-labeled COS-1 cell transfected with pJ3MOR1-599 (lanes 1-599) or deletion mutant lysates were precleared by incubating 150 Al of lysate with 10 pJ3MOR1-339 (lanes 1-339). ,ul of preimmune serum for 40 min at 0°C, followed by a further 40-min incubation at 0°C with 20 ,ul of a 50% slurry of Moreover, the deletion mutant lacking the hormone binding protein A-Sepharose (Pharmacia). After removal ofthe Seph- domain, MOR1-339, which is able to bind to DNA with high arose beads, specific immune (MP16) or preimmune serum (5 affinity only in the presence of the anti-MOR MP16 antise- ,ul) was added, and the mixtures were incubated for a further rum, was also expressed to similar levels in the presence or 1 hr on ice.